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 |
|---|---|---|---|---|---|
To replicate , viruses must gain access to the host cell's resources . Interferon ( IFN ) regulates the actions of a large complement of interferon effector genes ( IEGs ) that prevent viral replication . The interferon inducible transmembrane protein family members , IFITM1 , 2 and 3 , are IEGs required for inhibition of influenza A virus , dengue virus , and West Nile virus replication in vitro . Here we report that IFN prevents emergence of viral genomes from the endosomal pathway , and that IFITM3 is both necessary and sufficient for this function . Notably , viral pseudoparticles were inhibited from transferring their contents into the host cell cytosol by IFN , and IFITM3 was required and sufficient for this action . We further demonstrate that IFN expands Rab7 and LAMP1-containing structures , and that IFITM3 overexpression is sufficient for this phenotype . Moreover , IFITM3 partially resides in late endosomal and lysosomal structures , placing it in the path of invading viruses . Collectively our data are consistent with the prediction that viruses that fuse in the late endosomes or lysosomes are vulnerable to IFITM3's actions , while viruses that enter at the cell surface or in the early endosomes may avoid inhibition . Multiple viruses enter host cells through the late endocytic pathway , and many of these invaders are attenuated by IFN . Therefore these findings are likely to have significance for the intrinsic immune system's neutralization of a diverse array of threats .
The 2009 H1N1 pandemic provided a strong reminder of the threat that influenza A virus poses to world health ( http://www . cdc . gov/h1n1flu/cdcresponse . htm ) . The most effective means of protection against influenza is the seasonal vaccine . However , if the vaccine does not match the viral strains , its effectiveness can be reduced to 50% or less [1] , [2] . Among small molecules , only two approved influenza drugs remain effective , zanamivir ( Relenza ) and oseltamivir ( Tamiflu ) . Although resistance to zanamivir is rare , there has been an increase in oseltamivir-resistant flu strains [3] . Of concern , both drugs target viral neuraminidase ( NA ) , precluding combinatorial therapy to minimize resistance [4] , [5] . Thus , research to identify new anti-influenza strategies would be useful . The influenza A virus is 50–100 nm in size , encodes for up to 11 proteins , and contains eight segments of negative single-stranded genomic RNA ( 3 ) . Influenza A virus infection initiates with the cleavage and activation of the viral hemaglutinnin ( HA ) envelope receptor by host proteases [6] , [7] , [8] , [9] . HA then binds to sialylated proteins on the cell surface , eliciting endocytosis of the viral particle . Endocytosed viruses are transported through the early and late endosomes , with late endosomal acidification triggering a conformational change in HA which results in viral-host membrane fusion [6] , [10] . Fusion transitions from a hemifusion intermediate into a fusion pore through which the virus' eight viral ribonucleoproteins ( vRNPs ) enter the cytosol . The vRNPs are subsequently guided by the host cell's karyopherins into the nucleus [11] , [12] , [13] , wherein the viral RNA-dependent RNA polymerase synthesizes viral genomes ( vRNA ) and mRNAs , both of which are exported to the cytosol , culminating in the production of viral progeny . Genetic screens have identified multiple host factors and pathways which modulate influenza A virus infection in vitro [14] , [15] , [16] , [17] . Using such a genetic screen , we identified the IFITM protein family members IFITM1 , 2 and 3 as antiviral factors capable of blocking influenza A viruses [14] . We further tested the antiviral activity of IFITM3 protein using the seasonal influenza A strains , A/Uruguay/716/07 ( H3N2 ) and A/Brisbane/59/07 ( H1N1 ) , and found similar levels of IFITM3-mediated viral inhibition [14] . IFITM3 accounts for a significant portion ( 50–80% ) of IFN's ( type I or II ) ability to decrease influenza A virus infection in vitro , and IFITM3 resides in vesicular compartments that are IFN-inducible [14] . In addition , the IFITM family inhibits infection by the flaviviruses , dengue virus and West Nile virus [14] , [18] , as well as the filoviruses , Ebola and Marburg , and the SARS coronavirus [19] . The IFITM proteins also block vesicular stomatitis virus-G protein ( VSV-G ) -mediated entry , but do not substantially alter the replication of Moloney leukemia virus ( MLV ) , several arena viruses , or hepatitis C virus ( HCV , [14] , [20] ) . The human IFITM proteins were identified 26 years ago based on their expression after IFN stimulation [21] , [22] , [23] . The IFITM1 , 2 , 3 and 5 genes are clustered on chromosome 11 , and all encode for proteins containing two transmembrane domains ( TM1 and 2 ) , separated by a conserved intracellular loop ( CIL , [22] ) , with both termini extra-cellular or intra-vesicular [24] , [25] . TM1 and the CIL are well conserved between the IFITM proteins and a large group of proteins representing the CD225 protein family . CD225 family members exist from bacteria ( 125 members ) to man ( 13 members , with 156 members in chordata ) , with no in depth functional data available for any member other than the IFITM proteins . IFITM1 , 2 and 3 are present across a wide range of species including amphibians , fish , fowl and mammals . The IFITM proteins have been described to have roles in immune cell signaling and adhesion , cancer , germ cell physiology , and bone mineralization [25] , [26] , [27] , [28] , [29] , [30] . IFITM3 expression can inhibit the growth of some IFN-responsive cancer cells [31] . Genetic evidence also points to IFITM5/Bril being required for early bone mineralization [30] , [32] . IfitmDel mice , which are null for all five of the murine Ifitm genes , display a 30% perinatal mortality among null pups , but thereafter grow and develop normally in a controlled setting [26] . However , cells derived from these IfitmDel mice are more susceptible to influenza A virus infection in vitro [14] . IFITM3 inhibited infection by all influenza A virus strains tested including a 1968 pandemic isolate and two contemporary seasonal vaccine viruses [14] . We have found IFITM3 to be the most potent of the IFITM protein family members in decreasing influenza A virus replication [14] . Viral pseudoparticles are differentially inhibited by the IFITM proteins based on the specific viral receptors expressed on their surfaces [14] , [19] . Therefore , we have hypothesized that IFITM proteins inhibit susceptible virus families ( Orthomyxoviridae , Flaviviridae , Rhabdoviridae , Filoviridae , and Coronaviridae ) during the envelope-dependent early phase of the infection cycle , which extends from viral binding to cell surface receptors through the creation of the fusion pore between viral and host membranes [14] , [19] , [20] . In support of this notion , recent work demonstrated that IFITM protein overexpression did not prevent influenza A virions from accessing acidified compartments [19] . Consistent with its acting on endocytosed viruses , a portion of IFITM3 resides in structures that contain host cell endosomal and lysosomal proteins [19] . Furthermore , inhibition of influenza A virus infection depends on the palmitoylation of IFITM3 , a post-translational modification that targets proteins to membranous compartments [33] . Here we directly test the idea that IFITM3 restricts influenza A viral infection during the envelope-dependent early phase of the viral lifecycle . Consistent with previous studies , we find that IFITM3 inhibits influenza A viral infection after viral-host binding and endocytosis , but prior to primary viral transcription [19] , [20] . Moreover , using a combination of assays , we find that either IFN or high levels of IFITM3 impede influenza A viruses from transferring their contents into the host cell cytosol , and that IFITM3 is necessary for this IFN-mediated action . Therefore , we conclude that IFN is acting predominantly through IFITM3 to block viral fusion . We also find that IFN expands the late endosomal and lysosomal compartments , and that IFITM3 overexpression is sufficient for this phenotype . This study also presents data showing that IFITM3 overexpression leads to the expansion of enlarged acidified compartments consisting of lysosomes and autolysosomes . Interestingly , we observe that viruses trapped in the endocytic pathway of IFITM3-overexpressing cells are trafficked to these expanded acidified compartments . Based on these results and those of others [19] , [20] , we present a model whereby IFN acts via IFITM3 to prevent viral fusion , thereby directing endocytosed viruses to lysosomes and autolysosomes , for subsequent destruction . Collectively this study expands our understanding of how IFITM3 restricts a growing number of viruses by exploiting a shared viral vulnerability arising from their use of the host's endocytic pathway .
The inhibition of HA-expressing pseudoparticles by the IFITM proteins pointed towards restriction occurring during the envelope-dependent phase of the viral lifecycle [14] . Therefore we tested IFITM3's impact on the most proximal phase of infection , viral binding , by incubating influenza A virus A/WSN/33 H1N1 ( WSN/33 , multiplicity of infection ( moi ) 50 ) with A549 lung carcinoma cells either stably overexpressing IFITM3 ( A549-IFITM3 ) or an empty vector control cell line ( A549-Vector , Fig . 1A ) . Samples were incubated on ice to permit viral binding but prevent endocytosis . After incubation , cells were washed with cold media , fixed and stained for HA . When analyzed by flow cytometry , we observed no appreciable difference in surface bound HA between the vector and IFITM3 cells . There was also no difference in surface-bound virus over a series of ten-fold dilutions of viral supernatant ( data not shown ) . We also determined that the stable expression of IFITM3 did not alter the surface levels of ( α2 , 3 ) or ( α2 , 6 ) sialylated cell-surface proteins ( Fig . S1 ) . To investigate IFITM3's impact on initial viral mRNA production , we infected canine kidney cells , either expressing IFITM3 ( MDCK-IFITM3 ) or the empty vector ( MDCK-Vector ) , with influenza A virus ( A/Puerto Rico/8/34 H1N1 ( PR8 ) , moi 500 ) . We used PR8 because of the purified high titer stocks available . Next , the viral supernatant was removed and warm media was added ( 0 min ) . At the indicated times , cells were processed and stained for the positive stranded NP mRNA of PR8 using a specific RNA probe set ( red , Fig . 1B ) , then imaged on a confocal microscope . Based on NP mRNA staining , primary viral transcription begins by 60 min . p . i . in the vector control , with the NP mRNA signal increasing through to 180 min . , when the export of viral mRNAs to the cytosol can be observed . A decrease in primary viral transcription can be seen when comparing the IFITM3 cells to the vector control line . Therefore , IFITM3 inhibits influenza A viral infection after viral-host binding but before primary viral mRNA transcription . We next used confocal imaging to track the nuclear translocation of vRNPs ( Fig . 2 [34] , [35] ) . At the start of infection , the NP within infected cells is complexed with viral genomic RNA forming vRNPs . Therefore , immunostaining for NP permitted us to follow vRNP distribution intracellularly [16] , [34] , [36] . Normal diploid human lung fibroblasts ( WI-38 cells ) were stably transduced with empty vector ( Vector ) , IFITM3 cDNA ( IFITM3 ) , or short hairpin RNAs ( shRNA ) either against IFITM3 ( shIFITM3 ) or a scrambled non-targeting control ( shScramble , Fig . 2 , S2 ) . WI-38s were chosen because of their normal karyotype and relatively larger and flatter morphology . Cells were first incubated on ice with PR8 ( moi 500 ) . Next , the viral supernatant was removed and warm media was added ( 0 min ) . At the indicated times after warming , cells were fixed , permeabilized , stained for NP and DNA , and imaged on a confocal microscope . Image analysis software was used to create an outline of each cell's periphery ( white lines ) and nucleus ( blue lines ) . Based on NP staining , vRNPs arrive in the nuclei by 90 min in the vector control , shIFITM3 , and in the shScramble cells , with the NP signal increasing through to 240 min ( Fig . 2A , S2A–D ) . In contrast , we observed decreased nuclear and increased cytosolic NP staining in the IFITM3 cells ( Fig . 2 , S2C ) . Moreover , in the IFITM3 cells greater than 60% of the cytosolic NP colocalized with Lysotracker Red ( LTRed ) , a dye which marks acidic cellular compartments ( late endosomes , lysosomes , pH≤5 . 5 ) , and which was added to the warm media at time zero ( Fig . S2A , D ) . The increased NP in the cytosol of the IFITM3 cells likely arises in part from an increase in the local concentration of viruses because α-NP Western blots ( after trypsinizing the cells to remove adherent NP ) did not show substantial differences in internalized NP levels between cell lines for up to 90 min post infection ( p . i . , data not shown ) . Because IFITM3 is required for the anti-viral actions of IFN in vitro [14] , we performed a companion experiment with the WI-38 cells treated with IFN-α ( Fig . 2B ) . IFN-α treatment also decreased NP nuclear staining in the WI-38-Vector cells , however this block was not as complete nor was it associated with similar levels of cytosolic NP staining as those seen with high levels of IFITM3 . Consistent with the gain-of-function data , the depletion of IFITM3 decreased IFN's ability to block vRNP trafficking to the nucleus ( Fig . 2A and B , compare top and bottom rows ) . Similar results were obtained either using A549 cells ( Fig . S3 ) or using MDCK cells , with the latter experiments employing additional influenza A viral strains ( X:31 , A/Aichi/68 ( Aichi H3N2 ) , Fig . S4A–C , WSN/33 and A/Victoria/3/75 H3N2 , data not shown ) . It is important to note that the levels of IFITM3 protein in the A549-IFITM3 cells are higher than those seen after treatment with IFN-α or -γ ( Fig . S3C ) . However , we have not observed that other overexpressed proteins have either protected against viral infection or expanded the lysosome/autolysosome compartment ( data not shown ) , arguing that this is a specific effect . To better assess the expanded LTRed compartments observed with IFITM3 overexpression , we created MDCK cells stably expressing the lysosomal protein , LAMP1 , fused to a red fluorescence protein ( LAMP1-RFP ) and IFITM3 . As compared to control cells , the IFITM3 cells demonstrated extensive colocalization ( >60% ) between the NP and LAMP1-RFP signals , revealing that the entering viruses are trafficked to lysosomal compartments ( Fig . S5 ) . We extended this analysis by directly tracking the location of the vRNA contained in the incoming vRNPs . MDCK cells stably expressing an empty vector or IFITM3 , were used in time-course experiments as above ( Fig . 3A–D ) . At the indicated times , cells were processed and stained for the negative stranded NP vRNA of PR8 using a specific RNA probe set ( green ) . As seen with the WI-38 cells , we observed the nuclear translocation of vRNA by 80 min p . i . in the MDCK-vector cells ( Fig . 3A ) . The nuclear vRNA signal was strongly decreased with IFITM3 overexpression based on the average number of vRNA particles present per nucleus ( Fig . 3C ) . Consistent with the WI-38 results , the vRNAs accumulated in the cytosol of the IFITM3 cells , with >50% co-localizing with LTRed-staining acidic structures ( Fig . 3D ) . Similar levels of retained cytosolic vRNPs were observed in experiments without LTRed ( data not shown ) . Interestingly , we observed the loss of the vRNA signal in the acidic inclusions of the MDCK-IFITM3 cells between 80 and 240 min . p . i . ( Fig . 3B ) . By comparison , the vRNAs in the control cells increased in number in both the nucleus and cytosol , as would be expected with the nuclear export of newly synthesized viral genomes [36] . We next evaluated vRNP translocation in murine embryonic fibroblasts ( MEFs ) derived from animals that have had all five Ifitm genes deleted ( IfitmDel−/− , [14] , [26] ) . Compared to wild-type ( WT ) matched litter mate controls , the IfitmDel−/− MEFs displayed 5–10 fold more nuclear NP staining , with or without IFN-γ treatment ( Fig . 4 , S6C ) . IFN-mediated viral restriction was restored when we transduced the null MEFs with a retrovirus expressing Ifitm3 ( IfitmDel−/− Ifitm3 , Fig . S6 ) . Similar to what was observed with the IFITM3 overexpressing cell lines , the majority of the vRNP signal in the IFN-γ-treated WT and Ifitm3-rescued cells localized to acidic compartments ( red , Fig . S6B ) . An increase in acidic compartments occurred after IFN-γ treatment with either the WT or the IfitmDel−/−Ifitm3 MEFs , but not in the IfitmDel−/− cells , suggesting that Ifitm3 is required for this event ( Fig . 4 , S6 ) . Similar results were obtained with IFN-α ( data not shown ) . We conclude from these experiments using orthologous reagents ( cell lines and influenza A viruses ) and methods , that IFN impedes vRNP nuclear entry , and IFITM3 is necessary and sufficient for this activity . To further characterize the mechanism of IFITM3-mediated restriction , we used an established viral fusion assay [37] , [38] . Lentiviral pseudoparticles containing the β-lactamase protein fused to the HIV-1 accessory protein Vpr ( BLAM-Vpr ) and expressing either HA and NA ( H1N1 , WSN/33 ) , or VSV-G envelope proteins , were incubated for 2 h with cells , which were then loaded with the β-lactamase flourogenic substrate , CCF2 . Upon viral pseudoparticle fusion , BLAM-Vpr enters the cytosol and cleaves CCF2 , producing a wave length shift in emitted light ( from green to blue ) when analyzed by flow cytometry ( Fig . 5A , [37] ) . In MDCK-IFITM3 cells we observed a decrease in both HA- and VSV-G-directed fusion , which was comparable to the block produced by poisoning of the host vacuolar ATPase ( vATPases ) with a low dose of bafilomycin A1 ( Baf , Fig . 5B ) . The inhibition of vATPases prevents the low-pH activation required by these two viral envelope proteins to produce membrane fusion . A block to fusion of pseudoparticles expressing H1 ( PR8 ) , H3 ( A/Udorn/72 ) , H5 ( A/Thai/74 ) or H7 ( A/FPV/Rostock/34 ) subtypes of HA was also detected with MDCK cells or with chicken embryonic fibroblasts ( ChEFs ) , in which IFITM3 strongly inhibited viral replication ( Fig . S7A , B , C ) . In the case of the MDCK cells , the block to fusion closely paralleled the level of inhibition seen when the pseudoparticles were tested for productive infection using HIV-1 p24 expression as a readout ( Fig . S7E ) . Consistent with earlier findings , pseudoparticles expressing an amphotropic MLV envelope protein were insensitive to IFITM3 , showing the specificity of these results ( Fig . S7D ) . Similarly to its effect on H5-expressing pseudoparticles , IFITM3 inhibited replication of infectious avian H5N1 influenza A virus , A/Vietnam/1203/04 ( VN/04 ) , isolated from a fatal human infection ( Fig . S7F–H ) . To enhance our analysis , we tested two additional cell lines , WI-38 and HeLa cells . A strong block to fusion in WI-38-IFITM3 cells , similar to that of the Baf and uninfected control samples , was seen at a range of serial dilutions of pseudoparticles , as well as an increase in fusion with IFITM3 depletion ( shIFITM3 , Fig . 5C , D ) . IFN treatment inhibited fusion of the H1N1 pseudoparticles , albeit to a lesser extent than IFITM3 overexpression ( Fig . 5E ) , and this effect was largely absent when IFITM3 was stably depleted in HeLa cells ( Fig . S8 ) . Similar results were obtained with IFN-α ( data not shown ) . Based on these experiments using multiple cell lines and HA , VSV-G , and MLV envelope-expressing pseudoparticles , we conclude that IFITM3 is required and sufficient for an IFN-mediated block of viral pseudoparticle fusion . Importantly , the increase in pseudoparticle fusion seen when endogenous IFITM3 was depleted in either the HeLa or WI-38 shIFITM3 cell lines argues that fusion inhibition underlies the first line defense provided by endogenous , as well as overexpressed , IFITM3 . MxA is an IFN-inducible large GTPase which interferes with secondary transcription during influenza A viral replication [39] . A549 cells express MxA and have been used extensively in influenza A viral replication studies [40] . Therefore to clarify the antiviral roles of IFITM3 and MxA , we tested the levels of viral replication in A549 cells stably expressing one of three shRNAs targeting IFITM3 ( shIFITM3-1 , -2 , or -3 ) . All three shIFITM3 cell lines showed increased infection ( WSN/33 strain ) and strong IFITM3 knockdown , when compared to the negative control cell line expressing a shRNA against firefly luciferase ( shLuc ) , with or without IFN treatment ( Fig . S9A , B ) . The majority of the protective effect of either IFN-α or γ was lost in the shIFITM3 cell lines . We next confirmed both the baseline levels , as well as the IFN-inducibility of MxA in the A549 cells ( Fig . S9C ) . We also determined that MxA was both present and IFN-inducible in WI-38 normal fibroblasts , another cell line used in loss-of-function experiments in this work ( Fig . S9D ) . Furthermore , IF studies of WI-38 cells showed that MxA is expressed in an IFN-inducible vesicular pattern and that these structures did not appreciably co-localize with vesicles containing IFITM3 ( Fig . S9E , [39] ) . We conclude that MxA is expressed in the A549 and WI-38 cell lines , but cannot fully compensate for loss of the antiviral actions of IFITM3 . Our data demonstrate that IFN or IFITM3 inhibit viral fusion . Influenza A virus fuses with the host membrane in late endosomes when the pH decreases to 5 [6] , [7] , [41] . Rab7 is a late endosomal/lysosomal small GTPase that is required for the fusion of many pH-dependent viruses , including influenza A virus [6] , [41] . Previous reports have shown that IFITM3 colocalizes with LAMP1 and CD63 , components of lysosomes and multivesicular bodies , respectively [19] . However , the relationship of IFITM3 and Rab7 within the host cell infrastructure remains unknown . Therefore we investigated the location of IFITM3 , by undertaking immunoflourescence ( IF ) studies using antibodies that recognize IFITM3 , Rab7 , or LAMP1 [42] . Although the baseline level of IFITM3 in the A549-Vector cells was low , there was partial colocalization observed with either Rab7 or LAMP1 ( Fig . 6A–D , 7A , ) . IFITM3 also partially colocalized with LAMP1 and LTRed-containing structures seen with IFITM3 overexpression ( Fig . 6A , B , 7A ) . Interestingly , either IFITM3 overexpression or IFN increased the staining intensity of Rab7 and LAMP1 ( Fig . 7A , B , S10A ) . Partial colocalization of IFITM3 was also seen with either endogenous LAMP1 , or an exogenously expressed Rab7-yellow fluorescence fusion protein ( Rab7-YFP ) in MDCK cells ( Fig . 6E–I ) . However , in all cases , co-localization was not complete because cells contained areas that uniquely labeled for each of the proteins . Western blots indicated that IFITM3 over-expression led to modest increases in both LAMP1 and Rab7 proteins in the A549-IFITM3 cells ( Fig . 7C ) . However , these blots also showed that while IFN treatment of the A549-Vector cells increased IFITM3 protein levels as expected , the amount of Rab7 and LAMP1 remained unchanged . We conclude that IFITM3 partially resides in the late endosomal and lysosomal compartments along with Rab7 and LAMP1 , and that IFITM3 overexpression or IFN treatment expands these compartments through a mechanism that cannot be fully explained by increased protein expression alone . Our assays showed that incoming influenza A viruses were retained in the expanded acidic compartments of both the IFITM3 overexpressing cell lines as well as the IFN-γ-treated MEFs , and that IFITM3 partially localized to these structures ( Fig . 2–4 , S2–4 , S6 ) . Therefore , we extended our investigation of these compartments . An increase in acidic structures was seen in MDCK and A549 cells overexpressing IFITM3 as compared to control cell lines , using either the vital acidophilic stain , acridine orange ( AO ) , LTRed , or a cathepsin-L substrate that fluoresces only after it is proteolyzed , when compared to the corresponding vector control cells ( Fig . 8A , B , a , b ) . Cathepsins are a family of lysosomal zymogens active in acidic environments ( pH≤5 . 5 ) which are required for both the degradation of endocytic substrates and for the entry of several IFITM3-susceptible viruses [19] . Flow cytometry revealed an increase in the total LTRed fluorescent signal in both the MDCK and A549 IFITM3 cell lines when compared to controls ( Fig . 8C ) . This expanded compartment represents a heterogeneous population of lysosomes and autolysosomes , based on confocal imaging showing the colocalization of the autophagosome marker , microtubule-associated protein 1 light chain 3 ( LC3 ) , with either LTRed or with CD63 , with the latter being a resident of multivesicular bodies , amphisomes and autolysosomes ( Fig . 8D , E ) . Furthermore , MDCK-IFITM3 cells stably transduced with an LC3 protein fused to both a red fluorescent protein ( mCherry ) and an enhanced green fluorescence protein ( EGFP ) showed a predominantly red signal , which occurs when the mCherry-EGFP-LC3 protein resides inside the acidified interior of an autolysosome ( Fig . 8F , [43] ) . In keeping with previous reports that IFN-γ induces autophagy [44] , [45] , we detected enhanced LTRed staining in either IFN-γ treated MEFs or A549 cells ( Fig . 4A , S10A ) . We conclude that increases in IFITM3 levels expand the lysosomal/autolysosomal compartment .
Here we report several novel findings regarding the antiviral actions of IFN and the transmembrane IEG , IFITM3 . First , this study demonstrates that IFN inhibits the nuclear translocation of vRNPs , and that IFITM3 is required for this IFN-mediated block , with both endogenous and overexpressed IFITM3 inhibiting vRNP nuclear entry . Second , either endogenous or overexpressed IFITM3 , as well as IFN treatment , block the fusion of viral pseudoparticles expressing various influenza A virus envelope proteins ( H1 , H3 , H5 and H7 subtypes of HA ) , or the VSV-G envelope protein; this block is specific because the fusion of pseudoparticles expressing MLV envelope is not inhibited by IFITM3 . Third , our work reveals that IFITM3 partially resides with Rab7 in late endosomes , thus placing it in position to block influenza A virus' cytosolic access . Fourth , IFITM3 overexpression or IFN induce the expansion of late endosomal and lysosomal compartments containing Rab7 and LAMP1 . Fifth , we show that similar to IFN-γ treatment , IFITM3 overexpression expands the number and size of autolysosomes , and it is into these compartments that trapped viruses are trafficked and subsequently degraded . Consistent with previous reports , our data show that high levels of IFITM3 do not prevent viral access to acidified compartments and that IFITM3 colocalizes with CD63 and LAMP1 [19] . This is in contrast to a report noting the exclusion of overexpressed IFITM3 from LAMP1-containing structures [33] . Therefore , this work adds substantially to our interpretation of previous reports by demonstrating that key downstream events in the viral lifecycle , fusion and vRNP nuclear translocation , are prevented by either IFN or IFITM3 . IFITM3 thus represents a previously unappreciated class of anti-viral effector that permits viral entry into the endosomal compartment , but prevents egress into the cytosol . These studies also raise new questions including i ) how do IFN and IFITM3 prevent viral fusion ? ii ) how do IFN and IFITM3 alter the endosomal and autolysosomal compartments ? and iii ) is the latter action required for viral restriction , or alternatively does it arise as an outcome of IFITM3's potential cellular role ? Based on the substantial loss in IFN's potency observed when IFITM3 is depleted ( 50–80% loss of viral inhibition , Fig . S9A , B , [14] ) we conclude that inhibition of viral emergence from the endosomal pathway is a prominent component of IFN's antagonism of influenza A virus replication in vitro . Our data also show that MxA cannot fully compensate for the loss of IFITM3 in IFN-treated cells challenged with influenza A virus . Recent work by Dittmann et al . [46] and Zimmermann et al . [47] reveal that human influenza A viral strains have evolved a means to evade MxA , suggesting a possible explanation for the cellular reliance on IFITM3 for protection in vitro . Similarly the IEG , IFIT1 , prevents viral replication by targeting viral 5′ triphosphate-RNAs ( PPP-RNA ) for destruction [48] , [49] . Given that IFITM3 is necessary for the majority of IFN-mediated restriction of influenza A virus in vitro , it may be that the virus has also evolved a means to at least partially nullify IFIT1 , perhaps via the massive production of short “decoy” PPP-RNAs , as previously postulated [49] , [50] . IFITM3 primarily resides in the endosomal compartment and partly colocalizes with Rab7 and LAMP1 . IFITM3 overexpression or IFN stimulation caused the endocytosed viruses to accumulate in acidic compartments that contained both IFITM3 and LAMP1 . Together with the BLAM-Vpr fusion assay data , these results reveal that IFITM3 prevents viral-host membrane fusion within late endosomes , and likely within lysosomes as well , in light of studies showing IFITM-mediated restriction of filoviruses and coronaviruses , which depend on cathepsin-mediated activation prior to fusion [19] . In doing so , IFITM3 traps the virus on a path which terminates in a degradative environment [51] . In support of this , our experiments show the eventual loss of a detectable vRNA signal in the LTRed-positive compartments of the IFITM3-transduced cells , thus revealing the fate of viral fitness under those conditions . These studies also reveal that elevated levels of IFITM3 correlate with the expansion of host cell structures containing Rab7 and LAMP1 , and that IFITM3 was also present in these structures . In the MEF and A549 experiments , IFN produced increased Rab7 and LAMP1 immunostaining , in addition to an increase in acidic structures . At present , we cannot explain the increased Rab7 and LAMP1 signals seen after IFN stimulation or IFITM3 overexpression solely on the slight elevations in the abundance of these proteins detected by immunoblotting . Two possible explanations for the increased immunostaining observed , are that IFN stimulation induced these proteins to cluster together or alternatively unmasked sequestered epitopes; we find the latter possibility less likely since LAMP1 and Rab7 flourescent fusion proteins also showed larger and more intense signals under similar conditions . We envision that IFITM3-mediated clustering of organelles and their protein cargoes might contribute to the host cell's antiviral state . Earlier work reported no correlation between the size of the IFITM3-induced acidified compartments and the level of viral restriction [19] , however , we observe that increasing levels of IFITM3 result in both an expansion of lysosomes/autolysosomes and increased viral inhibition . These observations might be explained by a common mechanism underlying the increase in these structures and viral inhibition , in addition to raising the possibility that they play a role in IFITM-mediated viral restriction . Is there a common characteristic shared by IFITM3-susceptible viruses ? The late endosomal- and lysosomal-associated small GTPase , Rab7 , is required for influenza A virus infection [7] , [41] . The IFITM3-resistant viruses previously tested ( MLV , the arena viruses and the hepacivirus , HCV ) are all Rab7-independent , while the entry of the IFITM3-susceptible viruses ( influenza A , dengue , Ebola , Marburg , and SARS ) relies on Rab7 [14] , [19] , [41] , [52] , [53] , [54] . Standing against this hypothesis , is the lack of effect on VSV-G-mediated entry with expression of a dominant negative Rab7 [41] , [55] , [56] ) . However , additional studies have shown that VSV-G-directed entry is dependent on transport to the late endosomes [57] , [58]; these latter results , together with those of Huang et al . and Weidner et al . [19] , [20] , are consistent with the prediction that viruses that fuse in late endosomes or lysosomes are vulnerable to IFITM3's actions , while viruses whose genomes enter at the cell surface or in the early endosomes may avoid IFITM3's full effect . Of note , we have been unable to demonstrate that IFITM3 blocks HIV-1 replication using TZM-bl HeLa cells and are working to address these differences with a published study ( [59] , data not shown ) . This study , together with previous work , demonstrates that IFITM3 permits endocytosis of viruses , but prevents viral fusion and the subsequent entry of viral contents into the cytosol [19] , [20] . While the BLAM-Vpr fusion assay demonstrates inhibition of fusion by IFN or by IFITM3 , we note that this assay uses an indirect readout to assess entry of viral contents . Therefore several possibilities could explain the containment and neutralization of viruses within the endosomal pathway , including alterations in endosomal trafficking , acidification , or the host membrane's fusion characteristics ( bending modulus , elasticity ) . While additional work is required to further define the mechanism , the lack of toxicity seen with cells stably overexpressing high levels of IFITM3 suggests that gross alterations in endogenous trafficking or pH control are unlikely ( data not shown ) . Therefore overexpressing or activating IFITM3 to produce an enhanced antiviral state may be an effective prevention strategy during high risk periods in vulnerable populations . We propose that IFN causes the degradation of endocytosed viruses by preventing their contents from entering the host cytosol , and that IFITM3 is necessary and sufficient for this defense ( Figure 8G ) . IFITM3's mode of defense could be envisioned as an effective means to neutralize pathogens during an organism-wide threat . Such actions might confer an advantage to the host because if IFITM3 simply decreased viral attachment and/or entry , the repulsed viruses would be free to attack neighboring cells . Of course while there are considerable differences between this simple scenario and the directed phagocytosis of pathogens by specialized immune cells , i . e . macrophages , the similarities none-the-less suggest an early prototype for a more evolved defense mechanism .
U2OS , A549 , MDCK , HeLa cells ( all from ATCC ) , and chicken embryonic fibroblasts ( ChEFs , from Charles River Labs ) were grown in complete media ( DMEM , Invitrogen Cat#11965 ) with 10% FBS ( Invitrogen ) . WI-38 cells ( ATCC ) were cultured in DMEM ( Invitrogen Cat#10569 ) , containing non-essential amino acids ( Invitrogen Cat#11140 ) and 15% FBS . Wild type and matched IfitmDel−/− MEFs were from adult IfitmDel+/− mice [26] that were intercrossed and MEFs derived from embryos at day 13 . 5 of gestation , as described previously [14] . The MEFs were genotyped by PCR and Western blot , and the generation of the IfitmDel−/− Ifitm3 cells have been previously described [14] . The IFITM3 retroviral vector , pQCXIP-IFITM3 and empty vector control ( Clontech ) have been previously described [14] . The shRNA lentiviral vectors , pLK0 . 1-Scramble and pLK0 . 1-shIFITM3-3 ( clone ID HsSH00196729 ) are available from the Dana Farber DNA core , Harvard Medical School , Boston , MA . pCAGGS-HA WSN/33 and pCAGGS-NA WSN/33 were kind gifts of Dr . Donna M . Tscherne and Dr . Adolpho Garcia-Sastre , Microbiology Dept . , Mt . Sinai School of Medicine , NY , NY [38] . pBABE-mCherry-EGFP-LC3B was from Addgene ( Plasmid #22418 ) and was kindly deposited by Jayanta Debnath . pLZS-Rab7-YFP and pLVX-RFP-LAMP1 were generously provided by Walther Mothes , Section of Microbial Pathogenesis , Yale University School of Medicine . The following shRNA sequences ( sense strand sequence provided ) were cloned into the pAPM shRNA-expression lentiviral vector [60] , to create the viruses used to generate the A549 IFITM3 knockdown cell lines in Fig . S9: IFITM3-1: 5′-TCCTCATGACCATTCTGCTCAT-3′ IFITM3-2: 5′-CCCACGTACTCCAACTTCCATT-3′ IFITM3-3: 5′-TTTCTACAATGGCATTCAATAA-3′ Influenza A virus A/Puerto Rico/8/1934 ( H1N1 ) ( PR8 , Charles River Labs ) and A/WSN/33 ( H1N1 ) ( kind gift of Dr . Peter Palese , Microbiology Dept . , Mt . Sinai School of Medicine , NY , NY ) were propagated and assessed for viral infectivity as previously described [14] . Influenza A virus A/Vietnam/1203/2004 ( H5N1 ) was propagated and characterized as previously described [61] . Human interferon ( IFN ) -γ ( Invitrogen ) was used at 100–300 ng/ml , human IFN-αA2 ( PBL Interferon Source ) was used at 500–2500 U/ml . Cells were incubated with cytokines for 16–24 h prior to IF or viral infection experiments unless otherwise noted . Murine IFN-γ ( PBL Interferon Source ) was used at 100–300 ng/ml . Whole-cell extracts were prepared by cell lysis , equivalent protein content boiled in SDS sample buffer , resolved by SDS/PAGE , transferred to Immobilon–P membrane ( Millipore ) , and probed with the indicated antibodies . Cells were seeded on glass coverslips for Influenza A virus infection experiments . Cells were incubated on ice with PR8 for 40 min . At time zero , the viral supernatant was removed and 37°C media was added with or without Lysotracker Red DND-99 ( Invitrogen ) . At the indicated time points post-warming , cells were washed twice with D-PBS ( Sigma ) and incubated for 30 seconds with room temperature 0 . 25% trypsin ( Invitrogen ) . The cells were then washed with complete media twice and fixed with 4% formalin ( PFA , Sigma ) in D-PBS . Image analysis for quantitation of vRNP nuclear translocation was done using Imaris 7 . 1 ( bitplane scientific software ) . We generated a mask of the nucleus and applied this mask to the channel containing the viral signal ( puncta ) to determine vRNA puncta contained in each nucleus . Cells were incubated at 37°C and 5% CO2 for 60 min . with either Lysotracker Red DND-99 or acridine orange ( ImmunoChemistry Technologies ) . Hoechst 33342 ( DNA stain , Invitrogen ) was incubated ( 1∶10 , 000 ) with the cells for the final 15 min . The Cathepsin L flourogenic substrate assay was performed as per the manufacturer's instructions ( Cathepsin L -Magic Red , ImmunoChemistry Technologies ) . Cells were visualized live by confocal microscopy . Cells were fixed in 4% PFA in D-PBS , and then incubated sequentially in 0 . 25% Tween 20 ( Sigma ) , then 1% BSA with 0 . 3 M glycine ( Sigma ) , both in D-PBS . Primary and secondary antibodies are listed below . Slides were mounted in Vectashield with DAPI counterstain ( Vector Labs ) . Slides were imaged using a Zeiss LSM 510 , laser scanning inverted confocal microscope equipped with the following objectives: 40× Zeiss C-APOCHROMAT UV-Vis-IR water , 1 . 2NA , 63× Zeiss Plan-APOCHROMAT DIC oil , 1 . 4NA , and 100× Zeiss Plan-APOCHROMAT DIC oil , 1 . 46NA . Image analysis was performed using ZEN software ( Zeiss ) . Laser intensity and detector sensitivity settings remained constant for all image acquisitions within a respective experiment . Nuclear outlines were generated using Metamorph software suite ( Molecular Devices ) using the Kirsch/Prewitt filter to define boundaries and then subtracting out the original binary images . The following antibodies were used in this study for either Western blotting ( WB ) or immunoflourescence ( IF ) , or both as indicated , along with their respective source and catalogue number: Primary antibodies: Actin ( Sigma A5316 , WB ) , CD63 ( Developmental Studies Hybridoma Bank ( DSHB ) clone H5C6 , IF ) , Fragilis ( mouse Ifitm3 ) ( Abcam ab15592 , WB , IF ) , GAPDH ( BD Biosciences 610340 , WB ) , HA ( Wistar collection , Coriell Institute , clone H18-S210 , WC00029 , IF ) , IFITM3 ( Abgent AP1153a , WB , IF ) , IFITM3 ( Abgent AP1153c , IF ) , LAMP1 ( ( DSHB ) clone H4A3 , WB , IF ) , LC3 ( Nanotools Mab LC3-5F10 , WB , IF ) , MX1 ( Proteintech 13750-1-AP , WB , IF ) , NP ( Millipore clone H16-L10-4R5 MAB8800 , IF ) , RAB7 ( Abcam 50533 , WB , IF ) . Secondary antibodies for IF ( all from Invitrogen ) : Alexa Fluor 488 and 647 ( goat anti-rabbit and goat anti-mouse ) . The LAMP1 [H4A3] and CD63 [H5C6] antibodies were developed by J . T . August and J . E . K . Hildreth and were obtained from the DSHB and developed under the auspices of the NICHD and maintained by The University of Iowa , Department of Biology , Iowa City , IA . These experiments employ the QuantiGene ViewRNA slide-based assay kit from Affymetrix ( Cat #QV0096 ) with all components from that source unless noted . RNA was visualized following a modified manufacturer protocol; changes made include the omission of the ethanol dehydration step , and use of Vectashield mounting media . Post-fixation with 4% PFA , cells adherent on coverslips were incubated with 1× detergent solution or incubated in 0 . 25% PBS-Tween20 . Cells were then incubated with Proteinase K . Next cells were incubated at 40°C in hybridization solution A containing a viewRNA probe set designed against either the negative stranded RNA NP genome ( vRNA ) of PR8 ( Affymetrix VX1-99999-01 QG ViewRNA TYPE 1 Probe Set against NP Influenza A virus ( A/PuertoRico/8/34 ( H1N1 ) ) at 1∶100 ) or a probe set against the positive stranded NP mRNA . Cells were then incubated in hybridization preamplifiers ( 1∶100 in hybridization buffer B ) at 40°C . Finally cells were incubated with labeled probes ( 1∶100 in hybridization buffer C ) , washed and imaged as above . All steps were followed by two D-PBS washes . Pseudotyped lentiviral particles expressing the HA envelope were produced by plasmid transfection of HEK 293T cells with an HIV-1 genome plasmid derived from pBR43IeG-nef+ ( NIH AIDS Research and Reference Reagent Program ( Division of AIDS , NIAID , NIH , Cat#11349 , from Dr . Frank Kirchhoff ) modified with a deletion which abolishes expression of Env without disrupting the Rev-responsive element , pCAGGS-HA WSN/33 , pCAGGS-NA WSN/33 and pMM310 , which encodes a hybrid protein consisting of β-lactamase fused to the HIV accessory protein , Vpr ( NIH AIDS Research and Reference Reagent Program , Division of AIDS , NIAID , NIH ( Cat#11444 ) from Dr . Michael Miller ) . pCG-VSV-G together with pBR43IeG-nef+ and pMM310 were transfected to produce VSV-G pseudotyped lentiviral particles . For the H5N1 , H3N1 , and H7N1 pseudoparticles , pCAGGS-HA5 ( A/Thailand2 ( SP-33 ) /2004 ) pCAGGS-HA3 ( A/Udorn/72 ) , and pCAGGS-HA7 ( A/FPV/Rostock/34 ) expression plasmids were co-transfected with the pCAGGS-NA WSN/33 , pMM310 , and the pBR43IeG-nef+ lentiviral backbone . Cultures for pseudoparticle fusion assays , including stably transduced MDCK cells and WI-38 fibroblasts , were plated in 24-well dishes with 90 , 000 cells per well at the beginning of each assay . At the time of assay , 0 . 5 mL of virus stock was added to cells and incubated for 2–3 h ( depending on cell type ) at 37°C . In experiments using bafilomycin A1 ( Sigma ) , the inhibitor was added at 0 . 1 nM final concentration ( low dose ) at 37°C for 1 h prior to incubation with virus . After infection , viral media was then aspirated and replaced with complete DMEM containing CCF2-AM ( Invitrogen ) along with 1 . 7 µg/mL probenecid ( Sigma ) . Cells were incubated in the dark for 1 h , followed by dissociation from the dish using Enzyme Free PBS-based Dissociation Buffer , and fixation in 2% PFA . Flow cytometry was conducted on a Becton Dickinson LSRII using 405 nm excitation from the violet laser , and measuring 450 nm emission in the Pacific Blue channel and 520 nm emission in the Pacific Orange channel . Data was analyzed using FACSDiva and FlowJo8 . 8 . 7 . A549 cells stably transduced to overexpress IFITM3 or with empty expression vector ( pQCXIP , Clontech ) were grown to ∼50% confluency , dissociated with trypsin-free EDTA-based dissociation buffer ( Invitrogen ) for 10 min . at 37°C . Cells were incubated at 4°C with FITC-conjugated Sambucus nigra lectin ( SNA , Vector Labs #FL-1301 ) to detect ( α-2 , 6 ) sialic acid linkages , and biotinylated Maackia amurensis lectin II ( MAL , Vector Labs #B-1265 ) to detect ( α-2 , 3 ) sialic acid linkages , followed by streptavidin-PE-Cy7 ( Invitrogen ) . Cells were incubated with lectins individually and in combination , and the results of staining were indistinguishable . All cells were stained with violet cell-impermeable dye ( Invitrogen #L34955 ) , and cells were included in the analysis if viable by FSC/SSC and viability dye . A549 cells transduced with IFITM3 or the empty vector pQXCIP were detached using Enzyme Free PBS-based Dissociation Buffer , and then washed in cold PBS extensively . Cells and virus ( WSN/33 ) were pre-chilled on ice for 30 min . and mixed at a moi of 50 and incubated at 4°C for 1 h with rotation . Cells were washed extensively with ice cold PBS and then fixed using 4% PFA . The cells were then probed with anti-HA mouse monoclonal antibody ( Wistar collection , Coriell Institute , clone H18-S210 , WC00029 , IF ) for 1 h at room temperature , followed by anti-mouse AlexaFlour-488 conjugated antibody ( Invitrogen ) for 1 h with PBS washes in between , then analyzed by flow cytometry . | Influenza epidemics exact a great toll on world health . Thus research to identify new anti-influenza virus strategies would be useful . Each of our cells contains antiviral factors that work to inhibit infection . A large component of this antiviral program is regulated by the interferon family of signaling molecules . Here , we seek to better understand how one of these antiviral factors , IFITM3 , contributes to both baseline , as well as interferon-induced , antagonism of influenza A viral infection . We found that interferon prevents influenza A virus from entering our cells by blocking the virus' fusion with the cellular membrane . Furthermore , we learned that IFITM3 is required for this antiviral action of interferon , and that high levels of IFITM3 alone can produce a similar viral inhibition . Together , these results improve our understanding of how IFITM3 serves to defend us against viral invasion at a very early stage of infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2011 | IFITM3 Inhibits Influenza A Virus Infection by Preventing Cytosolic Entry |
Cyclic GMP-AMP ( cGAMP ) synthase ( cGAS , a . k . a . MB21D1 ) , a cytosolic DNA sensor , catalyzes formation of the second messenger 2’3’-cGAMP that activates the stimulator of interferon genes ( STING ) signaling . How the cGAS activity is modulated remains largely unknown . Here , we demonstrate that sentrin/SUMO-specific protease 7 ( SENP7 ) interacted with and potentiated cGAS activation . The small ubiquitin-like modifier ( SUMO ) was conjugated onto the lysine residues 335 , 372 and 382 of cGAS , which suppressed its DNA-binding , oligomerization and nucleotidyl-transferase activities . SENP7 reversed this inhibition via catalyzing the cGAS de-SUMOylation . Consistently , silencing of SENP7 markedly impaired the IRF3-responsive gene expression induced by cGAS-STING axis . SENP7-knockdown mice were more susceptible to herpes simplex virus 1 ( HSV-1 ) infection . SENP7 was significantly up-regulated in patients with SLE . Our study highlights the temporal modulation of the cGAS activity via dynamic SUMOylation , uncovering a novel mechanism for fine-tuning the STING signaling in innate immunity .
Cytosolic aberrant DNA is generally sensed as a danger signal to alert the host about the presence of invading microbes , which triggers the immediate immune responses to control microbial invasion and induces the subsequent adaptive immunity effective for ultimately eradicating infection . Elucidating the cytosolic-DNA-triggered signaling pathway and the relevant regulatory mechanisms represents a fast evolving field to understand the corresponding innate immunity . Cyclic GMP-AMP ( cGAMP ) synthase ( cGAS , a . k . a . MB21D1 or C6orf150 ) was recently characterized as a universal cytosolic-DNA sensor , triggering the immune and inflammatory responses in a DNA-sequence-independent manner [1] . The structures of cGAS-DNA complex have been reported [2 , 3] . cGAS-deficient cells display profound defects in the production of IFN-β and pro-inflammatory cytokines , when challenged by the DNA virus herpes simplex virus 1 ( HSV-1 ) or by the bacteria Listeria monocytogenes or Mycobacterium tuberculosis ( Mtb ) [4–8] . cGas-knockout mice were more susceptible to lethal infection after exposure to HSV-1 than wild-type mice [8] . In addition , cGAS is essential for innate immune responses against HIV and other retroviruses via detecting the reverse-transcribed retroviral DNA [9] . Interestingly , cGAS possesses the nucleotidyl-transferase catalytic activity , which could be activated by the dsDNA-induced oligomerization [2 , 3 , 10] . cGAS catalyzes from ATP and GTP the synthesis of a non-canonical cyclic dinucleotide c[G ( 2’ , 5’ ) pA ( 3’ , 5’ ) p] ( referred to as 2’3’cGAMP ) [11–15] . Notably , the 2’3’cGAMP is observed to transfer from the producing cells to the neighboring cells via cell-cell junctions , or to the newly infected cells by virions incorporated with cGAMP [16–18] . As a second messenger , cGAMP directly binds to and activates the ER-resident stimulator of interferon genes ( STING ) , a converging adaptor for all known cytosolic DNA sensors [19] . STING is triggered to dimerize and translocate from the endoplasmic reticulum ( ER ) , through the Golgi apparatus , to perinuclear microsomal compartments . Simultaneously , ER-resident autocrine motility factor receptor ( AMFR ) and insulin induced gene 1 ( INSIG1 ) protein complex catalyzes K27-linked poly-ubiquitination of STING [20 , 21] . TBK1 selectively binds to K27-linked poly-ubiquitin chains on STING , thus congregating at perinuclear microsomes in a STING-dependent manner . The DNA-induced assembly of the STING-TBK1 complex is required for the activation of TBK1 and the subsequent phosphorylation and nuclear translocation of the transcriptional factor IRF3 , ultimately leading to expression of interferon ( IFN ) and pro-inflammatory cytokines . However , the STING signaling pathway could adversely promote the autoimmune diseases , including systemic lupus erythematosus ( SLE ) , lupus-like diseases , and Aicardi Goutieres syndrome ( AGS ) , when the aberrant self-DNAs is not properly cleared out and accumulate in the cytosol under pathological conditions [22–24] . It is intriguing to dissect the regulatory mechanisms of the cGAS activity , and to understand how cGAS is spatially and temporally modulated to maintain immune balance . Protein posttranslational modifications ( PTM ) are effective means to dynamically shape the strength and duration of signal transductions . Ubiquitin and ubiquitin-like proteins ( Ubls ) are versatile molecular signatures for orchestrating the appropriate innate immune responses [25] . It remains to explore the regulatory role of SUMOylation in the cGAS-STING signaling pathway . Previous studies identified a family of sentrin/SUMO-specific proteases ( SENPs ) , which catalyze the biochemical reaction of de-SUMOylation and modulate the dynamic equilibrium of SUMOylation . SENP family has six members ( SENP1- 3 & SENP5-7 ) , each of which exhibits distinct expression patterns and substrate specificity [26 , 27] . In this study , we characterized SENP7 to specifically potentiate the cGAS activation . Silencing of SENP7 markedly attenuated the cGAS-mediated induction of antimicrobial genes . SENP7-knockdown mice were more susceptible to HSV-1 infection . Mechanistically , the small ubiquitin-like modifier ( SUMO ) was conjugated onto the lysine residues 335 , 372 and 382 of cGAS , which suppressed its DNA-binding , oligomerization and nucleotidyl-transferase activities . SENP7 reversed this inhibition via catalyzing the de-SUMOylation of cGAS . Our study reveals the novel function of SENP7 in the STING signaling , shedding new light on the regulatory role of SUMOylation in innate immunity .
To explore the functions of protein post-translational modifications in STING signaling , we screened a library of small interfering RNAs ( siRNAs ) that target the E3 ubiquitin ligases , deubiquitinases and SUMO-specific proteases . The 60-mer oligonucleotide dsDNA derived from the HSV-1 genome ( referred to as HSV-1 60-mer ) was used to stimulate the expression of the IFN-β-luciferase-reporter in the functional assay . As expected , the siRNA against STING dramatically impaired the activation of the IFN-β-luciferase-reporter . Notably , knockdown of Senp7 or Rnf133 respectively impaired the same activation to a similar extent . In contrast , knockdown of Senp2 or Usp18 displayed a 1 . 5 fold increase of the luciferase induction ( Fig 1A ) . It has been established that chronic activation of STING-dependent immune signaling and subsequent type I IFN production by aberrant DNA species play a crucial role in inflammatory disorders such as systemic lupus erythematosus ( SLE ) [28 , 29] . To assess the association of the genes fished out from the above screening ( Senp7 , Rnf133 , Senp2 and Usp18 ) with SLE , we collected peripheral blood samples from 27 patients with SLE and 28 healthy donors to check the mRNA abundance of these genes . Interestingly , the expression levels of SENP7 transcripts were highly elevated in the samples of SLE patients than those in healthy donors , while the expression levels of other genes are comparable between SLE patients and healthy donors ( S1A Fig ) . Also , the abundance of SENP7 mRNA in the samples of SLE patients was positively correlated with that of IFN-inducible genes ( S1B Fig ) , indicating that the expression of SENP7 was positively correlated with IFN production and SLE . These data implied that SENP7 might be important in DNA-triggered STING signaling . Two different Senp7 siRNAs ( Senp7 siRNA 440 and Senp7 siRNA 3880 ) were further employed to test whether SENP7 regulates the expression of IRF3-responsive genes induced by cytosolic DNA stimuli , as measured by qPCR ( quantitative PCR ) and ELISA ( enzyme-linked immunosorbent assay ) ( S1C Fig ) . It was observed that silencing of Senp7 drastically down-regulated the expression of the IRF3-responsive genes ( Ifnb , Ifna4 and Cxcl10 ) in MEFs , stimulated by poly ( dA:dT ) or ISD ( Fig 1B–1E ) . However , silencing of SENP3 had no such effect ( S1D Fig ) . Furthermore , Senp7-/- MEFs were generated by CRISPR-Cas9-mediated targeting ( Fig 1F , lower panel ) . Consistently , the absence of Senp7 markedly crippled the cytosolic DNA-triggered antiviral gene expression ( Fig 1F and 1G ) . In contrast , the SeV ( Sendai virus ) -induced activation of the IFN-β-luciferase reporter was barely affected when silencing SENP7 , indicating that SENP7 did not regulate the RIG-I/MDA5-mediated signaling ( Fig 1H ) . Silencing of SENP7 also displayed no effect on TNF- α -induced NF-κB activation ( Fig 1I ) . Collectively , these data suggest that SENP7 potentially promotes the STING signaling . Given that SENP7 is a SUMO-specific protease , we explored whether its regulation of the STING signaling was dependent on its de-SUMOylation activity . We generated the SENP7 C992S mutant ( cysteine to serine substitution at position 992 ) , which is deprived of the de-SUMOylation biochemical activity [30] . Ectopic-expression of SENP7 WT potentiates the ISD-triggered STING signaling , as evidenced by higher levels of IFNs transcription and secretion ( Fig 2A and 2B ) . In contrast , exogenous expression of the SENP7 C992S mutant failed to promote the production of IFNs . Alternatively , ‘rescue’ experiments were performed in Senp7-silenced MEFs . We generated two RNA interference ( RNAi ) -resistant SENP7 constructs , namely rSENP7 WT and rSENP7 C992S , in which silent mutations were introduced into the sequence targeted by the siRNA without changing the amino acid sequence of the corresponding proteins . MEFs were first transfected with control or SENP7 siRNAs , followed by transfection of the control or indicated rSENP7 plasmids , respectively . As shown in Fig 2C and 2D , the attenuation of IFNs production in Senp7-silenced cells was relieved by introducing rSENP7 WT . Notably , rSENP7 C992S was unable to rescue the deficiency of the DNA-induced IFNs expression . Consistently , reconstitution of the wild-type SENP7 , rather than SENP7 C992S , is able to rescue the deficiency of the DNA-induced IRF3-responisve genes expression in Senp7-/- MEFs ( S1E and S1F Fig ) . Taken together , these results suggest that the de-SUMOylation activity of SENP7 is essential for modulating the STING signaling pathway . To identify the potential target of SENP7 , we observed that exogenous expression of cGAS could trigger the expression of IFN-β-luciferase reporter , and this activation was markedly crippled when silencing SENP7 . In contrast , knockdown of SENP7 did not affect the activation of IFN-β-luciferase reporter , when ectopic-expressing STING , TBK1 or IRF3-5D ( Fig 3A ) . Likewise , SENP7 had no apparent effect on cGAMP-induced activation of the IRF3-responsive genes ( S2A Fig ) . Given the hierarchical relationships among these signaling molecules , we reasoned that SENP7 modulates the STING signaling pathway upstream of STING . In addition , knockdown of SENP7 barely affected the expression of IFN-β-luciferase reporter , stimulated by ectopically expressing the CARD domains of RIG-I or MAVS , indicating that SENP7 did not regulate the RIG-I-mediated signaling ( Fig 3A ) . To corroborate , knockdown of SENP7 attenuated the ISD-induced phosphorylation of TBK1 ( Fig 3B ) . Consistently , knockdown of SENP7 led to an apparent decrease in the phosphorylation and dimerization of IRF3 , when stimulating cells with ISD ( Fig 3C ) . The ISD-induced nuclear translocation of IRF3 was markedly diminished when silencing SENP7 ( Fig 3D and 3E ) . In contrast , knockdown of SENP3 displayed no such inhibitory effects ( Fig 3B–3E ) . Neither knockdown of SENP7 had any effect on either phosphorylation of TBK1 and IRF3 or dimerization of IRF3 when cells were treated with RNA stimulus , emphasizing the specific role of SENP7 in cytosolic DNA-triggered innate immune signaling , but not RNA-induced immune response ( S2B Fig ) . Co-immunoprecipitation ( co-IP ) assay revealed that SENP7 specifically associated with cGAS , whereas it did not interact with STING , TBK1 or IRF3 ( Figs 3F and S2C ) . cGAS associated with the SENP7 C992S mutant as well as with SENP7 WT , suggesting that the catalytic domain of SENP7 is dispensable for its association with cGAS ( Fig 3G ) . We then mapped the N-terminal domain of SENP7 ( 1–300 aa ) ( S2D Fig ) and the middle region of cGAS ( 240–380 aa ) ( S2E Fig ) to respectively mediate this interaction . The association between SENP7 and cGAS was also confirmed endogenously ( Fig 3H ) . Notably , we observed the dynamic association between cGAS and SENP7 , first strengthened upon HSV-1 challenge and then weakened around 12 hours after HSV-1 infection ( Fig 3H ) . The observation was confirmed by a semi-quantitative immunofluorescence time-course of the endogenous SENP7 and cGAS stimulated by HSV-1 ( Fig 3I ) . Confocal microscopy revealed that cGAS displayed a punctate staining pattern and partially co-localized with SENP7 in the cytoplasm of resting cells . Interestingly , their co-localization was strengthened shortly after HSV-1 infection . Unexpectedly , extended HSV-1 infection led to the nuclear-translocation of cGAS and the obvious spatial separation of SENP7 and cGAS . Taken together , these data establish that SENP7 interacts directly and dynamically with cGAS upon HSV-1 infection . The above observations led us to hypothesize that cGAS could potentially be modified by SUMO . To explore this possibility , we co-expressed Flag-tagged cGAS together with HA-tagged SUMO-1/2/3 , respectively . The cell lysates were subjected to the denaturing immunoprecipitation with an anti-Flag antibody . Immunoblot analysis with anti-HA antibody revealed multiple bands that migrated more slowly than cGAS , suggesting that cGAS was robustly modified by SUMO proteins ( Fig 4A ) . In contrast , the cytosolic DNA sensor DAI ( a . k . a . DLM-1/ZBP1 ) was not modified by SUMO ( Fig 4B ) . In vitro SUMOylation assay further confirmed that cGAS was bone-fide substrate of SUMO modification , in the presence of recombinant E1 enzyme ( SAE1/SAE2 ) and E2 enzyme ( Ubc9 ) ( Fig 4C ) . In addition , endogenous cGAS was observed to be SUMOylated ( Fig 4D ) . Notably , the level of the cGAS SUMOylation was markedly decreased at the early phase of HSV-1 infection , and gradually restored later , which inversely echoed the pattern of the dynamic interaction between cGAS and SENP7 during HSV-1 infection ( Figs 4D , 3H and 3I ) . To map the SUMOylation sites on cGAS , we carried out a systematic lysine ( K ) to alanine ( A ) mutation scanning . Lysines 335 , 372 and 382 were identified to be indispensable for the cGAS SUMOylation , as the SUMOylation of cGAS K335A , K372A or K382A mutants was largely compromised ( Figs 4E and S3A ) . To substantiate , we generated a cGAS ( 3A ) mutant , in which all of the three lysines ( 335 , 372 and 382 ) were mutated to alanines . As expected , cGAS K3A could barely be modified by SUMO ( Fig 4F ) . On the background of this cGAS K3A mutant , we generated three more cGAS mutants ( K335 , K372 or K382 respectively ) , in which only a lysine residue was re-introduced back to the original site . It was observed that the SUMOylation of cGAS reappeared in K335 , K372 or K382 cGAS mutants ( Fig 4F ) , indicating that K335 , K372 or K382 is also sufficient for cGAS SUMOylation . Consistent with that of cGAS K335A , K372A or K382A mutants , the SUMOylation of cGAS K335R , K372R or K382R mutants ( lysine to arginine mutations ) was also largely compromised ( S3B and S3C Fig ) . We further performed the sequence alignment of cGAS orthologs and uncovered Lys335 , Lys372 , Lys382 and their surrounding sequences to be highly conserved across species ( S4A Fig ) . Collectively , these data established that the three lysines ( K335 , 372 and 382 ) of cGAS are the major SUMOylation sites on cGAS . Next , a cell-based de-SUMOylation assay was performed to address whether SENP7 could deconjugate the SUMOylated cGAS . Wild-type SENP7 or catalytically dead SENP7 mutant ( SENP7 C992S ) was individually co-transfected with cGAS and SUMO . The cell lysates were subjected to immunoprecipitation of Flag-tagged cGAS and then the precipitates were probed for the SUMOylation signal . As expected , cGAS was robustly SUMOylated in the presence of SUMO-1 or SUMO-3 ( Fig 4G ) . Notably , this modification was drastically reduced when expressing SENP7 ( Fig 4G ) . In contrast , the catalytically inactive SENP7 C992S as well as wild-type SENP3 could not influence the SUMOylation status of cGAS ( Fig 4G and 4H ) . Furthermore , knocking down endogenous SENP7 could enhance the basal SUMOylation of Flag-cGAS , whereas knocking down endogenous SENP3 displayed no such effect ( Fig 4I ) . SENP7 knockdown apparently potentiated the SUMOylation of the endogenous cGAS ( Fig 4J ) . In vitro assay confirmed that wild-type SENP7 , but not SENP7 C992S , could remove the SUMO moiety from the SUMOylated cGAS ( Fig 4K ) . Collectively , these data demonstrate that SENP7 catalyzes the de-SUMOylation of cGAS . cGAS functions as a cytosolic DNA sensor by binding to dsDNA and catalyzing the synthesis of the second messenger 2’3’cGAMP . Biophysical and biochemical analysis revealed that the nucleotidyl-transferase ( NTase ) catalytic activity of cGAS is activated by the dsDNA-induced oligomerization of cGAS . According to the structural data of cGAS and cGAS/dsDNA , we noticed that the SUMOylation sites on cGAS ( Lys335 , Lys372 and Lys382 ) are located in either DNA-binding surfaces or dimerization interface of cGAS ( Fig 5A ) . Consistent with previous reports , mutating the three lysines deprived cGAS of its affinity to DNA as well as its catalytic activity ( S4B and S4C Fig ) . To explore the functional consequence of the cGAS SUMOylation , we expressed and purified cGAS or SUMOylated cGAS , and checked their affinity to the biotin-labeled ISD by pull-down assay with streptavidin-conjugated beads ( Fig 5B ) . Strikingly , the biotin-labeled ISD failed to pull down the SUMOylated cGAS , whereas it could pull down the unmodified cGAS ( Fig 5B ) , indicating that the cGAS SUMOylation impaired its DNA-binding ability . Moreover , as shown in Fig 5B , SENP7 rather than SENP7 C992S could reverse the effect of SUMOylation and restore the binding of cGAS with Biotin-ISD . Consistently , ectopically expressing SUMO abrogated the co-localization of cGAS with ISD ( Fig 5C and 5D ) . We next performed the fluorescence resonance energy transfer ( FRET ) assay to explore the cGAS self-association , using green fluorescent protein ( GFP ) - and mCherry- tagged cGAS . It was observed that there was significant FRET between GFP-cGAS and mCherry-cGAS ( ~12% ) , an index for cGAS homo-oligomerization ( Fig 5E–5H ) . However , cells expressing SUMO or silencing SENP7 displayed a significantly lower FRET signal ( Fig 5E–5H ) . Consistently , cells expressing cGAS K3A exhibited a weak FRET signal ( Fig 5E and 5F ) . In addition , immuno-precipitation experiments substantiated that the Flag-tagged cGAS could interact with its HA-tagged counterpart , whereas expressing SUMO crippled this interaction ( Fig 5I ) . To assess the effect of cGAS SUMOylation on its NTase activity , we conducted the standard enzymatic assay by stimulating cGAS with salmon sperm DNA , followed by ion exchange chromatography . The purified cGAS was subjected to the in vitro SUMOylation reaction and enrichment . As expected , cGAS without SUMOylation could efficiently catalyze the synthesis of 2’3’-cGAMP in the presence of ATP , GTP , and salmon sperm DNA ( Fig 5J ) . In contrast , cGAS enriched from the reaction in the presence of SUMO modification system synthesized much less 2’3’-cGAMP , indicating that cGAS catalytic activity was inhibited by the SUMOylation ( Fig 5J ) . Importantly , this inhibition was reversed by SENP7 , but not by the catalytic-dead SENP7 ( C992S ) ( Fig 5K ) . Collectively , these data uncovered the inhibitory function of the SUMOylation for cGAS activation , which is reversed specifically by the de-SUMOylation enzyme SENP7 . Robust inductions of IFN-β and interferon-stimulated genes represent one of the immediate responses to the microbial infections . qPCR and ELISA assays revealed that silencing of endogenous SENP7 markedly impaired the production of IFN-β , IFN-α4 or CXCL10 , when infecting cells with HSV-1 ( Figs 6A , 6B and S5 ) or Listeria monocytogenes ( Fig 6C and 6D ) . Since IFN-β protects host cells against viruses , we further assessed if SENP7 could restrict HSV-1 infection . MEFs were respectively pretreated with culture supernatants from ISD-stimulated Senp7-silenced MEFs or control MEFs , followed by HSV-1 infection . Fresh cells pretreated with culture supernatants from Senp7-silenced MEFs were more sensitive to HSV-1 infection ( Fig 6E ) . Next , Senp7-silenced MEFs or control MEFs were challenged by HSV-1 , and then the titers of HSV-1 were analyzed by standard plaque assay . As shown in Fig 6F , SENP7 knockdown resulted in a 20-fold increase in virus titer as compared with controls . Consistently , knockdown of SENP7 increased the number of HSV-1-GFP-positive cells ( Fig 6G ) . Unlike HSV-1 , knockdown of SENP7 has no effect on RNA virus infection , suggesting that SENP7 plays an essential role in restricting DNA virus infection ( S6 Fig ) . In addition , knockdown of SENP7 markedly enhanced the replication of Listeria monocytogenes ( Fig 6H ) . These data indicate that SENP7 is important for the host antimicrobial responses . Finally , we employed the HSV-1 infection model to investigate the in vivo function of SENP7 in innate immunity . First , the endogenous SENP7 was knocked down in mice , via tail vein injection of the SENP7-specific shRNA or control shRNA . The efficiency of the in vivo ‘knockdown’ was confirmed , and the depletion occurred most efficiently in the liver , kidney and spleen ( Fig 7A ) . Next , mice were injected intravenously with HSV-1 ( the sub-lethal dose ) , and their survival rates were monitored . As expected , SENP7 knockdown mice ( ‘SENP7-KD mice’ ) were more susceptible to HSV-1 than mock knockdown mice ( ‘control mice’ ) ( Fig 7B ) . All the SENP7-KD mice died within 3 days , whereas 83% of the infected control mice remained alive until 8 days after HSV-1 infection ( Fig 7B ) . Moreover , the IFN production of the mice was examined in vivo . Notably , the SENP7-KD mice displayed a more severe defect in the production of IFN-β and IFN-α4 in sera upon HSV-1 invasion , as compared with the infected control mice ( Fig 7C ) . The expression of Ifnb , Ifna4 and Cxcl10 mRNAs was significantly lower in SENP7-KD liver tissues ( Fig 7D ) . Consistently , HSV-1 virions were significantly more abundant in the brains of the SENP7-KD mice than those in control mice ( Fig 7E and 7F ) . In addition , the expression of Ifnb was also decreased in spleen or kidney tissues of SENP7-KD mice , compared to the control mice ( Fig 7G and 7H ) . Collectively , these data indicate that SENP7 is indispensable for protecting mice against HSV-1 infection .
Recent breakthroughs have uncovered the essential function of the cGAS-STING axis in monitoring cytosolic microbial DNAs , which triggers the expression of type-I interferons ( IFNs ) and pro-inflammatory cytokines . Given that cGAS senses DNAs in a sequence-independent manner , any aberrant activation of the cGAS by cytosolic DNAs contributes to the pathogenesis of chronic inflammation and autoimmune diseases . Hypothetically , the strength and duration of the cGAS-STING signaling is subjected to multiple layers of stringent modulations , so that the balance of the immune homeostasis could be appropriately maintained . It remains intriguing how cGAS is spatially and temporally modulated in immunity . In this study , we characterized the SUMO-specific protease SENP7 to potentiate cGAS activation via controlling its dynamic SUMOylation in immune responses . First , silencing of the endogenous SENP7 markedly attenuated the DNAs-triggered expressions of IFNs and ISGs; and this effect could be rescued by exogenously expressing siRNA-resistant SENP7 ( rSENP7 ) , but not by the catalytically dead rSENP7 ( C992S ) mutant . Second , silencing of SENP7 crippled the antimicrobial responses against HSV-1 and Listeria monocytogene in host cells . Third , the phosphorylation , dimerization and nuclear-translocation of IRF3 induced by the cytosolic DNAs were markedly impaired when silencing the endogenous SENP7 . Fourth , in vivo ‘knockdown’ of SENP7 made the mice more permissive to the HSV-1 infection and accelerated the death rate of the infected mice , due to the decreased production of interferons and cytokines . Fifth , the anti-microbial function of SENP7 depended on its de-SUMOylation catalytic activity . Mechanistically , we observed that cGAS could be modified by SUMO in vitro and in vivo , whereas another cytosolic DNA sensor DAI could not . This modification was further mapped onto the lysine residues 335 , 372 and 382 of cGAS . To substantiate , this SUMOylation markedly impaired the DNA-binding , oligomerization and nucleotidyl-transferase activities of cGAS , suggesting that the SUMO-moiety ( ~12 kDa ) might directly mask the DNA-binding surfaces or oligomerization interfaces . The X-ray crystal structure of the cGAS-dsDNA complex revealed that cGAS interacts with dsDNA through two binding sites ( site A and site B ) [2] . The DNA-binding triggers cGAS conformational changes to facilitate the subsequent formation of dimers and/or higher oligomers of cGAS through the dimer interface . Notably , the lysine residues 335 and 372 are located respectively in the site B and site A , while the lysine residue 382 is important for the cGAS dimerization . Our cell-based binding assays and functional assays further confirmed the importance of these residues for the cGAS activation . We speculated that the SUMOylation probably creates the steric hindrances for the cGAS recognition of DNAs and/or the cGAS dimerization . Another possibility is that the SUMOylation triggers a corresponding conformational change of cGAS , which influences the charge distribution of cGAS and abolishes the DNA access to cGAS . Future structural analysis of the SUMO-modified cGAS will hopefully provide insights to the specific mechanism . During revising this manuscript , Hu et al reported that SENP2 deactivates cGAS-STING signaling via modulating the equilibrium of cGAS and STING SUMOylation [31] . Notably , there are some discrepancies between the Hu et al’s and our studies concerning distinct SUMOylation pattern and the role of dynamic SUMOylation in cytosolic DNA sensing pathway . In Hu’s paper , SUMOylation study was performed in the presence of E3 ligase TRIM38 , whereas our study was carried out without specific E3 ligases . E3 ligase TRIM38 may catalyze and guide the SUMOylation reaction of cGAS in favor of SUMO-1 modification and covalently linking to specific lysine residues . It is intriguing to screen for potential other E3 ligases catalyzing cGAS SUMOylation and study the spatial and temporal modulations of cGAS activity by the different ligases-mediated SUMOylation . Additionally , it is notable that SENP2 not only possesses the deSUMOylation activity to remove SUMO moiety from cGAS , STING and IRF3 , but also involves in processing newly synthesized SUMOs into their mature forms [32] . Thus , it’s unknown whether SENP2 affects cGAS signaling in a direct or indirect manner . In contrast , our study showed SENP7 specifically targets cGAS to regulate the DNA-sensing pathway . In general , SUMO modification is highly dynamic that is potentially reversed by a family of the deconjugating enzymes ( SENP1-3 & SENP5-7 ) . SENPs are emerging as the versatile modulators for orchestrating innate and adaptive immune responses . SENP1 and SENP2 have been reported to regulate the transcriptional activities of IRF8 and IRF3 , respectively [33 , 34] . SENP1 also modulates the T cell synapse organization and T cell activation via controlling the sumoylation status of the kinase PKC-θ [35] . Recently , we characterized SENP6 to attenuate the activation of the inhibitor of NF-κB ( IκB ) kinase ( IKK ) via targeting NEMO , thus dampening TLR-induced inflammation [36] . The physiological functions of other SENPs remain largely unknown . In the current study , SENP7 was demonstrated to selectively interact with cGAS and catalyze the de-SUMOylation of cGAS . In contrast , SENP3 did not bind to cGAS or affect the cGAS SUMOylation . Furthermore , SENP3 influenced neither the STING signaling pathway nor the corresponding induction of IFNs and ISGs , highlighting the specific function of SENP7 in cGAS-STING axis . Interestingly , we observed that the interaction between cGAS and SENP7 was transient and dynamic , first increased to a maximum upon HSV-1 challenge and then dropped down after a while . This is consistent with our observation that the level of the SUMOylated cGAS inversely oscillated upon HSV-1 infection , revealing the vital role of SENP7 in fine-tuning the cGAS activity . The cGAS-STING signaling is recently established as instrumental for the onset and progression of autoimmune diseases [24 , 37–39] . Interestingly , we observed that SENP7 was significantly up-regulated in patients with autoimmune disease , indicating that SENP7 might synergize the action of cGAS in the pathogenesis of certain autoimmune disorders . The cGAS-STING signaling is also indispensable for the immune sensing of the immunogenic tumors , which ultimately induce the IFN-β production and activate dendritic cells , thus promoting the cross-priming of CD8+ T cells against tumor in vivo [40–42] . We speculate that SENP7 might modulate the tumor-related immune and inflammatory responses . Future investigation will focus on generating the SENP7 conditional knockout mice and analyzing its potential immunoregulatory function in autoimmunity and cancer immunology . Taken together , the current study demonstrated that the DNA sensor cGAS is dynamically modified by SUMO . This modification dampens the cGAS activation via suppressing its DNA-binding , oligomerization and nucleotidyl-transferase activities . To counterbalance , SENP7 relieves this inhibition by removing SUMO from cGAS . As an analogy , several recent studies demonstrated the delicate regulations of the adaptor protein STING [43] . For examples , STING is modified by K11 , K27 and K48 poly-ubiquitin chains in different contexts [20 , 44 , 45] . These poly-ubiquitinations are catalyzed by different ubiquitin E3 ligases and regulate a diverse aspects of the STING action . In addition , STING is phosphorylated by ULK and TBK1 , displaying antagonistic functional consequences [46 , 47] . It is currently unknown about the functional link between ubiquitination and phosphorylation of STING . We assume , a priori , that a more intricate network of the post-translational modifications of cGAS will appear in the coming years to deepen our understanding on how cGAS is fine-tuned or mis-modulated in physiological and pathological conditions .
The mice were maintained under specific pathogen-free ( SPF ) conditions at the Shanghai Institute of Biochemistry and Cell Biology . Animal experiments were carried out in strict accordance with the regulations in the Guide for the Care and Use of Laboratory Animals issued by the Ministry of Science and Technology of the People's Republic of China . The protocol was approved by the Institutional Animal Care and Use Committee of the Shanghai Institute of Biochemistry and Cell Biology , Chinese Academy of Sciences ( Permit Number: IBCB0027 Rev2 ) . The study of human peripheral blood samples was approved by the Research Ethics Board of Renji Hospital , Shanghai Jiao Tong University School of Medicine . All adult subjects provided informed consent , and a parent or guardian of any child participant provided informed consent on their behalf . Written informed consent was obtained before sample collection . The plasmid encoding shRNA was delivered into C57BL/6 mice with cationic liposomes PEI transfection reagent ( sigma ) according to the manufacturer’s instructions [48 , 49] . The Senp7-specific shRNA ( or control shRNA; 500 nM ) and PEI was each diluted into 100 μl of 5% glucose , then mixed and incubated for 15 min at room temperature at a final N/P ratio of 8 . Finally , the mixture ( 200 μl ) was injected into each mouse via tail vein . RNA derived from the PBMCs of patients with SLE , obtained from the Department of Rheumatology at Renji Hospital ( Shanghai , China ) , was used to quantify the gene expression of SENP7 and IFN-inducible genes . All patients met at least 4 of the American College of Rheumatology revised criteria for the classification of SLE [50] . Individuals with active virus infection were excluded . The patient group composed of 1 man and 26 women , the medium age was 29 years , range from 14 to 70 . RNA from the PBMCs of age and gender matched healthy volunteers were used as control . Peripheral blood samples ( 5 ml ) were collected from each subject and stored in tubes containing EDTA . PBMCs were isolated by Ficoll-Paque PLUS ( GE Healthcare ) gradient centrifugation and the RNA was extracted using TRIzol reagent according to the manufacture’s instruction . IFN scores were calculated as described in previous studies [51 , 52] . Briefly , the mean and standard deviation ( SD ) for the expression level of each IFN-inducible genes Gene ( i ) in the healthy donor ( HD ) group ( mean Gene ( i ) HD and SD Gene ( i ) HD , respectively ) were used to obtain a standardized expression level ( ST ) of each gene for each SLE patient . Then , the standardized expression levels for each patient were summed to obtain IFN score: ∑ ( i=1 ) 5ST[Gene ( i ) ]=[Gene ( i ) SLE−mean Gene ( i ) HD]/SD Gene ( i ) HD; where i = each of the 3 IFN-inducible genes ( ISG15 , MX1 , OAS1 ) , Gene ( i ) SLE is the expression level of a particular gene in a given SLE patient . HEK293T ( ATCC ) , MEF ( ATCC ) , L929 ( ATCC ) and HEK293 ( ATCC ) cells were cultured using DMEM ( Invitrogen ) plus 10% FBS ( Gibco ) , supplemented with 1% penicillin-streptomycin ( Invitrogen ) . Vero cells ( ATCC ) were cultured in MEM ( SAFC Biosciences ) supplemented with 10% FBS and 1% penicillin-streptomycin . Lipofectamine 2000 ( Invitrogen ) was used for transient transfection of HEK293T and HEK293 Cells . MEF cells were transfected with X-Gene HP ( Roche ) . Small interference RNA was transfected with Lipofectamine 2000 ( Invitrogen ) according to the manufacturer’s instructions . The siRNA duplexes targeting SENPs were chemically synthesized by Gene-Pharma . The sequences of siRNAs are shown as follows: mouse Senp7-440# , 5′- GGA CGA GAA UUC AGA AAG ATT -3′; mouse Senp7-3880# , 5′- GAG CUC AUC UGU UCA UAU ATT -3′; human SENP3 , 5′-GCU UCC GAG UGG CUU AUA ATT-3′; human SENP7 , 5′-CAA AGU ACC GAG UCG AAU AUU-3′; The nonspecific siRNA ( N . C . ) , 5′-UUC UCC GAA CGU GUC ACG UTT-3′ . Human codon-optimized Cas9 ( hCas9 ) and Senp7-targeting gRNA-expressing plasmid ( pLentiCRISPR ) was obtained from GenScript . The target sequence used is 5′-ATC ACC AGC TGA TTT ACA GA-3′ . To construct the knockout cell line , MEFs were infected with the lentivirus within which pLentiCRISPR was packaged for two days . Single clones were sorted into a 96-well plate and cultured in the presence of puromycin ( 8 μ g/mL ) . On-target CRISPR/Cas9 events were detected by the T7 endonuclease I-cutting assay and the candidate knockout clones were confirmed by sequencing . SENP7 , cGAS , RIG-I , STING , TBK1 , IRF3 cDNAs were constructed using standard PCR techniques from the thymus cDNA library and subsequently inserted into mammalian expression vectors as indicated . The reporter plasmids ( IFN-β-luciferase , PRDIII-I-luciferase , NF-κB-luciferase and pTK-Renilla ) have been described previously [53 , 54] . SUMO-1 , SUMO-3 constructs were kindly provided by Dr . Jinke Cheng ( School of Medicine , Shanghai Jiao Tong University , Shanghai , China ) . The SENP7 siRNA-resistant form was generated with silent mutations introduced into the siRNA target sequence . All point mutations were generated by using a QuickChange XL site-directed mutagenesis method ( Stratagene ) . All the plasmids were verified by sequencing . Recombinant GST-fusion proteins were purified from Escherichia coli ( BL21 ) by using glutathione-Sepharose 4B resin ( GE Healthcare , Piscataway , NJ ) . HSV-1 and HSV-1-GFP were kindly provided by Dr . Wentao Qiao ( Nankai University ) and Dr . Chunfu Zheng ( Suzhou University ) , respectively . HSV-1 was propagated and titered by plague assays on Vero Cells . Listeria monocytogenes ( 10403 serotype ) was a gift from Dr . Youcun Qian ( Institute of Health Sciences ) . Listeria monocytogenes was cultured in 3 . 7% Brain-Heart Infusion broth ( BD Biosciences ) . For immunoprecipitation assay , cells extracts were prepared by using RIPA buffer ( 50 mM Tris-HCl pH 7 . 4 , 150 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% SDS , 0 . 5% deoxycholate ) supplemented with a complete protease inhibitor cocktail ( Roche ) , a PhosSTOP phosphatase inhibitor cocktail ( Roche ) and 20 mM N-ethylmaleimide ( NEM ) . Lysates were incubated with the appropriate antibody for four hours to overnight at 4°C before adding protein A/G agarose for 2 hr . The immunoprecipitates were washed three times with the same buffer and eluted with SDS loading buffer by boiling for 5 min . For denaturing immunoprecipitation , cells were lysed in 1% SDS buffer ( 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 1% SDS , 10 mM DTT ) and boiled for thirty minutes . The lysates were centrifuged and diluted by 10-fold with Lysis buffer ( 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 1 mM EDTA , 1% Triton X-100 ) . The diluted lysates were immunoprecipitated with the indicated antibodies for four hours to overnight at 4°C before adding protein A/G agarose for 2 hr . After extensive wash , the immunoprecipitates were subjected to immunoblot analysis . For immunoblot analysis , the samples were subjected to SDS-PAGE . The resolved proteins were then electrically transferred to a PVDF membrane ( Millipore ) . Immunoblotting was probed with indicated antibodies . The protein bands were visualized by using a SuperSignal West Pico chemiluminescence ECL kit ( Pierce ) . Signal intensities of immunoblot bands were quantified by Image J software . Cells were transfected with reporter plasmids . Luciferase activity was assessed with a dual luciferase assay kit ( Promega ) and a Luminoskan Ascent luminometer ( Thermo Scientific ) [55] . The polyclonal antibody against SENP7 was generated by immunizing rabbit with recombinant mouse SENP7 ( 716–1010 aa ) . The rabbit polyclonal antibody against cGAS was obtained by immunizing rabbit with recombinant mouse cGAS ( 362–643 aa ) . The antibodies against hemagglutinin ( HA ) , Myc and IRF3 were purchased from Santa Cruz Biotechnology . Mouse monoclonal Flag antibody and β-actin antibodies were obtained from Sigma-Aldrich . The SUMO antibody was from Abcam . Phospho-IRF3 antibody , Phospho-TBK1 antibody , and rabbit DYKDDDDK tag antibody were from Cell Signaling Technology . TBK1 antibody was from Abcam . Anti-Flag ( M2 ) -agarose and EZview red anti-HA affinity gel were from Sigma . Poly ( dA:dT ) was obtained from Sigma . Interferon stimulatory DNA ( ISD ) was prepared by annealing equimolar amounts of sense and antisense DNA oligonucleotides at 95°C for 10 min before cooling to room temperature . Oligonucleotides used are as follows: ISD ( sense ) , 5′-TAC AGA TCT ACT AGT GAT CTA TGA CTG ATC TGT ACA TGA TCT ACA-3′; ISD ( antisense ) , 5′-TGT AGA TCA TGT ACA GAT CAG TCA TAG ATC ACT AGT AGA TCT GTA-3′ . cGAMP was from InvivoGen and was delivered into cultured cells by digitonin permeabilization method as previously described ( Girardin et al . , 2003 ) . Total RNA was isolated from indicated cells by using TRIzol reagent ( Invitrogen ) according to the manufacturer’s instructions , and then subjected to reverse transcription . The quantifications of gene transcripts were performed by real-time PCR using Power SYBR GREEN PCR MASTER MIX ( ABI ) . GAPDH served as an internal control . PCR primers used to amplify the target genes are shown as follows: Gapdh: sense ( 5′-GAA GGG CTC ATG ACC ACA GT-3′ ) , antisense ( 5′-GGA TGC AGG GAT GAT GTT CT-3′ ) ; Ifnb: sense ( 5′-AGA TCA ACC TCA CCT ACA GG-3′ ) , antisense ( 5′-TCA GAA ACA CTG TCT GCT GG-3′ ) ; Ifna4: sense ( 5′-ACC CAC AGC CCA GAG AGT GAC C-3′ ) , antisense ( 5′-AGG CCC TCT TGT TCC CGA GGT-3′ ) ; Cxcl10: sense ( 5′-CCT GCC CAC GTG TTG AGA T-3′ ) , antisense ( 5′-TGA TGG TCT TAG ATT CCG GAT TC-3′ ) ; OAS1: sense ( 5′-GAA GGC AGC TCA CGA AAC-3′ ) , antisense ( 5′-TTC TTA AAG CAT GGG TAA TTC-3′ ) ; MX1: sense ( 5′-GGG TAG CCA CTG GAC TGA-3′ ) , antisense ( 5′-AGG TGG AGC GAT TCT GAG′ ) ; ISG15: sense ( 5′-TGT CGG TGT CAG AGC TGA AG-3′ ) , antisense ( 5′-GCC CTT GTT ATT CCT CAC CA′ ) . Concentrations of the cytokine in culture supernatants were measured by ELISA kit ( R&D Systems ) according to the manufacturer’s instructions . Native gel electrophoresis for IRF3 dimerization was carried out as described previously [56] . Cells grown on coverslips were fixed for 15 min with 4% paraformaldehyde in PBS , permeabilized for 20 min in 0 . 1% Triton X-100 in PBS and blocked using 5% BSA for 1 hr . Then , the cells were stained with the indicated primary antibodies followed by incubation with fluorescent-conjugated secondary antibodies ( Jackson ImmunoResearch ) . Nuclei were counterstained with DAPI ( 4 , 6-diamidino- 2-phenylindole ) ( Sigma-Aldrich ) . Slides were mounted using Aqua-Poly/ Mount ( Dako ) . Images were captured using a Leica laser scanning confocal microscopy . The acquiring software was TCS ( Leica , Solms , Germany ) . For Ni-nitrilotriacetic acid resin ( NTA ) pulldown analysis , cells were lysed in His-Lysis Buffer ( 50 mM Tris-HCl pH 7 . 4 , 300 mM NaCl , 1% Triton X-100 , 20 mM imidazole , 10 mM β-ME ) supplemented with 1 mM PMSF . After centrifugation , the supernatants were collected and incubated with 20 μL Ni-NTA agarose beads ( Qiagen ) for 4 hr at 4°C . After extensively washing with His-Lysis Buffer containing 20 mM imidazole , the precipitates were subjected to SDS-PAGE followed by immunoblot analysis or eluted with TBS containing 300 mM imidazole and subsequently subjected to biotin-pulldown analysis . Purified His-Flag-cGAS , His-SUMO was mixed with E1 ( 50 nM ) and E2 ( 0 . 3 μM ) ( Boston Biochem ) in a reaction buffer containing 50 mM Tris-HCl , pH 7 . 5 , 5 mM MgCl2 , 2 mM ATP . The reaction was carried out at 37°C for 120 min and then resolved by SDS-PAGE . SUMOylated products were detected by immunoblotting with indicated antibodies . A total of 10 μM purified cGAS or mutant variants were incubated with the indicated DNA and reaction buffer ( 20 mM HEPES , pH 7 . 5 , 5 mM MgCl2 , 2 mM ATP , 2 mM GTP ) at 37°C for 2 hr . Samples were centrifuged at 16 , 000 g for 10 min . The product in the supernatant was separated from cGAS and DNA by ultrafiltration . The samples were diluted by 5-fold and analyzed on a MonoQ ion exchange column ( GE Healthcare ) equilibrated with the running buffer ( 50 mM Tris-HCl pH 8 . 5 ) and eluted with a NaCl gradient of 0 to 0 . 5 M in the running buffer . The plasmids GFP-cGAS /mCherry-cGAS , which express GFP- and mCherry-tagged proteins at a ratio of 1:1 , were transfected into HeLa cells . To calculate the apparent efficiency of FRET , we used the following two spectra obtained during the process of generating the FRET emission spectrum . The GFP-emission spectrum was obtained once before and once after photobleaching cherry . Each data set was based on >20 individual cells . FRET efficiency was calculated with the following formula: FRET% = ( GFP after bleaching − GFP before bleaching ) / GFP after bleaching ) × 100 . Statistical significance was determined with an unpaired t-test . Tissues were fixed in 4% paraformaldehyde , embedded in paraffin , cut into sections , and placed on adhesion microscope slides . Sections were subjected to immunohistochemical ( IHC ) staining according to standard procedures . The anti-HSV-1 antibody ( Abcam ) was used for staining . Student’s t test was used for the statistical analysis of two independent treatments . Mouse survival curves and statistics were analyzed using the Mantel-Cox Log-rank test . A non-parametric Mann-Whitney test was used to compare the gene expression in SLE patients and normal healthy subjects . Non-parametric correlation test ( Spearman's rank correlation coefficient ) were used to measure the degree of association between the expression of SENP7 and IFN score . For all tests , a P value of < 0 . 05 was considered statistically significant . | The Cyclic GMP-AMP ( cGAMP ) synthase ( cGAS , a . k . a . MB21D1 ) is critical for monitoring the pathogen-derived DNA upon microbial infection . Its activity should be dynamically modulated in case the inadvertent recognition of the aberrant self nucleic acids in cytosol leads to severe autoimmune diseases . Protein posttranslational modifications dynamically shape the strength and duration of the immune signaling pathways . It is intriguing to explore whether SUMOylation could modulate the cGAS-initiated signaling . In this study , we characterized sentrin/SUMO-specific protease 7 ( SENP7 ) to specifically potentiate the cGAS activation . Upon microbial DNA challenge , the small ubiquitin-like modifier ( SUMO ) was conjugated onto cGAS , which suppressed its DNA-binding , oligomerization and nucleotidyl-transferase activities . SENP7 reversed this inhibition via catalyzing the de-SUMOylation of cGAS . Our study sheds new light on the dynamic function of the SUMOylation in cytosolic DNAs-triggered innate immunity response . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"immune",
"physiology",
"chemical",
"compounds",
"gene",
"regulation",
"immunology",
"organic",
"compounds",
"fluorophotometry",
"sumoylation",
"immunoprecipitation",
"basic",
"amino",
"acids",
"amino... | 2017 | SENP7 Potentiates cGAS Activation by Relieving SUMO-Mediated Inhibition of Cytosolic DNA Sensing |
Our ability to control diseases caused by parasitic nematodes is constrained by a limited portfolio of effective drugs and a paucity of robust tools to investigate parasitic nematode biology . RNA interference ( RNAi ) is a reverse-genetics tool with great potential to identify novel drug targets and interrogate parasite gene function , but present RNAi protocols for parasitic nematodes , which remove the parasite from the host and execute RNAi in vitro , are unreliable and inconsistent . We have established an alternative in vivo RNAi protocol targeting the filarial nematode Brugia malayi as it develops in an intermediate host , the mosquito Aedes aegypti . Injection of worm-derived short interfering RNA ( siRNA ) and double stranded RNA ( dsRNA ) into parasitized mosquitoes elicits suppression of B . malayi target gene transcript abundance in a concentration-dependent fashion . The suppression of this gene , a cathepsin L-like cysteine protease ( Bm-cpl-1 ) is specific and profound , both injection of siRNA and dsRNA reduce transcript abundance by 83% . In vivo Bm-cpl-1 suppression results in multiple aberrant phenotypes; worm motility is inhibited by up to 69% and parasites exhibit slow-moving , kinked and partial-paralysis postures . Bm-cpl-1 suppression also retards worm growth by 48% . Bm-cpl-1 suppression ultimately prevents parasite development within the mosquito and effectively abolishes transmission potential because parasites do not migrate to the head and proboscis . Finally , Bm-cpl-1 suppression decreases parasite burden and increases mosquito survival . This is the first demonstration of in vivo RNAi in animal parasitic nematodes and results indicate this protocol is more effective than existing in vitro RNAi methods . The potential of this new protocol to investigate parasitic nematode biology and to identify and validate novel anthelmintic drug targets is discussed .
Lymphatic filariasis is a disease caused by filarial nematodes including Wuchereria bancrofti and Brugia malayi , transmitted through the bite of infected mosquitoes . These parasites perpetuate socioeconomic instability in developing countries by inflicting crippling morbidity and debilitating stigmatization . The impact of this disease is vast - over 120 million people are infected in 81 endemic countries [1] . In an effort to alleviate morbidity and eliminate transmission of this disease , the Global Program for the Elimination of Lymphatic Filariasis ( GPELF ) has orchestrated a systematic mass drug administration ( MDA ) program centered on the repeated dosing of either diethylcarbamazine citrate ( DEC ) and albendazole or albendazole and ivermectin in areas where the other filarial parasites , Onchocerca volvulus and Loa loa are co-endemic . This strategy has reduced prevalence in many areas [2] but lymphatic filariasis remains a significant global health concern . Many factors contribute to continued transmission , but central is the inadequate portfolio of effective drugs; none of the MDA drugs are effective against all life stages of the parasite with notable inefficacy against adult worms [3]–[5] . This means MDA must be provided annually for the duration of the lifespan of adult parasites . This situation is compounded by gaps in our understanding of mechanisms of drug action and pharmacology – the site of action of DEC is unknown despite being the drug of choice for lymphatic filariasis control for decades , and the filaricidal mechanism of ivermectin at therapeutic concentrations is also equivocal . There is a very real and significant need for additional and more effective antifilarial drugs , and a better understanding of the mode of action of existing drugs [6] . A major obstacle to the rational development of such drugs is the experimental intractability of parasitic nematodes . An example of this complication is RNA interference ( RNAi ) , a reverse genetic tool that allows researchers to rapidly and specifically ‘turn off’ genes of interest . RNAi has fast become a standard tool in rational drug discovery for the identification and validation of potential new drug targets [7] , [8] . By suppressing specific genes and examining the resulting phenotype , it is possible to delineate gene function and appraise the potential value of encoded proteins as drug targets . Successful applications of present RNAi protocols to parasitic nematodes have been sporadically reported , limited in their effectiveness and seldom repeated [9] . Some success has been achieved with Nippostrongylus brasiliensis [10] , Ascaris suum [11] , Trichostrongylus colubriformis [12] , Ostertagia ostertagi [13] and Haemonchus contortus [14] , [15] . Germane to the study of filarial worms , RNAi has been described in B . malayi [16] , [17] , Onchocerca volvulus [18] , [19] and Litomosoides sigmodontis [20] . The conclusion has been reached , however , that successful RNAi “only works on a limited number of genes , and in some cases the effect is small and difficult to reproduce” [14] . The inability to depend on present RNAi protocols with parasitic nematodes has proved a major stumbling block to the identification and validation of new drug targets , to a better understanding of anthelmintic mode of action , and to advancing our comprehension of parasite biology . The recalcitrance of animal parasitic nematodes to RNAi is perplexing , given that Caenorhabditis elegans , a free-living nematode , and plant parasitic nematodes are readily susceptible to the technique [21]–[28] . One hypothesis advanced to explain this recalcitrance is that because present RNAi protocols employ in vitro approaches including soaking nematodes in an RNAi trigger , feeding nematodes bacteria producing the trigger , or electroporating of the trigger into the parasite , the RNAi trigger is not provided in a manner conducive to systemic gene suppression [29] . Implicit in the use of these protocols is the removal of a parasite from the host and its maintenance in a liquid culture . Therefore these protocols have distinct drawbacks such as difficulty maintaining healthy , viable worms that behave normally in vitro , limitation of use of parasites or life stages for which in vitro culture is defined , and poor efficacy in RNAi trigger delivery methods that can prove lethal to the parasite [30] . The aim of this study was to develop an innovative in vivo approach to RNAi in parasitic nematodes that overcomes the drawbacks associated with present in vitro experimental paradigms . Our approach is based on the filarial nematode B . malayi . We establish a B . malayi infection in an intermediate host , the mosquito Aedes aegypti , and then initiate suppression of parasite genes by injecting an RNAi trigger directly into the mosquito . The mosquito acts as an ideal culture and delivery system , ensuring the RNAi trigger is exposed to healthy , developing parasites . Using this approach we have effectively and quantifiably suppressed expression of Bm-cpl-1 , a B . malayi gene encoding a cathepsin L-like cysteine protease . Dramatic aberrant phenotypes accompany this suppression , including a marked retardation of motility , an inhibition of normal parasite migration behavior within the mosquito and impaired parasite growth and development . Suppression is specific; non-target RNAi has no effect on nematode viability or behavior , and the level of gene suppression and extent of the resultant phenotypes suggest this new protocol is more effective than previous methods . The development of an in vivo RNAi protocol to reliably suppress gene expression in filarial worms has great potential for the identification and validation of novel drug targets , and more broadly , to explore parasitic nematode biology and host-parasite interactions .
Our hypothesis is that mosquitoes provide an optimal culture and delivery system for an RNAi trigger targeted to developing Brugia malayi parasites . Healthy , viable , developing parasites are subjected to the RNAi trigger because the parasites undergo growth and development in the mosquito intermediate host . In order to test the extent of dissemination of the RNAi trigger from the site of intrathoracic injection , 150 ng of an equimolar mix of four 3′ Cy 3-labelled Bm-cpl-1 siRNAs was injected into adult Aedes aegypti mosquitoes as described . The dissemination of this RNAi trigger through the mosquito was tracked over 15 d post-injection by periodic microdissection of the mosquito and evaluation of internal fluorescence compared to saline injected controls . The labeled siRNA mix spread rapidly from the site of injection and maximal fluorescence was observed 24 h post-injection ( Fig . 1 ) . The intensity of fluorescence slowly decreased until reaching basal levels at five d post-injection after which fluorescence intensity was not appreciably different from control mosquitoes . Our observations closely parallel those of a previous report that describes the spread of 140 ng AlexaFluor 555-labeled siRNA in the mosquito Anopheles gambiae from an injection site to the midgut and pericardial cells 36 h post-injection [31] . Systemic dispersion and persistence of RNAi signal from the site of injection suggests B . malayi larvae are likely to be exposed to the RNAi trigger in our experimental model . Recently it has been shown that B . malayi genes encoding cathepsin L-like enzymes can be suppressed in vitro by soaking adult parasites in culture media containing siRNA [17] . We tested the capacity of our novel methodology to suppress larval stage B . malayi gene expression in vivo by injecting mixed siRNAs specific to the cathepsin L-like Bm-cpl-1 gene directly into Ae . aegypti mosquitoes harboring L3 stage B . malayi parasites . Gene suppression was assayed 48 h post-injection using a semi-quantitative RT-PCR in which the intensity of Bm-cpl-1 amplification in the linear phase of the reaction was compared to an internal B . malayi reference gene ( Bm-flp-14 ) that is expressed stably and at comparable levels to Bm-cpl-1 . Control mosquitoes were injected with equal volumes of Aedes physiologic saline . This methodology was optimized to amplify Bm-cpl-1 from a heterogeneous mosquito/parasite total RNA preparation from a single mosquito . Suppression was concentration-dependent because injection of 0 . 15 ng siRNA did not appear to reduce Bm-cpl-1 transcript levels . However , injection of 15 ng or 1 . 5 ng of siRNA decreased transcript levels , and injection of 150 ng mixed siRNA into mosquitoes profoundly suppressed Bm-cpl-1 expression; the target parasite gene could not be amplified ( Fig . 2 ) . This suppression was also specific; expression of the Bm-flp-14 reference gene was unaffected by siRNA injection and target gene expression was normal in saline-injected controls . Application of dsRNA is the commonly used method for triggering RNAi in parasitic nematodes and has advantages over siRNA; dsRNA can be generated in-house more quickly than commercially produced siRNAs at lower cost . B . malayi-infected mosquitoes were also subjected to treatment with dsRNA as an RNAi trigger . The effect of dsRNA was concentration-dependent such that injection of 15 ng dsRNA results in Bm-cpl-1 suppression but 1 . 5 ng dsRNA had no appreciable effect . Injection of 150 ng of dsRNA potently suppressed Bm-cpl-1 transcript abundance and suppression appeared specific , with Bm-flp-14 expression unaffected by dsRNA ( Fig . 2 ) . RT-qPCR was used to quantify the level of Bm-cpl-1 suppression relative to two reference genes ( Bm-flp-14 and Bm-tph-1 ) using the efficiency-corrected ( EΔΔCq ) relative quantification method [32] . PREXCEL-Q software was used to optimize the performance of the RT-qPCR assay; and important data pertinent to PCR efficiency , linear dynamic range and normalization of the assay are documented in Table 1 . Bm-tph-1 showed stable Cq values across the experiment and therefore was the most appropriate reference gene for these studies , as shown previously [33] . The suppressive effects of both RNAi treatments were almost identical; injection of 150 ng siRNA reduced Bm-cpl-1 transcript by 83% compared to saline-injected controls ( P<0 . 0001 ) and 150 ng dsRNA also reduced Bm-cpl-1 transcript by 83% ( P<0 . 0001 ) ( Fig . 3 ) . Bm-flp-14 reference gene transcript was slightly reduced by both RNAi treatments but these reductions were not significant ( siRNA , 9% , P = 0 . 38; dsRNA , 12% , P = 0 . 17 ) . These data support the gel-based semi-quantitative RT-PCR experimental findings and demonstrate the efficacy of this novel method of RNAi delivery . Previous studies have described aberrant filarial worm phenotypes associated with cathepsin L-like gene suppression in vitro including decreased microfilariae ( mf ) release from adult B . malayi [17] and an inhibition of the L3 to L4 molt in Onchocerca volvulus [18] . Based on these data , we predicted that Bm-cpl-1 suppression would produce a phenotype in vivo in the mosquito host . Mosquitoes were injected with 150 ng Bm-cpl-1 dsRNA 10 d post-infection ( dpi ) then microdissected four d post-injection to harvest L3-stage parasites . Worm motility was digitally recorded and scored according to a five-point schema of one ( immobile ) , to five ( all parts of the worm in constant motion ) . 100% of control worms from mosquitoes injected with Aedes physiologic saline were categorized as four or five on this scale . Bm-cpl-1 suppression significantly inhibited this normal worm motility ( P<0 . 001 ) , with only 67% of worms ranked as four or five on the scale ( Fig . 4 ) . To confirm that this effect was Bm-cpl-1 specific and not due to exogenous dsRNA impairing worm viability , this experiment was repeated with dsRNA for enhanced GFP ( eGFP ) as a random exogenous RNA . These worms were phenotypically identical to saline-injected controls ( 100% category four or five ) , confirming the specificity of the aberrant phenotype in Bm-cpl-1 suppressed worms . The effect of changing the timing of Bm-cpl-1 suppression on worm motility was also examined . Bm-cpl-1 transcript levels are elevated in L3 stage parasites , such that this gene has a purported role in the L3 to L4 molt [18] . The temporal expression of Bm-cpl-1 was reported to be up-regulated during the L2 to L3 transition , at six to seven dpi [34] . Based on the timing of Brugia development in Ae . aegypti [35] , infected mosquitoes were injected with Bm-cpl-1 dsRNA at 10 dpi in order to target L3-stage worms ( described above ) and at seven dpi to target the L2 to L3 transition . Parasites exposed to Bm-cpl-1 dsRNA at seven dpi showed significantly inhibited motility compared to saline controls ( P<0 . 001 ) with only 31% of worms displaying normal motility . The difference between parasites exposed to dsRNA at seven and 10 dpi was significant ( P<0 . 001 ) , and may reflect an important biological role for Bm-cpl-1 during the transition from L2 to L3 stages . More explicitly , earlier exposure to the RNAi trigger could impose more significant detrimental impact on the parasite by disrupting the L2 to L3 molt , or it may simply be a consequence of the longer period of time from gene suppression to phenotype assay , allowing Bm-CPL-1 rundown and maturation of the phenotype . In addition to depressed activity , other morphological and motility phenotypes were apparent in Bm-cpl-1 suppressed worms . A highly active , convoluted body form characterizes motility of healthy B . malayi L3s , both the heads and tails of the parasites in particular are conspicuously tortuous – curvature we described as ‘knotted’ . Control worms from saline-injected mosquitoes frequently ( 86% of worms ) displayed knotting at both ends . Suppression of Bm-cpl-1 10 dpi significantly inhibited this motility , because only 14% of worms presented with both ends knotted ( P<0 . 001 ) ( Fig . 5 ) . This phenotype was enhanced by an early suppression of Bm-cpl-1 at seven dpi such that no Bm-cpl-1 suppressed parasites exhibited this knotting morphology . The difference between L2 and L3 Bm-cpl-1 suppression was significant ( P = 0 . 005 ) . Worms exposed to the exogenous eGFP dsRNA control confirmed that this phenotype was gene-specific because parasite motility was not significantly different from saline controls ( 85% knotted at both ends , P = 0 . 2 ) . Another aberrant motility observed was the presence of a perturbed section of body wall slightly caudal to the midpoint of the worm . This abnormal kinked morphology was absent from control worms ( 4% of worms from saline-injected and 0% from eGFP-injected mosquitoes displayed this morphology ) , but evident with significantly greater frequency in 10 dpi Bm-cpl-1 suppressed worms ( 47% , P<0 . 001 ) ( Fig . 6 ) . This kink rate increased with Bm-cpl-1 suppression at seven dpi ( 63% ) , but compared to 10 dpi this was not significant ( P = 0 . 08 ) . Finally , partial paralysis of Bm-cpl-1 suppressed worms was evident , presenting as immobility in the caudal third of the worm . This paralysis was observed in 61% of 10 dpi Bm-cpl-1 suppressed worms , and 83% of seven dpi suppressed worms ( this increase was significant , P = 0 . 005 ) but was generally absent from control worms ( 5% of worms from saline-injected mosquitoes and 3% of worms from eGFP-injected mosquitoes ) ( Fig . 6 ) . To examine the consequence of this aberrant motility on B . malayi development , mosquitoes were injected with 150 ng Bm-cpl-1 dsRNA 10 dpi then microdissected four d post-injection , partitioning the mosquitoes into head , thorax and abdomen preparations . Control worms from mosquitoes injected with either saline or eGFP dsRNA were found exclusively ( 100% ) in head preparations as expected . Bm-cpl-1 suppressed worms were most frequently observed escaping from the thorax and abdomen ( Fig . 7 ) . Parasites in Bm-cpl-1 dsRNA-injected mosquitoes , however , did not leave the thorax ( 94% of worms were found here ) or abdomen ( 6% ) . Bm-cpl-1 suppression , therefore , prevents worm migration to the head of the mosquito , effectively preventing normal progression of the parasite life cycle and thus abolishing the potential for parasite transmission . A significant negative effect also was seen on growth and development of parasites subjected to Bm-cpl-1 suppression . Parasitized mosquitoes were injected with 150 ng Bm-cpl-1 dsRNA seven dpi then microdissected 14 dpi and the length , width and appearance of the worms recorded . Mosquitoes were injected seven dpi because the previous motility experiments dictated that this experimental timing generated the most pronounced motility phenotypes . Bm-cpl-1 suppression significantly reduced the length of L3 worms by 48% ( P<0 . 0001 ) ( Fig . 8 ) . The mean length of control L3 , removed from mosquitoes seven d after saline injection and 14 dpi , was 1347±18 µm . This was reduced to 700±49 µm after RNAi treatment . Unlike parasite length , width was not significantly affected by Bm-cpl-1 suppression , although a slight decrease of 5% was observed ( P = 0 . 39 ) from 31±1 µm in control worms to 30±2 µm in RNAi worms . In addition to worm length , the majority of Bm-cpl-1 dsRNA parasites also presented with additional aberrant developmental phenotypes . Most evident was a disruption of the cuticle ( Fig . 8B ) , which extended significantly beyond the body of the worm . Some degree of this cuticular sloughing was noticed in most worms but the severity of this phenotype was variable . Finally , the integrity of the gut appeared compromised in Bm-cpl-1 worms . In such instances , the gastrointestinal tract of the parasites appeared incomplete and porous when examined at the light microscope level . Phenotype data resoundingly demonstrate that Bm-cpl-1 suppression decreases B . malayi viability . It is logical to predict that this decreased viability would also have an impact on mosquito survival and prevalence of parasite infection . To examine this , mosquitoes were injected with 150 ng Bm-cpl-1 dsRNA 10 dpi and the number of mosquitoes that survived through the development of parasites to the infectious stage , 14 dpi , was counted . Surviving mosquitoes then were microdissected to determine the proportion that harbored parasite infections . Bm-cpl-1 suppression increases host mosquito survival . The survival rate of control mosquitoes injected with saline or eGFP dsRNA was 62% and 65% respectively ( P = 0 . 6 ) as compared to 80% in Bm-cpl-1 RNAi-exposed mosquitoes ( P<0 . 001 ) ( Fig . 9A ) . Early suppression of Bm-cpl-1 at seven dpi enhanced the phenotype even more significantly such that 93% of mosquitoes were alive at the termination of the experiment ( P = 0 . 007 ) . This increased mosquito survival rate after Bm-cpl-1 suppression may be as a result of the parasite's compromised ability to feed on , and migrate through , the host or may result from a more successful or effective host response against parasites with decreased viability . This hypothesis is supported by our observation that Bm-cpl-1 suppression also decreased prevalence of infection – fewer surviving mosquitoes harbored parasites after Bm-cpl-1 RNAi ( Fig . 9B ) . Every surviving mosquito injected with saline or eGFP was found to contain parasites 14 dpi , but 14 dpi Bm-cpl-1 suppressed parasites ( exposed to dsRNA at 10 dpi ) were found in just 76% of mosquitoes , a significant reduction in prevalence ( P<0 . 001 ) . Prevalence was further reduced to 62% in parasites exposed to Bm-cpl-1 dsRNA at seven dpi , a statistically significant decrease compared to worms exposed to the dsRNA trigger at 10 dpi ( P = 0 . 03 ) .
Here we report the development of a novel in vivo approach to RNAi in the filarial nematode Brugia malayi , and describe its application first to suppress the expression of Bm-cpl-1 , a B . malayi gene encoding a cathepsin L-like cysteine protease , then to validate this gene as a potentially potent anthelmintic drug target . To the best of our knowledge , this is the first description of in vivo RNAi in parasitic nematodes and represents an advance in the study of filarial nematode biology that may aid in the development of drugs to combat parasitic nematode infection . The rationale for developing an in vivo RNAi protocol stems from the hypothesis that RNAi is ineffective in animal parasitic nematodes because the supply of an RNAi trigger to the worms is inappropriate [29] . Our overarching hypothesis was that RNAi would work effectively and robustly if a trigger is supplied to healthy , viable worms in a host environment . Supporting this hypothesis , we were able to specifically reduce target gene transcript abundance in B . malayi larvae by 83% by supplying an RNAi trigger to parasites developing within the mosquito host . This level of transcript knockdown has not previously been reported using current in vitro RNAi soaking methods . The ‘in squito’ approach to RNAi we describe is effective for the specific suppression of cathepsin genes in Brugia larval stages as they develop within their cognate mosquito host; it is therefore possible that this in vivo approach may represent a more effective means of eliciting gene suppression in filarial nematodes . The mechanism by which the RNAi trigger is delivered to the parasites ‘in squito’ is unclear , but could be a result of bathing the parasite in the trigger within a cell , or as a result of uptake by tissue ingestion . In support of the former , Cy3-labeled siRNA injected into the haemocoel rapidly disseminates throughout the mosquito supporting a hypothesis that the developing parasites are effectively incubating in a host milieu containing an RNAi trigger , essentially a scenario analogous to in vitro RNAi by soaking . If this is the case , the ‘in squito’ approach represents an efficient way to generate gene suppression by soaking . Most successful animal parasitic nematode in vitro soaking protocols use large amounts of ds- or siRNA with concentrations of 1 mg/ml typical , meaning anywhere between 25 µg and 2 mg of RNAi trigger are required per suppression experiment [10]–[19] , with the exception of one report showing that lower trigger concentrations could still be effective at producing gene suppression by soaking [20] . Here we showed that gene suppression can be achieved using just 150 ng of ds- or siRNA per RNAi event , and indeed , a reduction in transcript abundance was observed after injecting as little as 15 ng dsRNA . In addition to the obvious cost saving advantages to performing RNAi experiments in this manner , such low RNA concentrations may also improve the specificity of gene suppression . Soaking plant parasitic nematodes in serial dilutions of ds- and siRNAs has been shown to reduce off-target effects in RNAi experiments [28] , [36] . A second delivery hypothesis is that the developing parasites are ingesting the RNAi trigger . Microfilariae taken in during the blood meal rapidly penetrate the mosquito midgut [37] , and migrate to the thoracic musculature where they grow and develop to the L3 stage [38] , [39] , a process completed in under two weeks [35] . From the L2 stage , the developing larvae are active feeders and consume host tissue [35] , [40] , [41] , a behavior that would lead to the ingestion of an injected RNAi trigger in our experimental model . RNAi by feeding is a well-established method in free-living nematodes [22] , [23] , [42] , [43]; by feeding these worms bacteria expressing dsRNA , systemic gene suppression can be effected in a relatively simple and efficient manner . This approach has not been successful with parasitic nematodes , however , as most parasitic species are not bacteriotrophic , and even for those species with bacteriotrophic life stages , this method is unreliable [30] . Resolution of the RNAi trigger delivery mechanism afforded by our in vivo protocol may come through targeting B . malayi L1 worms in the mosquito . If target gene expression can be reduced in this non-feeding stage , this would support soaking as the prime mechanism . The in squito suppression of Bm-cpl-1 reveals new phenotypes associated with molting , growth and development , and motility that shed light on the important biological functions of this gene family in larval stages of B . malayi . Nematode molting is a three-stage process characterized by a shedding or separation of the old cuticle from the epidermis ( apolysis ) , generation of a new cuticle , then the shedding of the old cuticle ( ecdysis ) . The use of specific cysteine protease inhibitors markedly inhibits the L3 to L4 molt in filarial worms implicating cysteine proteases in general in this process [34] , [44] , [45] . More explicitly , both apolysis and ecdysis are disrupted giving rise to L4 parasites constrained within an L3 cuticle , termed an ‘accordion’ phenotype [34] . Multiple members of the cathepsin L-like family appear to be involved in molting as the specific suppression of cpl-1 alone in Onchocerca vovlvulus reduced but did not abrogate the L3 to L4 molt [18] . We show that Bm-cpl-1 is also involved in similar processes in B . malayi as its suppression manifested an aberrant cuticular phenotype in L3 worms . Examination of worms suppressed at seven dpi revealed an apparent sloughing of the cuticle without the accordion phenotype . As the L3 to L4 molt occurs in the vertebrate host this phenotype is not a disruption of the L3 to L4 molt , but rather a dysfunction in L3 cuticle formation , maintenance or development . Alternatively , we could be observing a disruption of the L2 to L3 molt . Bm-cpl-1 expression is up-regulated in the L3 stage but the exact timing of this up-regulation as it relates to the transition from L2 to L3 stages is unclear . Guiliano et al . [34] report Bm-cpl-1 up-regulation at six dpi , a window consistent with the L2 to L3 transition . If Bm-cpl-1 performs the same function for the L2 to L3 molt as for the L3 to L4 molt , then the sloughed cuticle we observe upon dsRNA injection at seven dpi could be that of the L2 , with Bm-cpl-1 suppression preventing ecdysis . Further examination of cuticle ultrastructure in these suppressed worms at the electron microscope level could provide evidence to this effect . We observed a stunting of Bm-cpl-1 suppressed L3 growth compared to control worms , a phenotype previously unreported either after chemical inhibition of cysteine proteases or gene suppression in other parasite stages . Normally at the end of the L2 stage parasites are 750–795 µm long and increase in length to approximately 1350 µm at the L3 stage within four d [38] . Our control L3 , taken from mosquitoes injected with saline , had a mean length of 1347 µm corresponding closely with the published data . The mean length of Bm-cpl-1 suppressed L3 , however , was significantly shorter ( 700 µm ) . Suppression of this gene at the L2/L3 interface ( seven dpi ) arrests parasite growth and the L3 worms remain L2-sized within the mosquito . One explanation for this observation is that Bm-cpl-1 suppression at the L2/L3 interface is inhibiting the L2 to L3 molt , L2 cuticle ecdysis is not successful and therefore the worms are constrained within it , unable to increase their length . Alternatively , the stunting may not be due to aberrant molting but rather an inhibition of normal CPL-regulated development or cellular remodeling post-molt as is seen in other nematodes [46] . RNAi suppression of cpl-1 in C . elegans L3 by soaking produced significantly shorter and thinner adults [46] and the localization of cpl-1 to the hypodermis in C . elegans , O . volvulus and B . malayi is consistent with a developmental role in nematodes [18] , [34] , [46] . Further , germline suppression of cpl-1 in C . elegans by dsRNA injection generated an embryonic lethal phenotype but some embryos did progress to the L1 stage and those had incomplete gut development [46] . A repeatedly observed phenotype in our Bm-cpl-1 suppressed L3 was a compromised gut that appeared fenestrated and poorly developed . Finally , Bm-cpl-1 suppression reduced normal L3 motility by up to 69% , increased atypical postural phenotypes including caudal paralysis , kinked appearance and reduced normal convolution at the head and tail of B . malayi L3 as compared to control worms . These behaviors made it impossible for the treated L3 to progress through to the culmination of development in the mosquito host , i . e . , transfer to the definitive host . The dystaxic behaviors produced by the suppression of Bm-cpl-1 suggest this gene has some role , directly or indirectly , in the neuromuscular activity of B . malayi L3 in the mosquito . It is certainly true that cathepsins are required for normal neuromuscular behavior in other helminths; suppression of a cathepsin L-like gene in the flatworm Fasciola hepatica generated several aberrant motile phenotypes including paralysis [47] . This study is the first to use the host as a delivery mechanism for animal parasitic nematode RNAi . The model of using the host as a delivery mechanism for RNAi has been established but has been restricted to plant pathology where the concept has an applied use with transgenic plants helping to control nematode infestation by RNAi mechanisms in planta [48]–[50] . An alluring corollary is that by generating transgenic mosquitoes capable of suppressing key nematode genes in vivo we may be able to abolish parasite transmission . We have already demonstrated here that Bm-cpl-1 suppression in vivo prevents parasites migrating to the mosquito head and proboscis thus eliminating transmission potential . Transformation of a mosquito with an inverted-repeat ( IR ) transgene derived from Bm-cpl-1 may result in endogenous transcription of a hairpin dsRNA , a trigger that conceptually would induce RNAi in vivo as described here and produce a mosquito incapable of transmitting lymphatic filariasis-causing worms . Methods to introduce transgenes into mosquito germlines are well established [51]–[55] and proof of this principle has already been demonstrated for a mosquito-borne virus; transgenic lines of Dengue virus-resistant mosquitoes were generated using a Dengue virus IR transgene driven by the carboxypeptidase A promoter , reducing virus transmission by an RNAi mechanism [56] . The viability of this approach is enhanced not only by the ability to transform important vector species but also by the identification of tissue-specific promoters to drive transgene expression in favorable tissues , for example , act88F [57] is a fly-specific promoter that drives gene expression in the flight musculature – the precise site of parasite development . Another positive impact this protocol may have on lymphatic filariasis control is as a means of better understanding the biology of current putative drug targets and generating new data that may validate proposed novel drug targets . This protocol introduces the ability to investigate mosquito-borne parasite life stages , allowing the critical examination of gene function in worms growing and developing in an optimum environment . This makes it possible to assay genes that encode known or proposed drug targets in a parasite within its native intermediate host , contextualizing the null phenotypes in vivo and accurately determining the consequences of target gene suppression producing a more valuable target validation . As an illustration , nematode cathepsins have been proposed as attractive novel drug targets [58] and we have further validated these drug targets in vivo , revealing new phenotypes , defining new biological roles and showing that B . malayi sans Bm-cpl-1 are incapable of completing their life cycle . These data enhance the appeal of cathepsins as novel anthelmintic drug targets . Beyond cathepsins , this technique will have most utility in the investigation of known and potential antifilarial drug target genes expressed in both the mosquito-borne life stages and those life stages that are vulnerable to chemotherapeutic intervention . In summary , we have developed an innovative RNAi protocol using B . malayi that differs conceptually from present RNAi protocols in that parasite gene expression is suppressed within the mosquito intermediate host . Using this protocol we suppressed a B . malayi gene in vivo , eliciting aberrant developmental and motility phenotypes in the parasite – phenotypes that eliminate transmission potential . In contrast to present RNAi methods , we have found the protocol to be reliable and effective , providing a major advancement in our capability to better understand filarial nematode gene function to the benefit of human health .
Aedes aegypti ( Liverpool strain ) , previously selected for susceptibility to filarial parasites [59] , were maintained in a contained environment at a constant temperature of 25°C , 80% relative humidity and a 14 h light to 10 h dark photoperiod . The mosquitoes were fed a diet of 0 . 3 M sucrose . Throughout the study mosquitoes to be injected were anaesthetized on ice and immobilized on a vacuum saddle before being microinjected intrathoracically at the base of the head using a pulled borosilicate glass pipette attached to a manual syringe for injection by volume displacement . A maximum volume of 0 . 5 µL can be injected using this approach with a high mosquito survival rate ( >95% ) . B . malayi microfilaria ( mf ) infected cat blood was obtained from the University of Georgia NIH/NIAID Filariasis Research Reagent Resource Center . To establish a consistent and repeatable parasitemia , mf were first purified using a filtration protocol [60] . Blood containing the parasites was diluted with phosphate buffered saline ( 1∶5 ratio , blood:PBS ) then syringe filtered through a 0 . 45 µm Millipore filter . Captured mf were washed three to five times with PBS then a further three to five times with Aedes physiologic saline [61] before centrifugation at 6 , 800× g for five min . Supernatant was removed and the pelleted mf resuspended in fresh Aedes saline to a concentration of 40 worms per µL . To inoculate mosquitoes , 20 mf were injected as described . Microdissection of the mosquitoes throughout a 14 dpi period confirmed this method established a controlled infection that progressed in a predictable and consistent manner . We also tried a blood feeding approach to establish infection but this produced an inconsistent worm burden that is too variable to reliably assess subsequent gene suppression experiments . Short interfering RNAs ( siRNA ) targeting a B . malayi cathepsin L-like gene ( Bm-cpl-1 AF331035 [34] ) were generated commercially ( Qiagen , CA ) and modified with a 3′-Cy3 fluorophore on the sense strand . The location of each siRNA was optimized using a proprietary algorithm and the sequence of each siRNA is as follows: BmCL1-1 , AAGGCTTAGTTTCTTATACAA; BmCL1-2 , CCGAATGGAAAGATTATGTAA; BmCL1-3 , CAGAAGTGCATTGAAGGAATA; and BmCL1-4 , CCGGTATTTACTCCAGTAATA . Equimolar amounts of each siRNA were combined and this mix was used for injection and gene suppression experiments . dsRNA duplexes were generated in-house using a T7 transcription-based approach . A 410 base pair transcription template was polymerase chain reaction ( PCR ) amplified from a B . malayi L3 stage cDNA library ( kindly provided by Dr . S . Williams , Smith College , MA ) using gene specific oligonucleotides designed to incorporate a T7 promoter sequence ( TAATACGACTCACTATAGGGTACT ) at both the 5′ and 3′ ends of the amplicon . For the Bm-cpl-1 template , oligonucleotide sequence was: L1T7dsRNAF 5′ TAATACGACTCACTATAGGGTACTACGGTTACCAAATTC 3′ and L1T7dsRNAR 5′ TAATACGACTCACTATAGGGTACTCGACAACAACAGGTC 3′ . The location of this transcription template was carefully chosen so as to exclude the pro region of Bm-cpl-1 , a domain with high sequence homology to other cathepsin L family genes , and consequently increase the specificity of this dsRNA duplex . Transcription templates were gel purified and dsRNA duplexes synthesized using the MEGAscript RNAi Kit ( Ambion , TX ) according to manufacturer's protocols . dsRNA species were quantified with a NanoVue spectrophotometer ( GE Healthcare , NJ ) prior to use . The timing of siRNA or dsRNA injection into B . malayi-infected mosquitoes coincided with the presence of the parasite stage of interest: to target second larval stage ( L2 ) parasites siRNA or dsRNA were injected five to eight dpi; to target third larval stage ( L3 ) parasites siRNA or dsRNA were injected nine to 12 dpi ( and for the lifespan of mosquito ) [35] . The mosquitoes were processed to confirm suppression of the target gene , as described below , 48 h post-injection of siRNA or dsRNA . Brugia infected , RNA-treated and control mosquitoes were cold-anesthetized on ice . Total RNA was extracted from individual mosquitoes using RNAqueous Kit ( Ambion , TX ) before DNase treatment using the TURBO DNA-free Kit ( Ambion , TX ) in thin-walled PCR tubes . The RNA was stabilized with RNase Out Inhibitor ( Invitrogen , CA ) and stored in RNase-free microcentrifuge tubes at 4°C . This RNA was used as a template for a relative semi-quantitative multiplex RT-PCR using the SuperScript III One-Step RT-PCR System with Platinum Taq DNA Polymerase ( Invitrogen , CA ) . The principle of this reaction is to amplify a target gene of interest and compare its intensity with a multiplexed and normalized internal standard during the linear phase of product amplification . A putative neuropeptide encoding gene , Bm-flp-14 ( Accession number AI508026 ) served this role . This gene was chosen as we had previously determined its stable transcript production during the B . malayi L3 stage by PCR ( C . Song , unpublished ) . The oligonucleotide primers used to amplify Bm-cpl-1 were: CPL-1 F 5′ ACAGGGCAATATGACGAGAC 3′ and CPL-1 R 5′ ATCGAAGCAACGTGGCACAT 3′ . These primer locations flank the region of Bm-cpl-1 homologous to the dsRNA construct . The oligonucleotide primers used to amplify the Bm-flp-14 internal standard were: FLP-14 F 5′ CTCGTCCACTCTTATCACTG 3′ and FLP-14 R 5′ ACCGCAATGATATACAACATATA 3′ . The profile for this PCR was: cDNA synthesis at 50°C for 30 minutes; an initial denaturation phase of 94°C for 2 min; 38 cycles of 94°C for 30 s , 60°C for 30 s , 68°C for 1 min and a final extension phase of 68°C for 5 min . Reactions were visualized on a 1 . 2% agarose gel containing ethidium bromide . Total RNA was extracted from individual mosquitoes and DNase-treated as described above for three replicated RNAi experiments and before addition of RNase Out Inhibitor and storage , each RNA sample was quantified spectrophometrically per a previous report [62] . This RNA served as a template in our RT-qPCR assays using the qScript One-Step Fast RT-PCR Kit with ROX ( Quanta BioSciences , MD ) . After confirmation of Bm-cpl-1 suppression , multiple assays were performed to describe worm phenotypes . Each phenotypic assay was performed 14 dpi and at either four or seven d post-injection . Mosquitoes were cold-anesthetized then the wings and legs removed and discarded using a dissecting microscope . The head , thorax and abdomen were partitioned and further dissected to release the parasites . The following characteristics of dissected parasites were observed: ( 1 ) Parasite location . In order to be successfully transmitted , these parasites have to actively migrate to the head of the mosquito and vigorously writhe free of the proboscis . Parasite migration through the mosquito was recorded and measured according to escape point from the mosquito body ( abdomen , thorax or head ) . ( 2 ) Worm motility . A scoring schema of: one ( immobile ) , two ( compromised motility , immobile for stretches of time ) , three ( sluggish , partial movement ) , four ( in motion , some straight segments ) , or five ( all parts of the worm in constant motion ) was used to quantify parasite movement in a blind fashion by an independent evaluator . Additional observations of aberrant motility included knotting at one or both ends , paralysis of caudal region and presence of a distinct angular kink were also recorded . ( 3 ) Parasite growth and development . Digital images of RNAi and control worms were captured so that length and diameter could be calculated using NIS Elements D 2 . 30 software ( Nikon , NY ) . ( 4 ) Parasite viability . The number of parasites that survived to the infectious stage was recorded so that infection prevalence and mean intensity could be calculated . ( 5 ) Mosquito viability . We documented the number of mosquitoes that survived through the development of parasites to the infectious stage because these parasites inflict significant pathology and decrease mosquito survival . Nikon Eclipse 50i fluorescence microscope under UV light ( EXFO , ON ) equipped with a Hy-Q FITC filter set ( Chroma , VT ) . Images were captured using a Digital Sight DS-2Mv camera and NIS Elements D 2 . 3 software ( Nikon , NY ) . t-tests were used to analyze the effect of RNAi treatment on gene expression in the RT-qPCR experiments and parasite size , and ANOVA to analyze the effect of RNAi treatment on worm motility based on our one through five blind-scoring schema . Chi square tests were used to analyze the effect of RNAi treatment on all other worm and mosquito behaviors assayed . In all tests , P values ≤0 . 05 were considered statistically significant . | Lymphatic filariasis is a debilitating tropical disease caused by parasitic nematodes such as Brugia malayi that are transmitted to humans through the bite of infected mosquitoes . Controlling lymphatic filariasis , and other parasitic nematode diseases , is made difficult by a limited repertoire of sub-optimal drugs and substantial experimental roadblocks to new drug development . We have developed a novel and highly effective in vivo RNA interference ( RNAi ) methodology to better understand gene function in the parasitic nematode , B . malayi . RNAi triggers are delivered to developing parasites in the mosquito intermediate host – an improvement on existing RNAi protocols , in which the parasite is removed from the host and maintained in liquid culture . Using this protocol we have significantly suppressed a cathepsin gene in B . malayi , which resulted in dramatic change to parasite movement , growth and development in the mosquito to the extent that worms cannot be transmitted to a new host . These data underscore the power of this protocol to 1 ) critically assess gene function in parasitic nematode biology , 2 ) to reveal the mechanism of action for existing drugs , and 3 ) for discovery of novel drug targets for drug development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"genetics",
"and",
"genomics/gene",
"function",
"biochemistry/drug",
"discovery"
] | 2010 | Development of an In Vivo RNAi Protocol to Investigate Gene Function in the Filarial Nematode, Brugia malayi |
The laboratory diagnosis of Chagas disease is challenging because the usefulness of different diagnostic tests will depend on the stage of the disease . Serology is the preferred method for patients in the chronic phase , whereas PCR can be successfully used to diagnose acute and congenital cases . Here we present data using a combination of three TaqMan PCR assays to detect T . cruzi DNA in clinical specimens . Included in the analysis were DNA extracted from 320 EDTA blood specimens , 18 heart tissue specimens , 6 umbilical cord blood specimens , 2 skin tissue specimens and 3 CSF specimens . For the blood specimens both whole blood and buffy coat fraction were analyzed . The specimens were from patients living in the USA , with suspected exposure to T . cruzi through organ transplantation , contact with triatomine bugs or laboratory accidents , and from immunosuppressed patients with suspected Chagas disease reactivation . Real-time PCR was successfully used to diagnose acute and Chagas disease reactivation in 20 patients , including one case of organ-transmitted infection and one congenital case . Analysis of buffy coat fractions of EDTA blood led to faster diagnosis in six of these patients compared to whole blood analysis . The three real-time PCR assays produced identical results for 94% of the specimens . The major reason for discrepant results was variable sensitivity among the assays , but two of the real-time PCR assays also produced four false positive results . These data strongly indicate that at least two PCR assays with different performances should be combined to increase the accuracy . This evaluation also highlights the benefit of extracting DNA from the blood specimen's buffy coat to increase the sensitivity of PCR analysis .
Chagas disease is a vector-borne infectious disease caused by the parasite Trypanosoma cruzi . It is endemic in several countries of Central and South America . In endemic areas the disease is spread by certain species of triatomine bugs that excrete the parasites in their feces while feeding on human hosts . Humans get infected when feces from infected triatomines contaminates wounds , allowing the parasite to enter the bloodstream . Other routes of infection include congenital transmission , blood transfusion , organ transplantation , accidental inoculation of the parasite during laboratory research and by consuming food and juice contaminated with the parasite . As efforts to control vector-borne and blood transmission are successful , congenital and oral transmission paths are becoming increasingly important [1] . After a short acute phase when the parasite can be found circulating in the blood , the disease enters the chronic phase when the amastigote stage develops and multiplies in organ tissues , primarily in the heart . The chronic phase is characterized by two forms; patients first develop the indeterminate form of chronic infection which can last for decades and the patients are typically asymptomatic during this time . An estimated 30–40% of patients may develop clinical disease , with manifestations such as cardiomyopathy or digestive megasyndromes [1] . Chronically infected patients that become immunosuppressed may experience a reactivation of the disease , a condition characterized by increasing parasitemia and atypical presentations such as epidermic lesions and compromise of the central nervous system [2] . The options for laboratory diagnosis of Chagas disease depend on the disease phase . Serology is the method of choice to diagnose chronic infections . Acute infections can be diagnosed by detecting motile organisms in fresh blood preparations , by culture or by detection of parasite DNA by PCR [3] . The latter methods are also recommended to detect increasing parasitemia in cases of reactivation following immunosuppression [4] . In cases of acute infections or reactivation of chronic disease it is important to use sensitive diagnostic methods since early detection and treatment results in a more favorable outcome . PCR-based methods are generally considered to be more sensitive than microscopy and have lately been increasingly used to diagnose Chagas disease [4] . However , the use of PCR is also challenging as there is no “gold standard” method for the diagnosis of Chagas disease [5] , [6] and the diagnostic performance can vary widely depending on the type of PCR assay . The most widely used PCR assays used for diagnostic purposes target either the kinetoplast genome ( kDNA ) , also called the minicircle , or a nuclear mini-satellite region designated TCZ [4] , [6] , [7] , [8] , [9] , [10] , [11] , [12] , [13] , [14] , [15] , [16] . Both of these targets are present in multiple copies in the parasite genome , which increases the sensitivity of detection [15] , [17] . However , assays that target these regions have been reported to cross-amplify non-T . cruzi DNA [10] , [16] , [18] , [19] , [20] . Assays that amplify other genes may show better specificity but they are generally less sensitive [4] , [9] , [16] , [21] . One important use of PCR as a diagnostic tool is to provide a sensitive method to detect reactivation in chronically infected patients with immunosuppression . Patients with chronic Chagas heart disease often require a heart transplant [22] . Current recommendations state that these patients should be monitored at regular intervals after the transplant for signs of increasing parasitemia [3] . Another category of patients for whom PCR testing is beneficial is patients who receive organs from chronically infected donors . Since only a fraction of organ recipients will develop an acute T . cruzi infection , preventive drug treatment is not recommended . In such cases the use of PCR can allow for early detection of those cases where transmission has occurred . Recently , an international collaborative study focusing on standardization and validation of PCR for diagnostic detection of T . cruzi DNA was conducted [13] . The study relied on the use of DNA specimens from genetically distinct cultured T . cruzi strains plus blood specimens from chronically infected patients . The specimens were coded at a coordinating laboratory and shipped to 26 participating laboratories that performed PCR testing according to their own standard operating procedures . Results were then sent back to the coordinating laboratory and performance characteristics were calculated for each PCR assay . The study found a high degree of variability in accuracy and performance among the included PCR tests and identified and further evaluated two DNA extraction methods and four PCR assays that performed better than the others . Two of the best-performing assays were real-time PCR assays . To continue these efforts we here present results from a diagnostic testing algorithm involving three of the real-time PCR assays included in the international validation study mentioned above . Real-time PCR has several advantages over conventional PCR , e . g . shorter turnaround times and less risk of amplicon carry-over contamination [23] , both of which can be advantageous in diagnostic laboratories . One of the real-time PCR assays included in this study was ranked among the four best-performing assays in the international validation study; a real-time PCR assay targeting the mini-satellite TCZ region . The second real-time PCR assay was selected because it was the best-performing real-time PCR assay targeting the kDNA included in the international validation study . The third real-time PCR assay was included in this study because it targets the small subunit ribosomal RNA ( 18 S rRNA ) gene , which is generally suitable for diagnostic assays because it is highly conserved . In contrast to the international validation study we mainly used specimens from patients with suspected acute or reactivating Chagas disease since PCR testing is more relevant for early diagnosis or monitoring in this patient group than in chronic patients , whose diagnosis relies on serological methods . The majority of the specimens tested were EDTA blood samples; we performed real-time PCR on DNA extracted from buffy coat preparations in addition to whole blood to determine the effect of buffy coat concentration on the sensitivity of the PCR analysis .
All the specimens used in this study were submitted to CDC for confirmatory diagnosis of Chagas disease during years 2008–2010 from state public health laboratories , hospitals and private clinics in the United States . The tests were performed on 349 laboratory specimens from 119 patients , who lived in the United States at the time of specimen collection . A breakdown of the specimen types and the conditions that prompted the diagnostic requests are presented in Table 1 . Samples analyzed in this study were anonymized by removing identifiers after diagnostic results were reported , in accordance with the CDC IRB , protocol number 3580 , entitled “Use of Human Specimens for Laboratory Methods Research” . All of the patients included in this study were evaluated for serology status using the Chagatest recombinante v . 3 . 0 ( Wiener Laboratorios , Rosario , Argentina ) and a CDC in-house IIF test . DNA extraction was performed from all specimens within 24 hours of arrival at the laboratory . DNA was extracted from whole blood specimens using the QIAamp blood mini DNA kit ( QIAGEN , Valencia , Calif . ) . The volume of whole blood used was 0 . 2 ml and if the remaining volume exceeded 1 ml , the buffy coat fraction was separated as follows: up to 2 ml of whole blood was centrifuged at 2 , 500×g for 10 minutes . The plasma was removed and the buffy coat layer plus some of the erythrocyte pellet were transferred to a clean tube . DNA was then extracted from that material in parallel with the whole blood aliquot using the same method mentioned above . Three EDTA blood specimens had enough volume left after initial DNA extraction to allow for one or more additional buffy coat preparations . Two-ml aliquots of these specimens were stored at 4°C for one , two or four weeks and then processed as described above . DNA from tissue specimens was extracted with the DNeasy blood and tissue DNA kit ( QIAGEN ) . For cerebrospinal fluids ( CSF ) , approximately half of the total volume received ( 0 . 5–1 ml ) was centrifuged for 5 minutes at 6000×g . Most of the supernatant was carefully removed until 0 . 2 ml remained and DNA was extracted from this remaining volume ( plus any pellet ) with the DNeasy blood and tissue DNA kit ( QIAGEN ) . All the DNA extraction procedures were performed following manufacturer's instructions for the different types of samples . Previous experiences with these methods in our laboratory had ensured that they efficiently removed potential PCR inhibitors from the specimen types included in this study ( data not shown ) . One negative extraction control was included in each batch of DNA extractions to monitor for potential cross- contamination among samples and contamination of kit reagents . All three real-time PCR assays were included in the international validation study [13] . Table 2 summarizes validation data for the PCR assays as presented in that study , plus specificity data for two other Trypanosoma spp . obtained in our laboratory . The real-time PCR assays were performed and analyzed in an Mx3000P QPCR system ( Agilent Technologies , Calif . ) . Each DNA sample was added to the PCR mix in two different concentrations ( corresponding to 5 µl and 1 µl of undiluted DNA ) . All PCR runs included two or more negative amplification controls ( adding water instead of template DNA ) plus two positive amplification controls ( DNA extracted from a culture of the Y strain in two different dilutions ) . The risk of false positive results due to contamination was minimized by the following procedures: using separate rooms for DNA extraction , pre-and post-amplification processes; having a uni-directional workflow; and using enzymatic removal of contaminating amplicons before real-time PCR amplification . TCZ TaqMan real-time PCR ( designated as method LbF1 in the international validation study [13] ) : This TaqMan assay was performed as described in Piron 2007 [11] , except that the Platinum qPCR supermix was used instead of the Universal mastermix from Applied Biosystems . kDNA TaqMan real-time PCR ( designated as method LbG/3 in the international validation study [13] ) : The reaction mix consisted of 1× Platinum qPCR supermix , 0 . 4 µM of each PCR primer 32F , 5′-TTT GGG AGG GGC GTT CA-3′ , and 148R , 5′-ATA TTA CAC CAA CCC CAA TCG AA-3′ , plus 0 . 1 µM of the LNA TaqMan probe 71P , 5′-CA TCTC AC CCG TACA TT-3′ , where the LNA nucleotides [24] are underlined . Total reaction volume was 20 µl . Thermocycling structure was as follows: 2 minute incubation at 50°C to activate UDG degradation , 2 minute incubation at 95°C to activate the hot-start DNA polymerase , and 40 cycles of 95°C for 15 seconds and 58°C for 60 seconds . 18 S rRNA TaqMan real-time PCR ( designated as method LbS/4 in the international validation study [13] ) : The reaction mix consisted of 1× Platinum qPCR supermix , 0 . 2 µM of each PCR primer TcF1042 , 3′-GCA CTC GTC GCC TTT GTG-3′ , and TcR1144 , 5′-AGT TGA GGG AAG GCA TGA CA-3′ plus 0 . 05 µM of the TaqMan probe TCP1104 , 5′-AA GAC CGA AGT CTG CCA ACA ACA C-3′ . Total reaction volume was 20 µl . Thermocycling structure was as follows: 2 minute incubation at 50°C to activate UDG degradation , 2 minute incubation at 95°C to activate the hot-start DNA polymerase , and 40 cycles of 95°C for 15 seconds and 60°C for 60 seconds .
Specimens from 25 patients with chronic Chagas disease ( as determined by positive serology ) were received for evaluation of reactivation disease during the study period . Eighteen of the patients had received a heart transplant , five were HIV infected and two had undergone a bone marrow transplant . Sixteen patients had one or more PCR-positive samples , including sporadic PCR-positive results in seven patients who had received a heart transplant prior to 2008 . Table 3 lists a selection of the samples analyzed from the remaining nine patients with at least one PCR-positive test result during this study . Seven patients were tested for reactivation following transplants; four of these were monitored on a regular basis by PCR . Two of the five HIV-positive patients were diagnosed with re-activated Chagas disease ( patients 8 and 9 ) ; one of them had cerebral Chagas , confirmed by the presence of T . cruzi DNA in CSF . Real-time PCR was used to test blood specimens from 14 previously non-infected transplant patients who received organs from a donor with suspected or confirmed chronic Chagas disease . It is recommended to closely monitor these patients with PCR or other sensitive technique in order to detect potential transmission as soon as possible . Three organ recipients had one or more PCR positive results ( see Table 4 ) . However , only patient 12 , a heart recipient , was actually infected with T . cruzi . The PCR positive results for the other two patients ( patients 10 and 11 ) were reported as equivocal and were most likely false positive results because of the following circumstances . Patient 10 received a kidney from a donor with borderline positive serology results with the Ortho T . cruzi ELISA test ( Ortho-Clinical Diagnostics , Raritan , New Jersey ) . Since this could have been interpreted as indicative of infection in the donor , regular PCR testing was started on patient 10 . However , subsequent serology testing of the donor associated with this case could not confirm the preliminary results; i . e . , the T . cruzi RIPA was indeterminate and both the Wiener and the IIF test were negative on repeated serum samples . It was therefore concluded that the donor was not infected with T . cruzi and additional PCR follow up of patient 10 was unnecessary . However , before the final donor serology status had been determined , weak positive signals in the kDNA and TCZ TaqMan assays were verified in blood samples from patient 10 . Unexpectedly , each subsequent specimen obtained from this patient showed a signal that was weaker than the signal obtained for the previous sample; i . e . the opposite of what was expected from an acute T . cruzi infection in an immunocompromised patient . At six weeks post-transplant patient 10 was no longer positive in any of the real-time PCR assays . Patient 11 received a kidney from a donor that was confirmed to be serologically positive for T . cruzi . The blood sample from patient 11 collected on the 3rd week post-transplant tested weakly positive in the kDNA and TCZ TaqMan assays , with only the whole blood aliquot being positive and not the buffy coat fraction . The blood collected a week later was PCR negative in both whole blood and buffy coat . Neither patient 10 nor 11 had any clinical signs of T . cruzi infection . Their blood smears were constantly negative for parasites and they did not receive anti-trypanosomal drugs . We received 48 blood samples from 13 healthy patients who had been bitten or in close contact with triatomine bugs plus 9 laboratory workers that had been accidentally exposed to T . cruzi via needle stick accidents or animal bites during research activities . None of these were PCR positive . We tested 22 specimens from 20 children ( aged newborn to 8 years ) with sero-positive mothers for possible congenital transmission and detected T . cruzi DNA in the blood of a 19-days-old infant ( patient 14 in Table 5 ) . Twenty-eight specimens were received from 18 adult patients with symptoms of acute T . cruzi infection ( fever and malaise after traveling to endemic region and/or having close contact with triatomine bug; three had a swollen eye that could be chagoma ) . Only one of these patients tested positive for T . cruzi by PCR and was treated for acute Chagas disease ( patient 13 in Table 5 ) . Follow-up specimens from this patient again tested positive in PCR after completed drug treatment but unfortunately the patient was lost to follow-up . Thirty-five of the PCR-positive blood specimens ( from 16 patients ) had enough volume to allow for DNA extraction from both whole blood aliquots and buffy coat fraction . Of these , 26 specimens ( from 10 patients ) had PCR-detectable levels of T . cruzi DNA in both whole blood and buffy coat , with a relatively higher concentration in the buffy coat based on the quantitative output ( the Cq value ) from the real-time PCR assays . The remaining 9 specimens ( from 6 patients ) were positive only in the buffy coat fraction . Thus , 26% of the PCR-positive specimens would have been reported as being negative for T . cruzi if no buffy coat analysis had been performed . For three patients the analysis of buffy coat was crucial: Chagas disease reactivation in two patients was detected two weeks earlier by testing the buffy coat sample as compared to whole blood ( patients 3 and 6 in Table 3 ) and the patient who acquired Chagas disease through transplantation ( patient 12 in Table 3 ) was identified as positive one week earlier by testing buffy coat as compared to whole blood . Three of the PCR-positive blood samples had enough volume to allow for analysis of more than one buffy coat preparation . Aliquots of these three samples were stored at 4°C for up to four weeks and then processed as described . Figure 1 depicts the quantitative real-time PCR results obtained from these samples over time . The results suggested that storage of EDTA blood for a limited time had minor effect on the quality of T . cruzi DNA obtained from buffy coat preparations , at least for the kDNA and TCZ genetic regions . Although these are only preliminary data that need confirmation with a larger set of samples , it removes some of the uncertainty whether to accept EDTA-blood samples that for various reasons are delayed in transport to the diagnostic laboratory .
The laboratory diagnosis of Chagas disease relies mainly on serology , microscopic identification of trypomastigotes in blood or buffy coat , hemoculture and PCR [3] . Several PCR assays with variable diagnostic sensitivity and specificity have been developed and used as diagnostic tests [4] , [7] , [8] , [9] , [10] , [11] , [15] , [16] , [21] . A complicating factor for PCR assays is the high genetic variability of T . cruzi strains; there are currently six genotype groups or discrete typing units ( DTU ) described that differ significantly in genetic content and gene copy numbers [25] , [26] . Since some DTUs are more common than others in various endemic regions , the same PCR assay can perform differently depending on the geographic origin of the specimen [16] , [25] , [27] , [28] , [29] . One way to circumvent these accuracy problems is to combine two or more PCR assays that target different genes . The reference diagnostic laboratory at CDC employs a multi-target PCR testing algorithm consisting of three real-time PCR assays that are performed in parallel on all specimens . The three assays target different genomic regions in T . cruzi and have therefore variable sensitivity and specificity . The rationale for including all three assays in the testing algorithm is to ensure the highest accuracy possible by combining assays that complement each other . The kDNA TaqMan assay seems to be the most sensitive assay but it can amplify non-T . cruzi DNA , e . g . T . rangeli , and thus lead to false positives . The TCZ TaqMan assay has better specificity but as shown in this study can produce false positive PCR results as well . The kDNA and TCZ TaqMan assays are both much more sensitive than the 18 S rRNA TaqMan but the main advantage of including the 18 S rRNA assay in the testing algorithm is that it seems to be 100% specific . According to the CDC protocol , if a specimen tests positive in all three real-time PCR assays it will be reported as positive for T . cruzi , but any specimen that is only positive in the kDNA and/or TCZ TaqMan assays and negative in the 18 S rRNA TaqMan assay require additional confirmation by other tests or clinical data in order to be reported as positive for T . cruzi . If confirmatory data is absent or does not support a diagnosis of T . cruzi infection , the PCR results are reported as equivocal and a new specimen is requested to repeat the molecular analysis . Diagnostic sensitivity can be enhanced by maximizing the amount of target DNA in the aliquot used for DNA extraction . For multi-copy PCR targets this can be obtained by mixing blood specimens with guanidine HCl-EDTA solution that lyses the parasites and releases their genetic content , thus making it possible to detect as little as one parasite in a large volume of blood [30] , [31] . It has also been reported that sensitivity could be enhanced if blood clot was used as starting material [32] . An alternative method is to concentrate the parasites in the buffy coat fraction [33] prior to DNA extraction; this has been reported to increase the sensitivity compared to analysis of frozen EDTA-blood and guanidine HCl-EDTA treated blood [32] , [34] . During this study , we compared the PCR results obtained from buffy coat concentration with results from fresh EDTA- blood and found that analysis of buffy coat allowed earlier detection of increasing levels of circulating parasite genome in three cases: two reactivation cases and one organ-transmitted acute infection . Thus , appropriate drug treatment for these patients could be initiated 1–2 weeks sooner . Analyzing both the buffy coat fraction and a whole blood aliquot in parallel can also give helpful information to ensure test validity and to troubleshoot suspicious false positive PCR results . DNA extraction from the buffy coat fraction of a blood sample containing T . cruzi trypomastigotes should produce more T . cruzi DNA than the corresponding volume of whole blood . If that is not the case , there could be a problem with the quality of the blood specimen , the DNA extraction process or the PCR accuracy . One of the false positive PCR results obtained in this study was immediately flagged as suspicious because only the whole blood fraction was positive while buffy coat was negative . Nevertheless , more data must be accumulated during a longer period of time for a more robust assessment about the advantages of analyzing both whole blood and buffy coat . In conclusion , we propose that in reference laboratories with the adequate infrastructure , the use of two or more real-time PCR tests with different performance characteristics combined with the analysis of buffy coat and whole blood can strengthen the use of PCR for accurate diagnosis of Chagas disease . | Chagas disease is endemic in several Latin American countries and affects approximately 8 to 11 million people . The protozoan parasite , Trypanosoma cruzi , is the agent of Chagas disease , a zoonotic disease that can be transmitted to humans by blood-sucking triatomine bugs . Other routes of infection include congenital transmission , blood transfusion , organ transplantation , accidental inoculation of the parasite during laboratory research and by consuming food and juice contaminated with the parasite . This study focused on the evaluation of three quantitative PCR ( QPCR ) assays for the diagnosis of Chagas disease . The evaluation was based on the analysis of 349 specimens submitted for confirmatory diagnosis of Chagas disease to the Centers for Disease Control and Prevention from 2008 to 2010 . By using such assays we were able to diagnose acute and Chagas disease reactivation in 20 patients , including one case of organ-transmitted infection and one congenital case . The paper also highlights the benefit of extracting DNA from the blood specimen's buffy coat to increase the sensitivity of diagnostic PCR analysis . The results obtained in this study strongly indicate that at least two QPCR assays with different performance characteristics should be combined to increase diagnostic accuracy . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"clinical",
"laboratory",
"sciences",
"diagnostic",
"medicine",
"chagas",
"disease",
"neglected",
"tropical",
"diseases",
"biology",
"microbiology",
"parasitic",
"diseases... | 2012 | Sensitive and Specific Detection of Trypanosoma cruzi DNA in Clinical Specimens Using a Multi-Target Real-Time PCR Approach |
Dengue can cause plasma leakage that may lead to dengue shock syndrome ( DSS ) . In approximately 30% of DSS cases , recurrent episodes of shock occur . These patients have a higher risk of fluid overload , respiratory distress and poor outcomes . We investigated the association of echocardiographically-derived cardiac function and intravascular volume parameters plus lactate levels , with the outcomes of recurrent shock and respiratory distress in severe dengue . We performed a prospective observational study in Paediatric and adult ICU , at the Hospital for Tropical Diseases ( HTD ) , Ho Chi Minh City , Vietnam . Patients with dengue were enrolled within 12 hours of admission to paediatric or adult ICU . A haemodynamic assessment and portable echocardiograms were carried out daily for 5 days from enrolment and all interventions recorded . 102 patients were enrolled; 22 patients did not develop DSS , 48 had a single episode of shock and 32 had recurrent shock . Patients with recurrent shock had a higher enrolment pulse than those with 1 episode or no shock ( median: 114 vs . 100 vs . 100 b/min , P = 0 . 002 ) , significantly lower Stroke Volume Index ( SVI ) , ( median: 21 . 6 vs . 22 . 8 vs . 26 . 8mls/m2 , P<0 . 001 ) and higher lactate levels ( 4 . 2 vs . 2 . 9 vs . 2 . 2 mmol/l , P = 0 . 001 ) . Higher SVI and worse left ventricular function ( higher Left Myocardial Performance Index ) on study days 3–5 was associated with the secondary endpoint of respiratory distress . There was an association between the total IV fluid administered during the ICU admission and respiratory distress ( OR: 1 . 03 , 95% CI 1 . 01–1 . 06 , P = 0 . 001 ) . Admission lactate levels predicted patients who subsequently developed recurrent shock ( P = 0 . 004 ) , and correlated positively with the total IV fluid volume received ( rho: 0 . 323 , P = 0 . 001 ) and also with admission ALT ( rho: 0 . 764 , P<0 . 001 ) and AST ( rho: 0 . 773 , P<0 . 001 ) . Echo-derived intravascular volume assessment and venous lactate levels can help identify dengue patients at high risk of recurrent shock and respiratory distress in ICU . These findings may serve to , not only assist in the management of DSS patients , but also these haemodynamic endpoints could be used in future dengue fluid intervention trials .
Dengue is a flaviviral infection that causes substantial morbidity in endemic areas , with 96 million clinically apparent cases each year [1] . Although the majority of infections result in a self-limiting febrile illness , 1–5% of cases can experience more severe manifestations , in the form of organ impairment , coagulopathy and plasma leakage which may lead to intravascular volume depletion and dengue shock syndrome ( DSS ) . The plasma leakage usually resolves around defervescence , and the extravasated fluid then gets reabsorbed , when fluid overload in the form of massive pleural effusions or pulmonary oedema can occur . The resulting respiratory compromise has been associated with an increased risk of death in adult [2] and paediatric severe dengue [3] . The current treatment of DSS is supportive , with careful intravenous fluid replacement . The majority of patients recover after a single crystalloid bolus and in experienced centres the mortality rate is less than 1% [4] . However , in approximately 30% of DSS cases , recurrent episodes of shock occur , which require more intensive treatment with larger volumes of intravenous fluids including colloid boluses; these patients have a higher risk of fluid overload , respiratory distress and poor outcomes [5] . Identifying such individuals early and investigating other potential contributing factors for recurrent shock is needed . Although the main mechanism of DSS is hypovolaemia , it is becoming increasingly recognized that myocardial impairment may play a role in the haemodynamic instability and potentially could contribute to recurrent shock [6 , 7] . Cardiac manifestations of dengue are diverse and include functional myocardial impairment , arrhythmias and myocarditis , however the clinical significance of these in DSS has not been well studied [8] . We have shown previously that systemic microvascular dysfunction occurs in more severe dengue infections , but the effect on end-organ perfusion has not been evaluated [9] . Lactate levels can be representative of tissue perfusion and elevated levels are associated with organ failure and predict mortality in septic shock [10 , 11] . Serum lactate levels and their prognostic significance in dengue shock syndrome have only been evaluated in small studies and in adults [12–14] . In this study we investigated the association of echocardiographically-derived cardiac function and intravascular volume parameters as well as lactate levels with the clinical outcomes of recurrent shock and respiratory distress in adults and children admitted to ICU with dengue . We hypothesised that DSS patients with cardiac dysfunction and elevated lactate levels would be more likely to develop recurrent shock and also to experience iatrogenic fluid overload and respiratory distress .
Ethical approvals were obtained from the Oxford Tropical Research Ethics Committee and the Ethics Review Committee at HTD , and written informed consent was obtained from all participants or the parent/guardian of children . A full blood count was performed daily and at follow-up . A biochemistry sample for liver and renal function was performed at enrolment and subsequently depending on clinical need . An un-cuffed venous blood sample was taken at enrolment for venous lactate which was processed within 30 minutes of collection . Dengue diagnostics: Commercial IgM and IgG serology assays ( Capture ELISA , Panbio , Australia ) were performed on batched acute and convalescent plasma . In addition RT-PCR was performed on the enrolment sample to identify the DENV serotype and measure plasma viraemia levels [16] . Patients were defined as having dengue if the RT-PCR was positive or if the IgM assays were positive at enrolment , or IgM seroconversion between paired specimens and on the basis of their clinical picture . Patients with negative tests at enrolment , but for whom convalescent plasma was not available , were considered unclassifiable . Echocardiograms were performed at the bedside by one of the investigators ( SY , HT , VN ) , using an M-turbo system ( FUJIFILM SonoSite , Inc , USA ) with cardiac settings . The Echocardiograms were performed daily and at follow-up 14 days later . The exam included two-dimensional , M-mode and Doppler studies . More detailed methodology can be found elsewhere [6] . All images were stored digitally and a selection reviewed by a cardiologist ( CB ) in the United Kingdom . The inter- and intra- user variability was checked at regular intervals and was consistently <10% . Linear regression models were used for the initial analysis , with each cardio-haemodynamic parameter as the outcome and shock status as covariate . The analysis was adjusted for age , gender , and day of illness at ICU admission . Logistic regression was used for the prognostic models , predefined candidate variables at enrolment were used to predict recurrent shock/respiratory distress . Associations between the parameters were assessed by partial correlations controlling for the following potential confounding variables: age , sex and day of illness at enrolment , study day of measurement . Significance of partial correlations was assessed based on their Fisher transformation and corresponding bootstrap standard errors . The cluster bootstrap which resamples patients rather than samples accounted for multiple measurements per patient . To informally adjust for multiplicity , a significance level of 0 . 01 was used for all comparisons . All analyses were performed with the statistical software R version 3 . 2 . 2 and the companion package geepack version 1 . 2–0 .
Patients who developed recurrent shock had a higher enrolment pulse than those with 1 episode of shock or no shock ( median: 114 vs . 100 vs . 100 b/min , P = 0 . 002 ) , and reduced pulse pressure ( PP ) ( median: 20 vs . 20 vs . 30 mmHg , P = 0 . 001 ) ( Table 2 ) . There was a significantly lower Stroke Volume Index ( SVI ) in the patients with recurrent shock versus patients with and without 1 shock ( median: 21 . 6 vs . 22 . 8 vs . 26 . 8ml/m2 , P = 0 . 002 ) . SVI was significantly lower at enrolment ( study day 1 ) for patients with recurrent shock compared with no shock ( median: 21 . 6 vs . 26 . 8mls/m2 , P<0 . 001 ) and also between patients with shock compared with no shock ( median: 22 . 8 vs . 26 . 8mls/m2 , P = 0 . 001 ) ( Table 3 , Fig 1 ) . There was a significantly lower cardiac index ( CI ) between patients with shock compared to no shock on the first study day . The SVI remained lower for patients with recurrent shock versus no shock on study day 2 ( median: 22 . 8 vs . 27 . 2 mls/m2 , P = 0 . 004 ) . A non-significant trend for higher CI and LMPI was observed on study day 3 in the recurrent shock patients compared to no shock . Higher SVI on study day 4 was associated with the secondary endpoint of respiratory distress , as well as a trend for higher CI and respiratory distress ( S1 Table ) . The majority of patients only received IV fluids on days 1–2 days ( S2 Table ) , so the higher SVI on days 4 and 5 likely represents fluid re-absorption rather than iatrogenic causes . On study days 3–5 , worse left ventricular function ( higher LMPI ) was associated with respiratory distress . There was an association between the total IV fluid administered during the ICU admission and respiratory distress ( OR: 1 . 03 , 95% CI 1 . 01–1 . 06 , P = 0 . 001 ) . Respiratory distress presented early ( study day 2 ) in half of the patients , all had evidence of bilateral pleural effusions , suggesting plasma leakage likely causes early respiratory distress in ICU which is later compounded by fluid re-absorption and myocardial impairment on study days 3–5 . Enrolment lactate levels predicted patients who subsequently developed recurrent shock compared to those who did not ( Table 4 ) . In addition , higher enrolment lactate levels were also found to predict patients developing respiratory distress ( 3 . 9 vs . 3 . 0 mmol/l , OR 1 . 46 , 95% CI 1 . 09–2 . 12 , P = 0 . 008 ) . The SVI correlated with other parameters of intravascular volume including inferior vena cava collapsibility index ( IVCCI ) with a negative correlation ( rho -0 . 491 , P<0 . 001 ) and left ventricular end-diastolic diameter ( LVEDD ) with a positive correlation ( rho 0 . 354 , P<0 . 001 ) . The IVCCI correlated with LMPI with a positive correlation ( rho 0 . 230 , P<0 . 001 ) . The LMPI , RMPI , LVEDD and IVCCI did not correlate with the amount of IV fluids in the preceding 24 hours . Enrolment lactate levels correlated positively with the total IV fluid volume received ( rho: 0 . 323 , P = 0 . 001 ) and also with enrolment ALT ( rho: 0 . 764 , P<0 . 001 ) and AST ( rho: 0 . 773 , P<0 . 001 ) , but not with any of the cardio-haemodynamic parameters .
We have shown that myocardial impairment was not associated with recurrent shock but was associated with the secondary endpoint of respiratory distress after 3 days of the ICU admission . Lower stroke volume indices during the first 2 days of ICU admission and tachycardia were associated with both recurrent shock and respiratory distress . Higher lactate levels at ICU admission were also predictive for recurrent shock and respiratory distress . These results suggest patients with evidence of severe volume depletion at ICU admission including lower SVI , higher heart rates and venous lactates were more likely to develop recurrent shock and require more intravenous fluids–resulting in respiratory distress from a combination of plasma leakage and myocardial impairment , exacerbated by volume overload from fluid reabsorption in the recovery phase . Cardiac functional assessment in patients with hypovolaemia is more challenging and hence our use of Doppler derived parameters , which have been shown to be less preload dependent [17] . Myocardial dysfunction was associated with respiratory distress but not with recurrent shock , suggesting the myocardial impairment was sufficient to play a role in fluid overload , following resuscitation and the associated respiratory compromise but not to contribute to the shock syndrome , which appears to be driven predominantly by intravascular volume depletion . These findings are comparable to other echo studies , including one study of Thai children which demonstrated 36% of patients with DSS had reduced systolic function and patients with cardiac impairment were more likely to have fluid overload [18] . The mechanisms underlying this transient myocardial dysfunction in dengue patients remain to be defined , but potential mechanisms may involve some or a combination of the following; myocardial depressant factors , myocardial interstitial oedema , abnormal coronary microcirculation and endothelial dysfunction and also abnormal calcium homeostasis [8 , 19 , 20] . Most patients admitted to ICU with severe dengue showed signs of intravascular volume depletion , as evidenced by low SVI , CI , and smaller LVEDD and higher IVCCI compared to follow-up . Ejection fractions however remained normal in all the groups , which may be explained by low end diastolic volume and/or diastolic dysfunction- both which may play a role in dengue . SVI was the most robust parameter associated with the severe outcomes of recurrent shock and respiratory distress . Heart rate was significantly higher at enrolment between patients with recurrent shock versus no shock . This confirms a previous study where higher heart rate was found to be useful in predicting children developing profound shock [21] . The IVCCI , although being higher in patients at enrolment compared to discharge , did not discriminate between shock and no shock and was not associated with clinical outcomes . IVCCI has also been shown to correlate with central venous pressure ( CVP ) and right atrial pressure ( RAP ) in children and adults [22 , 23] , and is useful in predicting fluid responsiveness , in mechanically ventilated patients and in spontaneously breathing patients [24 , 25] . Due to the coagulopathy and thrombocytopenia in the majority of severe dengue patients , CVP carries a significant risk of bleeding and other non-invasive methods of assessing intravascular volume and guiding fluid therapy are urgently needed [26] . Portable bedside echocardiographic assessment of haemodynamics , particularly the SVI are useful in identifying patients with recurrent shock and could be considered as an alternative to invasive CVP monitoring . We have shown venous lactates in dengue patients on the first day of admission to ICU is associated with severe outcomes of recurrent shock and respiratory distress . Lactate levels correlated with the total amount of IV fluids received , but did not correlate with other haemodynamic parameters . The higher lactates likely represent severe volume depletion from plasma leakage causing tissue hypoperfusion , hypoxia and anaerobic glycolysis . In addition to hypoperfusion and excess production of lactate , another mechanism for hyperlactatemia in severe dengue may involve reduced hepatic clearance as moderate hepatic dysfunction occurs in severe dengue [27] . The liver may play a role in the hyperlactatemia in critical illness with circulatory failure , not only by reduced metabolism but also because the liver itself can produce lactate due to hepatic ischaemia . This is supported by our study which showed lactate levels correlated positively with both ALT and AST levels . A study investigating patients with shock admitted to ICU , found higher lactate levels in patients with early hepatic dysfunction compared to those with no hepatic dysfunction , independent of haemodynamic severity parameters [28] . Altered microcirculation , which we have shown is worse in dengue patients with more severe plasma leakage [9] , may play a role in the increased lactate levels , although further studies are required to link the microcirculatory perfusion abnormalities with higher lactate levels . The current WHO guidelines for managing DSS recommend initial resuscitation with crystalloid fluids for compensated shock , followed by careful on-going assessment including serial HCT measurements and close monitoring of vital signs . Reassessing patients in shock and achieving predefined physiological targets has been a major focus of research in severe sepsis in the last 2 decades [29–31] . The ‘goals’ of resuscitation are currently being readdressed with emerging evidence that a conservative approach to fluid management has better outcomes in certain settings [32] . The balance of administering just sufficient intravenous fluid therapy to maintain haemodynamic stability while avoiding fluid overload and respiratory compromise is extremely difficult and additional cardiovascular monitoring using portable echocardiography in DSS would be beneficial . Stroke volume monitoring may provide improved targeted volume resuscitation . While serial HCT and vital sign monitoring are useful and widely applicable for resource constrained settings , intensive care facilities and associated technologies are improving in many dengue endemic areas , so additional non-invasive cardiovascular assessment is now possible and should be considered where available . There were some limitations to our study . In order not to interfere with emergency management , some patients had the first echo study after initial fluid resuscitation had commenced . This may therefore underestimate some of the cardiovascular parameters . Secondly , due to restrictions on research blood sampling in paediatric patients , we were unable to take daily lactates so it was not possible to study lactate clearance times . In addition , due to the coagulopathy in many of severe dengue patients , we were unable to take arterial blood gases for assessment of metabolic acidosis to explore relationship with the high lactate levels . As the majority of patients enrolled in this study were children and young adults , the results may not be generalizable to older adult populations with dengue .
In conclusion , this study has identified several simple non-invasive parameters that could assist risk prediction and help tailor management of dengue patients admitted to ICU . We have shown moderate cardiac dysfunction is common in ICU patients with dengue , particularly among adults . The cardiac dysfunction does not appear to play a major part in the haemodynamic instability of dengue shock but likely contributes to the development of fluid overload and respiratory compromise in some cases . Echo-derived volume assessment using stroke volume index combined with heart rate monitoring and venous lactate levels can help identify patients at high risk of recurrent shock . The clinical and therapeutic implications of these findings are potentially important , first as prognostic markers to guide fluid resuscitation and assist in the management of DSS as currently practiced , and second , these echo-derived haemodynamic endpoints could be used in future dengue fluid intervention trials designed to assess alternative strategies intended to improve DSS management and outcome . | Dengue is a viral illness that can lead to severe and potentially fatal complications . The most common complication is fluid leakage from blood vessels , which can cause low blood pressure or dengue shock syndrome ( DSS ) . The majority of patients recover with simple intravenous fluid replacement , however in approximately 30% of DSS cases , recurrent episodes of shock occur , and these patients have a higher risk of fluid overload , respiratory distress and death . We investigated whether using portable echocardiograms ( Echo ) in the intensive care unit ( ICU ) to assess cardiac function and intravascular volume parameters as well as blood lactate levels , can help identify these patients . We found patients who developed recurrent shock had higher heart rates and lower Stroke Volume Index ( SVI ) , and higher lactate levels at enrolment than those with 1 episode or no shock . Higher SVI and worse cardiac function after 3 days in ICU was associated with respiratory distress . Admission lactate levels predicted patients who subsequently developed recurrent shock and correlated positively with the total IV fluid volume received . These results demonstrate that Echo-derived intravascular volume assessment and venous lactate levels can help identify dengue patients at high risk of poor outcomes in the ICU , and could assist in the management of severe dengue . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusion"
] | [
"medicine",
"and",
"health",
"sciences",
"diagnostic",
"radiology",
"ultrasound",
"imaging",
"cardiovascular",
"physiology",
"tropical",
"diseases",
"echocardiography",
"health",
"care",
"pediatrics",
"cardiovascular",
"analysis",
"neglected",
"tropical",
"diseases",
"bioas... | 2017 | Cardio-haemodynamic assessment and venous lactate in severe dengue: Relationship with recurrent shock and respiratory distress |
Isoflavones comprise a group of secondary metabolites produced almost exclusively by plants in the legume family , including soybean [Glycine max ( L . ) Merr . ] . They play vital roles in plant defense and have many beneficial effects on human health . Isoflavone content is a complex quantitative trait controlled by multiple genes , and the genetic mechanisms underlying isoflavone biosynthesis remain largely unknown . Via a genome-wide association study ( GWAS ) , we identified 28 single nucleotide polymorphisms ( SNPs ) that are significantly associated with isoflavone concentrations in soybean . One of these 28 SNPs was located in the 5’-untranslated region ( 5’-UTR ) of an R2R3-type MYB transcription factor , GmMYB29 , and this gene was thus selected as a candidate gene for further analyses . A subcellular localization study confirmed that GmMYB29 was located in the nucleus . Transient reporter gene assays demonstrated that GmMYB29 activated the IFS2 ( isoflavone synthase 2 ) and CHS8 ( chalcone synthase 8 ) gene promoters . Overexpression and RNAi-mediated silencing of GmMYB29 in soybean hairy roots resulted in increased and decreased isoflavone content , respectively . Moreover , a candidate-gene association analysis revealed that 11 natural GmMYB29 polymorphisms were significantly associated with isoflavone contents , and regulation of GmMYB29 expression could partially contribute to the observed phenotypic variation . Taken together , these results provide important genetic insights into the molecular mechanisms underlying isoflavone biosynthesis in soybean .
Isoflavones are a group of secondary metabolites predominantly distributed in leguminous plants , including soybean [Glycine max ( L . ) Merr . ] [1] . In plants , isoflavones play important roles in microbial interactions , functioning as phytoalexins to protect plants from pathogen infection [2 , 3] . They also act as signal molecules in the formation of nitrogen-fixing root nodules in leguminous plants [4] . For humans , isoflavones have health benefits in the prevention of several diseases , such as cancer [5] , cardiovascular disease [6] , and climacteric syndrome [7] , which are associated with their phytoestrogenic and antioxidant properties [8] . However , isoflavones are undesirable in soy-based infant formulas [9] . In soybean breeding , an improved understanding of the mechanism of isoflavone biosynthesis would be of great value , as it may allow the manipulation of isoflavone biosynthesis and the production of cultivars that can meet various needs . In soybean , there are three core isoflavone aglycones: daidzein , genistein , and glycitein [10] . They are synthesized via the general phenylpropanoid pathway that exists in all higher plant species and a branch of the isoflavonoid biosynthesis pathway specific to leguminous plants [11] . Isoflavone biosynthesis begins with the deamination of phenylalanine by phenylalanine ammonia lyase ( PAL ) . After steps catalyzed by a series of enzymes , the critical branch point enzymes chalcone synthase ( CHS ) and isoflavone synthase ( IFS ) lead substrates to the isoflavone synthesis branch and finally generate isoflavones and their derivatives [12] . In addition to isoflavone biosynthesis , the phenylpropanoid pathway is also involved in the synthesis of lignins , stilbene , phlobaphenes , proanthocyanidins and anthocyanins via specific branches . Hence , the biosynthesis of isoflavones involves an intricate network reconciling many competing branch pathways . Thus , the modulation of a single gene does not necessarily alter the metabolic flux to target branch pathways [12–14] . The isoflavone biosynthesis pathway is complex , and functional differentiation is found in the isoflavone synthesis-related gene families due to two recent whole-genome duplication events: a soybean-lineage-specific duplication 13 million years ago and an early-legume duplication 59 million years ago [15 , 16] . Therefore , researchers have focused on the discovery and application of isoflavone regulation-related transcription factors ( TFs ) instead of the manipulation of a single gene . Various TFs have been identified to regulate the biosynthesis of phenylpropane substances in higher plants , such as MYB , bZIP , WRKY , MADS box and WD40 . Some MYB TFs involved in the regulation of the isoflavonoid biosynthesis pathway have been identified in soybean . For example , the R1-type MYB TF GmMYB176 has been shown to affect isoflavonoid synthesis by regulating CHS8 gene expression [1] . The R2R3-type MYB TFs GmMYB39 and GmMYB100 have been reported to negatively regulate isoflavonoid biosynthesis by suppressing the expression of structural biosynthesis genes [17 , 18] . The soybean genome contains 4343 putative transcription factors , which account for 6 . 5% of the total predicted genes [19] . It is therefore challenging to discover and identify the key isoflavone regulation-related transcription factors at the genomic level . Isoflavone content is a complex quantitative trait controlled by multiple genes and affected by both genetic and environmental factors . The primary mapping method for isoflavone-related quantitative trait loci ( QTLs ) is linkage analysis based on family lines , which is a classical method used to investigate complicated quantitative traits [19] . Previous studies have identified many QTLs controlling the biosynthesis of isoflavones in soybean seeds [20–25] . However , no isoflavone synthesis-related QTLs have been cloned due to limited allelic variation between recombinant inbred population parents . The application of genome-wide association study ( GWAS ) , a more accurate method than linkage analysis , could enhance the power of functional gene identification [26 , 27] . The development of the large genome-wide NJAU 355K SoySNP array in our previous study provides a useful tool facilitating GWAS in soybean [28] . In this study , we used this array to perform a GWAS for isoflavone contents and revealed a number of potential loci controlling isoflavone biosynthesis in soybean . We then demonstrated that one candidate gene , an R2R3-type MYB TF designated GmMYB29 , played an important role in the regulation of isoflavone biosynthesis in soybean . Transcription analyses revealed a close correlation between the expression of GmMYB29 and IFS2 under normal and stressed conditions as well as between the expression of GmMYB29 and the accumulation of isoflavones . Transient reporter gene assays and overexpression of GmMYB29 in soybean hairy roots also strongly supported its key roles in the regulation of IFS2 and CHS8 expression and the isoflavone accumulation . Additionally , by combining a GmMYB29-based association analysis with an analysis of GmMYB29 expression in seed samples of 30 natural soybean varieties , we confirmed the positive regulatory role of GmMYB29 in isoflavone biosynthesis .
To determine the range of variation of isoflavone contents in soybean , the total isoflavone contents ( TIC ) , daidzein contents ( DAC ) , genistein contents ( GEC ) and glycitein contents ( GLC ) in soybean seeds were determined using 196 soybean accessions . To address the potential environmental influence , the soybean accessions were grown in two locations: Nanjing and Nantong ( designed as two environments ) ( Table 1 ) . A broad variation in isoflavone contents was observed in the population . For example , the GLC varied from 10 . 36 μg g-1 to 1794 . 00 μg g-1 in Nanjing . The average TIC , DAC , GEC and GLC was 5445 . 74 μg g-1 , 3596 . 17 μg g-1 , 946 . 73 μg g-1 , and 423 . 85 μg g-1 , respectively . The isoflavone contents showed continuous variation and normal distribution ( S1 Fig ) , with skew and kurtosis less than one in the different environments . An analysis of variance ( ANOVA ) revealed that genotype and the genotype-by-environment interaction significantly influenced the major isoflavone contents ( P < 0 . 001 ) . This result supported the idea that isoflavone content is a complex trait controlled by multiple factors . However , the broad-sense heritability ( h2 ) values of TIC , DAC , GEC and GLC were 74 . 1% , 76 . 3% , 67 . 8% and 83 . 8% , respectively , indicating that isoflavone content was primarily affected by genetic factors . To identify the loci associated with isoflavone contents , GWAS was conducted using TIC , DAC , GEC , GLC and 207 , 608 SNPs with a minor allele frequency ( MAF ) > 0 . 05 . These SNPs were obtained from the genotyping results of the 196 soybean accessions acquired using the NJAU 355K SoySNP array [28] . Twenty-eight SNPs significantly associated with the major isoflavone components were not only detected in the NJ or NT environment but also repetitively detected in the best linear unbiased prediction ( BLUP ) data set under a threshold of P < 4 . 82×10−6 . These SNPs were considered as potentially reliable SNPs for further analysis ( Table 2 and Fig 1 ) . Additionally , 22 significant SNPs detected only once in the NJ environment , NT environment or the BLUP data set are presented in S1 Table . The 28 significant SNPs were located on chromosomes 5 , 6 , 11 and 20 and assembled into clusters on chromosomes 11 and 20 . Among the significant SNPs , 17 , 10 , and 11 SNPs were associated with TIC , DAC , and GLC , respectively . Notably , the 10 SNPs significantly associated with DAC were overlapped with those associated with TIC . Unfortunately , no detected SNPs were significantly associated with GEC . The phenotypic variation explained by each of these significant 28 SNPs ranged from 10 . 20% to 14 . 98% , suggesting that major QTLs for isoflavone contents may exist . Based on the linkage disequilibrium ( LD ) decay calculated previously , the genes within 130 kb flanking the significant SNPs were selected [28] . Among these genes , no known genes in the isoflavone biosynthesis pathway were identified . Therefore , it was speculated that there could be novel genes related to isoflavone biosynthesis or regulation in these loci . Consistently , numerous TF-encoding genes , including MYB , NAC , bZIP , and WRKY were identified ( Table 2 ) . These TF-encoding genes could function in the regulation of isoflavone biosynthesis . Among the 28 significant SNPs , there was only one SNP detected in NJ , NT and the BLUP data set . Notably , this SNP ( AX-93910416 ) was detected within the 5’-untranslated region ( 5’-UTR ) of Glyma20g35180 ( GmMYB29 ) . Interestingly , the homologous gene LjMYB14 has been characterized as a TF regulating isoflavonoid biosynthesis in Lotus ( Lotus japonicas ) [29] . These results suggested that Glyma20g35180 could be a candidate gene controlling isoflavone biosynthesis in soybean . The full-length open reading frame ( ORF ) of GmMYB29 was 819 bp and encoded a protein of 272 amino acid residues with a calculated mass of 31 . 15 kDa and a pI of 5 . 77 . The GmMYB29 protein was predicted to belong to the R2R3-type MYB subfamily . A multiple alignment of GmMYB29 with R2R3-type MYB TFs known to regulate isoflavonoids or flavonoids from various plant species showed a high homology in the N-terminal MYB domain ( S2 Fig ) . GmMYB29 was clustered with AtMYB13 , AtMYB14 , AtMYB15 , NtMYB2 , DcMYB1 , LjMYB13 , LjMYB14 , LjMYB15 , VvMYB14 , and VvMYB15 , and they shared a C-terminal conserved motif found in subgroup 2 of the R2R3-type MYB gene family in Arabidopsis ( the SG2 motif , DxSFW-MxFWFD ) , which has previously been described as a stress response motif ( S2 Fig ) [30–32] . Consistent with this finding , in the phylogenetic tree , the proteins from various plants that were grouped in the same cluster with GmMYB29 have been reported to respond to biotic or abiotic stresses ( S3 Fig ) [30] . Notably , LjMYB13 , LjMYB14 and LjMYB15 , which regulate isoflavonoid biosynthesis in Lotus [29] , were found to form a cluster with GmMYB29 . To determine the subcellular localization of the GmMYB29 protein , the GmMYB29 cDNA was fused with green fluorescent protein ( GFP ) under the control of the CaMV 35S promoter . This construct was then transformed into Arabidopsis mesophyll protoplasts using polyethylene glycol ( PEG ) and into onion epidermal cells using a gene gun . Consistent with the putative function of TFs , the GmMYB29::GFP fusion protein was localized in the nucleus , while in cells transformed with a GFP control plasmid , fluorescence was detected in both the cytoplasm and the nucleus ( S4 Fig ) . Previous studies have revealed that glutathione ( GSH ) treatment could induce isoflavonoid production [33] and that biotic stress could influence isoflavone content [34] . To determine whether the expression of GmMYB29 was induced by GSH and biotic stress and whether the expression pattern of GmMYB29 was consistent with that of isoflavone synthase 2 ( IFS2 ) , a key gene in isoflavone biosynthesis , we examined the expression of GmMYB29 and IFS2 in soybean leaves treated with GSH and common cutworms ( Fig 2 ) . After 3 h of GSH treatment , GmMYB29 and IFS2 showed 7- and 4-fold higher expression in treated leaves than in the control samples , respectively . After 7 h of treatment , these two genes showed 26- and 10-fold increases in expression , respectively . After insect-mediated induction , both GmMYB29 and IFS2 displayed marked up-regulation at 1 h and 6 h . Therefore , both GmMYB29 and IFS2 showed induced expression by GSH elicitation and insect feeding , and they were co-expressed under these two stresses . These results indicated that GmMYB29 and IFS2 could be involved in similar or the same biological processes . To further examine the correlation between GmMYB29 and IFS2 expression in different tissues and developmental stages of soybean , we investigated the expression patterns of GmMYB29 and IFS2 and the isoflavone content in different soybean tissues ( Fig 3 ) . The expression of GmMYB29 was closely associated with that of IFS2 . These two genes showed relatively higher expression in roots and seeds than in other tissues , and the expression noticeably increased with seed development . Notably , the expression of GmMYB29 and IFS2 was consistent with the isoflavone content in different tissues , suggesting that the expression of GmMYB29 and IFS2 is closely related with isoflavone accumulation in soybean . To examine whether GmMYB29 could regulate the expression of isoflavone biosynthesis-related genes , transient expression using Arabidopsis mesophyll protoplasts and a dual luciferase reporter gene assay was performed . The promoters of two critical genes ( IFS2 and CHS8 ) in the isoflavone biosynthesis pathway were amplified from 1790 bp and 1663 bp upstream of the start codons of IFS2 and CHS8 , respectively , to study the interaction between these promoters and GmMYB29 . Several MYB binding elements and stress-related cis-regulatory elements were predicted in the IFS2 and CHS8 promoters using the PLACE database ( http://www . dna . affrc . go . jp/htdocs/PLACE/ ) [35] . As shown in Fig 4 , transient expression demonstrated that overexpression of GmMYB29 increased the activity of both the IFS2 and CHS8 promoters by 100- and 200-fold , respectively , compared with the controls . These results suggested that GmMYB29 plays a critical role in the transcriptional regulation of key genes in the soybean isoflavone biosynthesis pathway . To further identify the GmMYB29 recognition regions in the IFS2 promoter , eight fragments ( IFS2ΔP1-IFS2ΔP8 ) were generated by gradual 5’ deletions of the promoter , which were then used to drive luciferase ( LUC ) expression ( S5 Fig ) . The reporters IFS2ΔP1-IFS2ΔP8pro:LUC and IFS2fullpro:LUC were co-transfected into Arabidopsis protoplasts with 35Spro:GmMYB29 , and the LUC activity was measured . The vectors containing IFS2ΔP1 and IFS2ΔP2 showed similar LUC activity to that containing IFS2full . However , the LUC activity dramatically decreased for IFS2ΔP3 and the further deletions . These results indicated that the 208 bp region between -885 and -1093 in the IFS2 promoter contained the motif required for promoter activity . In this region , a cis-regulatory element , MYBCORE ( containing the CNGTTR sequence ) , was predicted by PLACE as a MYB binding element , suggesting that GmMYB29 could bind the IFS2 promoter and activate IFS2 expression via the recognition of this element . To determine the role of GmMYB29 in isoflavone accumulation , overexpression and RNAi-mediated silencing of GmMYB29 were performed using a soybean hairy root system ( S6 Fig ) . The transgenic hairy roots were verified by PCR amplification , and the positive lines were selected to conduct further studies ( S7 Fig ) . We performed quantitative RT-PCR to study the effect of overexpression and RNAi silencing on the transcription levels of GmMYB29 and isoflavone biosynthesis-related genes , including PAL , cinnamate 4-hydroxylase ( C4H ) , 4-coumarate coenzyme A ligase ( 4CL ) , CHS8 , chalcone isomerase ( CHI ) , chalcone reductase ( CHR ) and IFS2 , in hairy roots obtained from several independent transgenic lines . The transcription level of GmMYB29 was significantly increased by 14- to 47-fold in GmMYB29-overexpressing transgenic hairy roots ( S8A Fig ) and significantly reduced by 3- to 7-fold in transgenic hairy roots with RNAi-mediated GmMYB29 silencing ( S8B Fig ) . The GmMYB29-overexpressing transgenic roots also showed increased transcription levels of PAL , 4CL , CHS8 , CHR , and IFS2 , but no significant change in C4H and CHI expression was observed between overexpressing and control roots ( S9A Fig ) . Interestingly , the transcription levels of all the monitored isoflavone biosynthesis genes were markedly decreased in GmMYB29-silenced transgenic roots ( S9B Fig ) . Furthermore , we measured the total isoflavone contents in GmMYB29-overexpressing and GmMYB29-silenced lines by high-performance liquid chromatography ( HPLC ) . The isoflavone content increased by 1 . 6- to 3 . 3-fold in GmMYB29-overexpressing roots ( Fig 5A ) and decreased by 2-fold in the gene-silenced roots ( Fig 5B ) ( P < 0 . 01 ) . To further investigate the association between the allelic variation of GmMYB29 and isoflavone contents , we sequenced the GmMYB29 gene in a subset of 30 soybean accessions , representing varieties with high , medium and low isoflavone contents . An approximately 2 . 4-kb genomic region , spanning the 5'- to 3'-UTR of GmMYB29 , was analyzed . A total of 12 SNPs and 11 indels ( insertions and deletions ) were identified and filtered out for the subsequent association analyses ( Fig 6A ) . The association study showed that 11 probable causative sites , including Site-102 ( located 102 bp upstream from the translation start codon , S-102 ) , S-46 and S-12 in the 5'-UTR , S99 in exon1 , S489 in exon2 , Indel645 , S679 and S1167 in intron2 , S1619 in exon3 and Indel2134 and S2135 in the 3'-UTR , were significantly associated with variations in the TIC ( Table 3 ) . S-12 ( corresponding to SNP AX-93910416 in our GWAS results ) and S-46 were significantly correlated with the TIC , both contributing to 49 . 99% of the phenotypic variation in the representative subset . A single-base transversion at S1619 resulted in an amino acid substitution of lysine to asparagine at amino acid position 133 , which contributed to 14 . 91% of the variation in TIC . Furthermore , based on the 11 significant variants with strong LDs ( Fig 6A ) , the 30 soybean genotypes were classified into seven haplotype classes ( Hap1-Hap7 ) ( Fig 6B ) . Haplotype 3 ( Hap3 , n = 19 ) is the largest group , and Hap5 ( n = 6 ) is the second largest group , whereas the other five haplotype classes are minor groups , each comprising one soybean accession . Statistically , Hap3 accessions had significantly higher TIC than Hap5 accessions ( Fig 6C ) . Among the different sites between Hap3 and Hap5 , the most significant variants are S-46 and S-12 , which are located in the 5'-UTR . Considering that the expression of GmMYB29 may cause phenotypic variation , we subsequently measured the expression of GmMYB29 in seeds from these 30 soybean accessions . The expression of this gene was positively correlated with the isoflavone content ( r = 0 . 63 , P < 0 . 01 ) ( S2 Table ) . Additionally , we observed that the Hap3 accessions exhibited higher GmMYB29 expression than the Hap5 accessions ( Fig 6D ) . Therefore , these data suggested that the expression of GmMYB29 could at least partially explain the phenotypic variation in isoflavone content .
Previously , a number of QTLs associated with soybean isoflavone-related traits have been identified by linkage mapping , but few of these have been cloned or functionally characterized , perhaps because of the QTL resolution , which is limited by lower density molecular markers [38] . GWAS with high-density markers can overcome this limitation and have recently been successfully applied in studies of Arabidopsis thaliana , rice and maize [39–41] . In the present study , using a diverse natural population genotyped with high-density markers ( 292 , 053 SNPs , one SNP/3 . 3 kb ) and phenotyped in various environments , we identified an important TF related to soybean isoflavone biosynthesis , clarified its molecular mechanism and determined the favorable alleles/haplotypes . Studies have shown that the selection of appropriate mapping populations genotyped with saturated markers is important for performing GWAS to identify complex QTLs [42] . The 196 accessions used in this study have been reported to identify QTLs associated with seed shape , phosphorus efficiency and yield , among other features [43–45] , suggesting that this panel might contain diverse genetic variations in complex quantitative soybean traits . As expected , many genetic variations in TIC , DAC , GEC and GLC were observed in the association mapping population . In addition , DAC was always the highest , followed by GEC and then GLC across various environments , which was consistent with a previous report [25] . Although isoflavone content was affected by both the genotype and the interaction between genotype and environment , isoflavone content also maintained a high heritability ( 0 . 68–0 . 84 ) , which agreed with recently reported results [22–25] . These studies reveal that the heritability of isoflavone content is high enough to be considered in breeding practices to genetically improve cultivars effectively . GWAS based on high-density SNP markers can be used to finely map quantitative trait genes , even to the genes themselves . Recently , an 82-bp ( MITE ) insertion in the promoter region of a NAC gene ( ZmNAC111 ) detected by a GWAS has been determined to be associated with maize drought tolerance [46] . In our study , a highly significant SNP , AX-93910416 , was identified to be associated with soybean isoflavone contents across two environments ( NJ and NT ) and their BLUP . A strong LD was detected in the region around this SNP ( S10 Fig ) , indicating the existence of artificial selection and a potential target gene responsible for phenotypes in this region . More importantly , this SNP was located in the 5’-UTR of the transcription factor GmMYB29 . In addition , GmMYB29 is homologous to LjMYB14 , a transcription factor reported to regulate isoflavone biosynthesis in Lotus corniculatus [29] , indicating that GmMYB29 is possibly involved in isoflavone regulation . It is known that similar and conserved protein functions are derived from conserved motifs from a common origin . Comparative genomic analyses have shown that GmMYB29 not only maintains the highly conserved R2R3 domain but also has a small amino acid motif in the C-terminal region . This small motif is SG2 ( DxSFW-MxFWFD ) , which has been reported to be related to stress resistance in plants [30–32 , 47 , 48] . For example , AtMYB15 , which contains an SG2 motif , reportedly can improve stress resistance by increasing the expression of genes related to ABA synthesis and the ABA signaling pathway in Arabidopsis plants exposed to drought and salinity [49 , 50] . Similarly , our results showed that the expression of GmMYB29 was significantly increased under abiotic and biotic stress . Interestingly , the expression pattern of GmMYB29 was similar to that of IFS2 , a key structural gene in the isoflavone biosynthesis pathway , which suggests that GmMYB29 could be involved in the same regulation pathway as IFS2 [29 , 30] . Furthermore , the expression profile of GmMYB29 determined by quantitative RT-PCR analysis in different tissues showed that GmMYB29 was expressed in every isoflavone-accumulating tissue [51] . However , the expression of GmMYB29 preceded IFS2 in different developmental stages of soybean seeds . For instance , the highest expression levels of GmMYB29 and IFS2 occur on the 40th and 50th day after flowering , respectively . This is consistent with the hypothesis that the expression of regulators occurs in advance of their target genes [30] . These results indicate that GmMYB29 may regulate isoflavone biosynthesis in soybean . Transcription factors often act on the promoter region of their target genes and regulate their expression [52 , 53] . As expected , we found that GmMYB29 can interact with the promoters of IFS2 and CHS8 and activate the expression of these two genes . Furthermore , co-transfection of promoter deletion fragments showed that a 208-bp fragment ( from -885 bp to -1093 bp ) , which contains the MYB TF binding cis-acting element MYBCORE ( CNGTTR ) , was necessary for the activation of IFS2pro:LUC , indicating the important role of this element in MYB recognition and gene transcriptional regulation . In addition , two other MYBCORE elements , located in the -1158 bp to -1790 bp and -1093 bp to -1158 bp regions of the IFS2 promoter , respectively , were identified by cis-acting element prediction software . However , gradual deletions of these two elements ( generating IFS2ΔP1 and IFS2ΔP2 , respectively ) showed no significant effects on LUC activity ( S5 Fig ) . The IFS2ΔP3 construct , in which all of the MYBCORE elements were deleted , exhibited almost no LUC activity , confirming that the MYB elements are the key sites recognized by GmMYB29 , thereby affecting IFS2 transcription . The promoter sequences of other structural genes ( PAL1 , C4H , 4CL , CHS8 , CHI , CHR and IFS1 ) in the isoflavone pathway were also investigated , and various MYBCORE elements were identified . Thus , further experiments are required to confirm whether GmMYB29 directly interacts with other isoflavone pathway-related genes . The isoflavone biosynthesis-related R2R3-type MYB TFs reported by previous studies have generally been negative regulators in soybean [17 , 18] . For example , GmMYB100 was found to inhibit isoflavonoid production by down-regulating the expression of CHS , CHI and IFS [17] . In addition to the negative regulators , the R1-type MYB TF GmMYB176 , which could activate the promoter activity of CHS8 , was also observed . RNAi-mediated silencing of GmMYB176 in transgenic soybean hairy roots resulted in reduced levels of isoflavonoids . Unfortunately , overexpression of GmMYB176 was not sufficient to increase CHS8 transcription and isoflavonoid levels in hairy roots [1] . In our study , overexpression of GmMYB29 increased the activity of IFS2 and CHS8 promoters; moreover , the isoflavone content was increased in GmMYB29-overexpressing hairy roots and decreased in GmMYB29-silenced hairy roots . These results imply that we identified a novel R2R3-type MYB TF , GmMYB29 , which acts as a positive regulator to activate isoflavone production . Surprisingly , the level of isoflavone production in GmMYB29-overexpressing hairy roots was found to be only 3 . 3-fold higher at most than the control lines . One possible explanation for this observation may be the possible phytotoxic effect of isoflavonoid accumulation on the growth of hairy roots , as reflected in the relatively slow growth of GmMYB29-overexpressing transgenic hairy roots and the loss of several lines with high GmMYB29 expression . Soybean hairy roots overexpressing GmMYB29 showed a marked increase in the expression of PAL , 4CL and CHS8 as well as CHR and IFS2 ( S9A Fig ) . This suggests that in addition to its role in the regulation of isoflavone biosynthesis , GmMYB29 may also be involved in the regulation of upstream phenylpropanoid pathway genes to ensure the availability of substrates for isoflavone biosynthesis . It has been reported that a single TF could regulate multiple genes in the phenylpropanoid pathway and that expression of a single target gene in the pathway might be regulated by multiple TFs [54–58] . The transcriptional regulation of the anthocyanin and proanthocyanidin pathway genes is conducted by a complex in which R2R3-type MYB TFs , WD40 proteins , and bHLH proteins all interact to activate gene transcription [59–61] . Thus , we cannot exclude the possibility that there are other TFs that can activate the biosynthesis of isoflavones . Further characterization of other TFs identified in this research would provide deeper insight into the regulatory mechanisms underlying isoflavone biosynthesis . In addition to transformation , the selection and accumulation of elite alleles of key genes functioning in isoflavone biosynthesis may be an effective strategy for soybean breeding . Similar investigations have been reported for maize , rice , soybean , and Arabidopsis , among others [46 , 62–65] . For example , a sequence analysis of the drought tolerance gene ZmVPP1 in 140 inbred maize lines identified a 366-bp insertion in the promoter , which was associated with maize drought tolerance and conferred drought-inducible expression of ZmVPP1 in drought-tolerant accessions . Although some isoflavone biosynthesis-related TFs have been characterized [1 , 17 , 18] , the polymorphism and haplotype analyses of these genes and the potential regulation mechanisms have not been reported . In this study , haplotype analysis showed that the GmMYB29 gene can be found as seven haplotypes ( Hap1-Hap7 ) , and Hap3 had higher levels of isoflavone content and GmMYB29 expression than the others , indicating that Hap3 might be significant to breeding soybeans with higher isoflavone content . Here , the 30 soybean accessions used to identify the favorable haplotype of GmMYB29 were selected to represent soybeans with different levels of isoflavone content . However , as more than 20 , 000 soybean accessions have been preserved [66] , other elite alleles of GmMYB29 might be discovered using the stored soybean germplasms . The optimal haplotypes and alleles of this gene could therefore be detected by investigating the genetic differences in GmMYB29 expression and transcriptional activity in additional soybean accessions . Taken together , these results could lead to the development of molecular markers for the breeding of soybeans with optimized isoflavone content .
A natural population of 196 representative cultivated soybean ( Glycine max ) accessions with broad variations in isoflavone contents was selected from Wang et al . to develop the association mapping panel ( S3 Table ) [28] . These accessions ( including 164 landraces , 24 improved accessions and 8 accessions with unknown evolution type ) originated from all three ecological habitats of soybean in China . The seeds of each accession were provided by the Germplasm Storage of the Chinese National Center for Soybean Improvement ( Nanjing Agricultural University , Nanjing , China ) . The trials were conducted in 2014 at two locations: Jiangpu Station of Nanjing Agricultural University in Nanjing ( 32°12'N , 118°37'E ) ( designated as environment NJ ) and the Experimental Farm of Jiangsu Yanjiang Institute of Agricultural Sciences in Nantong ( 31°58'N , 120°53'E ) ( designated as environment NT ) . In each environment , 196 soybean accessions ( corresponding to 196 plots ) were planted in a randomized complete block design with three replicate blocks . Each accession was planted in four rows per plot , and each row was 200 cm long , with a row-spacing of 50 cm . The inter-plot spacing was also 50 cm . All field management requirements during the growing period , including watering , weed management , and fertilization , were performed similarly at both test locations . After maturity , four individuals from each plot were randomly screened for isoflavone content analysis . The extraction and determination of isoflavones was performed according to the protocol described by Sun et al . [67] . First , approximately 20 g of dried seeds from each accession was ground to a fine powder using a cyclone mill . Fifteen milligrams of this powder was added to a 2 mL tube containing 1 . 5 mL of 80% methanol . The mixture was spun for 30 s , subjected to ultrasound treatment ( frequency 40 kHz , power 300 W ) for one hour at 50°C , and rotated every ten minutes . After centrifugation at 12 , 000 rpm for 10 min , the supernatant was filtered using a YMC Duo-filter ( YMC Co . , Kyoto , Japan ) with 0 . 22 μm pores . The sample was injected into a 2 mL Agilent auto sampler and stored at -20°C before use . Samples were analyzed with a high-performance liquid chromatography ( HPLC ) system ( Column: Zorbax SB-C18 , 5 μm , 80 Å , 4 . 6 mm×150 mm ) under the following conditions: solvent A was 0 . 1% aqueous acetic acid and solvent B was 100% methanol; the solvent system was 0–2 min 27% B ( v/v ) , 2–3 min 27%-38% B , 3–10 min 38% B , 10–12 min 38%-39% B , 12–14 min 39% B , and 14–15 min 39%-27% B . The solvent flow rate was 2 mL/min , and the UV absorption was measured at 254 nm . The column temperature was set at 36°C , and the injection volume was 10 μL . The identification and quantification of each isoflavone component was based on the standards of 12 isoflavone components provided by Sigma-Aldrich . The 12 isoflavone standards were daidzin ( D ) , glycitin ( GL ) , genistin ( G ) , daidzein ( DE ) , glycitein ( GLE ) , genistein ( GE ) , malonyldaidzin ( MD ) , malonylglycitin ( MGL ) , malonylgenistin ( MG ) , acetyldaidzin ( AD ) , acetylglycitin ( AGL ) , and acetylgenistin ( AG ) . Different concentration gradients ( 0 , 5 , 10 , 20 , 50 , 100 , 500 , 1000 ng/sampler ) of the 12 isoflavone standards in 80% methanol were produced . Twelve standard curves were generated to calculate the 12 kinds of isoflavone monomer content . The precise contents of these 12 isoflavone components per gram of dry seeds ( μg g-1 ) were calculated with the formula described in detail by Sun et al . [67] . The total isoflavone contents ( TIC ) were calculated as the sum of the 12 isoflavone components . The contents of daidzein ( DAC ) , genistein ( GEC ) and glycitein ( GLC ) were the sum of four corresponding components: malonyl glycosides , acetyl glucosides , β-glycosides and aglycones . An analysis of variance ( ANOVA ) of all phenotypic data was performed using PROC GLM of SAS/STAT 9 . 2 ( SAS Institute , 2002 ) with environment , replication within environment , genotype and genotype × environment as random effects . The ANOVA was based on the model yijr = μ+Gi+Ej+ ( GE ) ij+Br ( j ) +εijr , where yijr is the phenotype value of the ith genotype in the jth environment and the rth block , μ is the population mean; Gi is the effect of the ith genotype , Ej is the effect of the jth environment , ( GE ) ij is the genotype-by-environment interaction effect , and Br ( j ) is the effect of the rth replicate block in jth environment; and εijr is the random error . The broad-sense heritability values ( h2 ) were estimated as h2 = σ2g/ ( σ2g+σ2ge/n+σ2/nr ) , where σ2g is the genetic variance , σ2ge is the interaction of genotype with environment , σ2 is the residual error , n was the number of environments , and r is the number of replications [68] . The best linear unbiased predictor ( BLUP ) for each genotype across two environments was predicted with the lme4 package in R ( http://www . R-project . org/ ) and used as the phenotypic input in the genome-wide association study ( GWAS ) . The 196 soybean accessions used in this study were genotyped using the NJAU 355K SoySNP Array ( S1 File ) . After excluding SNPs with a MAF < 0 . 05 , 207 , 608 SNPs were left . The mean values of phenotypic data for all genotypes in the NJ environment and NT environment were separately used to perform the GWAS . Meanwhile , the BLUP values across these two environments were also used to perform the marker-trait association analysis . The GWAS was performed using a compressed mixed linear model ( CMLM ) , which accounted for the complex population structure and familial relatedness [69] . For the CMLM analysis , we used the equation y = Wv+Xβ+Zu+e , where y is a vector of a phenotype; v and β are unknown fixed effects representing marker effects and population structure effects , respectively; and u is a vector for unknown random polygenic effects . W , X and Z are the incidence matrices for v , 0 and u , respectively , and e is a vector of random residual effects . All analyses were conducted with an R package called Genome Association and Prediction Integrated Tool ( GAPIT ) [70] . The population structure was accounted for by a principle component analysis ( PCA ) , and the first five principal components were included in the GWAS model . The kinship matrix was calculated by the VanRaden method [71] and used as the covariance structure of random polygenic effects . The threshold for significant association was set to 1/n ( n is the marker numbers , P < 4 . 82×10−6 ) [72] . The expression pattern analysis of GmMYB29 and IFS2 in response to reduced glutathione ( GSH ) and insect as well as in different soybean tissues was conducted in NJAU-C041 ( Jianliniumaohuang ) , which was randomly selected from the 196 soybean accessions . The soybean seedlings used for both GSH and insect induction expression analyses were grown in growth chambers under the conditions of 16/8 h ( day/night ) , 28/23°C ( day/night ) , and 70% relative humidity . Soybean Jianliniumaohuang used for the tissue expression analysis was grown under natural conditions in the field at Nanjing Agricultural University . To analyze the expression of GmMYB29 and IFS2 in response to GSH , six pieces of healthy and fully expanded upper-third leaves from different individual 30-day-old soybean plants were excised and immediately submerged in 100 mL of a GSH preparation ( 10 mM , pH 5 . 8 ) containing 0 . 005% Silwet to reduce leaf surface tension in each beaker flask ( 250 mL ) . Leaves from the same location in the control plants were also detached and submerged in a solution ( pH 5 . 8 ) without GSH . Both control and treated leaves were incubated in an oven-controlled crystal oscillator at 25°C in the dark with gentle shaking ( 100 rpm ) . Samples were collected by filtration at four sampling times ( 0 , 3 , 6 , and 7 h after incubation ) [29 , 33] . The expression of GmMYB29 and IFS2 in response to insects was analyzed by placing two third-instar common cutworm larvae of a uniform size on each fully expanded upper-third leaf of intact 30-day-old soybean seedlings in growth chambers . Control plants were not exposed to common cutworms . The damaged upper-third leaves of treated plants and undamaged leaves at the same location on control plants were excised at eight sampling times ( 0 , 1 , 2 , 4 , 6 , 8 , 12 and 24 h after induction ) for the identification of induced expression [73] . To analyze the expression of GmMYB29 and IFS2 in different soybean tissues , RNA samples were isolated from roots , stems , leaves , and flowers during the full-blossom period , pod walls on 10th day after flowering ( DAF ) , seeds at 10 , 20 , 25 , 30 , 40 and 50 DAF , and mature seeds . The expression level of GmMYB29 was also detected at 40 DAF in the seeds of a subset of 30 soybean accessions , representing varieties with high , medium and low isoflavonoid contents from the 196 accessions . All collected samples were placed in 2 mL cryopreservation tubes , immediately frozen in liquid nitrogen and stored at -70°C . A total of 100 mg of each sample was used for RNA isolation with the plant RNA Extract Kit ( TIANGEN , Beijing , China ) . cDNA was synthesized using a TaKaRa Primer Script™ RT reagent kit with gDNA Eraser according to the manufacturer’s instructions . Gene expression was determined by RT-PCR assays using an ABI 7500 system ( Applied Biosystems , Foster City , CA , USA ) with SYBR Green Realtime Master Mix ( Toyobo ) . The constitutively expressed gene Gmtublin ( GenBank accession number: AY907703 ) was used as a reference gene for RT-PCR . Three replicates were performed for each reaction , and the data were analyzed using the ABI 7500 Sequence Detection System ( SDS ) software version 1 . 4 . 0 . The normalized expression , reported as the fold change , was calculated for each sample as ΔΔCT = ( CT , Target-CT , Tubulin ) genotype- ( CT , Target-CT , Tubulin ) calibrator [74] . The primers used are listed in S4 Table . Glycine max var . Williams 82 was the first whole-genome sequenced soybean with the most complete genome information . To obtain the accurate GmMYB29 sequence , we cloned it from this cultivar . To determine the subcellular localization of GmMYB29 in soybean , GmMYB29 was amplified from the cDNA , including the 5’- and 3’-UTRs , of Williams 82 and cloned into the pAN580 vector containing a GFP expression cassette ( pAN580-GFP ) to generate the recombinant plasmid pAN580-MYB29-GFP . The recombinant plasmid and the empty control plasmid pAN580-GFP were introduced into onion epidermal cells by gene gun and Arabidopsis mesophyll protoplasts by polyethylene glycol ( PEG ) . Transgenic cells were analyzed by a laser scanning confocal microscope using a Zeiss LSM780 camera . GmMYB29 amplified from Glycine max var . Williams 82 was inserted into the BamHI-NotI-digested GFP-removed pAN580 vector to generate the effector vector CaMV 35S::MYB29 . The open reading frame ( ORF ) of luciferase ( LUC ) was cloned from the pGL3 vector ( XbaI-XmaI ) ( Promega , Madison , WI , USA ) and introduced into the GFP-loss pAN580 vector to produce the CaMV 35S::LUC plasmid . The CaMV 35S::LUC plasmid was digested by SacI and NheI to remove CaMV 35S , and then the promoter sequence of IFS2 or CHS8 , amplified from Glycine max var . Williams 82 , was inserted to form the reporter vectors IFS2pro::LUC and CHS8pro::LUC . A Renilla luciferase reporter was used as an internal control for normalization . The transient promoter activity in protoplasts derived from Arabidopsis suspension cells was analyzed by following the Dual Luciferase Assay protocol ( Promega ) . GmMYB29 was inserted into pBA002 with the CaMV 35S promoter to produce the pBA002-MYB29 overexpression vector . The RNAi vector was constructed using the Gateway technology with a Clonase II Kit ( Invitrogen , Carlsbad , CA ) . A specific 442-bp fragment of the GmMYB29 cDNA sequence was amplified from Williams 82 , and attB1 and attB2 adapters were added . Next , through the BP and LR reactions , we cloned the specific fragment into the pB7GWIWG2 ( II ) vector to obtain the pBI-MYB29Ri vector . As the soybean cultivar Jack is known for its high transformation efficiency , the soybean hairy root transformation was performed using this accession with the pBA002-MYB29 overexpression vector , the pBI-MYB29Ri vector , and the respective empty vectors as controls . The positive hairy roots detected by PCR from several independent transgenic lines were harvested separately and used for gene expression or isoflavone content analysis . Using the 5'- and 3'-UTR sequences of GmMYB29 , which shared relatively low sequence similarity with a paralogous gene , a pair of gene-specific primers ( GmMYB29-SF and GmMYB29-SR ) were designed ( Prime 5 . 0 ) to amplify GmMYB29 from 30 soybean genotypes ( S3 and S4 Tables ) . The primers used to sequence the GmMYB29 PCR products are listed in S4 Table . The sequences were assembled and aligned using ContigExpress in Vector NTI Advance 10 ( Invitrogen , Carlsbad , CA ) and MEGA version 6 [75] , respectively . Polymorphisms , including SNPs and indels , with a MAF ≥ 0 . 05 were identified among these genotypes , and their association with isoflavone content and pairwise LDs were calculated by Tassel 5 . 0 [27 , 76] . Markers were defined as being significantly associated with the phenotype on the basis of a significant association threshold ( -LogP > 1 . 30 , P < 0 . 05 ) . | Isoflavones are bioactive substances with various benefits , and increasing isoflavone content is one of the major aims of soybean quality improvement . Isoflavone biosynthesis is regulated by multiple genes and complex metabolic networks . The modification of certain structural genes in the isoflavone pathway by genetic engineering has been unable to significantly improve isoflavone content . Thus , the identification and application of transcription factors specific to the isoflavone pathway may effectively resolve this problem . Here , a genome-wide association study ( GWAS ) was used to identify an R2R3-type MYB transcription factor , GmMYB29 , associated with soybean isoflavone contents . Transient expression and plant transformation results confirmed that GmMYB29 positively regulates isoflavone biosynthesis in soybean . A candidate-gene association analysis identified 11 probable causative GmMYB29 polymorphisms . The identification and functional characterization of GmMYB29 not only improves our understanding of the genetic molecular mechanisms underlying isoflavone synthesis but also provides a direct target for both genetic engineering and selection for the improvement of isoflavone content in soybean . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"genome-wide",
"association",
"studies",
"gene",
"regulation",
"regulatory",
"proteins",
"brassica",
"dna-binding",
"proteins",
"genetic",
"mapping",
"model",
"organisms",
"genome",
"analysis",
"transcription",
"factors",
"crops",
"experimental",
"organism",
"systems",
"p... | 2017 | An R2R3-type MYB transcription factor, GmMYB29, regulates isoflavone biosynthesis in soybean |
The source of symmetry breaking in vertebrate oocytes is unknown . Animal—vegetal oocyte polarity is established by the Balbiani body ( Bb ) , a conserved structure found in all animals examined that contains an aggregate of specific mRNAs , proteins , and organelles . The Bb specifies the oocyte vegetal pole , which is key to forming the embryonic body axes as well as the germline in most vertebrates . How Bb formation is regulated and how its asymmetric position is established are unknown . Using quantitative image analysis , we trace oocyte symmetry breaking in zebrafish to a nuclear asymmetry at the onset of meiosis called the chromosomal bouquet . The bouquet is a universal feature of meiosis where all telomeres cluster to one pole on the nuclear envelope , facilitating chromosomal pairing and meiotic recombination . We show that Bb precursor components first localize with the centrosome to the cytoplasm adjacent to the telomere cluster of the bouquet . They then aggregate around the centrosome in a specialized nuclear cleft that we identified , assembling the early Bb . We show that the bouquet nuclear events and the cytoplasmic Bb precursor localization are mechanistically coordinated by microtubules . Thus the animal—vegetal axis of the oocyte is aligned to the nuclear axis of the bouquet . We further show that the symmetry breaking events lay upstream to the only known regulator of Bb formation , the Bucky ball protein . Our findings link two universal features of oogenesis , the Bb and the chromosomal bouquet , to oocyte polarization . We propose that a meiotic—vegetal center couples meiosis and oocyte patterning . Our findings reveal a novel mode of cellular polarization in meiotic cells whereby cellular and nuclear polarity are aligned . We further reveal that in zygotene nests , intercellular cytoplasmic bridges remain between oocytes and that the position of the cytoplasmic bridge coincides with the location of the centrosome meiotic—vegetal organizing center . These results suggest that centrosome positioning is set by the last mitotic oogonial division plane . Thus , oocytes are polarized in two steps: first , mitotic divisions preset the centrosome with no obvious polarization yet , then the meiotic—vegetal center forms at zygotene bouquet stages , when symmetry is , in effect , broken .
Cell polarity is essential to epithelial tissue formation and function in both development and homeostasis . Correct cellular polarization is required for proper asymmetric cell division of stem cells , as well as the appropriate segregation of cell fate determinants to daughter cells in the generation and maintenance of functioning tissues [1–3] . Aberrant polarization in stem cells , their daughter cells , and differentiated cells causes tissue defects and cancer [3 , 4] . Tracing the origins of cell polarity in many systems has , therefore , been of great biological and clinical interest . In most vertebrates , oocyte polarization along the animal—vegetal ( AV ) axis is key to establishing the embryonic body axes , as well as specifying the germline . First , the embryonic dorsoventral axis is established by dorsal determinants localized to the egg vegetal pole during oogenesis . Following fertilization , these vegetally-localized dorsal determinants then translocate via the Syntabulin linker , Kinesin motor , and cortical microtubules to the future dorsal side of the embryo , where they activate a Wnt signaling pathway [5–11] . Nuclear localization of β-catenin then establishes the dorsal organizer , generating the embryonic dorsoventral axis [8 , 12] . Importantly , the mRNAs of syntabulin , as well as of dorsal determinants like wnt11 and wnt8 in Xenopus and zebrafish , localize to the oocyte vegetal pole [7 , 9–11] . Disrupting their localization and/or function results in ventralized embryos that lack dorsal structures [7–9 , 11 , 12] . Secondly , germ cells form by inheriting vegetally-localized germ cell determinants , termed germ plasm , during cleavage stages . The germ plasm initially adheres to the cleavage furrows and then segregates to daughter cells during asymmetric cell divisions , inducing germline fate in specific blastomeres . Importantly , prior to localization to furrows , the germ plasm components are aggregated at the egg and oocyte vegetal pole [8 , 9 , 13–19] . Finally , the AV axis of the oocyte aligns with and determines the embryonic anterior—posterior axis . Vegetal pole formation in the oocyte thus provides crucial positional information to the future developing embryo . The oocyte vegetal pole is specified by the Balbiani body ( Bb , also called the “mitochondrial cloud” in Xenopus ) . The Bb is an aggregate of mRNA protein granules ( mRNP ) that include embryonic patterning factors and germ plasm , as well as organelles such as mitochondria [16 , 17 , 20] . Following Bb formation adjacent to the nucleus , it associates with the oocyte cortex , where it unloads its mRNPs to specify the vegetal pole of the oocyte [8 , 14–16 , 21 , 22] . Demonstrating the developmental importance of the Bb , most of the dorsal determinants and germ plasm components discussed above localize to the Bb and then the vegetal pole . Moreover , the initial Bb pathway is likely further required for a second wave of vegetal mRNA localization in oogenesis , termed the “late pathway” , which utilizes a specific microtubule population and Kinesin motors to localize additional factors to the vegetal pole [23 , 24] . Loss of the Bb impedes mRNA localization by this pathway as well [21] . Despite the importance and universal conservation of the Bb from insects to mammals [16 , 20 , 25 , 26] , its formation is poorly understood . In particular , it is unknown whether its cellular position is determined stochastically or by a prepattern . The source of oocyte AV polarization is , therefore , unknown . Since the Bb forms intimately associated with the nucleus , we postulated that the nucleus may determine the position of the Bb . Such a mechanism would require polarized cues from the nucleus . Intriguingly , a polarized nuclear configuration , called the chromosomal bouquet , arises preceding Bb formation . During the zygotene stage of meiosis I , the chromosomes orient as a bouquet with their telomeres clustered to one pole of the nuclear envelope ( NE ) , while the chromosome loops face the opposing side [27] . The bouquet is a universal feature of meiosis and promotes the synapsis of homologous chromosomes to facilitate recombination [27–33] . We hypothesized that the bouquet plays an additional role in oocyte polarity , whereby the nuclear axis of the bouquet predicts the oocyte AV axis . Using quantitative image analysis and markers to define the early meiotic oocyte stages in zebrafish , we characterized the previously inaccessible early meiotic stages and analyzed the chromosomal bouquet configuration in zebrafish . We traced the first asymmetric localization of Bb precursor components to the telomere cluster of the bouquet as early as the onset of meiosis . We show that these early symmetry-breaking events lay upstream to the function of Bucky ball ( Buc ) , the only known regulator of Bb formation . We further show in vivo that microtubules are required for both telomere clustering and Bb precursor localization during the zygotene bouquet , indicating that these cytoplasmic and nuclear events are mechanistically coordinated . In addition , we found that meiotic oocytes in zebrafish develop in nests with specialized cytoskeletal features , revealing a higher order organization of the nest . This organization indicates that centrosome positioning in meiotic oocytes is predetermined by a previous mitotic division . Overall , our data show that oocytes are polarized in two steps: first , mitotic divisions preset centrosome localization , but no other asymmetry is apparent; then , in the zygotene stage , symmetry is broken in a microtubule-dependent manner that clusters the telomeres of the bouquet and localizes Bb precursors to the region of the centrosome . To our knowledge , this is the first dissection of the early dynamics of oocyte symmetry breaking in a vertebrate , revealing that oocyte polarity aligns with a conserved nuclear polarity .
To address when oocyte polarity is generated , we examined the localization of three core Bb components at earlier stages than previously studied: dazl mRNA , Buc , and the typical Bb-concentrated mitochondria ( the Bb is also called the “mitochondrial cloud” in Xenopus ) [21] . As previously shown [21 , 34 , 35] , all three components localize to the mature Bb at the nuclear periphery of mid-diplotene stages ( Fig 1A; oocyte size of 50–70 μm in diameter ) . In earlier diplotene stages , we detected the three Bb components aggregated within an indentation of the NE ( Fig 1A; 35–40 μm oocyte ) . Even earlier , at the onset of the diplotene stage , the nucleus exhibited a unique highly concave face , forming what we term a “nuclear cleft” ( Fig 1A; 20 μm oocyte; S1 Video ) . At this stage all three Bb components were aggregated in the cytoplasm within the nuclear cleft ( Fig 1A ) , which was also evident even earlier at the pachytene stage ( Fig 1A; 15–19 μm oocyte ) . We developed an algorithm to measure the Bb precursor intensities in the cleft versus noncleft cytoplasm of pachytene to early diplotene oocytes ( “cleft analysis” , Materials and Methods ) . Our cleft analysis shows a robust enrichment of dazl ( ×6 . 5 ) , Buc ( ×5 . 0 ) , and DiOC6 ( ×1 . 8 ) in the cytoplasm of the nuclear cleft ( Fig 1B; S1–S3 Figs , S1 Data ) . At all stages shown , mAb414 detected both the NE ( fine line ) and perinuclear granules ( spherules ) . The concave shape of the NE during pachytene to diplotene stages was confirmed by LaminB1 labeling , images of live oocytes labeled with vital dyes , and ultrastructural analysis of whole ovaries ( Fig 1C , S2 Fig , S1–S3 Videos ) . We found that the nuclear cleft was most pronounced at the onset of diplotene and then gradually recedes to a typical spherical nucleus as the mature , spherical Bb forms ( Fig 1D ) . Interestingly , cleft stage nuclei lack detectable A-type lamins ( LamA/C ) that are evident at postcleft stages ( S2 Fig , S4 Video ) . Nuclear lamin composition controls the rigidity of the nucleus [36 , 37] . The specific absence of LamA/C from cleft stages may provide the nuclear flexibility that allows the cleft to form and suggests that this transient nuclear morphology is developmentally regulated . We found that Bb precursor components are polarized in an aggregate within a nuclear cleft at the onset of pachytene much earlier than previously known . We thus investigated if Bb precursors aggregate earlier in meiosis . However , early stages of zebrafish oogenesis are not well characterized , and meiosis was not addressed [38–40] . We therefore characterized the stages preceding pachytene , the leptotene and zygotene bouquet stages . We investigated the telomere dynamics characteristic of these early meiotic stages , which is conserved across species [27] , using a telomere FISH marker . We also examined centrosome localization , which functions in mouse spermatocytes to cluster the telomeres of the bouquet stage ( Fig 2 , S3 Fig ) [41] . We found that in premeiotic oogonia , telomeres are randomly distributed in nuclei but become loaded radially on the NE at the leptotene stage ( Fig 2A ) ; the centrosome at these stages is positioned perinuclearly ( Fig 2B ) . During zygotene stages , telomeres are tightly clustered on the NE , forming the bouquet configuration ( Fig 2A , S5 and S6 Videos ) . We found that the centrosome localized to the cytoplasm facing the telomere cluster at this stage ( Fig 2B ) . At the pachytene stage , telomeres redistributed radially on the NE then detached from the NE at the onset of diplotene ( Fig 2A ) . At these two stages , the centrosome localized roughly to the center of the clearly evident nuclear cleft ( Fig 2B ) . At later diplotene stages ( >25 μm ) , we could not detect the centrosome , consistent with its known loss during oogenesis [42 , 43] . During our analysis , we found that early meiotic zebrafish oocytes develop in nests ( Fig 2C ) . A germline cyst is defined as a cluster of germ cells , which are connected via cytoplasmic bridges that remain from previous incomplete mitotic divisions and are collectively surrounded by somatic follicle cells . A nest defines the same organization as that of the cyst , but where cytoplasmic bridges were not directly demonstrated . Development of germ cells within cysts is conserved from insects to mammals [44] . However , in zebrafish only premeiotic oogonia , not meiotic oocytes , were described in nests [21 , 38 , 40] . We observed zygotene bouquet stage cells tightly adjacent to each other , with no follicle cells apparent between them , although they were detectable around the zygotene cell groups ( Fig 2C , S7 Video ) . β-Catenin staining of cell membranes was similar in premeiotic oogonia , leptotene , and zygotene cells consistent with these cells residing within a nest ( Fig 2D ) . Electron micrograph ( EM ) images confirmed the presence of only two membranes , tightly juxtaposed to each other , between zygotene oocytes , ruling out the presence of follicle cells between oocytes ( S4A Fig , red arrowheads; n = 19 oocytes in 7 nests of two ovaries ) . While this organization may represent a cyst , we did not directly examine cytoplasmic bridges between oocytes and therefore define these as meiotic nests . Having identified the oogonia , leptotene , and zygotene stages , we then examined Bb precursor components during these oogenesis stages . We first addressed mitochondrial aggregation with DiOC6 labeling , which detects the Bb-enriched mitochondria ( Fig 1A–1C ) . While the DiOC6 signal was radially distributed in premeiotic oogonia , it was enriched in zygotene stages in the perinuclear cytoplasm apposing the telomere cluster ( “telomere cluster cytoplasm”; Fig 3A , S1 Data ) . Since DiOC6 does not detect mitochondria specifically , we examined mitochondrial localization ultrastructurally . Transmission electron microscopy ( TEM ) images clearly showed radially distributed mitochondria in oogonia ( Fig 3A ) . At zygotene stages , the contact of synapsed chromosomes ( SCs ) with the NE marks the presumptive telomere cluster ( Fig 3A , arrowheads ) , consistent with our zygotene confocal observations ( Fig 3A , top ) . Counting the number of mitochondria in high-power images of the entire cytoplasm 360° around the nucleus ( S4A Fig ) confirms the enrichment of mitochondria to the cytoplasm apposing the presumptive telomere cluster ( Fig 3A , S1 Data ) . We conclude that mitochondria are first aggregated in the telomere cluster cytoplasm at zygotene . To measure potential enrichment of Bb precursor components in the telomere cluster cytoplasm , we developed a quantitative , high-throughput MATLAB program we term “zygotene nest analysis” ( Materials and Methods ) . This algorithm is based on our observation that the centrosome is a reliable marker of the telomere cluster cytoplasm ( Fig 2B ) and tests for enrichment of other components in the cytoplasm surrounding it . We first measured the enrichment of DiOC6 labeling at the zygotene bouquet stage ( Fig 3A ) . Indeed , our nest analysis results showed a 1 . 6-fold DiOC6 enrichment to the telomere cluster cytoplasm , specifically at zygotene stages ( Fig 3B , S1 Data ) . We next investigated localization of the Buc protein , the only protein known to be required for Bb formation [21] . Buc localizes to the mature Bb [34] and to the pachytene Bb precursor aggregate ( Fig 1A and 1B ) . We detected Buc radially distributed in oogonia cytoplasm ( Fig 3B ) . However , at zygotene Buc is enriched in the centrosome-telomere cluster cytoplasm ( Fig 3B ) . Nest analysis reveals a 1 . 7-fold enrichment versus the remaining cytoplasm ( Fig 3B , S1 Data ) , specifically at zygotene . As a control , we measured intensities of a transgene-driving DCLK-GFP ( Tg ( Ef1α:Dclk-GFP ) ) , a microtubule binding protein [6] . Nest analysis showed no specific enrichment in either oogonia or zygotene oocytes ( S4B Fig , S1 Data ) . We investigated Buc localization relative to the centrosome in zygotene oocytes , colabeling cell membranes with β-Catenin . Indeed , Buc localized to the centrosome cytoplasmic region of zygotene oocytes ( 100% , n = 14 oocytes; Fig 3C ) . Therefore , Bb precursor components become specifically enriched in the telomere cluster cytoplasm at the zygotene stage . GasZ is another protein resident to the Bb in mid-to-late diplotene stages [21 , 45] ( S4D Fig ) . GasZ serves as a scaffold protein for the piRNA machinery in perinuclear nuage in mice testis [46] . We found that perinuclear granules in early zebrafish oocytes contain the Vasa , Ziwi , and Zili proteins colocalized with the nuclear pore complex marker mAb414 ( perinuclear spherules; S4C Fig ) , similar to Vasa ( GLH-1 ) P granule localization to nuclear pores in Caenorhabditis elegans [47 , 48] . Consistent with its localization in mice testes , we detected GasZ colocalized with mAb414 to these perinuclear granules radially around oogonia nuclei ( Fig 3D , S4D Fig ) . GasZ immunostaining was not compatible with the Telo-FISH protocol , preventing a direct test for colocalization in the bouquet . However , we found that mAb414 spherule immunostaining localized to the telomere cluster at the bouquet stage ( Fig 3D ) . We therefore used mAb414 spherules as a reference for GasZ localization in the bouquet . GasZ exhibited a polarized pattern in the bouquet , similar to that of mAb414 spherules ( Fig 3D ) . Pixel-wise statistical analysis in zygotene stages confirmed colocalization of GasZ and mAb414 signals ( average Pearson correlation coefficient = 0 . 5; Fig 3D , S1 Data; ~75% of each signal was colocalized with the other , Manders coefficient ) . GasZ bouquet localization was also confirmed by costaining for GasZ and DiOC6 . Therefore , the Bb precursor GasZ protein localizes to the telomere cluster cytoplasm of the bouquet . Consistent with GasZ and the mAb414 spherule dynamics , the piRNA processing enzymes , Ziwi and Zili , and the RNA binding protein in the piRNA pathway Vasa , exhibited the same transition from radial granules in oogonia to high localization to the telomere cluster apposing cytoplasm ( Fig 3D ) . While a role for such a polarized localization of perinuclear granules in the piRNA pathway is unclear , these granules might have been coopted to contribute components of similar functions , like GasZ , to the forming Bb . GasZ bridges different mRNA binding proteins within the nuage in mice [46 , 49] and could function similarly within the forming Bb . Having defined the mAb414 spherules as markers for the telomere cluster position at zygotene ( Fig 3D ) , we utilized it to further analyze individual zygotene oocytes for their Buc localization . We identified zygotene oocytes blind to the Buc channel , and scored oocytes with clear mAb414-clustered localization , typical of its pattern in zygotene bouquet . We then turned on the Buc channel and examined its distribution in reference to the mAb414 signal position . Consistent with our previous analyses , this approach confirmed the transition of Buc from a radial distribution in oogonia to enrichment at the cytoplasm apposing the telomere cluster , as visualized by the mAb414 cluster ( Fig 3E , S1 Data ) . We conclude that Bb precursors transition from radial distribution in premeiotic oogonia to being enriched in the cytoplasm facing the bouquet telomere cluster ( Fig 3F ) . This association of Bb precursors , the centrosome and the telomere cluster , unveils the first symmetry breaking in the zebrafish oocyte . At this stage , the nuclear axis of the bouquet predicts the cellular AV axis in the oocyte , with the telomere cluster position marking the future vegetal pole . Before our current findings , early steps of Bb formation were unknown . A functional pathway for Bb formation has , therefore , yet to be constructed . We tested the relationship between the zygotene bouquet configuration , localization of Bb precursors in the nuclear cleft , and Buc function , the only known protein required for Bb formation . We found that telomeres in buc-/- zygotene oocytes cluster normally and the centrosome localizes to the telomere cluster in 100% of the cells similar to wild-type ( Wt ) ( Fig 4A ) . Concomitant with bouquet formation in buc-/- oocytes , DiOC6 was also normally enriched in the telomere cluster cytoplasm similar to Wt ( Fig 4B ) . Hence , Buc is dispensable for bouquet formation and early asymmetry formation . We next examined Bb precursor enrichment in the nuclear cleft stages that follow the bouquet stage in buc mutants . Most buc-/- oocytes from pachytene through early diplotene ( 17–25 μm ) showed normal morphology , with a nuclear cleft enriched for DiOC6 signal ( Fig 4B ) . Only at late diplotene stages ( 25–45 μm ) were aberrant DiOC6 patterns detected; in most oocytes , DiOC6 was expanded completely radially , while in the remaining oocytes DiOC6 was expanded partly outside the nuclear cleft ( Fig 4B ) . As expected [21] , all mid-diplotene buc mutant oocytes ( ≥50 μm ) lack a Bb based on both DiOC6 and dazl mRNA localization ( Fig 4B ) . These dynamics show that while the mitochondrial aggregate disperses at late diplotene stages , its formation is intact through the zygotene—pachytene stages . Following its normal bouquet localization , the centrosome properly localizes consistently to the nuclear cleft in buc mutant oocytes through pachytene to early diplotene stages . Therefore , early polarization appears normal in buc-/- oocytes . These results altogether demonstrate that Buc is required for Bb formation downstream of the zygotene bouquet configuration and early asymmetry formation , and in parallel to or downstream of nuclear cleft formation . Interestingly , dazl transcript is not enriched in the cleft of buc-/- pachytene to early diplotene ( 17–25 μm ) oocytes ( Fig 4C , S1 Data ) . We performed quantitative real-time polymerase chain reaction ( QRT-PCR ) analysis to compare dazl levels between Wt and buc-/- ovaries . No decrease in dazl levels was detected , indicating that dazl remains present but is no longer localized in buc-/- oocytes , as reported previously [21 , 50] . Since Buc does not bind mRNA itself , but does interact with at least one RNA binding protein [34] , it likely functions through such interactions to localize dazl to the cleft . A role for Buc in actively forming the large RNP Bb granule is consistent with Buc as an intrinsically disordered protein ( IDP ) [50 , 51] . IDPs provide adhesive scaffolds in mRNPs , promoting their phase separation from the soluble cytoplasm into distinct bodies [50–54] . Buc might serve a similar adhesive role in the Bb precursor aggregate . In the absence of Buc , the zygotene—pachytene mitochondrial aggregate does form but then disperses and fails to mature into a Bb . While Buc is required for Bb formation , Buc position is set upstream by the centrosome telomere-cluster cytoplasm during the symmetry-breaking events of the zygotene stage . We investigated further the relationship between bouquet formation and early asymmetry of Bb components . During the chromosomal movements that generate the bouquet , telomeres attach to Sun/KASH domain proteins on the NE that connect them with microtubules in the cytoplasm [28 , 29 , 31 , 55] . Importantly , microtubules are required for bouquet formation and synapsis in mice spermatocytes and C . elegans oocytes [31 , 41] . Therefore , we characterized microtubule organization during early oogenesis as a possible link between chromosomal bouquet formation and Bb component aggregation . We used a transgenic microtubule reporter line , EMTB-3GFP ( Fig 5A ) [56] and costained for γTubulin to visualize the centrosome . We found rather loose radially symmetric perinuclear microtubules in premeiotic oogonia without significant enrichment around the centrosome ( Fig 5A ) . In zygotene stages , a denser network of microtubules appeared around the centrosome at the cytoplasmic site apposing the telomere cluster . At later zygotene stages the network appeared wider , concomitant with the wider nuclear distribution of telomeres ( Figs 5A and 2A ) . During pachytene and early diplotene stages , a microtubule meshwork was found in the nuclear cleft cytoplasm surrounding the centrosome ( Fig 5A ) . The organization of microtubules around the centrosomes during zygotene suggests that the centrosome functions as a microtubule organizing center ( MTOC ) at this stage and that the role of microtubules in bouquet formation is conserved in zebrafish . We also examined acetylated microtubules during these early meiotic stages and found a specialized cytoskeletal organization unique to the meiotic nest . We observed long cables of acetylated tubulin ( Fig 5B and 5C ) . While varying in length , each cable emanated from a centrosome of one cell ( Fig 5B , S8 Video ) . The acetylated cables were specific to leptotene and zygotene nests: they were absent from premeiotic oogonia and were first detected in leptotene oocytes , where they appeared shorter and fewer ( Fig 5B ) , suggesting that they grow and fully mature at zygotene . We never observed the cables in later stage oocytes . Costaining for Acetylated tubulin and telomeres showed that the cables extend from the position of the telomere cluster , consistent with their centrosome association ( S6 Fig ) . Furthermore , costaining with β-Catenin indicates that some of these cables extend between cells in the nest ( Fig 5C , S9 Video ) . This observation also suggests that the cells in the meiotic nest are connected and some of the acetylated cables may extend through cytoplasmic bridges between oocytes . Interestingly at the zygotene to pachytene transition , we detected microtubules extending significant distances between pairs of cells in what appeared to be cytoplasmic bridges connecting the microtubule meshworks that surround the centrosome of each cell ( Fig 5D ) . We never detected acetylated tubulin staining at the zygotene to pachytene transition when these interoocyte connections were apparent . Only zygotene to pachytene transitioning oocytes displayed these microtubule connections , and they were always located in the periphery of a nest ( Fig 5D ) . These cells may be completing cytokinesis and separating from the nest . Such a release from the nest at the zygotene to pachytene transition is consistent with that of the mouse ovary , where early meiosis occurs in nests and individual oocyte folliculization initiates at pachytene [57] . Intriguingly , the acetylated and then nonacetylated microtubules we observed here strikingly resemble the typical events of late cytokinesis , where midbody microtubules that extend in the cytoplasmic bridges are acetylated but then their deacetylation is required for cytokinesis completion [58–63] . This suggests that zygotene—pachytene-transitioning oocytes complete cytokinesis for their release from the nest . Importantly , the nonacetylated microtubules connect the centrosome regions of the two connected oocytes , indicating that the final mitotic division of oogonia may set the centrosome position at zygotene . Remarkably , the centrosome region at this stage already defines the future vegetal pole of these oocytes . This indicates that the last mitotic division of oogonia may predispose the bouquet configuration and polarization . While polarization is , in effect , established at zygotene , these early events may preset its axis . Our data show that oocyte symmetry breaking can be traced back to the zygotene bouquet stage , when Bb precursors transition from a radial distribution to localization at the centrosome cytoplasm apposing the telomere cluster of the bouquet configuration . We next tested this correlation functionally . Microtubules are key in generating the telomere movements that establish the chromosomal bouquet in mice spermatocytes and in C . elegans oocytes [31 , 41] . Our data show that the centrosome—microtubule—telomere bouquet cellular organization found in these organisms is conserved in zebrafish zygotene oocytes . Microtubules are , therefore , strong candidates to link the nuclear telomere movements with the cytoplasmic localization of Bb precursors . To examine microtubule function and test the functional correlation of Bb precursor localization and telomere clustering , we aimed to disrupt microtubules and monitor Bb precursors and telomeres simultaneously . We established a protocol to isolate whole ovaries , culture them for a short period , and treat them for 80 min with the microtubule depolymerizing drug nocodazole or dimethyl sulfoxide ( DMSO ) as a control . The two ovaries from each fish were split between the nocodazole and DMSO groups , with the latter serving as an internal control . Nocodazole depolymerized microtubules of all oocyte stages , including zygotene bouquet oocytes ( Fig 6A ) . Unexpectedly , we also found that nocodazole disrupted most of the acetylated tubulin cables of zygotene oocytes ( Fig 6B ) . We then examined Bb precursors and telomeres in zygotene oocytes where both populations of microtubules are absent . To monitor Bb precursors we used DiOC6 to mark Bb mitochondria , since it is the best Bb precursor marker that is compatible with the telomere FISH staining ( the FISH protocol requires harsh conditions , which are deleterious for most antibodies and epitopes ) . We found that in control DMSO-treated ovaries , zygotene oocytes showed normal clustering of telomeres and typical localization of DiOC6 in the apposing cytoplasm ( Fig 6C ) . In contrast , nocodazole-treated zygotene oocytes showed partial to complete radial dispersion of telomeres . Moreover , the same oocytes showed a concomitant expansion of DiOC6 ( Fig 6C ) . We never observed an oocyte where one but not the other was affected . Fig 6D schematizes these effects . We next quantified these results . To avoid misinterpretation of results , we avoided early and late zygotene stage oocytes where telomere clustering is still incomplete or starting to disperse , respectively . Midzygotene oocytes display condensed chromosomes , a single peripheral nucleolus , and are approximately 12 μm in diameter , ( 10–11 μm are early and 14–15 μm are late zygotene ) . To test our ability to identify midzygotene oocytes based on these criteria , we scored the cells blind to the telomere channel and only selected oocytes of 11 . 5–13 μm in diameter . We then turned on the telomere channel and analyzed telomere clustering and DiOC6 patterns . Using this method , we found that 93% of DMSO-treated oocytes showed normal telomere clustering and apposing DiOC6 localization ( Fig 6E , blue dots , n = 45 oocytes in 7 nests of 2 ovaries ) . However , following nocodazole treatment , 88 . 5% of oocytes showed broad expansion of both telomeres and DiOC6 , as in Fig 6C and 6E ( Fig 6E red dots , n = 71 oocytes in 11 nests of 4 ovaries ) . Fig 6F shows the pooled results ( S1 Data ) . These results demonstrate that microtubules are required for both telomere clustering and DiOC6 localization during zygotene bouquet , strongly arguing that they are coregulated and functionally linked . In our experimental settings , we disrupted both the dynamic microtubule population and the acetylated tubulin cables . We cannot distinguish if one or both populations are required for telomere clustering and DiOC6 localization . However , we observed cases where nests in nocodazole-treated ovaries still showed intact acetylated tubulin cables , but DiOC6 was still dispersed . Since perinuclear microtubules were always disrupted , we believe that they may be more directly involved in telomere clustering and DiOC6 localization , and while the acetylated cables may be required , they are not sufficient .
We trace the first asymmetry in the zebrafish oocyte to chromosomal bouquet formation at the onset of meiosis , much earlier than previously known . Asymmetry is established when Bb precursors transition from radial distribution in premeiotic oogonia to being enriched in the cytoplasm facing the bouquet telomere cluster . Telomere clustering in the bouquet promotes homologous chromosomes to pair , facilitating recombination [28–33 , 64] . We now report that Bb precursor components localize to the associated centrosome and telomere-cluster cytoplasm of the bouquet , facilitating Bb formation and vegetal pole specification ( Fig 7A ) . To our knowledge , our finding that the bouquet nuclear axis aligns with the oocyte AV axis provides the first evidence for the bouquet being linked to a process outside of meiosis . We propose that the bouquet association of the centrosome and the telomere cluster comprise a meiotic—vegetal center that couples meiotic genetic events with oocyte patterning . The chromosomal bouquet is a universally conserved meiotic feature discovered in 1900 [27] , and the Bb is a universally conserved oocyte feature discovered in 1845 [20] . For the first time to our knowledge , we link these two fundamental oocyte features in the establishment of oocyte polarity . In most vertebrates , oocyte polarization is crucial in setting up the body axes and specifying cell fates in embryonic development [8 , 9 , 13–17] . In Xenopus , where the mature Bb has been well described , early dynamics of Bb precursors is lacking . While mitochondrial aggregates were described in premeiotic oogonia , it is not clear whether these are directly related to the mitochondria aggregates within the later mature diplotene Bb [14] . The potential roles of the mammalian Bb and oocyte polarity in general are less clear . Nonetheless , a Bb does form in oocytes of all mammals including human [16 , 26] , comprising at least a transient polarized oocyte morphology . While the mouse Bb is detected at diplotene stages between P3–P7 , localization of Bb precursors at zygotene stages ( E14 . 5 ) was not addressed [26] . The wide conservation of the bouquet and the Bb argues for the conservation of the meiotic—vegetal center in generating polarity generally in oocytes . To our knowledge , we provide the first evidence that a polarized nuclear morphology predicts cellular polarity . During asymmetric cell division in mitotic cells , interactions between the centrosome and the chromosomes or the nucleus serve as a platform for a polarized segregation of cell fate determinants [1 , 2 , 65] . We show an analogous example in meiotic cells where polarity factors localize to the centrosome—telomere cluster cytoplasm in the zygotene bouquet stage . Importantly , before the bouquet stage , the centrosome is at a perinuclear position , but telomeres are not clustered on the NE and Bb components are radially distributed in the cytoplasm . Only when the centrosome associates with the telomeres in the bouquet stage do Bb precursor components begin to aggregate in the telomere cluster cytoplasm . This suggests a mechanistic connection between the centrosome and the clustered telomeres of the bouquet that is also required for the initiation of Bb component recruitment . In the bouquet stage of various species , telomeres are connected to the centrosome via NE-resident Sun-KASH domain proteins and cytoplasmic microtubules [28 , 29 , 31 , 32 , 55] . However , the molecular nature of the centrosome—telomere association in the zebrafish meiotic—vegetal center , as well as the functional interactions that recruit Bb components to establish polarity in any vertebrate have been unknown . We analyzed microtubule localization in zebrafish zygotene oocytes and found that the bouquet centrosome—microtubule—telomere cellular organization is conserved . Furthermore , we find that microtubules are simultaneously required for both the telomere clustering and the apposing localization of Bb precursors of the bouquet . In their absence , telomeres and Bb mitochondria are dispersed . Thus , bouquet microtubules provide a mechanistic link between the nuclear events of the bouquet and the cytoplasmic events of Bb formation . How microtubules localize Bb precursors is still an open question . Microtubules could localize Bb precursors either by movements that stir cytoplasm towards the centrosome or by serving as routes for dynein/kinesin motors that localize Bb precursors as cargo . The ( − ) end motor dynein is a more likely candidate since localization is towards the ( − ) end of microtubules at the centrosome . Intriguingly , dynein localizes to telomere-bound Sun/KASH proteins on the NE and is required for chromosome synapsis in C . elegans zygotene oocytes [31] . More experiments are required to distinguish between these possibilities in zebrafish . Previous studies in Xenopus found no requirement for microtubules in localizing mRNA to the Bb [66] . However , these experiments were performed at much later stages of oogenesis addressing localization to the mature Bb . Our experiments show that microtubules are required for the initial localization of Bb precursors during its early formation . We also find a microtubule network at pachytene—early diplotene around the centrosome within the nuclear cleft , where Bb precursors are enriched . It will be interesting to address the roles of microtubules in Bb aggregation and cleft formation during these stages . The tracing of oocyte symmetry breaking to the meiotic—vegetal center in turn raises the question of upstream regulation of the activation and positioning of its components , i . e . , the centrosome and the site of clustered telomeres . Germ cells , from insects to humans , develop in germline cysts or nests [44] . While only premeiotic oogonia were described in zebrafish cysts [38] , we found that leptotene and zygotene oocytes develop within nests as well , some of which are clearly still connected by cytoplasmic bridges at the zygotene to pachytene transition ( Fig 5B–5D ) . The synchronized development and polarization of oocytes within the nest provides an environment wherein a higher order organization of the nest could regulate their polarization . The nest could , for example , be organized by intrinsic positioning of the centrosome by preceding oriented cell divisions and/or by paracrine signaling from surrounding follicle cells . We found a novel cytoskeletal feature in the meiotic nest that argues for a higher order nest organization . We find cables of acetylated microtubules in leptotene to zygotene stages and then nonacetylated microtubules at the zygotene to pachytene transition that can extend between two oocytes in the nest , likely reflecting cytoplasmic bridges . This is strikingly similar to the typical events of late cytokinesis , where acetylated midbody microtubules in the cytoplasmic bridge are then deacetylated in order to complete cytokinesis [58–63] . This suggests that the two connected oocytes that we observe constitute daughter cells from the last mitotic division . Remarkably , these microtubule extensions connect the two cells via their centrosome cytoplasm , indicating that the last mitotic division plane may localize the centrosomes and the meiotic—vegetal center , and predispose the axis of the bouquet and oocyte polarization . While no asymmetry is observed in oogonia , polarization may be established in two stages . In this model , the last mitotic division presets the polarization axis , but then polarization is only active at zygotene , when Bb precursors first asymmetrically localize ( Fig 7B ) . At zygotene stage , the cables of acetylated tubulin associate with centrosomes of the bouquet ( Fig 5B and 5C , S6 Fig ) , where dynamic perinuclear microtubules also reside . The structure of the acetylated cables strikingly resembles that of the axoneme of primary cilia [67] . Similar to primary cilia , these structures in the nest may serve mechanical and/or signaling roles [67] . Such cables may be affixed to the midbody ( as discussed above ) and serve as anchors or transmit forces important in the dramatic chromosomal movements that generate the zygotene bouquet configuration [31 , 41] . Forward genetic screens have not identified the oocyte polarity regulators in early meiotic stages in vertebrates [68 , 69] , and the accessibility of these stages to in vivo and quantitative imaging techniques has been challenging . Our dissection of early oocyte polarization in zebrafish now enables a reverse genetic approach , utilizing the advances in CRISPR/Cas9 techniques . Our work redefines the zygotene bouquet as a global cellular organizer at the nexus of oocyte differentiation . The juvenile zebrafish ovary arises as an attractive model to study the intersection of the fundamental biological processes of cell polarity and spatiotemporal tissue organization , encompassing mitosis and meiosis in a key developmental context .
Ovaries were collected from 6–8 wk postfertilization ( wpf ) juvenile fish: TU wild type , bucp43btmb/p43btmb ( referred to as buc-/- ) [50] , Tg ( Ef1α:Dclk-GFP ) [6] , and Tg ( βAct:emtb-3GFP ) [56] . Fish had a standard length ( SL ) measured according to [70] and were consistently ~12–20 mm . To fix the ovaries for immunostaining , RNA-FISH , and DNA-FISH , fish were cut along the ventral midline and the lateral body wall was removed . The head and tail were removed and the trunk pieces , with the exposed abdomen containing the ovaries , were fixed in 4% PFA at 4°C overnight with nutation . Trunks were then washed in PBS , and ovaries were finely dissected in cold PBS . Ovaries were washed in PBS and then either stored in PBS at 4°C in the dark , or dehydrated and stored in 100% MeOH at −20°C in the dark . For microtubule staining , fixation was carried out using a microtubule stabilizing buffer ( MSB ) , including 3 . 7% formaldehyde , 0 . 25% glutaraldehyde , and 0 . 5 μM Taxol ( MSB-fix; [71] ) . Post-fix washes and dissections were done in MSB without fix and taxol , and ovaries were stored in MeOH at −20°C in the dark . Ovaries were washed 2 times for 5 min ( 2 x 5 min ) in PBT ( 0 . 3% Triton X-100 in 1 x PBS; if stored in MeOH , ovaries were gradually rehydrated first ) , then washed 2 x 20 min , 1 x 40 min , 1 x 20 min in PBT . Ovaries were blocked for 1 . 5–2 h in blocking solution ( 10% FBS in PBT ) at room temperature and then incubated with primary antibodies in blocking solution at 4°C overnight . Ovaries were washed 4 x 20 min in PBT and incubated with secondary antibodies in fresh blocking solution for 1 . 75 hr and were light protected from this step onward . Ovaries were washed 4 x 20 min in PBT and then incubated in PBT containing DAPI ( 1:1 , 000 , Molecular Probes ) , with or without DiOC6 ( 1:5 , 000 , Molecular Probes ) for 50 min and washed 2 x 5 min in PBT and 2 x 5 min in PBS . All steps were carried out with nutation . Ovaries were transferred into Vectashield ( with DAPI , Vector labs ) . Ovaries were finally mounted between two #1 . 5 coverslips using a 120 μm spacer ( Molecular Probes ) . Primary antibodies used were mAb414 ( 1:1 , 000 , Abcam ) , LamB1 ( 1:400 , Abcam ) , γTubulin ( 1:400 , Sigma-Aldrich ) , Buc ( 1:500 ) [34] , LamA/C ( ready for use , Progen Biotech ) , GFP ( 1:400; Molecular Probes ) , Vasa ( 1:5 , 000 ) [72] , Zili ( 1:100 ) [73] , Ziwi ( 1:250 ) [74] , GasZ ( 1:100 ) [45] , Acetylated tubulin ( 1:200; Sigma-Aldrich ) , β-Catenin ( 1:1 , 000; Sigma-Aldrich ) . Secondary antibodies used were anti-rabbit IgG , or anti-mouse IgG1 , or anti-mouse IgG2b , Alexa 488 , and Alexa 594 ( all 1:500 , Molecular Probes ) . RNA-FISH was performed using the DNA-HCR-FISH technique ( Molecular Instruments ) [75] , following the company protocol , except for the hybridization temperature that was optimized for 50°C . DNA-FISH for telomeric repeats ( Telo-FISH ) was performed using the PNA technique ( PNA-Bio ) following the company protocol . Hybridization buffer was 70% Formamide , 1 mM Tris pH 7 . 2 , 8 . 5% MgCl2 buffer ( 25 mM magnesium chloride , 9 mM citric acid , 82 mM sodium hydrogen phosphate , pH7 ) , 1x Blocking reagent in Maleic acid buffer ( 100 mM Maleic acid , pH7 . 5 ) , 0 . 1% Tween20 , 88 nM probe ( 5’-CCCTAACCCTAACCCTAA-3’ , Cy3- conjugated ) . For a combination of IHC with RNA-FISH or Telo-FISH , IHC was performed first . At the end of the IHC procedure , ovaries were washed an extra time for 30 min in PBT and fixed quickly in 4% PFA for 15–20 min at room temperature . Ovaries were washed 3 x 5 min in PBS . If FISH procedures did not start immediately , ovaries were stored overnight in PBS at 4°C . During IHC that preceded RNA-FISH , all blocking solutions also contained RNAasin ( 1:100 , Sigma-Aldreich ) , to protect against potential RNAases in the blocking serum . All transitions between SSC buffers and PBS were done gradually to protect tissue morphology . After staining was complete , DAPI ( +/- DiOC6 ) staining and mounting was performed as described above . Images were acquired on a Zeiss LSM 710 confocal microscope using a 40X lens . The acquisition setting was set between samples and experiments to: XY resolution = 1 , 104 x 1 , 104 pixels , 12-bit , 2x sampling averaging , pixel dwell time = 0 . 59 sec , zoom = 0 . 8X , pinhole adjusted to 1 . 1 μm of Z thickness , increments between images in stacks were 0 . 53 μm , laser power and gain were set in an antibody-dependent manner to 7%–11% and 400–650 , respectively , and below saturation condition . Acquired images were not manipulated in a nonlinear manner , and only contrast and brightness were adjusted . All figures were made using Adobe Photoshop CC 2014 . Measurements were performed using a custom MATLAB ( Version 8 . 2 . 0 . 701 R2013b 64-bit , MathWorks ) script . A pixel-wise quantitative test for colocalization of GasZ with mAb414 signals was performed using the Coloc 2 plug-in on Fiji . We calculated a point spread function ( PF ) of ~2 . 5–3 pixels on average for each channel and image , based on PSF = [0 . 8 x Exitation wave length ( nm ) / objective’s NA] / pixel size ( nm ) . Min and max thresholds were automatically set by the plug-in for each image . Pearson correlation coefficient ( R ) was calculated to be ~0 . 5 . R values represent: −1 < R < 0 , anticorrelation; R = 0 , no correlation; 0 < R < 1 , correlation . The plug-in then scrambles the pixels to generate a random image and test for the probability to yield the same Pearson coefficient from a random image . Scrambling was set to 100 iterations , generating 100 different random images from the tested image’s pixels . The probability of receiving Pearson coefficient of R = ~0 . 5 was p = 1 . The plug-in also yielded Manders coefficient , which represents the colocalized cohort of each signal ( here , for each GasZ and mAb414 was ~75% ) . Ovaries were dissected from juvenile fish ( 6–8 wpf , SL ~12–20 mm ) into fresh warm HL-15 media ( Hanks solution containing 60% L-15 ( no phenol red ) , Sigma-Aldrich , and 1:100 Glutamax , Invitrogen , warmed to 28°C ) . Ovaries were then embedded in 0 . 5% low-melt agarose in HL-15 on a glass-bottom dish , and covered with warm HL-15 . After the agarose polymerized , ovaries were incubated in HL-15 containing Hoechst ( 6 . 66 μM , Molecular Probes ) , DiOC6 ( 1:5 , 000 ) , and/or , Mitotracker ( 500 nM , Molecular Probes ) , for 2 hr at 28°C . Still images of live ovaries were acquired using the Zeiss LSM 710 confocal microscope and a 40X lens with the settings described above . Ovaries were dissected from WT or Tg ( βAct:emtb-3GFP ) juvenile fish ( 6–8 wpf , SL ~12–20 mm ) into fresh warm HL-15 media . Then HL-15 was replaced with HL-15 containing either 100 μM nocodazole ( EMD Millipore ) or an equivalent volume of DMSO . The two ovaries from each fish were split between the nocodazole and DMSO groups as an internal control . Ovaries were incubated for 80 min at 28°C in the dark . Ovaries were then washed twice with plain HL-15 , and fixed either with 4%PFA/PBS , or with MSB-fix as described above . Ovaries were stored and stained as described above . To determine these conditions , we performed a time-lapse live imaging of Tg ( βAct:emtb-3GFP ) ovaries monitoring microtubules . We tested several nocodazol and DMSO concentrations and determined the incubation time for our experiments by the earliest time point in the movies that showed efficient microtubule depolymerization . Ovaries were still healthy and viable for at least 8 h after the 80-min time point . DMSO-treated ovaries always showed normal intact microtubules . Ovaries were dissected from juvenile fish ( 6–8 wpf , SL ~12–20 mm ) and fixed as described above in 2% PFA and 2 . 5% gluteraldehyde . To maximally preserve the tissue morphology , samples were prepared using a high pressure freezing technique [76] . Ovaries were vitrified under high pressure liquid nitrogen in an Abra HPM010 . Ovaries were freeze substituted in 2% OsO4 , 0 . 1% U acetate , in 100% acetone at −90°C for 72 hr . After slowly warming to room temperature , the ovaries were infiltrated and embedded in EMBed-812 ( Electron Microscopy Sciences ) . Thin sections were taken , counterstained with uranyl acetate and lead citrate , and examined with a JEOL 1010 electron microscope fitted with a Hamamatsu digital camera , and using AMT Advantage image capture software . For EM images , the “levels” and “brightness/contrast” functions on Adobe Photoshop , were mildly adjusted to improve the image tones and make the image generally crisper . These adjustments did not affect the biological properties of the imaged cellular features . WT or buc-/- ovaries ( SL = 17–22 mm ) were dissected and immediately transferred into Trizol ( Life Technologies; four ovaries were pooled of each genotype ) and snapfrozen and stored at −80°C . Total RNA was extracted using Trizol and chloroform extractions , followed by DNAaseI reaction . RT on WT , buc-/- and ( −RT ) control ( from a WT RNA sample ) sample was performed using random hexamer primers and the Superscript kit ( Invitrogen ) , followed by RNAaseH reaction according to the kit instructions . QPCR was performed using SYBR Green Jumpstart Taq Ready Mix ( Sigma-Aldrich ) , in Step One Plus PCR machine ( Applied Biosystems ) . Reactions on WT , buc-/- and ( −RT ) , and no-template controls were performed in duplicates , and each reaction was repeated twice . dazl levels , normalized to βActin , were compared using the ΔΔCT method . All statistical analysis and data plotting was performed using the GraphPad Prism 6 software . Data sets were tested with two-tailed unpaired t test . p-values were: *<0 . 05 , **<0 . 01 , ****<0 . 0001 , ns = not significant ( >0 . 05 ) . | In most vertebrates , an early event in egg development involves the establishment of the so-called animal—vegetal axis; this sets up the embryonic body axes and contributes to germ-line specification , and therefore , is key to embryonic development . The animal—vegetal axis is established during oogenesis by the Balbiani body ( Bb ) , an aggregate of specific mRNAs , proteins , and mitochondria , which forms adjacent to the nucleus and ultimately defines one pole of the oocyte , the vegetal pole . Despite its universal conservation , how the Bb forms and how its position is determined is unknown . Here , we show that Bb formation is initiated at the onset of meiosis , and its position coincides with a previously known meiotic polarized nuclear configuration , the chromosomal bouquet , which gathers the chromosome ends , the telomeres , asymmetrically on the nuclear membrane to assist in homologous chromosome pairing . We reveal that a global cellular organizer functioning via microtubules generates the bouquet and aggregates the Bb precursors asymmetrically towards the centrosome . We determined that these events lie functionally upstream to the Bb regulator Bucky ball . Further upstream , we found that the centrosome appears prepositioned by an intercellular cytoplasmic bridge derived from the last presumptive cell division plane of the premeiotic oogonial cell . Thus , oocyte polarity and the chromosomal bouquet are linked through a common cellular polarization mechanism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2016 | Oocyte Polarization Is Coupled to the Chromosomal Bouquet, a Conserved Polarized Nuclear Configuration in Meiosis |
Eosinophilic meningitis ( angiostrongyliasis ) caused by Angiostrongylus cantonensis is emerging in mainland China . However , the distribution of A . cantonensis and its intermediate host snails , and the role of two invasive snail species in the emergence of angiostrongyliasis , are not well understood . A national survey pertaining to A . cantonensis was carried out using a grid sampling approach ( spatial resolution: 40×40 km ) . One village per grid cell was randomly selected from a 5% random sample of grid cells located in areas where the presence of the intermediate host snail Pomacea canaliculata had been predicted based on a degree-day model . Potential intermediate hosts of A . cantonensis were collected in the field , restaurants , markets and snail farms , and examined for infection . The infection prevalence among intermediate host snails was estimated , and the prevalence of A . cantonensis within P . canaliculata was displayed on a map , and predicted for non-sampled locations . It was confirmed that P . canaliculata and Achatina fulica were the predominant intermediate hosts of A . cantonensis in China , and these snails were found to be well established in 11 and six provinces , respectively . Infected snails of either species were found in seven provinces , closely matching the endemic area of A . cantonensis . Infected snails were also found in markets and restaurants . Two clusters of A . cantonensis–infected P . canaliculata were predicted in Fujian and Guangxi provinces . The first national survey in China revealed a wide distribution of A . cantonensis and two invasive snail species , indicating that a considerable number of people are at risk of angiostrongyliasis . Health education , rigorous food inspection and surveillance are all needed to prevent recurrent angiostrongyliasis outbreaks .
Eosinophilic meningitis , a potentially fatal disease caused by Angiostrongylus cantonensis , is considered an emerging infectious disease in mainland China [1] , [2] . The first human case of angiostrongyliasis in mainland China was reported in 1978 , and a few more cases were diagnosed until the mid-1990s . Subsequently , several outbreaks have been recorded [1] . The first major angiostrongyliasis outbreak , involving 65 patients , was documented from Wenzhou in Zhejiang province in 1997 [3] . The biggest outbreak in China thus far could be attributed to a freshwater snail , i . e . , Pomacea canaliculata , and took place in the capital Beijing in 2006 [4] . Of the 160 infected individuals involved in this outbreak , 100 were hospitalized [5] . This outbreak also demonstrated that angiostrongyliasis had moved beyond its traditional endemic areas located in the southeastern coastal regions of China . The parasite was first described by Chen based on worm specimens collected from pulmonary arteries of rats in Guangzhou ( Canton ) [6] and Dougherty proposed the name A . cantonensis in 1946 [7] . Adult A . cantonensis live in the pulmonary arteries of its definitive hosts , i . e . , rodents , especially rats , which pass infective first-stage larvae ( L1 ) in their feces . The life cycle also involves mollusks , harboring larval stages . In humans , larvae fail to mature , and hence humans and their excreta play no role in the transmission and direct dissemination of the parasite . Humans become infected by ingesting third-stage larvae ( L3 ) in raw or undercooked intermediate host mollusks ( e . g . , snails and slugs ) or paratenic hosts ( e . g . , freshwater prawns , crabs , frogs and fish ) [8]–[10] . Lettuce and vegetable juice have also been identified as sources of infection when contaminated with intermediate or paratenic hosts [11] , [12] . Due to the low host specificity of A . cantonensis it is difficult to control this parasite [1] . Two snail species , i . e . , Achatina fulica and P . canaliculata , are believed to be closely associated with angiostrongyliasis in China . These snails were imported into mainland China in 1931 [13] , [14] and 1981 [1] , [15] , respectively , and have rapidly extended their geographic ranges . Indeed , these two snails are now listed as invasive species by the Chinese government . In response to the recent angiostrongyliasis outbreak in Beijing that had received considerable national and international attention and mass-media coverage , the Ministry of Health ( MoH ) of China launched the first national survey on A . cantonensis . Here , we report the design and key findings of this survey . Moreover , predictions are made for the spatial distribution of A . cantonensis and its intermediate hosts . Finally , recommendations are offered for the prevention of angiostrongyliasis .
The project entitled “The first national survey on Angiostrongylus cantonensis in China” has been approved by the institutional ethics committee of the National Institute of Parasitic Diseases , Chinese Center for Disease Control and Prevention in Shanghai ( ref . no . 2006030101 ) . Animal experiments were carried out in adherence to institutional guidelines for animal husbandry . The first national survey pertaining to A . cantonensis and its definitive and intermediate hosts in mainland China was implemented in two phases over a 1-year period , i . e . , between September and November 2006 , and between March and October 2007 . Considering that the distribution of A . cantonensis is affected by several environmental and ecological factors , the potential distribution of the parasite was first determined . Temperature was selected as the main factor to predict the potential distribution of the parasite and two invasive snail species in China . Since revealing the distribution of both the parasite and two snail species implicated in its transmission was the main aim of this survey , the widest potential distribution of P . canaliculata , which has previously been regarded as the most important intermediate host , delineated the survey region . The potential range of P . canaliculata in China was predicted using a degree-day model based on temperature data obtained from 149 observing stations across China [16] . A grid with a spatial resolution of 40×40 km was laid over the predicted area , and approximately 5% of the grid cells were randomly selected for sample collection . In each survey grid cell , one village was randomly selected for subsequent field work . The geographic coordinates of the survey villages were recorded using a hand-held global positioning system ( GPS ) device ( GPSmap 70; Kansas , USA ) . Rats are the definitive hosts of A . cantonensis . Some insectivores also serve as suitable definitive hosts [17] , [18] . Therefore , rats ( e . g . , Rattus norvegicus ) and insectivores ( Soricidae , e . g . , Suncus murinus ) were trapped in fields and in residents' houses . All captured animals were euthanized and dissected to determine the presence of adult A . cantonensis in their hearts and lung arteries . Freshwater snails ( e . g . , P . canaliculata and Bellamya aeruginosa ) , terrestrial snails ( e . g . , A . fulica ) and certain terrestrial slugs were collected from the surroundings of the study villages , and from restaurants and markets in the capital town of the counties , and snail farms across the study area , and examined for the presence of A . cantonensis larvae . Up to 100 specimens of each species were collected at each study site . The intermediate hosts were artificially digested using routine procedures ( incubation in a solution containing 0 . 2% pepsin and 0 . 7% hydrochloric acid at 37°C for 2 h ) [19] . Additionally , for the examination of P . canaliculata , a recently developed method relying on specific lung tissue features of this species was employed [19] , [20] . In brief , the lungs were separated from the snail body and opened . The nodules containing A . cantonensis larvae were then directly observed under a microscope . Paratenic hosts were also collected from markets and restaurants , and examined for L3 using an artificial digestion method . A . cantonensis larvae were identified based on distinct morphological criteria described elsewhere [21] . For quality control purpose , larvae identified as A . cantonensis from approximately 10% of the foci where A . cantonensis was found to be endemic were intragastrically injected into Sprague-Dawley ( SD ) rats . The animals were then maintained in the laboratory to confirm the identity of adult worms . An area was considered A . cantonensis endemic if the parasite was detected in any kind of animals captured in the field . The geographic locations of these sampling sites were linked to an existing geographic information system ( GIS ) , using the software ArcGIS version 9 . 1 ( ESRI , USA ) . Subsequently , ordinary kriging , a statistical technique for spatial prediction [22] , was performed , and thus a smoothed risk map of the A . cantonensis infection prevalence in P . canaliculata was produced .
The first national survey pertaining to A . cantonensis in China was implemented in 164 counties belonging to 19 provinces . A detailed list of the surveyed locations is available from the corresponding author upon request . Various mollusks were collected , belonging to one of the three following groups: ( i ) freshwater snails , ( ii ) terrestrial snails , and ( iii ) terrestrial slugs . All collected specimens were deposited in the “Preservation Center for Parasite Specimens in China” ( http://www . psic . cn ) , and further details are available from this center upon request . Overall , 11 , 709 P . canaliculata were screened , 6 . 8% of which harbored L3 of A . cantonensis . The prevalence among the other freshwater snails ( a total of 7 , 287 specimens were examined ) was only 0 . 05% . Of 3 , 549 A . fulica examined , 13 . 4% were infected . The infection prevalence among the 1 , 421 other terrestrial snail specimens was only 0 . 3% . Finally , 5 , 370 terrestrial slugs were dissected , revealing an infection prevalence of 6 . 5% . Hence , the endemicity of A . cantonensis in mainland China is primarily attributable to P . canaliculata , A . fulica and terrestrial slugs ( Figure 1 ) . Of the 711 potential host animals trapped during the field surveys , 32 were found to be infected with A . cantonensis ( 31 R . norvegicus and one R . flavipectus; overall prevalence: 4 . 2% ) . None of the 46 insectivores ( Suncus murinus ) were infected . The 652 paratenic hosts collected during the survey included frogs , shrimps , crabs , toads and fish . No A . cantonensis was found in any of these animals . The prediction map of the A . cantonensis prevalence in P . canaliculata in China , using an ordinary kriging approach with a spherical model , highlighted two potential clusters with prevalences of 19–28% in Guangxi province and 28–40% in Fujian province , respectively ( Figure 2 ) . Figure 3 shows the current distribution of A . cantonensis at county level in China . The parasite was identified in 59 of the 164 surveyed counties ( 36 . 0% ) . Most of the A . cantonensis-endemic areas were defined by infections in P . canaliculata and/or A . fulica snails . Only in three counties infected rats were found , but the presence of the parasite in intermediate hosts could not be ascertained . Seven provinces in southeastern China ( i . e . , Hainan , Guangdong , Guangxi , Fujian , Jiangxi , Hunan , and Zhejiang ) were identified as A . cantonensis endemic . P . canaliculata was introduced in Zhongshan city , Guangdong province in 1981 [15] . As shown in Figure 4 , P . canaliculata is now well established in southern China in a band spanning northeast-southwest . A separate endemic area is located in the Sichuan basin . The snail now colonizes almost the entire Pearl River valley , the Southern River system and the Southeast River system . The snail has also been observed in mountainous areas at high elevations in Yunnan province . Moreover , P . canaliculata snails have crossed from the Pearl River valley into the Yangtze River valley , and already inhabit the southeast section of the latter . Figure 5 shows that A . fulica has a more focal distribution than P . canaliculata , although the former species had been introduced into China half a century earlier than the latter . At present , A . fulica is known to occur in the provinces of Guangdong , Hainan , and Guangxi , in the southern areas of Yunnan and Fujian provinces , and in one county of Guizhou province . Unlike P . canaliculata , A . fulica occurs only south of 25° N latitude , and does not appear to be associated with major river networks . P . canaliculata snails were found on markets and/or in restaurants in 21 counties , whereas A . fulica snails were detected in three counties only . Infected P . canaliculata and A . fulica were found in nine and two counties , respectively . Additionally , two native freshwater snail species , i . e . , B . aeruginosa and Cipangopaludina chinensis , were commonly found to be on sale in many markets and restaurants . C . chinensis is one of the key intermediate hosts of A . cantonensis in Taiwan and , in the current survey , infected specimens were detected in one restaurant in Fujian province . To our knowledge , this is the first report of A . cantonensis-infected C . chinensis from mainland China . On two markets in Guangdong and Guangxi provinces , infected B . aeruginosa snails were detected . Only two commercial snail farms for P . canaliculata ( located in Jiangsu and Jiangxi provinces ) and one for A . fulica ( in Zhejiang province ) were identified during this survey . None of the snails collected in these farms was infected with A . cantonensis .
Eosinophilic meningitis caused by A . cantonensis is endemic in Southeast Asia , Australia , the Pacific Islands and the Caribbean . To date , more than 2 , 800 human cases have been reported [23] . It had been suggested that the parasite was dispersed from East Asia to other regions in two important hosts , i . e . , rats ( definitive host ) and A . fulica ( intermediate host ) especially during World War II [24] . Today , the parasite is still expanding its range and the associated disease is emerging in some regions , particularly China [1] , [25]–[28] . The results of the first national survey on the distribution of A . cantonensis and its hosts in China reported here indicate that there is a need for strengthening food safety inspections and food-borne disease surveillance . Long-distance trade , biological invasion and animal migration are contributing to the emergence of new diseases and the re-emergence of diseases that have previously been controlled [29]–[32] . Angiostrongyliasis in mainland China is an example of such an emerging food-borne disease . Its spread can be linked to the introduction , farming and consumption of certain snail species . Extrapolating from recent observations , the incidence of angiostrongyliasis is likely to further increase in China , although the 2006 outbreak in Beijing triggered considerable attention and a change in attitudes toward this parasitic infection not only in the medical and research community , but also the general public . The results of the national survey can be summarized as follows . First , the A . cantonensis-endemic area is very wide , covering seven southern provinces . However , not a single snail or rat infected with A . cantonensis was found in Yunnan province . This observation comes as a surprise , since the parasite was first documented in Yunnan some 20 years ago [33] , and several outbreaks have occurred subsequently [34] , [35] , most recently in Dali ( early 2008 ) . Hence , Yunnan must clearly be considered a potentially endemic province . Second , several freshwater and terrestrial snail species were found on local markets and in restaurants , and A . cantonensis-infected P . canaliculata and A . fulica clearly destined for human consumption were recorded . This observation suggests that the transmission of A . cantonensis to humans is ongoing , and that the health education and awareness raising campaigns initiated after the 2006 outbreak in Beijing – targeting consumers , health personnel and officials alike – must be improved since they appeared to have failed yet to stop the sale and consumption of infected snails . It follows that the impact of the previous health education campaigns through mass media to change human behavior has probably been overestimated , because angiostrongyliasis outbreaks continued in Guangdong province in 2007 [36] and Yunnan province in early-2008 , involving six and 41 patients , respectively . Third , culturally-routed dietary habits of certain ethnic groups increase the risk of A . cantonensis infection . For example , the consumption of raw or undercooked freshwater snails is held responsible for the early-2008 angiostrongyliasis outbreak in Dali . As a direct consequence of the booming inland tourism in China , the interest in minority dishes is growing , and ethnic dining has become popular among tourists and wealthy urban residents alike . Travelers to endemic regions with a tradition of preparing snails for human consumption should be better informed about the risks associated with certain dishes , and food inspection and hygiene regulations need to be enforced . Fourth , among the different factors facilitating the spread and transmission of A . cantonensis in China , the two invasive mollusk species P . canaliculata and A . fulica , play a central role . A range of mollusks can serve as intermediate hosts of A . cantonensis and were examined during the national survey . The prevalence of A . cantonensis infection among P . canaliculata , A . fulica and terrestrial slugs was found to govern the endemicity of this parasite in China . However , terrestrial slugs had rarely been found to be associated with human angiostrongyliasis; the only exception being their occasional use in local traditional medicine [37]–[39] . Thus , P . canaliculata and A . fulica are probably responsible for most angiostrongyliasis cases in China . Both snails not only expand their range , but also frequently go on the table for human consumption . Interestingly , P . canaliculata and A . fulica have facilitated the spread of the endemic area of A . cantonensis rather than the introduction of a new pathogen . Man-made ecological transformations and climate change are important drivers of the spread of exotic species and their establishment in new areas [29] , [40] , [41] . The emergence of several infectious diseases has already been attributed to the invasion of efficient vectors or hosts involved in their life cycle [42] . These two invasive snail species impact on the endemicity and transmission of A . cantonensis in at least two ways . First , the invasion of these snails facilitates the establishment of the life cycle of the parasite and thus increases the chances for an exposure of native mollusks to A . cantonensis in existing endemic areas . Previous experiments indeed documented a superior susceptibility of these snails to A . cantonensis compared to native snails [43] . Second , these invasive snails accelerate the spread of A . cantonensis since they rapidly expand their range , resulting in the local establishment of the snail and – sometimes – the parasite life cycle in previously non-endemic areas . A . fulica was recorded for the first time in mainland China in 1931 [14] . It has been suggested that eggs of A . fulica were accidentally imported from Singapore with shipments of plants , and that an initial snail population became established in Xiamen ( Amoy ) [13] . These terrestrial snails are nocturnal and become active under high-humidity conditions [44] . The snails feed on plants and deposit their eggs in the soil nearby . This behavior facilitates their dispersal through long-distance transportation of pot plants [13] . Since their unintentional introduction , A . fulica spreads across southern China , probably facilitated by the rapid expansion of long-distance trade and an increasing demand for farmed plants going hand-in-hand with China's ongoing economic development . It has even been speculated that A . fulica invaded China more than once . For example , the snail populations in Yunnan province might derive from trade with Indochina ( Mekong basin ) rather than eastern China [45] . The public health significance of A . fulica in mainland China was only noted when a parasitologically-confirmed case of angiostrongyliasis found in 1984 could be linked to this snail [46] . However , the consumption of A . fulica snails is generally less popular than that of P . canaliculata in mainland China . The freshwater snail P . canaliculata was deliberately introduced into China for human consumption . The invasion process can be stratified into three stages , i . e . , ( i ) introduction , ( ii ) establishment , and ( iii ) spread [47] . It was first imported into Zhongshan city in Guangdong province approximately 30 years ago [1] . Subsequently , the snail was farmed in most southern provinces with commercial aims [15] . However , within a few years , the snail also became established outside due to abandoning of farms and deliberate release [15] . Currently , the snails have reached 30° N latitude and had been found as high as 1 , 960 m above sea level in Yunnan province . It is conceivable that the dense river networks in eastern and southern China contributed to the dispersal of this snail . The isolated snail population in the Sichuan basin has expanded freely in this area for about 20 years . The easternmost natural colonies were observed in Zhoushan in Zhejiang province , suggesting a line from Zhoushan to the Sichuan basin south of which climate conditions are suitable for the snails to thrive . This line might move further northward as a consequence of global warming [48] . The public health significance of P . canaliculata was emphasized by the first major angiostrongyliasis outbreak in Wenzhou in 1997 [3] . The results of the national survey presented here suggest a close relationship between the endemicity of A . cantonensis and the area where P . canaliculata breed two or even three generations per year [16] , suggesting that A . cantonensis largely depends on this freshwater snail for its expansion in China . Although P . canaliculata in the whole endemic area of A . cantonensis were found to be infected , point prevalences of infection are heterogeneous: two heavily endemic areas could be identified in the provinces of Fujian and Guangxi , respectively . The snail is indeed responsible for many sporadic cases recorded throughout Fujian province . However , it remains to be investigated why no angiostrongyliasis cases have been observed thus far in Guangxi province . Although both A . fulica and P . canaliculata appear to have contributed to the emergence of angiostrongyliasis in China , several characteristics of P . canaliculata suggest that this species is mainly responsible for the spread of A . cantonensis . This claim is supported as follows . First , the aquatic P . canaliculata probably spread along waterways , accelerated through flooding events . This might partly explain why P . canaliculata more rapidly expanded its range than A . fulica , which appears to largely depend on human-facilitated transport . Second , the area colonized by P . canaliculata also expands far beyond that of A . fulica despite a considerably longer presence of the latter in China . Third , the consumption of P . canaliculata is more popular than that of A . fulica . During the national survey , for example , P . canaliculata was on sale in 21 counties , while A . fulica was only noted in three counties . The national survey shed light on different important aspects regarding the distribution of A . cantonensis and its hosts in China . The results indicate a need for more pointed attention to this emerging threat through awareness-raising campaigns among the medical community , the establishment of a hospital-based sentinel surveillance system , improved community-based health education and strengthening of food safety inspection . A number of pressing research questions could also be identified . For example , the model for predicting the prevalence of A . cantonensis within P . canaliculata identified two high-prevalence clusters . However , the accuracy of this prediction has not been assessed since no ground truthing of the predictions has been made thus far . The small-scale distribution , the range of hosts and the host-parasite compatibility should also be investigated to deepen our understanding of the transmission dynamics . In conclusion , the first national survey revealed the distribution of A . cantonensis and two invasive snail species , i . e . , P . canaliculata and A . fulica , and the pivotal role of these invasive snails for the transmission of this parasite . The results of the survey also suggest that people are at risk of angiostrongyliasis through consumption of raw or undercooked snails infected with A . cantonensis that are found in many markets and restaurants . Continued health education , rigorous food inspection , and hospital-based surveillance are needed to prevent recurrent angiostrongyliasis outbreaks in China . | Eosinophilic meningitis is caused by the rat lungworm ( Angiostrongylus cantonensis ) . This parasite is endemic in Southeast Asia , Australia , the Caribbean , and on Pacific Islands . Moreover , the disease is emerging in mainland China , which might be related to the spread of two invasive snail species ( Achatina fulica and Pomacea canaliculata ) . Thus far , the biggest angiostrongyliasis outbreak in China occurred in 2006 in Beijing , involving 160 patients . However , detailed information about the national distribution of A . cantonensis and its intermediate hosts is still lacking , and the importance of the two invasive snail species for disease transmission is not well understood . Therefore , a national survey on the distribution of A . cantonensis and its intermediate hosts in China was carried out in 2006/2007 . It was found that A . fulica and P . canaliculata were implicated in most angiostrongyliasis outbreaks , and that the distribution of A . cantonensis closely matched that of these snails . The two invasive snail species facilitated the expansion of the parasite , thus probably leading to the emergence of angiostrongyliasis , a previously rare disease , in mainland China . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"ecology"
] | 2009 | Invasive Snails and an Emerging Infectious Disease: Results from the First National Survey on Angiostrongylus cantonensis in China |
Lentiviral Nef proteins have multiple functions and are important for viral pathogenesis . Recently , Nef proteins from many simian immunodefiency viruses were shown to antagonize a cellular antiviral protein , named Tetherin , that blocks release of viral particles from the cell surface . However , the mechanism by which Nef antagonizes Tetherin is unknown . Here , using related Nef proteins that differ in their ability to antagonize Tetherin , we identify three amino-acids in the C-terminal domain of Nef that are critical specifically for its ability to antagonize Tetherin . Additionally , divergent Nef proteins bind to the AP-2 clathrin adaptor complex , and we show that residues important for this interaction are required for Tetherin antagonism , downregulation of Tetherin from the cell surface and removal of Tetherin from sites of particle assembly . Accordingly , depletion of AP-2 using RNA interference impairs the ability of Nef to antagonize Tetherin , demonstrating that AP-2 recruitment is required for Nef proteins to counteract this antiviral protein .
Human and simian immunodeficiency viruses encode several small , so called ‘accessory’ , proteins that do not appear to be required for viral replication in most in vitro replication systems . Nevertheless , it has become apparent that several of these accessory proteins play important roles in antagonizing host proteins , known as restriction factors , that inhibit viral replication . Specifically , Vif antagonizes members of the APOBEC3 family of cytidine deaminases whereas Vpu and Nef antagonize Tetherin ( reviewed in [1] ) . There is also emerging evidence suggesting that Vpx might also antagonize yet unidentified host restriction factors [2]–[4] . Tetherin ( BST-2/CD317/HM1 . 24 ) is a cell surface membrane protein with an unusual topology , consisting of a short N-terminal cytoplasmic tail ( CT ) , a transmembrane domain ( TM ) , an extracellular coiled-coil and a glycophosphatidyl inositol anchor at the C-terminus [5]–[8] . This topology , rather than primary sequence appears key for Tetherin's ability to retain nascent mature viral particles at the cell membrane [9] . Indeed , an artificial Tetherin assembled from domains of heterologous proteins with no sequence homology to natural Tetherins is active [9] . Tetherin appears to work by inserting either of its membrane anchors into the lipid envelope of nascent virions . In so doing , it physically bridges the nascent virion and cellular plasma membranes thereby preventing virions from disseminating to infect other target cells [10] , [11] . Thus , the spectrum of activity of Tetherin proteins against enveloped viruses is broad and includes retroviruses , filoviruses , arenaviruses , rhabdoviruses and herpes viruses [12]–[17] . Perhaps because Tetherin targets the lipid envelope , an almost invariant component of the virion , to block particle release , divergent viruses evolved various strategies and proteins to counteract Tetherin . HIV-1 uses Vpu , a type-I transmembrane protein [10] , [11] , Ebola uses its envelope protein [13] and Kaposi's sarcoma-associated herpesvirus uses the viral RING-CH E3 ubiquitin ligase K5 [14] , [15] . Interestingly , even among relatively closely related primate lentiviruses , three different viral proteins ( Vpu , Nef and Env ) have assumed the function of Tetherin antagonism [10] , [11] , [18]–[21] . Vpu is encoded by a subset of primate lentiviruses including HIV-1 , its direct chimpanzee-derived ancestor ( SIVcpz ) and the SIVgsn/mus/mon lineage whose 3′ portion of the genome , including Vpu , shares a common origin with SIVcpz/HIV-1 [22] . Vpu proteins from HIV-1 and SIVgsn/mus/mon antagonize Tetherin proteins from their respective hosts [20] , [23] , [24] . Even though both its direct descendent ( HIV-1 ) and its ancestors ( the SIVgsn/mus/mon lineage ) use Vpu to antagonise Tetherin , SIVcpz instead employs Nef for this function . In fact with the exception of HIV-2 [19] , that uses the envelope glycoprotein , all other primate that lack Vpu and have been tested encode Nef proteins that can counteract Tetherin [18] , [20] , [21] , [24] . The diverse nature of Tetherin antagonists that have arisen in primate lentiviruses is a consequence of the diversity in Tetherin sequence among primates , particularly at the target sites for Vpu and Nef , that the ancestors of modern viruses encountered as they were transmitted from species to species . Tetherin sequence diversity also means that the antagonists often exhibit species-specific activity . For example , SIVMAC Nef antagonizes macaque ( mac ) Tetherin but is less active against African Green monkey ( agm ) Tetherin and inactive against human ( hu ) Tetherin , whereas SIVagmSab Nef antagonizes both macaque and agmTetherin but is not active against huTetherin [21] . A key determinant of sensitivity to Nef proteins is a five amino acid motif in the Tetherin CT [18] , [21]; huTetherin is unique amongst primate Tetherins in that it lacks these five amino acids , and is resistant to all Nef proteins studied to date [20] . Nef is a 27–35 kD protein that is composed of ( i ) an N-terminal myristoyl membrane anchor , ( ii ) a flexible polypeptide chain that varies in length among Nef proteins , ( iii ) a polyproline helix type II that mediates interaction with SH3 domains ( iv ) a core domain that assumes a globular structure and is generally conserved among Nef proteins ( iv ) a C-terminal domain of unknown structure ( reviewed in [25] ) . Nef proteins have been shown to downregulate several cell surface molecules , exploiting distinct cellular protein partners that interact with distinct Nef sequences for downregulation of various targets ( [26] , [27] , reviewed in [28] ) . Thus , Nef recruits AP-1 to target MHC-I from the trans-Golgi to the lysosomes [29] , [30] and induces MHC-I endocytosis from the plasma membrane [31] . Conversely , Nef promotes endocytosis of CD4 from the cell surface and a critical dileucine motif in Nef was shown to be required for both this activity and Nef binding to AP complexes [32] , [33] , [34] , [35] . It was subsequently demonstrated that HIV-1 Nef employs two distinct motifs to form a complex with AP-2: an EXXXLL motif interacts with the AP-2 α-σ2 hemicomplex and a DD motif mediates additional interactions with AP-2 α and thus recruits AP-2 to mediate CD4 downregulation [36] , [37] . Notably , both AP-2 interaction motifs are conserved in Nef proteins from a variety of SIVs . The protein domains , cellular partners and mechanisms employed by Nef proteins to counteract Tetherin are unknown . Here we identify three key amino acid residues within SIVcpz Nef C-terminal flexible loop that are specifically required for Tetherin antagonism . We also demonstrate that Tetherin antagonism by Nef proteins from divergent SIVs is accompanied by the removal of Tetherin from sites of particle assembly and reduction of its levels at the cell surface , without effects on overall expression levels . We show that the two motifs in the Nef flexible C-terminal region that mediate interaction with AP-2 are also critical for these activities and , concordantly that AP-2 is required for Nef to antagonize Tetherin .
We and others have previously shown that five amino acids in the Tetherin CT are necessary and sufficient to confer sensitivity to antagonism by Nef [18] , [21] . To determine regions of Nef that are specifically required for Tetherin antagonism , we selected two Nef proteins that are closely related but differ in their ability to antagonize Tetherin . Specifically , SIVcpzGb1 Nef is a potent antagonist of cpzTetherin while HIV-1 Nef is completely inactive , even though the two proteins share ∼70% amino acid homology ( Figure 1A ) . We generated a series of chimeric SIVcpz/HIV-1 Nef proteins , depicted in Figures 1A and 1B , and assayed their ability to antagonize cpzTetherin when co-expressed , in trans , with an HIV-1 construct lacking Vpu and Nef . Chimeric Nef protein expression was confirmed using western blot analyses with an antibody that recognizes both proteins ( Figure 1C and 1D ) . Replacing increasingly larger portions of the N-terminal portion of the HIV-1 Nef protein by the corresponding regions from SIVcpzGb1 Nef gave rise to chimeras HIV/Gb-1 through -5 . HIV/Gb-1 , -2 , -3 and -4 , containing up to 147 N-terminal residues from SIVcpzGb1 Nef had the same phenotype as HIV-1 Nef in that they were unable to antagonize cpzTetherin . However , inclusion of the N-terminal 181 residues from SIVcpzGb1 Nef allowed the resulting HIV/Gb-5 chimera to antagonize cpzTetherin as efficiently as wild-type SIVcpzGb1 Nef ( Figure 1C and Figure S1A ) . In a reciprocal set of chimeras , HIV/Gb-6 to HIV/Gb-10 , SIVcpzGb1 Nef N-terminal sequences were replaced by those from HIV-1 Nef ( Figure 1B ) . Chimeras HIV/Gb-6 , -7 , -8 and -9 that contained up to 147 N-terminal residues from HIV-1 Nef maintained cpzTetherin antagonism activity . However , replacement of 181 N-terminal residues by those from HIV-1 Nef , in the HIV/Gb-10 chimera , abolished its activity ( Figure 1C and Figure S1A ) . There was only slight variation in the expression of these chimeric Nef proteins that did not correlate with their ability to counteract cpzTetherin ( Figure 1C and 1D ) and Nef expression levels were not affected by Tetherin expression ( Figure S1B ) . These data suggested that amino acids 148 to 181 of SIVcpzGb1 Nef contained key determinants of Tetherin antagonism . Thus , as expected , replacement of amino acids 148–181 in HIV-1 Nef by the corresponding residues from SIVcpzGb1 Nef resulted in chimera HIV/Gb-11 ( Figures 1A and 1B ) that was capable of antagonizing cpzTetherin restriction slightly less efficiently than wild-type SIVcpzGb1 Nef ( Figure 1D and Figure S1C ) . The reciprocal chimera HIV/Gb-12 , that contained HIV-1 Nef residues 148–181 , in an otherwise SIVcpzGb1 background was unable to antagonize cpzTetherin . Subsequently , smaller portions of HIV-1 Nef within this region were substituted by the corresponding residues from SIVcpzGb1 Nef , generating chimeras HIV/Gb-13 , -14 and -15 ( Figure 1A and 1B ) . Chimera HIV/Gb-15 was unable to counteract cpzTetherin . In contrast both chimeras HIV/Gb-13 and -14 that encoded SIVcpzGab1 Nef residues 148–158 and 159–167 , respectively , were able to antagonize cpzTetherin ( Figure 1D and Figure S1C ) . Compared to wild-type SIVcpzGb1 Nef , the activity of these chimeras was somewhat reduced ( Figure 1D and Figure S1C ) , suggesting that determinants within both regions 148–158 and 159–167 contribute to cpzTetherin antagonism . These chimeras were all expressed at comparable levels ( Figure 1D and Figure S1D ) . Importantly , all of the aforementioned Nef chimeras maintained CD4 downregulation activity ( Figure S1E ) , even though some chimeras exchanged Nef regions that have been shown to be required for CD4 downregulation [36] , [37] . This result demonstrates that Nef functions other than Tetherin antagonism were not grossly affected in the chimeras . CpzTetherin antagonism is a conserved property of Nef proteins from SIVcpz strains and absent from all the HIV-1 M-group strains studied thus far [20] . In amino acid region 148–158 several amino acids differed between the HIV-1 and SIVcpz Nef proteins used here , but only one of these ( K/Q at position 152 ) segregated according to whether the Nef proteins previously tested were from HIV-1 group M or SIVcpz strains ( Figure 1A and [20] ) . In region 159–167 , amino acids at positions 162 and 163 differed between HIV-1 Nef and SIVcpzGb1 Nef ( Figure 1A ) , although their identity did not segregate perfectly between HIV-1 and SIVcpz Nef proteins [20] . Based on this analysis , we generated three mutant HIV-1 Nef proteins: 1 ) HIV-1Q , harboring one amino acid change: K152Q , 2 ) HIV-1NC , harboring two amino acid changes: T162N and S163C and 3 ) HIV-1QNC , where all three substitutions K152Q , T162N and S163C were made . Both the HIV-1Q and HIV-1NC mutants antagonized cpzTetherin , albeit with reduced efficiency compared to SIVcpzGb1 Nef ( Figure 2 and Figure S2A ) . The combination of all three amino acid changes , in HIV-1QNC did not increase the efficiency of cpzTetherin antagonism over the HIV-1Q and HIV-1NC mutants suggesting that other residues within amino acid stretches 148–167 also contribute to anti-Tetherin activity . Protein expression levels were comparable for all mutant and wild type proteins ( Figure 2 and Figure S2B ) . In a reciprocal panel of SIVcpzGb1 mutants , introduction of the single mutation Q152K in SIVcpzGb1 Nef ( Gb1-K ) did not significantly affect its ability to inhibit cpzTetherin whereas a double mutation N162T and C163S ( Gb1-TS ) resulted in a very modest ( 2-fold ) reduction in activity ( Figure 2 and Figure S2A ) . However , when all three amino acid changes were combined , in Gb1-KTS , the resulting Nef protein was significantly impaired in its ability to antagonize cpzTetherin . The Gb1-KTS Nef protein retained its ability to downregulate CD4 although it was slightly less efficient than the parental and the other mutant Nef proteins ( Figure S2C ) . These data indicate that residues Q152 , N162 and C163 specifically affect the ability of SIVcpzGb1 Nef to counteract cpzTetherin , although other residues in the 148–167 domain also contribute to this activity . We have previously shown that a mutation in SIVMAC Nef ( D204R ) that inhibits CD4 downregulation also reduces its ability to counteract rhTetherin while a mutation that abolishes MHC-I downregulation ( Y223F ) does not affect Tetherin antagonism [21] . Because MHC-I downregulation by Nef requires AP-1 [30] while CD4 downregulation is mediated through AP-2 recruitment [36] , it seemed plausible that Nef might recruit AP-2 to remove Tetherin from the cell surface and rescue virus release . Two motifs in the HIV-1 Nef C-terminal flexible loop are required for interaction with AP-2 , namely EXXXLL165 and DD175 ( Figure 1A ) [36] , [37] . SIVMAC Nef proteins bearing mutations at the corresponding motifs ( EEHYLM195 to AEHYAA and D204 to R ) were expressed as fusion proteins with the fluorescent protein Venus at the C-terminus . The wild type SIVMACNef-Venus fusion protein efficiently antagonized rhTetherin , but as we have previously shown [21] , the D204R mutation abolished this activity ( Figure 3A and Figure S3A ) . Similarly , mutation of the leucine-based motif ( EXXXLM195/AXXXAA ) in SIVMAC Nef that mediates interaction with the AP-2 α-σ2 hemicomplex [36] also abolished rhTetherin antagonism . To determine potential roles in Tetherin antagonism of a number of SIVMAC Nef residues , previously reported to mediate interaction with several cellular proteins , we tested Nef proteins that were mutated at the following positions: 1 ) Y28 ( to A ) that has been suggested to affect interactions with AP-1 and AP-2 proteins [38] , [39] , 2 ) W70 ( to R ) that corresponds to residue W57 in HIV-1 Nef shown to be involved in CD4 binding [40] , 3 ) D155 ( to G ) that affects Nef dimerization [41] and 4 ) R138R139 ( to AA ) that affects Pak1/2 binding [42] . None of the above mutations had major effects ( <2–3-fold ) on particle release in experiments where SIVMACNef was asked to antagonize rhTetherin ( Figure 3A and Figure S3A ) . These data suggest that mutations that impair SIVMAC Nef interactions with other proteins do not generally affect its ability to antagonize rhTetherin . The notable exceptions were mutations that disrupt the leucine-based and diacidic motifs involved in AP-2 binding . These two AP-2-binding motifs are highly conserved among lentiviral Nef proteins and we therefore asked whether their mutation in other Nef proteins from SIVagm and SIVcpzGb1 that antagonize Tetherin , would also affect their activity . Indeed , mutation of either AP-2-binding motif in SIVagmSab Nef , EXXXLL183 or DD193 , significantly reduced the ability of SIVagmSab Nef to rescue particle release from inhibition by agmTetherin ( Figure 3B and Figure S3B ) . Similarly , mutation of the DXXXLL165 motif in SIVcpzGb1 Nef completely abolished its ability to counteract cpzTetherin ( Figure 3C and Figure S3C ) . AP-2 has been previously shown to interact with HIV-1 and SIVMAC Nef [36] , [38] , but not with SIVagmSab or SIVcpzGb1 Nef . We used a previously described yeast-3-hybrid assay to test for such an interaction [36] , [37] . The Nef proteins were fused to the Gal4 binding domain in a vector also expressing the AP-2 σ2 subunit and tested for interaction with a vector expressing the AP-2 α subunit and the VP16 activation domain . Like HIV-1 Nef , SIVagmSab Nef interacted with AP-2 α-σ2 hemicomplex , and mutation of either the EXXXLL183 motif or DD193 abolished this interaction ( Figure 3D ) . Unfortunately , in our hands both SIVMAC and SIVcpzGb1 Nef proteins induced activation of transcription even in the absence of AP-2 subunits and thus could not be tested in this assay . Therefore , we used an alternative assay to determine whether the Nef-Venus fusion proteins co-localized with endogenous AP-2 in mammalian cells . Indeed all three proteins SIVMAC , SIVagmSab and SIVcpzGb1 Nef showed a clear and obvious co-localization with AP-2 at the cell surface ( Figure S4 ) . Altogether , this data suggested that Nef residues that are important for interaction with AP-2 were necessary for Nef proteins to efficiently antagonize Tetherin . Since Nef recruits AP-2 to downregulate CD4 from the cell surface we determined whether Nef could downregulate Tetherin and whether the AP-2 binding sites were required for this function . Because we wished to analyze Tetherins from various species , but only antibodies that recognize the extracellular domain of mouse ( mo ) Tetherin that work well in flow cytometric assays are commercially available , we generated a panel of chimeric Tetherins . These contained the extracellular domain of moTetherin linked to the transmembane ( TM ) and CT domains from ( i ) huTetherin , ( ii ) a modified human Tetherin , termed hu ( GDIWK ) , in which 5 amino acids that confer sensitivity to antagonism by SIVMAC and SIVagm Nef [21] were reintroduced into the huTetherin CT and ( iii ) cpzTetherin ( Figure S5A ) . Notably the structure of moTeherin and huTetherin extracellular domains are nearly super-imposable [6]–[8] and clones of 293T cells stably expressing these Tetherins exhibited the expected phenotype in that they inhibited release of HIV-1 virions lacking Vpu and Nef , but expression of the appropriate Nef protein rescued particle release ( Figure S5B and C ) . The various Tetherin-expressing cell lines were transduced with HIV-1 based vectors expressing Nef-IRES-GFP cassettes and Tetherin cell surface expression was assessed by FACS using an antibody that recognizes the extracellular region of moTetherin . HIV-1 Vpu was used as a control in place of Nef; because it targets the TM domain of Tetherin which is well conserved between cpzTetherin and huTetherin , and caused downregulation of all the chimeric Tetherin proteins used ( Figure 4A and 4B ) . In agreement with previously published data [18] , we observed modest downregulation of the hu ( GDIWK ) but not the huTetherin chimera by SIVMAC Nef ( Figure 4A and Figure S6A ) . Notably , mutations of either the EXXXLM195 or D204 residues in SIVMAC Nef abolished this activity . The effects of SIVagmSab Nef on cell surface hu ( GDIWK ) Tetherin expression were more pronounced ( Figures 4A and Figure S6A ) and again downregulation was abolished by mutations in either the EXXXLL183 or DD193 motifs ( Figures 4A and Figure S6A ) . Similarly , SIVcpzGb1 Nef significantly decreased cell surface expression of the cpzTetherin chimera and mutation of the DXXXLL165 motif abolished this ability ( Figures 4B and Figure S6B ) . Importantly , the effects on Tetherin downregulation were specific , because none of the Nef proteins affected cell surface expression of the huTetherin chimera ( Figure 4A and 4B and Figure S6A and S6B ) . Thus the AP-2 binding side in three widely divergent Nef proteins was required for Tetherin downregulation from the cell surface . To determine whether Nef affects Tetherin endocytosis , cells expressing either chimeric hu-moTetherin or cpz-moTetherin proteins were transduced with vectors expressing no Nef or wild type SIVcpzGb1 Nef or the SIVcpzGb1 Nef DXXXLL165 ( AXXXAA ) mutant . Cells were incubated in the cold with a fluorescently labeled anti-mouse Tetherin antibody , washed , and then shifted to 37°C . The fraction of Tetherin that was internalized at various times after the temperature shift was then determined based on the amount of fluorescence that became resistant to an acid wash . These data suggested that a fraction ( about half ) of the Tetherin was rapidly and constitutively endocytosed within a few minutes , while the remainder was present at the cell surface for >40 minutes ( Figure S7 ) . Wild type SIVcpzGb1 Nef but not the DXXXLL165 mutant increased the fraction of cpz-moTetherin that was rapidly internalized within 10 min of the temperature shift to ∼100% but did not affect hu-moTetherin endocytosis ( Figure S7 ) . This data suggests that Nef downregulates Tetherin from the cell surface by increasing the amount of Tetherin that undergoes rapid endocytosis , to include the majority of cell surface molecules and that this activity is dependent on the Nef AP-2 binding site . Alternatively , this data is also compatible with the hypothesis that a pool of Tetherin is constitutively rapidly internalized and recycled to the cell surface , and that Nef interferes with this cycle by causing its entrapment at intracellular locations . In the presence of huTetherin and absence of HIV-1 Vpu , fluorescently labeled nascent HIV-1 particles are seen trapped on the cell surface and colocalizing with huTetherin ( [12] and unpublished data ) . To determine whether Nef proteins inhibited colocalization of Tetherin with nascent virions , YFP-labelled HIV-1 particles were generated in cells stably expressing HA-tagged rh- or cpzTetherin in the absence or presence of wild-type and mutant Nef proteins , and surface Tetherin protein revealed by immunostaining of non-permeabilized cells . In absence of Nef , nascent HIV-1 particles exhibited strong co-localization with both rhTetherin and cpzTetherin ( Figure 5A–D ) . Coexpression of HIV-1 Nef had no effect on this colocalization , but SIVMAC and SIVagmSab Nef proteins significantly inhibited the colocalization between viral particles and rhTetherin ( Figure 5A and 5B ) . Similar , but more pronounced effects were seen with SIVcpzGb1 Nef , which disrupted the co-localization between HIV-1 particles and cpzTetherin ( Figure 5C and 5D ) . For both SIVMAC and SIVcpzGb1 Nef proteins , mutation of the E/DXXXLM/L motif abolished their ability to remove their respective Tetherin targets from sites of particle assembly ( Figure 5A–D ) . Thus , Nef proteins removed Tetherins from sites of particle assembly and this activity was dependent on motifs critical for Nef-AP-2 interaction . To demonstrate that AP-2 is important for Tetherin antagonism by Nef , we used siRNA based approaches to deplete the α subunit of AP-2 , since Nef binds directly to this subunit [36] , [37] , [43] . Transfection of 293T cells with AP-2 α-specific siRNAs reduced AP-2 α subunit expression levels by ∼70–80% ( Figure 6A and 6B ) . AP-2 α depletion did not affect particle release in the absence of Tetherin , nor did it affect the ability of Tetherin to inhibit particle release in the absence of Nef . However , the ability of SIVcpzGb1 Nef to rescue particle release from inhibition by cpzTetherin was greatly attenuated when AP-2 α was depleted ( Figure 6A and Figure S8A ) . Similarly , AP-2 α depletion resulted in nearly complete loss of the ability of SIVMAC Nef to antagonize rhTetherin ( Figure 6B and Figure S8B ) . Thus , endogenously expressed AP-2 α is required for Nef proteins to inhibit Tetherin . Curiously , Tetherin proteins themselves contain a YXXV/M/I motif in their CT that could mediate binding to AP-2 , specifically the μ2 subunit [44] . Indeed rat Tetherin has been shown to bind to μ2 and to be subject to AP-2 mediated endocytosis [45] . huTetherin has also been reported to undergo AP-2-dependent endocytosis , and two tyrosine residues ( Y6 and Y8 ) in the huTetherin CT were found to be important for AP-2 α subunit binding [46] . To determine whether these residues were important for Tetherin activity , or for sensitivity to Nef , we mutated both tyrosines in cpz- and rh-Tetherin . Increasing amounts of unmodified or mutant cpzTetherin were co-expressed with a Vpu-defective HIV-1 virus that expressed either no Nef or SIVcpzGb1 Nef . In the absence of Nef , viral particle release was inhibited to similar extents by both the mutant and the unmodified Tetherin proteins ( Figure 7C ) . Similar results were obtained when increasing amounts of unmodified or mutant rhTetherin were co-expressed with SIVMAC or SIVMACΔNef ( Figure 7D ) . Importantly , for both cpzTetherin/SIVcpzGb1 Nef and rhTetherin/SIVMAC Nef , the presence or absence of the reported AP-2 binding site in the Tetherin CT did not affect sensitivity to antagonism by the corresponding Nef protein ( Figure 7C and 7D ) . Furthermore , cpzTetherin downregulation by SIVcpzGb1 Nef was also unaffected by mutation of the cpzTetherin AP-2 binding site ( Figure S9 ) . Thus , interactions between tyrosine motif in the Tetherin CT and AP-2 are not required for Tetherin activity , or sensitivity to Nef proteins . Rather , interactions between Nef and AP-2 are key for its Tetherin antagonist function .
Here , we show that Nef antagonizes Tetherin by decreasing its levels at the cell surface and particularly at sites of particle assembly . We also show that the ability of Nef to antagonize Tetherin is lost when residues that are important for AP-2 binding are mutated , or when AP-2 α expression is reduced using RNA interference . These results strongly suggest that the interaction of Nef with AP-2 is required for the removal of Tetherin from sites of virion assembly and antagonism of its antiviral function . Additionally , we identified three amino acids at the C-terminal loop of SIVcpzGb1 Nef , that are key determinants of its ability to antagonize Tetherin . However , our data suggest that other residues within the C-terminal loop and particularly within amino acids 148 to 167 also contribute to Tetherin antagonism . Together , these findings suggest a model in which Nef interacts with Tetherin and simultaneously with AP-2 via the Nef C-terminal loop and that the formation of this complex impairs the ability of Tetherin to be incorporated into virions , because it becomes sequestered by Nef and AP-2 away from sites of particle release , and either internalized or trapped at intracellular locations more efficiently than it would otherwise be . Interestingly , the decreased levels of Tetherin at the cell surface in the presence of Nef do not lead to an obvious reduction in the total Tetherin protein levels in the cell ( Figure 3 and Figure 4 and data not shown ) . This data suggests that in contrast to CD4 [32] , Tetherin internalization or intracellular retention by Nef-AP-2 does not lead to Tetherin degradation . The reciprocal specificity associated with the sequence requirements in both Nef and Tetherin for antagonism are most consistent with the notion that the two proteins directly interact with each other . Unfortunately , we were not able to demonstrate a specific physical interaction between Nef and Tetherin using in vitro binding assays with recombinant proteins , yeast two-hybrid assays , or co-immunoprecipitation experiments in mammalian cells . This suggests that the putative Nef-Tetherin interaction ( if it occurs ) is of low affinity or unstable outside the confines of a cell membrane . Interestingly , the Nef C-terminal flexible loop , in which key determinants of Tetherin antagonism reside , also mediates interaction with the AP-2 complex . In fact , two key amino acids identified in SIVcpzGb1 Nef as important for Tetherin antagonism ( T162 and C163 ) are embedded within the D/EXXXLL motif that is critical for interaction with the AP-2 α subunit ( specifically within the sequence DNNCLL ) . Mutations at these variable positions in the otherwise very well conserved D/EXXXLL motif do not grossly affect the ability of Nef to downregulate CD4 ( Figure S1 ) , indicating that AP-2 binding and Tetherin antagonism are separable activities . Nonetheless , the AP-2 binding site and residues required for Tetherin antagonism are in close physical proximity . It is also noteworthy that the cytoplasmic tails of primate Tetherins contain a classical sorting signal recognized by the AP-2 μ subunit [44] . Indeed , rat Tetherin has been shown to interact with the AP-2 μ subunit [45] but huTetherin endocytosis has been reported to be dependent on the AP-2 α subunit and to be dependent on two tyrosine residues , one of which is part of the YXXV motif [46] . Importantly , however , although Tetherin may interact directly with AP-2 , our findings indicate that this interaction is not required for virus restriction nor for sensitivity to Nef . This is in contrast with a recent report that demonstrates that the tyrosine residues in rhTetherin are critical for antagonism by an adapted SIVmac239 Env protein that causes rhTetherin downregulation , although the role of AP-2 in this activity was not determined [47] . Given the close proximity of amino acids putatively involved in Tetherin recognition to the AP-2 binding site , it might appear difficult to envisage how Nef might simultaneously bind to Tetherin and AP-2 . It is important to note however , that residues interspersed with those comprising the AP-2 binding site ( within the D/EXXXLL motif ) in Nef likely form only part of the determinant for Tetherin recognition . Indeed , mutation of the two residues alone within the D/EXXXLL motif that are important for Tetherin antagonism , is not sufficient to abolish the Tetherin antagonizing activity of SIVcpzGb1 Nef . Additionally , introduction of the three residues Q152/N162/C163 ( that include N162/C163 within the DXXXLL motif ) identified as critical for Tetherin antagonism in SIVcpzGb1 Nef to HIV-1 Nef was not sufficient to confer anti-Tetherin activity at the levels obtained with wild type SIVcpzGb1 Nef . Interestingly , residues N162/C163 are naturally found in a number of HIV-1 Nef proteins , including in Nef from strains JR-CSF and YU-10x that have been previously shown to have no activity against cpzTetherin [20] . Therefore , the context within which Nef residues are mutated determines whether they confer anti-Tetherin activity and several different residues within the C-terminal flexible loop of Nef contribute to Tetherin antagonism . It is possible that Nef forms multiple ( perhaps individually weak ) contacts with Tetherin , only one of which is mediated by an overlapping motif with the Nef-AP-2 binding site . One possible model , that is consistent with our findings , is that binding of the flexible C-terminal loop of Nef to AP-2 results in the formation of the site on Nef that binds to Tetherin . It is even possible that both AP-2 and Nef residues contribute to the formation of the Tetherin binding site . Alternatively , it has been reported that Nef can form dimers [41] , so it is possible that one molecule in a Nef dimer interacts with Tetherin and the other with AP-2 . However , arguing against this notion is our finding that mutation of a residue predicted to be required for SIVMAC Nef dimerization did not affect its ability to counteract rhTetherin ( Figure 3A ) . Finally , we cannot exclude the possibility that other cellular proteins , including other AP complexes , act as a bridge between the Nef-AP-2 complex and Tetherin , although this possibility seems unlikely due to the sequence specificity associated with the Nef-Tetherin antagonism . Of the known accessory proteins encoded by lentiviruses , none has been ascribed more functions than Nef ( reviewed in [28] ) . Nef is critical for efficient replication in vivo [48]–[50] and at least some of its reported functions , e . g . CD4 and MHC-I downregulation , are likely to play a significant role in pathogenesis . While it remains to be proven that Tetherin antagonism is required for efficient replication and dissemination of enveloped viruses in vivo , the finding that primate lentiviruses have evolved diverse strategies to counteract this antiviral protein suggests that it is . The identification of Nef domains and sequences that are required for Tetherin antagonism but not for other Nef activities could potentially allow the design of experiments to determine the importance of this function in viral pathogenesis in vivo . We note that Nef antagonizes Tetherin via a mechanism that is related to that which it employs to downregulate CD4 , namely recruitment of the AP-2 complex . Thus , it appears that Nef can parasitize the AP-2 complex for multiple ends . In a sense , the acquisition of Tetherin antagonism activity appears to be the result of the modification of an existing activity , namely CD4 downregulation , by broadening the array of target proteins to which AP-2 is recruited to include Tetherin . A similar concept might apply in the evolution of Vpu function , where ( again ) CD4 downregulation and Tetherin antagonism may proceed by similar mechanisms , with the viral accessory protein acting to recruit cellular factors ( in that case β-TRCP [51]–[53] ) to remove two different cellular molecules . Thus , lentiviruses can exploit Nef , and perhaps other accessory gene products , as somewhat plastic adaptors to recruit a given protein complex , in this case AP-2 , to diverse targets so as to manipulate host cells to provide a more permissive environment for virus replication and dissemination .
Plasmids expressing wild-type HIV-1 and SIVcpz Gb1 Nef proteins have been previously described [20] . Chimeric proteins were made by overlap-extension PCR using external primers introducing XbaI and MluI sites at the 5′end and 3′end of the Nef coding sequence , respectively and internal primers . PCR products were cloned into pCG-IRESGFP [21] . The HIV-1-based proviral plasmids , pBRHIV-1NL4-3ΔVpu , lacking Vpu and expressing wild-type versions of Nef proteins from HIV-1 , SIVcpz Gab , SIVMAC239 , or SIVagmSab Nef in cis , have been previously described [20] . Mutations in the Nef coding sequences of these plasmids were generated using overlap-extension PCR and BamHI ( in the Env coding region ) and MluI ( at the 3′-end of the Nef coding region ) sites in pBRHIV-1NL4-3ΔVpu plasmid . The same strategy was used to introduce wild type and mutant Nef proteins into a pBRHIV-1NL4-3ΔVpuΔNef plamids bearing YFP embedded in sequences encoding the stalk region of matrix ( MA ) . The pV1-derived plasmids encoding full length SIVMAC23 or SIVMAC239ΔNef proviruses have been previously described [21] . pCR3 . 1 plasmids were used to express wild-type and mutant SIVMAC239 and SIVagmSab Nef proteins fused to Venus at their C-terminus in trans . To express Nef proteins in 293T cell lines stably expressing Tetherin , Nef-IRES-GFP containing regions from pCG derivatives were transferred using SnaBI-NotI fragments ( encompassing the CMV promoter-Nef-IRES-GFP cassette ) into pCCGW , a HIV-based retroviral vector derived from pHRSIN-CSGW [54] by replacing the SSFV promoter with that from CMV . Plasmids expressing HA-tagged Tetherin proteins were constructed as previously described [12] , [55] . Mutations of the tyrosine residues in the CT of rh- and cpzTetherin were introduced by overlap-extension PCR . Chimeric hu-mo or hu-GDIWK-mo or cpz-mo Tetherin proteins , were generated by overlap-extension PCR , using hu , hu-GDIWK [21] or cpzTetherin as template for the CT and TM domains and moTetherin as template for the extracellular domain . PCR products were inserted into a retroviral vector pLHCX ( Clontech ) . All cloned coding sequences were verified by DNA sequencing . Oligonucleotide sequences are available upon request . 293T and TZMbl cells were maintained under standard conditions . Transfection protocols have been previously described [21] . Several 293T-derived cell lines stably expressing HA-tagged rh- or cpzTetherin and chimeric hu-mo , hu-GDIWK-mo and cpz-mo Tetherin were derived by transduction of 293T cells with the corresponding retroviral plasmids , followed by selection in hygromycin ( 5 µg/ml ) . To determine the ability of Nef proteins to counteract Tetherin , cells were transfected with 100-200 ng of pCG-IRES-GFP derived plasmids expressing various Nef proteins , 400–500 ng pBRHIV-1NL4-3ΔVpuΔNef or SIVMACΔNef and 20–25 ng of pCR3 . 1 Tetherin-HA expression plasmids or empty expression vector . To test the activity of tyrosine mutant rh- and cpzTetherin proteins , transfections were performed as above but with increasing amounts ( 0 ng , 5 ng , 10 ng , 20 ng , 40 ng , 80 ng/well ) of each Tetherin-HA expression plasmid . The total amount of DNA was held constant by supplementing the transfection with empty expression vector . To quantify Tetherin downregulation by Nef , viral stocks were generated by transfecting 5μg of pCCGW plasmids expressing various Nef proteins or HIV-1 Vpu , 5μg of an HIV-1 Gag-Pol expression plasmid and 1μg of VSV-G expression plasmid in 293T cells ( 10 cm dishes ) and used to inoculate cells stably expressing chimeric moTetherin proteins . At 48 h post-infection , cells were detached from plates with 5 mM EDTA in PBS and stained for cell surface Tetherin expression with anti-moTetherin antibody conjugated to APC ( anti-mPDCA-1-APC , Miltenyi Biotec ) . The amount of cell-associated APC and GFP fluorescence was measured with an LSRII flow cytometer ( BD ) . Experiments were performed as previously described [21] . Immunoblots were probed with the following antibodies: rabbit anti-HA antibody ( Rockland ) or mouse anti-AP2 α ( α-Adaptin 1/2 , Santa Cruz Biotech . ) , mouse anti-HIV-1-p24CA ( 183-H12-5C ) , mouse anti-HIV-1-Nef ( 1539 ) or mouse anti-SIVMAC-Nef ( 17 . 2 ) ( 2659 ) ( from the NIH AIDS Research and Reagents Program ) , followed by anti-rabbit or anti-mouse antibodies conjugated to IRDye680 or IRDye800 CW and scanned with an Odyssey Infrared Imager ( LICOR ) . Cells stably expressing HA-tagged cpzTetherin or rhTetherin were seeded on 3 . 5-cm , glass-bottomed dishes coated with poly-L-Lysine ( Mattek ) . Cells were transfected with 150 ng of an HIV-1 proviral plasmid ( pBRHIV-1NL4-3ΔVpuΔNef ) that was engineered to express various Nef proteins and 150 ng of an identical construct that expressed YFP embedded within the MA domain of Gag ( pBRHIV-1NL4-3ΔVpuΔNef-MA ( YFP ) ) using PEI ( PolySciences ) . At 48 h post-transfection , cells were fixed with 4% paraformaldehyde and incubated with mouse anti-HA . 11 monoclonal antibody ( Covance ) followed by anti-mouse IgG Alexafluor-594 conjugate ( Molecular Probes ) . A Z-series of images was captured from the top apical half of the cells using an Olympus IX70-based Deltavision microscope and were then deconvolved with SoftWorx software ( Applied Precision ) . For each cell , colocalization of Tetherin with virions was measured by tracing regions of interest in each Z-slice so as to analyze only the cell surface , and the Pearson's correlation coefficient for colocalization for each individual cell was calculated using SoftWorx software . For the Nef-AP-2 co-localization assays , 293T cells , seeded on 3 . 5-cm glass-bottomed dishes coated with poly-L-Lysine ( Mattek ) , were transfected with 100 ng of plasmids expressing Nef proteins fused to Venus at their C-terminus . At 48 h post-transfection , cells were fixed with 4% paraformaldehyde , permeabilized with 0 . 1% Triton and incubated with mouse anti-AP-2 α momoclonal antibody ( Santa Cruz Biotech . α-Adaptin 1/2 ) followed by anti-mouse IgG Alexa Fluor-594 conjugate ( Molecular Probes ) . A Z-series of images was acquired using Deltavision microscope , and colocalization between Nef and AP-2 was inspected using images of the cell surface acquired at the cell coverslip interface . The internalization of tetherin from cell surface was determined by flow cytometry assay as described previously for other cell surface markers [56] , [57] . Briefly , cells stably expressing chimeric moTetherin proteins were transduced with pCCGW Nef-IRES-GFP vectors plasmids . At 40 h post-infection , 107 cells were incubated with saturating amounts of anti-moTetherin antibody conjugated to APC ( anti-mPDCA-1-APC , Miltenyi Biotec ) in DMEM containing 0 . 5% bovine serum albumin on ice for 30 min . Following removal of excess unbound antibody , aliquots of 106 cells were shifted to 37°C for various periods of time . Then , each sample was split into two aliquots , which were diluted with DMEM under neutral ( pH 7 . 4 ) or acid ( pH 2 ) conditions and incubated for 1 min on ice . Samples were then washed and fluorescence measured with an LSRII flow cytometer ( BD ) . SIVagmSab Nef proteins were fused to the GAL4 DNA-binding domain in the pBridge vector ( Clontech ) also expressing the σ2 subunit of AP-2 ( a kind gift of Juan Bonifacino ) . The α subunit of rat AP-2 was cloned into pVP16/HA [58] . The protocol used subsequently has been previously described [36] , [37] except that Saccharomyces cerevisiae strain Y190 was used and transformed using the Gietz lab kit ( Molecular Research Reagents Inc . ) . | Primate lentiviruses express several small proteins which antagonize cellular proteins that inhibit virus replication . One such viral protein , Nef , has recently been shown to antagonize the cellular protein Tetherin that prevents newly formed viral particles from leaving the surface of infected cells . In this study we reveal the mechanism by which Nef overcomes inhibition by Tetherin . We show that three amino acids in the Nef C-terminal flexible loop are important for Tetherin antagonism . We also show that the interaction between Nef and AP-2 adaptor complexes is important for Tetherin downregulation from the cell surface , removal from sites of particle assembly and antagonism . Thus , our study demonstrates that AP-2 is important for the ability of Nef to antagonize Tetherin . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2011 | SIV Nef Proteins Recruit the AP-2 Complex to Antagonize Tetherin and Facilitate Virion Release |
The piRNA pathway is a surveillance system that guarantees oogenesis and adult fertility in a range of animal species . The pathway is centered on PIWI clade Argonaute proteins and the associated small non-coding RNAs termed piRNAs . In this study , we set to investigate the evolutionary conservation of the piRNA pathway in the hemimetabolous insect Rhodnius prolixus . Our transcriptome profiling reveals that core components of the pathway are expressed during previtellogenic stages of oogenesis . Rhodnius’ genome harbors four putative piwi orthologs . We show that Rp-piwi2 , Rp-piwi3 and Rp-ago3 , but not Rp-piwi1 transcripts are produced in the germline tissues and maternally deposited in the mature eggs . Consistent with a role in Rhodnius oogenesis , parental RNAi against the Rp-piwi2 , Rp-piwi3 and Rp-ago3 results in severe egg laying and female adult fertility defects . Furthermore , we show that the reduction of the Rp-piwi2 levels by parental RNAi disrupts oogenesis by causing a dramatic loss of trophocytes , egg chamber degeneration and oogenesis arrest . Intriguingly , the putative Rp-Piwi2 protein features a polyglutamine tract at its N-terminal region , which is conserved in PIWI proteins encoded in the genome of other Triatomine species . Together with R . prolixus , these hematophagous insects are primary vectors of the Chagas disease . Thus , our data shed more light on the evolution of the piRNA pathway and provide a framework for the development of new control strategies for Chagas disease insect vectors .
PIWI-clade Argonaute proteins have been implicated in a range of cellular and developmental events by regulating gene expression and imposing transposon silencing [1–3] . These proteins appear to be particularly critical for the maintenance of genomic stability during gametogenesis in a range of animal species including flies , worms and mice . The activity of the PIWIs is generally associated with the biogenesis and function of a specific class of 23-30nt small non-coding RNAs termed Piwi-interacting RNAs or piRNAs [4 , 5] . The details of the piRNA pathway have been mostly elucidated using the ovary of the fruit fly Drosophila melanogaster as model system . In this species , two branches of the pathway were shown to act in the germline and in the somatic follicle cells respectively . Drosophila harbors three PIWI proteins: Piwi , the founding member of this protein family , is mostly nuclear and acts in both arms of the piRNA pathway [4 , 6] , while Aubergine ( Aub ) and Argonaute3 ( Ago3 ) are expressed exclusively in the germ cells [7] . PIWI proteins belong to the Argonaute family and are characterized by typical PAZ , MID and Piwi domains . The PAZ and the MID domains interact with the mature piRNAs , which provide target specificity to the RNAse H slicing activity harbored in the Piwi domain . In Drosophila , piRNA precursor transcripts ( i . e . pre-piRNAs ) are mostly transcribed from genomic regions densely populated by transposon remnants and known as piRNA clusters[4] . A second source of piRNAs is provided by the transcripts generated by active transposable elements dispersed in the genome . The RDC complex , which is formed by the Cutoff ( Cuff ) , Rhino ( Rhi ) and Deadlock ( Del ) proteins , together with the transcription factor Moonshiner ( Moon ) and components of the THEO/Trex complex orchestrate that transcription of the piRNA clusters in the germline tissues and the transport of the pre-piRNAs from the nucleus to the cytoplasm [8–13] . The nuclei of Drosophila germ cells are surrounded by a membraneless organelle called the nuage , which hosts several enzymatic activities including the PIWI proteins Aub and Ago3 , the DEAD-box helicase Vasa ( Vas ) , and the tudor domain proteins Tudor , Krimper , Tejas and Papi [14] . These proteins act at different levels to process the pre-piRNAs and produce the mature 23-30nt long piRNAs . A critical role in the biogenesis of the piRNAs is exerted by Aub and Ago3 , which engage in feedforward amplification mechanisms termed the ping-pong cycle [4 , 15] . The ping-pong cycle amplifies the piRNA population and couples piRNA biogenesis with the degradation and , consequently , with the downregulation of transposon transcripts . Recent studies revealed that the Vas protein , a well conserved germline-specific DEAD-box helicase , provides an essential scaffold to anchor the Aub protein and to promote piRNA production [16] . Finally , antisense piRNAs are bound by Piwi , that translocates into the nucleus and employs the piRNAs as guide to locate and silence active transposable elements [17] . The second branch of the pathway acts in the somatic cells of the ovary . Slicing of the precursor transcripts from somatic piRNA clusters generates antisense piRNAs , which guide Piwi to silence transposons of the zam , gypsy and idefix families . piRNA biogenesis requires the Zucchini ( Zuc ) endonuclease , the helicase Armitage ( Armi ) and the Tudor domain proteins YB , SoYB , BoYB , Vreteno and T2RD2 , which accumulate in a cytoplasmic organelle known as the YB-body [18 , 19] . Both in the somatic as well as in the germline tissues of the fly ovary , the piRNA pathway protects the cells from the deleterious effects of massive transposon mobilization [2 , 20 , 21] . In Drosophila , mutations in PIWI proteins result in complete female adult sterility . Consistent with Aub and Ago3 being restricted to germline tissues , the absence of these factors culminates in a severe loss of stem and germ cells , a failure to assemble the chromatin in the oocyte nucleus ( i . e . the karyosome phenotype ) and the disruption of the dorsal-ventral polarity of the egg chamber and the future embryo [22–25] . Mutations in Piwi instead affect both the development of the follicular epithelium and of the germline . Also , this protein appears to act both in concert with or independent of the piRNAs [21] . A variable number of piwi genes originating from duplication events has been reported in various insect species . Among the Hemipteran insects , Rhodnius prolixus’ genome harbors 3 piwi genes and one ortholog of ago3 , while 8 piwis and 2 copies of the ago3 gene were observed in the aphid Acyrthosiphon pisum [26] . Thus , in various animals piwi undergoes gene amplification , and the different copies often display stage- and tissue-specific expression patterns suggesting functional specialization . For instance , the mosquito Aedes aegypti harbors 7 piwi orthologs , whereby the Piwi5 and Ago3 proteins have been connected to the biogenesis of viral piRNAs [27] . piRNAs and piwi orthologs have been also identified in mosquitos of the Anopheles genus [28 , 29] . piwi and vas have been extensively used to investigate the segregation of germ cell determinants in different species [30] . However , functional studies addressing the role of Piwi genes in insects other than Drosophila are still scarce . The blood-feeding insect Rhodnius prolixus is a primary vector of Trypanosoma cruzi , the etiologic agent of the Chagas disease [31] . The Chagas disease is life-threatening illness that currently affects 7–8 million people worldwide . Despite its medical relevance , the molecular events that drive oogenesis and guarantee adult fertility in Rhodnius are largely unknown . In this study , we employed transcriptome profiling to unveil the genetic and molecular basis underlying Rhodnius oogenesis . Our results reveal that central components of the piRNA pathway are conserved in this species and are expressed early during oogenesis . Furthermore , we show that Rp-piwi2 , Rp-piwi3 and Rp-ago3 , but not Rp-piwi1 , are expressed in Rhodnius ovaries , accumulate in germline tissues and are necessary for female adult fertility .
Rhodnius females were dissected 10 days after the feeding regimen and ovaries were immediately placed in cold Phosphate Buffered Saline ( PBS ) . Ovaries were fixed and immunostained as previously described [32] . The anti-γH2Ax ( Millipore ) and DAPI were diluted 1:1000 in PBS + Tween20 0 . 3% supplemented with 1% BSA . Ovaries were mounted in 70% glycerol and analyzed on a Leica Confocal Microscope . The evolutionary history of PIWI proteins in Rhodnius and Drosophila was inferred applying a Maximum Likelihood method [33] . The analysis included a total of seven amino acid sequences , which were aligned by the Multiple Sequence Alignment with Log Expectation ( MUSCLE , version 3 . 8 . 31 ) method [34] , employing standard parameters . The evolutionary history was inferred by Molecular Evolutionary Genetics Analysis version 6 . 0 ( MEGA6 ) , and visualized using interactive Tree of Life ( iTOL , v2 ) . The tree was validated by 1000 bootstraps replications . Values higher than 90% were indicated in nodes . The amino acid alignments performed to highlight the Rp-Piwi2 PolyQ stretch in Triatominae species included the following sequences: JAI55027 . 1 ( R . neglectus ) , JAP02788 . 1 ( T . dimidiata ) , JAC16725 . 1 ( T . infestans ) available in NCBI , and Rp-Piwi2 of R . prolixus . The aminoacid sequences of proteins from R . prolixus and D . melanogaster were obtained from VectorBase ( https://www . vectorbase . org/ ) and FlyBase ( http://flybase . org/ ) respectively . For the RT-PCR assays , total RNA was extracted from previtellogenic stages and , separately , from choriogenic stages dissected from 10/15 adult females . For qRT-PCR assays , total RNA was extracted from previtellogenic stages of wildtype , pRNAi and control ovaries 2 weeks after blood feeding . Tissues were ground in Trizol Reagent ( Invitrogen ) and processed as per manufacturer instructions . Total RNA was treated with Turbo DNA-free ( Ambion ) to remove genomic DNA traces . The resulting DNA-free total RNA was subjected to in vitro Reverse Transcription ( RT ) with Superscript III ( Invitrogen ) . 1 . 0μg of DNA-free total RNA was used for each reaction and assays were conducted in biological triplicates . The oligonucleotides used in RT-PCR and qRT-PCR assays , are listed in the S1 Table . For in vitro transcription , T7 promoter sequences in the appropriate orientation were added to the DNA templates through a PCR-based system using the set of oligonucleotides listed in S1 Table . The same DNA templates were adopted both for in situ hybridization assays and for dsRNA production . For parental RNAi assays , sense and antisense ssRNAs for each gene were produced using the Megascript kit ( Ambion ) . Approximately equal amounts of sense and antisense RNAs for each target were mixed in annealing buffer , precipitated and resuspended in water to a final concentration of approximately 1 . 5μg/μL . Two microliters of each dsRNA were injected in the abdomen of adult females three days prior blood feeding . A total of 10 adult females were injected with each dsRNA preparation . Ovaries were dissected two weeks after the feeding regimen . The DNA templates used to generate in situ hybridization probes were obtained by PCR using oligonucleotides carrying T7 promoter sequences at the 5'-end . The templates were subjected to in vitro transcription with the DIG RNA labeling kit ( Roche ) . In situ hybridization conditions have been described elsewhere [35] . Oligonucleotides used in this study are listed in the S1 Table . Ovaries of 10 blood-fed Rhodnius females were dissected in cold PBS and previtellogenic stages of oogenesis were manually separated from vitellogenic stages and chorionated eggs . Total RNA from two biological replicates was isolated with Trizol reagent ( Invitrogen ) and subjected to paired-end RNA library preparation as per manufacturer's instructions ( Illumina , Truseq paired-end RNA library prep . kit ) . The libraries were sequenced on Illumina HiSeq platforms at the Lactad Facility ( University of Campinas , Brazil ) . RNAseq datasets are available at NIH SRA ( SRP158580 ) . We employed the GEMtools pipeline ( https://github . com/gemtools/gemtools ) of the GEM mapper [36] to align 128 , 411 , 588 paired-end reads sequenced from two replicates ( 76 , 963 , 784 vs . 51 , 447 , 804 reads in each of the libraries vit1 and vit2 ) of the ovary samples to the Rhodnius genome assembly RproC1 , using the subsequently quantified transcriptome annotation RProC1 . 1 as a guide . Overall ~90% ( 89 . 1 vs . 89 . 6% ) of these reads mapped , and ~80% ( 80 . 9% vs . 79 . 2% ) of the in total sequenced reads were considered informative for the quantification at a proportion of multi-mappings of <1% . These mappings exhibited a fidelity of on average ~3 mismatches and indels ( 3 . 2 respectively 3 . 5 ) with the RproC1 reference genome sequence and were used for subsequent quantification of the RproC1 . 1 transcriptome as annotated by the Vectorbase community ( PMID: 22135296 ) and obtained from the Ensembl Metazoa database ( v88 ) . The EnsMart annotation of this database to maps 1 , 029 of the 1 , 467 most abundant mRNAs bidirectionally ( i . e . , orthology type "one-to-one" ) to protein-coding loci of the Flybase RefSeq annotation [37] , further 208 Flybase proteins can be rescued through "one-to-many" and "many-to-many" ortholog mappings . After evaluating the number of occurrences for each term in comparison to their occurrence in the entire Flybase reference annotation , a p-value for the statistical overrepresentation is computed according to the model implemented in the DAVID tool [38] . Based on the distribution of p-values , we control the rate of false-discoveries to be not higher than 0 . 05 and group the remaining terms by a fuzzy clustering procedure on their co-occurrence in gene products , calling clusters with at least 5 members . The RproC1 version of the Rhodnius genome is available at the following URL: https://www . ebi . ac . uk/ena/data/view/GCA_000181055 . 2 .
Adult Rhodnius females develop two ovaries , each composed of groups of seven ovarioles ( Fig 1A and 1B ) [39–41] . Germline stem cells are present in nymphal stages and the adult females inherit a discrete number of oocyte arrested in meiosis I and aligned at the anterior region of a lancet-like structure termed the tropharium . During oogenesis , each oocyte is surrounded by a layer of somatic follicle cells to form the mature egg chamber . Different from the meroistic polytrophic ovary found in Drosophila and other species , in the meroistic telotrophic ovary of Rhodnius the egg chambers do not harbor nurse cells . Instead , the nurse cells or trophocytes populate the tropharium , where they form a syncytium around a central region termed the trophic core ( Fig 1B ) . The tropharium can be divided in three regions with typical cell populations ( Fig 1A and 1B ) . Actively dividing germ cells are observed only in the Zone1 at the very anterior region [42] . These cells migrate to the Zone2 , where they lose their proliferation ability , begin the endoreduplication program and turn into trophocytes ( Ts ) , which are functionally analogous to the Drosophila nurse cells . In the Zone3 , that anticipates the previtellogenic egg chambers , the Ts display large nuclei with prominent nucleoli . In this region , the cell also start to degenerate and are progressively lost and replaced by new Ts migrating from Zone2 . Nutrients and possibly RNAs produced by the trophocytes accumulate in the trophic core and are subsequently transported to the growing oocytes through specialized cytoplasmic bridges termed trophic cords . The previtellogenic phase of oogenesis starts in the tropharium and ends when the egg chambers reach a diameter of 0 . 5mm [43] . During vitellogenesis , the egg chambers grow up to 1mm in length , the trophic cords are severed and choriogenesis begins . Mature eggs produced by Rhodnius adult females are protected by a compact and resistant chorion , which regulates the fertilization process and prevents dehydration ( Fig 1A ) . To investigate the molecular mechanisms that coordinate and drive Rhodnius oogenesis , we performed transcriptomic profiling in ovaries of blood-fed females . Our study focused on the previtellogenic phase of oogenesis . Total RNA extracted from these tissues was used to prepare and sequence paired-end RNAseq libraries and in total 12 . 84 Gigabases were sequenced in two replicates ( vit1 and vit2 ) . Sequence reads were then mapped to the Rhodnius genome RproC1 ( Methods ) as obtained from the Ensembl Metazoa database ( v83 ) [41] . On average ~86% of informative mappings to the genome ( 55 , 455 , 772 in vit1 , and 36 , 324 , 980 in vit2 ) superimposed in the correct orientation to the RProC1 . 3 transcriptome annotation . These mappings provide a deep interrogation of the annotated R . prolixus transcripts , with ~84% of the 14 , 840 transcripts detected by >10 read mappings , ~67% by >100 mappings , ~48% by >1 , 000 mappings , and ~10% by 10 , 000 mappings . Besides six rRNA loci ( RPRC015844 , RPRC015846 , RPRC016406 , RPRC016579 , RPRC016706 and RPRC016876 ) , these highly expressed loci comprise the RNAseP ( RPRC016972 ) and two SRP genes ( RPRC017200 and RPRC017302 ) . Interestingly , we also found a putative ortholog of the Drosophila squid gene to be highly expressed in Rhodnius ovaries . The Squid protein controls the localization and translation of the gurken mRNA , which encodes a TGFa-like morphogen involved in the axial polarization of the egg and future embryo . Of the remaining 7 , 172 highly expressed protein-coding ( pseudo- ) genes , we investigated the 1 , 467 genes that exhibited > 10 , 000 reads for common functional patterns . Since only 36 of these were annotated with known protein functions in the RproC1 . 3 genome version , we employed for our functional study 9 , 188 orthology mappings to the Drosophila melanogaster proteome . Five major groups obtained by clustering 626 functional terms are obtained from these most abundant mRNAs ( Fig 1C ) . These have been annotated in Flybase orthologs , collected from different databases and integrated into the DAVID functional annotation platform [44 , 45] . Our analysis reveals that genes with high expression levels in Rhodnius ovaries are orthologs of Drosophila proteins annotated with functions related to the ribosome ( group 1 ) and translation ( group 3 ) , to mRNA processing and splicing ( group 2 ) , to proteasome activity ( group 4 ) , and also to helicases that pave the way for transcription of genes and ATP metabolism ( group 5 ) . In agreement with our samples being depleted of vitellogenic and choriogenic egg chambers , the functional classes related to vitellogenin biogenesis and uptake as well as chorion synthesis displayed low expression levels in our datasets . Our results demonstrate that cells in the previtellogenic phase of Rhodnius oogenesis invest the major part of their energy in the biogenesis ( i . e . , at the level of transcription , mRNA processing and translation ) and turnover of the existing proteome by elevated proteome activity . We employed our transcriptome profiling to determine the extent of evolutionary and functional conservation of the piRNA pathway in R . prolixus . Using Blast tools , we interrogated the Vectorbase platform to identify genes with homology to the Drosophila factors involved in the biogenesis and function of the piRNAs . For each putative ortholog , we then computed the expression levels as per RNAseq , as average RPKM between two biological replicates ( Fig 2 ) . We immediately noticed that the heat shock protein 83 ( hsp83 ) and uap56 gene , which encode a nuclear-cytoplasmic RNA export factor , are expressed at higher levels ( >500 RPKM ) than other piRNA pathway components in Rhodnius ovaries ( Fig 2A ) . The majority of the putative piRNA pathway genes however could be grouped in two classes: intermediate and low expression levels ( Fig 2B and 2C ) . The first group is composed of 14 genes , whose steady state expression levels ranged between 50 and 250 RPKM ( Fig 2B ) . These genes encode putative orthologs of several cytoplasmic factors involved in the biogenesis of the piRNAs in Drosophila , including Vas , Tudor , Maelstrom , and two putative orthologs of the Zuc endonuclease belong to this class . The remaining 15 genes displayed average RPKM lower than 50 ( Fig 2C ) . Among them , armitage , the dSetDB1 methyl-transferase encoding eggless gene and the putative orthologs of krimper , papi and tejas , which encode Tudor domain proteins . Surprisingly , spn-E , a critical Helicase for the production of piRNAs in Drosophila , as well as the piwi ortholog Rp-piwi1 seem to be either expressed at very modest levels or not expressed in Rhodnius ovaries ( average RPKM<5 ) ( Fig 2C ) . The results of the RNAseq analysis were validated by qRT-PCR assays with oligonucleotides specific to a selected group of genes ( S2 Fig ) . In the fruit fly , the expression of the piRNA clusters in germline tissues is regulated by the RDC complex , which consists of the HP1 variant Rhino ( Rhi ) , the Rai1-like factor Cutoff ( Cuff ) and the Deadlock ( Del ) protein . Blast alignments of the Rhi aminoacid sequence with proteins encoded in the Rhodnius genome did not return a clear match . Since this protein is a member of the Heterochromatin Protein ( HP ) family , several putative Rhodnius HP proteins share comparable aminoacid sequence similarity with Rhi . Conversely , the Del protein appears to evolve rapidly and is restricted to the Drosophilids . In Drosophila , the cuff and the CG9125 genes code for proteins with aminoacid sequence similarity to the yeast transcription co-factor Rai1 . Interestingly , blast search analyses retrieve one single gene in Rhodnius encoding a putative protein displaying 20 . 7% and 28 . 1% aminoacid sequence identity with Cuff and with the protein encoded by CG9125 respectively . The Rhodnius Rai1-like gene , which we named Rp-rai1l , is not annotated in the current version of the Rhodnius genome and lies within the first intron of the gene RPRC008241 ( supercontig KQ034693 ) ( S1 Fig ) . Our transcriptomic analysis reveals that Rp-rai1l displays intermediate expression levels during Rhodnius oogenesis ( Fig 2B ) , although its role , if any , in the piRNA pathway needs to be elucidated . For several genes that have been linked to the piRNA pathway in Drosophila , our blast search did not return homologous sequences in the Rhodnius genome . For instance , the Rhodnius genome does not appear to host homologs of the Drosophila moon and squash genes , which are expressed in the germline tissues , as well as of YB , SoYB , BoYB and T2RD2 that act in the somatic branch of the piRNA pathway in the fruit fly . We then focused our study on the Rhodnius orthologs of the vas and piwi genes , which are central components of the piRNA pathway in Drosophila . Vas is a DEAD-box RNA helicase related to the translation factor eIF-4A and has been extensively used as a marker of germline tissues in distantly related organisms [46] . Importantly , studies in Drosophila showed that Vas elicits piRNA production in concert with Aub and Ago3 [16] . We found that the Rhodnius genome harbors a putative homolog of the vas gene ( Vectorbase ID RPRC009661 ) , which encodes a protein 75% identical to DmVas . Our RNAseq profiling reveals that Rp-vas is expressed at intermediate levels during Rhodnius oogenesis ( Fig 2B ) . Interestingly , while the D . melanogaster vas hosts the vig and solo genes in its intronic sequences , this arrangement is absent in Rhodnius , where the Rp-vas intron 1 harbors an Open Reading Frame ( ORF ) encoding a transposase enzyme from a mariner-like element ( Fig 2D ) . Transcripts of this ORF are readily detected in our ovarian transcriptome datasets . Previous studies reported that Rhodnius displays an amplification of the piwi genes , whereby three putative orthologs of piwi , namely Rp-piwi1 , Rp-piwi2 and Rp-piwi3 , in addition to the Rp-ago3 ortholog of the Drosophila ago3 gene are present in the genome [26] . The analysis of the normalized RNAseq reads shows that Rp-piwi2 , Rp-piwi3 and Rp-ago3 are expressed at intermediate or low levels in Rhodnius oogenesis ( Fig 2B–2D ) , while the Rp-piwi1 transcripts are barely detectable . Based on the absence of intronic sequences , it has been proposed that Rp-piwi1 is a pseudogene and might not be required for Rhodnius development . Our transcriptomic analyses seem to support this hypothesis , although we cannot rule out the Rp-piwi1 might be expressed in tissues other than the ovary . According to the current genome annotation , Rp-piwi2 contains a small intron of approximately 100bp . Our RNAseq analysis however reveals that this sequence is included in the mature transcript and the resulting ORF encodes a putative protein of 882aa . Furthermore , the 5' and 3' untranslated regions of the Rp-vas , Rp-piwi2 , Rp-piwi3 and Rp-ago3 extend beyond the limits annotated in the current version of the genome . Thus , our datasets not only provide information on the steady-state expression levels for all the genes and loci expressed in previtellogenic stages of Rhodnius oogenesis , but will also contribute to improve gene annotation and discovery . The Rp-piwi1 , Rp-piwi2 and Rp-piwi3 ( VectorBase IDs RPRC00252 , RPRC002460 and RPR001891 ) encode putative proteins with 38 . 7% , 36 . 1% and 43 . 3% aminoacid sequence identity with D . melanogaster Piwi respectively ( Fig 3A and S2 Fig ) . While the piwi orthologs appear to have originated from duplication of an ancestral piwi gene , the Rp-ago3 gene is homologous to the Drosophila ago3 gene ( Fig 3A and S2 Fig ) . Accordingly , the Rp-ago3 locus ( Vectorbase ID RPRC013054 ) encodes a putative protein 43 . 3% identical to DmAgo3 . We then analyzed the degree of sequence identity between the individual Piwi , MID and Paz domains across the PIWI proteins in Drosophila and Rhodnius ( Fig 3A and S2 Fig ) . All the domains appear to be well conserved in all the Rhodnius orthologs including the putative PIWI protein encoded by the Rp-piwi1 gene . Interestingly , the N-terminal region of the putative Rp-Piwi2 protein features a stretch of 18 Glutamine residues ( i . e . polyQ tract ) ( Fig 3A and S2 Fig ) . This characteristic has not been reported for any of the PIWI proteins so far analyzed in a range of animal species . We therefore wondered whether PolyQ stretches are present in PIWI related proteins from other insects . Using NCBI Blast search analyses , we found that the polyQ tract is present in Piwi-like proteins encoded in the genome of Rhodnius neglectus , Triatoma infestans and Triatoma dimidiata ( Fig 3B ) , but it is absent in PIWI proteins of Drosophila and other animals ( Fig 3A and S2 Fig ) . Thus , the acquisition of a PolyQ sequence is likely a recent evolutionary event and is restricted to certain Triatomine species . Next , we wondered whether the Rp-vas and the Rp-piwi genes display stage- or tissue-specific expression patterns during Rhodnius oogenesis . To answer this question , we performed in situ hybridization assays in fixed ovaries using antisense probes corresponding to specific sequences within the ORF of Rp-vas , Rp-piwi1 , Rp-piwi2 , Rp-piwi3 and Rp-ago3 ( Fig 4 ) . This approach revealed that the expression of Rp-vas is restricted to the germline tissues given that the Rp-vas probe generates a signal in the tropharium and in the oocyte , but not in the somatic follicle cells ( Fig 4A ) . Thus , RPRC009661/Rp-vas encodes a bona fide ortholog of DmVas , which allows distinguish the germ cell lineage from the somatic cell population . Previous studies reported the expression of a putative vas ortholog in the somatic follicle cells of the Rhodnius ovary [47] . Our in situ hybridization protocols did not allow preserve the vitellogenic egg chambers , thus we could not determine whether RPRC009661/Rp-vas transcripts are produced in the follicular epithelium in late stages of oogenesis . In accordance with the RNAseq data , the Rp-piwi1 probe did not produce any signals above the background levels ( Fig 4B ) . The Rp-piwi2 transcripts instead are clearly detected in the tropharium and in the developing oocytes ( Fig 4C ) . Interestingly , the Rp-piwi2 RNAs seem to unevenly accumulate in more mature oocytes , where they are enriched in the anterior region ( Fig 4C ) . In addition , the Rp-piwi2 transcripts are detected also in the somatic follicle cells , thus suggesting that the expression of this gene is not restricted to the germline tissues ( Fig 4C' ) . Similar to Rp-vas , the Rp-piwi3 and Rp-ago3 transcripts are detected in the region of the tropharium hosting the polyploid Ts and in the ooplasm of newly formed and mature egg chambers , but not in the follicle cells ( Fig 4D and 4E ) . As control assay , we generated a sense probe corresponding to a region of the Rp-ago3 ORF ( Fig 4F ) . This probe did not produce specific signals above the background levels . During Rhodnius oogenesis , RNAs and nutrients produced by the nurse cells hosted in the tropharium are transported to the growing oocytes through the trophic cords . In order to determine whether piwi and vas transcripts are maternally stored in the mature eggs , we performed RT-PCR assays in ovaries with oligonucleotides specific for the piwi and vas genes by separating the previtellogenic stages of oogenesis from the chorionated mature eggs . In both stages of Rhodnius oogenesis , a clear amplification product of the expected molecular weight was detected for the Rp-piwi2 , Rp-piwi3 , Rp-ago3 and Rp-vas genes , while no amplification signal was produced for Rp-piwi1 ( Fig 4G and 4H ) . These observations point to a direct role for Rp-piwi2 , Rp-piwi3 , Rp-ago3 and Rp-vas in Rhodnius germline development and possibly early embryogenesis . Parental RNAi ( pRNAi ) was previously shown to induce the efficient reduction of gene expression in Rhodnius , where genetic tools are still lacking [48] . In order to understand the function of the Rp-piwi genes in Rhodnius , we carried on pRNAi assays by injecting dsRNA molecules targeting portions of their coding regions in the abdomen of adult females ( Fig 5A ) . Injected females were blood-fed , their eggs were collected daily over a period of three weeks and let to develop until the first-instar nymphs emerged . After the 3-weeks period , the females were dissected and the eggs retained in the abdomen were also counted . This approach allowed us to investigate the oviposition and fertility of the injected females . Eggs were divided into three bins: 1 ) total number of eggs , which is given by the sum of the eggs retained in the abdomen and those that were actually oviposited , 2 ) oviposited eggs , 3 ) eggs that hatched to produce first-instar nymphs . Each group of control-injected females produced on average a total of 262 eggs of which 103 were oviposited and 67 developed into first-instar nymphs . Compared to these control animals , the Rp-piwi1 pRNAi females produced on average a slightly lower number of total eggs ( ~86% ) and oviposited eggs ( ~83% ) . However , the hatching rates were higher than the control ( ~111% ) . It is noteworthy that in our pRNAi assays , the Rp-piwi1-KD females consistently produced a slightly higher number of first-instar nymphs than the control females , although this gene is apparently not expressed in ovaries . In contrast , oogenesis seemed to be partially impaired by pRNAi-mediated downregulation of the Rp-piwi3 and Rp-ago3 genes . Compared to the control , females from these assays respectively produced ~63% and ~48% total eggs , ~40% and ~39% deposited eggs and ~33% and ~35% first-instar nymphs . The most striking result however was obtained upon injection of Rp-piwi2 dsRNA molecules in adult females . The total number of eggs and the number of oviposited eggs produced on average by these animals was ~19% and ~13% of the control values , respectively . More importantly , the first-instar nymphs generated by Rp-piwi2 pRNAi females were only ~2% of the control values . In order to investigate the specificity of our dsRNAs for the respective cognate Rp-piwi gene , we performed qRT-PCR assays in previtellogenic stages of pRNAi ovaries with oligonucleotides specific to Rp-piwi1 , Rp-piwi2 , Rp-piwi3 and Rp-ago3 ( Fig 5B ) . As internal control for this assay , we used oligonucleotides specific to Rp-rp49 ( RPRC014419 ) , a putative Rhodnius ortholog of the Drosophila rp49 gene . These assays revealed that each dsRNA specifically downregulates the expression of the cognate gene by ~30% for Rp-piwi3 , ~40% for Rp-ago3 and >50% for Rp-piwi2 . As expected , the injection of Rp-piwi1 dsRNA did not produce significant changes in the expression levels of Rp-piwi1 . Our results strongly point to a critical role for Rp-piwi2 , Rp-piwi3 and Rp-ago3 in Rhodnius oogenesis and female adult fertility . We then sought to determine the cellular basis of the reduced fertility observed in Rp-piwi2 , Rp-piwi3 and Rp-ago3 KD females . To this aim , we immunostained ovaries from pRNAi-injected adult females with DAPI to visualize the nuclei of the germ cells and of the follicle cells ( Fig 5 ) . We immediately noticed that Rp-piwi2 pRNAi tropharia displayed abundant DAPI-positive particles in Zone1 and Zone2 , which are rarely observed in the tropharia of control ovaries ( Fig 5C and 5D ) . In some cases , the anterior tip of the tropharium corresponding to Zone1 appeared severely atrophic ( Fig 5E ) . The γH2Ax histone variant was shown to accumulate at sites of DNA damage , induced for instance by meiotic recombination events or by transposable element mobilization [23 , 49] . We therefore monitored the occurrence of DSBs in control and Rp-piwi2-KD ovaries with the antibodies specific to γH2Ax ( Fig 5F , 5F’ , 5G and 5G’ ) . This assay revealed that the DAPI particles observed in Zone1 of Rp-piwi2-KD tropharia as well as some nurse cell nuclei are enriched in this histone variant . In contrast , control tropharia did not display any signal beyond the background levels . In Zone 3 of control tropharia , the Ts nuclei are arranged at the periphery of the tropharium and display apparent nucleoli ( Fig 5H ) . Instead , the Zone3 of the pRNAi-treated ovaries clearly displays a lower number of Ts nuclei and abundant nuclear debris in the trophic core ( Fig 5I ) . In Rhodnius , each egg chamber is formed by an oocyte surrounded by a follicular epithelium ( Fig 5J ) . pRNAi for Rp-piwi2 seems to strongly impair the progression through oogenesis and the growth of the egg chambers . We frequently observed smaller and apparently collapsed egg chambers during vitellogenesis ( Fig 5K ) . These atrophic egg chambers were still connected posteriorly to younger egg chambers emerging from the tropharium and anteriorly to more mature choriogenic egg chambers through bridges of stalk cells . Despite the clear impact of pRNAi on Rhodnius fertility , the analysis of Rp-piwi3 and Rp-ago3 pRNAi ovaries did not display obvious abnormalities by DAPI staining and anti-γH2Ax immunostainings and additional molecular tools will be necessary to dissect their function during germline development . Our results however , demonstrate that Rp-piwi2 in fundamental for Ts survival and egg chamber development during Rhodnius oogenesis .
PIWI proteins complexed with piRNAs coordinate a defense system that represses mobile genetic elements and protects the genome of animal germ cells . In this study , we show that central components of the piRNA pathway , first described in Drosophila are conserved in the hemimetabolous insect Rhodnius prolixus , which is 350mya distant from the fruit fly . Rhodnius harbors four putative piwi genes , and we show that Rp-piwi2 , Rp-piwi3 and Rp-ago3 , but not Rp-piwi1 , are expressed in ovaries . In order to investigate their expression patterns during oogenesis , we first identified RPRC009661 as a vas ortholog in Rhodnius and we showed that it is a germline-specific gene . The Rp-piwi3 and Rp-ago3 transcripts display a germline-specific expression patterns similar to Rp-vas and appear enriched in the growing oocytes . Interestingly , Rp-piwi2 transcripts can be detected both in the somatic as well as in the germ cells and seem to display an asymmetric localization pattern during oocyte development . This gene is expressed in the tropharium and its transcripts evenly accumulate in the budding egg chamber . In the neighboring and more mature egg chamber , however , Rp-piwi2 transcripts are enriched at the anterior pole of the oocytes . Our in situ hybridization assays suggest that Rp-piwi2 transcripts might diffuse from the ooplasm of the budding egg chamber into the neighboring more mature oocyte . Alternatively , Rp-piwi2 expression might occur in the invading follicle cells that form the boundary between the budding egg chambers and the transcript deposited in the adjacent oocytes . It will be of great interest to determine whether the Rp-piwi2 expression pattern impacts the axial polarization of the Rhodnius eggs and embryos . Our functional studies using pRNAi against the Rp-piwi2 gene resulted in oogenesis arrest and complete female adult sterility . In wild type ovaries , the Zone 1 of the tropharium hosts mitotically active trophocytes , which replenish the population of polyploid Ts in Zone 2 and 3 . Reduction of the Rp-piwi2 levels by pRNAi causes a severe loss of dividing cells in Zone 1 and of polyploid Ts in Zone 2 and 3 . The accumulation of γH2Ax-positive nuclear debris in the tropharia of injected females strongly suggests that Ts degenerate in Rp-piwi2 KD ovaries . The loss of Ts in turn likely results in dumping phenotypes , which explain the oogenesis arrest and the frequently collapsed egg chambers observed in these females . It is tempting to speculate that the DNA damage and the loss of Ts observed in the Rp-piwi2 KD ovaries might be caused by the deregulation of transposable elements . The percentage of transposable elements in the Rhodnius genome is approximately 6% and two thirds of the transposons in this species belong to the mariner family [41 , 50 , 51] . Rp-piwi2 might be required to silence these elements in the germline and , possibly , in the somatic tissues . The cloning and characterization of the piRNA population will be necessary to shed more light on the function of Rp-piwi2 and the piRNA pathway in this species . Remarkably , we found that the putative Rp-Piwi2 protein features a 18aa Polyglutamine ( PolyQ ) tract at its N-terminal region . PolyQ repeats have been identified in various proteins of organisms as distant as plants and vertebrates and are often found in transcription factors . Interestingly , the PolyQ stretch appears to be conserved in putative PIWI proteins of the closely related species Rhodnius neglectus , Triatoma infestans and Triatoma dimidiata , while it is not present in the PIWI proteins of other organisms including Drosophila . Thus , the PolyQ tract is likely a recent acquisition in the evolution of the PIWI proteins and , based on the available sequenced genomes , appears to be restricted to blood-feeding insects of the Triatomine family . Albeit to a lesser extent , Rp-piwi3 and Rp-ago3 KDs also affect egg production and female adult fertility in Rhodnius . The expression of Rp-piwi1 gene instead is negligible in Rhodnius ovaries and Rp-piwi1 dsRNA injection in adult females does not negatively affect oogenesis and fertility . In addition to the Rp-piwi genes our transcriptomic analysis revealed that several components of the piRNA pathway are conserved and expressed in the ovary of this species . We did not find evidence of an RDC complex in Rhodnius except for a putative protein ( i . e . Rp-Rai1l ) with similarity to Cuff . If piRNA clusters exist in this species , it is likely that their regulation relies on a set of proteins different from the one described in Drosophila . However , we provide evidence that several factors involved in the transport and processing of the pre-piRNAs , including Uap56 , Krimper and Maelstrom among others , are expressed during Rhodnius oogenesis . Yet , some critical germline factors , like the Helicase SpnE , are expressed at very low levels . Similarly , the somatic branch of the piRNA pathway might rely on the activity of the Rp-piwi2 gene and the zuc , armi and vret orthologs , while YB , BoYB , SoYB and T2RD2 , which associate with the YB bodies and catalyze the production of mature piRNAs in the Drosophila follicle cells , are not present in the Rhodnius genome . These genes have been reported to be absent also from the genome of other insect species , including the Honeybee Apis mellifera and Tribolium castaneum [52] . Thus , both branches of the piRNA pathway are partially conserved in insects and it will be a challenge for the future to fully understand the differences between Drosophila and Rhodnius . Rhodnius prolixus together with other Triatomine species are major vectors of the protozoan Trypanosoma cruzi , the causal agent of the Chagas disease . In this study , we shed light on the ovarian transcriptome of Rhodnius and unveiled the degree of evolutionary and functional conservation of the piRNA pathway in this species . Furthermore , we show that piwi genes are essential for oogenesis and adult fertility in Rhodnius and likely exert similar functions in other Triatomine species . Sterile Insect Techniques ( SIT ) have been extensively used to reduce natural populations of insects of medical or economic importance in many countries [53] . Thus , our results provide a framework for the development of novel strategies to control the natural populations of Triatomine insect vectors and reduce the spread of the Chagas disease . | Rhodnius prolixus together with other blood-feeding bugs of the Triatominae family are primary vectors of the protozoan Trypanosoma cruzi , the causative agent of the Chagas disease . It has been estimated that 7–8 million people are affected by this life-threatening illness worldwide , which makes the Chagas disease one of the most neglected tropical diseases . In this study , we describe the transcriptome of previtellogenic stages of Rhodnius oogenesis . Furthermore , by using a combination of molecular biology techniques and functional analyses we show that central components of the piRNA pathway are conserved in this species . The piRNA pathway guarantees genomic stability in the germ cells of organisms as distant as flies and mice . In accordance , we find that the knock-down of the piwi genes , which form the backbone of the pathway , results in partial or complete female adult sterility in Rhodnius . Our data will help improve the annotation of the Rhodnius genome and provide a framework for the development of novel techniques aiming at the eradication of Rhodnius prolixus and other Triatomine species from the infested areas . The achievement of this goal will ultimately prevent the transmission of trypanosomes to humans and reduce or eliminate the diffusion of the Chagas disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"animals",
"invertebrate",
"genomics",
"reproductive",
"physiology",
"animal",
"models",
"germ",
"cells",
"oocytes",
"drosophila",
"melanogaster",
"model",
"organisms",
"experimental",
"or... | 2018 | Transcriptomic and functional analyses of the piRNA pathway in the Chagas disease vector Rhodnius prolixus |
Differentiation of secretory cells leads to sharp increases in protein synthesis , challenging endoplasmic reticulum ( ER ) proteostasis . Anticipatory activation of the unfolded protein response ( UPR ) prepares cells for the onset of secretory function by expanding the ER size and folding capacity . How cells ensure that the repertoire of induced chaperones matches their postdifferentiation folding needs is not well understood . We find that during differentiation of stem-like seam cells , a typical UPR target , the Caenorhabditis elegans immunoglobulin heavy chain-binding protein ( BiP ) homologue Heat-Shock Protein 4 ( HSP-4 ) , is selectively induced in alae-secreting daughter cells but is repressed in hypodermal daughter cells . Surprisingly , this lineage-dependent induction bypasses the requirement for UPR signaling . Instead , its induction in alae-secreting cells is controlled by a specific developmental program , while its repression in the hypodermal-fated cells requires a transcriptional regulator B-Lymphocyte–Induced Maturation Protein 1 ( BLMP-1/BLIMP1 ) , involved in differentiation of mammalian secretory cells . The HSP-4 induction is anticipatory and is required for the integrity of secreted alae . Thus , differentiation programs can directly control a broad-specificity chaperone that is normally stress dependent to ensure the integrity of secreted proteins .
Cellular identity is largely defined by the proteins expressed in the cell or cellular proteome , whose functionality depends on successful folding , localization , and functional maintenance of expressed proteins . During cellular differentiation , rapid onset of new protein synthesis challenges the proteostasis and may result in the production of dysfunctional proteins and folding stress if not matched by corresponding increases in required chaperones [1] . This is especially evident for differentiating professional secretory cells because production of large quantities of secreted protein [2 , 3] makes them extremely sensitive to the folding stress in the endoplasmic reticulum ( ER ) . To accommodate the anticipated increase in newly synthesized proteins , the ER proteostasis networks are expanded during differentiation through activation of the ER stress response [4] , known as the unfolded protein response ( UPR ) [5] . In addition to their expansion , proteostasis networks may need to be remodeled during differentiation since different secreted proteins may require different chaperones for their biogenesis [6 , 7] . An early example of this was shown during differentiation of a B-cell line into antibody-secreting cells: while expression of the majority of ER proteins , including the Heat Shock Protein 70 ( HSP70 ) -family chaperone BiP , increased in proportion to the expansion of ER size , a small subset of ER proteins was preferentially up-regulated , resulting in their increased local concentration within the ER , presumably to support immunoglobulin folding and secretion [8] . How this selective up-regulation is achieved , and whether it requires the UPR machinery , is not well understood . The canonical UPR signaling includes three major branches—controlled by the Serine/Threonine-Protein Kinase/Endoribonuclease IRE1 and a basic leucine-zipper transcription factor X-Box Binding Protein 1 ( XBP-1 ) , by the Activating Transcription Factor 6 ( ATF-6 ) , or by the PRKR-Like Endoplasmic Reticulum Kinase ( PERK ) . IRE1 and/or XBP-1 are essential for differentiation of many secretory cells such as plasma cells [9 , 10] and eosinophils [11] and for biogenesis of exocrine pancreas and salivary glands in mice [12 , 13] . IRE1 is an ER transmembrane protein that , upon sensing folding stress in the ER , cleaves the mRNA of XBP-1; the resulting active spliced form of XBP-1 controls expression of molecular chaperones and other ER biogenesis genes [14] . Ectopic expression of spliced XBP-1 in cultured cells is sufficient to induce expansion of the ER size and cell’s secretory capacity , while deletion of xbp1 gene in the mouse B-cell lineage prevents development of antibody-secreting plasma cells [9 , 15] . In fact , XBP-1 , together with a transcriptional repressor BLIMP1 , are the two regulators required for plasma cell differentiation [15 , 16] . The xbp1 gene is repressed in resting B cells [17] , and BLIMP1 relieves this repression upon B-cell stimulation , leading to up-regulated xbp1 transcription [15 , 18] . Thus , plasma cell differentiation program directly regulates the UPR transcription factor responsible for the general increase in the secretory capacity . Indeed , activation of UPR during plasma cell differentiation appears to be in anticipation of increase in secretory load rather than in response to proteostatic stress [19 , 20] . Compared with the general ER expansion , much less is known about the second aspect of the ER proteostasis remodeling during differentiation: up-regulation of select chaperones to match the cell-type–specific folding needs . Many ER chaperones are expressed in cell- and tissue-specific patterns during development; however , there are only few examples in which the basis of this cell selectivity is understood at the molecular level . One specialized case are the client-specific chaperones , such as a collagen chaperone HSP47 , which is normally induced by heat stress but , during development , is co-regulated with its client collagens by developmental transcription factors [21] . An example for the induction of the major stress-responsive ER chaperone BiP can be seen during cardiac development [22] . Unlike the client-specific HSP47 , BiP is a broad-specificity chaperone , required for the general housekeeping functions in the ER . The Grp78 gene ( Glucose-Regulated Protein , 78 kDa ) , which encodes BiP , is a canonical UPR target whose promoter has been used to delineate the UPR signaling and to identify binding motifs for UPR transcription factors [23–25] . The induction of BiP during heart development reflects cooperation between the UPR transcription factor ATF-6 and the cardiac-specific transcription factor GATA-4 , which appears to bind Grp78 promoter through the ER stress element that is otherwise recognized by ATF-6 under stress conditions [22] . It remains unclear whether such cooperation between the UPR and developmental signaling is the rule and how the selective up-regulation of ER chaperones during differentiation is integrated with the cellular differentiation program . Here , we take advantage of the stereotypical timing and patterns of cell divisions and differentiation in C . elegans to examine the regulation of a broad-specificity chaperone , the BiP homologue HSP-4 , during differentiation of dedicated secretory cells that secrete cuticular ridges called alae . C . elegans possesses two homologues of BiP: HSP-3 , which is both constitutively expressed and stress-responsive , and HSP-4 , which has very low basal expression in most cells but is strongly induced by UPR signaling [26 , 27] . Using the well-characterized transcriptional reporter expressing green fluorescent protein ( GFP ) from the hsp-4 promoter ( phsp-4::GFP ) [28] , we find that hsp-4 is selectively and transiently induced during differentiation of the stem-like seam cells into alae-secreting cells . Asymmetric divisions of seam cells produce anterior daughters that differentiate into hypodermal cells and posterior daughters that continue stem-like divisions but differentiate into the alae-secreting cells after the last division . hsp-4 is induced only in these posterior cells prior to their differentiation , in an anticipatory fashion . Unexpectedly , this hsp-4 induction is neither dependent on the three canonical UPR signaling pathways—IRE1/XBP-1 , ATF-6 , and PERK—nor does it require the known ER stress elements in its promoter . On the other hand , repression of hsp-4 in the hypodermal-fated cells requires BLMP-1 , a C . elegans homologue of the B-cell differentiation factor BLIMP1 . The non-UPR induction of HSP-4/BiP may be selectively required for the folding or secretion of a specific client ( s ) in alae-secreting cells , as indicated by the abnormal alae structures and compromised barrier function of the cuticle when HSP-4/BiP induction is abolished . Our results demonstrate that a broad-specificity molecular chaperone that is a canonical UPR target can be selectively regulated by developmental signaling , independent of UPR pathways , to ensure the integrity of the secreted proteome and functionality of the cell postdifferentiation .
Although basal expression of the UPR-inducible BiP homologue HSP-4 is low in most tissues of C . elegans , the phsp-4::GFP transcriptional reporter is visibly induced in unstressed animals in two highly secretory tissues—spermathecae and the lateral seam . Because seam cells undergo stereotypical and well-characterized divisions and differentiate at defined developmental stages [29] , we used them to examine the regulation of BiP expression . During reproductive development , seam cells of V1–V4 and V6 lineages ( S1A Fig ) undergo two types of divisions—one symmetric division early in the second larval ( L2 ) stage and four rounds of asymmetric divisions [29] . The asymmetric divisions produce anterior daughters that differentiate and fuse with hypodermal syncytium after each cycle of divisions [30] and posterior daughters that continue dividing until the fourth larval ( L4 ) stage , when they differentiate , fuse with each other , and begin secreting proteins to make specialized cuticular structures , named alae [31 , 32] . In addition to this normal developmental sequence , early L2 animals under certain environmental stress conditions can enter into an alternative developmental program known as dauer diapause , resulting in formation of nonfeeding and long-lived dauer larvae [33] . During dauer development , the seam cells differentiate at the end of the predauer L2 stage , known as L2d stage ( S1A Fig ) , and secrete the dauer-specific cuticle and alae [34] . We observed that phsp-4::GFP reporter was visibly induced in seam cells during two developmental stages—weakly in the late L4 stage and strongly in L2 stage animals on starved crowded plates ( Fig 1A ) —while it was undetectable in other larval stages . Since starved L2 animals on crowded plates often initiate the dauer program , we also tested predauer animals . A mutant allele ( sa191 ) of an insulin/insulin-like growth factor ( IGF ) -like protein DAF-28 causes animals to enter the L2d stage even in the presence of food and to remain in that stage for several hours [35 , 36] . We observed a strong and persistent induction of the phsp-4::GFP reporter in the seam cells of daf-28 ( sa191 ) animals at the L2d stage ( Fig 1A ) . The reporter induction in the late L4 and predauer animals indicates that seam-specific hsp-4 expression is triggered at developmental stages that result in differentiation of the alae-secreting cells ( S1A Fig ) . Closer examination suggested that hsp-4 expression is indeed induced in the posterior daughter cells , fated to differentiate into alae-secreting cells after the last asymmetric division . To confirm this , we employed two commonly used reporters—pegl-18::H1-mCherry , which is preferentially expressed in the posterior cells after asymmetric divisions [37] ( S1B Fig ) , and pdpy-7::HIS-24-mCherry , expressed specifically in the anterior cells differentiating into hypodermal cells [38 , 39] ( Fig 1B and 1C ) . Unexpectedly , the pegl-18::H1-mCherry reporter lost its asymmetry in the predauer animals ( Fig 1B ) . However , the pdpy-7::HIS-24-mCherry reporter was strongly expressed in hsp-4–negative cells and only weakly in hsp-4–positive cells in predauers ( Fig 1C ) , confirming that hsp-4 expression is induced in the posterior seam cells as they are differentiating into the alae-secreting cells . While the asymmetric expression pattern showed cell-selective induction of the chaperone BiP/HSP-4 during differentiation , it was not clear whether it was triggered by the postdifferentiation increase in the secretory load or was induced in anticipation of it . To determine how early during the last asymmetric division and differentiation hsp-4 is induced , we used AJM-1::GFP ( Apical Junction Molecule ) protein that localizes to apical junctions in epithelial cells and outlines seam-cell boundaries [40] . Immediately after the asymmetric division , AJM-1::GFP is present in both daughter cells , but it is lost from the anterior daughters as they differentiate and fuse with the hypodermal syncytium [41] . In contrast , posterior stem-like daughters continue expressing AJM-1::GFP until they differentiate , when they fuse and begin secreting proteins necessary for the formation of alae [42] . Based on AJM-1::GFP pattern in L2d animals , we determined that induction of hsp-4 expression in posterior daughters happens already in the early stages after the last division , when anterior daughters have just started to lose their boundaries and have not yet migrated away ( Fig 2A ) . This timing is consistent with anticipatory induction . To further confirm the anticipatory nature of hsp-4 induction , we examined its timing in sa191 animals . Under normal growth conditions at 20 °C , sa191 animals that do activate the dauer program enter the L2d stage by 41 hours postgastrula . Because the dauer activation is only partial in these L2d animals , most of them ( approximately 70% ) [36] return to reproductive development several hours later instead of entering dauer [35] . Therefore , most of the sa191 animals do not complete the seam-cell–differentiation program and do not secrete dauer cuticle or form dauer alae . We found that the seam-cell–specific induction of hsp-4 expression was readily detectable in 100% ( n > 100 ) of sa191 animals that did enter the L2d stage , assayed at 41 hours postgastrula , and was still present in the same animals at 46 hours postgastrula ( see also control animals in blmp-1 RNA interference [RNAi] experiment below , n = 63 ) , after which time many animals return to reproductive development without secreting dauer cuticle or alae proteins . To ask whether such anticipatory induction early in differentiation is peculiar to the predauer stage , we examined seam cells in L4 animals . The last asymmetric division occurs around the time of the molt from the third larval ( L3 ) stage to the L4 stage; alae-fated cells then differentiate and fuse at the end of the L4 stage , prior to the onset of alae secretion [29 , 30] . The fusion is detectable by the change in the AJM-1::GFP pattern from outlines of individual seam cells to the outline of the syncytium running along the body length of the nematode ( Fig 2B ) . Because the L4 stage lasts nearly 10 hours at 20 °C , we imaged the seam lineage in young L4 animals after the last asymmetric division and in mature L4 animals prior to and after the fusion . Expression of HSP-4::GFP reporter was evident already in the very young L4 animals ( Fig 2B , left panel ) well before the fusion event . Collectively , these data show that hsp-4 expression is selectively and anticipatorily induced during differentiation of the alae-secreting cells . Since hsp-4 induction was strongest in predauers and seemed to follow the initiation of the dauer signaling , we asked whether it was responding to a specific dauer-inducing signal . We found that hsp-4 reporter was similarly induced in predauer animals whether the dauer signaling was induced through the insulin/IGF pathway ( daf-2 ( 1370 ) animals ) or the transforming Growth Factor β ( TGFβ ) pathway ( daf-7 ( e1372 ) animals ) ( S2 Fig ) . The transcription factor DAF-16 downstream of the insulin/IGF pathway , a C . elegans homologue of the mammalian FOXO3 ( Forkhead Box Protein O3 ) , was recently shown to have an impact on UPR [43] , prompting us to ask whether hsp-4 induction was dependent on DAF-16 . Animals bearing a hypomorphic allele daf-16 ( mu86 ) are dauer deficient; however , those animals that did initiate the predauer program upon starvation/crowding had a phsp-4::GFP induction pattern indistinguishable from the WT ( S2 Fig ) . Finally , dauer induction requires the heat-shock transcription factor HSF-1 [44] . The hsp-4 gene is heat-inducible , and the hsp-4 promoter has predicted HSF-1 binding sites ( S4A Fig ) . However , animals carrying the heat-shock-response–deficient hsf-1 ( sy441 ) allele were still able to induce hsp-4 in seam cells of starved L2 stage animals ( S2 Fig ) . Together , these data show that HSP-4/BiP is selectively induced in stem-like seam cells prior to their differentiation into alae-secreting cells . The chaperone induction is anticipatory and is triggered by specific developmental programs—dauer entry or the L4-stage-to-adult transition . Because anticipatory induction of the mammalian homologue of HSP-4 protein , BiP , during differentiation of B cells and other secretory cells is controlled by activation of the UPR signaling , we asked whether other UPR-responsive genes were also activated in the differentiating seam cells . We tested available transcriptional reporters of three genes—hsp-3/BiP , enpl-1 , encoding the orthologue of GRP94 ( Glucose Regulated Protein , 94 kDa ) , and cnx-1 , encoding the orthologue of calnexin—known to be induced by ER stress in C . elegans in an IRE-1/XBP-1–dependent manner [45] . Neither enpl-1 nor cnx-1 reporters showed detectable induction in the seam cells of either L2d stage ( S3A Fig ) or late L4 stage animals . The hsp-3 reporter was constitutively expressed in most tissues and was not induced beyond its basal levels during seam-cell differentiation . Interestingly , expression of cnx-1 was induced in the V5 seam-lineage–derived neuroblast cells in early L2 animals ( S3B Fig ) . The lack of induction of hsp-3 or enpl-1 was not due to the seam lineage being refractory to UPR signaling since we detected induction of both in seam cells when ER stress was induced by treatment with the glycosylation inhibitor tunicamycin ( S3C Fig ) . We next asked whether UPR pathways were required for selective hsp-4 induction during seam-cell differentiation . We examined the expression of phsp-4::GFP reporter in starved L2 animals deficient for each of the three canonical UPR pathways by using loss of function alleles ( Fig 2C ) . These alleles were previously characterized as UPR-deficient and were shown to affect the expression of hsp-4 and other UPR target genes under both ER stress and basal conditions [45 , 46] . Surprisingly , phsp-4::GFP reporter was induced normally in seam cells despite inactivating mutations of ire-1/IRE1 or xbp-1/XBP-1 or deletions of pek-1/PERK or atf-6/ATF-6 ( Fig 2C ) . Mammalian ATF-6 and XBP-1 are both bZIP transcription factors , binding to similar DNA elements and capable of heterodimerization [47] . Genetic inactivation of each is well tolerated in C . elegans , but loss of both is larval lethal because of the degeneration of the intestine [26] . Thus , it is possible that they compensate for each other in the singly deficient backgrounds . To test this , we used feeding RNAi to down-regulate atf-6 expression in xbp-1–deficient animals . To avoid the possible complications of combining feeding RNAi with starvation , we scored phsp-4::GFP induction during differentiation of seam cells in the L4 stage . All scored ( n = 20 ) xbp-1 ( zc12 ) ;atf-6 ( RNAi ) animals had normal induction ( Fig 2D ) , despite being unhealthy and with patchy coloration in their intestines , which indicated that RNAi treatment was effective [26] . We could not completely exclude the possibility that a small amount of ATF-6 protein was still expressed in RNAi-treated xbp-1 ( zc12 ) animals . To address this , we thought to mutate the ER stress elements in the promoter of the hsp-4 reporter . hsp-4 promoter was previously found to contain two ER stress element-II–like elements and a putative XBP-1/ATF-6 ( cAMP response element [CRE]-like ) element [26] ( S4A Fig ) . The hsp-4 ER stress element-II–like elements ERSE-II , ATTGG-N ( 6 ) -CCACA , show some deviation from ERSE-II consensus sequence ATTGG-N ( 1 ) -CCACG/A , as well as from ERSE consensus CCAAT-N ( 9 ) -CCACG/A , where CCAAT or ATTGG is a recognition site for the transcription factor Nuclear Transcription Factor Y ( NF-Y ) , while CCACG/A is recognized by XBP-1 or ATF-6 [23 , 24] . In the hsp-4 promoter , the two ERSE-II–like elements and their flanking regions contain perfect reverse-complementary sequences such that the region containing these elements , from residue ( −584 ) to ( −742 ) , can form a highly stable stem–loop structure ( S4A Fig ) . Because of this unusual arrangement , we chose to delete , rather than mutate , this region . We found that deletion of the ERSE-II-like–containing region did not prevent the induction of hsp-4 reporter in differentiating seam cells ( S4B Fig ) . The second ER stress element , between nucleotides ( −243 ) and ( −269 ) , is located on the reverse strand ( S4A Fig ) and contains the TGACGTGT XBP-1/ATF-6 ( CRE-like ) element , with the core XBP-1 motif underlined . We mutated this element to gGggGTGT ( mutated residues in lower case ) in the promoter with a deleted ERSE-II–like region , thus eliminating both types of the known ER stress elements in this promoter [26] . In agreement with the lack of effect from deleting UPR transcription factors , elimination of ER stress elements from hsp-4 promoter did not prevent its induction in posterior daughter cells during seam-cell differentiation ( S4B Fig ) . Surprisingly , this double-mutant promoter was still responsive to induction by ectopically overexpressed spliced XBP-1 ( XBP-1s , in neurons [48] ) . It is possible that additional binding sites , distinct from the known XBP-1 site , exist in this promoter or that XBP-1s activates the mutant promoter through interaction with another transcriptional regulator . However , because of the data from xbp-1 ( zc12 ) ;atf-6 ( RNAi ) animals ( Fig 2D ) and the lack of induction of other UPR target genes ( S3A Fig ) , we favor the conclusion that induction of hsp-4 expression during differentiation of the seam cells is independent of the three canonical UPR branches . In addition to the UPR transcription factor XBP-1 , the transcriptional regulator BLIMP1 is involved in differentiation of many secretory cell types in mammals , as well as in promoting and maintaining stem cell identity [49] . The C . elegans orthologue , BLMP-1 , is necessary for formation of both adult and dauer alae [50] . Interestingly , the seam-cell divisions themselves are normal in blmp-1 mutants , suggesting that it only contributes to the postdifferentiation cell fate [51] . Thus , we thought to determine whether BLMP-1 has a role in regulating hsp-4 induction during seam-cell differentiation . Examination of Model Organism Encyclopedia Of DNA Elements ( modENCODE ) data [52] showed a strong binding peak for BLMP-1 on the hsp-4 promoter ( S4C Fig ) . This is likely to represent a true binding peak for two reasons: First , this site does not overlap with the extreme highly occupied target ( xHOT ) regions , which represent redundant and likely nonspecific binding of multiple transcription factors [53] . Second , we identified a sequence , TAAGAAAGCTCTCGAAAAGTC , which is homologous to the known interferon regulatory factor ( IRF ) elements , near the XBP-1/ATF-6 ( CRE-like ) element and within the modENCODE peak ( S4A Fig; see Materials and methods ) . Because the mammalian BLMP1 is known to bind with high affinity to the subset of IRF elements containing GAAAG [54] , we designate it as a putative BLMP-1–binding site ( S4A Fig ) . To determine whether the developmental induction of hsp-4 is dependent on BLMP-1 function , we down-regulated blmp-1 in sa191;phsp-4::GFP animals by RNAi . Under normal growth conditions , sa191 animals that do initiate the dauer program enter the L2d stage by 41 hours postgastrula . We found no effect of blmp-1 RNAi on this initial phsp-4::GFP reporter induction at 41 hours postgastrula in all animals that had L2d morphology ( Fig 3A , top row ) . However , by 42 hours , blmp-1 RNAi caused increased reporter fluorescence in seam cells of these animals , and by 46–47 hours , approximately half of blmp-1 RNAi animals ( n = 81 ) exhibited induction of the reporter in the lateral hypodermis ( Fig 3A and 3B ) . None of the control RNAi animals ( n = 63 ) induced hsp-4 reporter in the hypodermis at any point during the L2d stage . The induction level of phsp-4::GFP reporter in the hypodermis of blmp-1 RNAi animals was similar to that in seam cells , except for occasional one or few seam cells per animal that exhibited a much brighter further induction ( Fig 3 ) . We considered a possibility that the increase in fluorescence in hypodermal tissue resulted from redistribution of the diffusible GFP protein from posterior seam cells to the hypodermis , if blmp-1 RNAi caused defects in the seam–hypodermis boundary . However , the GFP fluorescence was contained within the strongly induced cell , and AJM-1::GFP outlines of the seam cells appeared intact ( Fig 3C and 3D ) . Together , these data suggest that the activity that induces the hsp-4 gene in predauer animals may in fact be triggered in both the hypodermal-fated and alae-fated lineages at this point in development , but the induction of hsp-4 may at the same time be repressed in the anterior , hypodermal-fated daughter cells by BLMP-1 . We asked whether down-regulation of blmp-1 would result in induction of other ER chaperone genes . We examined same set of reporters as in S3A Fig , and found that while hsp-3 , encoding the second BiP homologue , was indeed weakly induced in seam cells of sa191 L2d animals after blmp-1 RNAi , the UPR targets enpl-1/GRP94 and cnx-1/calnexin were unaffected ( S5 Fig ) . Thus , removal of BLMP-1-mediated suppression is not sufficient for the induction of general UPR target genes in the differentiating seam cells , and a BiP-specific inductive factor appears responsible for this developmentally controlled expression of hsp-4 . The logic of anticipatory and selective ER chaperone induction during differentiation would suggest that the up-regulated chaperone is required for the specific secretory function of the resulting cell . Yet , BiP is considered to be a broad-specificity rather than client-selective chaperone , consistent with its global induction under folding stress conditions . We asked whether induction of hsp-4/BiP expression in differentiating alae-producing cells is important for the postdifferentiation function of these cells by examining the requirements for hsp-4 for cuticular structure . A GFP-tagged cuticular collagen , COL-19 , is expressed starting from the late L4 stage and is normally detected in evenly aligned circumferential pattern , as well as in the longitudinal linear structures of adult alae [55] ( Fig 4A–4C ) . Down-regulation of hsp-4 by RNAi resulted in a disrupted circumferential pattern in young adults , such that 46% ( n = 13 ) of animals contained large gaps between the COL-19::GFP fibers overlaying the lateral hypodermis and those overlaying the ventral/dorsal hypodermis ( Fig 4A and 4B ) . In contrast , only 8% ( n = 12 ) of control RNAi animals had gaps in the cuticle ( Fig 4A ) . Furthermore , hsp-4 RNAi caused occasional areas of disorganization of the longitudinal linear pattern , with COL-19::GFP being deposited in a “spaghetti-like” fashion in some animals ( Fig 4B ) . Similar large gaps and disorganization are known to be caused by mutations in proteins involved in cuticle synthesis and molting [56 , 57] . Because hsp-4 and hsp-3 genes both encode BiP homologues and share a high degree of sequence homology , the hsp-4 RNAi may target both genes . Thus , we confirmed the HSP-4 requirement for the cuticle using hsp-4 deletion allele gk514 . Unstressed gk514 animals are phenotypically normal and have normal dauer entry rates [36] , presumably because of the stress-related role of HSP-4 and because the second BiP homologue , HSP-3 , is functionally redundant with HSP-4 [27] . Yet , we found that deletion of hsp-4 resulted in defects in COL-19::GFP deposition in young adults: the levels of COL-19::GFP over the lateral hypodermis overlaying the seam cells were strongly reduced , and the protein was absent in the longitudinal areas underlying the forming alae ( Fig 4C ) . Because hsp-4 induction can be triggered by the dauer developmental program , we examined the alae in dauer cuticles . Because COL-19 is not expressed at these larval stages , we visualized the dauer alae by differential interference contrast ( DIC ) microscopy . We found that deletion of hsp-4 resulted in defective formation of dauer alae , with reduced number of ridges and visible gaps in the ridges in all examined dauers ( n = 5 ) ( Fig 4D ) . Finally , we asked whether the structural defects affected the function of the cuticle . Mutations that disrupt cuticle integrity cause it to become permeable to fluorescent dyes , such as Hoechst stain . We found that hsp-4 ( gk514 ) deletion caused increase in the cuticle permeability: although the degree of dye penetration was lower than that in the known leaky-cuticle bus-8 ( e2887 ) mutant strain [58 , 59] , twice as many hsp-4 ( gk514 ) as WT N2 animals took up the dye ( Fig 4E ) . Thus , HSP-4 protein is required for the formation of the structurally and functionally intact cuticle , and this function of HSP-4 is not compensated for by HSP-3 .
The importance of UPR signaling in the general expansion of ER biosynthetic capacity during differentiation of secretory cells has been firmly established . However , it is unclear whether UPR coordinates the repertoire of ER chaperones to the folding needs of specific secretory cell types . We find that during differentiation of the alae-secreting cells , induction of the stress-responsive C . elegans BiP homologue , HSP-4 , bypasses the requirement for the canonical UPR signaling . Instead , HSP-4/BiP is induced by specific developmental programs—the dauer program or L4-stage-to-adult transition . Interestingly , induction of HSP-4/BiP in the hypodermal-fated cells is repressed at the same developmental stage by a known transcriptional regulator of development , BLMP-1/BLIMP1 , which also regulates differentiation of many secretory cell types in mammals . Importantly , induction of HSP-4 is not required for the differentiation of alae-secreting cells , per se , but is essential for the secretory function of these cells postdifferentiation . Sharp increases in BiP expression are often interpreted to indicate activation of the UPR and are thought to require ER stress elements in its promoter . However , under some pathological conditions , the regulatory mechanisms differ from this expectation . For example , increased expression of BiP and other chaperones during acute-phase response in mice with bacterial infection was regulated by binding of Signal Transducer and Activator Of Transcription 3 ( STAT3 ) directly to Gpr78 promoter [60] . Even under conditions of ER stress , caused by limitation of specific folding resources , induction of the canonical UPR target proteins including BiP can be either dependent or independent of UPR signaling [61] . Induction of hsp-4 expression in our study may indicate the action of a non-UPR transcription factor ( s ) , specific to the differentiation of alae-secreting cells . Another possibility is that a member of stress-responsive family of CREB ( CRE binding ) / ATF transcription factors , other than ATF-6 , is involved in regulating hsp-4 , similar to the regulation of Grp78 by ATF-4 during translation block [62] or involvement of another CREB transcription factor , Old Astrocyte Specifically-Induced Substance ( OASIS ) , in bone development [63] . However , mutating the CRE-like element in hsp-4 promoter did not disrupt the pattern of its induction , making this possibility unlikely . Finally , it is possible that hsp-4 is regulated during alae-secreting cell differentiation by a UPR transcription factor binding to an element other than the known ER-stress elements . This is an intriguing possibility since we do see induction of the mutant hsp-4 transcriptional reporter , lacking ER-stress elements , in neurons expressing spliced XBP-1 . The transcriptional induction of hsp-4 gene during differentiation of alae-secreting cells appears to be anticipatory relative to its putative client protein ( s ) . In L4 stage larvae , hsp-4 induction precedes the fusion and differentiation by several hours , while in predauers , we first detect the phsp-4::GFP reporter fluorescence soon after the last asymmetric division , before the anterior daughters move away from the seam . Considering the time needed to accumulate the fluorescent signal to detectable levels , activation of hsp-4 promoter is likely to occur even earlier . Even more strikingly , most sa191 animals will have returned to reproductive development at 20 °C instead of entering dauer and thus will not have finished differentiation of the alae-secreting cells; yet , all sa191 L2d animals strongly induce the hsp-4 reporter . The anticipatory induction here parallels the regulatory logic of the UPR induction during differentiation of secretory cells in mammals [19 , 20] , even though it appears to bypass the UPR . It would be interesting in the future to understand whether this difference is an example of different organisms or even different cell types using different routes to achieve the same goal ( timely increase in the necessary chaperone ) , or whether it reflects the difference between the need for the generic expansion of ER capacity versus the need to match the chaperone repertoire to the cell-specific proteome . HSP-4 induction is also not simply a consequence of asymmetric divisions because it is not induced after asymmetric seam-cell divisions in other developmental stages . Together with the absence of a generic UPR in these cells and with apparent independence of hsp-4 induction from the canonical UPR signaling , these data suggest that the early differentiation program that determines the identity of the posterior daughter cell is able to directly regulate this chaperone . This phenomenon is similar to the recently reported regulation of some cytoplasmic chaperones , required for the myofilament formation , by the helix–loop–helix protein HLH-1 ( a C . elegans orthologue of the Myogenic Differentiation transcription factor MyoD ) during embryonic muscle differentiation in C . elegans [64] . Similarly , the transcription factor Kruppel-like zinc finger protein 9 ( Zf9 ) that regulates collagen-specific chaperone HSP47 in fibrotic tissues is capable of binding a collagen promoter [21] . In these examples , the chaperones and their clients are regulated by the same transcription factor ( s ) . While we do not know whether a similar regulatory logic applies to the developmental regulation of the hsp-4 gene since transcription factors that specify the identity of the alae-secreting cells are unknown , our data do show that HSP-4 function is specifically required for the alae-secreting function of these cells postdifferentiation . Another aspect of the observed temporal control of hsp-4 transcription is its repression in the lateral hypodermis by BLMP-1 . Silencing of blmp-1 resulted in hsp-4 reporter expression in the anterior , hypodermal-fated daughter cells . Interestingly , this ectopic induction of hsp-4 in blmp-1 ( RNAi ) animals was observed only following the last division before alae-secreting cells are specified but not during asymmetric seam-cell divisions in other larval stages . The most facile explanation for such pattern of induction is existence of a positive inductive signal that is activated in the entire seam lineage at the onset of differentiation in L2d or L4 stage animals . In such a case , the combination of this inductive signal in the entire seam lineage with the repressive action of BLMP-1 in the hypodermal-fated cells may explain the selective hsp-4 induction in the posterior daughter cells . This is also consistent with de-repression of hsp-4 in the anterior , hypodermal-fated daughters in blmp-1 ( RNAi ) animals . Alternatively , the inductive signal may be specific to the posterior daughters as they assume the alae-secreting fate . In this case , hsp-4 induction in the hypodermis upon blmp-1 RNAi may reflect de-repression of a different factor that can induce hsp-4 expression . Because our promoter sequence analysis suggests possible direct binding of BLMP-1 to the hsp-4 promoter , we favor the former scenario . The dependence of the cuticle functionality on HSP-4 is somewhat surprising , given that the second BiP homologue , HSP-3 , is basally expressed in seam cells . BiP is considered to be a broad-specificity chaperone , capable of binding the majority of proteins that are folded in the ER [65 , 66] , and HSP-4 protein in most C . elegans cells is only induced under folding stress conditions , further supporting the idea of nonselectivity of its function . Yet , its induction specifically in the alae-secreting cell precursors , and the alae and cuticle defects seen with its deletion suggest a unique cell-specific requirement for this chaperone . One possibility is that certain secreted proteins expressed in these cells require HSP-4 , but not HSP-3 , for their folding and secretion . Although HSP-4 and HSP-3 proteins are highly conserved and thought to be largely functionally redundant [27] , they are not identical , with 83% identity and 97% similarity in their peptide-binding domains . Another , less likely , possibility is that HSP-4 has a unique function in these cells , unrelated to its binding of unfolded proteins . While the lack of hsp-4 induction has clear negative consequences for the cuticle secretion , the functional importance of hsp-4 repression by BLMP-1 in the hypodermal-fated cells is not immediately clear . Deletion of blmp-1 was previously shown to cause defective formation of alae , and blmp-1–deficient animals have oxidative-stress–sensitive cuticles and a dumpy appearance [67 , 68] , indicating global cuticle defects . Because the blmp-1 deletion is not cell-specific , we do not know whether these defects stem from functional deficiencies in the lateral hypodermis , where hsp-4 is de-repressed in the absence of BLMP-1 . However , it is possible that inappropriate induction of HSP-4 in hypodermal cells results in their decreased ability to secrete proteins because overexpression of a broad-specificity BiP chaperone under nonstress conditions and in the absence of a high-affinity client may nonspecifically stabilize folding intermediates and decrease rates of folding in the ER . Indeed , overexpression of BiP in Chinese hamster ovary ( CHO ) cells blocks secretion of a subset of proteins , while overexpression of its cytosolic counterpart , HSP70 , causes developmental delays in Drosophila [69 , 70] . In addition to individual chaperones , ectopically increased UPR activity can be detrimental to animal development [71] , and different tissues may have different tolerance levels [48] . The integration of developmental and stress signaling is emerging as an important contributor to multiple aspects of metazoan biology [5 , 72 , 73] . UPR signaling pathways can be specifically activated in the absence of ER stress , for example , by growth factor signaling or infections: IRE1 can be activated by internalized Vascular Endothelial Growth Factor ( VEGF ) receptor 2 through direct interaction [74] , while Toll-like receptors ( TLRs ) in macrophages activate it by a NADPH oxidase-dependent signal [75] . Interestingly , the TLR-induced IRE1 activation does not result in chaperone expression or ER expansion , as would be expected from stress-activated IRE1 , but rather promotes sustained production of inflammatory mediators [75] . Similarly , a canonical UPR transcription factor , ATF-6 , and other members of the CREB/ATF family respond to extracellular cues in osteoblasts and odontoblasts by regulating expression of collagens and other matrix-forming proteins [63 , 76] , presumably by interacting with cell-type–specific transcriptional machinery . Thus , physiological processes can not only induce the generic UPR activation but can also trigger specific UPR pathways and , remarkably , control their outcomes . Our data show that , in addition , developmental signals can control the repertoire of induced chaperones directly , bypassing the UPR . Delineating the mechanisms integrating the physiological and stress signaling will thus be instrumental to further our understanding of the regulation of development , the pathogenesis of developmental disorders , and the mechanisms that maintain organismal homeostasis .
Standard methods were used for worm culture and genetic crosses [77] . After crosses , strains were confirmed by PCR and restriction digest or sequencing . Animals were synchronized by picking gastrula-stage embryos from well-fed uncrowded plates . The following strains were obtained from the Caenorhabditis Genetics Center ( CGC ) : SJ4005 ( zcIs4[phsp-4::GFP] ) , SJ30 ( ire-1 ( zc14 ) II; zcIs4[phsp-4::GFP] ) , SJ17 ( xbp-1 ( zc12 ) III ) , RB772 ( atf-6 ( ok551 ) X ) , RB545 ( pek-1 ( ok275 ) X ) , VC1099 ( hsp-4 ( gk514 ) II ) , JT191 ( daf-28 ( sa191 ) V ) , RW11606 ( unc-119 ( tm4063 ) III; stIs11606 [egl-18a::H1-wCherry + unc-119 ( + ) ] ) , SD1546 ( ccIs4251 I; stIs10166 [dpy-7p::HIS-24::mCherry + unc-119 ( + ) ] ) , PS3729 ( unc-119 ( ed4 ) III; syIs78[AJM-1::GFP + unc-119 ( + ) ] ) , CB1372 ( daf-7 ( e1372 ) III ) , CF1038 ( daf-16 ( mu86 ) I ) , CB1370 ( daf-2 ( e1370 ) III ) , PS3551 ( hsf-1 ( sy441 ) I ) , TP12 ( kaIs12[COL-19::GFP] ) , and CB6208 ( bus-8 ( e2887 ) X ) . BC10514 ( dpy-5 ( e907 ) I; sEx10514 [rCesT05E11 . 3::GFP + pCeh361] ) and BC10700 ( dpy-5 ( e907 ) I; sEx10700 [rCesZK632 . 6::GFP + pCeh361] ) strains and a strain expressing phsp-3::YFP were a gift from the Morimoto lab ( Northwestern University , Evanston , IL , USA ) . WT ( N2 ) animals were a subclone of N2Bristol from the Morimoto Lab . Primers used for PCR or to sequence-verify crosses with UPR mutant alleles were as follows . For hsp-4 ( gk514 ) : hsp-4_Ext_F:CCTCCGATTACTCCTGCTTG; hsp-4_Int_F:GTTTGATGCTGGGTTGACAAAG; and hsp-4_Ext_R:GAGTCTTCAAGAATGGGCGAG . For ire-1 ( zc14 ) : ire-1 ( zc14 ) _F:ATCAGCCAACGACCAATCTGC and ire-1 ( zc14 ) _R:GAAGCTTTGGATGGGCGAATAG; the mutation was confirmed by digesting the PCR product with BstBI . For atf-6 ( ok551 ) : atf-6_Ext_F:ATACCGCGTCAAGGAATCAC; atf-6_Int_R:TTAAATCTCACGCAGGCAAG; and atf-6_Ext_R:AATTGGCCAGTCCCTGTCAC . For pek-1 ( ok275 ) : pek-1_Ext_F:TCGGAGCACACGATTTCTCG; pek-1_Int_R:CTTGTGGACCCGGAGATACG; and pek-1_Ext_R:CTGAGCACATCTGACGTAAG . To generate L2d stage animals in daf-28 ( sa191 ) genetic background , appropriate strains were grown at 20 °C under noncrowded/noncontaminated conditions on fresh plates seeded with OP50 Escherichia coli for at least 2 generations . 20–40 YA animals were then picked to fresh plates , and L2d stage larvae were picked among their progeny based on their morphology [33 , 36]: L2d animals are radially constricted , although to a lesser extent than dauers; they are larger than L2 animals but with the germline morphology of L2 stage; they have a uniformly dark intestine; and they exhibit slow pharyngeal pumping . A similar procedure was used for other developmental stages . To generate L2d animals by starvation/crowding , parents were placed on fresh plates seeded with OP50 E . coli at 20 °C , and plates were examined daily until there was no food left . Predauers were imaged 1–2 days later . phsp-4::GFP-containing plasmid ( #21896 ) was obtained from Addgene ( Watertown , MA , USA ) . A 54-bp vector-derived region between the end of the hsp-4 promoter and the start of the GFP , which incidentally contained a PQM-1/DAE-like element , was removed using restriction enzyme PpuMI . To construct the phsp-4-ER stress element-II ( del ) ::GFP transgene , 171 bp of the ER stress element-II–like region was deleted using Q5 Site-Directed Mutagenesis Kit ( New England Biolabs , Ipswich , MA , USA ) . To construct phsp-4-ER stress element ( del ) -xbp-1 ( mut ) ::GFP , the XBP-1/ATF-6 element in the ER stress element ( del ) promoter was mutated from GATGACGTGT to GAgGggGTGT . All constructs were verified by sequencing ( Macrogen , Rockville , MD , USA ) . The mutagenesis primers were as follows: ESERII_del_F:CGGGTCTCTAAGGAAAGGATTC; ESERII_del_R:CCCAGTTGGACATCGGGTC; XBP_1_F:CCTCTCCGATAAGTACACGTTGC; XBP_1_R:GGGTGTATTAGTGCTGGAGAAATC . Transgenes were injected as a mix of 20 ng/μL plasmid DNA and 80 ng/μL sonicated salmon sperm DNA . The RNAi clones were from the Ahringer library . For RNAi experiments , animals were grown for one or two generations on 0 . 4 mM-IPTG–containing plates , spotted with designated RNAi bacteria . For atf-6 RNAi , xbp-1 ( zc12 ) ;phsp-4::GFP animals were imaged at the L4 stage . For hsp-4 RNAi , COL-19::GFP deposition was examined in young adult animals . For blmp-1 RNAi , 15–20 L4 stage progeny of RNAi-treated daf-28 ( sa191 ) ;phsp-4::GFP parents were placed on fresh RNAi plates , gastrula-stage embryos were picked 1–2 days later , and L2d stage animals were scored 41–46 hours later . All experiments were repeated with a different population of animals 2–3 times . Prior to the experiments , the RNAi plates were tested for the expected phenotypes , such as , for example , larval arrest of second generation of xbp-1 mutant animals on atf-6 RNAi , to ensure proper induction of the RNAi . A sequence TAAGAAAGCTCTCGAAAAGTC is located within the modENCODE BLMP-1 peak on the hsp-4 promoter , near the XBP-1/ATF-6 ( CRE-like ) element ( S4A and S4C Fig ) . This sequence is homologous to the known IRF binding site . The sequence contains a perfect match to the IRF consensus sequence GAAAG/CT/C found in the MHC class I promoter ( underlined ) and partial matches ( in bold ) to an interferon-stimulated response element , found in most interferon-inducible promoters ( A/GNGAAANNGAAACT ) , and to the positive regulatory domain ( PRD ) element found in the INF-β promoter ( G ( A ) AAAG/CT/CGAAAG/CT/C ) [78] . Because the mammalian BLMP1 is known to bind with high affinity to the subset of these elements containing GAAAG [54] and the IRF consensus sequence in the hsp-4 promoter contains this core sequence , we designated it as a putative BLMP-1-binding site ( S4A Fig ) . | During differentiation , cells that specialize in secretion of proteins , such as antibody-secreting B cells , prepare for the onset of secretory function by expanding the size of the major secretory organelle , the endoplasmic reticulum ( ER ) , and by increasing the expression of molecular chaperones and folding enzymes . This pre-emptive expansion of the ER depends on activation of the ER stress response pathways and is required for the secretory phenotype . In addition , cells may also need to up-regulate a selected subset of chaperones because different secreted proteins may require different chaperones for their folding and secretion . Except in specialized cases , how this selective up-regulation is achieved , and whether it depends on the ER stress pathways , is not well understood . Using Caenorhabditis elegans , we find that a chaperone BiP/HSP-4 , which is usually induced in most cells by stress , is selectively induced during differentiation of stem cells into the alae-secreting cells while being repressed in their sister lineage , the hypodermal cells . We find that induction of this chaperone is independent of the known ER stress pathways , while its repression requires a known regulator of development in mammals , BLIMP1/BLMP-1 . The pre-emptive induction of BiP/HSP-4 is important for the integrity of secreted alae and cuticle , suggesting that a general molecular chaperone that is a canonical target of ER stress pathways can be selectively regulated by development to ensure the quality of secreted proteome and functionality of the cells postdifferentiation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"skin",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"rna",
"interference",
"integumentary",
"system",
"caenorhabditis",
"gene",
"regulation",
"cell",
"cycle",
"and",
"cell",
"division",
"regulatory",
"proteins",
"cell",
"processes",
"endoplasmic",
"reticul... | 2019 | HSP-4/BiP expression in secretory cells is regulated by a developmental program and not by the unfolded protein response |
Many unicellular organisms live in multicellular communities that rely on cooperation between cells . However , cooperative traits are vulnerable to exploitation by non-cooperators ( cheaters ) . We expand our understanding of the molecular mechanisms that allow multicellular systems to remain robust in the face of cheating by dissecting the dynamic regulation of cooperative rhamnolipids required for swarming in Pseudomonas aeruginosa . We combine mathematical modeling and experiments to quantitatively characterize the integration of metabolic and population density signals ( quorum sensing ) governing expression of the rhamnolipid synthesis operon rhlAB . The combined computational/experimental analysis reveals that when nutrients are abundant , rhlAB promoter activity increases gradually in a density dependent way . When growth slows down due to nutrient limitation , rhlAB promoter activity can stop abruptly , decrease gradually or even increase depending on whether the growth-limiting nutrient is the carbon source , nitrogen source or iron . Starvation by specific nutrients drives growth on intracellular nutrient pools as well as the qualitative rhlAB promoter response , which itself is modulated by quorum sensing . Our quantitative analysis suggests a supply-driven activation that integrates metabolic prudence with quorum sensing in a non-digital manner and allows P . aeruginosa cells to invest in cooperation only when the population size is large enough ( quorum sensing ) and individual cells have enough metabolic resources to do so ( metabolic prudence ) . Thus , the quantitative description of rhlAB regulatory dynamics brings a greater understating to the regulation required to make swarming cooperation stable .
Cells can cooperate as multicellular populations and impact their environment in ways that would not be possible for an individual cell . This strength in numbers is observed in many natural populations of unicellular microbes and can be leveraged in engineered systems for synthetic biology [1] . While cooperation helps a population as a whole , natural selection acts at the level of individual cells , which makes cooperation susceptible to cheating [2] . The potential exploitation by non-cooperator cells ( cheaters ) that benefit from cooperation without participating in it threatens the robustness of multicellular systems both natural and synthetic [3–7] . However , the potential for cheating can drive the evolution of molecular mechanisms capable of effectively regulating cooperative traits [8] . Investigating the natural mechanisms that prevent cheating can reveal design principles underlying the robustness of multicellular systems . In bacteria , the expression of cooperative genes is often regulated by density-dependent signaling systems , called quorum sensing , that detect the transition from a unicellular to a multicellular state [9–11] . Quorum sensing works by the secretion of small signaling molecules , called autoinducers , which accumulate in the extracellular space in a density-dependent manner . Cells sense the extracellular concentration of these molecules and use it as a proxy for population density . Quorum sensing is used to regulate the expression of cooperative multicellular traits such as bioluminescence [9] , biofilm formation [12] and virulence factors [13–16] . The ubiquity of quorum sensing across bacterial species suggests a range of applications for this circuitry as a design principle [1 , 17] . Although quorum sensing can provide a robust benefit in changing conditions [18] , it is vulnerable to cheating . Cheater cells can take advantage of quorum sensing autoinducers or public goods that are regulated by quorum sensing [3 , 19 , 20] . Cooperative genes regulated by quorum sensing can also be sensitive to nutrient conditions , suggesting that metabolic information is integrated into the decision to cooperate [21–28] . Integrating metabolic information with quorum sensing offers a possible mechanism to prevent cheating , as cells can only cooperate when they have the appropriate nutritional resources to do so , reducing the cost of cooperation to the individual cell . The opportunistic pathogen Pseudomonas aeruginosa secretes massive amounts of rhamnolipid biosurfactants in order to move collectively over surfaces , a phenomenon known as swarming [29–33] . Swarming provides a benefit at the population level , enabling cells in a colony to disperse over wide areas and grow to large numbers . However , rhamnolipid production can be a significant cost to the individual cell since it requires an investment of carbon that might otherwise be used for cell growth and division . Non-producing cells can exploit secreted rhamnolipids , which makes the trait vulnerable to cheating . Appropriate regulation of rhamnolipid synthesis is therefore crucial to prevent cheating and make cooperation stable . The rhamnolipid synthesis operon rhlAB is regulated by a quorum-sensing cascade composed of lasI/lasR followed by rhlI/rhlR [34 , 35] . Although quorum-sensing regulation is necessary for rhlAB expression , it is not sufficient . Expression only occurs when the bacteria have carbon in excess of that needed for growth [23 , 36] . The use of metabolic signals to trigger expression of cooperative genes , in this case excess carbon triggering expression of rhAB , is termed metabolic prudence . This native regulation of quorum sensing and metabolic prudence prevents exploitation by cheaters and stabilizes cooperation [36 , 37] . The production of both mono- and di-rhamnolipids requires the function of three enzymes RhlA , RhlB and RhlC . RhlA converts B-hydroxyacl-ACP , an intermediate from fatty acid biosynthesis , into B-hydroxyalkanoyl-B-hydeoxyalkanoyl ( HAA ) [31] . RhlB and RhlC are required for the addition of rhamnose groups to produce mono- and di-rhamnolipids , respectively . RhlA is the rate-limiting enzyme and is required for any rhamnolipid production by the cell [38] . Tracking the activity of the rhlAB promoter therefore serves as a reporter for when a cell has made the decision to commit carbon to rhamnolipid production . Although synthetic constitutive expression of rhlAB can result in rhamnolipid production and enable swarming , this synthetic rhamnolipid regulation severely impacts P . aeruginosa fitness and makes cooperation cheatable [36 , 37] . To understand how the native circuitry allows P . aeruginosa to produce rhamnolipids without compromising fitness we construct a quantitative picture of rhlAB regulatory dynamics . We combine quantitative experiments with mathematical modeling to systematically probe P . aeruginosa growth behavior and rhlAB promoter activity as a population transitions between different nutrient levels and population densities . We find that a classical Monod model cannot explain growth under starvation of nitrogen or iron and that internal pools of these nutrients sustain growth during starvation . Utilizing our understanding of growth in different nutrient limitations we quantitatively analyze rhlAB promoter activity and find that there are sufficient signals of excess carbon during exponential phase to trigger rhlAB expression in a density-dependent manner . We also find that although the limiting nutrient governs the qualitative behavior of promoter activity during starvation , starvation-induced activity is also scaled by population density . Together these results suggest a supply-driven activation that continually integrates metabolic prudence with quorum sensing in a non-digital manner . These results support the view that P . aeruginosa cells express rhamnolipids prudently to reduce fitness costs and prevent cheating and add details about the nuances of this regulation under different conditions .
We analyzed the timing of rhlAB promoter activity directly in swarming colonies using fluorescent imaging and time-lapse video using a PrhlAB-gfp reporter strain ( Fig 1A ) [39] . A colony of P . aeruginosa inoculated on a swarming plate first grows without moving until it reaches a certain critical size at ~5 h ( Fig 1B ) . During the 2–5 h period expression at the colony center coincides with a decrease in growth rate that could be due to local nutrient depletion ( Fig 1C and 1D ) . GFP levels continue to increase until ~5 h when we can observe a translucent ring of secreted rhamnolipids around the colony by eye . This is followed by the appearance of motile swarming tendrils shooting out from the colony . The secreted rhamnolipids lubricate the agar surface and allow the colony to slide over it . In order to probe rhlAB expression more systematically , we turned to a batch culture system using shaken liquid media in a microtiter-based assay . This way we can simultaneously assess population density and gene expression using OD ( optical density ) and the PrhlAB-gfp reporter respectively . Bacterial growth curves are typically described by four phases: lag phase , exponential phase ( sometimes called “log” phase ) , stationary phase , and death phase . The definition of stationary phase often includes qualitatively distinct sub-phases of slowed growth and no growth , which can make analyzing responses to nutrient starvation difficult . To facilitate our analysis , we separate time series into three phases after lag phase . Phase I begins when the population becomes detectable by absorbance at 600 nm ( OD , for optical density; detectable at 0 . 01 OD in this study ) and grows exponentially at its maximum rate , μmax . During this initial period , the cells have all nutrients required for biomass synthesis and thus achieve balanced growth . The start of phase II is defined by growth limitation , where an essential nutrient runs out and the population growth rate has slowed below μmax . Phase III is when population density stops increasing and may actually decay . Fig 1E shows a representative growth curve with all three growth phases . The PrhlAB-gfp construct enables tracking of rhlAB expression throughout the different phases of growth with high time resolution ( Fig 1F ) . We corrected for P . aeruginosa secreted products that fluoresce in the GFP detection wavelengths , thus generating a compensated GFP signal from the PrhlAB promoter over time ( S1 Fig and Correction of Autofluoresence in the gfp Signal ) . Using population density and population level measurements of GFP assumes that rhlAB expression is homogenous across the population . To test if this assumption is valid , we used microscopy to measure single-cell expression levels at different stages of growth ( S2 Fig ) . The up-regulation of rhlAB was simultaneous across the population rather than bimodal . Therefore , we concluded that the population-level measurements could be used to probe expression dynamics . To better understand how a population responds to the entry into starvation we first analyze growth behavior in different limiting conditions . Our first growth limitation experiments set carbon as the growth-limiting nutrient in the media . We used varying concentrations of the carbon source ( glycerol ) and added a nitrogen source ( ammonium sulfate ) and iron ( iron ( II ) sulfate ) in excess . In these conditions , the population density and the length of phase I have a dose dependency with the initial amount of carbon , confirming that carbon is indeed the limiting nutrient ( Fig 2A ) . Each growth curve follows the same phase I ( exponential growth ) with identical μmax values ( 0 . 33 h-1 ) . We observed that once the carbon in the media is fully consumed , cell growth stops abruptly and the population shifts sharply from phase I to phase III ( decay ) without going through a period of slowed growth ( phase II ) . We carried out additional growth experiments , now in limiting concentrations of nitrogen . As before , population density and phase I time scale with initial nitrogen concentration , confirming that nitrogen is the limiting nutrient ( Fig 2B ) . In nitrogen starvation the growth rate drops at the end of phase I but , unlike carbon starvation , population density continues to increase until the end of our observation period ( approximately 45 hours ) . Throughout this period of slowed growth ( phase II ) , the growth rate is continually decreasing . Finally , we conducted experiments with iron as the growth-limiting nutrient . Increasing the amount of supplemented iron in the media increases both the population density and phase I , confirming that iron is the limiting nutrient ( Fig 2C ) . Interestingly , the behavior under iron-limiting conditions was qualitatively distinct from both carbon and nitrogen limitation . In iron-limited growth there is still a transition from phase I to phase II , however this transition is gradual . The growth rate continually slows from μmax , unlike the abrupt transition to phase II seen in nitrogen limitation . In the iron limitation titration , there is an uneven spacing between the curves; doubling the supplemented iron did not result in a doubling of total population density . This can be explained by a constant yield for iron ( YFe ) and the presence of trace iron in the growth media . If trace iron is present then doubling the amount of supplemented iron will not double the total amount of iron in the system , but rather add to the trace amount already present . To test this explanation , we used the population density data from the iron titrations ( Fig 2C ) , which fell on a line with slope YFe ( S3C Fig ) , to calculate the level of trace iron present in our non-iron supplemented media , Fe0 = 1 . 4 x 10–5 gFe/L . To additionally confirm the presence of trace iron in the media , we grew bacteria in media with no iron supplemented . As expected , the population was able to sustain a phase I ( S4 Fig ) in this medium and grew to the population density predicted by our trace iron calculation ( S3C Fig ) . See S1 Table for information regarding biological and technical replicates . In order to investigate how the nutrient levels experienced by cells influence their growth , we created a mathematical model of P . aeruginosa growth kinetics . Using the data from the experiments described above where we manipulated the media composition such that carbon , nitrogen , or iron become limiting , we were able to construct a kinetic model based on mass conservation , where biomass production is a function of these three nutrients . Central to this model was the calculation of the yields of biomass produced ( in units of OD ) per amount of carbon ( YC ) , nitrogen ( YN ) and iron ( YFe ) consumed ( S3 Fig ) . One benefit of this properly calibrated model is the calculation nutrient levels at any given time in the growth curve , which could not be directly measured with our assay . Bacterial growth limitation due to depletion of an essential nutrient is commonly modeled using Monod kinetics , which requires two parameters: μmax , the maximum specific growth rate , and Ks , the half-saturation constant [40 , 41] . In all of our nutrient limitations and titrations we observe the same μmax in phase I independently of the initial nutrient concentrations . In order to satisfy this constraint , a Monod model would need to have a Ks value for each nutrient that is significantly below our lowest titration value ( Ks<<0 . 063 gC/L , Ks<<0 . 0078 gN/L and KS<<1 . 4 x 10–5 gFe/L for carbon , nitrogen and iron respectively ) . In a Monod model , such a low Ks will give a very sharp transition from maximal growth to practically no growth . For carbon limitation we do observe this sharp transition behavior . This behavior is somewhat unexpected as the population might be predicted to slow in growth rate as carbon becomes increasingly scarce . This suggests that the half saturation constant value for carbon is indeed Ks<<0 . 063 gC/L and in order to measure its actual value we would need to monitor population density at OD values below the detection limit of our growth curve assay ( OD = 0 . 01 ) . Therefore , we instead model carbon ( C ) consumption as a step function with a constant yield YC ( S3A Fig and Eq ( 2 ) ) . This model recapitulates the growth behavior observed in carbon limitation media ( Fig 2D ) . Nitrogen and iron limitation growth curves , on the other hand , are inconsistent with a Monod model even with a Ks value below the lowest titration concentration . Although there is a sharp transition in nitrogen limitation from phase I to phase II , the sustained growth in phase II is incompatible with a Monod model . A Monod model with a low Ks also cannot explain the gradual slowdown in growth rate observed in iron limitation . There are a few possible explanations for the observed curvature in the nitrogen and iron limitations that we can exclude based on the data . Firstly , consistent phase II behavior across titrations excludes the possibility of a toxic product accumulating , as the populations with higher cell densities are not more severely affected . For nitrogen , the yield calculations predict no contaminating trace nitrogen amounts ( S3B Fig ) . The possibility of a contaminating trace nitrogen that is used only in phase II is also eliminated as this would result in the populations with lower cell densities growing much more than the populations with higher cell densities for a given amount of trace nitrogen , and instead phase II behavior is consistent across the titrations . The nutrient source for phase II growth in nitrogen and iron limitation must scale with population density and is independent of the starting nutrient concentration . A model that fits these criteria is one where the transition from phases I to II represents a switch in cellular metabolism from growth on extracellular nitrogen or iron ( upon complete depletion of the limiting nutrient from the media ) to intracellular nitrogen or iron [42] . Such growth behavior has been observed before in the yeast Saccharomyces cerevisiae . S . cerevisiae could grow in the absence of extracellular nitrogen by using nitrogen-rich intracellular biopolymers , presumably protein , and decreasing the nitrogen-to-carbon ratio of its biomass composition [43] . To determine if a model of P . aeruginosa utilizing internal nutrient pools for growth , like S . cerevisiae , could explain the observed behavior , we created a mathematical model to account for an internal nitrogen pool , internal iron pool , and trace iron . In this model , the cells consume extracellular nutrients to produce biomass and maintain homeostatic levels of intracellular nutrient pools while growing exponentially ( Ni , Fei ) ( Eq ( 3 ) and ( 5 ) ) . When the nitrogen or iron in the media is fully depleted , the cells switch to growth on the intracellular pool of the depleted nutrient , which then gradually decreases over time as the cells grow ( Eq ( 4 ) and ( 6 ) ) . Because we cannot directly measure the size of the internal pool of nitrogen or iron , we normalize it by the size of the pool during balanced growth; both Ni , and Fei are dimensionless . Each cell enters phase II with Ni or Fei = 1 and this internal pool is then depleted for cell growth and diluted through cell division . The current growth rate of the population , μ ( t ) , is dependent on the fraction of the internal pool in each cell . As the internal pool is depleted , growth slows from μmax ( Eq ( 7 ) ) . The kinetics of biomass ( X ) ( Eq ( 1 ) ) , growth and nutrient consumption of our different limiting nutrients are therefore given by: dXdt={μ ( t ) XifC>0 ( μ ( t ) −kd ) XifC=0 ( 1 ) dCdt=−1YCμ ( t ) XifC>0 ( 2 ) dNdt={−1YNμ ( t ) XifN>00ifN=0 ( 3 ) dNidt={− ( 1YNi+Ni ) μ ( t ) ifN=0andNi>00ifN>0 ( 4 ) dFedt={−1YFeμ ( t ) XifFe>00ifFe=0 ( 5 ) dFeidt={− ( 1YFei+Fei ) μ ( t ) ifFe=0andFei>00ifFe>0 ( 6 ) where μ ( t ) is the current specific growth rate . Nitrogen and iron use the same internal nutrient pool model . We find that in addition to accounting for trace iron in the media and internal nutrient pools of nitrogen and iron , we also must postulate that the maximum growth rate while using internal nitrogen is lower than μmax , here termed μmax´ . This postulation is required to explain the sharp transition from phase I to phase II observed in nitrogen limitation ( Fig 2B ) . As the transition in iron limitation is more gradual , this postulation is not required for growth on internal iron and thus we do not assume a switch from μmax to an alternative value ( μmax´ ) in iron limitation . Using this model , we are able to capture the dynamics observed in the experimental data ( Fig 2D–2F ) . The equation set that determines μ ( t ) for the different nutrient conditions is given below To test our internal nutrient model experimentally , we designed a nutrient depletion experiment ( Fig 2G ) . The experiment was only performed for carbon and nitrogen as trace iron prevents the depletion experiment for iron . Cells were grown to exponential phase in rich synthetic media ( Fig 2H ) and then harvested while still in balanced growth , washed , and separately inoculated in media lacking either carbon or nitrogen . As predicted by our model , the cells in media lacking nitrogen were able to grow and increase in population density ( Fig 2I ) whereas in the absence of carbon , the population transitioned immediately to phase III and population density started to decrease ( Fig 2I ) , supporting our model that total carbon depletion causes transition to phase III . These results are consistent with the model for phase II of nitrogen limitation and phase III of carbon limitation . We were unable to perform this experiment for iron limitation in the same manner due to the fact that medium without supplemented iron contains trace amounts of this nutrient ( S3C Fig ) . Nonetheless , the model with an internal iron pool is still able to describe the observed growth behavior in iron limitation ( Fig 2F ) . In summary , we developed a model of P . aeruginosa kinetics that successfully captures the observed growth dynamics . In total , the model has eight free parameters , which we were able to parameterize using growth curve experiments to derive the nutrient yields , YC , YN , and YFe , as well as kinetic parameters ( S5 Fig and S3 Table ) . We find that for carbon limitation , a Monod model ( where growth slows with decreasing concentration of the limiting nutrient ) with a very low Ks can explain the data , leading to a very sharp transition from exponential growth to no growth with virtually no slow down due to decreasing availability of carbon in the media ( Fig 2D ) . The growth model also reveals that growth behavior in both nitrogen and iron limitation , under these conditions , is incompatible with an explanation using a Monod model . However , the behavior observed in these limiting conditions can be explained using a model of intracellular nutrient pools ( Fig 2E and 2F ) , which is experimentally supported for nitrogen limitation ( Fig 2I ) . A model using these intracellular nutrient pools is required to accurately capture bacterial growth dynamics and recapitulate the observed growth rate . This phenomenological model can be used to constrain the mechanisms of rhlAB expression responses we observe in nutrient limitations . Rhamnolipid production is a dynamic process and changes in rhlAB expression coincide with transitions between growth phases , which are difficult to capture experimentally . Our growth model provides us with an understanding of the conditions cells experience throughout growth and the growth rate response to changes in the nutrient environment . We can use this understanding to interpret the expression response of rhlAB under these conditions . Our measurements of GFP driven by the rhlAB promoter ( Fig 3A–3C ) were taken simultaneously with the population density measurements ( Fig 2A–2C ) in the different limiting nutrient conditions . Using GFP ( Fig 3A–3C ) and OD measurements ( Fig 2A–2C ) , we calculated the promoter activity for rhlAB throughout the time series ( Fig 3D–3F ) , with compensation for GFP dilution by cell division ( see Promoter Activity Calculation ) . Note that promoter activity fluctuates more at the early time points because there is more noise at low OD measurements due to technical limitations of the equipment . We use our mathematical model of growth to systematically identify when the population exits phase I ( exponential phase ) . In Fig 3 , expression during phase I is shown in solid lines while expression that occurs after phase I ( phase II or III ) is shown in dashed lines . Our study and model of growth behavior in this media revealed that entry into phase II or III indicates starvation by the limiting nutrient , therefore promoter activity that occurs after phase I occurs during nutrient starvation . Previous work suggests that rhlAB expression occurs exclusively when carbon is in excess [23 , 36 , 44] . However , we unexpectedly observe appreciable rhlAB expression and promoter activity in carbon-limited media ( Fig 3A and 3D and 3G ) . Expression in carbon-limited media occurs during phase I and promoter activity drops to zero when the population enters phase III . ( The negative promoter activity values computed in carbon limitation media experiments ( Fig 3D ) are artifacts caused by a rapid increase in OD consistently occurring immediately before carbon starvation , coinciding with a shut off of rhlAB expression and beginning decay in the GFP signal . ) It is interesting that promoter activity increases throughout phase I , as it could be expected that there would be a constant level of expression during balanced growth and promoter activity would equilibrate quickly . However , increasing promoter activity during phase I is observed in all three limitation conditions ( Fig 3D–3F , solid lines ) . This phase I promoter activity could be due to the fact that even during exponential growth there is carbon available in excess of what is needed for biomass production , and this carbon can be dedicated to rhamnolipid synthesis . The promoter activity in phase I increases with population density in all conditions suggesting that this expression is density dependent ( Fig 3G–3I ) . By overlaying phase I of all conditions and plotting against population density we confirm a consistent slope across all conditions , suggesting that the mechanism driving the density dependent expression is the same in all three limitation conditions ( Fig 4A ) . In nitrogen limited media there is not a complete shutdown of expression after phase I . Instead , population level GFP continues to increase in phase II ( Fig 3B ) . The promoter activity drops , but is sustained , with a second peak of activity occurring before tapering off to near zero ( Fig 3E ) . In contrast , rhlAB promoter activity during phase II of iron limitation rapidly increases without an initial drop when the population enters phase II . Promoter activity of rhlAB decreases over time , but never reaches zero in our iron limitation condition ( Fig 3F ) . The level of promoter activity reached in iron limitation is much higher than that in nitrogen limitation when the highest titrations are compared directly ( Fig 4B ) . We also observe that even in nitrogen and iron starvation when promoter activity is sustained or induced , activity eventually shuts down , most likely due to the prolonged starvation experienced by the cell . Similar behavior is observed for all titrations within each limiting media across a wide range of population densities ( Fig 3G–3I ) . Therefore the limiting nutrient and the duration of starvation appear to be the main drivers of qualitative promoter behavior: shut down of activity in carbon starvation , sustained activity in nitrogen starvation , and induction of activity in iron limitation ( Fig 4B ) . Although the qualitative behavior is determined by the limiting nutrient , a closer examination of promoter activity during nitrogen starvation reveals that populations with a higher density have higher rhlAB promoter activity even during starvation ( Fig 4C ) . This suggests that the cells continue to use density-dependent information to modulate the dynamics of rhlAB promoter activity in nitrogen limitation . To determine if our current understanding of the rhlAB promoter response was sufficient to explain the observed dynamics , we derived a mathematical expression of PrhlAB-gfp promoter activity . The model contains three components . The first component implements the density-dependent up-regulation observed during balanced growth ( Fig 4A ) . The second component implements the observation from here , and in previous work , that rhlAB is expressed under nutrient starvation when growth is limited , but not halted ( Fig 4B , nitrogen and iron starvation ) [23 , 36] . Since promoter activity under starvation is also a function of population density ( Fig 4C ) , we use an additive model to describe the integration of population density and nutrient starvation induction . The third component implements a decrease in promoter activity that is observed under prolonged starvation by either nitrogen or iron ( Fig 4B ) . Prolonged starvation shutdown is implemented using Hill kinetics . Our mathematical model of bacterial growth predicts not only nutrient levels over time , but also the effect of these nutrient levels on growth rate . With an accurate prediction of μ ( t ) , from the growth model , and an understanding of the nutrient environment the cells experience in phases II and III , we are able to use growth rate as an indicator of cell starvation . Therefore , expression under starvation is implemented by induction when μ ( t ) falls below μmax ( μmax’ for nitrogen starvation ) . Promoter activity only occurs when carbon has not been depleted in the media . By using μ ( t ) as an indicator of starvation , rather than absolute nutrient values , our model remains flexible and can be adapted to multiple limitations of these nutrients or additional limiting nutrients . Two variables are required for converting the expression to GFP units and scaling the different components . The variable qD is used to scale the density dependent activity component and is the same value for all nutrient conditions . The variable qR is used to scale the starvation induced promoter activity . Consistent with our observations that expression under nutrient starvation depends on the limiting nutrient ( Fig 4B ) , different values of qR are required to achieve the observed levels of activity in nitrogen or iron limitation ( qRN and qRFe respectively ) Activity during prolonged starvation shuts down progressively when μ ( t ) has decreased further and falls below a threshold fraction of μmax , define here as kg . We find again that the limiting nutrient has a great effect on promoter activity and to implement the appropriate shutdown in both nitrogen and iron limitation the values for kg and h must be adjusted for each limitation ( kgN , hN and kgFe , hFe respectively ) . Although we must adjust for the different starvation conditions of nitrogen and iron , a parameterized three-component model of rhlAB expression ( Eq ( 8 ) ) is able to recapitulate the observed expression dynamics under our different nutrient limitations ( Fig 4D–4F ) . Although it is perhaps at first unsatisfying that the fitted parameters must be changed to account for behavior in both nitrogen and iron starvation , this ultimately supports that the internal state of the cell , which drives rhlAB promoter activity , is significantly different in these two conditions . By eye it is not clear if the growth rate and population size differences could be sufficient to explain the different behavior in the two starvation conditions . However , even by taking the observable differences into account we found no model or single parameter set that could sufficiently explain promoter activity in both starvation conditions simultaneously . We find that this model is capable of capturing the observed dynamics and indicates that the metabolic signal for expression is potentially at different levels during nitrogen and iron starvation or that downstream regulation is different in these two limitations . The use of μ ( t ) from our growth model functions as an indicator of starvation and accurately predicts the response of the rhlAB promoter to nutrient starvation . Importantly , both density-dependent regulation and metabolic regulation are required in the model as non-digital regulatory components to recapitulate the observed expression dynamics in all nutrient conditions . Previous work has shown that quorum sensing is required for rhlAB expression , however our observation of a gradual increase in rhlAB promoter activity as population density increases ( Fig 4A ) suggests a more nuanced role for quorum sensing in rhlAB expression . To confirm that quorum sensing does mediate the density-dependent component of rhlAB expression and to explore the effect of perceived density on rhlAB promoter activity , we utilized a quorum sensing mutant that does not produce the lasI/lasR nor rhlI/rhlR system autoinducers C12HSL ( N- ( 3-oxododecanoyl ) -L-homoserine lactone ) and C4HSL ( N-butyryl-L-homoserine lactone ) , respectively and we manipulated the levels of autoinducers in the medium . This strain ( PA14 ∆lasI ∆rhlI attB::PrhlAB-gfp ) has the same PrhlAB-gfp reporter as our wild-type strain ( Fig 5A ) . We first performed an extensive test of rhlAB induction across a wide range of autoinducer concentrations by varying the concentrations of each autoinducer independently ( S6 Fig ) . The data revealed that although one autoinducer can compensate for a lack of the other , this requires very high levels of that single autoinducer , which are likely not biologically relevant . We also observed that when both autoinducers are increased in concentration , but kept in the same proportion , there is a consistent increase in rhlAB expression . We proceeded with a fixed 1:5 ratio of C12HSL to C4HSL used previously [36] . To isolate the density-dependent component of expression , the ∆lasI∆rhlI strain was grown in carbon-limited media complemented with different concentrations of C12HSL and C4HSL kept at a 1:5 ratio ( Fig 5B ) . Higher total levels of rhlAB expression were observed with higher concentrations of quorum sensing signal ( Fig 5C ) . Importantly , promoter activity of rhlAB became constitutive during phase I ( exponential growth ) confirming that quorum sensing modulates expression during that phase ( Fig 5D ) . In contrast to promoter activity of the wild-type PA14 strain ( WT ) during exponential growth , phase I promoter activity in the mutant is decoupled from population density ( Fig 5D , compared with Fig 4A ) . The constitutive level of promoter activity in the mutant scales with the concentration of the autoinducers in the medium , as predicted from the relationship between population density and rhlAB promoter activity we observed and modeled in the WT . We hypothesize that this phase I promoter activity behavior is the result of quorum sensing signals inducing promoter activity when there is a constant level of carbon-rich metabolites present inside the cell during balanced growth in this medium . Because the level of carbon-rich metabolites is constant , promoter activity is modulated only by changes in population density , sensed by quorum sensing signals . Quorum sensing regulation of rhlAB is confirmed here to not be a digital switch , but instead produces a graded rhlAB expression response . Also , the shutdown of rhlAB promoter activity due to carbon depletion occurs even in the presence of high levels of quorum sensing autoinducers ( Fig 5D ) demonstrating that carbon starvation is capable of overriding the quorum-sensing regulated induction . Using this mutant strain we were also able to confirm that quorum sensing signals scale rhlAB promoter activity during starvation by nitrogen or iron ( S7 Fig ) . We were able to identify the differential responses of the rhlAB promoter to different nutrient limitations and population densities in our liquid culture system . To test the effects of the identified rhlAB promoter responses on swarming cooperation , we grew swarming colonies of the ∆lasI∆rhlI quorum sensing null strain in different media conditions . Unlike in liquid culture experiments , the media used in swarming assays has to be a complex media where casamino acids serve as the carbon and nitrogen source . In our liquid culture system , we found iron limitation to be a potent inducer of rhlAB activity . To test if the integration of quorum sensing signals and iron limitation is required for swarming colony formation we tested several conditions with and without iron limitation and with and without quorum sensing signals . Without quorum sensing signals , the colony is unable to swarm regardless of whether iron is limiting ( Fig 6A and 6B ) . As predicted from our liquid culture experiments and mathematical models , if a population has quorum sensing signals but lacks iron limitation , the colony does not have normally branching and does not travel far from the inoculation site ( Fig 6C ) . This confirms that the significant induction of rhlAB promoter activity we observe under iron limitation in our liquid culture system is also key for rhamnolipid production in swarming colonies . Only when a population is provided with both quorum-sensing signals and iron limitation does successful swarming occur ( Fig 6D ) . Swarming behavior in these four conditions supports our liquid culture data; significant production of rhamnolipids requires both quorum sensing signals and iron limitation . The observation that iron starvation facilitates swarming cooperation is consistent with our experiments showing that iron starvation induces higher rhlAB promoter activity than nitrogen starvation or quorum sensing signals alone ( Fig 3F ) and with previous reports [45 , 46] . Given these data the induction of rhlAB expression in swarming colonies ( Fig 1D ) is likely induced by iron limitation . We also tested the effect of nitrogen limitation , iron limitation , and additional quorum sensing signals on the WT . We found again that iron limitation is necessary for successful swarming while nitrogen limitation and additional quorum sensing signals only moderately affect swarming colony morphology ( S8 Fig ) .
Here we investigated the molecular circuitry underlying the regulation of genes required for a model multicellular trait , swarming in P . aeruginosa . We performed this analysis using a combination of quantitative growth curve experiments and mathematical models . Swarming requires cooperation between cells and the production and secretion of rhamnolipids [32] . Expression of the rhlAB operon results in expression of the rate-limiting enzyme for rhamnolipid synthesis and commits the cell to cooperation [31 , 38] . We carried out liquid batch culture growth experiments in shaken microtiter plates , which allows for a high-throughput investigation of gene expression during periods of changing nutrient conditions and cell densities [47] . Shaken liquid-culture neglects the spatial gradients of rhamnolipids , quorum sensing signals and nutrients that may occur in swarming colonies [48 , 49] . In turn , the conditions experienced by cells can be more precisely manipulated in liquid culture allowing us to build a quantitative picture of how metabolic prudence and quorum sensing are integrated into the cellular decision to cooperate . Our data and mathematical model support previous metabolic prudence models where the expression of rhlAB is triggered by excess carbon . During balanced growth internal levels of carbon-rich metabolites are constant and rhlAB promoter activity increases proportionally to population density ( quorum sensing signals ) . Since the population maintains the same growth rate in spite of increasing rhlAB promoter activity , the uptake rate of carbon would be predicted to also increase , to compensate for the increasing demand of rhamnolipid synthesis . This suggests that the rate of carbon uptake is not limiting the growth rate during balanced growth and that even in exponential growth cells can increase carbon uptake to allow for rhamnolipid synthesis . When carbon is fully depleted , rhlAB expression stops abruptly , potentially to reduce the demand on intracellular carbon-rich metabolites when the lack of carbon has become growth limiting . When extracellular iron is depleted , growth slows and rhlAB expression increases . We predict that the decreased growth rate reduces the demand for carbon in biomass production , leading to excess carbon-rich metabolites , which in turn trigger rhlAB expression . Growth also slows during nitrogen starvation , which should also decrease the demand for carbon in biomass production . However , cells can actively decrease carbon uptake due to nitrogen starvation [50] , which would balance the decreased demand for carbon in biomass production . In support of this balancing of supply and demand we observe that the expression of rhlAB during nitrogen starvation is sustained at a low level , but does not increase . Also in support of nitrogen starvation actively decreasing carbon uptake more significantly than iron starvation , our mathematical model reveals that maximum growth rate must undergo a dramatic decrease in nitrogen starvation , but not in iron starvation . Quorum sensing regulates multicellular traits in natural and synthetic bacterial systems , but it is not sufficient for robustness against cheating [3 , 18–20] . Promoter activity of rhlAB during exponential growth and non-carbon nutrient starvation scales with population density , supporting that quorum sensing signals do not act as a checkpoint , but instead continually modulate promoter activity and the decision to cooperate through swarming . Our results demonstrate that these two regulatory mechanisms are continually integrated throughout the entire period of expression and may allow cells to adapt to fluctuations in nutrient conditions and population concentration . The ability to respond to changing population density after the decision to express rhlAB has been made could play a role in maintaining an individual’s fitness while cooperating in a mixed population [37] . Furthermore , metabolic prudence was recently demonstrated as a regulatory mechanism in multiple other secreted products of P . aeruginosa [23] , suggesting that integration of population density and nutrient environment information may be a more widespread regulatory strategy . Taken together our data suggests that rhamnolipid synthesis is regulated by feed-forward supply-driven activation , similar to the coupling of end-product inhibition and supply-driven activation reported to regulate amino acid pools in E . coli [51] . To ensure maintenance of intracellular carbon rich metabolites , the uptake rate of carbon is in turn regulated using a feedback end-product inhibition ( Fig 7 ) . The data presented here supports that quorum sensing is continually integrated into the decision to express rhlAB and that the cell is capable of maintaining a constant rate of biomass production even while expressing rhlAB during exponential phase . The latter observation suggests that the cell can increase the uptake of carbon during balanced growth to compensate for the production of rhamnolipids . Our analysis provides new insights into metabolic prudence , but many details of its molecular implementation remain to be discovered . For example , what is ( are ) the molecule ( s ) that indicates that intracellular carbon is in excess ? Which proteins detect these metabolic signals and how do they interact with quorum sensing regulation at the molecular level ? Answering these and other questions will improve our knowledge of rhamnolipid synthesis by P . aeruginosa with implications for medicine and industry . Rhamnolipids have the power to disperse infectious biofilms of P . aeruginosa and other bacteria , and could thus be used in medical applications [52 , 53] . They also have commercial value as biodegradable surfactants [54 , 55] . Beyond P . aeruginosa , metabolic prudence emerges as a design principle to stabilize cooperation in multicellular groups . Like other design principles of biology [56] metabolic prudence may be applied to synthetic biology where quorum sensing modules are already used to regulate population level traits [57 , 58] .
Cells of the PrhlAB-gfp reporter were taken from a growth curve in a 96 well plate using the growth curve synchronization method described previously [59] . Cells were taken from the 96 well plate wells and mounted on agar pads of low melt agar ( approximately 1 . 75% agar ) with no nutrients . Cells were imaged for bright-field , GFP , and DsRed . A ratio of GFP to DsRed was used to determine expression level of rhlAB by GFP . To correct for autofluorescence in the GFP signal , we measured the fluorescence of the unlabeled PA14 strain , which lacks the PrhlAB-gfp reporter , in numerous experiments across multiple days and a range of conditions , and performed a linear regression on the logarithm of these data ( S1A Fig ) . We found the OD data ( OD600 ) to be a good predictor of the autofluorescence . The fit of the autofluorescence ( AF ) is then a function of the OD600 of the form: We found a correlation coefficient r2 = 0 . 97 using all data points that were OD600>0 . 01 . Using this function , we estimated the amount of autofluorescence in our reporter strains by using the OD600 . We then subtracted the calculated autofluorescence value from the total GFP signal to obtain a corrected GFP ( S1B Fig ) . All synthetic media utilized glycerol as the sole carbon source , ammonium sulfate as the sole nitrogen source and Fe ( II ) sulfate ( Acros Organics , Geel , Belgium ) for iron supplementation . Base media contains 64 g/L of Na2HPO4 . 7H2O , 15 g/L of KH2PO4 , 2 . 5 g/L of NaCl , 1 mM of MgCl2 , 0 . 1 mM of CaCl2 and carbon , nitrogen and iron concentrations depending on the limiting nutrient . Carbon limitation media: 0 . 5 gN/L , 2 . 79 *10–4 gFe/L , carbon at listed concentration . Nitrogen limitation media: 3 . 0 gC/L , 2 . 79 *10–4 gFe/L , nitrogen at concentration listed . Iron limitation media: 3 . 0 gC/L , 0 . 5 gN/L , iron at concentration listed . Promoter activity was calculated as the change in population GFP per unit time divided by OD leading to the following expression [40 , 60]: OD and GFP data are smoothed before calculating promoter activity using the 1-D digital filter using the function filter in Matlab with a window size of 5 timepoints . This calculation accounts for GFP dilution by cell division . Swarming motility and GFP expression were monitored at 37°C at 10 min intervals using a custom-made colony visualizer . The images acquired were processed using Matlab to quantify GFP signal and colony density . Starter cultures are inoculated into 3 mL of LB Miller from glycerol stock and incubated overnight at 37°C with shaking . 1 mL of this LB culture is taken , pelleted at 6000 rpm and re-suspended with 1x PBS . Pelleting and re-suspension in PBS is repeated twice before inoculation into the growth media at an OD600 of 0 . 0025 . All growth curves are performed at 37°C in clear flat-bottom BD Falcon 96 well microtiter plates with 150 μL of media per well . Measurements were taken using a Tecan M1000 plate reader ( Mannedor , Switzerland ) every 10 minutes for the duration of the experiment . Quorum sensing autoinducers HSL and C4HSL used to induce rhlAB expression were obtained from Sigma-Aldrich , St . Louis , MO . Unless otherwise indicated , all growth curves are aligned to have OD 0 . 01 occur at 10 hours . The lag time for each growth curves is determined during this analysis and reported in S2 Table . The mathematical model was implemented as a system of ordinary differential equations in Matlab ( the Mathworks , Natick , MA ) and solved numerically using the ode45 function . Parameter fitting was carried out using the fminbnd function in a step-wise manner . The OD time series were fitted initially , since these data are independent of the GFP time series data . The goal function to be minimized was defined as: errOD=∑i=1N ( log ( XiODi ) ) 2 where i ∈ {1 , N} represents all the data points used for the fit . The GFP data was fitted by first calculating the promoter activity from the data as explained in the main text followed by fitting with fminbnd to minimize the following function errGFP=∑i=1N ( Pi−pi ) 2 where Pi is the promoter activity calculated from the data and pi is the promoter activity predicted by the model . Parameter sensitivity was performed for both the bacterial growth and rhlAB expression models and results are reported in S4 and S5 Tables . Starter cultures and inoculation were performed as in Batch Culture Growth curves . Cells were grown to exponential phase in media with 3 . 0 gC/L , 0 . 5 gN/L and 2 . 79 *10–4 gFe/L in 88 wells of a 96 well plate . All wells were harvested and the cells were pelleted and washed with PBS as in Batch culture growth curves . The harvested cells were then split and re-suspended either 1 . 2 mL of media without nitrogen ( 3 . 0 gC/L , 2 . 79 *10–4 gFe/L ) or 1 . 2 mL media without carbon ( 0 . 5 gN/L , 2 . 79 *10–4 gFe/L ) and grown in a 96 well plate in a Tecan M1000 plate reader as described in Batch Culture Growth curves . Swarming assays were performed as previously described [36] . Nitrogen was supplemented with ammonium sulfate at 0 . 5gN/L . Iron was supplemented using iron ( II ) sulfate at 2 . 79*10–4 gFe/L . Quorums sensing autoinducers were added at the listed concentrations from liquid stock solutions . | Although bacteria are not multicellular organisms , they commonly live in large communities and engage in many cooperative behaviors . Cooperation can allow bacteria to access additional nutrients , but it requires the secretion of products that will be shared by the community . How bacteria make the molecular decision to cooperate within a community is still not completely understood . The bacterium Pseudomonas aeruginosa regulates the secretion of one of these shared products , rhamnolipids , using information about population density and nutrient availability in its environment . Expression of the operon rhlAB is required for the bacteria to produce rhamnolipids . We use a combined computational and experimental approach to investigate how P . aeruginosa continually combines current information of population density and nutrient availability to determine if it should express rhlAB . We find that when conditions are nutrient rich , P . aeruginosa uses population density to modulate the amount rhlAB expression , however when the bacteria are starved for nutrients the starvation condition largely determines how the bacteria will express rhlAB . Because the bacteria continually adjust expression based on the current conditions , the molecular decision to produce rhamnolipids can be adjusted if either population density or nutrient conditions change . Our combined computational and experimental approach sheds new light on the rich regulatory dynamics that govern a cellular decision to cooperate . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Integration of Metabolic and Quorum Sensing Signals Governing the Decision to Cooperate in a Bacterial Social Trait |
Efforts to engineer synthetic gene networks that spontaneously produce patterning in multicellular ensembles have focused on Turing's original model and the “activator-inhibitor” models of Meinhardt and Gierer . Systems based on this model are notoriously difficult to engineer . We present the first demonstration that Turing pattern formation can arise in a new family of oscillator-driven gene network topologies , specifically when a second feedback loop is introduced which quenches oscillations and incorporates a diffusible molecule . We provide an analysis of the system that predicts the range of kinetic parameters over which patterning should emerge and demonstrate the system's viability using stochastic simulations of a field of cells using realistic parameters . The primary goal of this paper is to provide a circuit architecture which can be implemented with relative ease by practitioners and which could serve as a model system for pattern generation in synthetic multicellular systems . Given the wide range of oscillatory circuits in natural systems , our system supports the tantalizing possibility that Turing pattern formation in natural multicellular systems can arise from oscillator-driven mechanisms .
Genetic networks which enable communication and coordination of behavior among cells in an ensemble have held the attention of developmental biologists and theoreticians [1]–[6] for over half a century . In particular , a vast body of literature – both theoretical [6] , [7] and experimental [1]–[4] – exists which focuses on the production of patterns in gene expression , a phenomenon central to the development of multicellular organisms . A particularly well-studied mechanism for pattern formation is diffusion-driven instability , originally proposed by Turing [8] , where a homogeneous steady state is destabilized in the presence of diffusion . Recently , attempts have been made to build synthetic gene networks which generate spatio-temporal patterns in gene expression mediated by diffusible signals [9]–[13] . To obtain pattern generation , these efforts have relied either on the external spatio-temporal manipulation of the cell's chemical environment [9] , [10] , [13] or the precise positioning of cells containing different gene networks which secrete or respond to diffusible signals [11] , [12] . To date , there have been no experimental demonstrations of a robust , tunable system which can break symmetry and spontaneously generate predictable gene expression patterns ( spatio-temporal inhomogeneities ) as in the Turing mechanism . What is specifically lacking in the community is an experimentally tractable model system for studying spontaneous pattern formation . Such a system would catalyze the engineering of complex cellular ensembles , ranging from engineered microbial communities [11] , [13] to auto-differentiating multicellular systems . In the synthetic biology community , efforts to achieve spontaneous generation of spatial patterns in gene expression have been centered around networks similar to the one originally proposed by Turing [8] , and expanded into activator-inhibitor theory by Meinhardt and Gierer [5] , [6] , [14] , [15]: two diffusible species interact with each other via chemical reactions that produce positive and negative interactions as in Figure 1A . For an appropriate range of kinetic parameters and diffusion constants , these topologies produce spatial or spatio-temporal patterns spontaneously from a homogeneous initial condition perturbed by small variations in concentration due to stochastic effects . However , this type of architecture has proven very difficult to implement using genetic networks because: ( a ) Turing instability requires that the steady state occur in the linear regime of the activator-inhibitor interactions away from saturation , and severely restricts the parameter range to meet the instability criteria; ( b ) when using systems with two diffusible components , either the diffusion constants [8] or the uptake rates [16] must be sufficiently different to allow unstable spatial modes , and significant differences are difficult to engineer; ( c ) the addition of intermediate protein steps to two-molecule activator-inhibitor models further restricts the parameter set for patterning; and ( d ) stochasticity plays a significant role in the behavior of these systems , but most analyses rely on continuum partial differential equation ( PDE ) models , making it difficult to reconcile theoretical predictions with observed experimental results . Although the activator-inhibitor model is the canonical example of a system demonstrating Turing instability , many other possible network structures exist . Indeed , the essential structural requirement for the emergence of the Turing phenomenon is that the network contain an unstable subsystem , which is stabilized by a feedback loop . The diffusion of molecules participating in this feedback loop then unleashes the inherent instability and allows growth of spatial modes . In the activator-inhibitor network in Figure 1A , the activator plays the role of the unstable subsystem and the inhibitor provides the stabilizing feedback . Although it is well known that the Turing mechanism is not restricted to the activator-inhibitor network ( see , e . g . , [17] for Turing instability conditions for general reaction-diffusion models ) , to the best of our knowledge , no other biologically plausible network has been proposed . Systems that contain more than two species have been studied , but their reactions conform to the essential structure of the activator-inhibitor paradigm [6] . This paper breaks away from the activator-inhibitor model and proposes a new network which we call a “quenched oscillator” system . This system uses one diffusible component and an oscillator circuit serving as the unstable subsystem that is quenched by a second feedback loop , as depicted in Figure 1B . To our knowledge , this is the first demonstration that oscillator-driven gene networks can exhibit Turing instability and spatial patterning of gene expression across fields of cells . Moreover , the network can be implemented with a variety of published oscillator circuits [18]–[20] using known genes and promoters . It is important to stress that the mechanism pursued here – Turing instability – is fundamentally different from the traveling wave trains and spiral waves in diffusively coupled oscillators [7] , [21] . The proposed architecture bears resemblance to the diffusively coupled repressilator model in [22] , where a second loop is integrated with the repressilator to incorporate a diffusible molecule , and the diffusively coupled oscillator model in [23] . However , in both of these systems , the oscillator is not quenched , but is simply allowed to communicate between cells to ensure synchronization , which is contrary to the pattern formation task studied here . Although we employ an oscillator as a subsystem , the full system in isolation is not an oscillator , instead exhibiting a stable steady state as in the Turing mechanism , and is fundamentally different from out-of-phase oscillator systems . The patterns presented in this paper are oscillatory in both time and space ( see Text S2 ) . While they still fall under the category of diffusion-driven instability as proposed by Turing , some researchers associate the term “Turing patterning” with stationary spatial non-uniformities [24] . However , in the remainder of this paper , the use of the term “Turing pattern” incorporates oscillatory Turing patterns . Systems which produce oscillating patterns have previously been reported [22] , [23] , [25] , [26] , but there are fundamental differences between these systems and our own . In Turing's diffusion-driven instability , the biological system exists in a population of homogeneous cells . In isolation , each individual system is stable and reaches a steady state over time . In the presence of diffusion , the steady state is destabilized and spatial inhomogeneities arise . The systems presented in [22] , [23] , and Figure 2C of [25] are not stable individually , so the patterns they produce do not fall under the general category of “Turing patterns . ” The systems presented in [26] and Figure 2A and 2B of [25] are stable , but do not contain an oscillatory subsystem . Our system utilizes an oscillator subsystem , which does not constitute positive feedback; this is a crucial point that separates it from previous systems . In addition to the new architecture presented in this paper , we believe the methodology used to find and tune new pattern-generating systems may prove of significant value to practitioners . We recognize that Turing patterning is just one possible method to achieve the generation of gene patterning across a population of cells , but it is a phenomenon that is well-characterized mathematically , allowing us to develop the combined PDE/stochastic simulation approach presented here . Patterning has proven difficult to produce experimentally , so these analysis tools should aid in the search for more reliable experimental systems . Below , we first provide an analysis of the system which predicts the range of kinetic parameters over which patterning should emerge . We show the architecture produces patterning for parameters within the range of values present in the literature for our molecules . Using both continuous , deterministic simulations ( henceforth called “PDE simulations” ) and discrete , stochastic simulations ( henceforth called “stochastic simulations” ) of fields of cells , we demonstrate how stochastic molecular interactions affect pattern formation in the limit of very low concentrations of molecular species per cell . The primary goal of this paper is to provide a circuit architecture which can be implemented with relative ease by practitioners and which provides an alternative implementation strategy for reaction-diffusion pattern generation in synthetic multicellular systems . Lastly , given the wide range of oscillatory circuits in natural systems , our system supports the tantalizing possibility that Turing-like pattern formation in natural multicellular systems can arise from oscillator-driven mechanisms .
The first feedback loop in our design is an oscillator . The second feedback loop is designed to quench these oscillations , meaning that , in the presence of the second loop , the first loop ceases to oscillate and the full system instead approaches a steady-state solution . If the oscillator design is based on a phase lag mechanism as in Figure 1B , then it is essential that the second loop with the diffusible molecule ( in blue ) have smaller phase lag than the first loop ( in pink ) , so that it is stable by itself and that it stabilizes the oscillator when interconnected . Smaller phase lag can be achieved with fewer reaction steps or with faster degradation rates in the second loop . As an illustrative example , consider the following “toy” model , possessing both an oscillator loop and a quenching loop : ( 1 ) where the concentrations , , and all other variables and parameters are non-dimensional . In particular , the time variable is scaled to bring the degradation constants ( assumed to be identical for each species for simplicity ) to one , and the one-dimensional length variable is scaled so that the spatial domain is . We assume only the fourth species is diffusible ( represented with wavy arrows in Figure 1B ) with diffusion coefficient and is subject to zero-flux boundary conditions , meaning there is no diffusion at the ends of the line of cells at and . For analysis , we take the linearized form of the reaction-diffusion system above:where is the Jacobian matrix of the vector field of reaction rates evaluated at the steady state of the reaction system , is the diagonal matrix of diffusion coefficients , and is the vector Laplacian . For our toy model , the Jacobian matrix about the steady state is: ( 2 ) where:and . The dynamical behavior of this reaction-diffusion system is determined from the matrices , where are the eigenvalues of the Laplacian operator on the given spatial domain , and the subscripts denote the wave numbers . On our one-dimensional domain , and the eigenfunctions are the cosine waves [27] . If the matrix is stable ( that is , if all of its eigenvalues have negative real parts ) , then the corresponding spatial wave decays to zero asymptotically in time . If is unstable ( at least one of its eigenvalues has positive real part ) , then the corresponding spatial wave grows . Let be the upper-left submatrix of , corresponding to the oscillator loop . For diffusion-driven instability to arise in this network , the following three conditions must be met: Condition 1 . The oscillator loop by itself would produce oscillations ( is unstable ) . For the oscillator subsystem to be unstable , we need: ( 3 ) so that the characteristic polynomial of , given by , has a pair of complex conjugate roots with positive real part . Condition 2 . The quenching loop ceases oscillations in the full system ( is stable ) . For stability of the full reaction network , we need: ( 4 ) so that has all roots with negative real parts . Condition 3 . Diffusion will weaken the quenching loop's influence on the oscillator loop for high wave numbers , allowing spatio-temporal oscillations to emerge ( is unstable for some ) . For diffusion-driven instability of the spatial mode , the polynomial: ( 5 ) must have at least one eigenvalue with positive real part . Indeed , when the product is sufficiently large , three roots of ( 5 ) approach those of , which contain roots with positive real part due to ( 3 ) . This means that the inhomogeneous modes grow over time if exceeds the threshold for instability of the polynomial ( 5 ) , which we will call . This implies that , for diffusion-driven patterning , we need a large diffusion coefficient or a large wave number . More generally for , we need , meaning patterning can also be achieved for a small spatial domain . See Text S1 for details . The parameters , , , , and in the system ( 1 ) satisfy conditions ( 3 ) and ( 4 ) with . The polynomial ( 5 ) becomes unstable when . PDE Simulations with indeed exhibit growth of the spatial inhomogeneity when the steady state is perturbed by adding the second wave ( ) with amplitude steady state peak-to-peak to ( Figure 2B , top ) . The PDE system does not include noise , so a perturbation must manually be added to the system for cells to leave the steady state . This Turing behavior is contrasted to the decay of the initial inhomogeneity for wave numbers below the instability threshold ( , in Figure 2B , middle ) and in the absence of diffusion ( , in Figure 2B , bottom ) . We now propose a novel network that can be synthesized from existing components . Consider the system of two interconnected loops shown in Figure 3 . The first ( top ) loop is the repressilator [18] , which is a ring oscillator , comprised of three pairs of transcriptional repressors ( TetR , cI , LacI ) and promoters , which match up with the three-component oscillator of the toy model ( -- ) . The second ( bottom ) feedback loop consists of V . fischeri quorum sensing genes luxI and luxR . The luxI gene is regulated by the promoter , and is transcribed in the absence of TetR . LuxI is the synthetase that catalyzes the formation of the membrane-diffusible signaling molecule acyl-homoserine lactone ( AHL ) . AHL binds to the constitutively produced protein , LuxR . The LuxR-AHL complex forms a homodimer that binds to the promoter and activates transcription . TetR production closes the second loop by repressing the second promoter . This quenching loop is much longer than that of the toy model ( - ) , but still contains a single diffusible molecule , AHL , and we ensure that it has smaller phase delay than the oscillator loop by using faster degradation rates . Even though the bottom loop has a single inhibitory interaction , this loop does not oscillate because the phase delay is small . The two loops interact through TetR and the first loop ceases to oscillate in the presence of the second loop . We represent the dynamics of the network in Figure 3 with the following set of partial differential equations: ( 6 ) ( 7 ) ( 8 ) ( 9 ) ( 10 ) ( 11 ) ( 12 ) ( 13 ) ( 14 ) ( 15 ) ( 16 ) where are mRNA concentrations , are protein concentrations , are velocity constants , are copy numbers , are dissociation constants , are Hill coefficients , are leakage rates normalized to , are degradation rates , and are protein translational rates . The parameters are subscripted according to their corresponding species ( C = [ cI] , T = [tetR] , L = [lacI] , I = [luxI] , A = [AHL] , R = [luxR] , RA = [luxR-AHL complex] ) except for velocity and leakage constants , which are subscripted by promoter , and copy numbers , which are subscripted by the gene being transcribed . The concentration of the mRNA for tetR is split into those produced by the oscillator loop ( O ) and the quenching loop ( Q ) . The variable is the total amount of LuxR protein in the system , which is assumed constant , thus the amount of free LuxR is represented by . The parameter C is the concentration level generated by a single molecule in an E . coli cell and is the diffusion coefficient of AHL . We take . The system is subject to zero-flux boundary conditions on the one-dimensional spatial domain . Following the same procedure as outlined for the toy model , we solve for constraints on kinetic and diffusive parameters by deriving expressions used to satisfy the three conditions for Turing instability ( see Text S1 ) . These expressions are complex combinations of the many parameters in our system , but still reveal ways of manipulating parameter choices to meet the Turing instability conditions . In this analysis , the protein translation rates were taken to be the most readily tunable , which made the big challenge finding protein translation rates for this system to meet the Turing conditions for patterning . To show the viability of this system for experimental implementation , we modeled the system behavior using parameter values from the literature that fit the constraints found in the analysis ( “Value for PDE Simulation ( Parameter Set 1 ) ” column of Table S1 ) . Expected steady-state values and instability measurements can be found in Tables S2 and S3 . We ran PDE simulations in MATLAB with and without AHL diffusion using an initial perturbation in of amplitude steady state peak-to-peak and wavelength , which was predicted to be unstable ( Figure 4 ) . The imprinted wave grows with diffusion and it decays without diffusion , exhibiting similar behavior to that of the toy model ( Figure 2 ) . While the simulation results produce spatio-temporal patterning as desired , the expected experimental behavior will be impacted by stochastic properties that stem from concentrations in our system approaching a few molecules per cell . Taking the concentration of a single molecule in an E . coli cell to be [28] , a number of steady-state values fall near or below this threshold ( Figure 5 ) , particularly and . This implies that: a ) stochastic simulations are necessary for examining experimental plausibility , and b ) Parameter Set 1 would need to be modified to produce pattern due to the behavior of certain species in our system being dominated by noise . In this limit , stochastic models better capture the behavior of in-vivo systems because of their inability to respond to aphysical concentration changes of less than one molecule per cell . Given complete freedom in choosing parameter values , our analysis would allow us to methodically identify regions in the parameter space that should produce patterning . It is encouraging that even when restricting ourselves to literature values for all of the parameters , we were able to demonstrate spatio-temporal patterning in PDE simulation . Here we show that with other parameter values that are still biologically realistic ( “Value for Stochastic Simulation ( Parameter Set 2 ) ” column of Table S1 ) , we can improve the system performance to also produce patterning in a discrete , stochastic environment . All of these values are physically possible based on information in the literature ( see references in Table S1 ) . To accommodate the change in the ratio , LuxI has been replaced in our system with the AHL synthetase RhiI from P . aeruginosa . A more thorough explanation of the origin of these two parameter sets can be found in Text S3 and Text S5 . As more biological parts are characterized or created , parts are likely to be found that match our chosen parameter values . The steady-state concentrations for Parameter Set 2 can also be seen in Figure 5 and do not fall below ( 4 molecules/cell ) . We also verified the desired system behavior of this parameter set in PDE simulation ( Figure 6 ) . This new parameter set results in growth of additional wave numbers other than the imprinted one , highlighting the nonlinear nature of our system . These effects arise when oscillations start to reach near-maximal amplitudes and would likely be seen for Parameter Set 1 if the simulations were run for a much longer time . We developed a set of reactions for stochastic simulation that , using the law of mass action and the quasi-steady-state approximation , would exactly match our set of PDEs . The full set of reactions used in our stochastic simulations can be found in Text S4 . To compare the behavior of PDE and stochastic simulations , we first ran single cell simulations to verify that the general expected behavior was maintained . While not indicative of the system's ability to generate pattern , these simulations allow us to draw comparisons between our PDE and stochastic models . To observe both an oscillating cell and a quenched cell , we used a single cell in the center of a long , empty volume . Without AHL diffusion , the cell remains isolated and we expect oscillations to decay to the steady state . With diffusion , AHL diffuses into the empty volume and weakens the quenching loop , meaning oscillations are expected to grow . Both PDE and stochastic simulations confirmed these expectations ( Figure 7 ) . The simulations exhibited similar behavior but oscillations in the stochastic environment are slower and more irregular , due to stochasticity and our modeling assumption that the dimerization and binding reactions are at equilibrium in the PDE model . Oscillations in the stochastic simulations are significantly slower – about 5 times slower in the decaying case , and 10 times slower in the growing case – which lead us to choose faster degradation rates for Parameter Set 2 . In a cell without diffusion , stochasticity keeps the system oscillating at a small amplitude with occasional “firing events , ” where a few cycles of increased oscillation amplitude occur before the system settles again . Both PDE and stochastic simulations exhibit the same phase relationship between the proteins in the oscillator loop and a slower period of oscillation when growing as opposed to decaying ( Figure 7 and S7 , S8 , S9 , S10 ) . As expected , stochastic simulations with Parameter Set 1 in a line of cells were unable to produce patterning due to the low steady-state concentration values ( results not shown ) , but did yield some insights . In particular , any initial imprint we imposed would very rapidly ( ) decay into noise , likely due to low copy numbers . With only four or five promoter binding sites per cell and the fact that almost all of them are bound in steady state , a large change in a single species of the system is unlikely to be able to propagate quickly enough throughout the system due to the bottlenecks at the promoter binding sites . Thus we avoided imprinting and used the ability of the stochasticity in our system to naturally excite high wave numbers . Indeed , stochastic simulations with Parameter Set 2 in a line of cells exhibit growing oscillations and eventually produce spatio-temporal patterning ( Figure 8 ) . Large amplitude oscillations emerge around 20 hours and an obvious pattern emerges as time goes on . Visually , patterning is most evident in AHL due to the effects of diffusion . Without diffusion , no spatial patterns emerge with single cell oscillations occurring randomly ( results not shown ) . To quantify the patterns produced by our system , we use the discrete cosine transform ( DCT ) to check the relative presence of the different emerging wave numbers . All wave numbers higher than a threshold ( for Parameter Set 2 ) should grow in the presence of noise according to our analysis , but a number of factors , including stochasticity and the discrete nature of only having 100 cells in our simulations , prevent them from growing uniformly . The exact wave numbers vary from simulation to simulation , but the averaged DCT over time frames late in simulations ( beyond the “start-up” phase ) always shows a number of spikes that are prominent across most species in the system ( see Figures S11 , S12 ) . The exceptions to this are AHL and subsequent species in the quenching loop , where diffusion acts as a low-pass filter and attenuates high wave numbers . This filtering effect is what accounts for the visual “bleeding” effect of diffusion .
The process of producing a set of parameters which produce pattern in the stochastic regime provided several insights which can inform implementation decisions as new promoters , proteins , and parameter manipulation techniques become available . These findings may also be of use when searching for putative natural systems which exhibit this behavior . Two of the most restrictive parameters that we had to change significantly from our initial solution set were promoter leakage rates and dissociation constants . High amounts of leakage makes it simultaneously more difficult to make the oscillator subsystem unstable and more difficult for the quenching loop to stabilize the overall system . The dissociation constants directly affected the steady-state concentrations of the protein species in our system; the system fails to produce patterning when these values are too small . These observations were made from studying the form of the expressions for and ( see Text S1 ) and many other such observations and insights can be drawn from the analysis . A few considerations only became relevant when performing stochastic simulations , the biggest of which was the bottleneck of promoter binding sites . In the PDE model , new mRNA would be produced at a rate that was a function of the amount of the appropriate activator or inhibitor in the system . By enumerating the number of promoter binding sites , we decrease the sensitivity of the system to very large concentrations of the activators and inhibitors and increase the importance of each binding and unbinding event . Analytically , we can maintain the same system behavior by holding the product in each mRNA differential equation constant . Arbitrarily increasing the copy numbers this way has its own drawbacks . We assume the concentration of LuxR is constitutively produced and is constant . At our current value of ( 12 molecules/cell ) , we can only bind at most six promoters with LuxR-AHL dimers , so having a large will not change the amount of being produced , which deviates from what our PDE model predicts . Assuming proper parameter values can be chosen for our system , our analysis generates a testable hypothesis for a possible experimental implementation . When setting up the experiment , the following additional concerns should be taken into account . Beyond finding parameters that meet the Turing instability conditions , system speed is very important because it determines the visibility of changes in the system over the course of a normal experiment duration . System speed is most directly affected by the degradation rates of every species in the system . These change the period of oscillations as well as the growth and decay rates of wave modes . Very slow growth and decay would delay the emergence of visible patterns and make experimental debugging difficult because any activity would be hard to observe . Very long experiments are problematic in terms of collecting data and dealing with cell division and lifespan . A reporter gene was unnecessary in simulation , but one would need to be used in experiments . As seen in Figure 8 , there are two distinct types of qualitative behaviors: the proteins cI , LacI , TetR , and LuxI exhibit brief bursts localized to single cells while AHL and subsequent quenching loop species exhibit more spread out behavior due to diffusion . It is possible to attach a fluorescent protein to the appropriate loop to follow either type of behavior . While AHL may produce a more visually-pleasing patterning , the oscillator loop species undergo larger swings in number of molecules , which would be easier to discern in units of fluorescence . The engineering of cooperative ensembles of cells , whether in the context of designer microbial communities or other synthetic multicellular systems will require tractable model systems which exhibit spontaneous symmetry breaking and pattern formation , both fundamental prerequisites for any kind of replicating or “programmed” heterogeneity of form or function . Attempts to produce spontaneous pattern formation using Turing's canonical system have proven difficult ( see Introduction ) . This paper breaks away from the activator-inhibitor model and alleviates some of the difficulties encountered by using oscillating subsystems . To our knowledge , this is the first attempt of this kind and significant effort was devoted to providing researchers with an experimentally tractable road map towards implementation . This work also implicitly suggests that natural systems may have arisen where oscillating subsystems , initially evolved for other purposes , provide the backbone not just for coordinated oscillation ( as in the diffusively coupled systems demonstrated by others [7] , [22] , [23] ) but for robust Turing-type pattern formation phenomena . It is not difficult to find examples in the recent literature of naturally-occurring coupled negative feedback oscillators , both in prokaryotes [29] and eukaryotes [30] , [31] . A function as fundamental as cell cycle oscillation appears to be maintained in yeast and other eukaryotes by coupled oscillators ( a negative feedback oscillator coupled to a relaxation oscillator ) [31] . Going further , these motifs are also present in protein-protein systems [32]; while outside the scope of the present work , the general results presented ( i . e . coupled multi-step negative feedback oscillators with one diffusible component can exhibit Turing instability ) would likely apply to kinase loops [32] . Lastly , in our model the relative phase lag between the oscillator loop and the quenching loop affect both the emergence and wave numbers of pattern; these , in turn , depend on the relative number of “steps” around the loops . It is tempting to suggest that the alteration of the number of steps , or the total delay around the loop , could provide a mechanism by which adaptation and evolution could generate systems ( and variants ) capable of pattern formation .
Continuous , deterministic models are useful because of the wide variety of analysis tools we can apply to them to generate predictions of system behavior and workable parameter spaces , which we cannot do for stochastic models . These models are accurate when the number of molecules for all species in the system are very large , but generally need to be supplemented with stochastic simulations for systems with small numbers of molecules . PDE simulations were run in MATLAB Version 7 . 10 . 0 . 499 ( R2010a ) with the function ode15s , which is a multi-step , variable order solver based on numerical differentiation formulas . For line of cell simulations , diffusion was handled using a finite difference approximation with 101 evenly-spaced grid points and zero-flux boundary conditions . For single cell simulations , the long empty volume was represented using a finite difference approximation with Dirichlet boundary conditions of zero AHL concentration . Stochastic simulations of the network were performed using the Stochastic Simulator Compiler ( SSC ) v0 . 6 [33] . The output from SSC was reformatted with custom Perl scripts and then plotted in MATLAB . SSC handles concentrations in units of molecules , so all parameter values were scaled appropriately , but the output values were converted to units of molarity in the figures given in this paper for ease of comparison . Reported values for protein concentrations are the totals of all forms of the protein: monomer , dimer , and bound to promoter . We represented cells with cubes of edge length . For single cell simulations , the cell was located at the center of a volume of . All multi-cell simulations consisted of a line containing 100 directly adjacent cells . A discrete cosine transform ( DCT ) expresses a finite sequence of data as a sum of cosine functions of different frequencies [34] . The eigenfunctions of the Laplacian operator on a one-dimensional spatial domain with zero-flux boundary conditions are cosine functions [27] , which are represented more accurately by the DCT than by the discrete Fourier transform , which is appropriate for periodic boundary conditions . The DCT is useful for our analysis because it allows us to examine the presence of certain spatial wave numbers in a line of cells simulation relative to the other wave numbers and how these relations change over time . Because the amplitudes of a DCT are changing in time and can be both positive and negative , we take the average of the absolute values of spatial DCTs over an interval of time . This was handled in MATLAB using the function dct . Because concentrations are non-negative , there is always a significant offset component , which we omit from our figures for better scaling of the remaining wave numbers . | The production of patterns in gene expression in an ensemble of cells is a phenomenon central to the development of multi-cellular organisms . Here we provide an exciting new result regarding diffusion-driven instability , a mechanism for spontaneous pattern formation originally proposed by Alan Turing . Efforts along this front have focused almost exclusively on Turing's original model and the “activator-inhibitor” models of Meinhardt and Gierer , but have yet to yield an experimental demonstration of a robust , tunable system that can break symmetry and spontaneously generate gene expression patterns . In this paper we propose a new family of oscillator-driven gene network topologies capable of Turing pattern formation . We believe this would be of significant impact to both emerging efforts at engineering multicellularity in the synthetic biology community as well as new guidance for those groups looking for similar phenomena in natural systems . Given the wide range of oscillatory circuits in natural systems , our system supports the tantalizing possibility that Turing pattern formation in natural multicellular systems can arise from oscillator-driven mechanisms . We provide an analysis of the system that predicts the range of parameters over which patterning should emerge and demonstrate the system's viability using stochastic simulations of a field of cells using realistic parameters . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"prokaryotic",
"models",
"model",
"organisms",
"synthetic",
"biology",
"molecular",
"genetics",
"biology",
"computational",
"biology",
"microbiology",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | A Feedback Quenched Oscillator Produces Turing Patterning with One Diffuser |
Clustered , regularly interspaced , short palindromic repeat ( CRISPR ) –CRISPR-associated 9 ( Cas9 ) genome editing is revolutionizing fundamental research and has great potential for the treatment of many diseases . While editing of immortalized cell lines has become relatively easy , editing of therapeutically relevant primary cells and tissues can remain challenging . One recent advancement is the delivery of a Cas9 protein and an in vitro–transcribed ( IVT ) guide RNA ( gRNA ) as a precomplexed ribonucleoprotein ( RNP ) . This approach allows editing of primary cells such as T cells and hematopoietic stem cells , but the consequences beyond genome editing of introducing foreign Cas9 RNPs into mammalian cells are not fully understood . Here , we show that the IVT gRNAs commonly used by many laboratories for RNP editing trigger a potent innate immune response that is similar to canonical immune-stimulating ligands . IVT gRNAs are recognized in the cytosol through the retinoic acid–inducible gene I ( RIG-I ) pathway but not the melanoma differentiation–associated gene 5 ( MDA5 ) pathway , thereby triggering a type I interferon response . Removal of the 5’-triphosphate from gRNAs ameliorates inflammatory signaling and prevents the loss of viability associated with genome editing in hematopoietic stem cells . The potential for Cas9 RNP editing to induce a potent antiviral response indicates that care must be taken when designing therapeutic strategies to edit primary cells .
Clustered , regularly interspaced , short palindromic repeat ( CRISPR ) –CRISPR-associated ( Cas ) genome editing has rapidly become a widely used tool in molecular biology laboratories . Its ease of use and high flexibility allows researchers to modify and edit genomes in cell lines [1] , stem cells [2] , animals and plants [3 , 4] , and even human embryos [5] . The Cas protein complexes with a target-specific CRISPR RNA ( crRNA ) and a trans-activating crRNA ( tracrRNA ) , which keeps the Cas protein catalytically active [6] . In experimental procedures , the two RNAs are often combined to generate a single guide RNA ( gRNA ) , which means that at least two components must be successfully delivered into cells during genome editing: the Cas protein , such as Cas9 , and gRNA to direct the Cas9 protein to its target site . For in vitro–cultured cells , this can be done by transfecting plasmids encoding gRNA and Cas9 protein . However , transfection of plasmid DNA into sensitive cell types such as primary and stem cells is challenging and inefficient . The introduction of plasmids can also lead to undesired integration of DNA at the cut site [7] , increased off-target activity through prolonged expression of the CRISPR-Cas9 components [8] , and a delay in editing while the cell expresses gRNA and Cas protein [9] . The delivery of gRNA and Cas9 protein as a precomplexed ribonucleoprotein ( RNP ) sidesteps issues related to plasmid expression and has proved to be a successful strategy to edit human primary cells , including T cells [10 , 11] , hematopoietic stem cells [12–15] , and neurons [16] . This makes RNP editing a particularly attractive approach for therapeutic applications , but relatively little is known about the nonediting consequences of introducing a foreign gRNA and Cas9 protein . Human cells have evolved multiple defense mechanisms to guard against foreign components , and genome editing reagents have the potential to activate these systems . For example , recent data suggest that humans may have a preexisting adaptive immune response to the Cas9 protein [17] . But cellular responses to the gRNAs used to program Cas9 editing have so far not been well explored . Cells respond to infection by RNA viruses with an innate immune response that protects the host cell from invading foreign genetic material [18] . Foreign RNAs are recognized by pathogen-associated molecular pattern ( PAMP ) binding receptors in the cytosol that include retinoic acid–inducible gene I ( RIG-I ) and melanoma differentiation–associated gene 5 ( MDA5 ) [19] . This triggers a cascade of events mediated by the mitochondrial antiviral signaling ( MAVS ) protein , resulting in the transcriptional activation of type I interferons and interferon-stimulated genes ( ISGs ) [20–22] . RNA PAMPs usually contain exposed 5’-triphosphate ends [19] , which may also be present in gRNAs made via T7 in vitro transcription [23 , 24] . Given that Cas9 has a picomolar affinity for targeting gRNA [25] , it is not clear that the 5’-triphosphate would be available to stimulate an innate immune response . We asked whether in vitro-transcribed ( IVT ) gRNAs complexed with Cas9 cause an innate immune response and here show that introduction of RNPs into cells induces up-regulation of interferon beta ( IFNβ ) and interferon-stimulated gene 15 ( ISG15 ) in a variety of human cell types . This activity depends upon RIG-I and MAVS but is independent of MDA5 . The extent of the immune response depends upon the protospacer sequence , but removal of the 5’-triphosphate from gRNAs avoids stimulation of innate immune signaling . The potential for Cas9 RNP editing to induce an antiviral response indicates that care must be taken when designing therapeutic strategies to edit primary cells .
To investigate if mammalian cells react to IVT gRNA/Cas9 with an innate immune response , we first performed genome editing in human embryonic kidney 293 ( HEK293 ) cells using Cas9 RNPs . To separate innate immune response from genome editing , we performed these experiments with a nontargeting gRNA that recognizes a sequence within blue fluorescent protein ( BFP ) and has no known targets within the human genome [26] . Constant amounts of recombinant Cas9 protein were complexed with different amounts of nontargeting IVT gRNA , and RNPs were transfected into HEK293 cells using CRISPRMAX lipofection reagent [27] . We harvested cells 30 h after transfection and measured transcript levels of interferon beta 1 ( IFNB1 ) and ISG15 by quantitative real-time PCR ( qRT-PCR; Fig 1A ) . Introduction of gRNAs caused a dramatic increase in both IFNB1 and ISG15 levels , and the presence of Cas9 protein did not have an effect on the outcome . Cas9 on its own did not induce IFNB1 or ISG15 expression . To our surprise , as little as 1 nM of gRNA was sufficient to trigger a 30–50-fold increase in the transcription of innate immune genes . We further found that a commonly administered amount of 50 nM gRNA can induce IFNB1 by 1 , 000-fold , which is equal to induction by canonical IFNβ inducers such as viral mRNA from Sendai virus [28] or a hepatitis C virus ( HCV ) PAMP [21 , 29] ( Fig 1B ) . RNPs can be delivered into cells via different transfection methods , and while lipofection is cost-effective and easy to use , many researchers prefer electroporation for harder-to-transfect cells . We wondered if the transfection method would affect the IFNβ response and compared gRNA transfection via lipofection ( Lipofectamine 2000 and RNAiMAX ) to nucleofection ( Lonza ) ( Fig 1C ) . Lipofection led to a strong increase in IFNB1 and ISG15 transcript levels after as little as 6 h posttransfection , and the response was sustained for up to 48 h . Nucleofection also caused an increase in innate immune signaling at early time points , but the response was milder than in lipofected samples and was greatly diminished by 48 h . Next , we asked if the innate immune response to gRNAs is a common phenomenon across different cell types and compared IFNβ activation in seven commonly used human cell lines of various lineages: human embryonic kidney cells 293 SV40 large T antigen ( HEK293T ) , HEK293 , Henrietta Lacks cells ( HeLa ) , Jurkat , HCT116 , HepG2 , and K562 ( Fig 2A ) . While the magnitude of induction varied between cell lines , all tested cell lines responded to IVT gRNA transfection with activation of IFNB1 expression . The sole exception was K562 cells , which have a homozygous deletion of the IFNA and IFNB1 genes [30] . We also measured transcript levels of two major cytosolic pathogen recognition receptors , RIG-I ( DExD-H-box helicase 58 [DDX58] ) and MDA5 ( interferon induced with helicase C domain 1 [IFIH1] ) , and noticed that all cell lines except K562 up-regulated these transcripts in response to introduction of gRNAs . We also confirmed these results on the protein level in HEK293 cells ( Fig 2B ) . The RIG-I and MDA5 receptors complement each other by recognizing different structures in foreign cytosolic RNAs , but the exact nature of their ligands is not yet fully understood [31 , 32] . To investigate if IVT gRNAs are recognized via RIG-I or MDA5 , we generated clonal knockout ( KO ) cell lines for RIG-I , MDA5 , and their downstream interaction partner MAVS in HEK293 cells using CRISPR-Cas9 . As the expressions of both RIG-I and MDA5 are themselves stimulated by IFNβ , we confirmed successful KO after transfection with gRNAs by genomic PCR , Sanger sequencing , and western blot ( S1A–S1C Fig ) . MAVS KO cells were confirmed by western blot ( S1D Fig ) . Strikingly , activation of IFNB1 expression after introduction of gRNAs was absent in RIG-I and MAVS KO cells , while MDA5 KO cells did not show a significant decrease in IFNB1 transcript levels ( Fig 2C ) . This indicates that IVT gRNAs are exclusively recognized through RIG-I to trigger a type I interferon response . As the structural requirements of RIG-I ligands are still not completely understood , we wondered if different 20-nucleotide protospacers in gRNAs vary in their potency to trigger an innate immune response via RIG-I . We designed 10 additional nontargeting gRNAs that we in vitro transcribed and transfected into HEK293 cells . Surprisingly , we found that the cells responded to different protospacers with a wide range of IFNB1 expression . Several gRNAs produced very little innate immune response , and one gRNA ( gRNA11 ) yielded no IFNB1 activation at all ( Fig 3A ) . We speculated that the differential response may be correlated with the purity of the RNA product after in vitro transcription or the stability of the secondary structure of the RNA [33 , 34] . However , we found that there was no obvious correlation between the immune response to certain gRNAs and their purity; predicted protospacer secondary structure; full secondary structure , including the constant region; or predicted disruption of the constant region by mispairing with the protospacer ( S2 Fig ) . When we separately nucleofected five of these gRNAs into primary CD34+ human hematopoietic stem and progenitor cells ( HSPCs ) , we found that all gRNAs induced a strong immune response . Only gRNA11 , which showed no immune stimulation in HEK293 cells , resulted in half the amount of ISG15 transcript ( Fig 3B ) . These results indicate that RIG-I recognition patterns of IVT gRNAs are complex and difficult to anticipate a priori based on predicted properties of the variable protospacer and cell type . One well-established structural requirement of RIG-I ligands is the presence of a 5’-triphosphate group [35] . We asked if preparations that remove the 5’ triphosphate might avoid or reduce the innate immune response to IVT gRNAs . We first used a synthetic gRNA that lacks a 5’-triphosphate and verified that this gRNA does not induce IFNB1 expression when transfected into HEK293 cells ( Fig 3C ) . Synthetic gRNAs are becoming more commonplace but are still an order of magnitude more expensive than in vitro transcription of gRNAs . This limits their application for high-throughput interrogation of gene function in primary cells . We therefore asked if treatment of IVT gRNA with phosphatases that remove the 5’-triphosphate would reduce IFNB1 induction . We tested calf intestinal alkaline phosphatase ( CIP ) , shrimp alkaline phosphatase ( SAP ) , 5’-RNA polyphosphatase ( PP ) , and thermosensitive alkaline phosphatase ( AP ) and found that phosphatase treatment with CIP and AP abolished the IFNB1 response , while SAP and PP treatment only resulted in a reduction of the response ( Fig 3C ) . We also compared purification of IVT gRNAs by solid-phase reversible immobilization ( SPRI ) beads to column purification and established that SPRI bead cleanup is not sufficient to completely avoid an immune response , even when more phosphatase is used ( S3A–S3B Fig ) . Taken together , these results indicate that 5’-triphosphate is a necessary requirement for gRNA-induced IFNB1 activation through RIG-I but that additional structural properties of the gRNAs also influence the magnitude of the immune response . Next , we asked if phosphatase treatment alters the genome editing potential of gRNAs . As gRNA1 targets the BFP gene , we used a HEK293T cell line with a stably integrated BFP reporter [26] , nucleofected cells with phosphatase-treated gRNA-Cas9 RNPs , and monitored editing outcomes by T7 endonuclease I assay ( S3C Fig ) . We did not observe any significant difference in editing outcomes between synthetic , IVT , and phosphatase-treated gRNAs , suggesting that phosphatase treatment does not affect the function of the gRNA . When a cell initiates an antiviral immune response , it also undergoes cellular stress that can affect cell viability [36 , 37] . Hence , we asked if there is a correlation between the IFNβ response and cell viability after transfection with synthetic , IVT , or CIP-treated IVT gRNA . Not surprisingly , the viability of the very robust HEK293 cell line was not affected by the antiviral immune response ( S3D Fig ) . We then turned to HSPCs , which are a much more sensitive cell type . We first nucleofected HSPCs with RNPs targeting the hemoglobin subunit beta ( HBB ) gene [12] and compared synthetic and IVT gRNA interferon stimulation and cell viability posttransfection . Double-strand breaks ( DSBs ) have been reported to cause innate immune stimulation and can themselves cause decreases in cell fitness [38 , 39] . Therefore , we performed controls using nuclease-dead Cas9 ( dCas9 ) to form RNPs and confirmed by Sanger sequencing and TIDE analysis that dCas9-RNPs did not induce DSBs [40] ( S3E Fig ) . We found a significant decrease in HSPC viability using both of the IVT gRNA RNPs that had an increase in IFN-stimulated genes ISG15 and RIG-I ( Fig D-E ) . We did not see a substantial difference in viability or ISG expression between Cas9 and dCas9 RNPs , suggesting that nuclease activity leading to DNA damage did not cause the immune response . Next , we asked if CIP treatment of gRNAs could reverse the decrease in viability in HSPCs . We nucleofected HSPCs with dCas9 RNPs targeting a noncoding intron of Janus kinase 2 ( JAK2 ) or Cas9 RNPs targeting the HBB gene and compared synthetic , IVT , and CIP-treated IVT gRNAs . Strikingly , CIP treatment restored viability in HSPCs ( Fig 3F ) . We were also interested in editing outcomes in these samples and performed amplicon next-generation sequencing ( NGS ) for the HBB locus . While the phosphatase-treated gRNA performed similarly to the synthetic gRNA , the IVT gRNA resulted in slightly fewer insertions and deletions ( indels ) ( Fig 3G ) .
We have found that IVT gRNAs used with Cas9 RNPs for many genome-editing experiments can trigger a strong innate immune response in many mammalian cell types ( Fig 4 ) . Lipofection results in a stronger and longer-lasting response than nucleofection , possibly because lipofection delivers gRNAs to the cytosol , while nucleofection delivers mainly to the nucleus . Using isogenic KO clones , we found the gRNA-induced response is mediated via the antiviral RIG-I pathway and results in expression of genes that initiate an antiviral immune response . While introduction of IFN-stimulating gRNAs does not affect viability in HEK293 cells , we found that viability of primary HSPCs is negatively affected by the antiviral immune response . While DSBs have on their own been reported to induce an innate immune response [38] , we found triphosphate-containing gRNAs complexed with dCas9 induce an immune response and cell death in HSPCs . Only removal of the triphosphate is sufficient to reduce gRNA-induced innate immune signaling . These results have several implications . We suggest that the gene signature associated with type I interferon stimulation should be considered when studying the transcriptome of recently edited bulk populations of cells . Furthermore , all mammalian cells can both produce type I interferons and also respond to them through the ubiquitously expressed receptor interferon alpha and beta receptor subunit 1 ( IFNAR1 ) [41] . Even cells that have not been successfully transfected with RNPs could sense the IFNβ produced by neighboring cells and activate downstream antiviral defense mechanisms . This could be an important consideration during in vivo genome editing applications , as RNP delivery into one set of cells could provoke a widespread innate immune response in the surrounding tissues . We found that synthetic gRNAs completely circumvent the RIG-I mediated response , offering a valuable path to avoid innate immune signaling during therapeutic editing . However , synthetic gRNAs can become expensive when performing experiments that require testing or using many gRNAs . We found that a cost-effective phosphatase treatment to remove the 5’-triphosphate before transfection reduces the immune response and increases posttransfection viability in HSPCs . Furthermore , editing outcomes in cell lines with phosphatase-treated gRNA were comparable to those of IVT gRNAs , suggesting that removal of 5’-phosphate groups does not abolish gRNA function . In fact , in sensitive HSPCs , phosphatase-treated gRNA slightly outperformed IVT gRNA , which is possibly due to reduced viability in samples transfected with IVT-RNPs . Thus , consideration of a potential innate immune stimulation prior to choice of genome editing reagents , study design , and implementation of controls is critical when performing genome editing using RNPs in mammalian cells . While we were preparing this manuscript for submission , the Kim group reported similar results in HeLa cells and primary human CD4+ T cells [42] . They confirmed that the type I interferon response is dependent on the presence of a 5’-triphopsphate group and that CIP treatment can increase viability by avoiding the antiviral response . These results are very much in alignment with our findings and extend the potential problem of innate immune signaling to additional cell types . Our study adds extra depth by further outlining the mechanisms by which gRNAs are sensed . We show that gRNA sensing depends upon RIG-I and MAVS , but MDA5 KO cells are fully capable of inducing IFNβ after IVT gRNA transfection . Hence , gRNA sensing is independent of the MDA5 PAMP receptor , consistent with RIG-I’s preference for short double-stranded RNA ( dsRNA ) structures and MDA5’s preference for long dsRNA fragments [43] . Furthermore , we show that in addition to a 5’-triphosphate , the protospacer sequence is also critical to determine the intensity of the IFNβ response . Not only do different gRNAs induce different innate immune responses , but some gRNAs induce no response at all . However , this seems to be cell-type specific , as we found that sensitive cells such as primary HSPCs react to the same gRNAs with a strong immune response independently of the protospacer . It has been proposed that 5’-base-paired RNA structures are required to activate antiviral signaling via RIG-I , but we found no correlation between signaling and a variety of predicted RNA properties , including secondary structure [33] . Our results therefore suggest that the mechanism of gRNA sensing by the RIG-I pathway is relatively complex in that it requires 5’-triphosphates but that this moiety is not sufficient to induce the response . Additionally , we have not ruled out the possibility that gRNAs could be recognized by Toll-like receptors ( TLRs ) , though we and others [42] have found that KO of RIG-I is sufficient to completely abrogate gRNA-induced signaling in multiple cell contexts . The role of TLR recognition could be addressed in future work to delineate the full set of molecular features responsible for gRNA activation of innate immunity , which might yield accurate predictors of innate immune signaling in general .
gRNA was synthesized by assembly PCR and in vitro transcription as previously described [12] . Briefly , a T7 RNA polymerase substrate template was assembled by PCR from a variable 58–59 nt primer containing T7 promoter , variable gRNA guide sequence , the first 15 nt of the nonvariable region of the gRNA ( T7FwdVar primers , 10 nM , S1 and S2 Tables for gRNA sequences ) , and an 83 nt primer containing the reverse complement of the invariant region of the gRNA ( T7RevLong , 10 nM ) , along with amplification primers ( T7FwdAmp , T7RevAmp , 200 nM each ) . The two long primers anneal in the first cycle of PCR and are then amplified in subsequent cycles . Phusion high-fidelity DNA polymerase was used for assembly ( New England Biolabs ) . Assembled template was used without purification as a substrate for in vitro transcription by T7 RNA polymerase , using the HiScribe T7 High Yield RNA Synthesis kit ( New England Biolabs ) following the manufacturer’s instructions . Resulting transcription reactions were treated with DNAse I ( New England Biolabs ) , and RNA was purified by treatment with a 5X volume of homemade SPRI beads ( comparable to Beckman-Coulter AMPure beads ) and elution in RNAse-free water . gRNAs were treated with phosphatases as follows: CIP ( New England Biolabs , 30 U ) , SAP ( New England Biolabs 10 U ) , PP ( Lucigen , 20 U ) , and FastAP AP ( Thermo Fisher Scientific , 10 U ) were added per 20 μl in vitro transcription reaction , and samples were incubated at 37°C for 3 h before proceeding to purification and DNAseI treatment . gRNA was purified using a Qiagen RNeasy Mini Kit ( Qiagen ) or by 5X volume of homemade SPRI beads ( comparable to Beckman-Coulter AMPure beads ) . The detailed protocol and additional notes are available online ( dx . doi . org/10 . 17504/protocols . io . nghdbt6 ) . HCV PAMP in vitro transcription template [21] was generated by annealing HCV fwd and rev ( 5 μM each ) oligos ( S1 Table ) . In the subsequent in vitro transcription reaction , 2 μl of the annealed product was used as DNA template , using HiScribe T7 High Yield RNA Synthesis kit ( New England Biolabs ) . The plasmid containing the SeV DI RNA[28] was a gift from Prof . Peter Palese , Icahn School of Medicine at Mount Sinai , New York . Plasmid was digested with HindII/EcoRI before in vitro transcription with HiScribe T7 High Yield RNA Synthesis kit ( New England Biolabs ) . The sequence of the IVT DI , including the T7 promoter , hepatitis delta virus ribozyme , and the T7 terminator , is TAATACGACTCACTATAACCAGACAAGAGTTTAAGAGATATGTATCCTTTTAAATTTTCTTGTCTTCTTGTAAGTTTTTCTTACTATTGTCATATGGATAAGTCCAAGACTTCCAGGTACCGCGGAGCTTCGATCGTTCTGCACGATAGGGACTAATTATTACGAGCTGTCATATGGCTCGATATCACCCAGTGATCCATCATCAATCACGGTCGTGTATTCATTTTGCCTGGCCCCGAACATCTTGACTGCCCCTAAAATCTTCATCAAAATCTTTATTTCTTTGGTGAGGAATCTATACGTTATACTATGTATAATATCCTCAAACCTGTCTAATAAAGTTTTTGTGATAACCCTCAGGTTCCTGATTTCACGGGATGATAATGAAACTATTCCCAATTGAAGTCTTGCTTCAAACTTCTGGTCAGGGAATGACCCAGTTACCAATCTTGTGGACATAGATAAAGATAGTCTTGGACTTATCCATATGACAATAGTAAGAAAAACTTACAAGAAGACAAGAAAATTTAAAAGGATACATATCTCTTAAACTCTTGTCTGGTGGCCGGCATGGTCCCAGCCTCCTCGCTGGCGCCGGCTGGGCAACATTCCGAGGGGACCGTCCCCTCGGTAATGGCGAATAGCATAACCCCTTGGGGCCTCTAAACGGGTCTTGAGGGGTTTTTTG . The sequence of the SeV DI is highlighted in boldface . Both HCV PAMP and SeV DI RNA were purified by treatment with a 5X volume of homemade SPRI beads ( comparable to Beckman-Coulter AMPure beads ) and elution in RNAse-free water . Chemically synthesized gRNAs , which were purified using high-performance liquid chromatography ( HPLC ) , were purchased from Synthego . IVT gRNAs were analyzed using a Bioanalyzer . This was performed by the UC Berkeley Functional Genomics Laboratory ( FGL ) core facility . gRNAs were denatured for 5 min at 70°C before analysis on bioanalyzer . The Cas9 construct ( pMJ915 ) contained an N-terminal hexahistidine-maltose binding protein ( His6-MBP ) tag , followed by a peptide sequence containing a tobacco etch virus ( TEV ) protease cleavage site . The protein was expressed in Escherichia coli strain BL21 Rosetta 2 ( DE3; EMD Biosciences ) grown in TB medium at 16°C for 16 h following induction with 0 . 5 mM IPTG . The Cas9 protein was purified by a combination of affinity , ion exchange , and size exclusion chromatographic steps . Briefly , cells were lysed in 20 mM HEPES pH 7 . 5 , 1 M KCl , 10 mM imidazole , 1 mM TCEP , 10% glycerol ( supplemented with protease inhibitor cocktail [Roche] ) in a homogenizer ( Avestin ) . Clarified lysate was bound to Ni-NTA agarose ( Qiagen ) . The resin was washed extensively with lysis buffer , and the bound protein was eluted in 20 mM HEPES pH 7 . 5 , 100 mM KCl , 300 mM imidazole , 1 mM TCEP , 10% glycerol . The His6-MBP affinity tag was removed by cleavage with TEV protease , while the protein was dialyzed overnight against 20 mM HEPES pH 7 . 5 , 300 mM KCl , 1 mM TCEP , 10% glycerol . The cleaved Cas9 protein was separated from the fusion tag by purification on a 5 ml SP Sepharose HiTrap column ( GE Life Sciences ) , eluting with a linear gradient of 100 mM–1 M KCl . The protein was further purified by size exclusion chromatography on a Superdex 200 16/60 column in 20 mM HEPES pH 7 . 5 , 150 mM KCl , and 1 mM TCEP . Eluted protein was concentrated to 40 uM , flash-frozen in liquid nitrogen , and stored at −80°C . Cells were obtained from ATCC and verified mycoplasma-free ( Mycoalert LT-07 , Lonza ) . HEK293 , HEK293T , HCT116 , HepG2 , and HeLa cells were maintained in DMEM supplemented with 10% FBS and 100 μg/mL penicillin-streptomycin ( all Gibco ) . K562 and Jurkat cells were maintained in RPMI supplemented with 10% FBS and 100 μg/mL penicillin-streptomycin . All transfections in cell lines were performed in 12-well cell culture dishes using 2 × 105 cells per transfection . For lipofection , we used Lipofectamine CRISPRMAX-Cas9 , Lipofectamine RNAiMAX , or Lipofectamine 2000 Transfection Reagent ( all Invitrogen ) in reverse transfections according to the manufacturer’s protocols . Unless stated otherwise , 2 × 105 cells were transfected with 50 pmol of RNA to a final concentration of 50 nM and harvested 24–30 h posttransfection for RNA extraction . HSPCs from mobilized peripheral blood ( Allcells ) were thawed and cultured in StemSpan SFEM medium ( StemCell Technologies ) supplemented with StemSpan CC110 cocktail ( StemCell Technologies ) for 48 h before nucleofection with dCas9 or Cas9 RNP ( 75 pmol of dCas9 , 75 pmol of gRNA ) . Then , 1 . 5 × 105 HSPCs were pelleted ( 100 × g , 10 min ) and resuspended in 20 μl Lonza P3 solution , mixed with 10 μl dCas9 or Cas9 RNP , and nucleofected using ER100 protocol in Lonza 4D nucleofector . Viability of the cells was measured 24 h postnucleofection using trypan blue exclusion test . RNA was harvested 16 h postnucleofection . Cell cultures were washed with PBS prior to RNA extraction . Total RNA was extracted using RNeasy Miniprep columns ( Qiagen ) according to the manufacturer’s instructions , including the on-column DNAseI treatment ( Qiagen ) . One μg of total RNA was used for subsequent cDNA synthesis using Reverse Transcription Supermix ( Biorad ) . For qRT-PCR reactions , a total of 20 ng of cDNA was used as a template and combined with primers ( see S3 Table ) , and EvaGreen Supermix ( Biorad ) and amplicons were generated using standard PCR amplification protocols for 40 cycles on a StepOnePlus Real-Time PCR system ( Applied Biosystems ) . Ct values for each target gene were normalized against Ct values obtained for GAPDH to account for differences in loading ( ΔCt ) . To determine “fold activation” of genes , ΔCt values for target genes were then normalized against ΔCt values for the same target gene for mock-treated cells ( ΔΔCt ) . For CRISPR/Cas9 genome editing , we used a plasmid encoding both the Cas9 protein and the gRNA . pSpCas9 ( BB ) -2A-GFP ( px458 ) was a gift from Feng Zhang ( Addgene plasmid #48138 ) . We designed gRNA sequences using the free CRISPR KO design online tool from Synthego . Two different gRNA sequences were designed for RIG-I and MDA5 , respectively ( see S3 Table ) . Using a Lonza 4D nucleofector ( Lonza ) with the manufacturer’s recommended settings , 2 × 105 HEK293 cells were nucleofected with 2 μg of px458 plasmids containing both targeting gRNAs in a 1:1 ratio . After 48 h , cells were harvested and subjected to fluorescence-activated cell sorting ( FACS ) . Cells expressing high levels of GFP were single-cell sorted into 96-well plates to establish clonal populations . For the screening process , genomic DNA ( gDNA ) from clonal populations was extracted using QuickExtract solution ( Lucigen ) . For KO of RIG-I and MDA5 , we screened clones by genomic PCR , looking for a PCR product that is significantly smaller in size than that of WT HEK293 cells ( see S4 Table for primers ) . PCR products were then Sanger sequenced by the UC Berkeley DNA Sequencing facility using the forward primers of the PCR reaction as sequencing primers . Cells were harvested and washed with PBS . Cells were lysed in 1x RIPA buffer ( EMD Millipore ) for 10 min on ice . Samples were spun down at 14 , 000 × g for 15 min , and protein lysates were transferred to a new tube . Fifty μg of total protein was separated for size by SDS-PAGE and transferred to a nitrocellulose membrane . Blots were blocked in 4% skim milk in 50 mm Tris- HCl ( pH 7 . 4 ) , 150 mm NaCl , and 0 . 05% Tween 20 ( TBST ) and then probed for RIG-I , MDA5 , MAVS , or GAPDH protein using antibodies against RIG-I ( D14G6 ) , MDA5 ( D74E4 ) , MAVS ( D5A9E ) , or GAPDH ( 14C10 ) , respectively ( all Cell Signaling Technologies ) . This was followed by incubation with secondary antibody IRDye 800CW Donkey anti-Rabbit IgG ( Li-Cor ) . Protein standards ( GE Healthcare ) were loaded in each gel for size estimation . Blots were visualized using a Li-Cor Odyssey Clx ( Li-Cor ) . Cells were harvested 24 h after transfection and washed with PBS . gDNA was extracted using QuickExtract solution ( Lucigen ) following the manufacturer’s protocol . PCR across the target site in the BFP gene was run using the BFP amplicon primer set ( S4 Table ) . Two hundred ng of PCR product was heated to 100°C and slowly cooled down to let DNA reanneal . Annealed DNA was digested with T7 endonuclease I ( NEB ) for 20 min at 37°C . DNA was then analyzed by agarose gel electrophoresis . PCR products were generated with target-specific HBB primer set 1 , sequenced , and Sanger traces were then analyzed with the TIDE webtool ( http://tide . nki . nl ) . Using primer set 1 , 50–100 ng of gDNA from edited CD34+ cells was amplified at HBB sites ( S4 Table ) . The PCR products were SPRI cleaned , followed by amplification of 20–50 ng of the first PCR product in a second 12-cycle PCR using primer set 2 ( S4 Table ) . Then , the second PCR products were SPRI cleaned , followed by amplification of 20–50 ng of the second PCR product in a third 9 cycle PCR using illumina-compatible primers ( primers designed and purchased through the Vincent J . Coates Genomics Sequencing Laboratory [GSL] at University of California , Berkeley ) , generating indexed amplicons of an appropriate length for NGS . Libraries from 100–500 pools of edited cells were pooled and submitted to the GSL for paired-end 300 cycle processing using a version 3 Illumina MiSeq sequencing kit ( Illumina , San Diego , CA ) after quantitative PCR measurement to determine molarity . Samples were deep sequenced on an Illumina MiSeq at 300 bp paired-end reads to a depth of at least 10 , 000 reads . A modified version of CRISPResso [44] was used to analyze editing outcomes . Briefly , reads were adapter trimmed and then joined before performing a global alignment between reads and the reference sequence using NEEDLE [45] . Indel rates were calculated as any reads in which an insertion or deletion overlaps the cut site or occurs within 3 base pairs of either side of the cut site , divided by the total number of reads . | Clustered , regularly interspaced , short palindromic repeat ( CRISPR ) –CRISPR-associated 9 ( Cas9 ) genome editing is transforming fundamental research , as it allows researchers to make targeted changes to the genome of cells . For efficient editing , the Cas9 protein ( a DNA nuclease ) and a guide RNA ( gRNA ) , which leads the nuclease to the correct location in the genome , have to be introduced into cells . One recent advancement is the delivery of a Cas9 protein and an in vitro–transcribed ( IVT ) gRNA as a precomplexed ribonucleoprotein ( RNP ) . This approach allows editing of more sensitive cell types such as immune cells and hematopoietic stem cells . However , the consequences of introducing foreign Cas9 nuclease and gRNA into mammalian cells are not fully understood . Here , we show that in many cell types , the IVT gRNAs trigger a potent innate immune response—a natural defense mechanism against RNA viruses . We show that the innate immune response causes cell death in primary hematopoietic stem cells but that removal of the 5’-triphosphate from gRNAs by phosphatase treatment can ameliorate the immune response and prevent the loss of viability . Hence , CRISPR-Cas9 RNP editing has the potential to induce a potent antiviral response , and we suggest that care must be taken when designing therapeutic strategies to edit primary cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"transfection",
"medicine",
"and",
"health",
"sciences",
"nucleic",
"acid",
"synthesis",
"enzymes",
"immunology",
"rna",
"extraction",
"enzymology",
"phosphatases",
"molecular",
"biology",
"techniques",
"rna",
"synthesis",
"extraction",
"techniques",
"chemical",
"synthesi... | 2018 | In vitro–transcribed guide RNAs trigger an innate immune response via the RIG-I pathway |
Shp2 is a cytoplasmic protein-tyrosine phosphatase that is essential for normal development . Activating and inactivating mutations have been identified in humans to cause the related Noonan and LEOPARD syndromes , respectively . The cell biological cause of these syndromes remains to be determined . We have used the zebrafish to assess the role of Shp2 in early development . Here , we report that morpholino-mediated knockdown of Shp2 in zebrafish resulted in defects during gastrulation . Cell tracing experiments demonstrated that Shp2 knockdown induced defects in convergence and extension cell movements . In situ hybridization using a panel of markers indicated that cell fate was not affected by Shp2 knock down . The Shp2 knockdown–induced defects were rescued by active Fyn and Yes and by active RhoA . We generated mutants of Shp2 with mutations that were identified in human patients with Noonan or LEOPARD Syndrome and established that Noonan Shp2 was activated and LEOPARD Shp2 lacked catalytic protein-tyrosine phosphatase activity . Expression of Noonan or LEOPARD mutant Shp2 in zebrafish embryos induced convergence and extension cell movement defects without affecting cell fate . Moreover , these embryos displayed craniofacial and cardiac defects , reminiscent of human symptoms . Noonan and LEOPARD mutant Shp2s were not additive nor synergistic , consistent with the mutant Shp2s having activating and inactivating roles in the same signaling pathway . Our results demonstrate that Shp2 is required for normal convergence and extension cell movements during gastrulation and that Src family kinases and RhoA were downstream of Shp2 . Expression of Noonan or LEOPARD Shp2 phenocopied the craniofacial and cardiac defects of human patients . The finding that defective Shp2 signaling induced cell movement defects as early as gastrulation may have implications for the monitoring and diagnosis of Noonan and LEOPARD syndrome .
Shp2 ( PTPN11 ) , a nonreceptor protein-tyrosine phosphatase ( PTP ) with two Src homology 2 ( SH2 ) domains , has a central role in cell signaling . Shp2 is essential for embryonic development and has a role in human disease [1] . Dominant negative Shp2 inhibits mesoderm formation in Xenopus laevis [2] . Gene targeting of Shp2 in Mus musculus leads to truncation of the Shp2 protein resulting in embryonic death around day 8 . 5 , with a range of defects consistent with abnormal gastrulation [3] . Chimeric mice derived from Shp2 ex3−/− embryonic stem cells display defective morphogenetic cell movements during gastrulation [4] . Interestingly , bona fide Shp2 null mouse embryos die peri-implantation and Shp2 is required for trophoblast stem cell survival [5] . Activating mutations in Shp2 cause Noonan Syndrome ( NS ) in humans , whereas LEOPARD syndrome ( LS ) is caused by dominant negative mutations in Shp2 . NS is an autosomal dominant disorder affecting around 1 in 2 , 000 live births , and is characterized by multiple defects , including short stature , facial abnormalities , and congenital heart defects [6] . Around 50% of NS cases are caused by mutations in Shp2 [7] and 39 different mutations have been identified [8 , 9] . Most of the NS mutations are localized in the N-SH2 domain or in the PTP domain , and result in activation of Shp2 catalytic activity [10] . The mouse model for NS with an activating D61G mutation bears striking similarities to NS patients , with defects such as short stature , facial dysmorphism , and multiple cardiac defects . Mice homozygous for the mutated gene die prenatally from severe cardiac edema and liver necrosis [11] . LS is an autosomal dominant disease characterized by defects such as lentigines , electrocardiographic defects , ocular hypertelorism , pulmonary stenosis , abnormal genitals , retarded growth leading to short stature , and deafness [12] . Many of the LS symptoms overlap with those seen in NS patients , and LS is also caused by mutations in Shp2 . These mutations occur exclusively in the PTP domain of Shp2 disrupting its catalytic activity and leading to dominant negative forms [13 , 14] . This leads to the conundrum of how mutations in Shp2 with opposing effects on activity induce syndromes in humans with similar symptoms . Shp2 is known to be involved in many different signaling cascades [15] . Initially , Shp2 was found to have a role in growth factor signaling , downstream of growth factor receptors and upstream of MAP kinase signaling . Zhang et al . [16] demonstrated that Shp2 controls phosphorylation and activation of Src family kinases ( SFKs ) , in that SFK activity is reduced in the absence of Shp2 , due to hyperphosphorylation of the inhibitory phosphorylation site . Two SFKs , Fyn and Yes , have a role in vertebrate convergence and extension ( CE ) cell movements during gastrulation [17] . Moreover , the C-terminal Src kinase , Csk , a negative regulator of SFKs , is crucial for normal gastrulation cell movements as well [18] . Morpholino ( MO ) -mediated Fyn/Yes knockdown in zebrafish embryos phenocopies Silberblick/Wnt11 and Pipetail/Wnt5 morphants both morphologically and molecularly . Although Fyn and Yes act in a synergistic manner with Wnt5 and Wnt11 , they do not function in a linear pathway; instead , they operate in parallel , converging downstream on the small GTPase RhoA . Given the role of Fyn and Yes in gastrulation cell movements , the function of Shp2 as an indirect activator of SFKs , and evidence from Xenopus and mouse that Shp2 may have a role in gastrulation cell movements , we hypothesized that Shp2 is involved in CE cell movements during gastrulation . Here , we show that MO-mediated knockdown of Shp2 in zebrafish embryos resulted in defective CE cell movements during gastrulation , but not cell specification . Genetic epistasis analyses indicated that Shp2 acts through Fyn and Yes , upstream of the small GTPase RhoA . Expression of mutant NS- or LS-Shp2 in zebrafish embryos resulted in overlapping phenotypes , in that normal gastrulation was impaired without affecting cell specification . Specifically , we found that NS-Shp2 induced CE cell movement defects . Moreover , craniofacial and cardiac development were also impaired . The NS- and LS-Shp2s did not act synergistically upon coinjection , which is consistent with the two mutants acting in the same signaling pathway with opposing effects . The defects resulting from expression of mutant Shp2s correspond to symptoms in human NS/LS patients . The notion that defective Shp2 signaling induced cell movement defects as early as gastrulation is important and may have implications for the monitoring and diagnosis of NS and LS .
Zebrafish Shp2 ( Ensembl database ENSDARG00000020334 , http://www . ensembl . org ) is highly homologous to human and mouse Shp2 ( 91 . 2% and 90 . 3% protein sequence identity , respectively ) ( Figure 1A ) . In situ hybridization experiments show that shp2 is ubiquitously expressed during early zebrafish development ( Figure 1B–1F ) . At 24 h post fertilization ( hpf ) and 3 d post fertilization ( dpf ) , shp2 was broadly expressed , with enhanced levels of expression in the anterior parts of the embryo ( Figure 1G and 1H ) . We designed a Shp2-MO targeting the start codon and injected it at the one-cell stage . We found that 1ng Shp2-MO consistently produced specific defects in embryonic development . The first visible defect is a failure of the embryo to extend properly around the yolk at 10 hpf ( Figure 2A–2C ) . At later stages ( 4 dpf ) , embryos were shorter and developed a hammerhead phenotype similar to Wnt5 morphants ( Figure 2D–2F ) . Alcian blue staining of 4 dpf embryos indicated that the cartilaginous structures in the head of Shp2 and Wnt5 morphants reside more posteriorly than in uninjected controls; compare Meckel's cartilage ( black asterisk ) and the ceratohyal ( red asterisk ) ( Figure 2G–2I ) . We quantified the hammerhead phenotype by assessment of the ratio of the distance to the tip of the nose divided by the distance between the eyes ( as illustrated in Figure S1 ) . There was a significant difference in this ratio between wild-type ( WT ) ( 1 . 87 ± 0 . 03 , n = 4 ) and Shp2-MO injected ( 1 . 40 ± 0 . 20 , n = 10 ) embryos ( p < 0 . 001 , student's t-test ) , demonstrating that there was a significant craniofacial defect following Shp2 knockdown . Coinjection of 300 pg human shp2 RNA rescued all Shp2-MO-induced defects ( Figure 2J ) . Higher amounts of Shp2-MO induced more severe defects ( unpublished data ) , but not all defects were rescued by coinjection of synthetic RNA . Therefore , we used 1 ng Shp2-MO for all subsequent knockdowns . Reduced extension of the body axis upon Shp2 knockdown is consistent with CE cell movement defects . Therefore , we directly investigated whether Shp2 knockdown affected CE cell movements by cell tracing . Caged fluorescein was ( co- ) injected at the one-cell stage . At 6 hpf , a cluster of cells within the dorsal shield was labeled by uncaging the caged fluorescein with a short , localized pulse of UV light . This group of cells was monitored every 2 h during gastrulation . The distance the cells migrated is directly proportional to embryonic extension . Repeating the process at 90° to the shield gives an effective measurement of convergence of the mesendodermal cells . Shp2 knockdown resulted in a significant reduction in CE during gastrulation ( Figure 2K , L ) . These results demonstrate that Shp2 knockdown induced defects in CE cell movements during gastrulation . Shp2 is known to be involved in many different signaling cascades [15] . In fact , in a recent study in Drosophila , over 40 different genes associated with at least four separate pathways ( EGFR , Notch , DPP , and Jak/Stat ) were found to interact with gain-of-function Shp2 mutants [19] . Some of these pathways are important for proper cell specification and defective signaling leads to a variety of phenotypes , including defects in the A-P axis . We used a panel of well-characterized in-situ markers that have been used before to assess effects on cell specification [20–23] . Bone morphogenetic protein 2b ( bmp2b ) is involved in the specification of ventral cell fates but was not affected by knockdown of Shp2 ( Figure 3A and 3B ) . Expression of the dorsalizing factor chordin ( chd ) remained the same ( Figure 3C and 3D ) , as did the dorsal specific gene goosecoid ( gsc ) , which is expressed in the organizer ( Figure 3E and 3F ) . Expression of the mesendodermal marker notail ( ntl ) also remained unchanged in Shp2 morphants ( Figure 3G and 3H ) . These markers indicate that knockdown of Shp2 did not alter cell fate in early zebrafish embryos , suggesting it has a role in CE cell movements , rather than cell specification . Expression of the anterior brain markers , six3 ( forebrain ) , pax2 ( mid-hindbrain boundary ) , and krox20 ( rhombomeres 3 and 5 ) persisted in the Shp2 morphants , indicating that the structures these markers delineate were present . However , the expression patterns of these three genes were shifted posteriorly ( Figure 3Il–Ll ) , which—in the case of pax2 and six3—was accompanied by a broader expression pattern ( Figure 3Ka–La ) . At 8 hpf , cyclops ( cyc ) is expressed in axial mesendodermal cells of gastrulating embryos . Shp2 morphants clearly have a shorter expression pattern when compared to uninjected controls ( Figure 3O and 3P ) . These results are consistent with disrupted convergence and extension cell movements upon knockdown of Shp2 . Shp2 might regulate CE cell movements by modulation of expression of the noncanonical Wnts , or by Fyn and Yes , known regulators of CE cell movements . However , the expression of wnt11 , wnt5 , fyn , and yes remained unaffected in Shp2-MO injected embryos ( Figure 3Q–3X ) , indicating that Shp2 has a more direct role in CE signaling . In order to assess the mechanism underlying Shp2-mediated cell movements in gastrulation , we coinjected MOs with synthetic mRNAs and assessed CE cell movement defects by measurement of the angle between the most anterior and the most posterior tissues at 10 hpf . The Shp2-MO by itself induced an increase in the angle between the extremes of the developing embryo , which was rescued by coinjection of synthetic human shp2 mRNA ( Table 1 ) . Coinjection of RNA encoding constitutively active Fyn and Yes with Shp2-MO also rescued the Shp2 morphants ( Table 1 ) , indicating that Fyn and Yes are genetically downstream of Shp2 . Fyn- and Yes-MOs induced severe reductions in embryo body axis extension , and coinjection with synthetic shp2 mRNA did not rescue ( Table 1 ) , confirming that Shp2 is upstream of Fyn and Yes . Low amounts of Shp2-MO together with Wnt5-MO , which did not induce defects when coinjected alone , induced a hammerhead phenotype at 4 dpf ( Figure S2 ) , indicating that Shp2 and Wnt5 interact genetically . However , Wnt5 mRNA did not rescue the Shp2 morphants and shp2 mRNA did not rescue Wnt5 morphants ( Table 1 ) , indicating that Shp2 and Wnt5 do not operate in the same linear genetic pathway . Active RhoA rescued the Shp2 morphants ( Table 1 ) , which is consistent with Shp2 being upstream of Fyn and Yes , which , in turn , act upstream of RhoA . Shp2 is most commonly associated with the Ras/MAPK pathway , which regulates many developmental processes , such as cell proliferation and cell specification [24 , 25] . However , the phenotype we observed in the Shp2 knockdown embryos was not consistent with a massive reduction in cell proliferation , nor with changes in cell specification ( Figure 3 ) . Given that the Shp2 knockdown phenotype was rescued by coinjection of active RhoA , the Ras/MAPK signaling pathway appeared not to be essential for Shp2 signaling during gastrulation . Instead , we implicate SFKs and RhoA downstream in the Shp2 signaling cascade . In Xenopus laevis , mutant , active Shp2 induces elongation of animal cap explants , which is blocked by coexpression of dominant negative RhoA [26] , suggesting involvement of RhoA rather than Ras/MAPK , similar to what we observed in early zebrafish embryos . In a recent report , Shp2 knockdown was reported to induce craniofacial hypoplasia and heart malformations , similar to Raf1 knockdown [27] . Mutations in Raf1 were linked to NS [27 , 28] . Other Ras/MAPK signaling components were identified in NS as well , including KRAS and SOS1 [29–31] . Although we cannot exclude that Ras/MAPK signaling has a role in Shp2 signaling in gastrulation cell movements , we demonstrate here that we can rescue the Shp2 knockdown phenotype with active SFKs or active RhoA , indicating that SFKs and RhoA are downstream of Shp2 in gastrulation cell movements . To investigate the use of zebrafish as a model for NS and LS , we generated two NS-Shp2 and two LS-Shp2 mutants by introducing mutations into zebrafish Shp2 , as found in NS and LS patients , respectively ( Figure 4A ) . For NS , we substituted Asp61 with Gly ( D61G ) or Thr73 with Ile ( T73I ) . For LS , Ala462 was mutated to Thr ( A462T ) , or Gly465 to Ala ( G465A ) . The two NS proteins showed a 6-fold increase in activity compared to WT Shp2 in in vitro PTP assays , whereas the two LS-Shp2s did not exhibit detectable PTP activity ( Figure 4B ) . These results are consistent with catalytic activity data of mammalian NS and LS Shp2 mutants [10 , 14] . To determine how mutant Shp2 affects the development of zebrafish , we injected synthetic RNA encoding NS-Shp2 or LS-Shp2 into embryos at the one-cell stage . We titrated the amount of RNA down to amounts that reproducibly induced specific phenotypes ( D61G , 150pg; T73I , 100pg; A462T , 75pg; and G465A , 50pg ) . These phenotypes were not observed in embryos injected with green fluorescent protein ( GFP ) RNA ( 300pg ) . Similar amounts of RNA encoding WT Shp2 ( 150–300 pg ) did not induce defects . Very high amounts of WT Shp2 RNA ( >800 pg ) induced phenotypes , similar to NS- and LS-Shp2 , albeit the phenotypes were not as severe ( unpublished data ) , indicating that NS- and LS-Shp2 had strong , dominant functions . Injection of NS- or LS-Shp2s resulted in significantly shorter embryos at 4 dpf when compared to noninjected or GFP-injected controls ( Figure 4C and 4D ) . Body axis extension was already reduced at 10 hpf as the angle between the most anterior and posterior tissues was significantly increased upon injection of each of the NS- and LS-Shp2s ( Figure 4E ) . Cell tracing experiments demonstrated that both extension ( Figure 4F ) and convergence ( Figure 4G ) were reduced significantly upon injection of mutant T73I NS-Shp2 . In situ hybridization with ntl and gsc markers on NS- and LS-injected embryos demonstrated that cell specification was not affected ( Figure S3 ) . These results demonstrate that expression of NS-Shp2 induced defective CE cell movements during gastrulation without affecting cell specification . Embryos injected with either NS- or LS-Shp2 RNA developed craniofacial abnormalities that were apparent at 4 dpf . Notably the eyes were set wider apart and anterior structures had not extended normally ( Figure 5A–5D ) . Alcian blue stainings of cartilaginous structures showed that structures , including Meckel's cartilage ( black asterisk ) and the ceratohyal ( red asterisk ) , resided more posteriorly than in WT controls . Failure of anterior structures to extend normally and wider spacing of the eyes in NS- or LS-Shp2 expressing zebrafish embryos was similar to the Shp2 knockdown zebrafish embryos ( Figure 2 ) . Moreover , the facial abnormalities in NS- and LS-Shp2 expressing zebrafish were reminiscent of the symptoms that are observed in NS/LS patients and the NS mouse model . There is no evidence to suggest that this phenotype was caused by defective gastrulation cell movements . Mutants with disrupted gastrulation such as wnt5 [32] and knypek [33] develop similar craniofacial anomalies . Rescue of the knypek mutant by RNA injection led to rescue of the gastrulation defects , but not of the craniofacial defects , indicating that these defects are independent [33] . Injection of NS- or LS-Shp2 RNA caused defects in heart development . Similar defects were observed upon injection of NS- or LS-Shp2 , and the defects varied in penetrance from mild to grossly edematous , as illustrated for NS-Shp2 ( Figure 5E–5H ) . In situ hybridization using the heart-specific probe cmlc2 at 24 hpf demonstrates that the heart of NS/LS-injected embryos failed to jog to the left in approximately 30% of the NS- or LS-Shp2–injected embryos ( Figure 5I–5K ) . Homozygous NS mutant mice develop a grossly edematous heart [11] , similar to NS and LS zebrafish , indicating that the injected zebrafish phenocopy the symptoms observed in human patients and in gene-targeted mice . The CE cell movement defects that we observed are most likely resulting from defective directional cell movements , which in turn result from impaired cell polarization . We hypothesize that the craniofacial and cardiac defects we observed may also result from defective cell movements of neural crest cells shaping the face and myocardial cells shaping the heart , respectively . Future work should focus on the role of Shp2 in polarization and migration of these cells . Finally , we investigated how activating and inactivating mutations in Shp2 might produce similar phenotypes in early zebrafish development . To this end , we coinjected suboptimal amounts of combinations of either NS-Shp2 , LS-Shp2 , or GFP , which did not induce phenotypes by themselves , and embryos were assessed at 4 dpf . Coinjection of the two NS-Shp2s ( D61G and T73I ) resulted in a significant increase in the observed phenotypes compared to coinjection of D61G or T73I with GFP ( Figure 6 ) . Likewise , coinjection of the two LS-Shp2s ( A462T and G465A ) induced an increase in phenotypes as compared to control A462T or G465A coinjections with GFP . In contrast , coinjection of suboptimal amounts of combinations of NS-Shp2 RNA with LS-Shp2 RNA did not lead to a significant increase in the number of affected embryos ( Figure 6 ) . These results demonstrate that combinations of NS-Shp2 and LS-Shp2 do not act synergistically , whereas suboptimal amounts of NS-Shp2 mutants or LS-Shp2 mutants cooperate to induce defects in zebrafish embryos . The observation that activation and inhibition of the same signaling pathway induces similar phenotypes is not unprecedented . Microinjection of RNA encoding Rok2 or Galpha12/13 induces similar gastrulation defects as knockdown of Rok2 or Galpha12/13 , respectively [21 , 22] . An activity window exists for these factors . If overall activity falls outside of this window , the resulting phenotypes are very similar . We established that coinjection of two different NS-Shp2s or LS-Shp2s led to an increase in the number of affected embryos . Our results are consistent with NS-Shp2 and LS-Shp2 acting in the same pathway , with one activating and the other inhibiting signaling . We show here that knockdown of Shp2 induced CE cell movement defects , but did not affect cell specification . These defects were rescued by active Fyn and Yes and active RhoA , indicating that SFK-Rho signaling , rather than MAPK signaling was required for this function of Shp2 . Expression of NS- or LS-mutant Shp2 RNA in zebrafish embryos led to overlapping phenotypes , characterized by craniofacial and cardiac defects , reminiscent of the symptoms observed in NS/LS patients . The notion that defective Shp2 signaling induced cell movement defects as early as gastrulation is important and may have implications for the monitoring and diagnosis of NS and LS .
Zebrafish were kept and the embryos were staged as described before [34] . In situ hybridizations were done essentially as described [35] using probes specific for bmp2b [36] , chd [37] , cmcl2 [38] , cyc [39] , gsc [40] , krox20 [41] , ntl [42] , pax2 [43] , six3 [44] , wnt5 [45] , wnt11 , fyn , and yes [18] . The mutations D61G , T73I , A462T and G465A were introduced into zebrafish Shp2 by site-directed mutagenesis and cloned into EcoR1/BamH1 sites of pBSK11 and verified by sequencing . Fusion proteins were expressed from pGEX-based bacterial vectors encoding GST fusion proteins of WT Shp2 and all four NS/LS mutated constructs . Fusion proteins were produced in bacteria and purified using standard procedures . Antisense MOs were designed to include the start ATG of the respective cDNAs and ordered from GeneTools ( Philomath ) : Shp2 , 5′-GGTGGAACCACCTTCGGGATGTCAT . The Fyn , Yes , and Wnt5 MOs were described previously [17] . 5′ capped sense RNAs were synthesized using the mMessage mMachine kit ( Ambion ) . The amount of RNA that was injected at the one-cell stage was optimized for each synthetic RNA . For the rescue experiments , we used mutant , constitutively active Fyn and Yes , derived by mutagenesis of their C-terminal inhibitory tyrosine phosphorylation sites to phenylalanine . Wnt5 and active RhoAV12 were described previously [17] . The NS- and LS-Shp2 RNAs contained their respective mutations , D61G , T73I , A462T , or G465A . Phenotypes were assessed at the indicated stages . Embryos were ( co- ) injected at the one-cell stage with 0 . 25% 4 , 5-dimethoxy-2-nitrobenzyl ( DMNB ) -caged fluorescein dextran ( 10 , 000 MW , Molecular Probes ) . Uncaging was done as described [17] at shield stage ( 6 hpf ) using an Axioplan microscope , equipped with a UV light source , adjustable pinhole , and 40× objective . Pictures were taken immediately following uncaging , at 80% epiboly ( 8 hpf ) and tailbud stage ( 10 hpf ) . The angles for dorsal convergence and anterior extension were determined using NIH imaging software . Purified GST-fusion proteins were directly incubated in PTP assay buffer ( 20 mM MES buffer pH 6 . 0 , 1 mM EDTA , 150 mM NaCl , 1 mM dithiotreitol , and 10 mM p-nitrophenylphosphate ) for 45 min at 30 °C . The reactions were quenched with 0 . 4 M NaOH , and optical density was measured with a spectrophotometer at 415 nm ( wavelength ) .
The National Center for Biotechnology Information ( NCBI ) Entrez database ( http://www . ncbi . nlm . nih . gov/sites/gquery ? itool=toolbar ) accession numbers of the genes and proteins used in this study are: Fyn ( Danio rerio ) , EMBL AJ620748; RhoA ( Homo sapiens ) , swissprot P61586; Shp2 ( Danio rerio ) , NCBI NM_199846; Shp2 ( Homo sapiens ) , swissprot Q06124; Wnt5a ( Mus musculus ) , swissprot P22725; and Yes ( Danio rerio ) , EMBL AJ620749 . Identifiers of the human Noonan and LEOPARD syndromes are OMIM 163950 and OMIM 151100 , respectively . | Shp2 is a protein-tyrosine phosphatase and mutations in Shp2 cause the related Noonan and LEOPARD syndromes in humans . We used the zebrafish to investigate the cell biological role of Shp2 in early development . Shp2 knockdown and expression of mutant Shp2 that contained mutations corresponding to those found in human Noonan and LEOPARD patients , induced similar convergence and extension cell movement defects during gastrulation without affecting cell specification . Active Src family kinases and active RhoA rescued the Shp2 knockdown , indicating that signaling downstream of Shp2 was mediated by Src family kinases and RhoA . Expression of the Noonan and LEOPARD Shp2s in zebrafish induced craniofacial and cardiac defects that were reminiscent of the symptoms observed in human patients . Coinjections demonstrated that Noonan and LEOPARD Shp2s did not cooperate , which is consistent with the two mutants acting in the same signaling pathway with opposing effects . The finding that defective Shp2 signaling induced cell movement defects as early as gastrulation may have important implications for the monitoring and diagnosis of Noonan and LEOPARD syndromes in humans . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"developmental",
"biology",
"cell",
"biology",
"danio",
"(zebrafish)",
"in",
"vitro",
"homo",
"(human)",
"genetics",
"and",
"genomics"
] | 2007 | Shp2 Knockdown and Noonan/LEOPARD Mutant Shp2–Induced Gastrulation Defects |
In animals , circadian rhythms in physiology and behavior result from coherent rhythmic interactions between clocks in the brain and those throughout the body . Despite the many tissue specific clocks , most understanding of the molecular core clock mechanism comes from studies of the suprachiasmatic nuclei ( SCN ) of the hypothalamus and a few other cell types . Here we report establishment and genetic characterization of three cell-autonomous mouse clock models: 3T3 fibroblasts , 3T3-L1 adipocytes , and MMH-D3 hepatocytes . Each model is genetically tractable and has an integrated luciferase reporter that allows for longitudinal luminescence recording of rhythmic clock gene expression using an inexpensive off-the-shelf microplate reader . To test these cellular models , we generated a library of short hairpin RNAs ( shRNAs ) against a panel of known clock genes and evaluated their impact on circadian rhythms . Knockdown of Bmal1 , Clock , Cry1 , and Cry2 each resulted in similar phenotypes in all three models , consistent with previous studies . However , we observed cell type-specific knockdown phenotypes for the Period and Rev-Erb families of clock genes . In particular , Per1 and Per2 , which have strong behavioral effects in knockout mice , appear to play different roles in regulating period length and amplitude in these peripheral systems . Per3 , which has relatively modest behavioral effects in knockout mice , substantially affects period length in the three cellular models and in dissociated SCN neurons . In summary , this study establishes new cell-autonomous clock models that are of particular relevance to metabolism and suitable for screening for clock modifiers , and reveals previously under-appreciated cell type-specific functions of clock genes .
In mammals , many aspects of daily behavior and physiology such as the sleep-wake cycle , body temperature , and liver metabolism are regulated by endogenous circadian clocks [1] , [2] . The circadian time-keeping system is a hierarchical , multi-oscillator network , with the hypothalamic suprachiasmatic nucleus ( SCN ) acting as a central pacemaker at the top of the hierarchy . The SCN integrates external time cues and , through complex signaling cascades , synchronizes and coordinates extra-SCN oscillators in the brain and in peripheral clocks throughout the body , culminating in overt , coherent circadian rhythms at the organismal level [3] , [4] . This time-keeping system is critical for normal physiology and behavior , and its disruption can lead to sleep disorders , metabolic syndrome , premature aging , and cancer ( reviewed in [2] , [5] ) . Virtually all cells in our body have circadian oscillators [6]–[8] . Despite tissue-specific physiological differences , these oscillators share a highly conserved molecular mechanism – a negative feedback loop . This consists of transcriptional activators BMAL1 and CLOCK , which bind to E-box enhancers and activate the transcription of the Per and Cry families of repressors . These repressors then feed back to inhibit BMAL1/CLOCK activity and their own expression [9] . Each molecular component in the core clock loop is represented by multiple paralogs ( Bmal1 , Bmal2; Clock , Npas2; Per1 , Per2 , Per3; Cry1 , Cry2 ) , which provides the potential for functional redundancy and cell type specificity . In addition , post-translational modifications play critical roles in clock function . For example , the ubiquitin ligases FBXL3 and FBXL21 regulate period length and amplitude through ubiquitin-mediated degradation of CRY proteins and regulation of REV-ERBα activity [10]–[15] . This core clock loop integrates with other transcriptional systems such as the ROR/REV-ERB ( via RORE ) and DBP/E4BP4 ( via D-box ) accessory loops [16] . In the RORE loop , retinoic acid receptor-related orphan nuclear receptors ( RORA , RORB , and RORC ) act as activators , and REV-ERBs ( REV-ERBα known as NR1D1 and REV-ERBβ known as NR1D2; referred to hereafter as NR1D1 and NR1D2 ) act as repressors to regulate rhythmic Bmal1 expression via the RORE cis-element in the Bmal1 promoter [17]–[19] . Similarly , DBP/TEF/HLF and E4BP4 serve as activators and repressors , respectively , to regulate D-box-mediated transcription of genes such as Per3 [16] , [19] . These interlocking loops mediated by E-box , RORE , and D-box cis-elements form a complex clock network . These loops act individually or in combination to give rise to distinct waves of gene transcription [16] , [20] . For example , while Nr1d1 , Bmal1 , and Per3 transcription are each mediated primarily by a single cis-element ( i . e . , primarily E-box , RORE , and D-box , respectively ) , many other clock genes ( e . g . , Cry1 ) are regulated via a combinatorial mechanism involving multiple circadian elements [21] . Cell-based models were instrumental in the identification and characterization of clock gene function in mammals [22] , [23] . These studies relied on immortalized cell lines that display circadian rhythms of gene expression in a cell-autonomous manner ( i . e . , without systemic cues ) . We and others have used fibroblasts derived from clock component mutant mice expressing a clock gene reporter [17] , [21] , [24] . Cellular clock models for comprehensive genetic analysis , however , have so far been limited to 3T3 mouse fibroblasts and U2OS human osteosarcoma cells [22] , [25] , [26] . In the U2OS model , knockdowns of all clock components have been evaluated for impact on period length and amplitude [27] . In mice , 3T3 fibroblasts [22] , [28] and more recently MMH-D3 hepatocytes [29] have been introduced as cellular clock models; however , unlike the U2OS model , these models haven't been fully characterized genetically . An implicit assumption in all these studies is that the clock works the same way in all cell and tissue types , such that gene function determined in one cell or tissue type applies to all cells , regardless of local physiological inputs to the clock . However , while 3T3 cells may be an appropriate model of the fibroblast clock , it is likely not an appropriate model for other cells . Recent studies point to bidirectional interactions between circadian clocks and other cellular and physiological processes . Thus , the circadian system is integrated with , and influenced by , the local physiology . Of particular interest is the reciprocal interaction between clock function and metabolism [5] , [30]–[33] . However , as yet , there aren't any characterized cellular models appropriate for the study of clock control of metabolism . Therefore , to reveal cell type-specific molecular , cellular , and physiological mechanisms of circadian clocks , new cell-autonomous , physiologically relevant peripheral clock models are needed . To facilitate cell type-specific genetic characterization , we explored mouse cell lines relevant to major metabolic functions , focusing on 3T3-L1 adipocytes and MMH-D3 hepatocytes . The mouse 3T3-L1 adipocyte cell line reflects adipose tissue function and has been pivotal in advancing the understanding of basic cellular mechanisms associated with diabetes , obesity , and related disorders in thousands of studies ( e . g . , [34] , [35] ) . In recent years , mouse MMH-D3 hepatocytes have become a prominent model reflecting hepatocyte function in the liver [36] , [37] . Both cell lines have been shown to exhibit rhythmic expression of clock genes and other genes that are involved in and modulated by local physiology [29] , [38] , [39] . In this study , we used luciferase reporters of clock gene expression and established three high amplitude cell-autonomous clock models: 3T3 fibroblasts , 3T3-L1 adipocytes , and MMH-D3 hepatocytes , with 3T3 fibroblasts as a reference model . These reporter cells displayed persistent , high amplitude rhythms , which allowed longitudinal recording of clock gene rhythms with high temporal resolution using an inexpensive off-the-shelf microplate reader . For genetic perturbations , we developed a pipeline to produce high-quality lentiviral shRNAs to knock down any gene of interest , and validated these cellular models with shRNAs against a selected panel of known clock genes . We show that knockdown of many clock genes resulted in expected phenotypes in all tested cell lines . Unexpectedly , however , we also observed cell type-specific knockdown phenotypes , particularly within the Per gene family . This study has important implications for the tissue-specific mechanisms of circadian clocks .
As an initial effort to develop new cellular clock models pertinent to metabolism , we screened cell lines for robust rhythms and chose 3T3-L1 adipocytes and MMH-D3 hepatocytes . We introduced a lentiviral reporter harboring the rapidly degradable firefly luciferase ( dLuc ) gene under the control of either mouse Per2 or Bmal1 gene promoters into cells [23] . Whereas the 3T3 reporter cells were directly used in bioluminescence recording , 3T3-L1 and MMH-D3 cells were first differentiated into mature adipocytes and hepatocytes , respectively , prior to recording . These cells displayed persistent bioluminescence rhythms in 35 mm culture dishes monitored in a LumiCycle luminometer ( Figure 1A ) . In each cell line , Per2-dLuc and Bmal1-dLuc reporters displayed anti-phasic rhythms of bioluminescence , consistent with the function of E-box- and RORE-containing promoters in regulating distinct and opposite phases of gene expression . Next , we adapted the LumiCycle reporter assay to high-throughput screening ( HTS ) formats on 96 well plates . For this , we performed single cell cloning and selected clonal cell lines that expressed high levels of bioluminescence . These reporter lines displayed persistent rhythms under optimized growth conditions when monitored on a microplate reader ( Synergy 2 SL ) with highly consistent period lengths ( Figure 1B ) . These highly reproducible rhythms seen in 96 well plates were similar to those in the LumiCycle , a lower throughput but much more expensive recorder . Therefore , these lines represent a tangible advantage to many labs interested in exploring circadian biology in these metabolically relevant cell lines . For genetic perturbations , we developed a pipeline to produce high-quality , validated lentiviral shRNA vectors to knock down any mouse gene . We chose lentiviral shRNAs over transfected siRNAs because lentivirus-mediated delivery mediates potent transduction and stable integration in both dividing and non-dividing cells of various types in vitro and in vivo , thus circumventing low transfection efficiency for certain cells . We designed 6 target oligonucleotide sequences for each gene and cloned the shRNA expression cassette into the lentiviral pLL3 . 7GW Gateway vector , in which shRNA expression is under the control of the mouse U6 promoter , as we reported previously ( Figure S1 ) [17] . Infectious lentiviral particles were produced in 293T cells using standard procedures and used to infect reporter cells . Infection efficiency was estimated by observing GFP co-expressed from a separate expression cassette under control of the CMV promoter ( Figure S1 ) . We used this pipeline to generate a panel of shRNA constructs targeting the following selected 13 clock genes: Bmal1 , Bmal2 , Clock , Npas2 ( core loop activators ) ; Per1 , Per2 , Per3 , Cry1 , Cry2 ( core loop repressors ) ; Fbxl3 ( core loop post-translational modifier ) ; Nr1d1 , Nr1d2 ( RORE repressors ) ; and E4bp4 ( D-box repressor ) . Because of the more prominent roles of repressors in clock function , we chose to examine Nr1d1 , Nr1d2 , and E4bp4 , the RORE and D-box negative factors , rather than the corresponding activators [17] , [40] , [41] . We tested shRNA knockdown ( KD ) efficiency for the 13 clock genes . Co-transfection of shRNA with Flag-tagged cDNA in 293T cells followed by Western blot analysis showed efficient KD of each gene at the protein level by at least two shRNAs ( Figure S2 ) . To check the KD efficiency for endogenous gene expression , mRNA levels of targeted clock genes were also measured using qPCR . For each gene , at least two shRNAs were effective in knocking down gene expression , a requirement to filter out off-target effects of shRNAs [42] . The versatility and efficiency of lentiviral shRNA allowed us to study all 13 known clock genes in all three cell type-specific clock models in parallel , which allows direct phenotypic comparison . Knockdown of Bmal1 , Clock , Cry1 , Cry2 , and Fbxl3 in all three cellular models resulted in expected phenotypes similar to those in LumiCycle assays using 35 mm dishes and consistent with previous knockout and knockdown studies using human and mouse cellular models [17] , [27] , [31] , [43]–[45] . Specifically , KD of Bmal1 or Clock results in rapid damping or arrhythmicity ( Figure 2A and Tables 1 , S1 , S2 , S3 ) ; Cry1 KD leads to low amplitude or rapid damping depending on KD efficiency , whereas Cry2 KD lengthens period and increases rhythm amplitude ( Figure 2B ) . The phenotypic defects correlate with KD efficiency of the endogenous genes by the individual shRNAs as determined by qPCR analysis . Taken together , our data demonstrate that Bmal1 , Clock , Cry1 , and Cry2 play similar roles in the clock mechanism across tested cell types , which provides validation for the three cellular models . Knockdown of Bmal2 and Npas2 did not exhibit any obvious circadian phenotypes in any of the three cellular models ( Figure S3 ) , even though their expression was knocked down to levels similar to those for Bmal1 or Clock ( Figure 2A and Table 1 ) . These results are consistent with absence of observable circadian phenotypes of liver and lung tissues from Npas2−/− mice [46] , though these mice do have deficits in circadian behavior and sleep homeostasis [47] . Despite potential functional redundancy between Bmal1 and Bmal2 [45] , [48] , and between Clock and Npas2 [46] , our data suggest that Bmal2 and Npas2 are not necessary for the clock to operate in these cells . We show that Fbxl3 KD caused long period and low amplitude in 3T3 , long period in 3T3-L1 , and low amplitude and rapid damping in MMH-D3 cells ( Figure 2C ) . This cell-autonomous phenotype is much more extreme compared to the relatively modest period-lengthening phenotypes seen at the SCN tissue and behavioral levels [10]–[12] , or in human U2OS cells [27] , [44] . Notably , although KD of Fbxl3 or Cry2 both produced long periods , Cry2 down-regulation increased rhythm amplitude , whereas Fbxl3 silencing resulted in low amplitude , consistent with its dual role in ubiquitin-mediated degradation of CRY proteins and in regulation of NR1D1-mediated transcriptional suppression [10]–[13] . Nr1d1 and Nr1d2 play overlapping but essential functions in regulating RORE-mediated transcription , and knockdown of either gene results in low amplitude rhythms , and in some cases , short period [17] , [27] . We examined the effects of Nr1d1 or Nr1d2 KD on clock function in our cellular clock models . Knockdown of Nr1d1 resulted in largely normal rhythms in these cells ( Figure 3A and Table 1 ) , indicating potential overlapping functions of Nr1d2 [17] . This is different from the period lengthening produced by siRNA knockdown in U2OS cells [27] , [31] or the period shortening and greater variability seen in behavioral rhythms of Nr1d1−/− mice [18] , [49] . Nr1d2 KD , on the other hand , resulted in period shortening in MMH-D3 and 3T3-L1 cells but low amplitude in 3T3 cells ( Figure 3A and Table 1 ) . The Nr1d2 impact on overall rhythms was at least similar to , if not substantially stronger than , for Nr1d1 . This is consistent with the reported redundant functions of Nr1d1 and Nr1d2 [17] , but contrasts with previous studies showing that NR1D2 deficiency has no observable rhythm phenotype in U2OS cells [27] or in mice at the behavioral level [18] . Thus , Nr1d1 and Nr1d2 play different roles in clock function depending on tissue or cell type , and Nr1d2 may be more important than previously recognized . Despite the widely accepted role of E4BP4 as the repressor of D-box-mediated transcription , definitive genetic evidence of clock function has been lacking . We show here that E4bp4 KD resulted in short period and rapid damping in 3T3 fibroblasts , low amplitude in 3T3-L1 cells , and short period in MMH-D3 cells ( Figure 3B and Table 1 ) . These data are in line with recent studies suggesting a prominent role of E4bp4 in regulating the phase of Cry1 transcription in mouse embryonic fibroblasts [21] and period length in Rat-1 fibroblasts [41] . Studies of E4BP4's function in the clock mechanism using E4bp4−/− mice are therefore needed to validate our findings in cellular clock models . The shRNA constructs against Per1 , Per2 , and Per3 down-regulated mRNA and protein expression ( Figures S2 and 4 ) . However , unlike Cry1 and Cry2 KDs , knockdown of the Per genes in our clock models resulted in cell type-specific clock phenotypes ( Figure 4 ) . First , compared to the dramatic circadian defects observed in peripheral tissue explants and fibroblasts of Per1−/− mice , or upon siRNA-mediated Per1 KD in U2OS cells [31] , [43] , [50] , Per1 KD in our clock models had milder effects on clock function . Interestingly , these less dramatic phenotypes are cell type-specific: WT phenotype in 3T3 and 3T3-L1 cells , but significantly shorter period and low amplitude in MMH-D3 cells ( Figures 4A and 4D; Table 1 ) . Similarly , Per2 KD did not cause arrhythmicity in any of the three cellular clock models: only a modest reduction of amplitude in 3T3-L1 cells , but significantly shorter period in 3T3 and short period and low amplitude in MMH-D3 cells ( Figures 4B and 4D; Table 1 ) . The amplitude reduction in MMH-D3 cells was evident in both the subtracted data and raw data ( compare Figure 4 with Figure S5 ) . Although it is possible that the more modest phenotypes may be due to incomplete silencing , the knockdown levels were comparable to those of Bmal1 , Clock , and Fbxl3 ( Figures 2 and 4B ) ; and as in the case of Per1 KD , the different phenotypes resulted from similar Per2 KD efficiency in different cells . This is unexpected given its essential role in circadian rhythms of mice at the behavioral level and in cultured fibroblasts and U2OS cells [27] , [43] , [51]–[53] . Interestingly , however , our finding of non-essential but cell type-specific role of Per2 is in line with a recent report showing that Per2−/− SCN explants displayed persistent rhythms with short periods , whereas Per2−/− pituitary explant rhythms were normal and lung explants displayed slightly long periods [54] . Thus , even though genetic knockout and knockdown ( incomplete silencing ) may cause variations in phenotypes , the loss-of-function phenotypes of Per1 and Per2 are largely consistent and are cell or tissue type specific . While Per2 appears to be more important than Per1 for normal clock function in mice [52] , [55] , deletion of Per3 has only subtle effects on the SCN clock and is often not considered part of the core clock mechanism [43] , [54] , [56] . However , we show here that KD of Per3 in all three models produced significantly shorter periods than in control cells ( Figures 4C and 4D; Table 1 ) . These results are consistent with recent reports showing that tissue explants of Per3−/− mice , including liver , lung , and pituitary , also displayed short periods [43] , [54] . In addition , knockdowns of Per1 , Per2 , and Per3 showed similar phenotypes in cells expressing a different reporter ( Figures 4 and S4 ) , confirming that the knockdown effects are reporter independent . Furthermore , the Per1 and Per3 knockdown effects are largely consistent with data of liver explants from Per1 and Per3 knockouts [43] , [57] , [58] . The prominent role of Per3 in peripheral clock function led us to examine its function more carefully in both intact SCN explants and dissociated SCN neurons derived from Per3−/−:mPer2Luc mice . We detected persistent mPer2Luc rhythms in Per3−/− SCN explants with a slightly shorter period than WT ( Figure 4E ) . This is consistent with the original study of Per3−/− mice , where a slightly shorter period of behavioral rhythms was reported [56] . We then dissociated Per3−/− SCN neurons and examined mPer2Luc bioluminescence from dispersed neurons at the single-cell level , as we have done previously for other genotypes [43] . We found that dissociated WT and Per3−/− neurons generally exhibited persistent rhythms with high amplitude . However , the mean period of rhythms in Per3−/− neurons was substantially shorter than in WT cells ( Figure 4E ) . The weaker circadian defect at the SCN tissue level than in cell-autonomous preparations ( either our cellular models or dissociated SCN neurons ) is consistent with the principle that the SCN network confers robustness against genetic perturbations [43] , [59] , such that Per3 plays a less prominent role in the intact SCN due to compensation by the SCN network . Based on these results , we conclude that Per3 plays an important role in the SCN cellular clock as well as in peripheral oscillators and thus represents a bona fide clock component . The more prominent role of Per3 in peripheral oscillators is in line with several recent studies showing that disruption of Per3 resulted in internal phase misalignment or desynchrony , or aberrant metabolic and sleep phenotypes [58] , [60]–[62] , all pointing to the role of Per3 in coherence of circadian organization . Thus , our findings suggest that tissue-specific function or dysfunction of clock genes in peripheral tissues can be an important contributing factor to human diseases , even when the behavioral effect of gene knockout is subtle . In this context , it is interesting to note that human polymorphisms in PER3 are associated with sleep and metabolic disorders [63] , [64] . Taken together , our study expands our knowledge of the distinct functions of known clock genes across tissues . In particular , Per3 plays a more important role in both SCN and non-SCN cells than previously appreciated , and Per1 and Per2 appear to have different roles in different cell types . Results from this study are broadly consistent with previous findings from loss-of-function studies and collectively point to the previously under-appreciated cell type specificity of Per gene function in circadian physiology ( Figure S6 ) . Compared to other cell-autonomous models , the short period length in MMH-D3 cells after knockdown of each of the Per genes is unique ( Figure S6 and Table 1 ) , and therefore we sought to perform single and composite knockdowns for further phenotyping using the LumiCycle assay . Consistent with the Synergy assay , Per1 , Per2 , and Per3 single gene KD each gave short period phenotypes , about 2 hrs shorter than the control cells ( Figures 5A and 5B; Table S4 ) . Composite Per1/Per2 double KD and Per1/Per2/Per3 triple KD caused complete arrhythmicity ( Figure 5A ) , indicating the prominent roles of Per1 and Per2 in the hepatocyte clock . Interestingly , Per1/Per3 and Per2/Per3 double KDs did not cause any further period shortening over single Per gene KDs ( Figures 5A and 5B ) . As an initial effort to probe the network features of the hepatocyte clock , we examined effects of Per KD on the expression of endogenous clock genes by qPCR . Per1 , Per2 , and Per3 were each knocked down in both single and composite KDs ( Figure 5C ) . Compared to the U2OS model , paralog compensation among the Per genes in MMH-D3 cells is more pervasive . In our MMH-D3 model , both Per1 KD and Per2 KD upregulated Per3 , and Per3 KD upregulated Per2 . While composite Per1/Per2 KD upregulated Per3 and Per1/Per3 KD greatly upregulated Per2 , Per2/Per3 KD did not increase Per1 expression . Interestingly , Per1 and/or Per2 KD had milder effects on the expression of E-box-containing genes ( e . g . , Nr1d1 and Nr1d2 ) than on RORE-containing genes such as Bmal1 , E4bp4 , and Cry1 ( Figure 5C ) , in line with the notion that the PERs can directly and indirectly affect Bmal1 transcription [65] , [66] . Overall , the network interactions in MMH-D3 cells appear to differ from those in the U2OS model in which PER1 plays a more dominant role than PER2 and PER3 [27] , and is expected to differ from those in 3T3 and 3T3-L1 models . Thus , extensive investigation into the network features of these cellular models will require additional experiments and is warranted in future studies , as we have done with the U2OS model [27] . Cell type-specific function of clock genes may result from their differential tissue expression and activity ( i . e . , expression levels , ratio of repressors to activators , rhythmicity , and relative amplitudes ) , compensatory mechanisms , alternative splice variants , and post-translational modifications ( PMTs ) , all of which can be rendered cell type specific by local physiology . Recent studies have suggested a role for stoichiometric balance among clock proteins in circadian clock robustness and periodicity , and call for mechanistic studies in a tissue specific manner [67]–[69] . In the context of the liver and adipose tissue function , it is plausible that the basic core clock mechanism incorporates cell type-specific factors and forms distinctive functional networks to regulate ( and in turn be regulated by ) different local physiologies . It is interesting to note , as PTMs of clock factors ( e . g . , phosphorylation , ubiquitination , ADP-ribosylation , acetylation , and O-GlcNAcylation ) represent critical regulatory mechanisms [33] , [70] , tissue-specific cellular functions and metabolic states that affect the PTMs would provide important inputs to adjust local circadian clocks , and vice versa . Thus , cell type specific clock gene function starts to make sense when local physiology is considered as inputs to the clock . This challenge surely provides an opportunity for deeper insights into mechanism of tissue specific clocks , akin to the recent realization that cyclin-dependent kinase networks in the cell cycle control program are tissue specific [71] . In summary , we established three new mouse cellular clock models: fibroblasts , adipocytes , and hepatocytes . These cellular clock models offer experimental tractability , efficiency , and versatility , which are more difficult or impossible to apply to traditional tissue or animal models . Of note , in contrast to previous cellular clock models , the new clock models are amenable to high throughput experiments with inexpensive off-the-shelf recording systems , making these lines especially suitable for screening small molecules or genomic entities for impacts on cell autonomous clocks relevant to metabolism . We validated these models by developing and testing an shRNA panel of selected known clock genes . Results from this study and others point to the previously under-appreciated cell type specificity of clock gene function in circadian physiology ( Figure S6 ) . The prevalence of tissue-specific clock gene function will have important implications for future studies of clock factors that affect local clock function . It is our hope that our findings in this study , along with the new cellular clock models , approaches , and tools developed here , can be applied to a greater variety of cell types in future studies , to reveal the full range of tissue-specific clock properties underlying local circadian biology .
Per3−/− mice were obtained from David Weaver at the University of Massachusetts . Knockout mice were bred with mPer2Luc reporter mice to obtain homozygous knockouts harboring the mPer2Luc reporter . Wheel-running assays were performed and analyzed as described previously [8] . All animal studies were conducted in accordance with the regulations of the Committees on Animal Care and Use at University of Memphis and UCSD . All cell culture media were from HyClone . 3T3 ( also known as NIH 3T3 ) and 3T3-L1 cells were cultured in regular medium in which DMEM was supplemented with 10% FBS and 1× penicillin-streptomycin-glutamine ( PSG ) . For 3T3-L1 differentiation , pre-adipocytes were first grown to confluence ( Day 0 ) . On Day 2 , cells were fed with induction medium ( regular medium with 1 µM dexamethasone , 0 . 5 mM isobutylmethylxanthine , and 2 µg/ml insulin ) . On Day 4 , cells were changed to maintenance medium ( regular medium containing 2 µg/ml insulin ) . From day 6 onward , cells were grown in regular medium until use . For bioluminescence recording , 3T3 and 3T3-L1 cells were grown in 25 mM HEPES-buffered regular medium ( pH 7 . 4 ) containing 1 nM forskolin and 1 mM luciferin . Fully differentiated 3T3-L1 cells were used in all experiments . MMH-D3 cells were grown in regular medium in which RPMI medium was supplemented with 10% FBS , 1× PSG , 10 µg/ml insulin , 55 ng/ml epidermal growth factor ( EGF ) , and 16 ng/ml insulin like growth factor-II ( IGF-II ) . For differentiation , pre-hepatocytes were first grown to 100% confluence . Two days later , cells were replaced with differentiation medium ( regular medium with 2% DMSO ) . Medium change was repeated every 48 hours for 6–8 days for cells to be fully differentiated for use . Circadian rhythms of differentiated cells were synchronized with 200 nM dexamethasone followed by bioluminescence recording in 25 mM HEPES-buffered serum-free explant medium ( pH 7 . 4 ) containing B-27 and 1 mM luciferin , as we have done previously [43] . Fully differentiated MMH-D3 cells were used in all experiments . SCN explants and dissociated neuronal cells were prepared and cultured as previously described [43] . Bioluminescence recording of explants , single cell-imaging of individual SCN neurons , and respective data analysis were performed as previously described [43] . Lentiviral luciferase reporters of the Per2 or Bmal1 promoter were described previously [17] , [43] . Reporter cells and clonal lines were generated as previously described [23] . Briefly , reporter viral particles of high titer ( >108 viral particles/ml ) were obtained by ultracentrifugation and used to infect 3T3 , 3T3-L1 , and MMH-D3 cells . Clonal cell lines of homogenous cell populations were obtained by single cell sorting and cloning in 96 well plates . We then selected the clones that expressed high levels of luciferase and exhibited circadian properties comparable to infected parental cell populations . These brighter cells were used in high-throughput assays on 96 well plates . Knockdown phenotypes were confirmed to be independent of the reporter , either Per2-dLuc or Bmal1-dLuc , and thus phenotypic differences across cell types are unlikely due to different reporter insertion sites . We used an optimized shRNA design algorithm adapted from Aza-Blanc et al . [72] for target sequence prediction . This adapted algorithm selects for optimal target sequence for knockdown , and against homologous sequences to minimize off-target effects . We selected 6 target sequences for each gene as listed in Table S5 . Each shRNA construct contained a sense and an antisense target sequence of 19 nucleotides ( nts ) in length , separated by 9 nts for a hairpin loop , and flanked by TTTG at 5′ and GATC at 3′ ends for cohesive end cloning . All oligos ( 55 nts ) were synthesized by Integrated DNA Technologies ( IDT ) . The annealed oligonucleotides were first cloned into the BbsI and SpeI sites of a pGWL-si2/U6 vector , in which the shRNA expression cassette is driven by an RNA polymerase III- based mouse U6 promoter . Subsequently , the U6-shRNA cassette was cloned into the lentiviral pLL3 . 7GW vector ( modified from pLL3 . 7 ) [17] , [73] in a Gateway LR Clonase reaction ( Life Technologies ) , according to manufacturer's instructions . Viral particles were prepared using standard methods in 293T cells on 12-well plates as previously described [23] , [74] . Culture medium containing viral particles ( ∼106 viral particles/ml ) were collected and used for subsequent infection of reporter cells . Transfection and infection efficiency were estimated by observing GFP co-expressed from a CMV promoter . To produce high titer viruses , crude viral particles were concentrated through ultracentrifugation and appropriate titers were used , as described previously [74] . This pipeline allowed us to generate a panel of shRNA constructs against all known clock factors for genetic perturbation and phenotyping . We used a LumiCycle luminometer ( version 2 . 31 , Actimetrics ) for bioluminescence recording of cells grown on 35 mm culture dishes , as described elsewhere [17] , [23] , [43] . The LumiCycle Analysis program version 2 . 53 ( Actimetrics ) was used to determine circadian parameters . Briefly , raw data were fitted to a linear baseline , and the baseline-subtracted data were fitted to a sine wave ( damped ) , from which period length and goodness of fit and damping constant were determined . For samples that showed persistent rhythms , goodness-of-fit of >80% was usually achieved . Due to high transient luminescence upon medium change , the first cycle was usually excluded from rhythm analysis . Damping rate = 1/damping constant . For amplitude analysis , raw data from day 3 to day 5 were fitted to a linear baseline , and the baseline-subtracted ( polynomial number = 1 ) data were fitted to a sine wave , from which the amplitude was determined . We used a Synergy 2 SL microplate reader ( Bio Tek ) for bioluminescence recording of cells grown on 96 well plates , as previously described [23] . Synergy data were analyzed with the MultiCycle Analysis program ( Actimetrics ) , in which bioluminescence data were baseline-subtracted and fit to a damped sine wave to determine period length , goodness of fit , and amplitude , as with LumiCycle Analysis . Due to the various reporter expression levels and for direct comparison of different rhythms , baseline subtracted data were plotted . Because there is no damping rate output function in the MultiCycle Analysis , we used a curve fitting program of “CellulaRhythm” to determine damping rate from Synergy data as previously developed [75] . For testing shRNA knockdown efficiency targeting each clock gene , the cDNA was cloned into a p3XFlag-CMV-14 vector . Flag-tagged cDNA was co-transfected with the indicated shRNA in 3T3 or 293T cells . Forty eight hours post-transfection , cells were lysed in RIPA buffer containing complete protease ( Roche ) and phosphatase inhibitors ( Sigma ) . Protein expression was determined by Western blot analysis using an anti-Flag monoclonal antibody ( Sigma ) . For all Western assays , PVDF membrane was used in protein transfer , and SuperSignal West Pico substrate ( Thermo Scientific ) was used for chemiluminescent detection . For qPCR analysis , parallel infection experiments were performed as with bioluminescence recording . Cells were harvested prior to medium change and were therefore unsynchronized . Total RNAs were prepared using the RNeasy 96 kit ( Qiagen ) , as previously described [31] . Reverse transcription was performed using a high-capacity RNA to cDNA kit ( Applied Biosystems ) , and qPCR was performed using SYBR Green PCR master mix ( Thermo Scientific ) on an iCycler thermal cycler ( BioRad ) . The primers used in qPCR analysis are listed in Table S6 . Transcript levels for each gene were normalized to Gapdh and values were expressed as percentage of expression in NS control cells , as previously described [17] . | Various aspects of our daily rhythms in physiology and behavior such as the sleep-wake cycle are regulated by endogenous circadian clocks that are present in nearly every cell . It is generally accepted that these oscillators share a similar biochemical negative feedback mechanism , consisting of transcriptional activators and repressors . In this study , we developed cell-autonomous , metabolically relevant clock models in mouse hepatocytes and adipocytes . Each clock model has an integrated luciferase reporter that allows for kinetic luminescence recording with an inexpensive microplate reader and thus is feasible for most laboratories . These models are amenable to high throughput screening of small molecules or genomic entities for impacts on cell-autonomous clocks relevant to metabolism . We validated these new models by RNA interference via lentivirus-mediated knockdown of known clock genes . As expected , we found that many core clock components have similar functions across cell types . To our surprise , however , we also uncovered previously under-appreciated cell type-specific functions of core clock genes , particularly Per1 , Per2 , and Per3 . Because the circadian system is integrated with , and influenced by , the local physiology that is under its control , our studies provide important implications for future studies into cell type-specific mechanisms of various circadian systems . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology",
"gene",
"regulatory",
"networks",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"molecular",
"genetics",
"computational",
"biology",
"molecular",
"cell",
"biology",
"gene",
"function"
] | 2014 | Cell Type-Specific Functions of Period Genes Revealed by Novel Adipocyte and Hepatocyte Circadian Clock Models |
The most common cystic fibrosis ( CF ) causing mutation , deletion of phenylalanine 508 ( ΔF508 or Phe508del ) , results in functional expression defect of the CF transmembrane conductance regulator ( CFTR ) at the apical plasma membrane ( PM ) of secretory epithelia , which is attributed to the degradation of the misfolded channel at the endoplasmic reticulum ( ER ) . Deletion of phenylalanine 670 ( ΔF670 ) in the yeast oligomycin resistance 1 gene ( YOR1 , an ABC transporter ) of Saccharomyces cerevisiae phenocopies the ΔF508-CFTR folding and trafficking defects . Genome-wide phenotypic ( phenomic ) analysis of the Yor1-ΔF670 biogenesis identified several modifier genes of mRNA processing and translation , which conferred oligomycin resistance to yeast . Silencing of orthologues of these candidate genes enhanced the ΔF508-CFTR functional expression at the apical PM in human CF bronchial epithelia . Although knockdown of RPL12 , a component of the ribosomal stalk , attenuated the translational elongation rate , it increased the folding efficiency as well as the conformational stability of the ΔF508-CFTR , manifesting in 3-fold augmented PM density and function of the mutant . Combination of RPL12 knockdown with the corrector drug , VX-809 ( lumacaftor ) restored the mutant function to ~50% of the wild-type channel in primary CFTRΔF508/ΔF508 human bronchial epithelia . These results and the observation that silencing of other ribosomal stalk proteins partially rescue the loss-of-function phenotype of ΔF508-CFTR suggest that the ribosomal stalk modulates the folding efficiency of the mutant and is a potential therapeutic target for correction of the ΔF508-CFTR folding defect .
Cystic fibrosis ( CF ) , caused by mutations in cystic fibrosis transmembrane conductance regulator ( CFTR ) , is characterized by multiorgan pathology , mainly affecting the upper and lower airways , gastrointestinal tract , and endocrine system [1 , 2] . To date ~2 , 000 mutations have been identified in the CFTR gene with widely variable disease severity [3–5] . The gene product , CFTR , is an ATP-binding cassette ( ABC ) transporter , which functions as a cyclic AMP-regulated chloride and bicarbonate channel in secretory epithelia [2 , 6] . Deletion of the phenylalanine at position 508 ( Phe508del , designated as ΔF508 ) in the nucleotide binding domain 1 ( NBD1 ) , the most common CF-causing mutation , results in misfolding and premature degradation of the mutant via the endoplasmic reticulum ( ER ) -associated degradation pathway ( ERAD ) [7–9] . The small amount of ΔF508-CFTR molecules that escape the ER are functionally , conformationally , and biochemically unstable and are rapidly removed from the plasma membrane ( PM ) via the endolysosomal associated degradation pathway [10 , 11] . To rescue the folding defect of ΔF508-CFTR , several strategies have been pursued with limited success so far [12–14] . Small molecule correctors that act as pharmacological chaperones , like VX-809 , can directly bind to and promote the folding of ΔF508-CFTR [15–19] . In combination with the gating potentiator VX-770 , VX-809 achieved only modest benefit in CF patients homozygous for the ΔF508 mutation [20] , which might be attributed in part to the destabilization of ΔF508-CFTR upon chronic exposure to VX-770 [21 , 22] . Modifier genes may also facilitate the ΔF508-CFTR functional rescue by enhancing the mRNA or protein expression , folding , stability , or by inhibiting its degradation at the ER and post-ER compartments [11 , 12 , 23–25] . Candidate modifier genes have been isolated by genome-wide SNP studies [26 , 27] , identification of the CFTR interactome [28–30] , and phenotypic screens of targted siRNA libraries [11 , 31] . As a complementary approach , strategies focusing on reverting the maladaptive stress response in CF have been proposed [32] . None of these approaches , however , appear to attain sufficient functional correction in preclinical studies to be therapeutically robust in patients with the most common CF mutation , particularly in individuals carrying only one copy of ΔF508-CFTR , representing 40% of US CF patients [3] . Chimeras between the yeast ABC transporter STE6 and ΔF508-CFTR were used as a homology model to identify revertant mutations [33] , but these chimeras are not recognized by the ER quality control [34] . Recently , we have employed a genome-wide screen to identify modifiers of CFTR misfolding , utilizing high-throughput yeast phenomic analysis of Yor1 , a member of the ABC transporter superfamily , with deletion of phenylalanine 670 ( Yor1-ΔF670 ) . The ΔF508-equivalent mutation , Yor1-ΔF670 , results in protein misfolding , ER retention , and proteasomal degradation similar to that of ΔF508-CFTR in mammalian cells [35–37] . Oligomycin , which inhibits the ATP synthase , is extruded by Yor1 across the PM , enabling a screen of the yeast gene deletion strain library [37 , 38] for modulators of Yor1-ΔF670 processing as determined by oligomycin sensitivity . Our phenomic screen provided a comprehensive gene interaction network that can potentially modulate ΔF508-CFTR biogenesis [39] . Evolutionary conservation in the ΔF-biogenesis network was demonstrated by the identification of many yeast homologs of published human genes that modulate ΔF508-CFTR biogenesis similarly to that of Yor1-ΔF670 function [39] . Here , we validated a subset of genes that were identified by quantitative high throughput cell array phenotyping ( Q-HTCP ) in the yeast model , including components of the cytoplasmic exosome , rRNA biogenesis pathway , and most notably , the ribosomal stalk , in human respiratory cells . We show that silencing of ribosomal stalk proteins , in particular RPL12 , increases the rescued ΔF508-CFTR PM density , function , and thermal stability , suggesting that ribosomal stalk proteins have a determinant role in the folding of the mutant channel and may represent a possible therapeutic target for correction of the ΔF508-CFTR folding defect . Furthermore , the results highlight the capacity of yeast phenomic screen as a systematic approach to investigate disease modifier genes , adding CF to a list of human diseases where yeast can function as a powerful model system [40–42] .
To select hits from our previous yeast screen [39] for further study in human cells , the top 180 gene deletions that increased oligomycin resistance of Yor1-ΔF670-associated oligomycin sensitivity ( or deletion suppressors ) were prioritized by retesting and secondary screening with additional experimental controls ( described below ) . The hits were also subjected to analysis by the DAVID bioinformatics tool [43] . Functional annotation clustering identified two groups , ribosomal or ribosome-associated genes and genes involved in RNA degradation , with an enrichment score of > 2 ( S1A Fig ) . These analyses revealed several gene interaction modules ( e . g . multiple subunits of a protein complex ) and , when analyzed in parallel with the yor1-Δ0 ( null ) allele , verified that oligomycin resistance required interaction with Yor1-ΔF670 ( Fig 1 ) . Additionally , Yor1-ΔF670 interactions were confirmed by expressing the protein from a different promoter and without the C-terminal GFP fusion , that was used in our original screen . Double mutants were remade in quadruplicate using modified synthetic genetic array ( SGA ) methodology [44 , 45] , and the new panel of double mutants was subjected to Q-HTCP to obtain growth curves at multiple oligomycin concentrations [37] . Growth curves were fit to a logistic growth function , and gene interaction was measured with respect to the cell proliferation parameter , L ( Fig 1A ) , which corresponds to the time at which a culture reaches half of the carrying capacity [46] . Multiple components of the cytoplasmic exosome ( SKI2 , SKI3 , SKI7 ) ( Figs 1C and S1B ) , genes involved in ribosomal RNA processing and the nuclear exosome ( LRP1 , RRP6 , and RRP8 ) ( Figs 1D and S1B ) , and ribosomal structural proteins ( RPL12A , RPP2B , RPS7A , and RPL19A ) ( Figs 1B , 1E and S1B ) were identified based on increased resistance to oligomycin upon the gene deletion , in the context of Yor1-ΔF670 , but not yor1-Δ0 ( Fig 1 ) . On this basis , we focused on the cytoplasmic and nuclear exosome , ribosomal RNA biogenesis , and ribosomal proteins in order to test for the evolutionarily conserved influence of their knockdown on ΔF508-CFTR biogenesis in mammalian cells ( Table 1 ) . To test for functional conservation among human homologs of S . cerevisiae genes found to regulate Yor1-ΔF670 biogenesis , we used the human CF bronchial epithelial CFBE41o- ( or CFBE ) cell line with CFTRΔF508/ΔF508 genetic background with no detectable CFTR protein expression as a model [47] . CFBE cells were engineered to express inducible WT- and/or ΔF508-CFTR-3HA as described [17 , 48] . As an indirect measure of biogenesis and/or peripheral stability of ΔF508-CFTR , first the PM density of the mutant channel was determined by cell surface ELISA in polarized CFBE after siRNA-mediated knockdown of the putative target genes . Two or three nonoverlapping siRNA sequences were used individually for each candidate gene to discriminate possible off-target effects . SiRNA silencing of Yor1-ΔF modifier homologs increased the PM density of the low-temperature ( 26°C ) rescued ΔF508-CFTR ( rΔF508-CFTR ) by up to ~3-fold , representing ~8% of the WT CFTR PM density , similar to the level achieved by the FDA-approved corrector , VX-809 ( Fig 2A ) . A 50% increase in the PM density of rΔF508-CFTR for at least two siRNAs in comparison to nontargeted ( NT ) siRNA were considered as criteria for further investigation . These were met by siRNA-mediated silencing of SKIV2L ( SKI2 in S . cerevisiae ) , TTC37 ( SKI3 ) , EXOSC10 ( RRP6 ) , RPL12 ( RPL12A ) , POMP ( UMP1 ) , and TBC1D22B ( GYP1 ) ( Fig 2A ) . To determine whether the biochemical rescue of rΔF508-CFTR correlates with a gain-of-function phenotype , the PM conductance of CFBE cells was measured by the halide-sensitive YFP quenching assay in CFBE cells stably expressing the halide-sensitive YFP-H148Q/I152L/F46L , using a fluorescence plate reader [19] . The iodide influx-mediated YFP quenching was determined after maximal activation of the temperature rΔF508-CFTR with cAMP-dependent protein kinase A ( PKA ) agonists forskolin ( Frk ) , 3-Isobutyl-1-methyl-xanthine ( IBMX ) , and 8- ( 4-Chlorophenylthio ) -adenosine-3' , 5'-cyclic monophosphate ( cpt-cAMP ) in combination with the potentiator genistein ( gen ) ( Fig 2B ) . Induction of rΔF508-CFTR expression strongly increased the halide conductance of CFBE epithelia that was further augmented by knockdown of some of the Yor1-ΔF modifier gene homologs ( Figs 2B , 2C , S2A and S2B ) . Knockdown of RPL12 or SKIV2L , by two independent siRNAs , increased the mutant channel PM function by >50% . Notable , RPL12 knockdown induced ~270% transport activity of the NT control ( Fig 2C ) . RPL12 knockdown , similar to VX-809 , proportionally increased the PM density and function of rΔF508-CFTR ( Fig 2D ) . In contrast , silencing of SKIV2L and most of the other genes tested , evoked a more substantial gain in PM density in comparison to function , suggesting the preferential escape of partially folded , poorly functional ΔF508-CFTR molecules from the ER ( Fig 2D ) . The inference that RPL12 knockdown enhanced the rΔF508-CFTR conformational stability was supported by measuring the PM turnover of rΔF508-CFTR . The rΔF508-CFTR removal from the PM was ~5-fold faster ( T1/2 ~2 h ) in comparison to its WT counterpart ( T1/2 of >10 h ) after exposure to 37°C , which can be attributed to the channel unfolding , reflected by its increased protease susceptibility and ubiquitination that largely accounts for the accelerated internalization , lysosomal targeting , and attenuated recycling [11 , 49] . RPL12 knockdown considerably slowed down the rΔF508-CFTR biochemical and functional turnover at the PM ( Fig 2E and Fig 2F ) . Jointly , these observations suggest that RPL12 deficiency promotes rΔF508-CFTR functional PM expression by facilitating the biogenesis and enhancing peripheral stability of the mutant . Considering that the modifier genes of the Yor1-ΔF/ΔF-CFTR processing defect can contribute to the regulation of translation , RNA processing , and vesicle transport , we tested the possibility that silencing of TBC1D22B , RPL12 , EXOSC10 , SKIV2L , POMP , or TTC37 genes can alter the cellular expression of native and conformationally defective membrane proteins in general . Ablation of these genes , however , did not increase the Ca2+-activitated TMEM16A Cl- channel activity , nor the transferrin receptor ( TfR ) or the conformationally defective , mutant megalencephalic leukoencphalopathy with subcoritical cyst 1 ( MLC1-S280L ) PM densities , as determined by the halide-sensitive YFP quenching assay [48] , transferrin-HRP binding , or cell surface density measurement of MLC1-S280L , respectively ( S2C–S2E Fig ) . Thus , silencing of these Yor1-ΔF deletion suppressor proteins does not universally influence PM protein biogenesis . To examine the effects of RPL12 silencing at physiological temperature , the ΔF508-CFTR function at the PM was studied in polarized CFBE monolayers . Both siRNAs decreased the RPL12 protein and mRNA expression by ~50% and ~40% , respectively ( S3A and S3B Fig ) . In parallel , the ΔF508-CFTR PM density was increased by 2 . 5–5-fold relative to the NT siRNA treated cells ( Fig 3A ) . The ΔF508-CFTR PM density was comparably augmented in nonpolarized HeLa cells , suggesting that the RPL12 knockdown effect is not CFBE-specific and independent of CFTR polarized expression ( S3C Fig ) . RPL12 knockdown also enhanced the PM density of WT-CFTR in CFBE and HeLa cells ( Figs 3A and S3C ) , probably by increasing the limited maturation efficiency of the WT channel [17 , 50] . To distinguish whether RPL12 knockdown promotes the accumulation of the mature , complex-glycosylated ( band C ) ΔF508-CFTR in post-ER compartments , or causes the channel redistribution from intracellular pools to the PM , the expression level of complex-glycosylated ΔF508-CFTR was determined by immunoblot ( IB ) analysis . Since we were unable to visualize the complex-glycosylated form in the inducible CFBE expression system ( Fig 3B ) , these experiments were also performed in CFBE cells constitutively expressing the channel at higher level , while preserving the hallmarks of ΔF508-CFTR misprocessing ( CFBEC ) [51 , 52] . In this model , the RPL12 knockdown increased the abundance of the complex-glycosylated ΔF508-CFTR ( Fig 3C ) . Interestingly , RPL12 knockdown led to the steady-state accumulation of the core-glycosylated ΔF508-CFTR ( band B ) as well ( Fig 3B and Fig 3C ) . In accord with the increased PM expression , RPL12 silencing augmented the halide conductance ( Fig 3D ) and the Frk-stimulated short-circuit current ( Isc ) by up to 5-fold ( Fig 3E ) . This was further enhanced by addition of the potentiator gen in the constitutive ΔF508-CFTR expressors ( Fig 3E ) . In contrast , RPL12 knockdown had little effect on the function of WT-CFTR in CFBE ( S3D Fig ) . To confirm the relevance of RPL12 silencing on the rescue of ΔF508-CFTR misprocessing , we used primary human bronchial epithelia ( HBE ) isolated from five CF patients with CFTRΔF508/ΔF508 genotype . CF-HBE were cultured on permeable filter supports at air–liquid interface for 3 wk and transfected with either NT or RPL12_6 siRNA every 7 d , resulting in a decrease of Rpl12 protein expression by ~25% ( Fig 4A ) . Alternatively , RPL12 silencing was achieved by a single transfection with a double-stranded Dicer-substrate siRNA [53] ( S4A Fig ) . The chronic knockdown of RPL12 did not alter tight junction formation , the differentiation of globlet and ciliated cells and the development of transepithelial resistance of the monolayers ( S4D Fig and S4E Fig ) . The effect of RPL12 siRNA alone or in combination with VX-809 was determined on Frk plus gen-activated and inhibitor 172 ( Inh172 ) -sensitive Isc of ΔF508-CFTR in CFBE and CF-HBE . VX-809-mediated ΔF508-CFTR correction was additive with RPL12 knockdown , increasing the maximal Isc to ~50% of WT-CFTR in CFBE ( Figs 3E and S3D ) . As expected , a large variation of the Isc was observed in CF-HBE cells from individual patients ( Figs 4B , 4C , S4B and S4C ) . RPL12 silencing enhanced the PKA-activated current by a mean ∼2 . 2-fold ( range 1 . 1–3 . 1-fold ) in the presence of the potentiator gen . This corresponds to ∼35 . 7% of the WT-CFTR current , measured in HBE isolated from non-CF lungs ( Figs 4C , S3E and S4C ) . VX-809 treatment alone increased the ΔF508-CFTR current by ∼2 . 1-fold ( range 1 . 6–2 . 7-fold ) , representing ∼27 . 7% of the WT-CFTR current as observed before [54] ( Figs 4C , S3E and S4C ) . Combination of RPL12 knockdown with VX-809 treatment augmented the Frk- and gen-stimulated Isc by ∼4 . 0 fold ( range 3 . 1–4 . 9-fold ) , representing 54 . 4% of WT-CFTR in HBE cells ( Figs 4C , S3E and S4C ) . RPL12 knockdown also increased the amount of complex-glycosylated ΔF508-CFTR in HBE , similar to that in CFBE ( Fig 4A ) . Treatment with VX-809 alone led to ∼2-fold enhanced expression of the complex-glycosylated form in both CFBE and CF-HBE cells . The rescue effect of VX-809 was at least doubled upon RPL12 silencing , indicated by ~6- and ~4-fold increase of the band C abundance of ΔF508-CFTR in CFBE and HBE , respectively ( Figs 3C and 4A ) . These findings support the notion that distinct mechanisms are responsible for the RPL12 knockdown-mediated rescue and VX-809-dependent correction of the ΔF508-CFTR functional expression defect in both CFBE and HBE cells . We postulated that RPL12 knockdown may enhance the conformational maturation efficiency of non-native ΔF508-CFTR at the ER and/or stabilize the native-like conformation of the mutant , delaying its cell surface removal . The low folding efficiency of the nascent ΔF508-CFTR chains were measured by the conversion efficiency of the core-glycosylated ΔF508-CFTR into complex-glycosylated form during an extended radioactive pulse labeling ( 3 h ) and chase ( 2 h ) at 37°C . To measure the total [35S]-methionine and [35S]-cysteine incoporation during the pulse , while minimizing the degradation of core-glycosylated forms , the pulse duration was reduced to 30 and 20 min in parallel samples of CFBE and HeLa cells , respectively , ( Fig 5A ) . RPL12 knockdown increased the ΔF508-CFTR ER folding efficiency from ~1 . 9% to ~3 . 2% in CFBE and from ~1 . 4 to ~4 . 4% in HeLa cells ( Fig 5A and Fig 5B ) , while it decreased the radioactive labeling of the core-glycosylated ΔF508-CFTR by 50–80% both at 26°C and 37°C ( Fig 5A–5C ) without reducing the CFTR transcript level ( S5A Fig ) . We obtained similar results by using a shorter pulse-labeling time ( 10 min ) ( S5C Fig ) . Combination of RPL12 knockdown with VX-809 additively increased the ΔF508-CFTR maturation efficiency from 4 . 5% to ~7 . 5% relative to that in the presence of VX-809 alone , at 37°C ( Fig 5D ) , consistent with nonoverlapping mechanisms of action . Selectivity of the RPL12 knockdown effect was determined by measuring its impact on the cell surface density and turnover of a panel of PM proteins as surrogate readouts of their conformational stability [11 , 50 , 55 , 56] . We chose mutations in the vasopressin 2 receptor ( V2R-Y128S ) [56 , 57] , MLC1-P92S and -S280L [58] , and the human Ether-à-go-go-Related Gene ( hERG-G601S ) [55] , which cause diabetes insipidus ( V2R ) , megalencephalic leukoencephalopathy ( MLC1 ) , and long QT type 2 syndrome ( hERG ) , respectively , due to conformational defects , misprocessing and accelerated PM turnover of the respective membrane proteins [55–60] . Silencing of RPL12 had no significant effect on PM density and stability of the V2R-Y128S and MLC-P92S or -S280L in HeLa cells ( Figs 6A , 6B , S6A and S6B ) . In contrast , knockdown of RPL12 increased the WT-hERG ( ~125% of NT ) and hERG-G601S ( 140% of NT ) PM densities ( Fig 6C ) , without stabilizing the channels at the PM ( S6C Fig ) . The effect on steady-state PM density of WT and mutant hERG may be due to their inefficient conformational maturation , estimated to be ~40% and ~15% determined by metabolic pulse chase experiments [61] ( Fig 6E ) . RPL12 knockdown increased the hERG-G601S maturation efficiency from ~15% to ~25% ( Fig 6E ) . Notably , incorporation of radioactive amino acids into newly synthesised hERG-G601S was decreased in RPL12 knockdown cells ( Fig 6F ) . This could be attributed to translational inhibition rather than cotranslational degradation , since the amount of core-glycosylated hERG was increased after a 3 h chase at 37°C ( Fig 6E ) . The PM density of the TfR , a type I transmembrane protein [56] , was only modestly decreased by RPL12 knockdown ( Fig 6D ) . These results support the hypothesis that the RPL12 knockdown can preferentially influences the folding efficiency of metastable , multidomain membrane proteins . Compelling evidence indicates that synonymous mutations can influence the folding of polypeptide via defining the available time frame for the cotranslational folding and unfolding during translation in vivo and in vitro [62–65] . If translational elongation slow down can shift the folding equilibrium of newly synthesized multidomain ΔF508-CFTR towards the WT-CFTR conformer by kinetic or thermodynamic means , it is predicted that the thermostability of complex-glycosylated mutant would be increased as compared to the temperature rΔF508-CFTR . The conformational stability of CFTR channels was measured by determining the thermal denaturation temperature , which induced 50% ( T50% ) conversion of detergent solubilised CFTR into SDS-resistant aggregates [66] . Cell lysates , obtained from rΔF508 or WT CFTR CFBE expressors were heat-denatured at 20°C–80°C and the aggregation resistant , complex-glycosylated CFTR pool was quantified by immunoblotting . RPL12 knockdown increased the T50% of the complex-glycosylated rΔF508-CFTR from ~58°C to 67°C , a value that is close to the T50% of WT-CFTR ( Fig 7A and Fig 7B ) . This observation suggests that RPL12 silencing promotes the accumulation of conformationally more stable complex-glycosylated rΔF508-CFTR , conceivably by shifting the mutant folding energetic towards the WT , an inference substantiated by increased metabolic stability of the rΔF508-CFTR , determined by CHX chase and immunoblotting in HeLa cells ( Fig 7C and Fig 7D ) . RPL12 silencing also delayed the rΔF508-CFTR biochemical and functional turnover at the PM in CFBE , as determined by cell surface ELISA and halide-sensitive YFP assays ( Fig 7F and Fig 7G ) . Surprisingly , RPL12 knockdown delayed the turnover of the immature ΔF508-CFTR . The core-glycosylated form T1/2 was decreased from ~30 to ~50 min in HeLa and CFBE measured by CHX chase and immunoblotting . ( Figs 7C , 7E and S5D ) . Comparable results were obtained by the metabolic pulse-chase technique . After 2 h chase ~14% of the initially labeled immature ΔF508-CFTR remained in the RPL12 knockdown CFBE cells as compared to ~7% in the NT control at 37°C ( Fig 7H ) . Thus , impeded degradation of the core-glycosylated form may contribute to the enhanced ER maturation of ΔF508-CFTR in CFBE cells ( see also Fig 3B and 3C ) . Collectively , these observations suggest that the increased folding propensity of ΔF508-CFTR in the presence of RPL12 down-regulation account for the increased PM density , function , biochemical , and functional stability , as well as reduced potentiator-dependent fractional activation of the Isc carried by ΔF508-CFTR ( S5C Fig and S5D Fig ) . RPL12 knockdown causes translational elongation slow down in yeast [67 , 68] . Consistently , the reduced pulse labeling of ΔF508-CFTR and HERG-G601S suggested that RPL12 silencing may also impede the translational elongation rate in eukaryotes ( Figs 5B , 5C , 6F and S5C ) . Due to insufficient radioactive signal incorporation into the ΔF508-CFTR nascent chain in CFBE cells , we monitored the global kinetics of run-off elongation [69] in HeLa cells after RPL12 silencing . To this end , translation initiation was inhibited with harringtonine [70] . Then run-off elongation was measured in the presence of [35S]-methinonine and [35S]-cysteine and terminated with cycloheximide ( CHX ) [71] after 1–3 min ( Fig 8A ) . Quantification of radioactive nascent chains synthesis by phosphorimage analysis showed a considerable reduction in the global translational elongation rate in RPL12 knockdown cells ( Fig 8B ) . Importantly , slowing down global translational elongation in Fisher rat thyroid ( FRT ) cells by low concentration of CHX also induced the accumulation of the complex-glycosylated ΔF508-CFTR and partial restoration of the PM chloride conductance [72] . To confirm that pharmacological inhibition of translational elongation can partially rescue ΔF508-CFTR processing in CFBE cells , we used low-concentrations CHX or emetine . Both translational elongation inhibitors elicited a dose-dependent increase in the PM density of ΔF508-CFTR , which correlated with the decreased expression of the core-glycosylated form as a consequence of partial translational inhibition ( Fig 8C and Fig 8D ) . These results confirmed and extended previous observations [72] , suggesting that the channel folding is favoured at a slower translational elongation rate regardless of whether it is achieved by pharmacological [72] or genetic means as shown by RPL12 ablation ( Fig 8B ) . Deletion of the RPL12A and RPL12B genes in yeast leads to a defective assembly of the ribosomal stalk , disrupting the normal stoichiometry of P1 and P2 isoforms , which along with P0 , and the 26S ( 28S in humans ) rRNA , interact with Rpl12 to form the GTPase-associated center ( GAC ) [67 , 73 , 74] . The stalk has been shown to participate in the translocation mechanism and to bind the eukaryotic elongation factor 2 ( eEF-2 ) , leading to the hypothesis that P-proteins act as GTPase-activators in conjunction with eEF-2 to increase the translational elongation rate [74–77] . To support the notion that compromised integrity of the ribosomal stalk could increase the functional expression of ΔF508-CFTR , the expression levels of P0 ( RPLP0 ) , P1 ( RPLP1 ) , P2 ( RPLP2 ) , and eEF-2 proteins were reduced by siRNA-mediated knockdown in CFBE . Two or three siRNA sequences were used for each gene , reducing the mRNA expression by ~10%–60% after 5 d in polarized CFBE ( S7A–S7D Fig ) . Transfection of these siRNAs increased the PM density and function of rΔF508-CFTR by up to 5-fold as compared to NT siRNA and indicated a good correlation between the knockdown efficiencies and the gain of PM density and function of the mutant ( Figs 9A , 9B , S7F and S7G ) . Importantly , P0 , P1 , P2 , and eEF2 knockdown conformationally stabilized the mutant , as indicated by the slower biochemical and functional PM turnover of the rΔF508-CFTR ( Fig 9C and Fig 9D ) . Considering that reduced translational elongation rate following RPL12 knockdown is associated with the reduction of the ER load , and this may influence the protein homeostasis ( proteostasis ) network activity [23] , we measured the expression level of molecular chaperones and cochaperones , critical determinants of the proteostasis network folding capacity [11] . The expression level of HSP90A , HSPA8 , AHA1 , STIP1 , DNAJA1 , and BAG1 however , remained unaltered upon RPL12 ablation ( S8 Fig ) , suggesting that the ratio between the total pool of synthesized proteins and these chaperones and cochaperones was not affected by RPL12 knockdown . As an alternative method to test the influence of the ER load on cotranslational folding of ΔF508-CFTR , we partially inhibited translation initiation by reducing one of the core translation initiation factors , the cap-binding protein eIF4E , expression using siRNA-mediated knockdown . Reduction of the eIF4E mRNA by ~50% ( S7E Fig ) decreased the pulse-labeled pool of ΔF508-CFTR by ~30% ( S7L Fig ) . This led to a 20%–30% increase in the rΔF508-CFTR PM density and function without altering its fast PM turnover or functional stability ( S7H–S7K Fig ) . These results are at variance with observation obtained in FRT cells upon inhibiting initiation with hippuristanol , which increased the ΔF508-CFTR activity by ~2 . 5-fold at the PM [72] . Thus , significant reduction of protein synthesis , including ΔF508-CFTR and the ER load by impeding translational initiation without translational elongation has a modest contribution to the conformational rescue of ΔF508-CFTR , as compared to that of RPL12 silencing in CFBE and failed to provoke the accumulation of native-like , conformationally stabilized rΔF508-CFTR at the PM .
While technology developments significantly improved the identification of the interactome of conformationally defective polypeptides , including the ΔF508-CFTR [28–30] , establishing the functional significance of interacting proteins remains a daunting task . One possible approach is offered by global , quantitative analysis of functional gene interaction in yeast models . While it is appreciated that the molecular processes involved in the translation , folding , maturation , and trafficking of integral membrane proteins are evolutionarily conserved [78] , the extent to which global , quantitative analysis of gene interaction in a yeast model yields correct predictions of human genetic modifier that enhance the functional rescue of disease-causing target proteins , remains poorly understood [79 , 80] . Here , we show that gene interaction studies using the yeast ATP-binding cassette family C ( ABCC ) protein Yor1 incorporating the ΔF508-CFTR homologous mutation ΔF670 can predict novel modifier genes of ΔF508-CFTR biogenesis in respiratory epithelia . Silencing six of the seven genes selected to probe various steps during Yor1-ΔF670 biogenesis that acted as deletion suppressors to augment Yor1-ΔF670-dependent oligomycin resistance [39] , also increased PM expression of ΔF508-CFTR in the human CFBE cells . Knockdown of RPL12 , a component of the large ribosomal subunit stalk , increased both the ER folding efficiency as well as the metabolic and functional stability of the mature , complex-glycosylated ΔF508-CFTR in post-ER compartments and at the PM , respectively . These results jointly indicate the conservation/concordance of gene interaction modules influencing the early biogenesis of Yor1-ΔF670 and ΔF508-CFTR , including mRNA turnover , rRNA processing , and ribosome structural assembly . In contrast to mammalian cells , the Rpl12 protein in S . cerevisiae is encoded by a duplicated gene , RPL12A and RPL12B , which are remnants of a genome duplication event [81] . In the case of RPL12 and Yor1-ΔF biogenesis , both RPL12A and RPL12B acted as deletion suppressors , but RPL12A knockout yielded a quantitatively greater effect . Deletion of both genes is associated with a severe growth , but not a viability defect and results in a profound decrease in the translational elongation rate , while deletion of either single gene slightly retarded cell growth and elongation [67 , 68] . Knockdown of Rpl12 protein expression by ~50% considerably decreased the global translational elongation rate as determined by the kinetics of run-off elongation in human CFBE cells ( Fig 8B ) . The concomitant increase of ΔF508-CFTR PM density and function is reminiscent of a recent report by Meriin et al . , demonstrating that pharmacological attenuation of translational initiation or elongation rate by hippuristanol or CHX , respectively , can partially rescue the PM functional expression defect of ΔF508-CFTR in FRT cells [72] . While the cellular and molecular mechanism of ΔF508-CFTR rescue is not known in FRT cells , here we show that RPL12 silencing increases the ER folding efficiency of the nascent chain and the stability of both the core- and complex-glycosylated forms , as well as the PM resident ΔF508-CFTR channel . The enhanced ER folding efficiency can be attributed , at least in part , to favourably changes in the folding energetics of the mutant , as reflected by the increased thermostability of the final fold , represented by the complex-glycosylated ΔF508-CFTR in RPL12 ablated cells ( Fig 7B ) . While we do not have a definitive explanation for the significantly increased co- and post-translational folding efficiency and stabilization of ΔF508-CFTR upon translational elongation slow down , similar phenomena have been demonstrated for various polypeptides in silico , in vitro as well as in vivo [62–65 , 82 , 83] . Translational rate slow down may help to reduce the amount of misfolded domain intermediates that are resistant to conformational rescue by the proteostasis machinery , influence the stability of folding intermediates of the nascent chain and/or the folding trajectory of ΔF508-CFTR by altering the binding to profolding and/or prodegradative constituents of the ribosome associated quality control machinery [84–87] . The longer residence time of the nascent chain on the ribosome may also shield against degradation and facilitate folding . The hypothesis that reduced translation elongation increases ΔF508-CFTR folding is consistent with earlier observations showing that slowing the global translational elongation rate by various interventions can improve the folding efficiency of specific proteins in eukaryotic cells . Synonymous codon changes that result in ribosomal pausing can lead to alternate folding pathways , distinct conformations , and can facilitate cotranslational targeting of membrane proteins to the translocon [63 , 65 , 88] , as has been shown for the multidrug resistance 1 gene product P-glycoprotein [62 , 64] . In contrast , accelerating translation rate by synonymous codon substitutions in the α-subdomain of WT-CFTR NBD1 resulted in aggregation of the full-length channel [89] . Codon optimization also resulted in conformational changes and impaired function of Neurospora , a clock protein [90] . The possibility that global translational attenuation by RPL12 knockdown can exert a general profolding effect is unlikely but cannot be ruled out . RPL12 silencing did not shift the expression levels of known CFTR modifier genes such as HSP90A , HSPA8 , AHA1 , STIP1 , DNAJA1 , and BAG1 , which are molecular chaperones and cochaperones that are known modulators of ΔF508-CFTR folding and PM expression [11] . Furthermore , whereas RPL12 siRNA increased the PM expression of hERG-G601S and ΔF508-CFTR , and to some extent their WT-counterparts , RPL12 ablation had no effect on the PM density of V2R-Y128S , as well as MLC1-P92S or -S280L , despite the documented recognition of these misfolded membrane proteins by the ER and PM protein quality control systems [11 , 55 , 56 , 58 , 91] . Thus , intrinsically slow and inefficient folding and domain assembly of the WT CFTR and hERG , which is further compromised by missense mutations , may sensitize these complex multidomain proteins to rescue upon translational elongation slow down . Rpl12 is localized in the 60S subunit GAC and interacts with the GTP-bound translation factors [92–94] . Rpl12 binds to the 28S rRNA together with ribosomal protein P0 to constitute the base of the lateral ribosomal stalk ( P-stalk ) , which serves as a binding platform for ribosomal proteins P1 and P2 [67 , 73 , 74 , 95 , 96] . The stalk is present in all eukaryotic ribosomes and plays an integral role in translation elongation , due to specific interactions of P1 and P2 with eukaryotic elongation factor 1α ( eEF-1α ) and Rpl12 with eEF-2 [92–94] . Based on evidence in yeast showing disruption of stalk components reduces translation without changing translational fidelity [68 , 75 , 76 , 97] , we hypothesize that similar to the knockdown of RPL12 , silencing of other ribosomal stalk proteins , which inhibits elongation factors recruitment and slows down the translational elongation , allows the development of a native-like ΔF508-CFTR conformation . Consistent with this hypothesis , silencing of P1 , P2 , P0 , or eEF-2 increased the ΔF508-CFTR functional expression and PM stability ( Fig 9 ) . In contrast , silencing of one of the core translation initiation factors , the cap-binding protein eIF4E , decreased ΔF508-CFTR synthesis but only slightly increased ΔF508-CFTR PM density without stabilizing the rescued mutant channel biochemically or functionally at the PM . Studies in FRT cells , however , showed that partial inhibition of translation initiation with hippuristanol leads to a substantial increased ΔF508-CFTR function [72] . Thus , the concomitant reduction of the overall ER protein load might be a contributing factor to the increased PM expression and function of ΔF508-CFTR , but is likely insufficient to explain enhanced ER folding efficiency of the mutant channel . Our results showing that knockdown of different ribosomal stalk proteins increases the folding propensity and function of ΔF508-CFTR provide new insight into targeted slow-down of translational elongation as a possible therapeutic strategy for treating a conformational disease resulting from mutations in an evolutionarily conserved amino acid residue . Whether other ribosomal proteins and ribosome-bound soluble factors are involved in this mechanism or small molecule inhibitors can be identified to mimic the genetic effect of RPL12 knockdown remain to be established in future work . From a translational perspective , our findings show that RPL12 knockdown is additive with the small-molecule corrector VX-809 in primary HBE cells with CFTRΔF508/ΔF508 genotype . Moreover , the treatment combination restores ΔF508-CFTR function to ~50% of the WT level , a value that is deemed sufficient to alleviate CF clinical manifestations , since heterozygous carriers do not show disease symptoms [98] . Knockdown of other ribosomal stalk proteins and the translational elongation factor eEF-2 , similarly result in improved ΔF508-CFTR biogenesis . Thus , ribosomal stalk perturbation represents a potential target for rescuing the ΔF508-CFTR biogenesis in combination with VX-809 , which represents one the most efficacious strategies for the correction of ΔF508-CFTR thus far [20] . Future work will clarify the mechanism by which the reduced translational elongation rate via reduced function of the ribosomal stalk could be used to treat ~90% of CF patients afflicted with this most common allele . Furthermore , application of yeast phenomics as part of an integrative research approach should be considered for its potential to discover genotype-phenotype networks that can guide discovery of new therapeutics for patients with CF and possibly other diseases that can be modeled in S . cerevisiae .
Human lung tissues were obtained from CFTRWT/WT non-CF and CFTRΔF508/ΔF508 CF individuals under the protocol and consent form approved by the Institutional Review Board at University of Alabama Birmingham ( IRB #X080625002 ) . All adult participants provided informed consent , or a parent or guardian of any child participant provided informed permission on their behalf . All consent was obtained in written form . SiRNA was purchased from Qiagen , dsiRNA were obtained from Integrated DNA Technologies , and the target sequences are listed in S1 Table . The following antibodies were used: monoclonal mouse anti-HA ( MMS101R , Covance ) , monoclonal mouse anti-CFTR antibodies ( 10B6 . 2 , 570 and 596 , Cystic Fibrosis Foundation Therapeutics , Inc . ) , polyclonal rabbit anti-Rpl12 ( AP16275c , Abgent ) , polyclonal rabbit anti-Rpl12 ( ab157130 , abcam ) , monoclonal rat anti-HSP90A ( 9D2 , Enzo ) , monoclonal rat anti-HSPA8 ( 1B5 , Enzo ) , mouse monoclonal anti-HSPA4 ( C92F3A-5 , Enzo ) , monoclonal mouse anti-AHA1 ( 1A2-A8 , Abnova ) , monoclonal mouse anti-STIP1 ( DS14F5 , Enzo ) , polyclonal rabbit anti-DNAJA1 ( ADI-SPA-405 , Enzo ) , monoclonal mouse anti-BAG1 ( 4A2 , Enzo ) , monoclonal mouse anti-Mucin5A/C ( 45M1 , ThermoScientific ) , monoclonal mouse anti-acetylated tubulin ( 6-11B-1 , Sigma ) , polyclonal rabbit anti-occludin ( Zymed ) , monoclonal mouse anti-Na+/K+-ATPase ( H3 , Santa Cruz Biotechnology ) , monoclonal mouse anti-βactin ( ab8226 , abcam ) , and monoclonal mouse anti-βactin ( AC-15 , Sigma ) . The TfR PM density was detected using HRP-conjugated transferrin ( 090-030-050 , Jackson ImmunoResearch ) , and actin was labeled for immunofluorescence microscopy using A555 conjugated Acti-stain ( Cystoskeleton ) . Yeast deletion mutant strains were obtained from Research Genetics . To retest selected hits , four single-colony deletion mutant clones were selected for construction of double mutants by the SGA , as described [44 , 45] . The query allele of the RL8 strain ( described in [39] ) was modified by substitution for the Tet promoter with the ACT1 promoter ( Pact1 ) and without the C-terminal GFP tag to obtain the Pact1-yor1-F670-R1116T-HA allele . To construct this allele , the pJ023 plasmid was modified by replacing the tetracycline promoter element with 680 bp of the 5’ UTR of ACT1 promoter and then used to construct Pact1-yor1-ΔF670-R1116T-HA ( the R1116T second site mutation is described in [36] ) by integrating it at the YOR1 promoter of LMY789 , containing the yor1-ΔF670-R1116T-HA allele ( gift from Elizabeth Miller ) . After SGA , frozen glycerol stocks were made for subsequent Q-HTCP analysis . YPEG media was made with 10 g/L yeast extract , 20 g/L peptone , 3% ethanol , and 3% glycerol . Q-HTCP was performed as previously described [46] , assessing gene interaction by comparison of the cell proliferation parameter , L ( the time ( hours ) after which half carrying capacity is reached ) , obtained from growth curves for single and double mutant cultures across a range of perturbation intensity using different growth-inhibitory concentrations of oligomycin , the primary substrate of Yor1 . Hundreds of replicates of the single mutant , Yor1-ΔF670 , were used to obtain the range of L at each condition tested . Interaction was calculated as the difference between the L values for the double mutant and the Yor1-ΔF670 single mutant , after normalizing all data by the difference in L between the double mutant and the Yor1-ΔF670 single mutant reference median in media not containing oligomycin , as previously described [39] . Human lung tissues were obtained from CFTRWT/WT non-CF and CFTRΔF508/ΔF508 CF individuals ( patient code 21 , 22 , and 48 ) under the protocol and consent form approved by the Institutional Review Board at University of Alabama Birmingham ( IRB #X080625002 ) . Cell isolation was performed as described [51 , 99] . After expansion , first or second passage cells were transferred to permeable filter supports and allowed to differentiate under air–liquid interface conditions for 3 wk as described in [100 , 101] . Alternatively , HBE cells ( patient code BCFr43 and BCF121209 ) were purchased from the Cystic Fibrosis Translational Research center ( CFTRc ) at McGill University , expanded following the conditional reprogramming protocol described in [102] followed by differentiation on permeable filter supports under air–liquid interface using the Vertex conditions [100] . The generation and maintenance of the CFBE cell lines , expressing WT- or ΔF508-CFTR with a 3HA-tag in the 4th extracellular loop under the control of a tetracycline-inducible transactivator was described before [17 , 48] . Generation of stable CFBE cell lines constitutively expressing WT- or ΔF508-CFTR has been described and these cell lines are depicted as CFBEC [51 , 52] to distinguish them from the inducible CFBE expression system . The HeLa cells constitutively expressing extracellular 3HA-tagged WT- or ΔF508-CFTR have been described previously [11] . In CFBE cells , the target genes were silenced by forward transfection using 50 nM siRNA ( Qiagen ) and Lipofectamine RNAiMAX ( Invitrogen ) . Transfected cells were allowed to polarize for 5 d . For studies in filter grown CFBE cells , 100 nM siRNA duplexes were introduced by reverse transfection using an established protocol [53] . HeLa cells were transfected using Oligofectamine ( Invitrogen ) as described previously [11] , and the experiments were performed 4 d after transfection . HBE cells were either reverse transfected with dsiRNA ( Integrated DNA Technologies ) using Lipofectamine RNAiMAX [53] or were forward transfected for 6 h 1 d postseeding with 50 nM NT or RPL12_6 siRNA and the transfections were repeated every 7 d . Monolayers were cultured for a total of 21 d . PM densities of extracellular HA-tagged proteins were determined by cell surface ELISA [11] . The TfR PM density was measured by the specific binding of HRP-conjugated transferrin . The relative amount of transferrin-HRP or HRP-conjugated secondary antibodies was measured by luminescence , using a VICTOR Light plate reader ( PerkinElmer ) after addition of 50 μl/well HRP-Substrate ( SuperSignal West Pico , Thermo Fisher Scientific ) . PM density measurements were normalized for cell number , determined by Alamar Blue assay ( Invitrogen ) or protein concentration , determined by bicinchoninic acid assay ( BCA , Pierce ) . The halide-sensitive YFP quenching assay has been used before to determine the PM function of CFTR [19] or TMEM16A [48] . Briefly , ΔF508-CFTR was activated by well-wise injection of 50 μl/well activator solution ( 20 μM Frk , 0 . 5 mM IBMX , 0 . 5 mM cpt-cAMP and 100 μM gen in PBS ) followed by 100 μl PBS-iodide in which NaCl was replaced with NaI after a delay of 60 s . TMEM16A was activated by well-wise injection of 50 μl/well 100 μM ATP in PBS followed by 100 μl PBS-iodide after a delay of 6 s . For immunoblot analysis , cells were lysed either with NP-40-based lysis buffer ( 150 mM NaCl , 1% NP-40 , 50 mM Tris-HCl , pH 8 . 0 ) or RIPA ( ThermoScientific ) in the presence of Halt protease inhibitor cocktail ( ThermoScientific ) . Following total protein quantification using the BCA assay ( ThermoScientific ) , equal amounts of protein were mixed with 4X loading buffer , incubated at 37°C for 10 min , resolved by SDS-PAGE , and blotted onto PVDF membranes . Following antibody binding , signals were detected using the SuperSignal West Femto ( ThermoScientific ) or Luminata Crescendo ( EMD Millipore ) substrates and quantified on a ChemiDocXRS ( Bio-Rad ) or by densitometry using ImageJ . Thermoaggregation assays were performed as previously described [66] . Briefly , following cell lysis with RIPA buffer , the lysates were cleared by centrifugation and the aggregation tendency of WT- and ΔF508-CFTR was compared after exposing the lysates to 20–80°C for 15 min . Macromolecular aggregates were eliminated by centrifugation ( 15 , 000 × g for 15 min ) . The remaining soluble WT- and ΔF508-CFTR , and HSPA4 in the supernatant was measured by quantitative immunoblotting . CFBE or HBE cells expressing WT- or ΔF508-CFTR were grown to confluence on permeable filters at an air-liquid interface and mounted in modified Ussing chambers . Isc measurements were obtained under voltage clamp conditions using MC8 equipment and P2300 Ussing chambers ( Physiologic Instruments ) as previously described [51 , 52 , 103] . Cells were equilibrated for 5–10 min in regular Ringer solution ( 115 mM NaCl , 25 mM NaHCO3 , 2 . 4 mM KH2PO4 , 1 . 24 mM K2HPO4 , 1 . 2 mM CaCl2 , 1 . 2 mM MgCl2 , 10 mM D-glucose , pH 7 . 4 ) . In some measurements , this was followed by the exchange of low chloride Ringer ( 115 mM NaCl reduced to 1 . 2 mM NaCl and 10 mM D-glucose replaced with 115 mM Na-gluconate ) to the apical surface . After addition of the sodium channel inhibitor amiloride ( 100 μM ) , the CFTR agonists Frk ( 10 μM ) and gen ( 50 μM ) were sequentially added , followed by inhibitor172 ( 10 μM ) at the conclusion of each experiment , in order to specifically inhibit CFTR activity . Changes in CFTR-mediated ion transport were calculated using the highest current value for each sample after achievement of a stable plateau for several minutes . Experiments were performed essentially as described [104] . Briefly , 4–5 d after NT or RPL12 siRNA transfection , CFBE ( or HeLa ) cells expressing ΔF508-CFTR were pulse-labeled with 0 . 2 mCi/ml ( or 0 . 1 mCi/ml for HeLa cells ) 35S-methinonine and 35S-cysteine ( EasyTag Express Protein Labeling Mix , PerkinElmer ) in cysteine and methionine-free medium for 30 min for CFBE ( or 20 min for HeLa ) at 26°C or 37°C ( incorporation efficiency ) or labeled for 3 h at 26°C and then chased 2 h at 37°C in full medium ( maturation efficiency ) . Radioactivity incorporated into the core- and complex-glycosylated glycoproteins was visualized by fluorography and quantified by phosphoimage analysis using a Typhoon imaging platform ( GE Healthcare ) . For depiction of representative images , autoradiographs were acquired by film-exposure . For the experiments involving VX-809 , NT , or RPL12 , siRNA-transfected HeLa cells expressing ΔF508-CFTR were treated with 3 μM VX-809 or DMSO for 24 h before the experiment . Cells were pulse-labeled with 0 . 1 mCi/ml for 20 min and chased 2 h at 37°C in the presence or absence of VX-809 . The maturation efficiency was determined by calculating the percent of pulse-labeled immature , core-glycosylated ΔF508-CFTR conversion into the mature , complex-glycosylated form . To allow for the detection of the low percentage of complex-glycosylated ΔF508-CFTR , the total labeling for 3 h was extrapolated from values obtained for 20 or 30 min pulse labeled samples . HeLa cell transfected with either NT or RPL12_6 siRNA were treated with harringtonine ( 2 μg/ml ) to stop translation initiation [69] . Ongoing elongation was quantified by the metabolic pulse chase technique , using 35S-methinonine and 35S-cysteine . The run-off elongation was terminated after 1–3 min with CHX ( 100 μg/ml ) . The cell lysates were separated by SDS-PAGE and the total radioactive signal of fluorographs was measured by phosphoimage analysis . The signals were normalized for the amount of protein loaded and expressed as percentage on non-CHX-treated controls . Q-PCR was performed as described previously [48] . The primers are listed in S2 Table . Filter-grown differentiated HBE cells were washed with PBS containing 5 mM DL-dithiothreitol for 5 min on ice to remove mucin . Cells were fixed in 4% paraformaldehyde and permeabilized with 0 . 2% Triton X-100 . For occludin staining , cells were pre-extracted with 0 . 2% Triton X-100 for 2 min before fixation . After blocking in PBSCM ( PBS containing 1 mM MgCl2 , 0 . 1 mM CaCl2 ) with 0 . 5% bovine serum albumin , cells were incubated with primary antibodies overnight at 4°C , washed with PBSCM , and stained with Alexa-Fluor-conjugated secondary antibodies ( 1:1000 ) or Acti-stain ( 1:300 ) for 1 h at room temperature . Nuclei were stained with DAPI . Cut-out pieces of filter were mounted on glass slides , and optical stacks were acquired using a laser confocal fluorescence microscope ( LSM780 Carl Zeiss ) equipped with a 63X/1 . 40 oil DICIII Plan apochromat objective . Typically 20–30 optical xy-sections were acquired and reconstituted using the Zen 2012 software , and representative xz-sections are shown . Results are presented as mean ± SEM with the number of experiments indicated . Statistical analysis was performed by two-tailed Student's t test with the means of at least three independent experiments , and the 95% confidence interval was considered significant . | Cystic fibrosis ( CF ) is one of the most common autosomal recessive diseases in Caucasians . It is caused by mutations in the CF transmembrane conductance regulator ( CFTR ) , which functions as an anion channel at the apical plasma membrane of secretory epithelia . The most common CF mutation , a deletion of the phenylalanine residue at position 508 ( ΔF508 ) , results in the channel misfolding and subsequent intracellular degradation . Our previous genome-wide phenotypic screens , using a yeast variant , have predicted modifier genes for ΔF508-CFTR biogenesis . Here , we show that silencing of one of these candidate genes , RPL12 , a component of the ribosomal stalk , increased the folding and stabilization of ΔF508-CFTR , resulting in its increased plasma membrane expression and function . Our data suggest that reducing the translational elongation rate via RPL12 silencing can partially reverse the ΔF508-CFTR folding defect . Importantly , RPL12 silencing in combination with the corrector drug VX-809 ( lumacaftor ) , increased the mutant function to 50% of the wild-type CFTR channel , suggesting that the ribosomal stalk perturbation may represent a therapeutic target for rescuing the ΔF508-CFTR biogenesis defect . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"transfection",
"hela",
"cells",
"molecular",
"probe",
"techniques",
"gene",
"regulation",
"biological",
"cultures",
"immunoblotting",
"membrane",
"proteins",
"fungi",
"luminescent",
"proteins",
"model",
"organisms",
"yellow",
"fluorescent",
"protein",
"cell",
"cultures",... | 2016 | Ribosomal Stalk Protein Silencing Partially Corrects the ΔF508-CFTR Functional Expression Defect |
Hepatitis C virus ( HCV ) infection constitutes a significant health burden worldwide , because it is a major etiologic agent of chronic liver disease , cirrhosis and hepatocellular carcinoma . HCV replication cycle is closely tied to lipid metabolism and infection by this virus causes profound changes in host lipid homeostasis . We focused our attention on a phosphatidate phosphate ( PAP ) enzyme family ( the lipin family ) , which mediate the conversion of phosphatidate to diacylglycerol in the cytoplasm , playing a key role in triglyceride biosynthesis and in phospholipid homeostasis . Lipins may also translocate to the nucleus to act as transcriptional regulators of genes involved in lipid metabolism . The best-characterized member of this family is lipin1 , which cooperates with lipin2 to maintain glycerophospholipid homeostasis in the liver . Lipin1-deficient cell lines were generated by RNAi to study the role of this protein in different steps of HCV replication cycle . Using surrogate models that recapitulate different aspects of HCV infection , we concluded that lipin1 is rate limiting for the generation of functional replicase complexes , in a step downstream primary translation that leads to early HCV RNA replication . Infection studies in lipin1-deficient cells overexpressing wild type or phosphatase-defective lipin1 proteins suggest that lipin1 phosphatase activity is required to support HCV infection . Finally , ultrastructural and biochemical analyses in replication-independent models suggest that lipin1 may facilitate the generation of the membranous compartment that contains functional HCV replicase complexes .
Millions of humans are chronically infected by hepatitis C virus ( HCV ) worldwide [1] . Chronic HCV infection is a major biomedical problem as it causes liver inflammation and fibrosis , which can lead to severe liver disease , such as cirrhosis and hepatocellular carcinoma [2 , 3] . There is no vaccine against HCV and , although blood-screening tests and other prophylactic measures have reduced the dissemination of this pathogen , a number of newly acquired infections still occur associated with risk behavior or with unknown origin [4 , 5] . However , chronic HCV infection can be successfully eradicated from chronically infected individuals through specific direct-acting antiviral ( DAA ) combination therapies , virtually in all treated patients [6] . Since these specific treatments have only been in place recently , there are no sufficient clinical data on the long-term benefit of these treatments in relieving the severity of advanced liver disease [7 , 8] . HCV is a Hepacivirus ( Flaviviridae ) with a positive sense , single-strand RNA genome that encodes a single open reading frame ( ORF ) flanked by untranslated regions ( UTR ) , which are essential for viral polyprotein translation and viral genome replication . HCV ORF is co- and post-translationally processed by cellular and viral proteases to produce ten major proteins . These have been functionally classified in a replication module , that includes the minimal viral components of the RNA replicase ( NS3 , NS4A , NS4B , NS5A and NS5B ) and an assembly module , which comprises the major structural components of enveloped HCV virions , the capsid protein ( core ) and the envelope glycoprotein complex formed by E1 and E2 heterodimers; as well as the polypeptides p7 and NS2 , which are not structural components of virions but contribute to infectious particle assembly in a concerted action with the viral replicase [9] . HCV utilizes key aspects of cellular lipid metabolism for essentially every aspect of the virus replication cycle and strongly interferes with host cell lipid homeostasis [10–12] . In fact , chronic HCV patients display high rates of liver steatosis , severity of which inversely correlates with serum liver derived-lipoprotein [13] . Thus , although host immune response remains a major component in HCV pathogenesis , direct interference of HCV infection with hepatocyte lipid metabolism may contribute to overall disease progression [13] . One of the salient manifestations of HCV interference with lipid metabolism at the cellular level is the formation of a distinctive membranous web in HCV-replicating cells [14] . Within this structure , HCV RNA replication is thought to occur in double-membrane vesicles ( DMVs ) that emerge from the endoplasmic reticulum [15–17] . HCV replicase formation requires not only viral replicase components NS3 through NS5B [16] , but also recruitment and subversion of different key cellular factors that cooperate to provide an optimal membrane microenvironment for the assembly of HCV replicase complexes [11] . In this sense , HCV replicase complexes are located in characteristic detergent-resistant membranes ( DRMs ) that co-fractionate with caveolin-2 [18–20] and the sigma-1 receptor ( SIGMAR1 ) [21] . These endoplasmic reticulum ( ER ) microdomains containing the viral replicase are enriched in cholesterol , probably by the direct involvement of cellular proteins involved in non-vesicular cholesterol transport [11] . Assembly of progeny virions is thought to occur at the interphase between ER and lipid droplets ( LD ) [12] and HCV virion assembly and secretion processes are strongly dependent on host factors involved in biogenesis of triacylglycerol ( TAG ) -rich lipoproteins [22–25] . Thus , HCV infection requires de novo biosynthesis of phospholipids , like phosphatidylcholine ( PC ) , in order to generate its membranous RNA replicase compartments [26] and TAG biosynthesis to ensure progeny virion production [12] . This , together with the fact that PC and TAG biosynthesis are altered by HCV infection [26–28] , suggests an important role for glycerophospholipid metabolism in HCV infection . Lipins are key players in glycerophospholipid metabolism , as they catalyze phosphate removal from phosphatidic acid ( PA ) to produce diacylglycerol ( DAG ) , which is a precursor of TAG , but also a precursor for PC and phosphatidylethanolamine ( PE ) biosynthesis in mammalian cells [29] . Lipins may also translocate to the nucleus to directly interact with promoter DNA-binding transcription factors to stimulate or repress transcription of lipogenic genes [30–32] . Two conserved motifs are required for phosphatidic acid phosphatase ( PAP ) catalytic activity ( DXDXT ) and transcriptional regulation ( LXXIL ) respectively [31] . Substrate specificity is restricted to PA , a characteristic that , together with the lack of integral membrane domains and the dependence on Mg2+ for catalytic activity , differentiates lipins from other mammalian PAP [33] . Lipin1 , the best-characterized member of the family , was discovered as a consequence of the genetic analysis of a mouse strain displaying fatty liver dystrophy phenotype ( fld mice ) [34 , 35] . Interestingly , lipin1 deletion in the liver of fld mice , results in accumulation of lipin2 protein , a second member of the family predominantly expressed in the liver , which maintains normal cellular PAP activity and compensates some , but not all , aspects of glycerolipid metabolism [36] . Conversely , LPIN2 knock-out mice display increased lipin1 protein expression , as compared with wild type ( wt ) littermates , suggesting a similar homeostatic compensatory mechanism to maintain liver PAP activity [37] . Thus , lipin1 and lipin2 coordinate lipid homeostasis in the liver [37] . While individual LPIN1 or LPIN2 deletions are tolerated in mice , double knockout mice are embryonically lethal [37] . Despite this apparent functional redundancy , lipin1 is emerging as a key player in ethanol-induced steatosis [38] and a single nucleotide polymorphism in LPIN1 is associated with the severity of liver damage and fibrosis progression in pediatric human patients with histological non-alcoholic fatty liver disease ( NAFLD ) [39] , suggesting that liver lipin1 dysfunction may contribute to steatosis-related liver pathogenesis . In this study , we focused our attention on lipin1 , as analysis of previously published transcriptomic profile datasets revealed that this gene is consistently regulated at the mRNA level during HCV infection . Silencing experiments indicate that HCV infection efficiency is strongly dependent on lipin1 expression . Using different cell culture models of infection , we identified a limiting role of lipin1 PAP activity in the generation of HCV replicase complexes . Defective replicase assembly leads to strong inhibition of HCV propagation in lipin1-deficient cells but not that of another ( + ) strand RNA virus , indicating a specific role for lipin1 in HCV infection .
Given the relevance of lipin1 for lipid metabolism and its steatogenic potential , we set out to independently verify data from published differential transcriptomic profile studies that suggested that HCV infection may alter LPIN1 mRNA abundance in cell culture [40 , 41] . A specific and statistically significant LPIN1 mRNA induction ( Fig 1A ) was observed in Huh-7 cells after single cycle infection experiments [multiplicity of infection ( MOI ) 10] with a cell culture-adapted genotype 2a HCV ( D183 ) variant [42] at the peak of the infection ( 48 and 72 hours post-infection ) and as compared with mock-infected cells ( Fig 1B ) . LPIN1 mRNA induction was prevented when infected cells were treated with 1μM sofosbuvir ( Fig 1C ) , an HCV RNA polymerase inhibitor [43] that reduced viral RNA accumulation by more than two orders of magnitude ( Fig 1D ) , indicating that active HCV replication is required to induce LPIN1 mRNA accumulation . LPIN1 mRNA has been shown to be upregulated by mechanisms that involve induction of reactive oxygen species ( ROS ) , as treatment of the cells with ROS-scavenger molecule N-acetylcysteine ( NAC ) is capable of preventing LPIN1 mRNA induction under glucose deprivation [44] or during H2O2 treatment [45] . Given that transcriptional activation of a subset of lipogenic genes during HCV infection is also prevented by addition of antioxidants [46] , we sought to determine if LPIN1 mRNA induction by HCV infection is mediated by ROS production and therefore dampened by NAC treatment . LPIN1 mRNA induction was prevented in the presence of the antioxidant ( Fig 1E ) , despite comparable HCV RNA accumulation in mock-treated and NAC-treated HCV-infected cells ( Fig 1F ) , suggesting that virus replication-induced ROS production is required to induce LPIN1 mRNA accumulation . Western-blot analysis confirmed that the observed transcriptional change leads to a correlative protein accumulation ( Fig 1G and 1H ) ; reinforcing the notion that acute HCV infection alters lipin1 expression . In order to determine if lipin1 subcellular localization was altered during HCV infection , we performed confocal microscopy studies in control and HCV-infected Huh-7 cells . Lipin1 staining was observed as cytoplasmic punctated structures both in control and HCV-infected cells . To study if lipin1 signal colocalized with viral antigens , we performed double staining with antibodies against lipin1 and double-stranded RNA ( dsRNA ) or replicase subunits NS3 and NS5A . None of the viral antigens strictly colocalized with lipin1 ( Pearson´s<0 . 5 ) ( S1A Fig ) . However , Mander´s coefficients indicate that the majority of lipin1 overlapped with a small fraction of NS3 and NS5A ( S1C Fig ) . In contrast to lipin1 , lipin2 signal did not overlap with that of viral proteins NS3 and NS5A ( S1B Fig ) . Our results indicate that no major lipin1 rearrangements are observed after HCV infection and that only a minor fraction of NS proteins colocalize with lipin1 . To study if lipin1 plays any role in HCV infection , lipin1-deficient cells were generated by transducing human hepatoma ( Huh-7 ) cells with lentiviral vectors expressing shRNAs targeting LPIN1 mRNA or a control vector expressing an irrelevant shRNA . Lipin1 expression silencing was verified by Western-Blot , typically 7 days post-transduction ( Fig 2A ) . A partial ( 50%; shLPIN1-1 ) and a more profound ( >95%; shLPIN1-2 ) reduction in lipin1 accumulation was observed after transduction with specific shRNAs as compared with the control ( Fig 2B ) . As expected , lipin1 and lipin2 expression is inversely correlated in lipin1 shRNA-expressing cells ( Fig 2B ) . The viability of lipin-deficient cells , as determined by MTT assay [47] was comparable to that of control cells ( Fig 2B ) . These results illustrate that lipin1 shRNA-expressing cells respond homeostatically to functionally compensate partial ( shLPIN1-1 ) and more pronounced ( shLPIN1-2 ) loss of lipin1 ( Fig 2B ) . Control and lipin1-deficient cells were infected with a genotype 2a D183 at MOI 0 . 01 to study viral spread by determining extracellular infectivity titers at different times post infection . Fig 2C shows limited propagation of this virus in lipin1-deficient cells , with a statistically significant reduction of up to two ( shLPIN1-1 ) and three orders of magnitude ( shLPIN1-2 ) in viral titer between the control and lipin1-deficient cell lines at day 5 post-infection . These results indicate that lipin1 is required for efficient HCV propagation and that homeostatic lipin2 accumulation ( Fig 2B ) is not sufficient to support efficient HCV infection . Since important differences in the interaction of different HCV genotypes with cellular lipid metabolism have previously been described [48] , we set out to determine if the observations made with a JFH1-derived virus ( genotype 2a ) were extensive to other HCV genotypes . We performed low multiplicity infections ( MOI 0 . 05 ) with genotype 1a TNcc virus strain in lipin1-deficient Huh-7 . 5 cell lines , which are susceptible to infection by this recombinant virus [49] . First , we verified that Huh-7 . 5 . 1 also display a significant reduction in genotype 2a infection efficiency when lipin1 is silenced ( S2B Fig ) . In order to determine TNcc infection efficiency , intracellular HCV RNA accumulation was determined 72 hours post-inoculation in control and lipin1-deficient Huh-7 . 5 cells . Viral RNA detected under these conditions reflects the ability of genotype 1a to infect and replicate viral RNA in the different cell lines , as treatment of control cells with an HCV polymerase inhibitor 2´-C-methyladenosine ( 2mAde; 10 μM ) [50] reduced viral RNA content by two orders of magnitude ( Fig 2D ) . Lipin1 silencing consistently and significantly reduced TNcc replication by approximately 3-fold both in shLPIN1-1 and shLPIN1-2 cell lines ( Fig 2D ) , indicating that lipin1 is also limiting for genotype 1a HCV infection . To determine the specificity of these observations , identical lipin1-deficient and control Huh-7 cultures were inoculated at MOI 0 . 01 with a human alpha-coronavirus CoV-229E bearing a GFP reporter gene ( hCoV-229E-GFP ) [51] . Inoculation of control and lipin1-deficient cells with this virus resulted in comparable progeny virus production , as determined by infectivity titration in cell supernatants 48 hours post-infection ( S3 Fig ) . These results suggest that lipin1 is not rate limiting for CoV-229E-GFP infection and that lipin1 expression is particularly limiting for HCV . To determine which aspects of the HCV replication cycle are limited by lipin1 silencing , single cycle infection experiments were conducted by inoculating control and lipin1-deficient cell cultures at MOI 10 with genotype 2a D183 virus . Infection efficiency was measured by titration of progeny virus infectivity present in the supernatant of infected cells and intracellular HCV RNA accumulation at 48 and 72 hours post-infection . Infection of lipin1-deficient cells resulted in a significant reduction of progeny infectious virus production in shLPIN1-1 and shLPIN1-2 cells as compared with the titers observed in the supernatants of control cells ( Fig 3A ) , reinforcing the notion that lipin1 silencing interferes with HCV infection . Reduced virus production is likely due to parallel reduction of intracellular HCV RNA levels observed in lipin1-deficient cells as compared with the control cell line ( Fig 3B ) . This reduction was observed at all time points , except for that at 5 hours , indicating that the size of the inoculum and initial virus adsorption is comparable among the different cell lines ( Fig 3B ) . Thus , lipin1 silencing suppresses HCV infection by interfering with a step of the HCV lifecycle preceding intracellular HCV RNA accumulation . Next , we set out to determine if lipin1 silencing has any impact on persistent HCV infections to verify if lipin1 is also limiting for late aspects of the virus lifecycle . Persistently infected cells continuously replicate viral RNA , express viral antigens and secrete infectious virions . Thus , it is a valuable system to measure steady-state HCV RNA replication as well as infectious particle assembly and secretion . Persistently infected cultures were transduced with the lentiviral vectors described above to produce persistently infected , lipin1-deficient cells ( S4 Fig ) . Analysis of extracellular infectivity titers revealed that infectivity titers in lipin1-deficient and control cells were comparable , with the exception of a marginal reduction in shLPIN1-2 expressing cells , indicating that lipin1 silencing does not strongly interfere with infectious virus production ( Fig 3C ) . Intracellular HCV RNA levels in lipin1-deficient cells were comparable to that of the control cells ( Fig 3D ) , indicating that lipin1 expression is not rate limiting for HCV RNA replication once infection has been established . Taken together , the results shown above indicate that lipin1 is only limiting at early steps of HCV infection leading to viral RNA accumulation . As an independent verification of the hypothesis that lipin1 is limiting for early aspects of HCV infection , we used a single cycle surrogate infection model based on the production of HCV virions bearing a defective reporter genome encapsidated by trans-complementation ( HCVtcp ) . HCVtcp are capable of producing abortive single-cycle infections , efficiency of which is proportional to the luciferase activity found in the target cells [52] . As expected from the results shown in Fig 3 , infection of lipin1-deficient cells with HCVtcp resulted in a 75% reduction in the reporter luciferase activity in shLPIN1-1 cells and 90% in shLPIN1-2 determined at 48 hours post-infection , proportionally to the degree of silencing in these cells ( Fig 4A and 4B ) . These results underscore the role that lipin1 plays at early steps of HCV infection and indicate that either viral entry or a step leading to efficient HCV RNA replication is impaired in these cells . LPIN-1 mRNA is alternatively spliced in human liver to produce isoforms α and β , which may differ in catalytic activity , subcellular localization and gene expression regulation [53] . To determine the relative contribution of these isoforms to HCV infection , isoform-specific shRNAs were generated and used to transduce Huh-7 cells , as described above . LPIN1α-specific shRNA ( shLPIN1-3 ) reduced total lipin1 protein by 70% , while LPIN1β-specific shRNA ( shLPIN1-4 ) reduced total lipin1 expression by only 30% ( Fig 4A ) . Infection of control and lipin1 isoform-deficient cells with HCVtcp revealed that lipin1α silencing resulted in a strong ( 85% ) reduction in HCV infection while lipin1β silencing resulted in a milder ( 55% ) but significant reduction in HCV infection efficiency ( Fig 4B ) . These results indicate that both isoforms alpha and beta are limiting for HCV infection and suggest that the total amount of lipin1 present in the cell determines HCV infection efficiency . Taken together , the results obtained with four different shRNAs indicate that total lipin1 expression levels strongly correlate with HCVtcp infection efficiency ( Fig 4C ) , underscoring a role for this host protein in early aspects of HCV infection . Reduced HCV RNA accumulation in a single cycle infection ( Fig 3 ) as well as reduced HCVtcp infection ( Fig 4 ) may be due to a defect in entry of incoming virions . HCV E1E2-pseudotyped retroviral vectors bearing a luciferase gene ( HCVpp ) were used to measure viral entry because they constitute a sound model to study viral adsorption , receptor-mediated internalization and E1E2-mediated fusion in endosomes [54] . To assess the specificity of these observations , parallel cultures were inoculated with VSV-G pseudotyped retroviral vectors ( VSVpp ) . Control and lipin1-deficient cell lines were infected with HCVpp ( genotype 2a; JFH-1 strain ) and VSVpp . As a positive control of inhibition of HCV entry , we used hydroxyzine ( 5 μM ) , which efficiently blocks HCV infection by interfering with viral entry [55] . As expected , hydroxyzine selectively inhibited HCVpp infection , as shown by reduced luciferase levels 48 hours post-inoculation only in HCVpp-infected cells ( Fig 5A ) . Interestingly , lipin1-deficient cells ( shLPIN1-1 and shLPIN1-2 ) were fully susceptible to HCVpp and VSVpp infection , as comparable luciferase activity levels were found in all cell lines 48 hours post-inoculation ( Fig 5A ) . These results indicate that lipin1 is not rate limiting for receptor binding , particle internalization or E1E2-mediated endosomal fusion , which are steps recapitulated in this model [54] . Based on the data presented thus far , we hypothesized that lipin1 is limiting for a step in the HCV lifecycle downstream of HCV entry , leading to HCV RNA accumulation . To verify the hypothesis that lipin1 silencing causes strong reduction in initial HCV RNA accumulation by interfering with a step downstream of viral entry , we bypassed this step of the viral replication cycle by transfecting a subgenomic HCV RNA replicon bearing a reporter luciferase gene into control and lipin1-deficient cells . First , we evaluated HCV-IRES driven primary translation of incoming genomes by transfection of a replication-deficient mutant replicon that bears an inactivation mutation in the catalytic site of NS5B RNA polymerase [56] . Luciferase activity measured at 5 hours post-transfection was not reduced in any of the cell lines , indicating that transfection efficiency and HCV IRES-dependent primary translation was not significantly affected by lipin1 silencing ( Fig 5B ) . A significant increase in Renilla luciferase activity was observed in shLPIN1-1 cells both when transfecting a replicon ( Fig 5B ) or a plasmid expressing Renilla luciferase under a minimal RNA polymerase II promoter ( S5A Fig ) , suggesting that the increase in luciferase activity observed in these cells is not related with HCV . HCV RNA replication was evaluated by measuring accumulation of a reporter luciferase gene at 5 and 48 hours post-transfection of a replication competent HCV subgenomic replicon . Under these experimental conditions luciferase accumulation at 5 hours also represents HCV IRES-driven primary translation of the input RNA , while reporter luciferase activity at 48 hours depends on effective HCV RNA replication . In contrast to primary translation , HCV RNA replication inferred by luciferase activity at 48 hours was strongly reduced in lipin1-deficient cells ( 90% in shLPIN1-1 and 99% in shLPIN1-2 cells ) ( Fig 5C ) , indicating that initiation of HCV replication is dependent on normal lipin1 expression . This reduction was not due to a non-specific defect in luciferase expression , as co-transfection of a plasmid expressing Renilla luciferase lead to comparable luciferase accumulation in all cell lines , discarding the possibility that death of transfected cells or other spurious effects are responsible for the reduced luciferase activity accumulation ( S5A Fig ) . Similar experiments were conducted in ATG4B-deficient cells , as this host factor was shown to be limiting for primary HCV translation [57] . Our studies confirmed that , while partial ATG4B silencing ( S5B Fig ) significantly interfered with primary translation ( S5C Fig ) and correlatively with HCV RNA replication efficiency , as expected [57] , lipin1-silencing only affected HCV RNA replication ( S5D Fig ) . Overall , these data indicate that HCV RNA replication is not initiated efficiently in lipin1-deficient cells and that blockade occurs at a step downstream translation of incoming HCV genomes . In order to determine if any of the known functions ascribed to lipin1 is required for HCV infection , we tested the ability to restore HCV infection susceptibility of silencing-resistant wild-type ( wt ) or mutant lipin1 versions bearing a mutation in the catalytic site responsible for its phosphatase activity ( DXDXT ) and a mutant in the LXXIL motif , which is inactive both for transcriptional activation as well as for phosphatase activity [31] . Control and lipin1-deficient cells were transfected with wt lipin1 beta cDNA as well as with DXDXT and LXXIL mutants . Comparable overexpression levels of the wt and mutant proteins was obtained in each cell line , although overexpressed lipin1 levels were consistently higher in lipin1-deficient cells ( Fig 6A ) . Cells were subsequently inoculated with HCV D183 at MOI 10 and relative infection efficiency was calculated by determining extracellular infectivity titers 48 hours post-infection . Overexpression of wt lipin1 did not significantly alter susceptibility to HCV infection in control cells , although we observed a small but consistent reduction in extracellular infectivity titers when overexpressing wt and mutant lipin1 constructs ( S6A Fig ) . In contrast to control cells , wt lipin1 overexpression in lipin1-deficient cells consistently increased infectivity titers as compared to mock-transfected cells ( S6B Fig ) and in clear contrast with the reduction observed in control cells ( S6A Fig ) or when overexpressing mutant lipin1 in lipin1-deficient cells ( S6B Fig ) . In order to take into account this divergent behavior in control and lipin1-deficient cells , we calculated the relative infection efficiency as the ratio of the infectivity titers found in lipin1-deficient cells and control cells ( S6C Fig ) . Ratios were subsequently normalized to that found in mock-transfected cells in order to average the data from different experiments ( S6 Fig ) . Using relative infection efficiency as readout of this set of experiments , we could clearly observe a statistically significant increase ( 2-fold ) in the relative infection susceptibility in cells overexpressing wt lipin1 cDNA as compared with mock-transfected cells or cells expressing similar or higher levels of the mutants ( Fig 6A ) , suggesting that wt lipin1 modestly , though significantly , rescues HCV infection while phosphatase ( DXDXT ) and transcriptional coactivation ( LXXIL ) mutants do not ( Fig 6B ) . These results suggest that lipin1 transcriptional co-activation capacity is not sufficient to support HCV infection while lipin1 phosphatase activity is essential . However , given that LXXIL mutant is deficient both in transcriptional co-activation and PAP activity [31] , we cannot determine if transcriptional co-activation by lipin1 is also required to support HCV infection . The data described above suggest a role for lipin1 phosphatase activity in a step of HCV replication cycle between translation of the viral genome and formation of functional replicase complexes . Data regarding primary translation where inferred from a surrogate model of translation based on a reporter luciferase gene ( Fig 5B ) . To address if indeed viral polyprotein is properly processed and inserted into detergent-resistant microdomains to form the characteristic membranous ultrastructures bearing the viral replicase , we used a replication-independent surrogate model of polyprotein expression . This system is based on a vector encoding the portion of the viral polyprotein corresponding to the replicase ( NS3-NS5B ) under the transcriptional control of the T7 polymerase and the translational control of encephalomyocarditis virus ( EMCV ) IRES ( pTM-NS3/5B ) [58] . Replication-independent polyprotein overexpression systems enable assessment of polyprotein processing as well as studying the formation of virus-derived membranous structures [16 , 59] . Control and lipin1-deficient cells were infected with a recombinant vaccinia virus expressing T7 RNA polymerase ( VacT7 ) and subsequently transfected with the plasmid pTM-NS3/5B to enable viral replicase expression . Sixteen hours post-transfection , cells were processed for Western-Blot using anti-NS3 . Accumulation of NS3 is comparable in control and both lipin1-deficient cells , underscoring the notion that lipin1 is not limiting for polyprotein translation and processing ( Fig 7A ) . Similarly , NS3 and NS5A expression and subcellular distribution was similar in all cell lines ( Fig 7B ) . These results suggest that there are no major differences in accumulation of viral proteins in lipin1-deficient cells and that a step downstream is affected in these cells . Transmission electron microscopy ( TEM ) of ultrathin cell sections of cells expressing HCV replicase components shows the expected accumulation of a mixture of characteristic double-membrane vesicles ( DMV ) as well as multiple membrane vesicles ( MMV ) ( Fig 7D , 7E and 7F ) that were not found in mock-transfected cells ( Fig 7C ) , as reported in previous studies using similar systems [16] . Individual vesicle diameter displays heterogeneous size distribution in which the predominant population is distributed between 125–150 nm ( median 150 nm; average ± SD: 167 ± 81 nm , n = 279 ) with larger vesicles being less predominant ( Fig 7G and S7 Fig ) . This size distribution is compatible with that observed similar replication-deficient systems and during HCV infection . Treatment of these cells with 100 nM daclatasvir ( DCTV ) , resulted in a strong reduction in the number of vesicles per section area ( S7B Fig ) , without significantly altering the size distribution of the remaining vesicles ( Fig 7G ) , as reported by Berger et al . [59] . Similarly , the diameter distribution of the vesicles found in lipin1-deficient cells was comparable to that in control cells ( Fig 7G ) . However , HCV-induced structures were significantly less abundant 16 hours post-transfection in lipin1-deficient cells than in controls cells ( S7C Fig ) . This reduced abundance is illustrated by a significant reduction in the fraction of cells displaying vesicular structures in lipin1-deficient cell cultures ( Fig 7H ) despite comparable transfection efficiency and viral protein expression levels , indicating that lipin1 may be required in a critical step leading to formation of the HCV-induced vesicular compartment . To validate the TEM results independently , we set out to establish a biochemical assay to evaluate replication-independent replicase complex formation . One of the characteristics of the HCV replicase complexes is that they are located in detergent-resistant membranes ( DRM ) [19 , 20] . In this sense , NS proteins are associated with replicase complexes that co-sediment with DRM markers such as caveolin-2 or sigma-1 receptor ( SIGMAR1 ) in low-density fractions in isopycnic gradient ultracentrifugation experiments [19 , 21] . Control and lipin1-deficient cells were infected with VacT7 and subsequently transfected with limiting doses of the plasmid pTM-NS3/5B [16] . Parallel samples were treated with DCTV ( 100 nM ) . Sixteen hours post-transfection , cell lysates were generated and subjected to equilibrium ultracentrifugation in 10–40% sucrose gradients . Gradient fractions were collected and subjected to Western-Blot analysis to determine the impact of lipin1 silencing on NS3 , SIGMAR1 and actin sedimentation profiles . Fig 8A shows how , as previously shown for replicon and JFH1-infected cells , NS3 can be detected in DRM fractions ( fractions 3–5 ) , as determined by the presence of SIGMAR1 in those fractions ( fractions 3–5; S8A Fig ) , although in this experimental system , unlike during viral infection [21] , most NS3 co-sediments together with solubilized proteins , as shown for actin ( fractions 9–12; S8A Fig ) . DRM-associated NS3 is reduced in DCTV-treated and lipin1-deficient cells , while total NS3 expression remains unchanged ( Fig 8A ) . Four independent experiments were performed in which relative-DRM associated NS3 , normalized to that found in solubilized fractions , was calculated for each experimental condition ( Fig 8B ) . Lipin1-deficient cells display a consistent and statistically significant reduction in the DRM-associated , but not total NS3 abundance ( Fig 8A and 8B; shLPIN1-1 and shLPIN1-2 ) , similar to that observed in control cells in the presence of DCTV ( Figs 8A and 8B; shControl+DCTV ) , consistent with the TEM data ( Figs 7 and S7 ) and the notion that NS3 in DRM fractions may reflect the abundance of replicase complexes formed in these cells . This reduction is not due to an overall reduction in cellular DRM abundance in lipin1-deficient cells , as lipin1-deficient cells display similar SIGMAR1 distribution pattern as the control cells ( S8 Fig ) . Overall , TEM data and DRM floatation assays strongly suggest that lipin1 is rate limiting for the generation of replicase complexes from fully processed polyprotein subunits .
Hepatitis C virus replication cycle is tightly linked to host cell lipid metabolism and interference with cellular lipid homeostasis contributes to viral pathogenesis [3] . One of the most evident consequences of this interference is the high prevalence of liver steatosis among chronically infected patients [13 , 60] . This clinical manifestation of the infection has been linked to , among others , chronic ER stress , mitochondrial dysfunction and metabolite depletion induced by HCV infection , which result in the activation of persistent homeostatic adaptation of the cellular lipid metabolism to permit cell survival , at the cost of pathogenic metabolic alterations [11 , 61–64] . Among the different regulatory networks that have been shown to be stimulated during HCV infection , PPARα [61] , PGC-1α [65] , HIF-1 [66] and SREBP [46 , 67] have also been shown to regulate transcription of LPIN1 mRNA [31 , 45 , 68 , 69] . Thus , it is likely that stimulation of one or several of these regulatory networks by HCV infection results in the LPIN1 mRNA transcriptional activation observed in this ( Fig 1A ) and other studies [40 , 41] . Importantly , prevention of LPIN1 mRNA accumulation with NAC ( Fig 1E ) did not significantly interfere with HCV RNA replication ( Fig 1F ) , suggesting that enhanced LPIN1 mRNA accumulation is not required for efficient HCV infection . We favor the hypothesis that ROS induced by HCV protein accumulation actively participates in LPIN1 induction , as treatment with the antioxidant NAC prevented LPIN1 mRNA accumulation , similar to what has been shown for other SREBP-regulated genes during HCV infection [46] . Accumulation of LPIN1 mRNA during HCV infection results in concomitant protein accumulation ( Fig 1G and 1H ) . However , post-translational mechanisms such as phosphorylation , acetylation or sumoylation regulate lipin1 protein stability , membrane association as well as subcellular localization thus influencing the activity of lipin1 as PA-phosphatase and as transcriptional coactivator [70] . Hence , it is difficult to predict the implications of lipin1 protein accumulation during HCV infection . Thus , future studies on the interference of HCV infection with cellular lipin1 functions will be required to determine its role in HCV-related pathogenesis , particularly in its contribution to steatosis . The data presented in this study provide evidence that lipin1 is rate limiting for HCV infection at an early step of the infection leading to formation of membranous HCV replicase complexes , downstream of viral polyprotein expression and processing . We provide evidence for reduced accumulation of viral RNA during single cycle infection experiments ( Fig 2D ) , that is reminiscent of a faulty initiation of viral replication , as suggested by reduced replication of a transfected subgenomic replicon in lipin1-deficient cells ( Fig 5D ) . Data obtained in replication-independent polyprotein expression models suggest that generation of the membranous compartment that contains functional replicase complexes is severely limited in lipin1-deficient cells , as suggested by a significant reduction of the fraction of cells where these structures could be visualized by TEM ( Figs 7 and S7 ) . This hypothesis is further supported by a significant reduction in DRM-associated NS proteins in lipin1-deficient cells ( Fig 8 ) , which may reflect limitations in the association of viral replicase subunits with cholesterol and sphingolipid-rich membranes in lipin1-deficient cells [19 , 20 , 71] . Four different shRNAs targeting LPIN1 mRNA decrease susceptibility to HCV infection proportionally to their ability to reduce total lipin1 protein accumulation ( Fig 4 ) . These results , together with the cDNA rescue experiments ( Fig 6 ) , strongly reduce the possibility of observing RNAi-associated off-target phenomena . Interestingly , homeostatic accumulation of lipin2 protein in lipin1-deficient cells ( Fig 2A and 2B ) is not sufficient to compensate for lipin1 loss to support efficient HCV infection . The notion that , despite being capable of mutually compensating basic liver functions [36 , 37] , lipin1 and lipin2 play non-redundant functions in the liver has previously been proposed [36 , 72 , 73] . Lipin1 is tightly regulated at many different levels and its activity accommodates PAP activity in response to different physiological situations such as fasting and insulin signaling [70] . Compelling evidence indicates that , while lipin1 and lipin2 cooperate to maintain liver lipid homeostasis , the two proteins differ in many aspects . For instance , lipin1 is transcriptionally induced by PGC-1α and it is also an inducible amplifier of this transcriptional network [31] , whereas lipin2 is not [36] . Lipin1 is sumoylated and sumoylation regulates its nuclear localization and function , whereas lipin2 sumoylation could not be demonstrated , despite the presence of a canonical sumoylation motif in its primary sequence [74] . Lipin1 enzymatic activity is blocked by mammalian target of rapamycin ( mTOR ) -dependent phosphorylation in response to different metabolic stimuli [30 , 75] , whereas lipin2 is constitutively active even when phosphorylated [76] . Thus , lipin2 is considered more as a constitutive phosphatidic acid phosphatase with lower specific activity than lipin1 [76] . In addition to these differential regulatory networks , it has been shown that in vitro PAP activity of purified lipin1 and lipin2 is differentially influenced by the composition of the substrates ( liposomes and lipid-detergent micelles ) as well as the pH at which the assay is performed [76] . This differential lipid substrate recognition may be reminiscent of the different preferential association with membranes of different subcellular compartments ( S1 Fig ) [73 , 76] . These differences suggest that , while lipin1 and lipin2 may share some common features , they are not functionally interchangeable , particularly not in the case of HCV infection [70] . Our data support the notion that lipin1 silencing has a strong impact on HCV infection without affecting basic cellular functions ( Fig 2B ) or significantly interfering with infection by an unrelated virus ( S3 Fig ) . Despite great efforts and different overexpression systems , functional rescue of lipin1 functions by wt lipin1 cDNA overexpression in lipin1-deficient cells only lead to a small but consistent rescue of virus infection efficiency , which was only observed when overexpressing wt lipin1 ( Figs 6 and S6 ) . Given the multiple transcriptional , post-transcriptional and post-translational regulation levels existing for lipin1 expression , it is conceivable that only a fraction of the overexpressed lipin1 is fully competent to sustain HCV infection . Moreover , high overexpression levels could only be achieved in lipin1-deficient cells ( Fig 6A ) , underscoring the notion that intracellular lipin1 levels are tightly regulated by the host . Nevertheless , overexpression of a mutant lipin1 lacking phosphatase activity ( DXDXT ) or a mutant inactive as transcriptional coactivator ( LXXIL ) were not capable of enhancing HCV infection in lipin1-deficient cells as compared with the wt lipin1 ( Fig 6B ) . These data reveal that lipin1 phosphatase activity is required for lipin1 to support HCV infection and suggest that , while transcriptional co-activation by lipin1 may be important , this function is not sufficient to support HCV replication . Lipins are important enzymes in the main pathway for de novo phospholipid biosynthesis by providing DAG derived from the glycerol-3-phosphate pathway to produce PC and PE through the Kennedy pathway [70] . Production of membranous replicase compartments likely requires de novo synthesis of PC and/or PE , which are major components of biological membranes that depend on DAG biosynthesis [29] . In fact , local PC biosynthesis is required for efficient replication of ( + ) RNA viruses and certain PC species accumulate in HCV-infected cells [26 , 28] . Although increased lipin2 accumulation may be sufficient to compensate lipin1 silencing at the whole-cell level [37] , it is possible that acute , local demand of de novo synthesized phospholipids is required at defined suborganellar compartments during early steps of HCV infection and that lipin1-deficiency shortens or alters the availability of different membrane components , demand that may not be satisfied by lipin2 , given the differential regulation[76] and subcellular localization of these two proteins ( S2 Fig ) . Remarkably , deletion of the yeast lipin homologs pah1/smp2 ( S . cerevisae ) or ned1 ( S . pombe ) gene , results in deregulated proliferation of the ER and nuclear envelope membranes [77 , 78] , with concomitant enhancement in ( + ) RNA virus replication [79 , 80] . The membranous alterations and elevation of total phospholipid content observed in pah1-deficient yeast have been ascribed to transcriptional activation of pah1-independent alternative phospholipid biosynthetic programs due to PA accumulation [77 , 80] . Our data in mammalian cells are more compatible with a shortage of phospholipid production , which may be at the basis of the reduced abundance viral membranous structure ( Figs 7 and 8 ) . This opposite outcomes of infection may derive from the fact that yeast use mainly PA ( lipin1 PAP substrate ) as a precursor for PC biosynthesis , while mammals mainly use DAG ( lipin1 PAP product ) as precursor for PC biosynthesis through the Kennedy pathway [77 , 80–82] . Moreover , in contrast to yeast and lower eukaryotes , which express only one lipin gene , three different lipin genes coordinate glycerolipid homeostasis in mammals [72] . Thus , interfering with expression of one of the members of the family may not be sufficient to observe the same effects observed when deleting pah1 , as transcriptional and posttranslational homeostatic compensations are in place in mammals , particularly between lipin1 and lipin2 in liver tissue [36 , 37] . In this sense , deletion of either lipin1 or lipin2 in mouse models results in a relatively balanced liver phospholipid content while simultaneous deletion of both lipins is embryonically lethal [37] . Accordingly , only minor alterations of ER membranes [83] and no significant alterations in total PC levels [36] have been reported in lipin1-deficient mouse liver . Given the fact that hCoV-229E is fully capable of replicating in these cells ( S3 Fig ) , it is unlikely that a general disruption of de novo phospholipid biosynthesis occurs in lipin1-deficient cells , particularly since hCoV-229E infection also induces profound ER membrane rearrangements required for replication [84] , some of which are structurally similar to DMVs observed during HCV infection [85] . Thus , we favor the hypothesis that a subcellular pool of glycerophospholipids is managed by lipin1 in Huh-7 cells and that lipin1 silencing perturbs local levels of PA and DAG , limiting local availability of precursors of structural components of virus-induced membranes . Alternatively , lipin1 deficiency may alter local amounts of important signaling molecules , in particular , that of its substrate ( PA ) or its product ( DAG ) . Deregulation of the local PA and DAG pools may cause important alterations for the host cell , as both metabolites are potent chemical messengers that regulate different aspects of cellular homeostasis [86–88] . Regarding PA conversion into DAG by lipins , it has been shown that pah1 ( yeast lipin1 homolog ) phosphatase activity is critical for transforming local pools of PA into DAG at the ER membrane to facilitate membrane fusion events mediated by SNARE complexes [89 , 90] . Mammalian lipin1 phosphatase activity is also critical for transforming local pools of PA that accumulate at the surface of mitochondria to promote mitochondrial fission [91] or at the surface of endolysosomes to facilitate autophagy [92] . Thus , lipin1 and probably other members of the lipin family modulate different aspects of intracellular membrane signaling . Given that the function of host factors known to be involved in functional HCV replicase biogenesis , like VAPA , VAPB and OSBP [11] are indirectly regulated by local PA/DAG pools [93] , it is tempting to propose that lipin1 silencing interferes with the function of one or several of these , or other yet uncharacterized cellular factors . In contrast to what has been reported for other host factors required for HCV replicase complex formation [58] , we did not find evidence of lipin1 protein relocalization during HCV infection ( S1 Fig ) . Thus , determining the precise mechanism by which lipin1 regulates HCV replicase formation is challenging , as association of lipin1 with different cell membranes is transient and highly regulated by posttranslational modifications [70] . Moreover , some of the known lipin1 cellular functions may be compensated by other lipins , particularly lipin2 in the liver . Nevertheless , our data clearly indicate that lipin1 participates at early stages of HCV replication and that the aforementioned homeostatic compensations by other lipins in regards to cellular metabolism may constitute an advantage when considering lipin1 as a host target for anti-HCV therapy .
HCV antiviral compounds 2´-c-methyladenosine ( 2mAde ) , sofosbuvir and daclatasvir were obtained from Boc Sciences ( NY , USA ) , Selleckchem ( Texas , USA ) and Medchem Express ( New Jersey , USA ) respectively and dissolved in DMSO to obtain 10mM stock solutions . N-acetylcysteine ( NAC ) and puromycin were obtained from Sigma-Aldrich ( Missouri , USA ) , dissolved in water to a final concentration of 0 . 5 M and 50 mg/ml respectively . Hydroxyzine pamoate was purchased from Sigma-Aldrich ( Missouri , USA ) and dissolved in DMSO to a final 10 mM concentration . Human hepatoma Huh-7 and derived subclones Huh-7 . 5 , Huh-7 . 5 . 1 ( clone 2 ) have been described [94–96] and were kindly provided by Dr . Chisari ( TSRI-La Jolla , CA ) . HEK-293T cells [97] were kindly provided by Dr . Ortin ( CNB-Madrid , Spain ) . Cell cultures were maintained subconfluent in Dulbecco´s Modified Eagle´s Medium ( Gibco ) supplemented with 10 mM HEPES ( Gibco ) , 100U/ml Penicillin/Streptomycin ( Gibco ) , 100μM non-essential amino acids ( Gibco ) and 10% fetal bovine serum ( Sigma-Aldrich ) . Lentiviral vectors expressing control and LPIN1-specific shRNAs were used to inoculate Huh-7 cells . Twenty-four hours later , cells were subjected to selection with 2 . 5μg/ml of puromycin to assess the lowest lentivirus dose capable of conferring puromycin resistance to 100% of the cell population . Selected cell populations were subsequently cultured in the presence of puromycin until LPIN1 silencing was ascertained by Western-Blot using anti-lipin1 antibodies , typically at day 6–7 post lentiviral transduction , time at which all experiments were performed in the absence of puromycin . Before execution of all the experiments shown in this study , lipin1 expression was assessed by Western-Blot . Cell viability was determined by a thiazolyl blue tetrazolium blue ( MTT ) formazan formation assay [47] . Control and lipin1-deficient cell lines ( 5 . 104 cells/well ) were plated onto 12-well plates and were inoculated with D183 virus at a MOI 10 FFU/cell . Samples of the cells and supernatants were collected 24 , 48 and 72 hours post-infection . For multiple cycle infection experiments ( MOI 0 . 01 ) , samples of the supernatants were collected at day 3 , 5 and 7 post-inoculation . Cells were split 1:3 in the multiple cycle infection experiments at days 3 and 5 to maintain the cultures subconfluent . Extracellular infectivity titers were determined by endpoint dilution and infection foci counting as previously described [99] . Intracellular HCV RNA was determined by reverse transcription and quantitative PCR ( RT-qPCR ) as previously described [99] . Total protein samples were prepared in Laemmli buffer and separated using polyacrylamide denaturing gel electrophoresis ( SDS-PAGE ) . Proteins were subsequently transferred onto PVDF membranes and incubated with 5% milk ( lipins ) or 3% BSA in PBS-0 . 25% Tween20 for one hour at room temperature ( RT ) . Primary antibodies against lipin1 ( clone B-12; Santa Cruz ) , lipin2 ( H-160; Santa Cruz ) , NS3 ( clone 2E3; Biofront ) , beta-actin ( ab8226; Abcam ) and tubulin ( clone AA2; Sigma-Aldrich ) were diluted in PBS-0 . 25% Tween20 and incubated for 1 hour ( four hours for lipins ) at RT . Membranes were subsequently washed for 20 minutes with PBS-0 . 25% Tween20 three times . Horseradish peroxidase-conjugated secondary antibodies were incubated for 1 hour at room temperature in 5% milk-PBS-0 . 25% Tween20 and subsequently washed three times for 20 minutes at room temperature . Protein bands were detected using enhanced chemoluminescence reactions and exposure to photographic films . Specific bands were quantitated using the ImageJ Software [100] on non-saturated , scanned films . Huh-7 cells were grown on glass coverslips and infected at high multiplicity ( MOI 10 ) with D183 virus . Forty-eight hours post infection cells were fixed for 20 minutes at RT with a 4% formaldehyde solution in PBS , washed twice with PBS and incubated with an incubation buffer ( 3% BSA; 0 . 3% Triton X100 in PBS ) for 1 hour . Antibodies were diluted in incubation buffer: rabbit anti-lipin1 antibody ( 1:50; Cell Signaling-Leiden , The Netherlands ) , rabbit anti-lipin2 antibody ( 1:100; H-160; Santa Cruz ) , mouse anti-dsRNA ( 1:200; J2 clone; Scicons ) , anti-NS3 ( 1:500; 2E3 clone; Biofront ) or anti-NS5A ( 1:200; 7E2 clone; Biofront ) . Primary antibodies were incubated with the cells for 1 hour ( 4 hours for lipin2 experiments ) time after which the cells were washed with PBS and subsequently incubated with a 1:500 dilution of a goat anti-mouse conjugated to Alexa 488 or Alexa 594 ( Invitrogen-Carlsbad , CA ) . Nuclei were stained with DAPI ( Life Technologies ) during the secondary antibody incubation using the manufacturer´s recommendations . Cells were washed with PBS and mounted on glass slides with Prolong ( Invitrogen-Carlsbad , CA ) . Confocal microscopy was performed with a Leica TCS SP8 laser scanning system ( Leica Microsystems ) . Images of 1024 × 1024 pixels at eight bit gray scale depth were acquired sequentially every 0 . 13–0 . 3 μm through a 63x/1 . 40 N . A . immersion oil lens , employing LAS AF v 2 . 6 . 0 software ( Leica Microsystems ) . Colocalization indexes were calculated using Jacop plugin for Image J [101] from a minimum of 10 regions of interest ( ROI ) . Images were processed using ImageJ , were medians of 1 pixel were obtained for the different channels , only for illustration , not for analysis . Color levels , brightness and contrast were manipulated for illustration using technical and biological controls as reference . Total RNA extraction was performed using the GTC extraction method [102] . Purified RNA ( 10–500 ng ) was subjected to RT-qPCR using random hexamers and a Reverse Transcription Kit ( Applied Biosystems ) . Quantitative PCR was performed using 2X Reaction Buffer from ( Applied Biosystems ) and specific oligonucleotides as previously described [99 , 103] . Standard curves were prepared by serial dilution of a known copy number of the corresponding amplicon cloned in a plasmid vector . Control and lipin1-deficient Huh-7 . 5 cells were inoculated with TNcc virus ( MOI 0 . 05 ) . Due to the relatively low propagation levels of the TNcc virus in this experimental setup , parallel cultures were infected and treated with 2´-C-methyladenosine ( 2mAde; 10μM ) to determine the levels of non-replicative , background HCV RNA . Cells were incubated for 72 hours at 37°C , time after which samples of the cells were collected to determine intracellular HCV RNA levels by RT-qPCR . To establish persistently infected cell cultures Huh-7 cells were inoculated at MOI 0 . 01 with JFH-1 HCV strain as previously described [24] . Cell cultures were maintained subconfluent for two weeks , time after which infection rates reach nearly 100% of the cells , as assessed by immunofluorescence microscopy . At this point cells were split and transduced with the corresponding lentiviral vectors in order to generate lipin1-deficient cell cultures as well as control cell lines . Once silencing had been verified by Western-blot , typically at day 6–8 post-transduction , cells were split and samples of the cells and supernatants were collected 24 hours later to determine intracellular HCV RNA levels by RT-qPCR and extracellular infectivity titers by end-point dilution and immunofluorescence microscopy . Infectious , spread-deficient HCV particles produced by trans-complementation ( HCVtcp ) have previously been described [52] . Briefly , Huh-7 . 5 . 1 clone 2 cells expressing core-E1 and E2-NS2 regions from JFH-1 by lentiviral transduction , were electroporated with a JFH-1 subgenomic dicistronic replicon bearing a firefly luciferase gene with reagents kindly provided by Dr . Ralf Bartenschlager ( U . of Heidelberg ) . Supernatants containing HCVtcp were collected 36 , 48 and 72 hours post-electroporation , pooled and assayed for viral infectivity . HCVtcp infection efficiency was determined by inoculating naïve Huh-7 cells with the electroporation supernatants and measuring luciferase activity 48 hours post-infection using a commercially available kit ( Promega ) . Retroviral particle production pseudotyped with different viral envelopes has previously been described [54 , 55] with the materials kindly provided by Dr . F . L . Cosset ( INSERM , Lyon ) . Control and lipin-deficient cell lines were inoculated with HCVpp and VSVpp and incubated for 48 hours , time at which total cell lysates were assayed for luciferase activity using a commercially available kit ( Promega ) . A selective HCV entry inhibitor , hydroxyzine pamoate ( HDX ) from Sigma-Aldrich ( Missouri , USA ) , was used as positive control of inhibition [55] . A plasmid containing the sequence corresponding to a subgenomic JFH-1 replicon bearing a firefly luciferase reporter gene was kindly provided by Dr . Ralf Bartenschlager ( U . of Heidelberg ) [52] . After digestion with the restriction enzyme MluI , the linearized plasmid was transcribed in vitro using a commercial kit ( Megascript T7; Ambion-Paisley , UK ) . The resulting products were digested with DNAse and precipitated with LiCl . Pelleted RNA was washed with 75% and 100% ethanol , and resuspended in nuclease-free water . In vitro transcribed RNA was transfected into the different cell lines together with a plasmid expressing Renilla luciferase under a minimal promoter ( pRL-null; Clontech-California , USA ) using Lipofectamine 2000 and the manufacturer´s recommendations ( Life Technologies- California , USA ) . Firefly and Renilla luciferase activities were measured in the sample using a commercial kit ( Dual Luciferase Assay System; Promega-Wisconsin , USA ) at different times post-transfection . Lipin1-deficient cells were generated by lentiviral transduction of shLPIN1-2 shRNA . At day 3 post-transduction , control and lipin1–deficient cell populations ( 5 X 104 cells/M24 well ) were transfected in suspension using lipofectamine 2000 with plasmids ( 800 ng/M12 well ) expressing wt lipin1beta isoform shLPIN1_2-resistant cDNA or DXDXT or LXXIL motif mutants [31] . Transfected cell cultures were incubated for 48 hours and subsequently inoculated at MOI 10 with D183 virus . Infection efficiency was determined by measuring extracellular infectivity titers 48 hours post-infection . Parallel cultures were used to determine relative wt and mutant protein expression efficiency by Western-Blot . Infectivity titers were measured as described above . The relative impact of cDNA expression was estimated by determining the ratio between the infectivity found in lipin1-deficient cells and the control cells transfected with the same plasmid . In order to average experiments with different raw infection efficiency , all the experiments were referenced to the ratio in the mock-transfected cells . Control and lipin1-deficient Huh-7 cells were inoculated with CoV-229E ( MOI 0 . 01 ) for 2 hours at 37°C . Cells were washed twice with warm PBS and replenished with DMEM-10%FCS . Extracellular infectivity titers were determined 48 hours post-infection by end-point dilution and fluorescence microscopy in Huh-7 cells . Huh-7 cells were inoculated at MOI 10 with a recombinant vaccinia virus expressing the T7 phage RNA polymerase ( VacT7 ) [104] . Two hours later , cells were transfected with the plasmid pTM-NS3/5B [16 , 58] ( kindly provided by Dr . Lohmann; U . of Heidelberg ) and Lipofectamine 2000 ( ThermoFisher-Massachussets , USA ) following the manufacturer´s recommendations in terms of total DNA per well ( typically 4μg per 35mm dishes with 7 . 5 X 105 cells/well ) and 50% of the recommended lipofectamine:DNA ratio . Transfected cells were cultured in the presence of the DNA replication inhibitor cytosine β-D-arabinofuranoside ( AraC; Sigma-Aldrich ) for 16 hours to prevent VacT7 replication [105] . When indicated , media was also supplemented with 100nM daclatasvir ( DCTV ) . Total cell extracts were used to determine viral protein accumulation by Western-blot using anti-NS3 antibody ( clone 2E3; Biofront ) and β-actin ( Abcam; ab8226 ) as loading control . For ultrastructural electron microscopy studies , control and lipin1-deficient cells expressing NS3-5B polyprotein ( see above ) were cultured on glass coverslips and fixed in situ after polyprotein expression with a mixture of 2% paraformaldehyde ( TAAB ) and 2 . 5% glutaraldehyde ( TAAB ) ( 1h at room temperature ) , post-fixed with 1% osmium tetroxide in PBS ( 45 min ) , treated with 1% aqueous uranyl acetate ( 45 min ) , dehydrated with increasing quantities of ethanol and embedded in epoxy resin 812 ( TAAB ) . Ultrathin , 70-nm-thick sections were cut in parallel to the monolayer , transferred to formvar-coated EM buttonhole grids and stained with aqueous uranyl acetate ( 10 min ) and lead citrate ( 3 min ) . Sections were visualized on a Jeol JEM 1200 EXII electron microscope ( operating at 100 kV ) . Quantitation of HCV-induced structures was performed as follows . To quantitate the differences in total vesicle abundance , TEM sections were visually inspected under the microscope for the presence/absence of vesicular structures . The number of positive cells and total number of cells were inserted in a 2 X 2 contingency table to determine the statistical significance of the differences between control and lipin1-deficient cells using two-tailed Fisher´s Exact Test or two-tailed Chi Square Test . In addition , we determined the frequency of HCV-induced vesicles in DCTV-treated control cells by dividing the number of structures per inspected area and calculating the average and SD of the frequencies found in the different images . The diameters of individual vesicles were determined manually using size-calibrated images and Image J software . Lipin1-deficient and control cells ( 7 . 5 X 105 cells ) were infected with VacT7 virus ( MOI 10 ) and transfected with limiting doses of pTM-NS3/5B plasmid ( typically 800 ng/ well ) , as higher plasmid doses may difficult observing the reported differences . Cells were lysed by adding 250 μl of TNE ( 50mM Tris-HCl pH 7 . 5 , NaCl 150 mM and EDTA 2 mM ) buffer containing 0 . 5% Triton X-114 and protease inhibitors ( Complete; Roche- Basel , SW ) . Lysates were incubated for 30 minutes on ice before clearing them by 10-minute centrifugation at 12 , 000 r . p . m . Clear supernatants were mixed 1:1 with 60% sucrose TNE solution . This mixture was applied on top of a 40% sucrose-TNE cushion and was overlaid with 20% and 10% sucrose-TNE until completing the discontinuous gradient . Gradients were centrifuged for 16 hours at 120 , 000 X g . Fourteen fractions were collected from the top and analyzed by SDS-PAGE and Western-Blot using antibodies against NS3 ( clone 2E3; Biofront ) , SIGMAR1 ( S-18; Santa-Cruz ) , caveolin-2 ( Epitomics; 3643–1 ) and beta actin as described previously [21] . NS3 signal was quantitated using ImageJ software and the fraction of DRM-associated NS3 was determined as the ratio of NS3 signal in fractions 3 , 4 and 5 to the total NS3 signal in the gradient . | Hepatitis C virus ( HCV ) infection is an important biomedical problem worldwide because it causes severe liver disease and cancer . Although immunological events are major players in HCV pathogenesis , interference with host cell metabolism contribute to HCV-associated pathologies . HCV utilizes resources of the cellular lipid metabolism to strongly modify subcellular compartments , using them as platforms for replication and infectious particle assembly . In particular , HCV induces the formation of a “membranous web” that hosts the viral machinery dedicated to the production of new copies of the viral genome . This lipid-rich structure provides an optimized platform for viral genome replication and hides new viral genomes from host´s antiviral surveillance . In this study , we have identified a cellular protein , lipin1 , involved in the production of a subset of cellular lipids , as a rate-limiting factor for HCV infection . Our results indicate that the enzymatic activity of lipin1 is required to build the membranous compartment dedicated to viral genome replication . Lipin1 is probably contributing to the formation of the viral replication machinery by locally providing certain lipids required for an optimal membranous environment . Based on these results , interfering with lipin1 capacity to modify lipids may therefore constitute a potential strategy to limit HCV infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"luciferase",
"vesicles",
"pathology",
"and",
"laboratory",
"medicine",
"hepacivirus",
"pathogens",
"enzymes",
"microbiology",
"enzymology",
"viruses",
"physiological",
"processes",
"rna",
"viruses",
"homeostasis",
"infectious",
"d... | 2018 | Host phosphatidic acid phosphatase lipin1 is rate limiting for functional hepatitis C virus replicase complex formation |
Alpha-Synuclein ( aSyn ) misfolding and aggregation is common in several neurodegenerative diseases , including Parkinson’s disease and dementia with Lewy bodies , which are known as synucleinopathies . Accumulating evidence suggests that secretion and cell-to-cell trafficking of pathological forms of aSyn may explain the typical patterns of disease progression . However , the molecular mechanisms controlling aSyn aggregation and spreading of pathology are still elusive . In order to obtain unbiased information about the molecular regulators of aSyn oligomerization , we performed a microscopy-based large-scale RNAi screen in living cells . Interestingly , we identified nine Rab GTPase and kinase genes that modulated aSyn aggregation , toxicity and levels . From those , Rab8b , Rab11a , Rab13 and Slp5 were able to promote the clearance of aSyn inclusions and rescue aSyn induced toxicity . Furthermore , we found that endocytic recycling and secretion of aSyn was enhanced upon Rab11a and Rab13 expression in cells accumulating aSyn inclusions . Overall , our study resulted in the identification of new molecular players involved in the aggregation , toxicity , and secretion of aSyn , opening novel avenues for our understanding of the molecular basis of synucleinopathies .
Aggregation of alpha-Synuclein ( aSyn ) is associated with a group of disorders known as synucleinopathies , that include Parkinson’s Disease ( PD ) , Dementia with Lewy Bodies and Multiple System Atrophy [1–3] . The common pathological hallmark among these disorders is the accumulation of aSyn in aggregates within neurons , nerve fibers or glial cells [4 , 5] . Moreover , multiplications [6] as well as point mutations ( A53T , A30P , E46K , H50Q , G51D and A53E ) are associated with familial forms of PD [7–13] . Recent findings suggest that aSyn can oligomerize into a tetramer under physiological conditions [14–18] , although this finding remains controversial [19–21] . In pathological conditions , it is widely established that aSyn can enter an amyloid pathway of aggregation , first as soluble , oligomeric species that , ultimately , can accumulate in insoluble aggregates [22] . The role of the large protein inclusions , such as Lewy bodies ( LBs ) , is unclear , but they may actually constitute a protective mechanism in neurons to neutralize and preclude the effects of more toxic aSyn intermediates [18 , 23–25] . Although the function of aSyn is still unclear , it interacts with lipid membranes [26 , 27] and seems to be involved in vesicle recycling and neurotransmitter release at the synapse [28 , 29] . Moreover , it is suggested that multimeric forms of aSyn physiologically bind to phospholipids at the synapse to chaperone SNARE-complex assembly required for neurotransmitter release , while monomeric forms are increased in disease and prone to aggregate [30–33] . Work in yeast and mammalian models suggests that aSyn-mediated cytotoxicity might be associated with alterations in vesicular trafficking , such as disruption of endoplasmic reticulum to Golgi trafficking [34 , 35] . This could be rescued by Rab ( Ras analog in brain ) GTPases , which play major roles in vesicular transport , tethering , docking and fusion [36] . Moreover , different studies have shown that dysregulation of Rab family members , such as Rab3a and Rab3b ( involved in exocytosis ) and Rab5 and Rab7 ( involved in the endocytic pathway ) , are associated with aSyn-induced toxicity in dopaminergic neurons of mammalian PD models [37 , 38] . Together with the Braak staging hypothesis , the finding that LB pathology might have spread in the brains of PD patients transplanted with embryonic nigral cells [39–41] , suggests that aSyn is able to spread in a prion-like manner in the brain . This theory has recently been supported by several studies in mouse models [42–45] . In neurons , secretion of aSyn follows a non-classical pathway [46] that is calcium-dependent and is up-regulated under stress conditions [47] . In addition , aSyn can be internalized through endocytosis or the classical clathrin-dependent pathway [48 , 49] . In LBs , aSyn is highly phosphorylated on Ser129 , contrasting with only 4% of the total protein phosphorylated at this residue in normal brain [50 , 51] . This suggests that phosphorylation might interfere with the aggregation process , although it is still unclear whether phosphorylation is a trigger or a consequence of aSyn aggregation . Thus , it is critical to understand whether modulating the activity of kinases and phosphatases can interfere with aSyn aggregation and/or toxicity . Here , we conducted an unbiased RNA interference ( RNAi ) screen to identify modulators of aSyn oligomerization , using the bimolecular fluorescence complementation ( BiFC ) assay as readout . We identified genes both encoding Rab GTPases and proteins involved in signal transduction . In addition to modifying oligomerization , the identified hits also altered aSyn toxicity and later stages of the aggregation pathway . Interestingly , we found that some of the trafficking-associated identified genes also modulated the secretion of different aSyn species . Altogether , our study brings novel insight into the molecular pathways involved in aSyn aggregation , toxicity and secretion , forming the basis for the testing of novel molecules with therapeutic potential in PD and other synucleinopathies .
In order to understand the contribution of different cellular pathways towards aSyn aggregation , we conducted an unbiased lentiviral vector-based RNAi screen in a cellular model of aSyn oligomerization , based upon a BiFC assay that we have previously described [18] . The screen comprised 1387 genes involved in trafficking and signal transduction-related pathways ( S1 Table and Fig 1A ) . We identified four genes encoding Rab proteins ( RAB8B , RAB11A , RAB13 and RAB39B ) and five genes encoding kinases or signal transduction proteins ( CAMK1 , DYRK2 , CC2D1A , CLK4 and SYTL5 ) that modulated aSyn oligomerization ( Fig 1B and 1D and S1A Fig ) . Interestingly , silencing of genes encoding kinases ( ALS2CR7 and STK32B ) , or phosphatases ( PSPH and PPP2R5E ) , did not affect aSyn oligomerization but altered the subcellular distribution of the oligomers . While silencing of ALS2CR7 or PSPH promoted aSyn aggregation , silencing of STK32B or PPP2R5E reduced the nuclear localization of aSyn oligomers ( S2 Fig ) . In the remainder of the study , we focused on the genes that modified aSyn oligomerization . Evidence of gene downregulation by the shRNAs was validated by qPCR ( S1B Fig ) and was confirmed by at least three different shRNAs targeting the same gene . Upon silencing of the Rab GTPase genes listed above , we observed a significant increase of aSyn-BiFC fluorescence intensity , similar to the effect of silencing CAMK1 and DYRK2 . Conversely , the silencing of CC2D1A , CLK4 and SYTL5 led to a significant reduction of aSyn oligomerization ( Fig 1B and 1D ) . To further characterize the role of the hits on aSyn oligomerization , we measured the levels of aSyn in aSyn-BiFC cells where each gene was stably silenced ( Fig 1C and 1D and S1C Fig ) . We found that aSyn protein levels were significantly increased upon silencing of RAB8B or CAMK1 . Silencing of RAB11A , RAB39B or DYRK2 did not change the levels of aSyn , but we found a decrease upon silencing of RAB13 , CC2D1A , CLK4 and SYTL5 . In order to correlate the levels of oligomerization with changes in the protein levels of aSyn , we compared the ratio between protein levels and fluorescence intensity ( Fig 1D ) . The increase in aSyn oligomerization upon silencing of RAB8B was accompanied by an increase in levels of aSyn , suggesting the effects might be related . On the other hand , in the case of CAMK1 silencing , the ratio of aSyn protein levels versus aSyn oligomers was <1 , suggesting that the increase in oligomerization was not simply due to an increase in the levels of aSyn . Moreover , the reduced oligomerization in cells silenced for CC2D1A , CLK4 or SYTL5 might be due to reduced levels of aSyn . In contrast , the increase in aSyn oligomerization upon RAB11A , RAB39B or DYRK2 silencing seems independent of the levels of aSyn . Interestingly , despite the observed increase in aSyn oligomerization upon RAB13 silencing , we found a reduction in aSyn levels relative to the control . To assess whether the silencing of the candidate genes was cytotoxic , we measured the release of lactate dehydrogenase ( LDH ) into the media as an indicator of cell-membrane integrity . We found that silencing of RAB8B , RAB13 or CLK4 resulted in an increase in cytotoxicity in cells with aSyn oligomers compared to cells with no aSyn ( Fig 1E and S1D Fig ) . Since aSyn oligomerization precedes the formation of larger inclusions , we next asked whether the hits identified in the screen would also modulate later stages of aSyn aggregation . To test this hypothesis , we used an established model of aSyn aggregation that results in the accumulation of LB-like inclusions in H4 cells [52–54] . We co-transfected a C-terminal modified version of aSyn ( aSynT ) and Synphilin-1 in cells stably transduced with lentiviruses encoding shRNAs targeting each of the identified hits , and then assessed inclusion formation using immunocytochemistry and fluorescence microscopy ( Fig 2A and 2B and S3A Fig ) . We quantified the percentage of cells according to the pattern of aSyn distribution , i . e . cells with no inclusions , cells with less than ten or cells with more than ten inclusions . In the conditions tested , more than 80% of control cells presented less than ten intracellular inclusions , 14% did not present inclusions , and less than 4% of the cells displayed ten or less inclusions . In contrast , the silencing of all the hits except CLK4 and DYRK2 resulted in an increase in the percentage of cells displaying aSyn inclusions ( Fig 2A and 2B and S3A Fig ) . Moreover , with the exception of RAB39 , the silencing of all hits caused an increase in the percentage of cells with more than 10 inclusions . This effect was stronger upon silencing of RAB8B ( approximately 60% of cells displaying more than ten inclusions ) , followed by SYTL5 ( 27% ) and RAB13 ( 15% ) . Together these data suggest that the hits can also modulate later steps of the aggregation process of aSyn . To assess whether the silencing of the different hits altered cytotoxicity in the aSyn aggregation model we measured cell membrane integrity , as described above . Only the silencing of RAB8B and RAB39B resulted in an increase in cytotoxicity ( Fig 2C and S3B Fig ) . Interestingly , the silencing of CLK4 resulted in the accumulation of inclusions with irregular shapes and silencing of SYTL5 resulted in the accumulation of elongated cells ( S3C Fig ) . In order to further evaluate whether trafficking indeed plays a role in aSyn aggregation , we silenced another traffic component involved in exocytosis ( RAB27A ) . We found that silencing of RAB27A increased aSyn oligomerization and did not affect aSyn levels or cytotoxicity ( S4 Fig and S2 Table ) . Furthermore , in the aggregation model , it increased the percentage of cells displaying aSyn inclusions , without affecting toxicity ( S4H and S4I Fig ) . Thus , the effects of RAB27A silencing are consistent with those observed with the hits selected in our screen . We and others have previously shown that aSyn can be secreted and affect multiple steps of membrane trafficking [47 , 55–57] . Therefore , we next investigated whether the trafficking of aSyn was affected upon silencing of four selected hits ( RAB8B , RAB11A , RAB13 or SYTL5 ) and RAB27A , involved in different steps of intracellular trafficking . To study cell-to-cell trafficking of aSyn , we used the aSyn-BiFC system with VENUS [58 , 59] . Firstly , RAB8B , RAB11A , RAB13 or SYTL5 were silenced in H4 cells and , 24 h later , cells were transfected with either VENUS1-aSyn or aSyn-VENUS2 plasmids , separately . 24 h later , an equal number of cells transfected with VENUS1-aSyn or aSyn-VENUS2 were mixed . 72 h later , mixed cultures were analyzed by flow cytometry and microscopy for the presence of fluorescence signal , which indicates bimolecular complementation of the VENUS fluorophore and , thus , cell-to-cell trafficking of aSyn ( Fig 3A ) . Scramble-mixed populations of VENUS1-aSyn and aSyn-VENUS2 were used to quantify cell-to-cell transfer of aSyn and used as a control to compare the effect of silencing of trafficking hits on aSyn intercellular transfer ( Fig 3B ) . Fluorescene of cells containing a single BiFC plasmid was identical to cells without any plasmid . In the mixed population of cells that were silenced for RAB8B , RAB13 and SYTL5 , we observed a significant number of fluorescent cells , indicating that transfer of aSyn between cells occurred . By knocking down RAB8B or RAB13 , we observed that the number of fluorescent cells increased to 4% , double of the control situation , while SYTL5 knock down resulted in fluorescence in 7% of cells in population , more than three times the fluorescence of scrambled cells , while silencing of RAB11A or RAB27A had no effect ( Fig 3B , 3C and 3D and S4E Fig ) . Next , we assessed whether overexpression of the hits selected would have the inverse effects to those observed upon silencing . For this , we expressed Rab8b , Rab11a , Rab13 and Slp5 in aSyn oligomerization model . In the case of Rab8b , Rab11a or Rab13 , we compared the effects of overexpressing wild type forms , constitutively active mutants ( Rab8b-Q67L , Rab11a-Q70L and Rab13-Q67L ) , or dominant-negative mutants ( Rab8b-T22N , Rab11a-S25N and Rab13-T22N ) . We found that overexpression of wild type forms or the constitutively active mutants of the Rab proteins significantly reduced aSyn oligomerization , while overexpression of Slp5 had no effect ( Fig 4A and 4B and S5 , S6 , S7 and S8 Figs ) . Overexpression of wild type or mutant forms of Rab13 reduced almost four times aSyn oligomerization . The dominant negative form of Rab8b , Rab8b-T22N , had a more attenuated effect on aSyn oligomerization compared to Rab8b-WT or Rab8b-Q67L . The dominant negative mutant of Rab11a did not change aSyn oligomerization . To investigate whether the endocytic recycling pathway was altered in the presence of aSyn oligomers , we monitored the distribution of fluorescently-labeled Transferrin ( Tf ) , which follows the endocytic recycling pathway and marks the endocytic recycling compartment ( ERC ) . As expected , Tf accumulated in ERC in control cells without aSyn . In contrast , in cells expressing aSyn-BiFC , Tf lost the preferential accumulation in the ERC , appearing at the periphery of the cell ( Fig 4A and S8B Fig ) . We measured the fluorescence intensity of Alexa-647-Tf and observed that the expression of wild type forms of Rab11a or Rab13 decreased the amount of the Tf within cells . In contrast , the dominant-negative forms of Rab11a and Rab13 showed a stronger Tf intracellular signal than control ( Fig 4C , S6 and S7 Figs ) . These results indicate that overexpression of the selected Rab hits restores endocytic recycling in cells accumulating aSyn oligomers . We also found that expression of Slp5 did not alter the intracellular Tf signal , while Rab8b increased it ( Fig 4C , S5A and S8A Figs ) . To further determine if aSyn oligomers were secreted from cells , we measured the levels of aSyn both in the media and in cell lysates . We observed no differences in the intracellular levels of aSyn . Moreover , although secretion was slightly increased upon overexpression of all hits , only Slp5 overexpression significantly increased aSyn secretion ( Fig 4E ) . Given that cells overexpressing each of the selected hits showed reduced cytotoxicity ( Fig 4D ) , we concluded that the release of aSyn was not due to cell death . Overall , our results suggest that the overexpression of Rab11a and Rab13 , but not Rab8b or Slp5 , promotes endocytic recycling of aSyn oligomers . Also , the Slp5-mediated release of aSyn oligomers does not appear to be due to increased trafficking of aSyn via the endocytic recycling pathway . To further explore the role of the hits identified on aSyn inclusion formation , we used the aSyn aggregation model and overexpressed Rab8b , Rab11a , Rab13 or Slp5 . We found that wild type and constitutively active forms of Rab8b , Rab11a and Rab13 significantly decreased the percentage of cells with aSyn inclusions , when compared with the dominant negative forms or control . A similar effect was verified with the overexpression of Slp5 ( Fig 5A and 5B ) . These results are consistent with those obtained upon silencing of the same genes ( Fig 2B ) . Interestingly , in cells lacking aSyn inclusions , Rab8b , Rab11a , Rab13 and Slp5 were normally distributed in the cell , as in the control situation ( Fig 4A , S5 , S6 , S7 and S8 Figs ) . However , we found that in cells with aSyn inclusions , these four proteins changed their subcellular localization and co-localized with the inclusions ( Fig 5A , S5 , S6 , S7 and S8 Figs ) . Together , these results suggest that Rab8b , Rab11a , Rab13 or Slp5 can modulate aSyn aggregation and can be recruited into inclusions . To investigate whether endocytic recycling was altered in the presence of aSyn inclusions , we monitored this process using Alexa-647-labeled Tf . Normally , Tf accumulates in the ERC . In our experiments , we found that Slp5 and wild type or constitutively active mutant forms of Rab8b , Rab11 and Rab13 decreased the intracellular fluorescence signal of Tf . In contrast , cells expressing the dominant-negative forms of these Rabs displayed similar fluorescence intensity to the controls ( Fig 5C ) . These results indicated that recycling through the endocytic recycling pathway was compromised in cells with aSyn inclusions , as more Tf accumulated inside the cells , and that overexpression of wild type and constitutively-active forms of Rab8b , Rab11a , and Rab13 , and Slp5 , could rescue this defect . To determine whether Rab8b , Rab11a , Rab13 or Slp5 played a role in aSyn secretion when this protein is aggregated , we measured the levels of aSyn in conditioned media . We found that aSyn secretion was not changed by Slp5 ( Fig 5E ) . However , wild type forms of the Rabs increased aSyn secretion . To further confirm that the increased levels of extracellular aSyn were not due to increased cell death , we measured the release of LDH , and found that all the hits tested were protective ( Fig 3D ) . Altogether these results show that overexpression of Rab8b , Rab11a , Rab13 or Slp5 reduces aSyn aggregation , and that the subcellular localization of these proteins is altered in the presence of aSyn inclusions , since they all co-localize . Overexpression of these Rabs also promotes aSyn secretion , which can occur through the endocytic recycling pathway . Thus , the increased aSyn secretion upon Rab8b , Rab11a and Rab13 overexpression can explain the decrease of aSyn inclusions within the cells , as this effect is not related with an increase in cell death .
Increasing evidence suggests that pre-fibrillar , oligomeric forms of aSyn are the toxic species that lead to pathology [23 , 24 , 60] . The main objective of this study was to identify regulators of aSyn oligomerization , an early step of the aggregation process that precedes the formation of larger protein assemblies typically referred to as protein aggregates . To do this , we performed an RNAi screen targeting 76 membrane trafficking and 1311 phosphotransferase genes using a cell model of aSyn oligomerization . Interestingly , given the uniqueness of our approach , based on live-cell imaging of aSyn oligomers , the screen also enabled us to identify genes that did not alter aSyn oligomerization but modified the subcellular distribution of the oligomeric species . With respect to the primary goal of the screen , we identified four genes encoding Rab proteins ( RAB8B , RAB11A , RAB13 and RAB39B ) and five genes encoding phosphotransferase proteins ( CAMK1 , DYRK2 , CC2D1A , CLK4 and SYTL5 ) that modulated both oligomerization and aggregation ( except DYRK2 ) of aSyn . Regarding the effect of the hits on aSyn oligomerization and protein levels , we identified hits that increased both parameters , as in the case of RAB8B and CAMK1 . Interestingly , silencing of RAB8B , but not CAMK1 , is toxic in the presence of aSyn oligomers . The fact that RAB8B silencing is also toxic in the presence of aSyn inclusions suggests this is a relevant modulator at two different stages of aSyn aggregation process . Camk1 is a Calmodulin-dependent kinase that plays a role in axonal growth [61] . Until now Camk1 activity had not been associated with aSyn . However , Camk2 seems to play an essential role in the redistribution of aSyn during neurotransmitter release at the synapse [62] . Moreover , Camk2 forms a complex with aSyn and seems to regulate its oligomerization status [63] . If Camk1 and Camk2 share some functionality , this might explain the stronger downstream effect of CAMK1 silencing , with a more pronounced effect on oligomerization rather than on the levels of aSyn . We also identified hits that decreased both aSyn oligomerization and protein levels; for example , the silencing of CC2D1A , CLK4 and SYTL5 decreased oligomerization probably because they reduce the levels of aSyn . Silencing of CLK4 , but not of CC2D1A and SYTL5 , is toxic to the cells . Thus , we can speculate that , at least for CC2D1A and SYTL5 , the effects observed are not due to cytotoxicity , as membrane integrity is preserved , and these hits can be further tested as candidate therapeutic modulators in synucleinopathies . Moreover , we also found hits that had a direct effect on oligomerization without changing the levels of aSyn; silencing of RAB11A , RAB39B and DYRK2 increased oligomerization without affecting the levels of aSyn . Silencing of RAB39B was toxic in the aSyn aggregation model but not in the oligomerization model . Thus , from a therapeutic perspective , hits that modify oligomerization/aggregation without altering the levels of aSyn are of great interest . Finally , we found one hit ( RAB13 ) that increased oligomerization while reducing the levels of aSyn . When overexpressed , this gene was protective against toxicity , reduced oligomerization and did not alter the levels of aSyn . In total , our findings reveal an intricate connection between aSyn aggregation , toxicity and levels that will need to be further investigated in future studies . Four out of the nine modifiers of aSyn oligomerization and aggregation were Rab small GTPases . Rab GTPases are a family of more than 60 members in humans that are master regulators of intracellular formation of vesicles , motility and release , thereby playing a key role in neuronal trafficking ( reviewed in [64 , 65] ) . Rab GTPases switch between GDP-bound ( inactive ) and GTP-bound ( active ) states to regulate downstream cellular functions . It is the activation by a guanine-nucleotide exchange factor ( GEF ) that converts an inactive Rab into the active GTP-bound form [66] . Active Rab GTPases can bind Rab effectors , which control the spatiotemporal regulation of Rab steps within cells . Given the importance of Rab GTPases and their effectors in the regulation of membrane trafficking , several human disorders have been associated with their dysfunction , in particular diseases affecting neuronal cells ( reviewed in [67] ) . Although the hits identified fall into several different functional classes , all but SYTL5 affect neuronal trafficking [61 , 68–74] . We focused on hits involved in secretion , as this process might underlie the spreading and transmission of aSyn pathology in the brain [39] . Thus , we further characterized the effect of Rab8b , Rab11a , Rab13 and Slp5 on aSyn aggregation . Rab8 is associated with actin and microtubule cell reorganization and polarized trafficking to dynamic cell surface structures [71] . Interestingly , Rab8 is able to reconstitute Golgi morphology in cellular models of PD [75] and , in addition , we recently reported that aSyn interacts with Rab8a . Moreover , we also found that Rab8 rescues the aSyn-dependent loss of dopaminergic neurons in Drosophila [76] . Here , we showed that silencing of Rab8b increased the accumulation of oligomeric or aggregated species of aSyn and was toxic to cells , while overexpression of Rab8b reverted those effects . Rab11a is ubiquitously expressed with preferential localization to ERC . Defective trafficking of Rab11 from the ERC has been implicated in AD , HD and PD [72 , 77 , 78] . Rab11a is involved in the process of exocytosis of aSyn via recycling endosome [78] . Silencing of Rab11a increased accumulation of oligomeric or aggregated aSyn , while overexpression of Rab11a was protective and reverted the oligomerization and aggregation of aSyn , as we previously reported in independent studies [55 , 79] . Rab13 mediates trafficking between the trans-Golgi network and recycling endosomes [80] and it is associated with neuronal plasticity , neurite outgrowth , cell migration and regulation of tight junctions . Interestingly , we found that Rab13 silencing was toxic to cells with aSyn oligomers but not to cells with aSyn inclusions . On the other hand , overexpression of Rab13 reduced aSyn toxicity in both cell models . We also found that Rab11a and Rab13 decrease the amount of intracellular Tf both in the models of aSyn oligomerization and aggregation . Moreover , secretion of aSyn is also differentially affected depending on the cell model , suggesting that the endocytic recycling pathway might be used to clear aSyn aggregates , possibly through secretion . Slp5 is a calcium-dependent protein that belongs to the Synaptotagmin-like protein family . Proteins from this family contain tandem C2 domains that bind phospholipids and proteins associated with the plasma membrane . Slp5 interacts with GTP-bound Rab27a , Rab3a and Rab6a , but not with Rab8 or Rab11a [81] . As vesicles an effector of Rab27a , Slp5 mediates the tethering/docking of Rab27a-positive vesicles to the plasma membrane [82] . Moreover , it can modulate the Rab27a-mediated transport of Cystic Fibrosis Transmembrane conductance Regulator ( CFTR ) to the membrane [83] . Slp5 can be found in the brain and in other tissues , and was shown to promote exocytosis of dense core in PC12 cells [84] . On other hand , SYTL5 was also identified in another RNAi screen as player in chemotaxis [85] , being potentially important in the generation of inflammatory responses . To the best of our knowledge , Slp5 had not been previously associated with brain disorders . Here , we showed that Slp5 silencing decreases aSyn oligomerization and increases the number of aSyn inclusions per cell . Moreover , the recycling endocytic pathway is active upon Slp5 overexpression in cells presenting aSyn inclusions , but the levels of aSyn secretion are not altered ( Fig 6 and S2 Table ) . Interestingly , RAB27A was not identified in our primary RNAi screen ( S1 Table ) . We hypothesize that this might be due to redundancy between Rab27 isoforms and also because this GTPase has at least eleven different effectors [82] that may mask the effects of RNAi-mediated silencing . Although silencing of RAB27A does not affect aSyn oligomerization or secretion , it promotes aggregation ( S4 Fig ) . This further supports the hypothesis that trafficking components are key players in aSyn homeostasis . Remarkably , we found that overexpression of Rab8 , Rab11a , Rab13 and Slp5 significantly increases the percentage of cells without inclusions to 50–75% . Although future studies will be important to further clarify the precise molecular mechanisms involved , it is possible that these proteins reduce aSyn aggregation by affecting its release . A second important observation showed that , in the remaining cells displaying aSyn inclusions , Rab8 , Rab11a , Rab13 and Slp5 localized in the inclusions together with aSyn . Therefore , and as previously suggested , the sequestration of the Rabs in the inclusions may affect their function [86] . Additional studies will be necessary to clarify the relationship between the genes we identified and their cellular functions , especially those related to the endocytic recycling pathway . For example , in addition to Rab11a , Rab8 also assists the transport of Tf within cells and colocalizes with Slp1 and Slp4 [87 , 88] . In fact , in cells with no inclusions , transferrin ( Tf ) labels the endocytic recycling compartment ( ERC ) . In cells with inclusions , the ERC location of Tf was maintained but the signal was weaker . However , if one of the selected traffic hits is overexpressed in cells with few aSyn inclusions , Tf can be seen at i ) the ERC ( as the traffic hit ) and ii ) in aSyn inclusions , co-localizing with the hit . If the number of inclusions is higher , Tf signal loses the ERC location ( as the overexpressed hit ) and is redistributed in inclusions . This sequential difference in Tf location reflects the possible redistribution of trafficking players and , thus , represents an alteration in the endocytic recycling machinery promoted by aSyn aggregation . This effect can synergistically be explained by a first cellular attempt to flow the excess of aSyn within the cell specifically when it is aggregated . The increase in aSyn secretion in the aSyn aggregation model ( upon expression of all selected hits ) suggests that endocytic recycling is being activated as less Tf signal is detected . However , when there are more inclusions in the cells there is a higher chance that the hits will be sequestered in the inclusions . aSyn is known to be secreted under physiological conditions , possibly via unconventional exocytosis , as it lacks an ER-targeting signal peptide . Although the precise mechanisms involved are still unclear , multiple secretory pathways have been described [56] . However , it was demonstrated that pathological and aggregated species of aSyn can also be secreted [89] . This suggests that misfolded and aggregated aSyn is a key agent for the propagation of PD pathology by a prion-like mechanism [90 , 91] . In this context , aggregation can be viewed as a protective mechanism , as it could arrest the toxic species that would otherwise be secreted . This is consistent with several observations by different groups , including our own , that protein aggregates ( or at least some types of aggregates ) appear to be less toxic than smaller , oligomeric species . It is possible that , after a certain threshold , the cumulative failure of cellular quality control systems , together with the secretion of aSyn , disrupts the initial cellular attempt to contain pathological aSyn species . As a result , toxic species of aSyn can spread in a prion-like manner . Since we found that silencing of RAB8B , RAB13 or SYTL5 augmented aSyn cell-to-cell transfer ( Fig 3B and 3C ) , these genes emerge as potential modifiers of the spreading of aSyn pathology . Transfecting independent cells with VENUS1-aSyn or aSyn-VENUS2 plasmids for 24h , and then mixing equal numbers of each cell population , enables the study of aSyn cell-to-cell transmission using the split-VENUS BiFC system . We observed a two to threefold increase in aSyn cell-to-cell transfer upon silencing of RAB8B , RAB13 or SYTL5 , while silencing of RAB11A or RAB27A had no effect when compared to scramble-infected cells . This is further supported by a stronger signal in the immunoblot of at least two of three shRNAs used to silence RAB8B , RAB13 or SYTL5 . Linking these results with the oligomerization state of aSyn , silencing of both RAB8B and RAB13 promoted aSyn oligomerization , probably because the balance between the entrance and exit of aSyn is increased in those cells . On the other hand , silencing of SYTL5 decreased oligomerization and increased cell-to-cell transfer of aSyn . Recently , aSyn was shown to be secreted by exosomes [47 , 58 , 92] . Given that Slp5 is an effector protein of Rab27a involved in exosome-mediated secretion [93] and that , upon silencing , intercellular trafficking of aSyn is increased , this confirms that aSyn transmission also occurs by pathways independent of exosomes , as previously reported [55 , 58 , 94] . Silencing of RAB11A did not affect the cell-to-cell trafficking of aSyn ( Fig 3 ) . Hence , the increase in aSyn dimerization induced by silencing of RAB11A might reflect the impairment of the endocytic recycling pathway , one of the routes through which aSyn oligomers can be released [58] . In this study we showed that traffic-related modulators of aSyn oligomerization can reverse toxicity and reduce aggregation by increasing secretion of aSyn . Altogether , the genetic screen we performed serves not only as a proof of concept for the identification of genetic modifiers of aSyn aggregation , but provides novel insight into the molecular underpinnings of PD and other synucleinopathies . Ultimately , future validation in animal models will establish which of the genes identified holds greater potential as targets for therapeutic intervention .
Human H4 neuroglial cells ( HTB-148—ATCC , Manassas , VA , USA ) were maintained in Opti-MEM medium supplemented with 10% of fetal bovine serum ( FBS ) ( Life Technologies ) , and incubated at 37°C , 5% CO2 . Cells were plated 24 h prior to transfection until 80% of confluence . Transfections were performed using Fugene 6 ( Promega ) according to the manufacturer’s instructions . aSyn-BiFC stable cell lines were obtained by transfecting H4 cells with GN-link-aSyn and aSyn-GC constructs [18] and maintained with G418 and Hygromycin B antibiotics ( both at 100 μg/ml , InvivoGen ) in Opti-MEM media with 10% FBS . Green fluorescence protein ( GFP ) reconstitution assay was made as previously described [18] and brightest cells were viably separated using a fluorescence activated cell sorter . After growth of these selected cells , sorting and regrowth was repeated until we obtained a homogenously fluorescent aSyn-BiFC cell line . To generate stable cell lines with hits silencing , H4 cells or aSyn-BiFC stable cells were seeded on 10 cm plates 24 h prior to infection . Cells were infected as described [95] with lentiviruses . Infected cells were selected with 5 μg/ml puromycin antibiotic ( Invivogen ) 48 h later and maintained with antibiotic in media . 1 . 5x106 aSyn-BiFC cells were plated in 10 cm plates and infected with selected hits as described [95] . 48 h post-infection , total RNA was extracted from cell lysates with Trizol reagent ( Invitrogen ) in accordance with the manufacturer’s instruction . 1μg of RNA was reverse transcribed into cDNA using Superscript First Strand Synthesis Kit ( Invitrogen ) . PCR amplification was performed by using 2μl of cDNA with SYBR Green master mix ( Sigma-Aldrich ) . Primers used for real time PCR were chosen using Primer 3 , Net Primer and BLAST software to ensure specificity . RT-PCR primers here used were: for RAB8B , forward 5’-ATGAGGCTGGAATCCACTTG , reverse 5’-ATGAGGCTGGAATCCACTTG; for RAB11A , forward 5’-CATGTTCCACCAACCACTGA , reverse 5’- GTCATTCGGGACAAGTGGAT; for RAB13 , forward 5’-CAAGACAATAACTACTGCCTACTACCG , reverse 5’-AAGCCTCATCCACATTCATACTG; for RAB39B , forward 5’-AGTTCCGGCTCATTGTCATC , reverse 5’- ATCTGGAGCTTGATGCGTTT; for CAMK1 , forward 5’-AAGAGCAAGTGGAAGCAAGC , reverse 5’-AGTGAGGAGTGGTAGGGAAGC; for DYRK2 , forward 5’-CCAGAAGTAGCAGCAGGACC , reverse 5’-CCCACTGTTGTAAGCCCATT; for CC2D1A , forward 5’-ATCTGGATGTCTTTGTTCGGTT , reverse 5’-TTGATGCCCTTGGTCTGG; for CLK4 , forward 5’-GGTTGGTCTCAGCCTTGTG , reverse 5’-TGTGTTGTGGTATGGGTCCTAA; for SYTL5 , forward 5’-AGCAAAGCCACCAAGCAC , reverse 5’-CTGAGAGTCCATCCAATCCAC; for ACTB ( beta-actin , endogenous control ) , forward 5’-GGACTTCGAGCAAGAGATGG , reverse 5’-AGCACTGTGTTGGCGTACAG . 1 . 5x106 H4 stable cells with hit silencing ( RAB8B , RAB11A , RAB13 or SYTL5 ) per dish were plated and transfected in 10 cm dishes . 24 h later , cells were transfected cells with VENUS1-aSyn or aSyn-VENUS2 vectors independently [58] . 24 h later , 0 . 5x106 of transfected cells with VENUS1-aSyn and aSyn-VENUS2 constructs were mixed . 72 h after , trypsin was added to each plate and neutralized with media ( Opti-MEM+10% FBS ) . Cell suspension was centrifuged at 1100 rpm for 10 min , the supernatant aspirated and the pellet reconstituted in phosphate buffered saline ( PBS ) . The resulting supernatant was filtered with cell strainer caps into polypropylene tubes ( both from BD Biosciences ) . VENUS Fluorescence was measured on a BD LSRFortessa ( BD Biosciences ) and detected also at Axiovert200M microscope ( Carl Zeiss MicroImaging ) . In order to generate pcDNA ENTR BP myc-mCherry-C2 mouse Rab8b , pcDNA ENTR BP V5-mCherry-C2 mouse Rab11a and pcDNA ENTR BP V5-mCherry-C2 mouse Rab13 , pcDNA ENTR BP myc-mCherry-C2 or pcDNA ENTR BP V5-mCherry-C2 , mammalian expression vectors were used . These mammalian expression vectors were previously generated by inserting a polylinker containing several restriction sites into pcDNA6 . 2GW/Em-GFP , a mammalian expression Gateway ( Invitrogen ) previously digested with DraI/XhoI followed by insertion of myc-mCherry-C2 or V5-mCherry-C2 , previously synthetized into NheI/BamHI . Rab8b , Rab11a and Rab13 mouse coding sequence and part of 3’ UTR were produced by RT-PCR amplification using total RNA isolated from at-T20 cell line as a template , digested with EcoRI/SalI and cloned with the same restriction enzymes into the mammalian expression vectors . The primers here used were: for Rab8b , forward 5‘-AGTGAATTCATGGCGAAGACGTACGATTATCTGTTC , reverse 5‘-GACCGTCGACTCACAGGAGACTGCACCGGAAGAA; for Rab11a , forward 5‘-TGAGGAATTCATGGGCACCCGCGACGACGAGTA , reverse 5‘-AATAGTCGACCATGCTGGTTGCTGAATATGGTG; for Rab13 , forward CCCGGCGCCCCCAGTGGAATTCATGGCCAAAG , reverse 5‘-GTGCGTCGACAGCCTCTCAGGACCCTAACC . Rab8b ( Q67L and T22N ) , Rab11a ( Q70L and S25N ) and Rab13 ( Q67L and T22N ) mutants were generated by PCR mutagenesis and using the following primers: for Rab8b-Q67L , forward 5‘-GGCCTGGAAAGATTCCGAACAATTACG , reverse 5‘-CGCCGTGTCCCATATCTGAAGTTTAAT; for Rab8b-T22N , forward 5‘-GACTCCGGCGTTGGCAAGAACTGC , reverse 5‘-GCCGATGAGCAGCAGCTTGAACAGATA; for Rab11a-Q70L , forward 5‘-GGGCTGGAGCGGTACAGGGCTATAAC , reverse 5‘-TGCTGTGTCCCATATCTGTGCCTTTAT; for Rab11a-S25N , forward 5‘-GGTGTTGGAAAGAATAACCTCCTGTCT , reverse 5‘-AGAATCTCCAATAAGGACAACTTTA; for Rab13-Q67L , forward 5‘-GGCCTAGAACGATTCAAGACAATAACT , reverse 5‘-AGCCGTGTCCCACACTTGCAGTTTGAT; for Rab13-T22N , forward 5‘-TCGGGGGTGGGCAAGAATTGT , reverse 5‘-GTCCCCGATGAGCAGCAACTTGAAGAG . In order to generate pENTR V5-C2 mouse Sytl5 and pENTR GFP-C2 mouse Sytl5 , Gateway mammalian expression vectors previously described were used [96] . Sytl5 mouse coding sequence was produced by RT-PCR amplification of total RNA isolated from mouse brain as template ( using the primers forward 5‘-TCGAAGCTTCGGATCCATGTCTAAGAACTCAGAGTTCATC and reverse 5‘-CTAGTCGACTCAGAGCCTACATTTCGCCATGCT ) , digested with HindIII/SalI and cloned into pENTR GFP-C2 with the same restriction enzymes . For overexpression assays , H4 cells or aSyn-BiFC stable cells were seeded 24 h prior to transfection ( on 35 mm glass bottom ibi-treated imaging dishes , ibidi GmbH ) for immunocytochemistry and cell imaging or on 6 well plates for immunoblotting or cytotoxicity assays ) . Cells were transfected with wild type , constitutively active and dominant negative mutants of pcDNA ENTR BP myc-mCherry-C2-RAB8B ( RAB8B-WT , RAB8B-Q67L , RAB8B-T22N ) , pcDNA ENTR BP myc-mCherry-C2-RAB11A ( RAB11A-WT , RAB11A-Q70L , RAB11A-S25N ) , pcDNA ENTR BP V5-mCherry-C2-RAB13 ( RAB13-WT , RAB13-Q67L , RAB13-T22N ) , pENTR V5-C2-SYLT5 constructs or empty vector ( plasmids were a kind gift of Dr . José S . Ramalho , Universidade Nova de Lisboa , Portugal ) . 48 h post-transfection , cells were washed with PBS and incubated with media with no serum for 1 h . 50 μg/ml of Alexa-633-Tf ( Life Technologies ) were added for 30 min . Cells were then washed with PBS and fixed with 4% paraformaldehyde for 10 min and washed again . Immunocytochemistry was performed only for SYTL5 construct , using primary antibody ( mouse anti-V5 , Cell Signaling ) and secondary antibody ( goat anti-mouse IgG-Alexa568 , Life Technologies ) . Nuclear staining was made using 1 μg/ml of Hoescht 33342 dye ( Sigma Aldrich ) for 2 min . Cells were washed and imaged in PBS . Cells were imaged using a Zeiss LSM 710 microscope with a 63× 1 . 4 NA oil immersion objective . Fluorescence emission was detected for Hoechst , GFP , mCherry and Far red: excitation at 405 nm ( band pass 420–480 ) , 488 nm ( band pass 505–550 ) , 561 nm ( band pass 575–615 ) , 633 nm ( 647–754 ) . Pinhole was at 160 μm for all channels and 2–10% of transmission was used . For loss of function assays , stable H4 cells for hit silencing were seeded in 35 mm glass bottom imaging dishes ( ibidi GmbH ) 24 h prior to transfection and were co-transfected with aSynT and Synphilin-1-V5 as previously described [55] and subjected to immunocytochemistry 48 h later . For overexpression assays , triple transfections were with aSynT , Synphilin-1-V5 and mCherry-Rabs or GFP-Slp5 plasmids and Tf incubation was made as described for aSyn-BiFC cells . Then , cells were permeabilized with 0 . 5% Triton X-100 in PBS for 20 min at RT , blocked for 1 h at RT with 1% normal goat serum in 0 . 1% Triton X-100 in PBS , incubated with primary antibody against aSyn ( mouse anti-aSyn 1:1000; BD Biosciences ) and Synphilin-1-V5 ( only for loss of function assays; mouse anti-V5 , 1:1000 , Cell Signaling ) at 4°C overnight followed by secondary antibody incubation ( 1:1000 , goat anti-mouse IgG-Alexa488 for aSynT ( or igG-Alexa 568 for aSynT when co-transfected with GFP-SYTL5 ) and goat anti-mouse IgG-Alexa568 ( Life Technologies ) for Synphilin-1-V5 ( only for loss of function assays ) , for 2 h at RT . Nuclear staining was made using 1 μg/ml of Hoechst 33342 dye ( Sigma Aldrich ) for 2 min . Cells were washed and imaged in PBS . Cells were then subjected to microscopy analysis using Zeiss Axiovert 200M for loss of function assays or Zeiss 710 confocal microscope for overexpression assays using the same settings used for dimerization model . The proportion of cells displaying aSyn-positive intracellular inclusions in the aSyn-positive cell population was determined by counting at least 100 cells in each condition . Moreover , for overexpression assays , Alexa-546-Tf fluorescence intensity was also determined using ImageJ . Total protein extracts were obtained 48 h post-transfection using standard procedures . Briefly , cells were washed twice in PBS and lysed in NP40 buffer ( glycerol 10% , Hepes 20mM pH7 . 9 , KCl 10mM , EDTA 1 mM , NP40 0 . 2% , DTT 1mM ) containing protease and phosphatase inhibitors cocktail ( 1 tablet/10ml , Roche Diagnostics ) . Cell debris was spun down at 2 , 500 rpm for 10 min and supernatant were sonicated at 10mA for 15 s ( Soniprep 150 ) . Protein concentration was determined using the BCA protein assay ( Thermo Scientific ) and 20 μg of protein lysates were resolved in 12% SDS-PAGE . Resolved proteins were transferred to nitrocellulose membranes . After quick washing in TBS-T ( Tris buffered saline and 0 . 1% Tween 20 ) , membranes were blocked either in 5% non-fat dry milk in TBS-T or in 5% BSA in TBS-T for 1 h and then incubated with primary antibodies in 5% BSA in TBS overnight at 4°C . The primary antibodies used were mouse anti-aSyn , 1:1 , 000 , BD Transduction; mouse anti-pS129aSyn 1:1 , 000 Wako Chemicals USA; mouse anti-beta actin , 1:4 , 000 , Sigma; mouse anti-V5 , 1:1 , 000 , Cell Signaling; mouse anti-myc , 1:1 , 000 , Santa Cruz . The membrane was then washed three times for 10 min each in TBS-T at room temperature and probed with IgG horseradish peroxidase-conjugated ( HRP ) anti-mouse secondary antibody ( 1:10 , 000 ) for 1 h at room temperature . The membrane was washed again four times for 15 min each with TBS-T and the signal was detected with an ECL chemiluminescence kit ( Millipore Immobilon Western Chemiluminescent HRP Substrate ) . 1 . 5x105 cells were plated in 6 well plates one day prior to transfection ( for overexpression assays ) . 48 h post-transfection or seeding ( for loss-of-function assays we used stable cell lines with hits silencing ) , media was extracted . Using a dot-blot apparatus with a nitrocellulose membrane , 380 ul of media were loaded into wells of the dot blot templates and proteins were trapped on the membrane by vacuum . Blocking , washing and detection was made as described above , by immunoblotting the membrane . For aSyn cytotoxicity assay , stable H4 cells for aSyn-BiFC system or H4 cells were transduced with lentiviruses ( for assays with hits loss-of-function ) or transfected with overexpression vectors as described above ( for overexpression assays ) . 48 h post-transfection or transduction , culture media was used to determine the levels of released LDH as described in the manufacturer’s instructions ( Clontech Laboratories ) . LDH levels in the culture media were measured and ratio of toxicity between cells with aSyn dimers versus cells with no aSyn was determined . Statistical significance was determined using the paired t-test with Wilcoxon matched pairs test and 95% confidence interval . Differences were considered statistically significant when p≤0 . 05 . Analyses were performed using the Graphpad Prism 5 . 0 software ( GraphPad Software , CA , USA ) . | Synucleinopathies are neurodegenerative diseases characterized by the abnormal accumulation of a neuronal protein called alpha-Synuclein ( aSyn ) . The normal function of this protein in the cell remains unclear , but plays a role in synaptic function and plasticity , cell differentiation and vesicular trafficking . During disease , aSyn is believed to establish aberrant protein-protein interactions , culminating with its accumulation and aggregation and the concomitant disruption of several cellular pathways . Therefore therapeutics aimed at preventing or reversing aSyn accumulation could have potential disease-modifying effects . Here , our aim was to investigate the molecular mechanisms underlying the initial steps that lead to aSyn accumulation . For this , we performed a gene silencing screen to identify novel genetic modifiers of aSyn accumulation . We identified four genetic modifiers of aSyn , including genes involved in intracellular transport and in signal transduction pathways . In the future , the biological contextualization of the early events underlying aSyn pathology , using the modifiers identified in our study , will guide us to a deeper understanding of the role of aSyn in health and disease . In conclusion , our study constitutes a significant step forward towards the understanding of the molecular mechanisms underpinning aSyn accumulation , paving the way towards the development of novel therapeutic strategies for synucleinopathies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunocytochemistry",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"molecular",
"probe",
"techniques",
"immunoblotting",
"light",
"microscopy",
"toxicology",
"physiological",
"processes",
"toxicity",
"microscopy",
"materials",
"sci... | 2016 | shRNA-Based Screen Identifies Endocytic Recycling Pathway Components That Act as Genetic Modifiers of Alpha-Synuclein Aggregation, Secretion and Toxicity |
Aging is an inherently stochastic process , and its hallmark is heterogeneity between organisms , cell types , and clonal populations , even in identical environments . The replicative lifespan of primary human cells is telomere dependent; however , its heterogeneity is not understood . We show that mitochondrial superoxide production increases with replicative age in human fibroblasts despite an adaptive UCP-2–dependent mitochondrial uncoupling . This mitochondrial dysfunction is accompanied by compromised [Ca2+]i homeostasis and other indicators of a retrograde response in senescent cells . Replicative senescence of human fibroblasts is delayed by mild mitochondrial uncoupling . Uncoupling reduces mitochondrial superoxide generation , slows down telomere shortening , and delays formation of telomeric γ-H2A . X foci . This indicates mitochondrial production of reactive oxygen species ( ROS ) as one of the causes of replicative senescence . By sorting early senescent ( SES ) cells from young proliferating fibroblast cultures , we show that SES cells have higher ROS levels , dysfunctional mitochondria , shorter telomeres , and telomeric γ-H2A . X foci . We propose that mitochondrial ROS is a major determinant of telomere-dependent senescence at the single-cell level that is responsible for cell-to-cell variation in replicative lifespan .
Central to current understanding of biological aging is the idea that limited evolutionary investment in mechanisms of cellular and molecular maintenance and repair causes a gradual accumulation of cellular damage , which in turn causes age-related frailty , disease , and death [1] . It follows that aging is not itself genetically programmed , although longevity is , through the genetic regulation of repair mechanisms . It also follows that multiple molecular mechanisms are expected to participate in cell aging and that the actions of these mechanisms are inherently stochastic , i . e . , subject to the variations of chance . The essential multiplicity of mechanisms of aging is widely acknowledged , although in practice , it has not yet been much studied , whereas the stochastic nature of aging is evident in the pronounced variability that is a hallmark of the aging process [2–5] . In this paper , we describe a study of the multiplicity of mechanisms contributing to one of the best-studied cellular models of aging , i . e . , the replicative senescence ( loss of proliferative capacity ) of human diploid fibroblasts ( HDFs ) , which occurs after a number of population doublings in vitro [6] . We show that interactions between mechanisms whose connections have not been previously addressed in this system provide new insights into the underlying biology of replicative senescence . We also show that these interactions explain the marked cell-to-cell variation in replicative lifespan within a culture , which is seen even if cultures are grown clonally from single cell founders [7] . Replicative senescence in HDF is known to be caused ultimately by a DNA-damage response that is triggered by uncapped telomeres [8 , 9] , which in turn result from replication-driven loss of telomere repeats in the absence of telomerase . This process is sometimes interpreted as “programmed aging” at the cell level , although the idea of an underlying program driving replicative senescence is hard to reconcile with its extensive , intrinsic heterogeneity . Furthermore , telomere-shortening rate and cell replicative lifespans can be greatly modified by DNA-damaging oxidative stress [10] via a telomere-specific repair deficiency , which causes stress-dependent accumulation of single-strand breaks [11] and accelerates telomere shortening during DNA replication [12] . This had led to the suggestion that telomere reduction is not strictly programmed , with telomere length acting as a mere cell-division counting device , but instead that telomeres act as sentinels for cumulative oxidative and/or environmental stress triggering division arrest when the damage burden ( detected through telomere length ) becomes too great [10] . Such an idea is fully consistent with the suggestion that replicative senescence serves a pleiotropic role in aging as an anti-cancer mechanism [13] for which there is growing evidence [14–17] , since the damage that triggers senescence would otherwise pose an increased risk of malignancy . It is also consistent with recent evidence linking environmental stress to telomere length in humans [18] , which could provide important insights into the nature of the connection between replicative senescence and organismal aging , as well as into the potential malleability of these processes . It has been recognized for some time that the heterogeneity in telomere-driven replicative senescence challenged the simple idea that in telomerase-negative cells , the telomeres shorten progressively through failure of complete chromosome-end replication . It has been suggested [19] that telomere uncapping , the process that leads to telomeres being recognized as inducers of a DNA-damage response triggering growth arrest , is itself stochastic , but such an effect does not appear sufficient to explain the observed heterogeneity in division capacity within cultures . Rubelj and Vondracek proposed a hypothetical model of “abrupt telomere shortening , ” but this amounted only to speculation embodied in equations [20] . A data-driven modelling approach has shown how the effects of oxidative stress on telomere shortening can be used to study the dynamics of replicative senescence in silico and , in particular , how damage-induced mitochondrial defects might underlie stochastic cell-to-cell variation in oxidative stress , which by acting on telomeres can account for the observed heterogeneity [21–23] . Most of the evidence to date for mitochondrial involvement in aging has come from post-mitotic cells , although recently it was shown that extensive , stochastic variability in the presence and level of mitochondrial DNA mutations occurs in human colon stem cells , one of the most actively proliferating cell types in the body [24] . Motivated by this background of theoretical and experimental work , we have examined the possible role of mitochondria and oxidative stress in telomere-driven replicative senescence , something which until now has been unclear [25] . We first showed that various indices of mitochondrial function revealed significant mitochondrial dysfunction in senescent fibroblast cultures as well as a global change in gene expression pattern indicative of enhanced retrograde signalling . The retrograde response includes down-regulation of mitochondrial membrane potential by reactive oxygen species ( ROS ) -dependent expression of the uncoupling protein UCP-2 and is thought to be at least partly a mechanism of adaptation to high cellular levels of ROS . We then tested whether mitochondrial ROS production plays a causal role in senescence by uncoupling of mitochondria in live cells with 2 , 4-dinitrophenol . This reduced mitochondrial superoxide generation , slowed the rate of telomere shortening , and delayed formation of telomeric DNA-damage foci containing activated H2A . X , thus extending replicative lifespan . These results indicate that mitochondrial superoxide production is a determinant of telomere-dependent replicative senescence . Finally , we examined whether cell-to-cell heterogeneity of mitochondrial ROS production might account for the variability of cell division capacity . We found that those cells in an actively growing culture that showed a senescent phenotype in terms of morphology and lipofuscin content and stained positive for senescence-associated β-galactosidase ( Sen-β-Gal ) and γ-H2A . X were similar to cells in a senescent culture in terms of mitochondrial dysfunction , short telomere length , and co-localisation between γ-H2A . X and telomeres . Our studies confirm an intimate connection in the normal replicative senescence of HDF between oxidative stress and telomere-dependent effects on cell proliferation . They suggest that mitochondrial dysfunction largely determines the age-related development of the extensive cell-to-cell variation in cell division potential .
Production of ROS , especially superoxide , is a necessary by-product of mitochondrial respiration . Under many conditions , mitochondrial superoxide production is positively coupled to the potential energy for protons ΔΨ across the inner mitochondrial membrane and the mitochondrial membrane potential [26] . Mitochondrial dysfunction is characterized by concurrent high superoxide production and breakdown of membrane potential , often associated with mitochondrial DNA ( mtDNA ) damage . To assess whether mitochondrial dysfunction is associated with human fibroblast senescence , we compared young and senescent cultures of MRC5 fibroblasts . Young cultures were defined as actively growing ( >0 . 25 population doublings [PD]/day ) and containing less than 10% of cells positive for markers of senescence and DNA double-strand breaks ( Sen-β-Gal and γ-H2A . X ) . These conditions were met for PDs up to about 40 . Senescent cultures were defined by a net growth rate below 0 . 025 PD/day and more than 50% of the cells positive for Sen-β-gal and γ-H2A . X staining . Using a fluorescent probe that specifically detects superoxide within the mitochondrial matrix ( MitoSOX; Figure 1A ) , we found that mitochondria in senescent fibroblasts produced more superoxide than young ones ( fluorescence increase by 863% ± 49% p = 0 . 003 ) ( Figure 1C ) . Moreover , cellular peroxides as measured by staining with dihydrorhodamine 123 ( DHR ) were increased in senescence ( fluorescence increase by 545% ± 52% , p = 0 . 01 ) ( Figure 1C ) . Surprisingly , the mitochondrial membrane potential ( MMP ) , measured as JC-1 fluorescence ratio by confocal microscopy ( Figure 1B ) and by flow cytometry ( Figure 1C ) , was decreased ( fluorescence ratio decrease to 54 . 6% ± 4 . 1% , p = 0 . 008 ) . There was an increase in mitochondrial mass and mitochondrial DNA copy number in senescent cells ( Figure S1 ) ; however , the increase in MitoSOX and DHR fluorescence levels remained significant even when expressed per unit of mitochondrial mass ( unpublished data ) . Increased mitochondrial ROS production , despite decreased MMP , indicated significant mitochondrial dysfunction in senescent cells . This is in accordance with increased mtDNA damage in senescent cells as shown by decreased amplification efficiency for a large ( 11 kilobase [kb] ) fragment of mtDNA ( Figure 1D; see Protocol S1 for measurement of mtDNA damage ) . Mitochondrial dysfunction induces a major reprogramming of nuclear gene expression , termed the retrograde response [27] , which is a feature of replicative aging in budding yeast [28] and can be induced by severe mtDNA depletion in human cells [29] . Retrograde response includes dysregulation of Ca2+-dependent signalling , up-regulation of mitochondrial biogenesis [30] , and major metabolic and anti-apoptotic adjustments [27] . However , induction of retrograde signalling in mammalian cellular senescence has not been demonstrated so far . An essential trigger for retrograde response is believed to be the increased cytoplasmic [Ca2+]i levels due to low Ca2+ storage capacity of dysfunctional mitochondria with low membrane potential [27] . Senescent fibroblasts displayed significantly higher cytoplasmic Ca2+ levels under basal conditions as well as slower recovery after a Ca challenge as measured by live cell confocal imaging using the Ca2+-sensitive dye Fluo3 ( Figure 1E and 1F ) . Senescent fibroblasts also show increased mitochondrial mass and number ( Figure S1 ) consistent with activation of mitochondrial biogenesis [30] . To further characterize retrograde response in fibroblast senescence , we performed an RNA microarray expression analysis ( see Protocol S1 ) . KEGG pathway analysis comparing young with senescent cells indicated significant enrichment of differentially expressed genes in various major Ca2+-related signalling pathways as well as in energy ( nitrogen and sulphur ) , nucleotide , and amino acid metabolism ( Table S2 ) . Expression of 2 , 610 gene spots changed at least 2-fold between young and senescent cells , with the majority of these ( 1 , 522 ) down-regulated in senescence as compared to young fibroblasts ( Figure S2 ) . A limited number of these changes ( mostly up-regulations ) were confirmed in cells that reached senescence under hyperoxia ( Figure S2 ) . From all genes that were changed similarly under both normoxic and hyperoxic senescence , 92 genes ( 120 probe sets ) could be assigned to one of four processes that are known to be activated in retrograde response: Ca2+-binding and Ca2+-mediated signalling ( Figure 1G ) , glycolysis and Krebs cycle enzymes ( Figure 1H ) , mitochondrial biogenesis and function ( Figure 1I ) , and stress response ( Figure 1J ) . The majority of these genes ( 69 ) are up-regulated in senescence . In particular , we noted activation of enzymes involved in calcium/calmodulin-dependent signalling , including PRKC A and H , PLCB4 , EREG , CEBPE , CREB3L1 and -L4 , TGFβ2 , and IGFBP3 ( Figure 1G ) . A number of enzymes involved in glucose metabolism , such as HK2 , PDP2 , and PDK4 ( Figure 1H ) , glutamate metabolism , such as GCLM and GLUL ( Figure 1H ) , and lipid metabolism , such as CPT1A , PTE1 , and FABP4 ( Figure 1H ) , were up-regulated , as well as enzymes with major functions in mitochondrial metabolism , such as AK3 , PDK4 ( Figure 1H ) , and MAOA ( Figure 1I ) . Up-regulation of some of these genes in senescence and under hyperoxia was confirmed by reverse-transcriptase ( RT ) -PCR ( Figure S3 ) . Increased resistance of cells to stress and apoptosis has also been described as part of the retrograde response . We observed increased mRNA expression of the antioxidant SOD2 and the anti-apoptotic BCL-2 family member BCL2L1 ( BCL-X ) , and a decrease in the expression of the pro-apoptotic BCL2L11 ( BIM ) ( Figure 1J ) . Taken together , the fact that mitochondrial dysfunction during senescence of human MRC5 fibroblasts is accompanied by increased mitochondrial biogenesis , compromised [Ca2+]i handling , and transcriptional up-regulation of genes involved in Ca-mediated signalling , stress response , glycolysis , and mitochondrial function suggests that cellular senescence might be associated with the induction of retrograde response , which in turn is likely to be caused by mitochondrial dysfunction . The finding that low MMP was associated with high mitochondrial superoxide levels in senescent cells was unexpected , because it is well established that lowering MMP ( and thus ΔΨ , allowing leakage of protons across the inner mitochondrial membrane ) decreases mitochondrial ROS production in isolated mitochondria [31] . However , it is also known that ROS and ROS-derived reactive aldehydes activate uncoupling proteins and can thus lower MMP [32] . To test whether MMP was determined by cellular ROS levels in human fibroblasts , we first manipulated ROS levels over about one order of magnitude . Cellular and mitochondrial ROS levels increased 2–3-fold when young human fibroblasts were grown under increased oxygen partial pressure ( 40% ) for 1 wk . Treatment with the free radical scavenger α-phenyl N-tert butylnitrone ( PBN ) blocked this increase . Low ambient oxygen ( 5% ) , treatment with PBN under normoxia [33] , or overexpression of the human SOD3 gene [34] all resulted in a 2–5-fold decrease of ROS levels ( Figure 2A ) . MMP correlated inversely with both mitochondrial superoxide ( n = 4 , r2 = 0 . 710 ) and cellular peroxide levels ( n = 8 , r2 = 0 . 812 , Figure 2A ) , resulting in p = 0 . 0005 ( analysis of variance [ANOVA] ) for the overall correlation . The same correlation was predictive for MMP levels in senescent and sorted early senescent ( SES ) MRC-5 cells , which had high ROS levels ( Figure 2A ) . Within the inner mitochondrial membrane , there is a family of anion carriers which , if expressed , uncouple mitochondria by allowing proton leakage across the membrane . This family includes the uncoupling proteins UCP-1 , UCP-2 , and UCP-3 , and the adenine nucleotide translocase ( SLC25A5A ) . We found that UCP-2 expression was increased at the mRNA level in cells that senesced under either normoxia or hyperoxia ( Figure 1J ) . Up-regulation of UCP-2 in senescent cells was confirmed by RT-PCR ( Figure 2B ) and Western blotting ( Figure 2C ) . To test whether UCP-2 regulated MMP and mitochondrial superoxide production in intact human fibroblasts , we knocked down UCP-2 by transient transfection with small interfering RNAs ( siRNAs ) . Reducing UCP-2 mRNA levels by about 50% resulted in an enhanced MMP , as measured by a 31% ± 6% ( mean ± the standard deviation [SD] , p = 0 . 016 ) increase of the JC-1 fluorescence ratio and in a 50% ± 5% ( p = 0 . 006 ) increase in MitoSOX fluorescence ( Figure 2D ) . There is good evidence that reducing cellular ROS levels can extend the replicative lifespan of human fibroblasts [35 , 36] , and delay telomere shortening [33 , 34 , 37 , 38] , suggesting that mitochondrial dysfunction contributes to replicative senescence . However , other data indicate that mitochondrial ROS production might occur as a result of senescence ( [39–41] , unpublished data ) . To establish the existence and direction of a causal relationship , we slowed down mitochondrial superoxide production by mild uncoupling in young cells , because mitochondrial uncoupling has been proposed as a mechanism of cellular adaptation to protect against oxidative stress [42] . Treatment of MRC-5 cells with a 250 μM concentration of the uncoupler 2 , 4-dinitrophenol ( DNP ) resulted in mild mitochondrial uncoupling as measured by a decreased JC-1 fluorescence ratio ( 85 . 2% ± 0 . 4% , p = 0 . 05; Figure S4A ) . This is important because severe uncoupling , for instance by overexpression of UCP-1 [43] or by using FCCP and DNP under conditions of high oxygen supply ( unpublished data ) , can increase mitochondrial ROS production . Cellular ATP levels were decreased following DNP treatment and reached 75% ± 10% of those in untreated cells after 2 wk . Moreover , long-term DNP treatment prevented the further increase of mitochondrial mass that is characteristic for cells approaching senescence ( Figure S4B ) . Importantly , uncoupling by DNP slowed the accumulation of superoxide in mitochondria of intact fibroblasts ( Figure 3A ) . In comparison to controls , the replicative lifespan was extended by 4–5 PD , amounting to an increase of about 50% of the treated period ( Figure 3B ) . Increased activity of Sen-β-Gal has been associated with cellular senescence [44] , and this increase is also delayed under treatment with DNP ( Figure S5 ) . To establish the mechanism by which uncoupling extends replicative lifespan , we first measured the frequency of cells with nuclear foci containing the phosphorylated form of the histone variant H2A . X , γ-H2A . X ( Figure 3C ) . Formation of foci containing γ-H2A . X can be initiated by DNA double-strand breaks or uncapped telomeres , and is a hallmark of the signalling pathway leading to senescence [8 , 45–47] . Mitochondrial uncoupling delayed foci formation in the majority of cells ( Figure 3D ) . We next investigated the effect of mitochondrial ROS production on telomere maintenance . Telomere shortening is the major route to telomere uncapping in senescing fibroblasts [9] , and the rate of telomere loss in telomerase-negative human somatic cells can be significantly modulated by either increasing oxidative stress [10] or by antioxidant protection [34 , 37] . However , no direct effect on telomere maintenance has been demonstrated before . Mitochondrial uncoupling by DNP resulted in improved telomere maintenance ( Figure 4A ) . The telomere shortening rate of 80 ± 14 base pairs ( bp ) /PD in controls decreased to 9 ± 29 bp/PD in treated cells , so that DNP-treated cells senesced with longer telomeres than controls ( Figure 4B ) . This effect of DNP was confirmed by measuring telomere length by quantitative fluorescence in situ hybridisation ( Q-FISH ) on metaphases ( Figure 4C ) . The telomere shortening rate measured here in DNP-treated cells is lower than what would be expected if telomere shortening was governed by simple overhang resection [48] . However , overhang length and telomere shortening are not correlated in human fibroblasts [49] . Moreover , telomere shortening was similarly slow in other fibroblast systems under comparatively low oxidative stress [12 , 33 , 50] . To further corroborate the association between oxidative stress levels and telomere shortening , ROS levels in three different human fibroblast strains were modified by growth under high or low oxygen concentrations , treatment with the free radical scavenger PBN , or overexpression of SOD , and telomere shortening was measured by two independent methods ( in-gel hybridisation and real-time PCR ) . A strong positive correlation was obtained ( Figure S6 ) , suggesting that lowering the levels of ROS is relevant for the decreased telomere shortening rate under DNP treatment . Because DNP-treated cells ( Figure 4B and 4C ) , similar to cells treated with PBN [33] , senesced with longer telomeres as controls , we next wanted to evaluate the extent of co-localisation between telomeres and DNA-damage foci . There was evident co-localisation between telomeres and γ-H2A . X–containing foci in DNP-treated cells at senescence ( Figure 4D ) . Quantitative evaluation of the deconvoluted images confirmed significant correlation between telomeres and foci equal to the extent of co-localisation seen in senescent control fibroblasts ( see below ) , indicating that the majority of these are telomere dysfunction-induced foci ( TIFs ) . This suggests the possibility that lowering cellular ROS levels might increase the probability of telomere uncapping . Alternatively , there might also an as yet unknown telomere-independent component of lowering ROS levels upon senescence . Together , these data indicate that mitochondrial uncoupling delayed replicative senescence by slowing down oxidative stress-dependent telomere shortening . Having established that , in human fibroblasts , ( 1 ) mitochondrial dysfunction is associated with replicative senescence , ( 2 ) such dysfunction induces a phenotype akin to retrograde response , ( 3 ) mitochondrial dysfunction and retrograde response are linked to mitochondrial uncoupling , and ( 4 ) uncoupling modulates stress-dependent telomere shortening , we were able to test whether the intrinsically stochastic nature of ROS production and mitochondrial dysfunction could account for the intrinsic heterogeneity of replicative senescence . Around 5%–10% of early passage fibroblasts appear senescent as indicated by various biomarkers [51 , 52] ( Figure 5A ) . Thus , lifespan heterogeneity can be characterized by the fraction of cells displaying a senescent phenotype . We noted first that conditions modulating mitochondrial ROS generation significantly altered the frequency of senescent cells . Increasing ambient oxygen led to a time-dependent increase of the fraction of γ-H2A . X– or Sen-β-Gal–positive cells , whereas reduction of ambient oxygen , treatment with the free radical scavenger PBN , or mild mitochondrial uncoupling all decreased this fraction ( Figure 5A ) . Confocal microscopy also revealed considerable cell-to-cell heterogeneity of mitochondrial superoxide levels and membrane potential among cells in proliferating culture ( Figure 1A and 1B ) . To test the hypothesis that cells with higher mitochondrial ROS production might be the ones to develop earlier senescence , we stained a proliferating fibroblast culture sequentially for MitoSOX and γ-H2A . X . This showed that in the vast majority of cases , cells with dysfunctional mitochondria harboured many nuclear γ-H2A . X foci ( Figure 5B ) . We next examined the extent of co-localisation between γ-H2A . X foci and telomeres in foci-positive cells at early passage ( Figure 5C ) . In these cells , foci and telomeres were significantly correlated; however , the correlation coefficient was lower than in senescent cells ( Figure 6H ) . Given that some of the foci-containing cells at early passage are not senescent , but S phase cells displaying ATR-dependent foci at replication forks [53] , this result confirms the idea that early senescence in young cultures is telomere dependent . Accordingly , by analysing metaphases from early passage HDFs by telomere Q-FISH ( Figure 5D ) , we found that some of these metaphases displayed telomere length distributions very similar to those in metaphases from near-senescent cultures ( Figure 5D and 5E ) . Together , data shown in Figure 5 provide evidence linking mitochondrial dysfunction and superoxide production to telomere loss and induction of senescence via TIF formation . To obtain more direct proof for the linkage between mitochondrial dysfunction , telomere maintenance , and early senescence , we used fluorescence-activated cell sorting ( FACS ) to physically sort cells with a senescent phenotype ( SES cells ) out of early passage using cell size and lipofuscin content ( autofluorescence ) as sorting parameters , because both are strongly and positively associated with human fibroblast senescence [51 , 54 , 55] ( see Figure 6A ) . SES cells sorted for large size and high lipofuscin content did not grow when replated , but showed the typical morphology of senescent fibroblasts ( Figure 6B ) . On the contrary , cells sorted for small size and low lipofuscin content grew actively ( PROL cells; Figure 6B ) . Staining of PROL and SES cells for the senescence marker Sen-β-Gal and γ-H2A . X confirmed a high degree of enrichment ( Figure 6C ) . SES cells had significantly higher DHR fluorescence ( 407 . 0% ± 125% p = 0 . 05 ) and lower JC-1 fluorescence ratios ( 63 . 0% ± 10 . 7% p = 0 . 03 ) than the proliferating population ( Figure 6D ) , and showed higher expression levels of UCP-2 ( Figure 6E ) . They also displayed higher mtDNA copy number ( 233% ± 82% , n = 4 , p = 0 . 029 , Figure 6D ) . The increase in mtDNA damage in SES cells did not reach statistical significance ( unpublished data ) . Together , these data indicate mitochondrial dysfunction in SES cells similar to that found in a senescent population . All analysed SES cells showed γ-H2A . X foci , and many of these co-localised with telomeres , whereas there were hardly any nuclei containing foci in PROL cells ( Figure 6F ) . To confirm the role of mitochondrial superoxide generation for telomere-dependent early senescence , we also sorted fibroblasts in early passage according to their MitoSOX fluorescence intensity . There were no γ-H2A . X–positive cells in the low-MitoSOX population , whereas all analysed cells in the high-MitoSOX group showed nuclear γ-H2A . X foci , with the majority of these containing TIFs ( Figure 6G ) . Quantitative correlation analysis again showed the same degree of foci-telomere co-localisation as in senescent cultures ( Figure 6H ) . Thus , cells separated according to either large size and high lipofuscin content or high mitochondrial superoxide production show the same frequency of TIFs as senescent cells .
Since the pioneering work of Harman [56 , 57] , a wealth of data has been accumulated that supports a causal role for mitochondrial dysfunction and ROS production in aging of postmitotic cells [26] . Mitochondrial dysfunction induces retrograde response , a pathway that signals electron transport chain disruption to the nucleus , thus causing wide-ranging adaptations . This is an important part of replicative aging in yeast [58 , 59] . Interestingly , induction of the retrograde response plays partially contradictory roles in yeast aging: as a part of the normal aging process , it can extend lifespan if activated early , but it also contributes to genomic instability and thus curtails longevity [58] . Retrograde signalling has been described in human cells deprived of mitochondrial DNA or following mitochondrial uncoupling [30 , 60] . Disruption of the MMP results in elevated cytoplasmic free [Ca2+]i and activation of Ca2+-dependent signalling including calcineurin and the NFkB pathway , PKC , CREB , and the JNK/MAPK pathway . Retrograde signalling includes up-regulation of mitochondrial biosynthesis via PGC-1 and PPAR-γ , major metabolic and anti-apoptotic adjustments , and possible interaction with TOR signalling [27] . However , there appears to be a large amount of plasticity in this response , depending on , among other things , the specific treatment and cell line examined . Moreover , the role , if any , of mitochondrial dysfunction and retrograde signalling in mammalian cellular senescence is poorly understood [25] . Our data showing decreasing mitochondrial membrane potential , increased mitochondrial biogenesis , decreased capacity for cytoplasmic [Ca2+]i regulation , up-regulation of various genes involved in Ca2+-dependent signalling pathways , metabolic readjustment , and activated anti-apoptotic response are consistent with induction of retrograde signalling in near-senescent human fibroblasts . It should be stressed that the specific subset of genes identified in Figure 1G–1K cannot be taken as an unbiased representation of retrograde response , because they haven been isolated a posteriori from a sample of differentially expressed genes , and they have not been shown to be individually related to this process . However , they might represent a first approach towards candidates for “signature” genes of retrograde response in senescent fibroblasts . Our data are in accord with recent data showing metabolic impairment in senescent cells , including a significant decline in ATP levels [61] . Thus , we propose that mitochondrial dysfunction and resultant retrograde signalling might be a conserved mechanism in cellular replicative aging from yeast to man . Interestingly , our results suggest that the disruption of the MMP in senescent fibroblasts is not simply the result of direct ROS-mediated damage . Rather , it is at least in part an adaptation mediated by increased expression of UCP-2 and , possibly , UCP-3 . Low MMP together with high ROS levels was found before in postmitotic cells [62] . Activity of uncoupling proteins , including UCP-2 , is induced in response to lipid peroxidation products [63] or to superoxide anion in vitro [32] and in vivo [64] . UCP-2 increases the proton leak and decreases MMP and mitochondrial ROS production [65] . This led to the hypothesis of mitochondrial uncoupling as an adaptive mechanism to oxidative stress with potential relevance for aging of postmitotic cells [42] . However , this hypothesis has recently been challenged by showing that physiologically controlled overexpression of SOD-2 in mice , although resulting in decreased mitochondrial superoxide production , did not alter mitochondrial coupling , MMP , or activity of uncoupling proteins in brown fat or skeletal mitochondria [66] . There are conflicting data regarding the interaction between mitochondrial uncoupling and ROS generation and its role for aging in vivo: Mitochondrial uncoupling levels correlated positively with lifespan in outbred mice [67] , and overexpression of the human UCP-2 in adult neurons decreased cellular oxidative damage and extended the lifespan of flies [68] . However , mitochondrial coupling decreased with age in mice skeletal muscle concomitant with altered cellular metabolism and energetics , but without observable changes in the expression of UCP-3 , the major uncoupling protein in muscle [69] . Moreover , long-term caloric restriction increased rat skeletal muscle UCP-3 content and decreased mitochondrial H2O2 production , but did not increase leak-dependent ( state 4 ) respiration [70] . Thus , a thorough study of the effects of ROS level variation on the expression of uncoupling proteins and MMP in a well-defined cellular system was warranted . We found here significant inverse correlations between both mitochondrial superoxide and cellular peroxide levels and MMP , which were independent of the specific type of intervention used to modify ROS levels . This confirms a primary signalling role for ROS in the regulation of mitochondrial uncoupling in human fibroblasts . Senescent and SES fibroblasts both had high ROS levels , high expression of UCP-2 , and an MMP as predicted by the ROS intervention experiments ( see Figure 2 ) . This suggests that up-regulation of UCP-2 , mitochondrial uncoupling , and retrograde response are part of an integral adaptation process of senescent fibroblasts to high ROS levels . Supporting this is the finding that knock-down of UCP-2 generates higher MMP and higher superoxide production in young cells . The association between mitochondrial dysfunction and senescence might be because mitochondrial dysfunction occurs after the cells have exited the cell cycle . However , reversible quiescent G0 arrest in young cultures did not lead to similar effects ( unpublished data ) . More important , artificial mild mitochondrial uncoupling by DNP not only improved telomere length maintenance , slowed down the rate of TIF formation , and elongated lifespan , but it also diminished the increase in mitochondrial ROS production and slowed down mitochondrial biogenesis , one major component of the retrograde response . Thus , the data strongly suggest that MMP-dependent ROS production is a determinant of telomere-dependent replicative senescence . This argument is strengthened by the observation of intra-clonal variation in replicative lifespan of human fibroblasts , which indicates that stochastic factors must determine the potential for each individual cell to divide [7] . This intra-clonal heterogeneity in lifespan has been explained both by theoretical models involving mitochondrial ROS production and telomere shortening [21] , and experimental work showing shorter telomeres in SES cells as compared to the proliferating ones in the same clone [51] . However , so far , the causal chain leading to variable telomere shortening in individual cells has not been elucidated . Heterogeneity in telomere length might simply be a result of intrinsic variation in leading/lagging strand replication and/or telomere end processing . However , our data show that cell-to-cell heterogeneity is governed by the levels of oxidative stress , both at the population level ( Figure 5A ) and at the single-cell level ( Figure 5B ) . Sorting cells with a senescent phenotype out of early passage cultures , we observed the same differences as between senescent and young cells in terms of mitochondrial superoxide production , UCP-2 levels , and MMP , indicating that cell-to-cell variation in the degree of mitochondrial dysfunction and mitochondrial ROS production determines heterogeneity of cellular senescence . In accordance with earlier data [10 , 71] and with those presented above regarding the ROS-dependency of replicative senescence , ROS level variation is a significant determinant of variable telomere shortening , both in terms of average telomere length ( Figure 4A ) and of telomere length at individual chromosomes ( Figure 5D and 5E ) . Thus , we suggest that stochastic variation , occurring at the level of mitochondrial ROS production , mitochondrial dysfunction , and retrograde response , determines the probability of telomere uncapping and thus , cell-to-cell heterogeneity of replicative senescence . Further work is necessary to decide whether mtDNA damage is a cause or a consequence of increased mitochondrial ROS production .
MRC-5 human embryonic lung fibroblasts were from ECACC and P43 , and P100 fibroblasts were obtained from adult skin [34] . Cells were grown in Dulbecco's Modified Eagle's Medium ( DMEM; Sigma , http://www . sigmaaldrich . com ) plus 10% fetal calf serum ( Sigma ) . Ambient oxygen partial pressure was modified using 3-gas cell culture incubators ( Zapf Instruments , Sarstedt , Germany ) to either 95% air , 5% CO2 ( termed normoxia ) , 40% O2 , 5% CO2 ( termed hyperoxia ) , or 5% O2 , 5% CO2 ( termed hypoxia ) . Early senescent cells were sorted by FACS out of young MRC-5 cultures as the quartiles of cells with the highest forward scatter ( for size ) and FL1 values ( for lipofuscin autofluorescence ) as described [51] . Low- and high-MitoSOX cells were sorted in a Becton-Dickinson FACSort ( http://www . bdbiosciences . com ) as the quartiles of cells with the highest and lowest FL3 values following MitoSOX staining . MRC-5 cells at a PD level of 25–30 were transiently transduced with either of three siRNAs specific for human UCP-2 ( si1-UCP: 5′-GCCUGUAUGAUUCUGUCAATT-3′ , si2-UCP: 5′-CCUGUAUGAUUCUGUCAAATT-3′ , and si3-UCP: 5′-CCCUUACCAUGCUCCAGAATT-3′ ) or a control siRNA ( scrambled UCP2 sequence ) . Cells were assayed 72 h after transfection . To measure mitochondrial superoxide , cells were stained in 5 μM MitoSOX Red ( Molecular Probes , http://www . invitrogen . com ) for 10 min at 37 °C , and FL3 median fluorescence intensity was measured by flow cytometry ( Partec PAS , http://www . partec . com ) . Specificity of MitoSOX for superoxide has been shown by the manufacturer , and its mitochondrial localisation was tested by co-staining with Mitotracker Green ( unpublished data ) . Cellular peroxide levels were assessed by staining with 30 μM DHR ( Molecular Probes ) for 30 min at 37 °C and analysis of FL1 fluorescence . As a positive control , cells were treated with H2O2 ( 0 to 400 μM for 30 min ) before DHR staining ( unpublished data ) . MMP was measured as the FL3/FL1 ratio after staining cells with 1-μg/ml JC-1 ( 5 , 5' , 6 , 6'-tetrachloro-1 , 1' , 3 , 3'- tetraethylbenzimidazolylcarbocyanine iodide; Molecular Probes ) in phenol-free RPMI 1640 ( Sigma ) for 30 min at 37 °C . Uncoupling with carbonyl cyanide p-trifluoromethoxyphenylhydrazone ( FCCP ) ( 20 μM for 30 min ) was used as positive control , resulting in a 60% decrease of the JC1 ratio . Mitochondrial mass was measured as FL1 fluorescence after staining of cells with 10 μM nonyl acridine orange ( NAO; Molecular Probes ) for 10 min at 37 °C in the dark . The flow cytometer was calibrated using fluorescent microspheres . All data are mean ± s . e . m . from n independent experiments with measurements in duplicate and 104 cells per measurement . Cells growing on coverslips were stained as above and observed using a Zeiss LSM 510 Meta confocal microscope ( Carl Zeiss , http://www . zeiss . com ) . Conventional transmission electron microscopy was used to confirm NAO staining . Twelve sections of cells with central nuclei per group were systematically random sampled to measure the volume density of mitochondria per cell . Twenty-four hours prior to imaging , cells were plated in DeltaT dishes ( Bioptechs , http://www . bioptechs . com ) at a density of 50 , 000 cells . Fluo3-AM ( Molecular Probes ) calcium dye ( 1 μM ) was loaded for 30 min then washed three times in medium . Fluorescence imaging was performed at room temperature using an LSM 510 Meta confocal microscope . Calcium concentrations were determined as described by [72] . Basal [Ca2+]i was determined from the average of ten consecutive images for each cell . Images were obtained from three independent experiments . Recovery to basal [Ca2+]i was determined as the time to reach 1/e from the maximal response to addition of exogenous 100 mM CaCl2 . Telomere restriction fragment length was measured by in-gel hybridisation following pulsed-field gel electrophoresis ( CHEF3; BioRad , http://www . bio-rad . com ) as described [51] . The amount of loaded DNA and its integrity was controlled by EtBr staining of total DNA in the gel and by re-hybridisation with a ( CAC ) 8 probe as described [34] . The average telomere length was calculated as weighted mean of the peak using AIDA v3 . 11 image analysis software ( Raytek , http://www . raytest . de ) . In some experiments , telomere length was additionally measured using real-time PCR as described [51] . The telomere shortening rate per PD was calculated as the slope of the linear regression between telomere length and PD . Cells grown on coverslips were fixed in 2% paraformaldehyde and incubated with mouse monoclonal anti-γ-H2A . X ( Upstate Biotechnology , http://www . upstate . com ) for 1 h at room temperature . Slides were analysed in a Zeiss LSM 510 Meta confocal microscope . For MitoSOX/γ-H2A . X co-staining , cells were grown on gridded coverslips and first stained with MitoSOX as described above . Cells were mounted in 50% glycerol , photographed by confocal microscopy , and then immunostained for γ-H2A . X . The same areas of the grid were identified and images for γ-H2A . X taken . Metaphases were prepared by treatment of subconfluent cells with 10-μg/ml Colcemid for 2 h at 37 °C , followed by 60 mM KCl for 15 min at room temperature and fixation in ethanol: acetic acid ( 3:1 ) . Slides were air dried and baked at 60 °C for 1 h , re-hydrated in 1 ml of 2× SSC at 37 °C for 2 min , and dehydrated . A total of 20 μl of Cy-3–labelled telomere-specific ( C3TA2 ) 3 peptide nucleic acid ( PNA ) probe ( 4 ng/μl ) ( DakoCytomation , http://www . dako . com ) was applied to the cells followed by co-denaturation at 85 °C and hybridisation for 2 h at room temperature in the dark . Cells were washed with 2× SSC/0 . 05% Tween for 10 min at room temperature . Telomere signals from at least 14 metaphases per group were quantified using TelomereQuant v 1 . 0 ( Dako ) . Cells grown on coverslips were fixed , and γ-H2A . X immunostaining was performed as described above . After application of the secondary antibody , cells were washed with PBS , and Q-FISH was performed as described . After washing with SSC/0 . 5% Tween for 10 min , the secondary antibody incubation was repeated , followed by DAPI staining , mounting , and imaging using a Zeiss LSM 510 Meta confocal microscope . For co-localisation analysis , confocal images were deconvolved and analysed using Costes approximation method [73] in ImageJ , v1 . 37a ( http://rsb . info . nih . gov/ij/ ) . For each nucleus , identified by DAPI staining , a Pearson's correlation coefficient ( R ) was determined between the fluorescein isothiocyanate ( FITC ) and Cy3 images ( demonstrating the amount of co-localisation between the two fluorescent images on a scale of +1 to −1 , representing perfect co-localisation to no co-localisation , respectively ) . The co-localised pixels obtained in the co-localisation analysis were represented in a third image as a white pixel overlay ( Figures 4D , 6F , and 6G ) . | After a limited number of cell divisions , somatic cells lose the capacity for proliferation , called cellular replicative senescence . Senescence , which is triggered by the loss of DNA sequences at the ends of chromosomes ( telomeres ) , is often seen as an example of a regular “biological clock . ” However , cell senescence is heterogeneous , with large differences in lifespan between individual cell lineages . This heterogeneity is clearly related to stress , specifically oxidative stress . It was not known , however , whether stress-induced “premature” senescence involves telomeres or is caused by telomere-independent DNA damage responses . Mitochondria are the most important source of reactive oxygen species ( ROS ) in cells under physiological conditions . We found that mitochondrial function deteriorated while cells approached senescence , leading to increased ROS production . Delaying mitochondrial dysfunction led to postponed replicative senescence and slowing of telomere shortening . Prematurely senescing cells sorted out of young cultures displayed mitochondrial dysfunction , increased oxidative stress , and short telomeres . We propose that replicative telomere-dependent senescence is not “clocked , ” but rather is a stochastic process triggered largely by random mitochondrial dysfunction . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology",
"eukaryotes"
] | 2007 | Mitochondrial Dysfunction Accounts for the Stochastic Heterogeneity in Telomere-Dependent Senescence |
The outer membrane ( OM ) of Gram-negative bacteria is a complex bilayer composed of proteins , phospholipids , lipoproteins , and lipopolysaccharides . Despite recent advances revealing the molecular pathways underlying protein and lipopolysaccharide incorporation into the OM , the spatial distribution and dynamic regulation of these processes remain poorly understood . Here , we used sequence-specific fluorescent labeling to map the incorporation patterns of an OM-porin protein , LamB , by labeling proteins only after epitope exposure on the cell surface . Newly synthesized LamB appeared in discrete puncta , rather than evenly distributed over the cell surface . Further growth of bacteria after labeling resulted in divergence of labeled LamB puncta , consistent with a spatial pattern of OM growth in which new , unlabeled material was also inserted in patches . At the poles , puncta remained relatively stationary through several rounds of division , a salient characteristic of the OM protein population as a whole . We propose a biophysical model of growth in which patches of new OM material are added in discrete bursts that evolve in time according to Stokes flow and are randomly distributed over the cell surface . Simulations based on this model demonstrate that our experimental observations are consistent with a bursty insertion pattern without spatial bias across the cylindrical cell surface , with approximately one burst of ∼10−2 µm2 of OM material per two minutes per µm2 . Growth by insertion of discrete patches suggests that stochasticity plays a major role in patterning and material organization in the OM .
The Gram-negative outer membrane ( OM ) is a complex [1] , non-uniform [2] , largely immobile [3] collection of lipids , lipopolysaccharides ( LPS ) , and membrane proteins . This asymmetric organelle is composed of phospholipids and lipoproteins in the inner leaflet , OM proteins spanning the membrane , and LPS in the outer leaflet . Proteins make up approximately two-thirds of the mass of the OM [4] , and several OM proteins exhibit distinct subcellular localizations on the bacterial surface ( polar , septal , or uniform ) [5] . While there is a growing appreciation for the large diversity of outer membrane protein localization patterns [5] , [6] , insertion patterns have been elucidated only in a few special cases , such as the polar secretion of IcsA in Shigella flexneri [7] , [8]; a general understanding of insertion patterning remains lacking . Although the spatiotemporal dynamics of OM protein insertion patterns have typically not been quantitatively measured , electron microscopy studies have shown that newly inserted porins in Salmonella typhimurium appear as discrete clusters [3] , and also that LPS arrives in localized patches [9] , [10] , indicating that growth is the product of discrete events in which many molecules are inserted in bursts . In Gram-negative bacteria , lipids , proteins , and LPS must traverse the inner membrane and the periplasmic space before insertion into the OM , and each step could potentially be spatially localized and/or occur in bursts . Many components of the molecular machinery implicated in OM protein and LPS transport have only recently been identified [11]–[20] . Secreted proteins are synthesized in the cytoplasm and tagged with an N-terminal signal peptide that targets them for transport across the inner membrane via either the Sec or Tat pathways [21] , both of which are widely conserved among bacteria . The Sec apparatus is uniformly distributed in the inner membrane , while the Tat pathway is concentrated at the poles; nevertheless , some polar-targeted proteins such as IcsA [22] are transported through the Sec apparatus . Finally , translocation across the cell wall and insertion of folded proteins into the outer membrane is mediated by the BAM ( β-barrel assembly machinery ) complex [21] , [23] . After delivery , the dynamic behavior of OM proteins varies according to subcellular position . Label-and-chase experiments , in which cells are imaged immediately after fluorescent labeling of OM proteins and again during growth without further labeling , show cells that initially have uniformly bright peripheries ( indicating that OM proteins are distributed reasonably uniformly across the surface at high density ) but transition over several generations to non-uniform fluorescence distributions , ultimately with only originally labeled “old" poles ( poles of progenitor bacteria , versus new poles synthesized during subsequent rounds of bacterial division ) remaining bright [24] . While general labeling of all outer membrane proteins using amine-reactive ( succinimidyl ester-linked ) fluorescent dyes revealed that a subset was freely diffusible , the non-uniform pattern after label-and-chase indicated that other proteins were far less mobile [25] . Similarly , lectin-labeled LPS molecules were virtually immobile on the time scale of growth [25] , although crosslinking by the multivalent lectin might have limited LPS mobility in this experiment . Since time intervals on the order of one cell cycle are required to produce a shift in cellular fluorescence distribution , growth itself may be intimately coupled to the localization of older OM proteins . The simplest interpretation of polar retention is that new OM material is inserted along the cylindrical portion of the cell but not at the poles . Thus , material at poles tends to remain at the poles , while older material in the cylindrical portion of the cell is spread out by the insertion of new material that results in growth . To elucidate the role of growth in OM organization , we examined the spatial pattern of initial secretion and subsequent redistribution of the abundant OM protein LamB ( maltoporin ) in live Escherichia coli cells . LamB is responsible for the uptake of maltose or maltodextrins , which are important carbon sources and the primary breakdown products of starches in the human intestine [26] . This channel protein also transports other carbohydrates including glucose , lactose , and glycerol [27] , [28] , and is the receptor for bacteriophage λ [29] , [30] . In this work , we study the underlying growth pattern of the E . coli OM on generational time scales by adding a single , small , covalently bound fluorophore to LamB . We found that LamB is secreted in discrete punctate spots that diverge from one another during cell elongation , but are virtually immobile in the absence of cellular growth . LamB puncta that are positioned at the poles during septal growth and subsequent division are retained at the poles , consistent with the polar retention characteristic of the OM population as a whole . We used computational modeling of OM growth to demonstrate that the motion of these puncta , including the dilution of older OM components in the cylindrical portion of the cell , may result from discrete insertion events that are temporally stochastic and randomly distributed over the cylindrical cell surface .
Our experiments employed a genetically encoded 20-residue peptide tag inserted into a surface-exposed external loop of the OM protein LamB [31] . This peptide is recognized and covalently labeled with a single tetramethylrhodamine-Coenzyme A ( TMR-CoA ) fluorophore by the enzyme Sfp [31] . Sfp labeling has previously been demonstrated to label proteins in solution and on mammalian cell membranes [31]–[33]; this is the first application of this technique to an E . coli surface protein . Fluorescent labeling of the bacterial surface expressing LamB under an IPTG-inducible promoter from a multi-copy plasmid required the presence of the peptide tag , Sfp enzyme , and TMR-CoA . SDS-PAGE analysis revealed covalent attachment of the fluorophore to only a single protein band , corresponding to the size of LamB ( data not shown ) . The importance of the biological processes associated with LamB has motivated a diverse collection of genetic and biochemical investigations of the function and behavior of LamB at the level of single proteins , including tracking and mobility of the LamB distributions in live cells [34]–[36] . The Sfp labeling technique is well suited to study the motion of bacterial OM proteins , for which the most common protein visualization technique , Fluorescent Protein ( FP ) fusions , fails because FPs do not fold properly when exported through the general secretory apparatus [37] . Like FP fusions , our labeling strategy uses a covalently attached molecule suitable for following proteins over generational timescales . This probe has the additional advantage that the peptide tag and covalently attached fluorophore are physically small compared to FPs and to other probes previously used to study LamB surface motion [34]–[36] , and is therefore minimally perturbing and offers higher spatial resolution . After IPTG induction of the modified LamB for at least five minutes and subsequent covalent labeling , we observed punctate fluorescence distributions of new protein across the cell surface ( Fig . 1A , B ) . The number of fluorescent puncta per bacterium initially increased linearly as a function of induction time ( Fig . 1C ) ; however , for induction times greater than 20 minutes , the number of distinguishable puncta began to plateau ( Fig . 1C ) due to overlap in the fluorescence distributions of newly appearing puncta with those already present on the bacterial surface . After 1 hour of induction , the fluorescent label distribution on the cell peripheries appeared to be nearly uniform ( Fig . 2A ) . When these cells were allowed to grow in the absence of inducer , dark zones of unlabeled material appeared and were accompanied by the emergence of fluorescent puncta from groups of older , labeled protein . Unlike puncta created by brief induction and subsequent labeling , these puncta were produced by the dilution of labeled material into optically resolvable regions by insertion of new , unlabeled material ( Fig . 2B , C ) . This unlabeled material , which includes LamB , other OM proteins , and LPS , must be inserted in discrete bursts to produce the observed puncta . These observations suggested that OM growth drives the movement of labeled LamB puncta . To examine this hypothesis directly , we used the RNA polymerase inhibitor rifampin to block new mRNA synthesis , thereby arresting cell growth . Bacteria grown in the presence of inducer , fluorescently labeled , and then treated with rifampin exhibited stationary fluorescence distributions for 60 minutes , with no apparent broadening or large-scale migration of the puncta ( Fig . 3B ) . Consistent with previous experiments [34]–[36] , [38] , LamB puncta were essentially immobile in non-elongating , rifampin-treated cells ( Fig . 3B , C ) . However , when driven by the dynamics of cell growth , puncta moved apart ( Fig . 3A , C ) . In addition , fluorescence recovery after photobleaching of rifampin-treated cells revealed that fluorescent puncta do not recover by 40 minutes after photobleaching ( data not shown ) . These observations suggest that puncta are structures that do not engage in extensive diffusive motion or molecular exchange , and thus indicate that there is no large-scale motion of LamB molecules on the surface of non-elongating E . coli . In contrast to rifampin-treated cells , the fluorescence distributions on the surfaces of growing cells changed substantially over time ( Fig . 3A , C ) . Initially , bacteria induced to express tagged LamB for 1 hour exhibited a uniform peripheral fluorescence distribution after labeling ( Fig . 2A ) . As the cells elongated , two notable shifts in the fluorescence distribution occurred . First , distinct fluorescent puncta emerged along the cylindrical portion of the cell periphery . Second , the cylindrical walls of the bacteria became progressively dimmer over successive rounds of division . In contrast , labeled OM material that either started in the polar regions , or that became incorporated into the polar regions during the division process , formed persistently bright puncta at the poles , indicative of the “polar retention" that is characteristic of the OM population as a whole ( Fig . 2C , 3A ) [2] . To measure the growth-dependent dynamics of fluorescent LamB puncta in single cells , we tracked puncta with time-lapse epifluorescence microscopy and measured the divergence of adjacent puncta around the cell periphery . Puncta along the cylindrical walls of the cell diverged from their neighbors faster than puncta in the polar regions ( Fig . 3A , C ) . Movements of puncta tended to be directed , as expected for the surface expansion of an elongating cell . In addition to pairwise divergence , many fluorescent puncta exhibited broadening , and occasionally a single punctum split into two distinguishable spots . We hypothesized that these effects were due to different degrees of insertion of dark material ( unlabeled LamB and other material ) into a diffraction-limited punctum of fluorescent material ( labeled LamB ) . In the case of broadening , if a small or finely distributed amount of dark material is inserted into a diffraction-limited region , the fluorescence distribution may broaden , but not to the point of becoming two individually resolvable puncta . In contrast , spot splitting may occur when enough dark material is inserted into a diffraction-limited fluorescence distribution to split the fluorescence into two distinguishable puncta . Given the negligible mobility of puncta in rifampin-treated cells ( Fig . 3B , C ) , our observations suggest that divergence of neighboring puncta , puncta broadening , and occasional puncta splitting are all due to the pattern of incorporation of new OM material , rather than due to diffusion . Likewise , our data suggests that material incorporated near midcell during the division process remains there over generational time-scales once this region becomes a new pole , due to the significantly reduced rate of insertion at the poles . Based on our observations that the motion of LamB puncta is dependent on growth , we hypothesized that the patchy incorporation of LamB in the OM could be representative of the pattern of OM incorporation as a whole . Our own measurements of LamB mobility as well as measurements by others [3] suggest that diffusion is not a major protein transport mechanism within the OM . In addition , on the time scales of interest for cell growth , we assume bilayer viscoelasticity is negligible . With these constraints in mind , we constructed a minimal model of the OM as a viscous , incompressible , two-dimensional fluid without diffusion that moves with laminar flow . In this model , the only motion relevant on the time scale of cell doubling is due to insertion . Insertion of new material causes elongation of the cell body , which we assume remains cylindrical with a fixed radius matching the shape of the cell wall . Our model of growth does not explicitly include the hemispherical poles , however , the cylindrical shape constraint dictates that there is necessarily a stable , convergent zone of material flow at each end of the cell , out of which material cannot flow in the absence of active insertion in that region . Simulations based on our model consist of a temporally stochastic sequence of insertion events occurring randomly across the cylindrical cell surface . The overall process of insertion is controlled by three kinetic parameters , kon , kins , and τ . We assume that each insertion event is comprised of three steps . First , insertion events are initiated at a fixed mean rate per unit area ( kon ) following Poisson statistics . Each insertion event occurs independently and is uniformly distributed over the cylindrical cell surface . Second , insertion at a particular site leads to the addition of new material at a fixed rate ( area per unit time ) of kins . Third , each insertion event proceeds for an exponentially distributed duration τ corresponding to a single kinetic step for terminating insertion . Note that these assumptions are not critical for the conclusions of the model , rather they were selected as the simplest possible set of assumptions . Each insertion event creates a patch of material with average area τ kins that is labeled ‘light’ during the simulated induction phase , or ‘dark’ during insertion after the induction phase ( see Materials and Methods ) . We note that our model can easily be modified to account for a non-uniform spatial distribution of insertion locations or a non-exponential distribution of insertion-event initiation and duration times; however , this model allows us to explore a wide range of spatial patterns that can be compared with our experimental data . The average behavior of this stochastic growth process is described by the coupled first-order differential equations ( 1 ) ( 2 ) where A ( t ) is the area of the cylindrical cell and n ( t ) is the number of insertion sites at time t . In Eq . 1 , the OM area increases at a rate proportional to the current number of insertion sites and the insertion rate . In Eq . 2 , the number of insertion sites increases proportional to the current cell area and the insertion initiation rate , equivalent to the assumption that the average concentration of insertion sites is constant , with insertion terminating in a single kinetic step . The solutions to Eq . 1 and 2 describe growth according to the exponential function with a doubling time , where Ao is the initial area of the cylindrical portion of the OM . To simulate the full flow field of the OM as the cell grows , our model of OM growth discussed above can be couched as a material flux conservation problem . Given the locations of point sources of new material insertion , , the material flux field follows the continuity equationwhere is the Dirac delta function , and is the vectorial material flux over the cell surface in time . Each insertion site creates a linearly-independent flow field given bywhere the image vectors map the 2D plane onto the cellular topology of a wrapped cylinder with radius R . The position of each insertion site , , is then updated by the total flux field , given byUsing this stochastic model , we performed simulations to mimic the conditions of our label-and-chase experiments in which light material was inserted for 15 minutes , followed by 75 minutes of dark-material insertion . To determine the kinetic parameters for which our simulations most closely match our experimental observations , we explored the growth dynamics in a large region of parameter space across more than four orders of magnitude in kins and three orders of magnitude in kon . To match our experimental measurements , we fixed the doubling time at tdouble = 90 minutes , which constrains the average duration for an insertion event to . A lower bound on the rate of initiation of insertion events was determined from our induction data ( kon>0 . 0005 µm−2 s−1 , Fig . 2C ) , and we estimated the lower bound on the rate of material insertion to be kins = 10 nm2 s−1 . Using our model , we scanned values of the average insertion event duration from 0 . 1 s<τ<200 s and average area per insertion event from 200 nm2<τ kins<1 µm2 . Thus , the range of mean insertion sizes spanned from well below to well above the diffraction limit . The average area added in each insertion event , τ kins , determined the degree of patchiness of the simulated fluorescent material in the OM . The effect of varying the mean insertion size over a modest range ( 20-fold variation ) is shown in Fig . 4; note that changes in τ can be complemented by changes in kins ( and vice versa ) , so that only their product is constrained . In Fig . 4A , newly inserted light and dark material for large , medium , and small mean insertion areas are depicted by the polygons representing each insertion event , with the interstitial white space occupied by the original material . The clear vertical boundaries on the left and right sides of Fig . 4A , where no new material was inserted , are the zones where flow created by growth converges symmetrically on the polar zones , hence the relatively clear delineation between those inert polar regions and the cylindrical regions of active growth . In order to determine whether our model was consistent with our experimental data , we developed software [39] to simulate the full 3D fluorescence light field from our modeling results , along with the corresponding 2D micrographs ( Materials and Methods ) . The configurations in Fig . 4A were convolved into simulated fluorescence images ( 15 min and 90 min time points shown in Fig . 4B ) , and a kymograph was generated using the fluorescence pattern along the top surface of the simulated cell ( Fig . 4C ) . If the average area added during an insertion event was large , the motion of OM material was highly stochastic , and the length scale of OM heterogeneity between newer and older material was relatively large ( Fig . 4 , top panels ) . Conversely , if the average area added per insertion event was small , many insertion events were required to achieve the same degree of elongation; these insertion events were uniformly distributed over the cell surface and hence led to relatively even spreading of the OM , and a relatively short length scale of OM heterogeneity ( Fig . 4 , bottom panels ) . Thus , by varying the average area of each insertion event , a wide range of simulated fluorescence patchiness was accessible in our model . To further constrain the values of the activation rate kon and the mean insertion size τ kins , we used our model to compare simulations and experiments using quantitative metrics motivated by our experimental punctum appearance data . We performed at least 21 independent simulations for 101 parameter sets to explore a wide range of values of kon and kins ( with τ constrained by the fixed doubling time ) , and calculated the average number of diffraction-limited puncta as a function of time , similar to Figs . 1C and 4D . We then fit a sigmoidal function to these curves to define the initial punctum appearance rate and maximum number of puncta ( Fig . 5 ) . When normalized for cell perimeter length and compared with the experimental data ( Fig . 1C ) , both of these metrics predicted an activation rate of 0 . 002 µm−2 s−1<kon<0 . 008 µm−2 s−1 and a mean insertion size of 0 . 015 µm2<τ kins<0 . 06 µm2 , equivalent to the area of ∼1000 typical OM proteins . The proximity of this mean insertion size to the spatial resolution of the optical system , limited by the PSF , is consistent with the observation that puncta both broaden and split during experiments . To determine whether our computational model successfully reproduced other qualitative features of the experimental data , we focused on parameters that matched our experimental punctum appearance data and performed identical analyses on both real and modeled data to study puncta spatial distributions and motion . For kon = 0 . 004 µm−2 s−1 , kins = 0 . 00064 µm2 s−1 , and τ = 50 s , our model produced punctate fluorescence distributions with growth causing divergence of puncta around the cell periphery ( Fig . 6C ) similar to experimental measurements ( Fig . 6A ) . Our simulations also revealed that puncta splitting and broadening can be explained by the blurring effects of the point-spread function ( PSF ) on the fluorescence of labeled material after the insertion of dark ( unlabeled ) material ( Fig . 6D ) . If a small amount of dark material whose size is below the diffraction limit was inserted within a bright region of the OM , the punctum subsequently appeared broader and dimmer , since the resulting pattern could not be resolved as two separate fluorescence distributions . However , if enough unlabeled material was inserted to create a dark region of size similar to the PSF width , the single fluorescent punctum split into two distinguishable puncta . Hence , the relative occurrence of punctum broadening and splitting was determined by the distribution of areas of insertion events . Taken together , our model suggests that patchy OM growth in the cylindrical portion of the cell can lead to the observed phenomena of punctate fluorescence , divergence of puncta , puncta broadening , puncta splitting , and the polar retention of OM proteins in general [24] .
Our observations and simulations support a model in which OM growth occurs in discrete insertion events that are uniformly distributed over the cylindrical cell surface , forming an inhomogeneous mixture of newer and older OM on a length scale set by the size of the insertion events . The appearance of fluorescent puncta can be attributed to a large average area per insertion event ( Fig . 4A ) , and simulations based on our fluid dynamics model indicate that the observed growth patterns do not require spatial organization of insertion events , mechanical stresses , or biochemical interactions between molecules . Similarly , the pattern of newly inserted porins in S . typhimurium observed using electron microscopy consists of patches with dimensions of 50–100 nm that increase in number over time but do not increase in size [3] . We observed this patchy insertion directly in our induction time-course experiments , in which longer induction times produced more puncta that were more densely packed on the cylindrical cell surface ( Fig . 1 ) , and indirectly in our growth experiments , in which new unlabeled material dispersed the older , labeled OM into punctate spots ( Fig . 2 ) . Additionally , other direct and indirect measurements indicate that active translocation events occur at discrete sites on the cylindrical cell surface [40] . While the cylindrical section of the cell undergoes patchy growth , at the division septum , OM material that was once part of the actively growing cylindrical portion of the cell becomes trapped in the inert polar region . Our experiments showed that at short induction times , where new insertion events resulted in new fluorescent puncta , LamB puncta appeared more often in the cylindrical portion of the cell than at the poles ( Fig . 1B ) , indicating that the rate of insertion was indeed lower at the poles . This observation is consistent with our bulk label-and-chase experiments that showed faster dilution of the fluorescence distribution in the cylindrical region , and , after a few generations , the emergence of relatively bright “old" poles that resulted from the lack of growth in the polar region after the initial labeling ( Figs . 2C and 3 ) . While our data cannot be used to rule out the possibility that groups of molecules much smaller than the diffraction limit are occasionally being inserted , when combined with the mechanistic insights from our model , our data suggests that the formation , evolution , and movement of fluorescent OM puncta is well explained by fairly large ( ∼10−2 μm2 ) discrete insertion events that are uniformly distributed over the cylindrical cell surface , with an exponentially distributed duration that has a single , fixed time constant . Within the production and transport pathway of proteins bound for the OM , several possible molecular mechanisms could produce the punctate growth pattern observed here . First , nearly all OM proteins ( with the known exception of PulD [41] , [42] ) are transported from the periplasm , through the rigid cell wall , and inserted into the OM by BAM complexes , and hence one possible interpretation of our data is that the appearance of fluorescent LamB puncta reflects the location of active BAM complexes and OM protein insertion in general . A second , independent possibility is that the uniformly distributed Sec apparatuses [22] , responsible for translocating proteins from the cytoplasm to the periplasm where BAM complexes then effect OM insertion , are only active at specific locations , and hence determine the pattern of protein insertion in the OM . A third interpretation arises from the recent discovery that chromosomally expressed mRNAs remain localized to their transcription sites and form diffraction-limited spots , with mRNA diffusion coefficients orders of magnitude lower than expected for free diffusion [43] . While this observation has not been replicated for plasmid-expressed mRNAs , several multicopy plasmids have been shown to localize to a handful of foci in E . coli [44] . The mean duration of an OM insertion event ( 2 to 30 sec ) is considerably shorter than the half-life of E . coli mRNA ( ∼400 sec ) [45] , thus bursts of cell growth are likely not due to transient mRNA production . However , we speculate that the combination of plasmid localization and mRNA localization to a few foci per cell could give rise to localization of secreted LamB . This list is by no means exhaustive; as data on the lifetimes of candidate molecular complexes ( e . g . BAM and Sec ) become available , the regions of parameter space we have identified through screening of computational results ( Fig . 5 ) could provide a useful means of distinguishing among possible mechanisms producing the patchy patterns of OM protein insertion . Our observations suggest that bacterial cell surface organization is strongly influenced by growth patterning . Calculations based on ferritin labeling of newly translocated LPS in Salmonella predict that an LPS molecule would diffuse less than 300 nm [9] , [10] , or approximately one twelfth of the cylindrical circumference , on the time scale of bacterial cell division . These ferritin-labeled LPS molecules initially appear in clusters assumed to be membrane translocation sites . Electron micrographs taken at regular intervals after labeling revealed that these clusters gradually disperse from each other across the surface of Salmonella , beyond the 300 nm diffusion limit . In comparison to the doubling of OM area that occurs on this time scale , it is reasonable to expect that the motion of OM proteins is dominated by patterns of OM incorporation , not diffusion . The spatial pattern of protein expression dictates not only the distribution of proteins within a single cell , but also the distribution of OM proteins in a population . The unequal partitioning of proteins into daughter cells can produce heterogeneity among isogenic bacterial populations , and in a growing bacterial population , the fraction of cells harboring ‘old poles’ shrinks exponentially . If an OM protein is expressed and subsequently repressed early in the growth curve , polar retention will cause it to be concentrated at these few old poles . In this manner , OM protein inheritance could generate heterogeneity based on generational age that has important phenotypic consequences for differential permeability to essential nutrients , drug concentration , or phage infection . The agreement between our simple model and the observed motion of LamB puncta suggests that heterogeneity may be a general outcome in any bacterium in which the OM grows in discrete bursts .
The background E . coli strain used in this study was a lamB::kan single-gene deletion strain from the Keio collection [46] . Strain JAT567 is this deletion strain transformed with plasmid pEHT1 encoding a LamB derivative under the lac promoter . pEHT1 was constructed from pSB2267 [47] . The 14-amino acid ybbR tag with flanking glycine residues ( GGGTVLDSLEFIASKLAGGG ) [31] is located between codons 155 and 156 of the mature LamB sequence , which were mutated to introduce PstI and XhoI sites , respectively [47] . Plasmid pSB2267 was linearized by digestion with PstI and XhoI and ligated with olignonucleotides containing the ybbR-tag sequence ( ggt ) 3accgttcttgattctcttgaatttattgctagtaagcttgcg ( ggt ) 3 . Purified Sfp protein was a generous gift from Chun Tsai . Sfp was expressed from plasmid pRSG56 [48] . Cells were induced with 0 . 4 mM IPTG in 2YT for 18 hours at room temperature and lysed in an EmulsiFlex-C5 ( Avestin Inc . ) . The cell lysate was centrifuged and the protein was purified from the filtered supernatant in two steps . Filtered lysate was purified by anion exchange with a HiTrap Q FF column ( GE Healthcare ) in 50 mM Bis-Tris ( pH 6 . 5 ) with 2 mM EDTA and a linear gradient from 0 to 500 mM NaCl . Fractions were analyzed by SDS-PAGE and Coomassie staining . Fractions containing Sfp were pooled , concentrated , and purified by gel filtration over a Superdex 200 column ( GE Healthcare ) in 50 mM Tris-Cl [pH 8 . 0] , 5% glycerol , and 10 mM MgCl2 . TMR-CoA was synthesized according to the protocol of Yin et al . [49] with the substitution of methanol buffers for HPLC purification . Purified TMR-CoA was a generous gift from Kierstin Schmidt . LamB was labeled using Sfp , which covalently attaches a TMR-CoA conjugate to the ybbR tag [31] . Mid-logarithmic phase JAT567 cells were labeled in a 50-µL volume with 1 . 5 µM Sfp , 5 µL of 12 µM TMR-CoA , and 10 mM MgCl2 at room temperature for 30 minutes . After labeling , the cells were washed and resuspended in Luria-Bertani broth ( LB ) at room temperature . Strain JAT567 was diluted 1∶100 , grown to mid-logarithmic phase in LB with 35 µg/mL chloramphenicol for 1 . 5 hours at 37°C , and induced with 1 mM IPTG . Cells were then washed in LB and labeled with TMR-CoA . For measurements of the effects of LamB induction time , cells were imaged immediately on a Zeiss Axioplan 2 ( Zeiss , Thornwood , NY ) fluorescence microscope , using a MicroMAX 512BFT ( Princeton Instruments , Trenton , NJ ) camera with 6 . 8 µm pixels and a 100X 1 . 4NA objective , and captured with the MetaMorph ( Molecular Devices ) software package . For label-and-chase experiments , cells were grown at room temperature for 1 hour before imaging . One microliter of labeled cells was mounted on an agar pad and sealed with 1∶1∶1 vaseline∶lanolin∶paraffin , and the cells were imaged under phase contrast and epifluorescence at room temperature as described in [36] . For observations of protein movement in the absence of cell growth , cells were induced and labeled as described above and then incubated in LB with 10 µg/mL rifampin at room temperature for ten minutes and imaged on rifampin agarose pads . We used the Matlab ( The Mathworks , Inc . , Natick , MA ) package PSICIC to define cell outlines and intracellular coordinate systems based on phase images [50] . Corresponding fluorescence images of each bacterium were input into the MTT-peak detection algorithm [51] to determine the positions of fluorescent puncta . Puncta further than four pixels ( ∼272 nm ) from the cell border were removed from further analysis . The remaining fluorescence peaks were linked into multi-frame tracks in two dimensions using the u-track software [52] . These tracks were ordered sequentially around the perimeter and projected onto the cell border for display in one-dimensional kymographs using ImageJ . Using a custom algorithm written in MatLab , we modeled the OM as a cylinder whose surface is a two-dimensional viscous Stokes fluid without diffusion . Each new insertion event began at a random location on the cylinder at a fixed rate per unit area . Each insertion event had an exponentially distributed duration , during which new material was inserted at a fixed rate of area per unit time . Newly inserted material was labeled as either ‘light’ or ‘dark’ to mimic labeling conditions in our experiments . Each new insertion event resulted in the creation of a small , regular polygon whose size and number of vertices were determined by the kinetic rates of area insertion and event duration . To model the effect of Stokes flow on the newly inserted material , the positions of the polygons' vertices were evolved in time using a time-dependent Green's function that was calculated using a series expansion of source images truncated symmetrically at 21 total terms . The simulation time step was chosen to ensure proper sampling of the exponential distribution of insertion-event duration times . For each configuration of polygons , simulated fluorescence images were generated by convolution with the three-dimensional point spread function for the appropriate optical properties of our experimental data acquisition ( 100X 1 . 4NA microscope objective at 575 nm on a charge-coupled device with simulated noise and a pixel size of 6 . 8 µm ) . | All Gram-negative bacteria share common structural features , including an inner membrane , a stiff cell wall , and an outer membrane . Balancing growth in all three of these layers is critical for bacterial proliferation and survival , and malfunctions in growth often lead to cellular deformations and/or cell death . However , relatively little is known about how the incorporation of new material into the outer membrane is regulated in space and time . This work combines time-lapse microscopy with biophysical modeling and simulations to examine potential mechanisms by which new material is added to the outer membrane of the rod-shaped Gram-negative bacterium Escherichia coli . Our results indicate that the outer membrane grows in discrete bursts randomly distributed over the cylindrical cell surface . Each insertion event adds a random amount of new material , pushing old material into new locations and thus expanding the cell membrane . Using our biophysical model , we generated simulated fluorescence images and directly compared analyses of our experimental and computational results to constrain the rate and size of bursts of growth . Together , this indicates that growth of the outer membrane does not require spatial regulation , and the stochastic nature of insertion may contribute to the establishment of cellular patterning and asymmetry . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"bacteriology",
"physics",
"bacterial",
"physiology",
"molecular",
"cell",
"biology",
"cell",
"biology",
"cell",
"growth",
"membranes",
"and",
"sorting",
"biology",
"microbiology",
"biophysics",
"simulations",
"biophysics"
] | 2012 | Analysis of Surface Protein Expression Reveals the Growth Pattern of the Gram-Negative Outer Membrane |
Morocco has achieved the goal of leprosy elimination as a public health problem several years ago ( less than 1 case/ 10 000 habitant ) . The aim of this study was to analyze trends of leprosy detection during the last 17 years taking into consideration the implementation of single dose rifampicin chemoprophylaxis ( SDRC ) started in 2012 . Time series of leprosy cases detected at national level between 2000 and 2017 . Variable collected for each year were leprosy per 100000 H , age category , gender , origin , regions , grade of disabilities and clinical forms . The detection time series was assessed by Joinpoint Regression Analysis . Annual percentage changes ( APCs ) were estimated to identify the years ( joinpoint ) when significant changes occurred in the trend . We therefore examined trends in leprosy detection according to epidemiological variables . Joinpoint regression showed a reduction in the detection rate between 2000 and 2017 . The APC for the period 2012–2017 ( -16 . 83 , 95% CI: -29 . 2 to -2 . 3 , p <0 . 05 ) was more pronounced than that of the previous period 2000–2012 ( - 4 . 68 , 95% CI: -7 . 3 to -2 . 0 , p <0 . 05 ) ; with a significant break in the same joinpoint year SDRC implementation . In stratified analysis , case detection decreased , but not significantly , after the joinpoint years in men , children , multi-bacillary cases , grade 0–1 disabilities , rural and urban cases and in ten regions . Leprosy detection was declining over years with a significant reduction by 16% per year from 2012 to 2017 . SDRC may reduce leprosy detection over the years following its administration .
Leprosy is a chronic infectious disease known since antiquity that mainly damages the skin and peripheral nerves . Leprosy can lead to severe forms of disability and definitive sequelae . Global control actions for leprosy have been a worldwide success since the introduction of antibacterial multidrug therapy ( MDT ) in 1981 . Indeed , global number of leprosy cases has decreased from 5 . 4 million in the early 1980s to 210 758 in 2015 [1] . Despite this progress , leprosy control actions based on a secondary detection strategy can not lead to leprosy elimination nor suppress the source of infection [2] . Additional control intervention toward high-risk groups is then needed to interrupt transmission . Household contacts of leprosy cases are considered the main source of infection [3–4] . Single dose Rifampicin Chemoprophylaxis ( SDRC ) of leprosy contacts is a preventive intervention that may stop the transmission of infection among household contacts [5] . A randomized controlled trial [6] including more than 28 000 contacts in Bangladesh showed evidence that SDRC reduces the incidence of leprosy by 57% in the first two years following the intervention . This effect is maintained after 4 to 6 years . A meta-analysis [2] confirmed also the effectiveness of SDRC in preventing new leprosy cases after two years . In Morocco , great efforts were undertaken over several decades to control leprosy . The country achieved the goal of leprosy elimination as a public health problem in 1991 [7] and the decreasing trend in new case detection is confirmed year by year . SDRC was introduced in Morocco in 2012 among household contacts of accumulated leprosy cases registered in the country since 2000 as well as the household contacts of new cases registered from 2012 . To our knowledge , this is the first time series on leprosy conducted in a country that adopted SDRC in its interventions and at a national level . The aim of this study was to examine the trends of new leprosy cases detected between 2000 and 2017 , as well as to generate a hypothesis suggesting any link between the decrease in the number of detected cases and the introduction of SDRC in Morocco .
We performed a time series study based on new leprosy cases data registered in Morocco between 2000 and 2017 . Morocco is a northwestern African country that includes 12 administrative districts or regions and more than 70 provinces . The ethical committee of the Faculty of Medicine and Pharmacy of Rabat approved the study protocol . All data analyzed were anonymized . Leprosy detection rate per 100 000 habitant was calculated from the data of the National Leprosy Control Program ( NLCP ) and the population data obtained from le Haut Commissariat au Plan for the period 2000–2017 . For each year , the detection rate was the number of new leprosy cases registered in the country reported per 100 000 habitant [7] . Data collected , for each year , were also age category ( adult , child ) , gender , origin ( rural , urban ) , grade of disability , regions and clinical leprosy forms ( multi-bacillary or pauci-bacillary ) of all the new cases registered during the same period . Data from SDRC surveys among household contacts of cumulative leprosy cases registered from 2000 to 2017 were also collected . Household contacts of a leprosy case index were defined as people living under the same roof with the case index for more than three months [8] . SDRC was administered during the leprosy case investigations conducted by regional NLCP teams . SDRC was administered after a counseling session and a careful clinical examination of contacts in order to eliminate contraindications to rifampicin [8] . SDRC was administered to contacts with respect for ethics and confidentiality . First , the observed trend of leprosy detection rate was described between 2000 and 2017 . Then , a Joinpoint Regression Analysis was conducted to assess the time trends [9] . Joinpoint regression detects years when a significant change in leprosy detection trend has occurred . This method is widely used in trend analysis of incidence rate or mortality of several diseases [10] . Joinpoint regression analysis was adjusted to estimate the Annual Percent Changes ( APCs ) and to identify the joinpoints or years where significant changes over time in the linear slope of the trend had occurred . Statistical significance was tested using the Monte Carlo Permutation method [9] . A maximum of one joinpoint was allowed in the model and each APC segment was calculated using a log-linear model . The 95% confidence intervals ( 95% CI ) were calculated for each APC and were used to determine whether the APC for each segment was significantly different from the previous time segment . A p value <0 . 05 was considered statistically significant . Moreover , a stratified analysis by age ( adult child ) , gender , clinical forms ( multibacillary , paucibacillary ) , origin ( rural , urban ) , regions and disability grades were also studied . Joinpoint regression analysis was performed using Joinpoint Regression Program , version 4 . 2 . 0 . 2 ( United states National Cancer Institute , Bethesda , MD , USA ) [11] .
A total of 801 new leprosy cases were recorded between 2000 and 2017 . Epidemiological characteristics of registered cases are shown in Table 1 . The number of new cases detected decreased from 61 cases in 2000 to 13 cases in 2017 . Leprosy detection rate was declining since the early 2000s , as the annual detection rate decreased from 0 . 21 per 100 000 H in 2000 to 0 . 04 per 100 000 H in 2017 . Since the introduction of SDRC into the routine activities of the NLCP in 2012 , 146 investigations were conducted by the regional teams . Among 5201 household contacts registered , 4019 ( 77% ) contacts were examined during investigations , of whom 3704 ( 93% ) received SDRC ( Fig 1 ) . The joinpoint regression analysis showed a significant reduction in leprosy detection between 2000 and 2017 ( Fig 2 ) . The leprosy detection rate decreased by 4% per year from 2000 to 2012 and by 16% from 2012 to 2017 . The APC for the period 2012–2017 ( APC = -16 . 83 , 95% CI: - 29 . 2 to -2 . 3 , p <0 . 05 ) was then more pronounced than the previous period 2000–2012 ( APC = -4 . 68 , 95% CI: -7 . 3 to -2 . 0; <0 . 05 ) . The joinpoint corresponded to the year 2012 . The stratified joinpoint regression analysis ( Table 2 ) showed that the detection rate decreased significantly after the joinpoint years in female cases ( APC = -12 . 3 , 95% CI: -18 . 1 to -6 . 3; <0 . 05 ) , adults ( APC = -17 . 0 , 95% CI: -29 . 7 to -2 . 0 , p <0 . 05 ) , paucibacillary forms ( APC = -25 . 0 , 95% CI: -37 . 1 to -10 . 8 , p <0 . 05 ) and in two regions: Béni-Mellal-Khenifra ( APC = -22 . 9 , 95% CI: -35 . 2 to -8 . 1 , p<0 . 05 ) and Marrakech-Safi ( APC = -8 . 9 , 95% CI: -16 . 3 to -0 . 8 , p<0 . 05 ) . The detection rate decreased also but not significantly after the joinpoint years in men , children , multibacillary cases , grade 0–1 disability cases and rural and urban cases . For grade 2 disability cases , the detection rate increased insignificantly from 2000 to 2006 in ( APC = 25 . 7 , 95% CI: -5 . 4 to 67 . 1 ) followed by a non-significant decrease in the second period ( Fig 3 ) .
This study shows that leprosy detection was declining over years with a significant reduction by 16% since the introduction of SDRC in Morocco . Leprosy control strategies in the country were implemented through NLPC . NLCP is one of the oldest programs in the Ministry of Health; it was set up in 1981 . Control actions for leprosy in Morocco started since the fifties by the establishment of a prophylactic and socio-medical policy . Today , NLCP achievements are evident since the epidemiological situation of leprosy confirms the decreasing trend in new case detection . An average of 26 new leprosy cases was detected during the last five years . NLCP control actions have given the opportunity to many patients , mostly of low socioeconomic level , to be treated free of charge and , if necessary , rehabilitated for a better reintegration into society . Leprosy in Morocco remains contagious and the genetic predisposition to this disease explain the fact that more than 50% of new cases diagnosed have either a leprosy family contact and more than 60% are from four high endemicity aeras for leprosy [8] . Given that , NLCP has accorded a close attention to contact surveillance . Several publications have demonstrated SDRC cost-effectiveness among contacts with 50–60% efficiency [12–14] . However , there is no recommendation of World Health Organization ( WHO ) to implement SDRC in low or high endemic countries because of low quality of evidence in the field . WHO experts in leprosy advise low endemic countries , such as Morocco , to implement SDRC into routine program activities since they already have a leprosy surveillance system in place . [1] . SDRC was then implemented at national level in Morocco among household contacts with the aim to reduce leprosy transmission among contacts . Our study showed that leprosy detection has been declining since 2000 with a significant break in the curve occurring in 2012 . Leprosy detection has dropped by an average of 16% per year since the introduction of the SDRC into NLCP . SDRC seems then able to reduce leprosy detection over the years following its administration by interrupting transmission . With regard to examined trends according to epidemiological characteristics of cases , leprosy detection was decreasing from 2000 to 2017 in the 12 regions of Morocco even in male , female , adult , children , urban and rural cases , all clinical forms and in cases with disabilities grade 0 and 1 . The decreasing trend of leprosy was more pronounced in the second period after joinpoints . Years corresponding to joipoints were different following the characteristics but were all over 2006 . With regard to grade 2 disabilities trend of leprosy detection was increasing between 2000 and 2006 followed by an inversion of the curve and a significant decreased in leprosy detection after 2006 . One fact may explain the pronounced drop of detection rate in the second periods: Morocco has reviewed its control strategy for leprosy in 2006 by adopting the WHO short protocol of multidrug therapy . This protocol is more effective , suppresses infection and facilitates observance . The new organization of NLCP after 2006 was accompanied as well by a decentralization of leprosy health care allowing better geographical access to services for rural areas , early case detection , early treatment and closed contact surveillance; leading then to the reduction of new case detection in the country . To our knowledge , national reporting protocols and trends have remained constant and rigorous over the period studied . Stratified analysis showed that leprosy detection rate decreased significantly prior to 2012 intervention in female cases , adults and in paucibacillary forms . These results are common findings in countries with declining leprosy like Morocco , as there is evidence of decreases in female to male sex-ratio and in paucibacillary forms among new cases . The proportion of paucibacillary cases has been found to decrease in several populations concomitant with declining incidence [15] . These patterns probably reflect an increase in the proportion of long incubation period cases as the disease disappears . Long incubation periods are associated more with multibacillary than with paucibacillary forms , and multibacillary cases are associated with males [16] . In the same way , if MDT affects the transmission of infection it should also reduce the detection rate of paucibacillary cases first because of their shorter incubation periods . Our findings can be then the results of MDT and decline in transmission . Moreover , differences in sex-ratios in low endemic countries can be influenced by susceptibility to infection or to disease and also by potential bias in case ascertainment . Social stigma and sociological patterns that confine females to houses in some settings may be related to low case reporting of females . On the other hand , 94% of leprosy cases registered in the study period were adult forms , those cases significantly decreases after 2012 . This can be due to general effect of SDRC . The present study is the first one that reported data on leprosy detection in Morocco in the last two decades . It has the strength to gather information comprising all cases reported nationally through NLCP; however , some limitations of the study have to be considered when interpreting our findings: 1/ Data of NLCP may present inconsistencies in the quantity and quality of information over time and between the regions . Case detection maybe under-reported , despite the progress achieved during the observational period in terms of active case findings ( examination of household contacts ) , early MDT treatment ( WHO protocols ) , social support , training of health care workers , integration of NLCP activities into primary healthcare services and implementation by general practitioners . 2/ Stratified analysis showed that leprosy declined significantly prior to 2012 intervention in female cases , adults and in paucibacillary forms . Other possible factors related to 2006 policy changes in NLCP may also contribute to the decreasing trend of leprosy in the country . 3/ There is no doubt that NLCP control measures have played an important role in leprosy decline; however , improvements in socioeconomic conditions over the country , implementation of laws reducing stigma and other cultural factors should also be considered , even if their contribution to leprosy decline remains difficult to assess . These issues deserve further investigations and may provide valuable insights . 4/ Stratified analysis by regions have showed a decrease in leprosy detection . The decreasing trend of leprosy after joinpoints was not significant in all regions except for two ones ( Béni-Mellal-Khénifra and Marrakech-Safi ) . This can be due to a less of statistical power of analysis because of limited recoil of the study ( only 5 years after the joinpoint ) . 5/ The joinpoint regression analysis used in the study is an analytical but exploratory method that does not take into account factors that may influence leprosy trends ( SDRC , level of regional implementation of NLCP , etc ) . A further study , using interrupted time series analysis with a multivariate approach is required and is currently on the way in order to assess , in statistical terms , how much an intervention changed an outcome of interest and whether factors other than the intervention could explain the change [17] . Despite these limitations , the data analyzed in our study are consistent and representative for a country over a period of 17 years . In conclusion , Morocco’ decline in leprosy reflects a global trend . This time series demonstrated a drop in case detection by 16% per year since SDRC implementation . This fact can allow us generating a hypothesis about a possible influence of SDRC on accelerating the reduction of leprosy detection in Morocco . An interrupted time series analysis is needed to confirm this hypothesis . | After the introduction of effective therapy for leprosy by the world health organization in early eighties , there has been a marked reduction in new case detection all over the world . Despite the decline in case detection of leprosy , incidence reduction to zero and suppression of infection source is still not possible . Addressed strategies to high-risk groups–contacts of leprosy cases- may be effective to eliminate the disease at global level . Chemoprophylaxis in contacts may lower the incidence of leprosy in contacts of patients diagnosed with leprosy . In Morocco , chemoprophylaxis was introduced at national level in 2012 . Morocco has achieved the goal of leprosy elimination as a public health problem in 1991 and the declining trend of leprosy detection is observed years after years . However , no study has been conducted in the country to analyze whether the declining trend of leprosy can be related or not to chemoprophylaxis in contact . This time series can be of interest for medical community as the authors analyses the trend in leprosy in the last 17 years and try to generate a hypothesis about any relationship with chemoprophylaxis implementation . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"morocco",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"disabilities",
"statistics",
"antigen-presenting",
"cells",
"immunology",
"tropical",
"diseases",
"geographical",
"locations",
"regression",
"analysis",
"bacterial",
"diseases",
"mathematics",
"neglecte... | 2018 | Trend analysis of leprosy in Morocco between 2000 and 2017: Evidence on the single dose rifampicin chemoprophylaxis |
Opisthorchis viverrini is distinct among helminth infections as it drives a chronic inflammatory response in the intrahepatic bile duct that progresses from advanced periductal fibrosis ( APF ) to cholangiocarcinoma ( CCA ) . Extensive research shows that oxidative stress ( OS ) plays a critical role in the transition from chronic O . viverrini infection to CCA . OS also results in the excision of a modified DNA lesion ( 8-oxodG ) into urine , the levels of which can be detected by immunoassay . Herein , we measured concentrations of urine 8-oxodG by immunoassay from the following four groups in the Khon Kaen Cancer Cohort study: ( 1 ) O . viverrini negative individuals , ( 2 ) O . viverrini positive individuals with no APF as determined by abdominal ultrasound , ( 3 ) O . viverrini positive individuals with APF as determined by abdominal ultrasound , and ( 4 ) O . viverrini induced cases of CCA . A logistic regression model was used to evaluate the utility of creatinine-adjusted urinary 8-oxodG among these groups , along with demographic , behavioral , and immunological risk factors . Receiver operating characteristic ( ROC ) curve analysis was used to evaluate the predictive accuracy of urinary 8-oxodG for APF and CCA . Elevated concentrations of 8-oxodG in urine positively associated with APF and CCA in a strongly dose-dependent manner . Urinary 8-oxodG concentrations also accurately predicted whether an individual presented with APF or CCA compared to O . viverrini infected individuals without these pathologies . In conclusion , urinary 8-oxodG is a robust ‘candidate’ biomarker of the progression of APF and CCA from chronic opisthorchiasis , which is indicative of the critical role that OS plays in both of these advanced hepatobiliary pathologies . The findings also confirm our previous observations that severe liver pathology occurs early and asymptomatically in residents of O . viverrini endemic regions , where individuals are infected for years ( often decades ) with this food-borne pathogen . These findings also contribute to an expanding literature on 8-oxodG in an easily accessible bodily fluid ( e . g . , urine ) as a biomarker in the multistage process of inflammation , fibrogenesis , and infection-induced cancer .
Over 750 million people ( 10% of the human population ) are at risk of infection with food-borne trematodes , with more than 40 million people currently infected with one of three of these parasites: Clonorchis sinensis , Opisthorchis felineus , and Opisthorchis viverrini [1 , 2] . O . viverrini is considered the most important of these food-borne trematodes due to its well-documented association with hepatobiliary pathologies that include advanced periductal fibrosis ( APF ) [3 , 4] and intrahepatic cholangiocarcinoma ( CCA ) [5–10] . In Northeastern Thailand ( Isaan ) , uncooked cyprinoid fish , which is the intermediate host for the parasite , are a staple of the diet , with O . viverrini infecting an estimated 10 million people in Isaan alone [8] . While infection with O . viverrini can be eliminated by chemotherapy ( praziquantel ) , regional culinary practices result in rapid re-infection after treatment , often leading to life-long infection with the parasite [5 , 7 , 11] and the highest incidence of CCA in the world ( 85 per 100 , 000 ) [7] . In our community-based ultrasound studies in Northeastern Thailand [8 , 9 , 12 , 13] , we have identified a series of pathologic changes that occur early and asymptomatically in the bile duct in individuals resident in O . viverrini endemic areas . As individuals can be infected with O . viverrini for years ( even decades ) , we hypothesize that a chronic cycle of tissue damage and repair ensues in the intrahepatic biliary ducts as a result of the constant immunological , mechanical and oxidative damage from the parasite , resulting in a persistent “smoldering and chronic inflammatory milieu” [14] . These processes stimulate the production of desmoplastic stroma ( i . e . , bile duct fibrosis ) , which has recently been shown to play a crucial role in promoting malignant transformation to CCA ( see [15] ) . In both the hamster and human models of O . viverrini infection , fibrosis in the biliary epithelia routinely precedes CCA [8 , 10 , 16 , 17] . The exact mechanism by which stromal desmoplasia ( bile duct fibrosis ) transforms to CCA is a topic of intense research [15 , 17–19] . There is some consensus that an important component in this transformation is the genomic instability that accompanies both fibrogenesis and carcinogenesis[20 , 21] . During these processes , cells are recruited to the site of damage and induce a “respiratory burst” from an increased uptake of oxygen , with the accumulation of reactive oxygen species ( ROS ) referred to as oxidative stress ( OS ) ( see [22] ) . Both fibrotic [16] and neoplastic transformation [20 , 23] have been linked to increased levels of OS by several mechanisms , including DNA damage , genomic instability , and cellular proliferation , The DNA base modifications caused by OS also result in oxidation of guanine residues to 8-oxo-7 , 8-dihydro-2′-deoxyguanosine ( 8-oxodG ) , which are excised into bodily fluids , such as urine , blood and saliva . Augmented 8-oxodG levels in urine have been used as a biomarker for oxidative DNA damage [20 , 21 , 24–28] in acute lymphoid leukemia , colorectal cancer , high grade cervical dysplasia , scleroderma fibrosis , liver fibrosis , renal cell carcinoma , lung cancers , and prostate cancer [15 , 22 , 29 , 30] . Elevated levels of 8-oxodG have been reported in the urine of individuals chronically infected with O . viverrini and in the urine of individuals with O . viverrini-induced CCA [31–33] . However , to our knowledge , there have been no studies that measure levels of urine 8-oxodG during the interval from initial infection with O . viverrini to neoplastic transformation to CCA , a progression which occurs over years and proceeds through several well-defined hepatobiliary pathologies [12] . The most precisely detected and well documented of these hepatobiliary pathologies is APF [9] . Our community-based ultrasound studies ( i . e . , Khon Kaen Cancer Cohort or the KKCC ) along the Chi River , Khon Kaen Province , Thailand have shown that APF is prevalent among otherwise apparently healthy residents in O . viverrini endemic areas [5 , 8–10 , 13] . The objective of the current manuscript was to determine if levels of 8-oxodG in urine increased when O . viverrini infection transitioned to APF [34] and if elevated levels of urine 8-oxodG associated with APF were comparable to the markedly elevated levels of urine 8-oxodG observed in individuals with O . viverrini-induced CCA [17 , 28 , 35] . As APF is a precursor stage to CCA , a threshold concentration of urinary 8-oxodG could then be used to risk stratify APF individuals into those more likely to develop O . viverrini-induced CCA and , as such , take advantage of recent therapeutic advances used to target bile duct fibrosis and the rich milieu it provides for neoplastic transformation to CCA [36] . More specifically , our objectives were as follows: ( a ) quantify the presence of urinary 8-oxodG in O . viverrini infected individuals with APF compared to O . viverrini-infected individuals without APF and O . viverrini-individuals with CCA and ( b ) determine the performance of urine 8-oxodG in predicting APF , including its sensitivity , specificity , and ability to predict diagnostic risk ( e . g . odds ratios ) . These studies form the “discovery phase” for a potentially critical and easily accessible candidate biomarker that could be advanced to biomarker validation in large-scale studies in Thailand and in the neighboring countries of the Mekong River Basin , where O . viverrini-induced CCA has the highest incidence in the world [7 , 12] .
The participants provided written informed consent using forms approved by the Ethics Committee of Khon Kaen University School of Medicine , Khon Kaen , Thailand ( reference number HE480528 ) and the Institutional Review Board ( IRB ) of the George Washington University School of Medicine , Washington , D . C ( GWUMC IRB# 020864 ) . The urine donated by participants from Group 4 was obtained from the biological specimen repository of the Liver Fluke and Cholangiocarcinoma Research Center , Khon Kaen University , Thailand using a protocol approved by the Ethical Committee on Human Research , Faculty of Medicine , Khon Kaen University , Khon Kaen , Thailand ( reference Nos . HE450525 and HE531061 ) . This study uses baseline data from a recently enrolled village in the Khon Kaen Cancer Cohort ( KKCC ) , which samples villages along the Chi River Basin , in Khon Kaen province , Thailand . A detailed description of the KKCC and the methods used to assemble this cohort have been extensively reported [8–10 , 13] . A modification to the original KKCC protocol was approved by the Division of Microbiology and Infectious Diseases ( DMID ) of the US National Institutes of Health , the Ethics Committee of Khon Kaen University School of Medicine , and the GWU IRB , allowing urine collection , which initiated in 2012 ( Fig 1 ) . The current sample of 221 individuals represents the most recent enrollment into the KKCC following this modified protocol ( Fig 1 ) . Participant inclusion criteria consisted of enrolling all males and females between 20 and 60 years of age ( inclusive ) registered with the village health outpost who were willing to participate in the study as evidenced by signing the informed consent form . Participants were excluded from the study if they attended school or worked full-time outside of the village ( n = 0 ) or had a positive urine β-hCG pregnancy test in the case of females ( n = 0 ) . Individuals infected with O . viverrini were referred to the local public health clinic for treatment with praziquantel ( PZQ ) regardless of their participation in the study . The data for the 181 KKCC participants were stratified during analyses to remove thirty two ( n = 32 ) individuals whose feces contained eggs or larvae from other helminths endemic to the region , including hookworm ( Necator americanus ) , Ascaris lumbricoides , or Strongyloides stercoralis ( Fig 1 ) . This reduced the study sample to 149 individuals assigned to three groups as follows: Group 1 ( n = 23 ) individuals negative ( uninfected ) upon fecal examination for O . viverrini ( also referred to as Endemic Controls or EN ) ; Group 2 ( n = 48 ) O . viverrini positive ( OV+ ) individuals who were negative for APF as determined by US ( also refereed to as Clinical Controls ) ; and Group 3 ( n = 78 ) O . viverrini positive individuals who were also positive for APF as determined by US ( also referred to as Clinical Cases ) ( Fig 1 ) . Group 4 consists of samples from 33 individuals who were not part of the KKCC , but had histologically proven O . viverrini-associated CCA , whose urine samples were stored at the biological specimen repository of the Liver Fluke and Cholangiocarcinoma Research Center , Khon Kaen University , Thailand . The samples from Group 4 were chosen from the biological specimen repository by matching them on age , sex , and nearest neighbor ( i . e . , residence at time of death in a village within 10 kilometers of the current study sample ) to the Clinical Cases in the KKCC . A detailed description of the ultrasonography methods used in this study can be found in the following references [8–10] . Briefly , a mobile , high-resolution ultrasound ( US ) machine ( GE model LOGIQ Book XP ) was used . Hepatobiliary abnormalities including portal vein radical echoes , echoes in liver parenchyma , indistinct gallbladder wall , gallbladder size , sludge and suspected CCA were graded and recorded . Individuals were classified as “Non-Advanced Periductal Fibrosis” ( APF- ) or “controls” if the US grade was 0 or 1 , and “Advanced Periductal Fibrosis” ( APF+ ) or “case” if the US grade was 2 or 3 . Individuals with alcoholic liver disease , which is seen as fatty liver by US exam , were excluded from this study . Also , individuals with marked hepatic fibrosis not related to OV infection ( e . g . , cirrhosis from HBV or HCV ) were also excluded from this study ( see Fig 1 ) . Data analysis was performed with SAS 9 . 2 ( SAS Institute Inc . , Cary , NC , USA ) . The raw data used to generate this manuscript are publicly available from the Dryad Digital Repository with the accession number doi:10 . 5061/dryad . pd6mn .
Table 1 shows the descriptive statistics for urinary 8-oxodG levels by sex , age strata , smoking status , alcohol consumption , and the criteria used to diagnose various stages of O . viverrini-induced infection . The levels were highest among persons in the 30–39 years of age group followed by individuals in the 20–29 years of age group , which consisted of only three individuals . Smokers had higher median levels of 8-oxodG at 185 . 93ng/mg creatinine compared to 147 . 55ng/mg creatinine in non-smokers . Similarly , individuals who indicated that they consumed alcohol had higher median levels of creatinine adjusted 8-oxodG compared to those who did not , with 154 . 83ng/mg creatinine and 133 . 43ng/mg creatinine , respectively . The 8-oxodG levels increased with advancing disease status . The Kruskal-Wallis one-way analysis of variance demonstrated that there were significant differences in the distributions of urinary 8-oxodG levels between participants separated into groups by disease progression ( p < . 0001 ) . Dunn's multiple comparisons procedure , for which the family-wise error rate was set at 0 . 05 , identified significant ( p < 0 . 05 ) differences in the distributions of urinary 8-oxodG between control individuals ( both EN and APF- ) and O . viverrini-infected individuals ( APF or CCA positive ) . No significant differences were observed between participants in EN and APF- groups , which , along with co-occurrence of these individuals in the same population , justified combining these participants into one group in subsequent analyses . No significant difference was observed between participants in the APF+ and CCA groups . Lack of statistically significant findings here may be explained by the presence of three individuals in the APF+ group who had extremely high levels of urinary 8-oxodG ( 4 . 66 , 4 . 33 , and 4 . 04 times greater than the group's median urinary 8-oxodG levels ) . However , the groups remained distinct in subsequent analyses due to the presence of the outlier values and also due to the specific pathological differences that exist between APF+ and CCA individuals . Fig 2 illustrates the distribution of urinary 8-oxodG in individuals participating in this study . Panel A of Fig 2 depicts the distributions of the levels of urinary 8-oxo-dG in each of the four groups of the study participants and Panel B illustrates the distributions following the amalgamation on participants from groups 1 and 2 . Superscripts above the group labels on the x-axis indicate significant findings . Following the combination of control participants into one group ( HBP negative ) , the following results were obtained: the concentrations of 8-oxodG in both cases and controls ranged between 14 . 25ng/mg creatinine and 1100 . 97ng/mg creatinine . The range was narrower in the control group ( 14 . 25 to 476 . 58 ng/mg creatinine , n = 71 ) than among cases , especially OV+ and APF+ ( 28 . 89 to 1100 . 96 ng/mg creatinine , n = 78 ) and OV+/CCA individuals ( 26 . 74 to 933 . 77 ng/mg creatinine , n = 33 ) . The median 8-oxodG levels of persons with APF or CCA were significantly higher ( 2 . 12x and 1 . 62x , respectively ) than the median level of the control group as determined by Dunn's multiple comparisons procedure applied to the Kruskal-Wallis one-way analysis of variance , with the family-wise error rate set at 0 . 05 . These data indicated progressively higher levels of 8-oxodG in the urine during advanced hepatobiliary disease during opisthorchiasis . Various candidate biomarkers , including urinary 8-oxodG , serum level of IgG and IgG1 against a crude adult O . viverrini antigen extract were modeled as potential relevant predictors of progression to advanced hepatobiliary pathology during chronic O . viverrini infection . In the initial stages of model development , control individuals , APF+ individuals , and CCA individuals were included in the model to identify the potential relevant predictors of disease progression . This model identified age ( p = 0 . 0065 ) , serum IgG ( p = 0 . 0002 ) , and urinary 8-oxodG ( p < 0 . 0001 ) as significant and these predictors were subsequently considered in the two models comparing control individuals to either APF+ participants or CCA participants . In this second stage of model development , urinary 8-oxodG was retained as the single significant biological predictor of an individual having ultrasound confirmed APF ( p ≤ 0 . 0001 ) or histologically confirmed CCA ( p ≤ 0 . 0014 ) ) . Age and serum IgG were no longer significant in the model of APF+ participants ( p = 0 . 4802 and p = 0 . 2890 , respectively ) . In the model of CCA individuals age remained a significant predictor ( p < 0 . 0001 ) , while serum IgG was eliminated as non-significant ( p = 0 . 0587 ) . The final parsimonious models retain the concentration of creatinine-adjusted urinary 8-oxodG as the only predictive biomarker ( for diagnosis of both APF and CCA ) along with age as a significant covariate in individuals diagnosed with CCA ( equation 2 in S1 Text—Supplementary Equations and Definitions ) . The results of a logistic regression model that included urinary 8-oxodG as a predictor were considered alongside the actual ability of the clinical ( gold standard ) methods to identify specific increases in the level of 8-oxodG in urine of participants in the study . Increasing odds ratios ( ORs ) were associated with progressively higher concentrations of 8-oxodG , indicting the increased likelihood of progressing to APF and CCA for individuals with higher measurable concentrations of creatinine-adjusted 8-oxodG in the urine . The ORs and their 95% confidence intervals ( 95% CIs ) are summarized in Table 2 . Receiver operating characteristic ( ROC ) curves constructed from levels of creatinine-adjusted urinary 8-oxodG measured in study participants are presented in Figs 3 and 4 . The area under the ROC curve ( AUC ) describes the probability of correctly identifying a positive individual as a ‘case’ and a negative individual as a ‘non-case’ . An AUC of 1 would describe a diagnostic test that would correctly identify all cases and all non-cases 100% of the time . The 45-degree line in the ROC plot marks the “chance diagonal” , which corresponds to an ROC curve with an AUC of 0 . 50 [50] . Figs 3 and 4 are labeled with the AUC values of each ROC curve; the ROC curve generated from a diagnostic model of APF+ individuals has an AUC of 0 . 74 and the AUC of the ROC generated from a diagnostic model of CCA individuals is 0 . 88 . This means that levels of urinary 8-oxodG in urine correctly identifies individuals with O . viverrini-induced APF 74% of the time and correctly identifies individuals with O . viverrini-induced CCA 88% of the time . The diagnostic positivity thresholds were established as described in the methods section and are presented in Table 3 along with the relevant diagnostic validity parameters ( sensitivity and specificity ) , the PPV , NPV , LR+ , and LR- for the urinary 8-oxodG assay .
O . viverrini-infected individuals with advanced periductal fibrosis ( APF ) as determined by abdominal ultrasound had markedly elevated levels of urinary 8-oxodG compared to O . viverrini-infected individuals without APF . Moreover , the concentrations of 8-oxodG in the urine of individuals with APF were comparable to the highly elevated levels of this oxidatively modified DNA lesion in individuals with O . viverrini-induced CCA [28 , 31–33] . These results clearly suggest that elevated levels of this metabolite in urine are indicative of hepatobiliary fibrogenesis and tumorogenesis from chronic O . viverrini infection . Moreover , levels of 8-oxodG in the urine of O . viverrini infected individuals with APF or CCA individuals corroborated the ‘gold’ standard diagnostics used to detect both of these hepatobiliary pathologies in a dose-dependent manner: e . g . , the highest 50 unit increment of 8-oxodG ( 200 units ) indicated an increased risk of diagnosis of APF or CCA by 354% and 408% , respectively , compared to individuals with no detectable levels of urinary 8-oxodG . Furthermore , the risk models used to evaluate the utility of 8-oxodG as a biomarker were also used to identify the diagnostically relevant levels of 8-oxodG in the urine of study participants . The identification of these diagnostic threshold levels of urinary 8-oxodG , if corroborated in additional larger scale validation studies , would have important implications for urine 8-oxodG as diagnostic tool in field settings , with special significance in resource-limited settings , since they establish benchmarks that may be used to identify individuals at-risk of CCA and refer them for further testing ( e . g . , confirmatory abdominal ultrasound diagnosis ) and preventive chemotherapy . In order to assess the utility of 8-oxodG as a candidate biomarker for O . viverrini-induced APF and CCA , we also constructed a logistic regression risk model that initially included creatinine-adjusted urinary 8-oxodG levels along with significant demographic , behavioral , and immunological covariates associated with chronic O . viverrini infection . In this risk model , creatinine-adjusted urinary 8-oxodG emerged as the only significant predictor of APF , especially at higher concentrations of 8-oxodG . Creatinine adjusted 8-oxodG was also a significant predictor of CCA status when adjusted for age . Moreover , the risk model showed that 50 unit increases in creatinine-adjusted urinary 8-oxodG ( in ng/mg ) increased its ability to corroborate APF status by 137% and CCA status by 142% ( after adjusting for age ) and also helped to establish diagnostically relevant threshold levels of this biomarker . Moreover , elevated levels of urine 8-oxodG accurately identified progression to advanced stages of the disease ( Figs 3 and 4 ) , with high odd ratios ( ORs ) of APF and CCA associated with higher levels of this candidate biomarker ( Table 2 ) . As APF is precursor stage to CCA [36] , a simple , non-invasive assay for 8-oxodG in urine as a biomarker for this O . viverrini-induced pathology would be of profound benefit in Southeast Asia , especially among populations residing in the resource-limited settings of the Mekong Basin region , where the incidence of intrahepatic CCA is the highest in the world [51] . The methods and standards for measuring urinary 8-oxodG have received considerable attention[24 , 40] , with a consensus on using urine 8-oxodG as a biomarker established by the European Standards Committee on Urinary ( DNA ) Lesion Analysis ( ESCULA ) [24] . When properly detected , measured , and analyzed , ESCULA determined urine 8-oxodG to be a robust , accurate , reproducible , and ( “remarkably” ) stable urine biomarker for OS [24] . The production of 8-hyrdroyguanine is almost exclusively elicited by OS , with the main attack site by oxidative radicals at the N7-C8 bond [26 , 52 , 53] . While there are findings that suggest that diet contributes to urinary levels of thymine glycol and 8-oxo-7 , 8-dihydro-guanine ( 8-oxoGua ) [40] , little or no information has been reported that this applies to urinary levels of oxidatively modified 2′-deoxyribonucleosides ( 8-oxodG ) , hence obviating diet as a potential confounder . As with other studies , the present findings showed a relatively strong association between levels of urinary creatinine and levels of urinary 8-oxodG , which ( as explained by Barregard et al . [40] ) are likely due to differences in body mass index ( BMI ) , since metabolic rate is associated with lean body mass and a higher metabolic rate generates a larger amount of modified 2-deoxyribonucleoside products in urine [40]: e . g . , in general , males excrete more 8-oxodG than females per kilogram body weight [40] . Hence , following the recommendations of Barregard et al . [40] for using 8-oxodG in cross-sectional studies , we normalized urinary 8-oxodG concentrations by individual creatinine concentrations as determined by the Jaffe method [38] . Oxidative DNA damage as a key event in O . viverrini-induced CCA has been studied extensively in an animal model ( see [23] for review ) , where oxidatively damaged DNA bases formed along the inflamed intrahepatic biliary ducts , at sites adjacent to the parasite , perhaps from persistent wound repair [31 , 32] . Immunohistochemical studies in a hamster model of O . viverrini-induced CCA showed that inflammatory cells surrounding the parasite in the bile duct , such as mononuclear cells and eosinophils , generate reactive oxygen species ( ROS ) , which induce oxidative stress ( OS ) and increased cleavage of 8-oxodG [31 , 32] . Additionally , immunohistochemical analyses on livers resected from humans with O . viverrini-induced CCA showed 8-oxodG in tumor tissue from the bile duct [20 , 33] . Moreover , Thana et al . [33] observed elevated levels of 8-oxdG in the urine of individuals infected with O . viverrini compared to healthy controls as also observed in the current study ( Fig 2 Panel A ) . However , the current study adds to the literature the observation that urine 8-oxodG can be detect in markedly higher concentration when O . viverrini infected individuals who have progressed to APF or CCA ( Fig 2 ) . The observation that urinary 8-oxodG concentrations were similar in APF and CCA individuals support our hypothesis that common mechanisms drive bile duct fibrosis and bile duct tumorogenesis from chronic O . viverrini infection [8–10 , 12 , 13] . It is also in keeping with findings from other groups that chronic bile duct inflammation leads to a desmoplastic stroma ( i . e . , fibrosis ) in the bile duct that precedes CCA ( reviewed by Sirica and Gores [36] . As depicted in S1 Fig , elevated levels of 8-oxodG in the urine of APF individuals reflects ongoing tissue repair and the smoldering inflammatory milieu” [14] in the hepatobiliary epithelia from persistent injury from the parasite [15 , 30 , 54 , 55] . This “desmoplastic reaction” to chronic O . viverrini infection provides a rich niche for cancer cells to develop and progress [1 , 36 , 56 , 57] . Currently , therapeutic targeting to reduce desmoplastic stroma ( periductal fibrosis ) to prevent CCA is being investigated [36 , 56 , 57] . Taken together , these data show that urine 8-oxodG may be an excellent biomarker for the advanced hepatobiliary pathology that occurs prior to O . viverrini–induced CCA . In keeping with a recent position statement from the European Group on Tumor Markers [58] , the current manuscript places urine 8-oxodG as a “candidate biomarker” for APF and CCA in the “discovery phase” , i . e . , “when differential expression of a specific marker is shown to associate with a ‘gold’ standard clinical outcome” . The next phase of biomarker development for urine 8-oxodG is the “verification stage” , when our analyses would be extended to a much larger sample ( i . e . , hundreds ) of individuals infected with O . viverrini and at risk of CCA . The objective of the verification stage would be to incorporate the broadest range of cases and controls in order to capture the environmental , genetic , biological , and stochastic variation in the population in the Mekong Basin Subregion . As the diagnostic sensitivity of the candidate biomarker has been established herein , the verification stage would focus on the specificity of the candidate biomarker and the utility of the diagnostic threshold levels set forth in this manuscript [58] . If validated , levels of urinary 8-oxodG may be an inexpensive way to identify at-risk individuals , who may be referred for subsequent , more demanding testing . This process would streamline diagnostics for APF and CCA and perhaps improve the utilization of the limited resources allocated to cancer screening in this region .
The findings herein confirm previous observations that severe hepatobiliary disease occurs early and asymptomatically among residents in O . viverrini endemic areas . A simple , non-invasive assay targeting 8-oxodG in urine would be of profound benefit to populations in Southeast Asia , especially in the resource-limited settings of the Mekong Basin region countries of Thailand , Laos and Cambodia , where the incidence of O . viverrini-induced CCA is the highest in the world [7] . The future plan for the candidate biomarker includes moving to a verification step to test its accuracy in a larger sample size [58] . | Opisthorchis viverrini is a food-borne helminth infection that drives a strong inflammatory response in the bile duct that can result in bile duct fibrosis and bile duct cancer ( intrahepatic cholangiocarcinoma ) . Extensive research shows that oxidative stress ( OS ) plays a critical role in chronic O . viverrini infection transitioning to cancer in the bile duct . OS also results in a modified DNA lesion , referred to as 8-oxodG , excreted in the urine , where it can be detected by an antibody-based test . We measured the concentrations of 8-oxodG in the urine of O . viverrini-infected individuals who had developed bile duct fibrosis or bile duct cancer and compared levels of this metabolite in urine to O . viverrini infected individuals who did not have bile duct fibrosis or cancer in Northeastern Thailand . We determined bile duct fibrosis by ultrasonography and bile duct cancer by immunohistochemistry on resected liver tissue . We then built a statistical model to quantify how well urinary 8-oxodG predicted bile duct fibrosis and bile duct cancer in O . viverrini-infected individuals . We found that individuals with elevated levels of 8-oxodG in urine had a greater probability of developing bile duct fibrosis or bile duct cancer from O . viverrini infection . This association occurred in a strongly dose-dependent manner: in other words , the O . viverrini-infected individuals who had the highest concentration of urinary 8-oxodG also had the highest risk of presenting with bile duct fibrosis or bile duct cancer . In summary , measuring levels of 8-oxodG in the urine offers a unique opportunity to develop a candidate biomarker for advanced O . viverrini induced hepatobiliary pathologies such as fibrosis and cancer . The findings also confirm our previous observations that severe liver pathology occurs early and asymptomatically in residents of O . viverrini endemic regions , where individuals are infected for years ( often decades ) with this food-borne neglected tropical diseases ( NTD ) pathogen . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion",
"Conclusion"
] | [] | 2015 | Levels of 8-OxodG Predict Hepatobiliary Pathology in Opisthorchis viverrini Endemic Settings in Thailand |
DnaK is a molecular chaperone that has important roles in protein folding . The hydrolysis of ATP is essential to this activity , and the effects of nucleotides on the structure and function of DnaK have been extensively studied . However , the key residues that govern the conformational motions that define the apo , ATP-bound , and ADP-bound states are not entirely clear . Here , we used molecular dynamics simulations , mutagenesis , and enzymatic assays to explore the molecular basis of this process . Simulations of DnaK's nucleotide-binding domain ( NBD ) in the apo , ATP-bound , and ADP/Pi-bound states suggested that each state has a distinct conformation , consistent with available biochemical and structural information . The simulations further suggested that large shearing motions between subdomains I-A and II-A dominated the conversion between these conformations . We found that several evolutionally conserved residues , especially G228 and G229 , appeared to function as a hinge for these motions , because they predominantly populated two distinct states depending on whether ATP or ADP/Pi was bound . Consistent with the importance of these “hinge” residues , alanine point mutations caused DnaK to have reduced chaperone activities in vitro and in vivo . Together , these results clarify how sub-domain motions communicate allostery in DnaK .
Escherichia coli DnaK is a member of the heat shock protein 70 ( Hsp70 ) family of molecule chaperones that assists in protein folding and minimizes protein aggregation [1]–[4] . Because of its central role in the proteostasis network , DnaK has been suggested as a promising new anti-bacterial target , and its human ortholog , Hsp70 , is a drug target for the treatment of cancer [5] , [6] and neurodegenerative disorders [7] , [8] . These observations have led to an increased interest in understanding the structure and function of Hsp70/DnaK . DnaK , like other Hsp70s , consists of a nucleotide-binding domain ( NBD ) and a substrate-binding domain ( SBD ) tethered by a flexible linker . The NBD is composed of four subdomains , I-A , II-A , I-B and II-B , arranged to form a nucleotide-binding cleft that belongs to the actin/hexokinase/Hsp70 superfamily ( Figure 1A ) [9] , [10] . The NBD binds and hydrolyzes ATP , and this activity is important for chaperone functions . Specifically , nucleotide turnover in the NBD regulates binding of misfolded proteins in the SBD through an inter-domain allosteric network [7] , [11]–[13] . The ATP-bound form of DnaK has an “open” SBD that binds loosely to misfolded proteins , while nucleotide hydrolysis in the NBD re-arranges the SBD and increases the affinity for proteins . Thus , cycling through the ATP- and ADP-bound states appears to be important in DnaK allostery and the productive refolding of denatured proteins [14] . Crystallography and NMR have been used to provide important insights into the effects of nucleotides on the structure and function of DnaK . For example , a comparison of the crystal structures of the NBD in the apo form ( 1DKG ) [15] and the ADP-bound form ( 1BUP [16] and 1KAZ [17] ) suggests a substantial , nucleotide-dependent movement in subdomain II-B ( Figure 1A ) . This motion appears to involve rotations of subdomain II-B in relation to subdomain II-A , which is mediated by a sheet-coil-helix element ( residues 222–234 ) . NMR studies have further suggested that this region may act as a hinge for subdomain motions [18] , [19] , and more recent structural studies on an ATP-bound form of DnaK [20] further support this idea . In order to understand the specific role of the sheet-coil-helix element in this hinge motion , Chang et al . [21] mutated a series of residues between subdomains II-A and II-B . This work identified a number of mutations that disrupted allostery and chaperone function in vitro and in vivo . However , the underlining molecular mechanisms of these mutations are not yet clear . Computational simulations have further advanced our knowledge of allostery in Hsp70/DnaK . For example , coarse-grained molecular dynamics of full-length Hsp70 demonstrated the collective motions that are essential in the allosteric communications between the NBD and SBD [22] , [23] . Likewise , all-atom simulations identified residues essential in the binding of DnaK to nucleotide-exchange factors ( NEFs ) [24] , and the molecular mechanism that relays the allosteric communication between the NBD and SBD [25] . Together , these studies have provided insight into the mechanism of DnaK and , more broadly , this system has contributed to our understanding of dynamics and allostery in biology . Despite these insights , the molecular mechanisms of hinge motions in DnaK's NBD are still not clear . Specifically , it is not known which residues are essential to the hinge motions and how these molecular motions are affected by nucleotides . It is also not clear whether disrupting these hinge motions would decrease inter-domain communication between the NBD and SBD . Here , we have used dynamics simulations and mutagenesis to examine the detailed communication between the protein and the nucleotide . Conformational behavior during the simulations pointed to the allosteric importance of certain residues . These residues were then compared to their evolutionary conservation , which further highlighted the importance of the identified residues . We found that hinge residues , including G228 and its neighbors , are key regulators of this transition . They were nearly invariant across bacteria , plants , and animals , which highlights the incredible importance of these residues for chaperone function . These results provide a detailed molecular mechanism linking ATPase activity to structural transitions in DnaK/Hsp70 .
To initiate these studies , models of DnaK in different nucleotide states ( apo , ATP- , and ADP/Pi-bound ) were constructed based on the crystal structure of the apo DnaK NBD ( PDB: 1DKG ) . The nucleotide-bound models were derived from simulations of the apo state with cofactors introduced because structures of these states were not yet available at the time of this study . Since 1DKG does not carry nucleotide and metal ions , these cofactors were introduced from structures of bovine Hsc70 ( PDB: 1BUP and 1KAZ ) . To enhance the conformational sampling , a Generalized Born , implicit-solvent model and Langevin dynamics ( LD ) were used for the simulations . Then , for each model , five independent trajectories were generated and confirmed to be structurally stable ( Figures S1 and S2 ) . Multiple , short trajectories were chosen to enhance sampling because it has been shown to be a more effective strategy than using a single , long-trajectory simulation [26] . When all restraints were removed , the NBD bound to either ATP or ADP/Pi spontaneously converted from the initial “open” conformation of an apo state to a “closed” conformation in which the subdomains I-B and II-B moved toward each other . These closed models agree well with recently reported structures that were unavailable at the start of the project . The RMSD were ∼3 . 2 Å and ∼4 . 0 Å to structures 4B9Q [27] and 4JNE [28] , respectively . These values are appropriate for a very large , flexible , multi-domain protein , but the variation does point to an important limitation . To achieve the large conformational changes necessary , continuum solvent was used , and the electrostatics and salt bridges could be over stabilized , which may slightly alter the closed conformation . Previous dynamics simulations of DnaK and its homologs bound to various nucleotides also show generally rigid motions of IIB relative to IB and IIA [22]–[25] , [29] . These motions were primarily driven by movement of the entire subdomain II-B , such that the global conformation of the NBD can be described by the relative position of the intact subdomains . Thus , the distances between the center-of-mass ( COM ) of II-B and subdomains I-A/B were used to quantify and compare the “openness” of each conformation ( Table 1 ) . Comparing the apo , ATP- , and ADP/Pi-bound states of NBD , we observed that the apo state adopts the most “open” conformation , with an average inter-subdomain distance of ∼32 Å . In contrast , the ATP- and ADP/Pi-bound states adopt closed conformations , with average distances of 27 . 3 and 26 . 1 Å , respectively . In some of the LD trajectories , the α-helix of subdomain II-B ( residues 230–280 ) was observed to “bend” near residues 262–267 ( Figure 1B ) . This bending did not appear to be dependent upon the nucleotide state because it was observed in the simulations of the apo , ATP- , and ADP/Pi-bound NBDs . Similar helix bending has also been described in actin and a phosphatase with a similar fold [9] . Although the functional importance of this transition is not clear , it may be involved in assisting with the full opening and closing of the NBD . To study conformational changes in the NBD , essential dynamics ( ED ) was used to examine the LD trajectories of each model . Previous studies have shown that combinations of several independent dynamics simulations often provide sufficient sampling of conformational space for ED [30] . Thus , five independent LD trajectories were combined for each model and analyzed with ED . In all cases , the first 5 eigenvectors ( first – fifth ) represented ∼80% of the principle motions of the protein ( Figure S3 ) . The apo and nucleotide-bound models yielded generally similar global motions , and the lowest mode of the ED was a shear motion between subdomains I and II ( Figure 2A ) . This motion involved two crossed α-helices ( 171–179 and 367–377 ) with inter-helical angles of approximately 70° to 80° . This shear motion occurred in the interface between subdomains I-A and II-A , and it manifested as a large displacement between subdomains I-B and II-B . Actin and glucokinase , which both have similar folds to DnaK , have been shown to undergo a similar shear motion [31] , [32] . The second eigenvector involved a rotational motion of subdomain II-B , in which II-B moves relatively independent to the other subdomains ( Figure 2B ) . This motion appeared to be a hinged-motion between subdomains II-A and II-B . Together , the first and second eigenvectors described motions involved in the opening and closing of the nucleotide-binding site , and these motions are consistent with experimental observations from NMR [19] . The movement of protein domains is often linked to protein function [33] , and these movements typically involve the motion of rigid domains relative to each other [31] , [34] . These structural changes cluster into two major mechanisms: shear movements in which domains slide along one another while maintaining closely packed interfaces , and hinged movements in which relatively disordered regions connecting the domains undergo significant backbone conformational changes [31] . Hinged movements usually involve a small set of residues that adopt distinct torsion angles of their φ ( C′-N-Cα-C′ ) and ψ ( N-Cα-C′-N ) bonds , enabling these individual “hinge” amino acids to mediate larger motions between domains or subdomains [34] . Therefore , it is often possible to identify the key hinge residues by examining which amino acids undergo changes in their φ-ψ torsion angles during dynamic simulations . One of the key questions in the DnaK system is which residues might be involved in the nucleotide-dependent hinge motions . To detect such hinge residues , we performed Cα torsion angle analysis on the NBD trajectories ( Figure 3A and 3B ) . Our Ramachandran results of the amino acids are similar to those described by Hovmöller et al . [35] , in which residues in the helix , sheet , and random coil region have different φ-ψ distributions . As expected , we found that the great majority of residues , such as A191 and K245 , do not have the properties of hinged movement , such that the Ramachandran plot of these residues show a single φ-ψ torsion angle cluster throughout the simulations . In contrast , a small number of residues , such as G74 , had more than one conformation evident from multiple φ-ψ clusters , which suggests that they may be involved in hinged movement ( Table 2 ) . Further , a subset of these residues ( N13 , N14 , P67 , A68 , I73 , R84 , D85 , T184 , T185 , G223 , L227 , G228 , and G229 ) had multiple conformations with distinct φ-ψ clusters . Further , some of these φ-ψ clusters were correlated with changes in nucleotide state . Next , we examined the location of the residues with multiple φ-ψ clusters on the DnaK NBD structure . These residues are found in all 4 subdomains with a bias for the flexible loops between subdomains I-A and I-B ( Figure 3C ) . Additionally , several clusters were grouped in subdomain II-A , including residues 183–186 and 340–342 , which form flexible loops , and residues 195 , 198 and 199 , which are found in a flexible β-hairpin that coordinates with the phosphates of the nucleotide ( Figure 3D ) . In DnaK , the lobe between I73/G74 and E104/V105 consists of a helix-loop-sheet structure . We found that this lobe is capped by two sets of residues with multiple φ-ψ clusters ( Table 2 and Figure 3C ) , suggesting that these residues might function as hinges . This movement might be interesting because this lobe is directly adjacent to the nucleotide-interacting residue R71 , which plays a critical role in enzymatic activity . Finally , another significant cluster of residues with multiple φ-ψ clusters was identified near the junction between the subdomains II-A and II-B . These clusters included residues in β-sheets ( I202 , S203 ) and random coils ( G223 , L227 , G228 , G229 ) . The β-sheet residues were not sensitive to nucleotide state ( Figure 4 ) , implying that they may be involved in intrinsic protein dynamics . However , residues G223 , L227 , G228 , and G229 had multiple φ-ψ states , and their states changed in response to nucleotide ( Figures 5A–E ) . These residues seemed like good candidates for playing a key role in nucleotide-dependent structural transitions and allostery , as suggested by Chiappori et al . [25] . As a complementary way to understand how nucleotides might affect motions in DnaK , the correlation matrices of each trajectory from the LD simulations were compared . Then , residues with motions strongly correlated with ATP or ADP were extracted and ranked according to their degree of absolute correlation ( 1 . 0 correlation was just as important as −1 . 0 anti-correlation ) . The 60 most-correlated residues were tallied for each LD trajectory . Residues that were highly correlated in at least 4 of the 10 independent trajectories were considered significant ( Table 3 ) . Residues with weak or no correlation to the nucleotides were largely found on the surface of the NBD and at the flexible loops at the distal ends of the subdomains . A number of residues that have multiple φ-ψ states , such as P67 , E104 , and A288 , do not have strong correlation to the nucleotides and were not considered for further study . In contrast , residues with strong correlation were found in the interior of all four subdomains . As expected , most of these residues were in the vicinity of the nucleotide-binding site ( <8 Å from the nucleotide ) , but a few were further away ( Figure 6A ) . Comparing these residues to those identified in the φ-ψ torsion angle analysis , we found that several residues are present in both lists , including G51 , A68 , L195 , G223 , G228 , G229 , and V340 ( Figure 6B ) . An examination of these residues suggested that they may have significant effects on the functionality of NBD . For example , A68 is adjacent to K70 and R71 , key residues that bind to nucleotides [36] . Likewise , L195 and V340 may interact with I202 and S203 to define the shape of the nucleotide-binding site . Residues G223 , G228 , and G229 , were especially interesting because ( as mentioned above ) they are located away from the nucleotide-binding site and situate in a random coil that tethers subdomains II-A to II-B . To further explore the potential importance of the “hinge” residues , we studied which residues had correlated motions with G223 , G228 and G229 . Dynamics simulations of NBD in the apo , ATP-bound , and ADP/Pi-bound states were examined , and among the top 60 strongly correlated and anti-correlated residues , we found strong enrichment of components of the known proline switch that consists of residues K70 , R71 , P143 , A144 , Y145 , F146 , R151 and E171 . The proline switch has previously been found to regulate allostery in Hsp70 chaperones [36] . In simulations of the NBD in all the apo- , ATP- and ADP/Pi-bound states , K70 and R71 were consistently correlated with each of the “hinge” residues ( i . e . , G223 , G228 , and G229 ) . In contrast , P143 , A144 , Y145 , F146 , R151 , and E171 correlated to G228 more frequently in the ATP-bound and ADP/Pi-bound states than in the apo-state . This result might be expected because NBD in the apo-state simulations had a wider range of movements ( Figure S1 ) . In all cases , correlations between the hinge residues and the proline-switch residues appeared more frequent in the ATP-bound state than in the ADP/Pi-bound state . These results suggest that interactions among ADP , Pi , the hinge residues , and the proline-switch residues may be looser than when ATP is bound . To understand whether the proposed φ-ψ “hinge” residues might be preferentially conserved among Hsp70 family members , BLASTP [37] was used to align the sequences of DnaK and all other available members of the Hsp70 family ( 3700 orthologs ) from the GenBank database . Several residues in the random coil that tether subdomains II-A and II-B ( G223 , L227 , G228 , and G229 ) were highly conserved among bacteria , with over 90% identity and nearly 100% similarity . Remarkably , high conservation was also observed across all kingdoms , particularly for L227 , G228 , and G229 ( Table 4 ) . This level of conservation was greater than expected for the average residue in these chaperones; the overall conservation between the prokaryotic E . coli DnaK and the eukaryotic human Hsp70 is ∼50% . These findings suggest that the highly conserved residues might be especially important for the function of Hsp70 family members . Next , we designed mutations of residues in DnaK to understand their potential role in allostery and chaperone functions . In this effort , we focused on residues that met two criteria: those that were highly conserved ( >90% identity in bacteria and animals , Table 4 ) and those that were also predicted by the simulations to be involved in φ-ψ hinge motions ( see Table 2 and Figures 4 and 5 ) . The resulting five residues ( Table 5 ) were placed into two groups , depending on whether their motions were correlated with nucleotide . We then expressed and purified wild type DnaK and a series of DnaK mutants in which each of the proposed hinge residues were replaced with an alanine [21] . All of the purified proteins were properly folded as determined by thermal stability and circular dichroism measurements ( Figure S4 ) , suggesting that the alanine mutations did not damage the overall fold or stability of the chaperone . To assess the effects of the mutations on function , we measured the ATPase activity of the mutants using a malachite-green assay [38] . We also measured the ability of the chaperone to refold denatured firefly luciferase , which is an assay that is commonly used as a measure of chaperone functions in vitro [39] . Changes in the ATPase activity of DnaK are expected to impact luciferase refolding because this process requires multiple cycles of binding between DnaK and the denatured luciferase [14] . Thus , ATP cycling that is either too fast or too slow might imbalance the dwell time of DnaK on luciferase and decrease overall folding efficiency [21] . Further , extensive studies have revealed that the NBD and SBD of DnaK communicate through inter-domain interactions [18] , [20] , [40] . Thus , if the putative hinge residues are important in controlling NBD allostery , mutations in these positions would be expected to disrupt normal ATP turnover and , in turn , decrease chaperone functions , such as protein refolding , both in vitro and in vivo . Importantly , this system appears to be highly sensitive to even modest changes in ATPase activity . For example , chemical compounds that impact steady-state ATP turnover by only 20–50% [41] , [42] or single turnover by only about 20% [43] , [44] have profound effects on Hsp70/DnaK biology in cellular assays measuring the stability of chaperone protein substrates . These experiments revealed that mutations I202A and L227A decreased ATPase activity by ∼30% and refolding function by ∼20% , suggesting that these highly conserved residues are important for allostery in DnaK . Mutation G223A had no effect on ATPase activity , but it did have a modest effect on refolding activity ( ∼80% of wild type ) . The S203A mutant had normal refolding activity , but it had accelerated ATPase function ( ∼180% of wild type ) . Finally , G228A had dramatically enhanced ATPase activity ( ∼160% of wild-type ) , yet it had only 16% of the wild-type refolding activity . The results with G228A were particularly striking , supporting the idea that G228 is an important hinge residue . G228 is located between subdomain II-A and II-B , and it likely dictates the relative movement of the two subdomains , as observed in the second mode of the essential dynamics of NBD ( see Figure 2B ) . It is situated between residue L227 and G228 , which also show altered φ-ψ behavior based on the nucleotide-bound state in Figure 5 . These experimental results confirmed the importance of the residues identified in the LD simulations . To test whether these mutations might disrupt the physiological functions of DnaK , we next expressed the hinge mutants in E . coli ( DE3 ) cells that lack endogenous chaperone ( Δdnak ) [21] . DnaK is normally required for growth of bacteria at elevated temperatures ( ∼37°C ) , so we measured their growth under stress conditions to assess whether the point mutant could recover wild-type DnaK activities . Compared to cells in which WT DnaK was restored , we found that the mutations that damaged luciferase-refolding activity ( I202A , L227A , G223A , and G228A ) had a significantly reduced ability to recover bacterial viability ( Table 5; growth was 75% , 25% , 22% , and 13% of WT , respectively ) . Conversely , the S203A DnaK mutant , which had little effect on luciferase refolding , had nearly normal viability ( 110% of WT ) . These results support the importance of the hinge residues , especially G228 , on DnaK functions . These results also support a model in which luciferase refolding activity is correlated with the global , in vivo functions of DnaK [21] . Nucleotide-dependent structural transitions in DnaK are linked to its chaperone functions . While there is much known about the allostery between the NBD and SBD in DnaK [18] , [22] , [45] , there is less known about the key residues involved in the subdomain motions of the NBD and how they impact NBD-SBD communication . Using dynamics simulations of DnaK's NBD in the apo , ATP- , and ADP/Pi-bound states , we identified shear motions between subdomains I-A and II-A and a hinge motion that is consistent with recent reports [19] , [20] . Together , these motions result in a dramatic movements of the nucleotide-binding cleft , which is likely important for nucleotide cycling and for communicating the nucleotide state to the SBD . Using the results of these simulations , we further identified residues , including G223 , L227 , G228 , and G229 , as being potentially important to the hinge motions . These residues are very highly conserved through evolution , and they are located at what appeared to be a critical hinge region . These results were supported by torsion angle analyses , which suggested that these residues adopt two distinct backbone conformations . For some of these residues , these clusters correlated with nucleotide state . Finally , mutation of these residues largely supported this model and showed that chaperone functions were misregulated when the hinge residues were mutated to alanine . Together , these studies point to specific hinge residues that are nearly invariant across all kingdoms of life and are important for the nucleotide-dependent motions in DnaK . These findings are important for understanding structure and function in DnaK .
Coordinates of E . coli DnaK NBD in complex with the nucleotide-exchange factor , GrpE ( PDB: 1DKG [15] ) were obtained from the PDB , and GrpE was discarded . The missing side chains and short loops of NBD were introduced using Molecular Operating Environment [46] version 2005 . 06 . AMBER 10 [47] was then used to perform unrestrained all-atom LD simulations . Models of DnaK NBD in different states ( apo- , ATP- , and ADP/Pi-bound ) were built for these simulations . The structure 1DKG was crystallized without bound cofactors , so the essential ions , ATP , and ADP/Pi were introduced by transferring the coordinates from the crystal structures of the closely related homolog , bovine Hsc70 . The cofactors ADP , PO43− , Mg2+ , and K+ were obtained from the 1BUP [16] structure whereas ATP , Mg2+ , and K+ ions were obtained from 1KAZ [17] . Note that the ions are not explicit counter ions; they are bound to the protein and play a structural role . The protein was modeled using the FF99SB force field [48] . Parameters for Pi and Mg2+ were generated using the ANTECHAMBER module of AMBER with GAFF [49] , [50] and AM1-BCC charges [51] , [52] . Parameters developed by Meagher et al . [53] were used for ATP and ADP . The SHAKE algorithm [54] was used to restrain hydrogen atoms . Model II ( igb = 5 ) of a modified Generalized Born approach [55] was used to implicitly model aqueous solvation . Collision frequency of 1 ps−1 was used . A non-bonded interaction cutoff of 999 Å was used . Default dielectric values were used: interior = 1 and exterior = 78 . 5 . Five independent LD simulations for each of the NBD states ( apo , ADP/Pi-bound , and ATP-bound ) were initiated with different random-number seeds . Hydrogen atoms were first minimized , followed by residue side chains , and finally an all-atom minimization . Heating and restrained equilibrations were followed , in which the system was heated gradually from 100 to 300 K during the first two equilibration phases , and the temperature remained at 300 K for the remaining equilibration and production phases . Restraints were placed on all heavy atoms and gradually relaxed over the first four equilibration phases , using force constants from 2 . 0 to 0 . 1 kcal/mol·Å2 . Restraints were maintained on backbone atoms in the fifth equilibration phase , using a force constant of 0 . 1 kcal/mol·Å2 . All restraints were removed in the sixth phase . The first three phases were performed for 20 ps each , followed by 50 ps for the fourth and fifth . The sixth phase , unrestrained equilibration , was run for 740 ps , and the production phase was run for 5 ns . A time step of 1 fs was used , and snapshots were collected every 1 ps . Results of the LD simulations were analyzed by calculating the variance-covariance matrix of the trajectory using the ptraj module of AMBER . In-house scripts were used to parse subsets of data and to plot the correlation results for visualization and analysis . The E . coli dnaK genes were amplified by PCR using Platinum Pfx DNA polymerase and inserted into the pMCSG7 plasmid as Ndel-HindIII fragment through ligation-independent cloning [56] . The primers for site-directed mutagenesis ( see the Supplemental Information ) were designed based on the report of Zheng et al . [57] , and QuikChange mutagenesis was carried out following the manufacturer protocol . Based on the simulations , the following dnaK mutants were made: I202A , S203A , G223A , L227A , and G228A . Each mutant and the wild-type ( WT ) DnaK was expressed as an N-terminal His-tagged protein and purified from BL21 ( DE3 ) cells as previously described [21] , [38] . Purity of the proteins was greater than 90% , as judged by Coomassie staining . Circular dichroism spectra were consistent with the structure of folded NBD . Proteins were frozen on liquid nitrogen and stored at −80°C until use . ThermoFluor is a fluorescence-based thermal shift assay system that assesses protein stability [58]–[60] . To test the relative stability of DnaK and its mutants , each protein was diluted to 0 . 2 mg/mL in either buffer A ( 100 mM sodium phosphate , 5 mM MgCl2 , 20 mM KCl , pH 7 . 2 ) or buffer B ( 100 mM HEPES , 5 mM MgCl2 , 20 mM KCl , pH 7 . 2 ) , as indicated . To these solutions was added ADP or ATP ( 1 mM ) and 1 , 8-ANS ( 100 µM ) . These samples were placed in 384 well plates and covered with silicone oil to minimize evaporation . The fluorescence of ANS was monitored with excitation/emission at 375/480 nm . Measurements were performed in continuous mode , using a temperature range of 30 to 85°C and increments of 1°C . ATPase activity , firefly luciferase refolding assays , and bacterial growth studies were carried out as previously described [21] . Briefly , steady-state phosphate release from ATP was measured using a malachite green assay . Chaperone-dependent refolding was measured by Steady Glo after incubation of chemically denatured firefly luciferase with DnaK ( or DnaK mutants ) , DnaJ and ATP . For the bacterial growth assays , the DnaK point mutants were expressed under control of a T7 promoter ( pMCSG7 ) in E . coli Δdnak ( DE3 ) cells . Mutant strains expressed approximately equal amounts of either WT or mutant DnaK , as judged by Coomassie stains . The growth of bacteria expressing the mutants was assessed by measuring optical density ( OD600 ) at 37°C for 4 hours . Results are reported as a percentage of the growth of ΔdnaK cells in which WT DnaK was restored from the pMCSG7 vector . | DnaK belongs to the highly conserved heat shock protein 70 ( Hsp70 ) family , a group of ATP-dependent molecular chaperones that regulates proteostasis . Studies have suggested that global movements of the subdomains in the nucleotide-binding domain ( NBD ) of DnaK regulate its catalytic activity . However , there is less known about the key residues involved in these subdomain motions and whether these residues might also regulate inter-domain allostery with the substrate-binding domain ( SBD ) . To examine the motions in the NBD , dynamics simulations of DnaK's NBD in the apo , ATP-bound , and ADP/Pi-bound states were performed . Through essential dynamics and torsion angle analyses , we identified motions and highly conserved hinge residues between subdomains IIA and IIB that are likely to be important for nucleotide cycling and for communicating the nucleotide state to the SBD . Supporting this model , mutating these conserved hinge residues affected ATPase activity and chaperone functions in vitro and in bacteria , suggesting their importance in the nucleotide-dependent motions in DnaK . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Identification of Key Hinge Residues Important for Nucleotide-Dependent Allostery in E. coli Hsp70/DnaK |
The relationships between heterogeneities in host infection and infectiousness ( transmission to arthropod vectors ) can provide important insights for disease management . Here , we quantify heterogeneities in Leishmania infantum parasite numbers in reservoir and non-reservoir host populations , and relate this to their infectiousness during natural infection . Tissue parasite number was evaluated as a potential surrogate marker of host transmission potential . Parasite numbers were measured by qPCR in bone marrow and ear skin biopsies of 82 dogs and 34 crab-eating foxes collected during a longitudinal study in Amazon Brazil , for which previous data was available on infectiousness ( by xenodiagnosis ) and severity of infection . Parasite numbers were highly aggregated both between samples and between individuals . In dogs , total parasite abundance and relative numbers in ear skin compared to bone marrow increased with the duration and severity of infection . Infectiousness to the sandfly vector was associated with high parasite numbers; parasite number in skin was the best predictor of being infectious . Crab-eating foxes , which typically present asymptomatic infection and are non-infectious , had parasite numbers comparable to those of non-infectious dogs . Skin parasite number provides an indirect marker of infectiousness , and could allow targeted control particularly of highly infectious dogs .
Studies of microparasites usually consider hosts as homogeneous infection units ( infected or uninfected ) , despite knowledge that infections progress through states of clinical severity , that clinical severity is often associated with the number of infecting microorganisms ( load ) , and that individual transmission potential may be related to infection load . The significance of “super-spreaders” responsible for spreading infection to a disproportionate number of secondary cases has long been recognised [1] , [2] , however the relationships between parasite load and transmission are rarely measured; even in well-studied macroparasites ( e . g . helminths ) infectiousness is assumed to correspond to worm burden and egg count [3]–[6] . Variations in individual infection loads tend to be characterised by right-skewed ( over-dispersed or aggregated ) frequency distributions . Over-dispersion translates into diminishing proportions of the host population harbouring disproportionately higher infection loads . Where transmission potential is directly related to infection load , over-dispersed distributions may be interpreted as a small fraction of the population being responsible for most transmission , giving rise to the “20/80 rule” ( whereby 20% of cases cause 80% of transmission ) , proposed for a number of parasitic agents ( e . g . [7]–[10] ) . Heterogeneity in transmission can increase the basic case reproduction number R0 of a pathogen compared to that under assumptions of homogeneous mixing or density-dependent contact networks [9] , [11] , and affect the effort required , and choice of strategy ( mass or targeted ) , to interrupt transmission [7]–[9] , [12] . Molecular techniques , such as real-time quantitative PCR ( qPCR ) , have been used recently to differentiate between infected individuals and to help understand the spread and treatment of emerging infectious diseases e . g . [2] , [9] , [13]–[15] , nevertheless few empirical studies relate individual infection loads to transmission . Zoonotic visceral leishmaniasis ( ZVL ) is a fatal disease of humans and canids caused by the protozoan parasite Leishmania infantum , and transmitted between hosts by Phlebotomine sandflies . The domestic dog is the only proven reservoir [16] , though severity of infection and infectiousness varies greatly between individuals; in humans and wild mammals the majority of infections are asymptomatic and non-infectious [16] . Control of ZVL focuses on the detection and elimination of infected dogs ( particularly in South America ) , indoor residual spraying of insecticide , and human case treatment [17] . Positivity to serum anti-Leishmania antibodies is the principal criterion for mandatory slaughter of dogs [17] . Analyses indicate that this policy has little impact on reducing ZVL incidence , though robust data are lacking [16] , and there have been calls to re-evaluate the ZVL control program in Brazil [16] , [18]–[21] . Contributing factors to the lack of effectiveness include delays between testing and slaughter , low test sensitivity [22] , and significant dog-owner non-compliance [21] . An alternative strategy could be to target infectious rather than infected dogs , providing infectious hosts can be identified . Direct measurement of infectiousness by xenodiagnosis requires blood-feeding of colony-reared sandflies on hosts followed by screening for parasite infections in the vector . Rearing large quantities of vectors for community surveillance however is not practical . Tissue parasite loads have the potential to provide a reliable indirect marker of infectiousness [23]–[30] , though no studies have tested these relationships through the time course of infection . Here we measure L . infantum loads in cohorts of naturally infected domestic dogs Canis familiaris and crab-eating foxes Cerdocyon thous in Amazon Brazil . This study is unique in being able to relate host tissue parasite loads to serial xenodiagnosis from time of natural infection . The aims were ( i ) to characterize the heterogeneities in L . infantum loads between sampled tissues and between individual hosts with different severity of infection , ( ii ) to investigate whether tissue parasite loads can predict infectiousness to the sandfly vector; ( iii ) to compare parasite loads between dogs and crab-eating foxes , and ( iv ) to evaluate the performance of qPCR and ELISA diagnostic assays to identify infectious animals in mixed populations .
Canine samples were collected with informed consent from dog owners . Sampling was performed in accordance with UK Home Office guidelines . Dog samples were available from −80°C archived material generated in a cohort study of naturally exposed dogs between April 1993 and July 1995 in the municipality of Salvaterra , Marajó Island , Pará State , Brazil , in which bone marrows aspirated from the iliac crest and 3 mm skin biopsy punches of the ear pinnae outer edge were sampled repeatedly at approximately 2 month intervals for up to 27 months post initial exposure [31] . Ear skin was the preferred skin sample since it is reported to be more infectious to sandflies than abdomen skin [23] , [30] . Both skin and bone marrow are reported to be more sensitive than blood for parasitological and molecular detection of L . infantum , and higher qPCR counts are recorded in bone marrow than in blood [32]–[34] . For the present study , 265 bone marrow samples were available from 82 infected dogs ( 1–10 samples per dog ) , and 185 ear skin biopsy samples were available from 64 infected dogs ( 1–6 samples per dog ) , of which 173 samples from 63 dogs had paired bone marrow samples . Fox samples were collected during a concurrent longitudinal study of sympatric marked-recaptured free-ranging foxes [35] . Here , 67 bone marrow samples from 34 infected foxes , and 51 ear biopsy samples from 30 infected foxes , were available; all ear biopsy samples had paired bone marrow samples . Dog samples were collected with informed consent from dog owners . Dog and fox samples were assayed at all , or at the majority , of time-points , for ( i ) anti-Leishmania IgG by ELISA using crude leishmanial antigen ( CLA ) , with antibody concentrations expressed as arbitrary units/mL relative to a positive control serum [31] ( n = 277 samples ) ; ( ii ) PCR on bone marrow biopsies using primers specific for kinetoplast DNA ( kDNA ) and ribosomal RNA [36] ( n = 277 samples ) ; ( iii ) rK39 Kalazar Detect Rapid Diagnostic Test ( RDT ) , Inbios International Inc . , WA . , USA [37] , ( iv ) qPCR primers for kDNA ( described below ) , and ( v ) clinical score , defined as the sum of the score of six typical clinical signs ( alopecia , dermatitis , chancres , conjunctivitis , onychogryphosis , and lymphadenopathy ) , each scored on a semi-quantitative scale from 0 ( absent ) to 3 ( intense ) [36] ( n = 266 samples ) . Animals were assessed for infectiousness to the sandfly vector by xenodiagnosis , using uninfected colony-reared Lutzomyia longipalpis , and following dissection 4–5 days post full engorgement [22] , [35] . Here , matching xenodiagnosis data were available for 103 dog bone marrow samples ( 36 infected dogs , 3 , 751 fed flies dissected ) , 58 dogs ear samples ( 26 infected dogs , 1 , 702 flies ) , 39 fox bone marrow samples ( 22 infected individuals , 1 , 309 flies ) , and 30 fox ear samples ( 18 foxes , 1 , 187 flies ) . DNA was extracted from 100 µL aliquots of bone marrow , using phenol-chloroform [38] . DNA from 3 mm ear skin punch biopsies ( average: 0 . 029 grams , range: 0 . 0144–0 . 0837 ) was extracted using a commercial kit ( DNeasy: Qiagen , UK ) . qPCR was performed using primers specific for a conserved region of Leishmania kDNA [27] . Quantification of Leishmania DNA was performed by comparison of Ct values with those from a standard curve constructed from 10-fold dilutions of L . infantum DNA extracted from cultured parasites , from 1×105 to 0 . 001 parasite equivalents/mL ( strain MHOM/MA/67/ITMAP-263 ) . Samples were tested in duplicate and standards in triplicate on every plate . The occasional duplicates giving one positive and one negative result were re-tested: none remained unresolved after re-testing . A non-template control ( NTC ) was run in triplicate on every plate . A plate of negative controls including DNA extracted from blood samples of 30 UK dogs with no history of foreign travel , and 40 endemic control dogs from São Paulo , Brazil was run every 5 plates . A standardised Ct threshold value of 0 . 01 was selected as cut-off value to define infection based on the NTC signal . The endogenous control was a eukaryotic 18S rRNA gene as a reference of total canine DNA quantified in a separate qPCR reaction to the Leishmania assay using pre-developed TaqMan Assay reagents ( Applied Biosystems , UK ) following the manufacturer's recommendations . Parasite loads were normalized ( d ) between animals to the eukaryotic 18S rRNA gene per reaction , where d = absolute Leishmania kDNA equivalents/ ( copy number of 18S rRNA gene/2 ) /ng tissue DNA extracted measured spectrophotometrically . Normalized log10 parasite numbers and absolute log10 parasite numbers per ml ( bone marrow ) or per gram ( ear skin ) were strongly correlated ( r2 = 0 . 93 and r2 = 0 . 98 respectively ) . Consequently , for ease of interpretation , we report the per unit absolute log10 parasite numbers . The date of patent infection for dogs and foxes was estimated as the first date at which animals were positive by any serological or parasitological assay; all samples thereafter were considered as infected based on previous analyses demonstrating a very low incidence of serological reversal [31] , [35] , [36] . At each bimonthly examination , dogs were classified according to their total clinical score as asymptomatic ( scores 0–2 ) , oligosymptomatic ( 3–6 ) and symptomatic ( >6 ) . Dogs with >8 months post infection follow-up and all bimonthly clinical scores <3 were considered long-term asymptomatic . Infectiousness was assessed as either positive ( ≥1 sandfly infected ) or negative , or as the proportion of sandflies infected at any single time point ( point xenodiagnosis ) . Dogs were also classified previously [22] , [35] as “highly infectious” ( >20% of total flies infected ) , “mildly infectious” ( >0% and <20% flies infected ) , and “non-infectious” ( no flies infected ) by serial xenodiagnoses ( n = 6 , 002 flies dissected from 173 independent trials ) : the highly infectious group were shown to be responsible for >80% of all transmission events [22] . All foxes were non-infectious ( n = 1 , 469 flies from 44 trials ) [35] . Parasite aggregation was characterised by the dispersion coefficient k of the fitted negative binomial distribution . Negative binomial models were used to test for differences in parasite loads between groups . Analysis of parasite loads against independent variables were conducted using negative binomial mixed models , with animal identity included as the random effect . The relationship between infectiousness and markers of infection was analysed by logistic regression . Receiver Operating Curves ( ROC ) were used to identify parasite load ( qPCR ) and anti-Leishmania antibody ( ELISA ) threshold values that maximised test sensitivity and specificity to differentiate currently infectiousness and non-infectious dogs . Areas under the ROCs were similar: 0 . 937 ( ear biopsies , n = 58 ) , 0 . 837 ( bone marrows , n = 103 ) and 0 . 846 ( ELISA , n = 173 ) ( χ2 = 72 . 0 , df = 2 , P = 0 . 699 , n = 52 ) , providing test threshold values of 4 . 64 log10 parasites/gram ( ear biopsies ) , 3 . 51 log10 parasites/mL ( bone marrows ) , and 4 . 59 log10 antibody units/mL , respectively . These values were then used to evaluate the performance of threshold-based qPCR and ELISA assays to detect dogs classified by longitudinal infectious status in the mixed population . The average times of detection by the threshold-based assays relative to infection were calculated using Kaplan-Meier survival analysis . Differences in Kaplan-Meier curves were compared by log rank test , and confidence limits calculated following [39] . All analyses were carried out in Stata v . 11 . 1 ( Stata Corporation , College Station , Texas , USA ) .
Parasite loads were quantified by qPCR in 265 post-infection bone marrow samples from 82 dogs , and 185 post-infection ear skin biopsies from 64 dogs ( Table 1 ) . The median parasite loads were 142 parasites/mL in bone marrow and 119 parasites/gram in ear skin ( Table 1 ) but the correlation was not strong ( Spearman's ρ = 0 . 56 , P<0 . 001 ) . Note that since the unit of measurement of these two samples differ , the magnitude of the parasite loads in skin and bone marrow were not directly compared . The frequency distributions of parasite loads in both tissues was highly skewed , with maximum burdens of 2 . 4×106 parasites/mL and 1 . 3×108 parasites/gram in bone marrow and ear skin , respectively ( Figure 1 ) . The degree of parasite aggregation , measured by the negative binomial parameter k , was very high , with loads in ear skin ( k = 0 . 066 ) showing greater aggregation than those in bone marrow ( k = 0 . 104 ) . Comparable degree of aggregation was observed for mean parasite loads in individual dogs ( Table 1 ) . Of the total L . infantum loads recorded in bone marrows biopsies , 90% of parasites were found in 8% ( 21/265 ) of samples and 16% ( 13/82 ) of dogs; for skin biopsies , the equivalent figures were 8% ( 14/185 ) of samples and 9% ( 6/64 ) of dogs . Parasite loads in both tissues increased on average with time since infection ( Table 2; Figure 2 ) . Ear skin loads increased at a faster average rate than bone marrow loads , reflected in the ear skin to bone marrow parasite load ratios being significantly greater in later infection ( Table 2 ) . However , the relationship between parasite load and time varied between individual dogs , showing positive to negative slopes for both tissues ( Figure 3 ) . Both bone marrow and ear skin loads were significantly higher in sick dogs , in infectious dogs and in dogs with higher anti-Leishmania antibody levels ( Table 2 ) . Severity of infection was also associated with greater ear skin to bone marrow parasite ratios ( Table 2 ) . However , in symptomatic dogs this ratio did not vary according to the type of symptom: dogs with skin symptoms had comparable ratios to those with only non-skin symptoms ( IRR = 0 . 67 ( 95% CL 0 . 28–1 . 62 ) , χ2 = 0 . 79 , P = 0 . 37 ) . The probability of a dog being infectious to sandflies at point xenodiagnosis was positively associated with parasite load , PCR status , IgG antibody titer , total clinical score , and time since infection; the strongest predictor of being infectious was ear skin parasite load ( Table 3 ) ; similar results were seen when analysis was restricted to only paired bone marrow and ear skin samples ( data not shown ) . Infectivity to sandflies was associated with high parasite loads in ear skin ( Figure 4 ) : the majority of dogs had loads <106 parasites per gram and were very rarely infectious . Highly infectious dogs had higher mean parasite loads than mildly infectious dogs ( ears: Wald χ2 = 7 . 36 , P = 0 . 0073; marrow: χ2 = 7 . 21 , P = 0 . 0067 ) , the latter showing greater average loads than non-infectious dogs ( ears: χ2 = 13 . 35 , P = 0 . 0003; marrows: χ2 = 14 . 56 , P = 0 . 0001 ) ( Figure 5 ) . L . infantum was detected in bone marrow of 50% ( 17/34 ) and skin of 67% ( 20/30 ) of infected foxes . Parasite loads showed similar over-dispersion as for dogs ( Table 1 ) . Of the total L . infantum loads recorded in bone marrows , 90% was attributed to 8% ( 5/67 ) of samples and to 12% ( 4/34 ) of the foxes . The equivalent figures for skin biopsies were 8% ( 4/53 ) of samples and 10% ( 3/30 ) of foxes . Bone marrow loads varied significantly with fox age , rising rapidly in the first 6 months of life ( age ( months ) : IRR = 1 . 25 ( 95% CL 1 . 11–1 . 42 ) , P = 0 . 0004 ) and declining thereafter ( months2: IRR = 0 . 997 ( 0 . 995–0 . 999 ) , P = 0 . 0015 ) ; a similar , though not significant , pattern was seen for ear skin samples ( P = 0 . 25 ) ( Supplementary Figure S1 ) . No parasites were detected in 15 bone marrow samples from 6 foxes over 6 years old , whereas 4/6 of these foxes ( 4/12 samples ) showed residual parasites in ear skin . In contrast , anti-Leishmania IgG titres did not decline in older age classes ( Supplementary Figure S1 ) . There were significant positive relationships between fox tissue parasite numbers and anti-Leishmania IgG titres ( marrow Wald χ2 = 16 . 0 , df = 1 , P = 0 . 0001; skin Wald χ2 = 5 . 68 , df = 1 , P = 0 . 017 ) , and ear skin to bone marrow parasite ratios were moderately higher in foxes with high titres ( Wald χ2 = 3 . 81 , df = 1 , P = 0 . 051 ) . Only one fox showed any clinical signs of disease ( alopecia ) but which was mild and transitory . Skin and bone marrow parasite loads of foxes were similar to those in non-infectious dogs ( P>0 . 10 ) ( Figure 5 ) . Seven long-term “truly” asymptomatic infected dogs were identified: they transmitted infection to 1/678 sandflies exposed in 24 xenodiagnosis trials on 4 dogs . Their parasite loads were similar to those in foxes ( P>0 . 18 ) , which were all asymptomatic by the same definition ( Figure 5 ) . None of the 22 infected foxes tested were infectious in 39 xenodiagnosis trials . Applying the model coefficients from analysis of dog infectivity ( Table 2 ) to fox ear skin parasite data ( n = 53 ) , foxes were predicted to have been infectious with ≥15% probability ( ≥104 . 64 parasites/gram in skin ) on 6 of 53 occasions for 4 foxes , equivalent to a total predicted number of infectious samples of 2 . 9 of 53 , compared to the observed 0/39 xenopositive trials of infected foxes . The performances of qPCR and ELISA to differentiate dogs of different infectious status in the mixed population were tested using positivity threshold values calculated by ROC analysis of the point xenodiagnosis data ( see Methods ) . PCR-based diagnostic tests showed a high sensitivity ( 94–100% ) to detect highly infectious dogs , though the sensitivities of serology-based tests were somewhat lower ( 78–100% ) ( Table 4 ) . The sensitivities of most tests to detect mildly infectious dogs were lower , but these dogs contributed <20% of transmission . Only tests based on qPCR thresholds showed high specificities for infectious dogs ( i . e . low sensitivities to detect non-infectious dogs ) ( Table 4 ) . Highly infectious dogs were detected by qPCR significantly earlier after patent infection ( 152 days [95% CI: 117–186] ) than either mildly infectious dogs ( 442 days [302–582] ) or non-infectious dogs ( 435 days [317–553] ) ( log rank tests: qPCR: χ2>17 . 3 , P<0 . 0003 ) ; estimates for the latter two groups were statistically indistinguishable ( P = 0 . 70 ) . Detection time of highly infectious dogs approximated their observed time to becoming infectious ( 134 days [68–201] ) .
We demonstrate pronounced heterogeneity in L . infantum loads between dogs , assessed by qPCR in bone marrow and ear skin . Loads were highly over-dispersed with evidence of greater aggregation in ear skin relative to bone marrow ( 9% vs 16% of dogs harboured 90% of total parasites ) . Parasite loads in the two tissues showed different dynamics: bone marrow loads increased rapidly reaching a peak 100–200 days after infection , while ear skin loads continued to increase over a 600 day period , resulting in increased skin to bone marrow load ratios in late infection . Dissemination to the skin varied between dogs , being greater in sick and infectious dogs . Evidence of L . infantum parasite over-dispersion has been reported in different dog tissues [27] , [40]–[42] and in human blood [15] , [26] , [43] , and greater variation in parasite loads in ear skin compared to paired bone marrows , lymph nodes , blood , and liver and spleen samples has been reported for Brazilian dogs [40] , [42] . However these studies did not evaluate parasite loads through time . One cohort study of Italian dogs noted a decrease in ear to lymph node parasite ratios during clinical development , in apparent contrast to results here . In that study , the time of infection was not established , so dogs may have been at a different stage and severity of infection [34] . Tissue parasite load , particularly in ear skin , was the best predictor of being currently infectious to vectors . L . infantum amastigotes in skin tissue or skin capillaries are directly accessible to sandflies , which are known to feed abundantly on ear pinnae; and ear skin appears to be more infective than abdomen skin [23] , [30] . Some of the variation in parasite loads between ear tissue samples may also reflect small scale spatial variation in parasite density within the ear . We did not restrict sandflies to feed only on ears , unlike other studies [23] , [29] , [44] . However , the proportion of dogs that were infectious was substantially lower than the proportion with detectable skin parasites , and only dogs with very high skin parasite loads were consistently infectious . Highly infectious dogs showed greater average loads compared to mildly infectious and non-infectious dogs , and also tended to fall within the top 20% parasite loads for each tissue . These data , and the observed high degree of parasite aggregation in ear skin , suggest that the majority of transmission events to vectors result from a small proportion of infectious dogs . Previously we reported for these dogs that 7 of 42 infectious dogs ( 17% ) were responsible for >80% of all sandfly infections [22] . Similar over-dispersion in infectiousness can be calculated from published xenodiagnosis studies , with 15% to 44% of dogs accounting for >80% of transmission events [23] , [44] . qPCR studies of canid tissue L . infantum loads relative to xenodiagnoses are not available elsewhere , but parasite estimates by immunohistochemistry of ear skin show moderate correlations with xenodiagnosis positivity [30] , [45] . Our current results suggest that high parasite loads in dog ear skin , rather than the simple presence of parasites , is the important metric to identify likely infectious individuals and potential reservoir populations . In the current study , all infections were shown to be L . infantum [36] . To identify super-spreaders in regions of mixed Leishmania co-infections , the specificity of qPCR methods would need to be fully validated . Current ZVL control strategy in Brazil includes mass test-and-slaughter of Leishmania antibody positive dogs [17] , which is criticised on theoretical , logistical and also on ethical grounds [18]–[22] . If the small fraction of dogs that are responsible for the majority of transmission could be identified ( e . g . by detection of high parasite loads ) and targeted , this would directly address many of these issues , and may be more cost-effective than mass interventions [9] , [12] . Canine infectiousness to sandflies is known to increase with the severity of disease and high anti-parasite antibody , but sensitive and specific markers of infectiousness have not been identified [22] , [23] , [29] , [30] , [46] . Here , we show that adopting quantitative test threshold values based on skin parasite numbers , highly infectious dogs can be distinguished from non-infectious dogs . These tests were highly sensitive for highly infectious dogs , equivalent to detection of 87–94% of sandfly infections in these samples ( data not shown ) , and importantly also showed high specificities ( 0 . 83–0 . 99 ) to detect non-infectious dogs , unlike conventional tests for infection . Since up to 50% of seropositive dogs may be asymptomatic in a single community survey , such a targeted approach should also raise dog-owner compliance . The crab-eating fox occurs widely in South America , and is commonly infected with L . infantum [16] , [35] , [47] , and thus often assumed to be a sylvatic reservoir . However , few infected foxes have been shown to infect sandflies [48] , [49] , and in our cohort study none of the foxes were infectious [35] . Here , we show that fox parasite loads , though heterogeneous , were significantly lower than those of infectious dogs , and similar to non-infectious dogs , providing further evidence that foxes are not likely to be important for maintaining transmission [22] , [35] . The results also provide a parasitological explanation for why the foxes here , and probably wild canids more generally , tend to present asymptomatic infections [16] , [50] , [51] . Relatively low parasite loads were also noted in the truly asymptomatic cohort dogs , as also reported in asymptomatic human infection [26] , [28] , [52] . Whether asymptomatic human infections with L . donovani is associated with low parasite loads and thus low transmission potential remains speculative , and further studies are needed [16] . Variation in parasite load between individuals of other potential reservoir hosts ( e . g . hares in Iberia [53] ) , and variation in parasite load in skin between different parts of the host , would also be informative . In conclusion , this study highlights the importance of quantifying heterogeneities in infection loads in relation to transmission potential through prospective studies , underpinning development of novel tools for parasitic disease management . Studies are now needed to confirm the efficacy of diagnostic threshold-based driven actions against transmission , and to develop diagnostic kits , based on the detection of parasite DNA ( e . g . isothermal amplification ) or parasite antigens , for practical field use . | Zoonotic visceral leishmaniasis is a sandfly-borne disease of humans and dogs caused by the intracellular parasite Leishmania infantum . Dogs are the proven reservoir . The disease is usually fatal unless treated , and is of global health significance . Diagnosis of canine infections relies on serum antibody-based tests that measure infection . In some endemic regions , a test-and-slaughter policy of seropositive dogs forms part of the national control policy to reduce human infection . However , this strategy is not considered effective . Since not all infected dogs are infectious to sandfly vectors , one option is to target control at infectious dogs , as only these dogs maintain transmission . We quantify Leishmania numbers in individual host tissues from time of infection using molecular methods . Comparing these results with their infectiousness to sandflies , we also evaluate the performance of molecular and immunological assays to identify infectious animals . Parasite numbers varied substantially between individuals , increasing with duration and severity of disease . Infectiousness to the sandfly vector was associated with high parasite numbers , and parasite loads in the skin was the best predictor of being infectious . The results suggest that molecular quantitation is useful in identifying individuals and populations responsible for maintaining transmission , with potential application in operational control programmes . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2014 | Heterogeneities in Leishmania infantum Infection: Using Skin Parasite Burdens to Identify Highly Infectious Dogs |
Phosphatidylinositol 3-phosphate ( PtdIns ( 3 ) P ) is a signaling molecule important for many membrane trafficking events , including phagosome maturation . The level of PtdIns ( 3 ) P on phagosomes oscillates in two waves during phagosome maturation . However , the physiological significance of such oscillation remains unknown . Currently , the Class III PI 3-kinase ( PI3K ) Vps34 is regarded as the only kinase that produces PtdIns ( 3 ) P in phagosomal membranes . We report here that , in the nematode C . elegans , the Class II PI3K PIKI-1 plays a novel and crucial role in producing phagosomal PtdIns ( 3 ) P . PIKI-1 is recruited to extending pseudopods and nascent phagosomes prior to the appearance of PtdIns ( 3 ) P in a manner dependent on the large GTPase dynamin ( DYN-1 ) . PIKI-1 and VPS-34 act in sequence to provide overlapping pools of PtdIns ( 3 ) P on phagosomes . Inactivating both piki-1 and vps-34 completely abolishes the production of phagosomal PtdIns ( 3 ) P and disables phagosomes from recruiting multiple essential maturation factors , resulting in a complete arrest of apoptotic-cell degradation . We have further identified MTM-1 , a PI 3-phosphatase that antagonizes the activities of PIKI-1 and VPS-34 by down-regulating PtdIns ( 3 ) P on phagosomes . Remarkably , persistent appearance of phagosomal PtdIns ( 3 ) P , as a result of inactivating mtm-1 , blocks phagosome maturation . Our findings demonstrate that the proper oscillation pattern of PtdIns ( 3 ) P on phagosomes , programmed by the coordinated activities of two PI3Ks and one PI 3-phosphatase , is critical for phagosome maturation . They further shed light on how the temporally controlled reversible phosphorylation of phosphoinositides regulates the progression of multi-step cellular events .
PtdIns ( 3 ) P is a phosphorylated phosphatidylinositol ( PtdIns ) species that is embedded in distinct membrane domains and plays important roles in many membrane trafficking events , including endocytic trafficking , retrograde trafficking , autophagy , and phagosome maturation ( reviewed in [1] , [2] ) . PtdIns ( 3 ) P activates downstream pathways through recruiting specific PtdIns ( 3 ) P-binding proteins , the PtdIns ( 3 ) P effectors , to the site of action [3] . The production and elimination of PtdIns ( 3 ) P on a particular membrane domain , such as the surface of phagosomes , are presumably under tight temporal regulation to allow the occurrence of multiple transient signaling events . However , the molecular mechanisms behind such regulation are not well understood . In metazoan , such as the nematode Caenorhabditis elegans , the fruit fly Drosophila melanogaster , and mammals , a large number of cells undergo apoptosis during development and adulthood and are subsequently engulfed by phagocytes and degraded inside phagosomes . Like other phagosomal cargos , dying cells are degraded through phagosome maturation , a process that relies heavily on the fusion between phagosomes and various intracellular organelles including endosomes and lysosomes , which leads to the delivery of digestive enzymes into phagosomes and the acidification of the phagosomal lumen [4]–[6] . Immediately after an apoptotic cell is internalized , a high level of PtdIns ( 3 ) P appears transiently on the surface of a nascent phagosome [7] , [8] . This prominent feature was observed on nascent phagosomes containing various kinds of cargos , including zymosan particles , latex beads , and invading pathogens in addition to apoptotic cells and is well conserved in different organisms [7]–[13] . In C . elegans embryos , during the maturation of phagosomes containing apoptotic cells , PtdIns ( 3 ) P is dynamically enriched on phagosomes in two consecutive waves: the initial burst , which appears upon the closure of a phagocytic cup and dissipates after 10–15 min , and a subsequent reappearance of a relatively weaker PtdIns ( 3 ) P signal approximately 10 min later , which lasts until an apoptotic cell is fully degraded [7] , [8] . Such a PtdIns ( 3 ) P oscillation pattern has also been observed on phagosomes containing other kinds of cargos [10] , [14] . However , the physiological significance of this biphasic PtdIns ( 3 ) P appearance on phagosomes remains unknown . PtdIns ( 3 ) P is a 3′-phosphorylated form of PtdIns . Among the three classes of known phosphoinositide 3-kinases ( PI3Ks ) , Vps34 , the sole Class III PI3K , is known to specifically produce PtdIns ( 3 ) P on intracellular membranes , such as endosomes , phagosomes , and autophagosomes , to regulate diverse membrane trafficking events ( reviewed in [2] , [15] ) . Specific inactivation of Vps34 , through injection of anti-Vps34 antibodies into mammalian cells or RNA interference ( RNAi ) treatment in C . elegans , reduces the efficiency of phagosome maturation , indicating that PtdIns ( 3 ) P is important for promoting phagosome maturation [12] , [13] , [16] . However , whether Vps34 is the only PI3K that produces PtdIns ( 3 ) P on phagosomal surfaces remains unknown . Furthermore , in the absence of a strategy capable of complete depletion of phagosomal PtdIns ( 3 ) P , it is impossible to quantitatively determine how fundamental the role of PtdIns ( 3 ) P is in initiating phagosome maturation . During the development of C . elegans hermaphrodites , 131 somatic cells and approximately 300–500 germ cells undergo programmed cell death [17] . These cells are easily recognizable within living animals under the Nomarski Differential Interference Contrast ( DIC ) optics as highly refractive , button-like objects referred to as “cell corpses” [18] , [19] . In C . elegans , after being rapidly engulfed and contained inside phagosomes , apoptotic cells are degraded via a pathway initiated by the phagocytic receptor CED-1 , which is transiently clustered on the surface of extending pseudopods and nascent phagosomes [8] . CED-1 triggers the robust production of PtdIns ( 3 ) P on phagosomal surfaces through recruiting the large GTPase DYN-1 , the C . elegans homolog of mammalian dynamins , to phagosomes [8] , [20] . In another study focusing on how PtdIns ( 3 ) P triggers phagosome maturation in C . elegans , we identified LST-4/SNX-9 , SNX-1 , and SNX-6 , three PX and BAR domain-containing sorting nexins , as PtdIns ( 3 ) P effectors recruited to phagosomal surfaces by PtdIns ( 3 ) P and subsequently acting in two parallel pathways to drive the incorporation of endosomes and lysosomes into phagosomes [21] . These findings revealed a signaling cascade initiated by phagocytic receptor CED-1 , mediated by PtdIns ( 3 ) P , and executed through these sorting nexins to degrade apoptotic cells [8] , [21] . Interestingly , the phagosome maturation-delay/arrest phenotype observed from C . elegans vps-34 mutants was much milder than that from the ced-1 or dyn-1 single mutants , or the lst-4; snx-1; snx-6 triple mutants [8] , [13] , [21] , suggesting that inactivating VPS-34 alone might not completely deplete PtdIns ( 3 ) P from phagosomes . We examined whether there existed additional PI 3-kinase ( s ) responsible for producing phagosomal PtdIns ( 3 ) P . Besides Vps34 , Class II PI3Ks are also able to produce PtdIns ( 3 ) P from PtdIns , whereas Class I PI3Ks primarily produce PtdIns ( 3 , 4 , 5 ) P3 [15] . In vitro , PtdIns is the favorable substrate for Class II PI3Ks [22]–[25] . Multiple lines of evidence have revealed that Class II PI3Ks produce PtdIns ( 3 ) P in vivo in response to certain extracellular and intracellular stimuli [26]–[29] . However , in comparison to Class I and III PI3Ks , relatively little is known about the physiological functions and regulation of Class II PI3Ks ( reviewed in [15] , [30] ) . In particular , it is not known whether any Class II PI3K ( s ) is involved in producing PtdIns ( 3 ) P on phagosomes . Here we revealed the function of PIKI-1 , the only C . elegans Class II PI3K , in phagosome maturation . In a closely related aspect , little is known about the factors that down-regulate the initial peak of phagosomal PtdIns ( 3 ) P , or the physiological significance of PtdIns ( 3 ) P oscillation during phagosome maturation . Promising candidates able to down-regulate PtdIns ( 3 ) P level include PtdIns ( 3 ) P phosphatases . The myotubularin phosphatases are a family of PI phosphatases that convert PtdIns ( 3 ) P or PtdIns ( 3 , 5 ) P2 into PtdIns or PtdIns ( 5 ) P , respectively [31] . They act to maintain the homeostasis of PtdIns ( 3 ) P on intracellular membranes [29] , [31] . Inactivation of MTM-1 , a member of the myotubularin family in C . elegans , rescued the endocytosis defect of vps-34 mutants , suggesting that MTM-1 antagonizes the PI 3-kinase activity of VPS-34 during endosomal trafficking [32] . However , it was not known whether MTM-1 was involved in the dephosphorylation of phagosomal PtdIns ( 3 ) P . In this report , we identified a novel role of the C . elegans Class II PI3K PIKI-1 in the degradation of apoptotic cells . We further revealed the differential and complementary roles of PIKI-1 and VPS-34 in the production and maintenance of phagosomal PtdIns ( 3 ) P . Moreover , we have uncovered a novel function of MTM-1 in modulating the dynamic pattern of PtdIns ( 3 ) P on phagosomes . We found that the prompt down-regulation of phagosomal PtdIns ( 3 ) P , like its robust production on nascent phagosomes , is pivotal for driving phagosome maturation . Our work revealed a regulatory system that issues a precise temporal control over the PtdIns ( 3 ) P cycling pattern on phagosomes and ensures the efficient degradation of apoptotic cells .
To evaluate whether inactivating vps-34 would result in a complete depletion of phagosomal PtdIns ( 3 ) P , we monitored the level of PtdIns ( 3 ) P on the surface of phagosomes using a previously established PtdIns ( 3 ) P reporter , 2xFYVE::GFP , which was expressed in engulfing cells under the control of Pced-1 and specifically associated with PtdIns ( 3 ) P [8] , [33] . We first monitored phagosomal PtdIns ( 3 ) P inside gonadal sheath cells , the engulfing cells for apoptotic germ cells [34] . In the gonad of wild-type adult hermaphrodites , 86% of germ cell corpses , recognized under the DIC optics , were labeled with bright 2xFYVE::GFP on their surfaces ( Figure 1A ( a , e ) and C ) , indicating the presence of PtdIns ( 3 ) P at high levels on phagosomal surfaces . Knocking down vps-34 by RNA interference ( RNAi ) using two non-overlapping RNAi constructs reduced the percentage of gonadal phagosomes labeled with 2xFYVE::GFP to 54% or 58% , respectively ( Figure 1A ( b , f ) and C ) . Furthermore , vps-34 ( RNAi ) animals displayed a mild cell-corpse removal defective ( Ced ) phenotype resulting from un-removed cell corpses , evident by the modestly increased numbers of germ cell corpses in the gonad ( Figures 1C and S1A ) , as reported previously [13] . These results suggest that inactivation of vps-34 only partially impairs the production of PtdIns ( 3 ) P on gonadal phagosomes . To examine VPS-34's contribution to the production of phagosomal PtdIns ( 3 ) P in a more strict manner , we created a strain that produced vps-34 ( h510 ) ( m−z− ) ( m , maternal gene product; z , zygotic gene product ) homozygous null mutant embryos that lost both the maternal and the zygotic expression of vps-34 , which is likely to represent the vps-34 complete loss-of-function phenotypes ( Figure S1B and Text S1 ) . In vps-34 ( h510 ) ( m−z− ) mutant embryos , the number of PtdIns ( 3 ) P-labeled cytoplasmic puncta , which represented early endosomal particles , was nearly abolished , consistent with the known function of VPS-34 in the production of PtdIns ( 3 ) P on early endosomes ( Figure 1D ( e , f , i , j ) ) [35] . Moreover , PtdIns ( 3 ) P was much less frequently detected on phagosomal surfaces than in wild-type embryos ( Figure 1D ( b , f , j ) and E ) . Consistent with this defect , mutant embryos displayed a modest Ced phenotype , retaining more cell corpses than wild-type embryos at all four embryonic stages examined ( Figure 1F ) . However , as in the gonad , PtdIns ( 3 ) P was still detectable on certain phagosomes ( 11 . 3% ) in embryos ( Figure 1D ( f ) and E ) , indicating that the inactivation of vps-34 reduces but does not completely abolish the production of PtdIns ( 3 ) P on phagosomes . Additional PtdIns ( 3 ) P-producing activity besides VPS-34 must exist on phagosomal membranes . Previously , we found that tm3171 , a deletion mutation of piki-1 , which encodes the only C . elegans Class II PI3K ( Figure 2A , Figure S2A , Text S1 ) , resulted in a strong Ced phenotype in the adult hermaphrodite gonad ( Figure 1A ( c ) and C ) [21] . We further found that the piki-1 ( tm3171 ) mutation also resulted in moderate Ced phenotypes in embryos throughout mid- and late-embryonic stages ( Figure 1F ) . These results indicate that piki-1 is required for the removal of both the somatic and germ cell corpses . In the gonad of piki-1 ( tm3171 ) mutant hermaphrodites , the percentage of phagosomes labeled with 2xFYVE::GFP was reduced to 15% of the wild-type level ( Figure 1A ( c , g ) , B–C ) . The majority of gonadal phagosomes were embedded inside the 2xFYVE::GFP-expressing cytoplasm of the host cells , yet lacked enriched GFP signal on their surfaces , appearing as “dark spheres” ( Figure 1A ( g , filled arrowheads ) and B ) . These observations revealed a novel activity of PIKI-1 in producing phagosomal PtdIns ( 3 ) P . Moreover , quantitative comparison of the percentage of PtdIns ( 3 ) P-labeled phagosomes in piki-1 ( tm3171 ) and vps-34 ( RNAi ) animals indicates that in gonadal sheath cells , PIKI-1 plays a major role in producing PtdIns ( 3 ) P on phagosomes ( Figure 1C ) . In mid-stage ( 2-fold stage ) piki-1 ( tm3171 ) mutant embryos , the percentage of PtdIns ( 3 ) P-labeled phagosomes was reduced to 40% of the wild-type level ( Figure 1D and E ) , indicating that , like in the adult gonad , piki-1 also plays a role in producing phagosomal PtdIns ( 3 ) P in somatic engulfing cells . Unlike vps-34 ( h510 ) ( m−z− ) mutant animals , which are embryonic and larval lethal ( Figure S1B ) , piki-1 ( tm3171 ) mutant animals undergo normal development , and are viable and fertile . In addition , unlike vps-34 ( RNAi ) animals , piki-1 ( tm3171 ) animals display normal endocytosis activities ( Text S1 and Figure S3 ) [36] . Consistent with this observation , similar numbers of the bright PtdIns ( 3 ) P-labeled puncta , which represent endosomal particles , were found in the cytoplasm in both piki-1 ( tm3171 ) mutant and wild-type embryos ( Figure 1D ( i , k ) ) , indicating that the function of PIKI-1 was dispensable for the production of PtdIns ( 3 ) P on early endosomes . These results indicate that PIKI-1 specifically produces PtdIns ( 3 ) P on phagosomes , and that this activity is required for the efficient removal of cell corpses . Inefficient removal of apoptotic cells might be caused by defects in either the internalization step or the degradation step . In piki-1 ( tm3171 ) mutant adult hermaphrodites , persistent germ cell corpses were all observed inside phagosomes: only 13% of these phagosomes were labeled with PtdIns ( 3 ) P on their surfaces that appeared like bright GFP circles , the rest of them lacked PtdIns ( 3 ) P on their surfaces and appeared as dark spheres inside the host cells ( Figure 1A ( c , g ) and 1C ) . This observation suggests that in piki-1 mutants , cell corpses were successfully engulfed but not efficiently degraded inside phagosomes . Using transmission electron microscopy ( TEM ) , we analyzed the internalization status of germ cell corpses ( Materials and Methods ) and found that 95% of germ cell corpses in the gonad of piki-1 ( tm3171 ) mutants were fully engulfed and remained undegraded inside phagosomes ( Figure 2B–G ) . This result confirmed that the function of PIKI-1 is essential for the degradation but not the engulfment of cell corpses . Since the inactivation of either VPS-34 or PIKI-1 only partially impaired the PtdIns ( 3 ) P production on phagosomes , we further monitored PtdIns ( 3 ) P in the gonad of piki-1 ( tm3171 ) ; vps-34 ( RNAi ) hermaphrodites and observed a near complete depletion of phagosomal PtdIns ( 3 ) P ( Figure 1A ( d , h ) and 1C ) . Similarly , in mid-stage vps-34 ( h510 ) ( m−z− ) ; piki-1 ( tm3171 ) double mutant embryos , PtdIns ( 3 ) P was no longer detected on phagosomal surfaces or on any cytoplasmic puncta ( Figure 1D ( d , h , l ) and E ) . These observations indicate that VPS-34 and PIKI-1 together produce all the PtdIns ( 3 ) P molecules on gonadal and somatic phagosomes . They also indicate that vps-34 ( RNAi ) is potent in inactivating VPS-34 in the gonadal sheath cells since only a residual PtdIns ( 3 ) P-production activity was left on phagosomes in piki-1 ( tm3171 ) ; vps-34 ( RNAi ) double mutant animals . Previously , without an effective method to completely block the production of PtdIns ( 3 ) P on phagosomes , it was impossible to determine whether PtdIns ( 3 ) P was absolutely essential for triggering phagosome maturation or merely contributed to phagosome maturation as one of multiple signaling molecules . The vps-34; piki-1 double mutant , in which PtdIns ( 3 ) P is no longer detectable on phagosomal surfaces , provides a suitable tool for addressing this question . In the adult gonads of piki-1 ( tm3171 ) ; vps-34 ( RNAi ) double mutants , a larger number of persistent germ cell corpses were observed than in any PI3K single mutants ( Figure 1C ) . As in piki-1 ( tm3171 ) single mutants , all cell corpses were engulfed inside phagosomes in piki-1 ( tm3171 ) ; vps-34 ( RNAi ) double mutants ( Figure 1A ( d , h , arrowheads ) , indicating a specific defect in phagosome maturation . Similarly , vps-34 ( h510 ) ( m−z− ) ; piki-1 ( tm3171 ) double mutant embryos displayed a much stronger Ced phenotype than any single mutants ( Figure 1F ) . At 4-fold stage , the number of persistent cell corpses in the double mutant embryos was on average 10-fold larger than that in single mutant embryos ( Figure 1F ) . These observations strongly indicate that PIKI-1 and VPS-34 act in combination to drive the efficient degradation of cell corpses . The tight quantitative correlations between the depletion of phagosomal PtdIns ( 3 ) P and the arrest of cell-corpse degradation demonstrate that PtdIns ( 3 ) P is an imperative factor for phagosome maturation . Previously , we identified SNX-1 , SNX-6 , and LST-4/SNX-9 , three BAR domain-containing sorting nexins , as novel PtdIns ( 3 ) P effectors that drive the degradation of apoptotic cells [21] . In vitro , SNX-1 and LST-4 display high affinity towards PtdIns ( 3 ) P [21] . To further determine whether PtdIns ( 3 ) P acts as an upstream regulator of SNX-1 and LST-4 , we examined the dynamic localization of SNX-1::GFP and LST-4::GFP reporters ( Figure S5 ) [21] on phagosomes in vps-34 ( h510 ) ( m−z− ) ; piki-1 ( tm3171 ) double mutant embryos . We used a previously established time-lapse recording technique to monitor the maturation process of three well-defined phagosomes , C1 , C2 , and C3 , which were formed inside three adjacent hypodermal cells at ∼330 min post-first cleavage ( Figure S6C ) [8] , [33] . This technique enabled us to compare the dynamic localization pattern of a reporter on the exact same phagosome in different genetic backgrounds . In wild-type embryos , as reported [21] , SNX-1::GFP and LST-4:GFP were rapidly recruited onto nascent phagosomes immediately after the sealing of phagocytic cups , and remained on phagosomes for ∼10 min ( Figure 3A–B ) . In contrast , in vps-34 ( h510 ) ( m−z− ) ; piki-1 ( tm3171 ) double mutant embryos , SNX-1 and LST-4 failed to accumulate on phagosomes ( Figure 3A–B ) , indicating that the phagosomal PtdIns ( 3 ) P is necessary for the recruitment of its effectors to nascent phagosomes . In adult hermaphrodite gonadal sheath cells , we found that the percentage of phagosomes labeled with SNX-1::GFP was reduced to 20% of the wild-type level ( Figure 3C–D ) . This observation indicates that PtdIns ( 3 ) P is also responsible for recruiting sorting nexins to phagosomes containing germ cell corpses . To further investigate the function of PtdIns ( 3 ) P in regulating phagosome maturation , we examined whether the inactivation of piki-1 would affect the recruitment of three small GTPases , RAB-5 , RAB-2 ( also named UNC-108 ) , and RAB-7 , to phagosomal surfaces . During phagosome maturation , these membrane-tethering factors are sequentially recruited from the cytoplasm of the host cells to phagosomal surfaces , where they facilitate the fusion of intracellular organelles to phagosomes [7] , [8] , [13] , [37] , [38] . In wild-type gonads , we observed that GFP::RAB-5 , GFP::RAB-2 , and GFP::RAB-7 were enriched on 68% , 85% , and 90% of phagosomes , respectively ( Figure 4A–D ) . In the gonads of piki-1 ( tm3171 ) mutant hermaphrodites , however , the percentages of phagosomes labeled with GFP::RAB-5 , GFP::RAB-2 , or GFP::RAB-7 were strongly reduced to 9% , 16% , or 23% , respectively ( Figure 4A–D ) , indicating that the function of PIKI-1 is essential for the recruitment of all three RAB GTPases to phagosomal surfaces . When piki-1 and vps-34 were simultaneously inactivated , the phagosomes labeled with GFP::RAB-7 were further reduced to 10% ( Figure 4C and D ) . These results indicate that PtdIns ( 3 ) P is essential for recruiting these three RAB proteins . To further determine the relationships between PtdIns ( 3 ) P and the RAB GTPases , we examined the production of PtdIns ( 3 ) P on phagosomes in each of the rab-5 , rab-2 , and rab-7 loss-of-function background . Whereas rab-2 and rab-7 mutations did not affect the presentation of PtdIns ( 3 ) P on phagosomes , inactivation of rab-5 by RNAi greatly reduced the percentage of PtdIns ( 3 ) P-labeled phagosomes ( Figures 4E and S4 , Text S1 ) , indicating that the production of phagosomal PtdIns ( 3 ) P relies on RAB-5 but not RAB-2 or RAB-7 . Together , our observations indicate that PtdIns ( 3 ) P and RAB-5 depend on each other for the enrichment on the surface of nascent phagosomes . Our observation that PtdIns ( 3 ) P and RAB-5 were concurrently enriched on nascent phagosomes further supports this conclusion ( Figure S7 and Text S1 ) . To determine why the activities of two PI3Ks , which produce an identical signaling molecule PtdIns ( 3 ) P on phagosomes , are both needed for efficient phagosome maturation , we monitored the pattern of PtdIns ( 3 ) P appearance on phagosomes over time in wild-type , each of the single PI3K mutant , and double PI3K mutant embryos ( Figures 5 ) . As reported previously [8] , in wild-type embryos , the level of PtdIns ( 3 ) P on phagosomes oscillated in a two-wave pattern: immediately ( 2–4 min ) after engulfment , a robust PtdIns ( 3 ) P signal rapidly appeared on nascent phagosomes and lasted for ∼15 min before it quickly diminished ( Figure 5A , E ) ; after a gap period ( ∼10 min ) , a PtdIns ( 3 ) P signal weaker than the initial one reappeared on the maturing phagosomes , where it lasted until the phagosomal content was completely degraded ( Figure 5A and E ) . During phagosome maturation , we also frequently observed PtdIns ( 3 ) P ( + ) puncta on phagosomes , which first attached to phagosomal surfaces and later merged into phagosomal membranes ( Figure 5A ) . The dynamic association of PtdIns ( 3 ) P puncta with phagosomes likely represents the docking and fusion process of endosomes with phagosomes . In vps-34 ( h510 ) ( m−z− ) ; piki-1 ( tm3171 ) double mutant embryos , PtdIns ( 3 ) P was completely absent from phagosome surfaces throughout the entire recording period ( 90 min ) ( Figure 5D and E ) , causing a complete arrest of phagosome maturation , indicated by the largely unchanged size of phagosomes after a long period of time . In vps-34 ( h510 ) ( m−z− ) single mutant embryos , PtdIns ( 3 ) P was robustly produced on nascent phagosomes immediately after engulfment , as in the wild-type background ( Figure 5B , E , and F ) . However , PtdIns ( 3 ) P disappeared from phagosomal surfaces much more promptly , lasting for only 6 min on average on phagosomes ( Figure 5B , E , and G ) . Moreover , PtdIns ( 3 ) P failed to reappear on phagosomes ( Figure 5B and E ) . In contrast , in most piki-1 ( tm3171 ) single mutant embryos , an obvious delay of PtdIns ( 3 ) P production on nascent phagosomes was observed ( Figure 5C , E , and F ) . Furthermore , the duration of the first PtdIns ( 3 ) P wave was significantly shorter than that in wild-type control ( Figure 5A , C , E , and G ) . Apart from these aberrations , the disappearance of the initial phagosomal PtdIns ( 3 ) P and the subsequent reappearance of PtdIns ( 3 ) P on phagosomes followed a relatively normal temporal pattern ( Figure 5C and E ) . These observations indicate that PIKI-1 and VPS-34 play differential roles in the production and maintenance of phagosomal PtdIns ( 3 ) P: whereas PIKI-1 is required for the initial production of PtdIns ( 3 ) P , VPS-34 is needed for the sustained production of PtdIns ( 3 ) P ( Figure 5H ) . Given that the Ced phenotype displayed by vps-34 and piki-1 single mutant embryos is similar in severity ( Figure 1F ) , the precise timing of initial PtdIns ( 3 ) P production must be as important as the proper time span of PtdIns ( 3 ) P appearance on phagosomes for phagosome maturation . To examine the subcellular localization of PIKI-1 , in particular to determine whether PIKI-1 acts on nascent phagosomes to produce PtdIns ( 3 ) P , we characterized a PIKI-1::GFP reporter expressed in engulfing cells under the control of Pced-1 . Pced-1piki-1::gfp fully rescued the Ced phenotype of piki-1 mutants ( Figure 6A ) , indicating that the function of PIKI-1 in engulfing cells was sufficient for driving phagosome maturation . piki-1-expression constructs lacking either the entire PI kinase domain ( Figure 2A ) or carrying a mutation ( K1059A ) that disrupted ATP-binding ( Figure S9 ) [39] lost most of the rescuing activity ( Figure 6A ) , indicating that the PtdIns ( 3 ) P-production activity of PIKI-1 is essential . In both embryos and the gonads of adult hermaphrodites , PIKI-1::GFP was primarily localized to the cytoplasm and was enriched on phagosomal surfaces ( Figure 6B–C ) . During cell-corpse removal , PIKI-1::GFP was first detected on extending pseudopods and subsequently further enriched on nascent phagosomes , prior to the rapid appearance of the bright PtdIns ( 3 ) P signal ( Figures 6D–E , S8 ) . PIKI-1::GFP remained on phagosomal surfaces for ∼6 min before dissociating from phagosomes ( Figures 6D–E , S8 ) . This dynamic phagosomal enrichment pattern is consistent with PIKI-1's role in producing the initial PtdIns ( 3 ) P molecules on nascent phagosomes ( Figure 5H ) . The large GTPase DYN-1 plays an essential role in the production of phagosomal PtdIns ( 3 ) P [8] , [13] . To determine whether DYN-1 plays this role through regulating PIKI-1 , we examined the phagosomal localization of PIKI-1 in dyn-1 ( n4039 ) null mutant embryos . We failed to observe any significant enrichment of PIKI-1::GFP on the surfaces of either the extending pseudopods or nascent phagosomes ( Figure 6F–G ) . This result indicates that DYN-1 triggers PtdIns ( 3 ) P production by recruiting PIKI-1 to phagosomes . In wild-type embryos , despite the sequential and combined activities of PIKI-1 and VPS-34 , an obvious temporal gap , during which PtdIns ( 3 ) P is almost undetectable on phagosomes , was observed between the two PtdIns ( 3 ) P peaks ( Figures 5A and S6F ) . In particular , the quickly vanishing PtdIns ( 3 ) P signal on phagosomes in vps-34 mutants suggests the existence of a previously unidentified activity that antagonizes PI3Ks by removing PtdIns ( 3 ) P from phagosomes . To identify the negative regulator of phagosomal PtdIns ( 3 ) P , we examined the function of C . elegans PI 3-phosphatase MTM-1 in regulating PtdIns ( 3 ) P dynamics during phagosome maturation . In embryos homozygous for mtm-1 ( op309 ) , a partial loss-of-function mutation [40] , we observed the normal initial appearance of PtdIns ( 3 ) P within 2–4 min of phagosome formation . However , the level of the PtdIns ( 3 ) P on phagosomes was significantly higher in mtm-1 ( op309 ) mutant embryos than in wild-type embryos ( Figures 7A–C and S10 ) . In addition , the appearance period of PtdIns ( 3 ) P on phagosomes was remarkably prolonged: the mean duration of the first PtdIns ( 3 ) P wave was more than doubled in mtm-1 ( op309 ) embryos , with 27% phagosomes lasting longer than 3 . 6 times of the average length in wild-type embryos ( Figures 7A–D and S10 ) . Since MTM-1 is able to dephopshorylate PtdIns ( 3 ) P to PtdIns in vitro [40] , our observations suggest that MTM-1 might directly dephosphorylate phagosomal PtdIns ( 3 ) P . Previously , whether the down-regulation of phagosomal PtdIns ( 3 ) P has any functional significance has not been explored . To address this question , we measured the rate of phagosome maturation in mtm-1 ( op309 ) embryos , aided by CED-1C ( the intracellular domain of CED-1 ) ::GFP expressed in the engulfing cells [8] . CED-1C::GFP , which is evenly distributed in the cytoplasm of host cells , allowed us to detect the GFP ( − ) phagosomes as dark spheres ( Figure 7F ) [8] . In mtm-1 ( op309 ) mutant embryos , phagosome duration was mildly yet significantly ( p = 2×10−9 , Student's t test ) prolonged , with the average duration 38% longer than the wild-type control ( Figure 7E–F ) . This result suggests that the magnified and prolonged action of PtdIns ( 3 ) P on phagosomes does not speed up but rather slows down phagosome maturation . The mtm-1 ( op309 ) mutation is a missense mutation ( G106E ) that only partially inactivates MTM-1 [40] . To inactivate MTM-1 more effectively , we performed mtm-1 ( RNAi ) ( Materials and Methods ) , which resulted in the accumulation of a large number of germ cell corpses in adult hermaphrodite gonads ( Figure 7G–H ) . This strong Ced phenotype is unlikely a result of an off-target effect of RNAi , since two non-overlapping RNAi constructs , each of which targeting the N- or C-terminal half of mtm-1 cDNA , respectively , caused the Ced phenotype to the same severity ( Figure 7H ) . Aided by a previously established phagocytosis assay that utilized a GFP::RAB-7 reporter specifically expressed in engulfing cells to distinguish engulfed versus unengulfed cell corpses [21] , we further determined that the majority of the germ cell corpses observed in mtm-1 ( RNAi ) gonads were engulfed inside gonadal sheath cells ( Figure 7I ) . This phenotype is in direct contrast to that displayed by ced-5 ( n1812 ) null mutants , which are primarily defective in the engulfment of germ cell corpses ( Figure 7I ) , and indicates that mtm-1 is primarily required for phagosome maturation . mtm-1 ( RNAi ) increased the percentage of PtdIns ( 3 ) P-labeled gonadal phagosomes from 86% to 100% ( Figure 7J ) , again suggesting that PtdIns ( 3 ) P is retained on phagosomes once MTM-1 is inactivated . Together , the above results strongly indicate that the turnover of phagosomal PtdIns ( 3 ) P is crucial for phagosome maturation . Using a GFP::MTM-1 reporter expressed specifically in engulfing cells ( Pced-1gfp::mtm-1 ) , we observed the transient association of MTM-1 with the extending pseudopods throughout engulfment and with nascent phagosomes for approximate 13 min ( Figure 8A ) . The phagosome-association pattern of GFP::MTM-1 overlaps with the first PtdIns ( 3 ) P wave ( Figures 5A and S6F ) . Furthermore , MTM-1 completely co-localizes with PIKI-1 on phagosomes until PIKI-1 is dissociated from phagosomes ( Figure 6D ) ; MTM-1 also overlaps with the PtdIns ( 3 ) P-producing activity of VPS-34 during the first but not the second PtdIns ( 3 ) P wave ( Figure 5H ) . To determine the functional relationship between MTM-1 and the two PI3Ks , we first analyzed mtm-1 ( op309 ) ; piki-1 ( tm3171 ) double mutant animals carrying the PtdIns ( 3 ) P reporter 2xFYVE::GFP . In both embryos and adult hermaphrodite gonads , mtm-1 ( op309 ) ; piki-1 ( tm3171 ) double mutants had significantly fewer numbers of persistent cell corpses than piki-1 ( tm3171 ) single mutants ( Figure 8B–C ) . Furthermore , the percentage of phagosomes labeled with 2xFVYE::GFP was significantly higher in mtm-1 ( op309 ) ; piki-1 ( tm3171 ) double mutants than in piki-1 ( tm3171 ) single mutants ( Figure 8B–C ) . These observations indicate that the partial loss of mtm-1 function significantly rescued the defects caused by piki-1 ( tm3171 ) mutation in both PtdIns ( 3 ) P production and phagosome maturation . The suppression effect of the Ced phenotype of piki-1 ( tm3171 ) mutants by the mtm-1 ( op309 ) mutation became more evident when examined in strains not carrying Pced-12xFYVE::gfp , which might compete with endogenous PtdIns ( 3 ) P effectors for PtdIns ( 3 ) P ( Figures 8D , S6 , and Text S1 ) . More strikingly , in piki-1 ( tm3171 ) ; mtm-1 ( RNAi ) animals , the number of persistent germ cell corpses was drastically reduced to 15% of that observed in mtm-1 ( RNAi ) animals and 29% of that in piki-1 ( tm3171 ) single mutant animals ( Figure 8E ) . Therefore , piki-1 ( tm3171 ) and mtm-1 ( RNAi ) , each of which caused a strong Ced phenotype , efficiently rescued each other's defect in phagosome maturation . These results clearly indicate that MTM-1 and PIKI-1 antagonize each other's activity and underscore the importance of a balanced PtdIns ( 3 ) P production on phagosomes . vps-34 ( RNAi ) resulted in a relative weak Ced phenotype in the adult hermaphrodite gonads ( Figure 8E ) . vps-34 ( RNAi ) modestly reduced the severity of the Ced phenotype of mtm-1 ( RNAi ) animals ( Figure 8E ) . The suppression effect of vps-34 ( RNAi ) is much weaker than that of piki-1 ( tm3171 ) mutation , suggesting that PIKI-1 is the major PI3K that counteracts the activity of MTM-1 . Importantly , the strong Ced phenotype caused by the simultaneous inactivation of piki-1 and vps-34 , which totally abolished PtdIns ( 3 ) P production , was no longer suppressible by mtm-1 ( RNAi ) ( Figure 8E ) , clearly demonstrating that MTM-1 specifically targets phagosomal PtdIns ( 3 ) P , the products of PIKI-1 and VPS-34 .
Among all PtdIns ( 3 ) P-mediated cellular events , phagosome maturation in C . elegans provides a unique opportunity for studying the molecular mechanisms and the functional significance of the temporal changes of PtdIns ( 3 ) P on intracellular membranes because the dynamics of PtdIns ( 3 ) P can be readily monitored , using time-lapse microscopy , on a relatively large object ( diameter >2 µm ) throughout the entire maturation process ( ∼50 min or longer ) . In this report , using genetic and live-cell imaging approaches , we identified a temporal regulation mechanism that programs the two-wave cycling pattern of PtdIns ( 3 ) P on phagosomal surfaces ( Figure 9 ) . This mechanism is executed by PIKI-1 and VPS-34 , two PI 3-kinases that sequentially produce PtdIns ( 3 ) P on phagosomes , and MTM-1 , a PI 3-phosphatase that dephosphorylates phagosomal PtdIns ( 3 ) P . Our findings demonstrate that the precisely regulated production and turnover of phagosomal PtdIns ( 3 ) P are both essential for phagosome maturation and the consequential degradation of apoptotic cells . The physiological functions of Class II PI3Ks are only starting to be understood . Mammalian Class II PI3Ks are implicated in the production of a dynamic PtdIns ( 3 ) P pool on the plasma membrane in response to external stimuli [27] , [41]–[43] . Certain Class II PI3Ks are also known to establish various intracellular PtdIns ( 3 ) P pools [28] , [29] , [44] . However , the function of Class II PI3Ks in phagosome maturation was not known . Previously , Class III PI3K Vps34 was thought to be the primary kinase that generated PtdIns ( 3 ) P on the surface of phagosomes [12] , [13] , [16] . Here we identified the novel function of PIKI-1 , the only C . elegans Class II PI3K , in producing a dynamic pool of PtdIns ( 3 ) P on nascent phagosomes and in initiating the degradation of apoptotic cells inside phagosomes . The combined activities of PIKI-1 and VPS-34 account for most if not all PtdIns ( 3 ) P molecules on the surface of phagosomes . To our knowledge , this is the first example that two PI3Ks , which belong to different classes , contribute to overlapping pools of PtdIns ( 3 ) P on a particular organelle and coordinately regulate the same cellular event . In piki-1 and vps-34 single mutants and vps-34; piki-1 double mutants , defects in PtdIns ( 3 ) P-production correlate quantitatively with defects in cell-corpse removal , indicating that the lack of phagosomal PtdIns ( 3 ) P is the primary cause for the Ced phenotype . Our observations further indicate that the contribution of PIKI-1 and VPS-34 to phagosome maturation depends on the tissue context: in the adult hermaphrodite gonad , PIKI-1 apparently plays a major role in producing phagosomal PtdIns ( 3 ) P . The accumulation of persistent cell corpses could result from defects in either the engulfment or the degradation of apoptotic cells , two consecutive cellular events executed through different mechanisms [4] , [45] . A previous report suggested that piki-1 mutants were specifically defective in the engulfment of apoptotic cells based on the observation of extra cell corpses in animals , which displayed the unique button-like morphology under DIC microscope [46] . However , persistent cell corpses , which could result from either inefficient engulfment or defective degradation , all display a similar DIC morphology [8] , [20] , [21] , [37] . Thus , the DIC phenotype of cell corpses does not allow the defects in engulfment to be distinguished from that in phagosome maturation . In this study , through both fluorescence and electron microscopy , which are capable of distinguishing unengulfed cell corpses from engulfed ones , we found that PIKI-1 and VPS-34 specifically control the degradation but not the engulfment of apoptotic cells . Our results indicate that the defect of piki-1 mutants that resulted in the accumulation of cell corpses was misinterpreted previously [46] . Our results are also consistent with the observation that PtdIns ( 3 ) P , the molecule produced by PIKI-1 and VPS-34 , appears on phagosomal surfaces only after engulfment is complete [7] , [8] . Mammals have three Class II PI3K isoforms , which display differential expression patterns ( reviewed in [30] ) . We propose that one or more mammalian Class II PI3Ks might act in phagocytes such as macrophages to promote the degradation of apoptotic cells and other kinds of phagosomal contents such as invading pathogens through producing phagosomal PtdIns ( 3 ) P . We have found that PIKI-1 and VPS-34 play differential roles in establishing the dynamic PtdIns ( 3 ) P pattern on phagosomal surfaces: whereas PIKI-1 initiates the production of PtdIns ( 3 ) P on nascent phagosomes , VPS-34 acts to maintain the level of phagosomal PtdIns ( 3 ) P for the subsequent period ( Figure 9A–B ) . The functional period of PIKI-1 on phagosomes corresponds to the first half of the initial PtdIns ( 3 ) P wave , whereas VPS-34 function is required afterwards , covering the time periods corresponding to the latter part of the initial PtdIns ( 3 ) P wave and the entire second wave of PtdIns ( 3 ) P ( Figure 9A–B ) . As inactivating either one of the two kinases perturbs the phagosome maturation , we propose that the proper timing of the initial production and the continuous presence of PtdIns ( 3 ) P on phagosomes are both crucial for this event . The molecular mechanism behind the sequential activation of PIKI-1 and VPS-34 is currently under investigation . PIKI-1::GFP is recruited to extending pseudopods and nascent phagosome , prior to the first appearance of PtdIns ( 3 ) P , consistent with the role of PIKI-1 in initiating PtdIns ( 3 ) P production on nascent phagosomes . The phagosome association of PIKI-1 relies on DYN-1 , the key organizer of phagosome maturation events , further suggesting that PIKI-1 might be recruited by DYN-1 , directly or indirectly , to its site of action ( Figure 9C ) . This finding places PIKI-1 under the control of the phagosome maturation pathway initiated by the phagocytic receptor CED-1 ( Figure 9C ) [8] . Mammalian dynamin 2 was reported to directly interact with Vps34 , and was proposed to recruit Vps34 to phagosomal surfaces [13] . On the other hand , mammalian Vps34 is one of the Rab5 effectors [2] , [47]–[49] . In C . elegans , we and others have found that RAB-5 is required for the production of phagosomal PtdIns ( 3 ) P ( Figure S4 and Text S1 ) [13] . We have also observed that RAB-5 is recruited to the surfaces of nascent phagosomes prior to the time period when VPS-34's activity is needed for PtdIns ( 3 ) P production ( Figures 5 and S7 ) . These observations indicate that RAB-5 might also participate in recruiting or activating VPS-34 on phagosomes ( Figure 9C ) . Vps34 controls multiple membrane trafficking events , some of which , such as endocytosis , are essential for viability ( reviewed in [2] ) . Unlike the vps-34 ( m−z− ) null mutants , piki-1 homozygous deletion mutants are viable and fertile , and do not display any obvious endocytosis defects ( Figures S1B , S3 ) , suggesting that PIKI-1 specifically functions in promoting phagosome maturation but not in other essential cellular events . Previous studies of Vps34 established the principle that PtdIns ( 3 ) P is important for phagosome maturation [12] , [13] , [16] . We went one step further by developing a strategy ( the vps-34 ( m−z− ) ; piki-1 double mutation ) that completely blocked phagosomal PtdIns ( 3 ) P production , which allowed us to determine quantitatively how fundamental the role of PtdIns ( 3 ) P is in initiating phagosome maturation . The vps-34 ( m−z− ) ; piki-1 double mutations resulted in a Ced phenotype as severe as that caused by mutations in known key regulators and executors of cell-corpse degradation , such as mutations in DYN-1 and triple mutations that completely inactivate SNX-9 , SNX-1 , and SNX-6 , three PtdIns ( 3 ) P effectors [20] , [21] . These observations demonstrate that PtdIns ( 3 ) P is absolutely required for the degradation of apoptotic cells during animal development , rather than acting as one of multiple parallel contributing factors . VPS-34 has been implicated in the recruitment of RAB-5 and RAB-7 to phagosomes [13] . We have revealed that the absence of phagosomal PtdIns ( 3 ) P prevents the phagosomal recruitment of SNX-1 and LST-4/SNX-9 as well as three Rab GTPases , RAB-5 , RAB-2 , and RAB-7 . These proteins play important roles in facilitating the fusion of particular types of intracellular organelles such as endosomes and lysosomes to phagosomes . [8] , [13] , [21] , [38] , [50] . Furthermore , LST-4/SNX-9 also helps stabilize the association of DYN-1 with phagosomes [21] . These findings further illuminate the molecular mechanisms used by PtdIns ( 3 ) P to initiate phagosome maturation . MTM1 , a member of the myotubularin PI phosphatase family , helps maintain the optimal level of PtdIns ( 3 ) P on the surface of multiple kinds of intracellular membranes by converting PtdIns ( 3 ) P to PtdIns ( reviewed in [29] , [31] ) . Here we have revealed that MTM-1 controls cell-corpse degradation through regulating the turnover of phagosomal PtdIns ( 3 ) P ( Figure 9 ) . That MTM-1 function is important for the degradation of apoptotic-cell has also been reported independently [40] . The mtm-1 ( op309 ) mutation slows down phagosome maturation in embryos . Likewise , mtm-1 ( RNAi ) results in a dramatic phagosome maturation arrest in adult gonads . These results demonstrate that despite being a negative regulator of PtdIns ( 3 ) P , MTM-1 is an essential positive regulator for phagosome maturation . Furthermore , the piki-1 deletion mutation and mtm-1 ( RNAi ) mutually suppresses the phagosome maturation defects of each other . In addition , vps-34 ( RNAi ) partially suppresses the phagosome maturation defects caused by mtm-1 ( RNAi ) . On the contrary , mtm-1 ( RNAi ) 's suppression effect no longer exists in piki-1; vps-34 ( RNAi ) double mutant backgrounds . Together , these observations demonstrate that MTM-1 directly antagonizes the function of the two PI3Ks on phagosomes by dephosphorylating their product , PtdIns ( 3 ) P . The association of MTM-1 with the extending pseudopods and nascent phagosomes correlates , completely and partially , with the PtdIns ( 3 ) P-producing activities of PIKI-1 and VPS-34 on phagosomes , respectively , further supporting the model that MTM-1 antagonizes the activities of PIKI-1 and VPS-34 and results in the gap period between the two PtdIns ( 3 ) P waves on phagosomes ( Figure 9A and B ) . During the PtdIns ( 3 ) P ( − ) gap period , a low-level activity of MTM-1 might remain on phagosomal surfaces despite that the enrichment of GFP::MTM-1 on phagosomal surfaces is below the detection capacity of our fluorescence microscope . We propose that the prompt dephosphorylation of phagosomal PtdIns ( 3 ) P might be critical for the timely dissociation of certain initial phagosome maturation factors from phagosomes as well as the subsequent association of certain other , perhaps yet-to-be identified maturation factors that act at later stages of phagosome maturation ( Figure 9B ) . The dynamic oscillation of phagosomal PtdIns ( 3 ) P thus would enable phagosomes to interact with various signaling modules and intracellular organelles in sequence , and promote the step-wise progression of phagosome maturation ( Figure 9B ) . It would be of general interest to identify each of the time-sensitive phagosome maturation factors subject to this regulation . Recently , MTM-1 was reported to negatively regulate the engulfment of apoptotic cells , one step prior to the degradation of apoptotic cells [40] , [46] . Zou et al . [46] proposed that MTM-1 negatively regulates the level of PtdIns ( 3 ) P on the plasma membrane produced by VPS-34 and PIKI-1 during engulfment , based on the observations that inactivating piki-1 and vps-34 resulted in the accumulation of persistent cell corpses and that the ability of MTM-1 to suppress the engulfment defect of ced-1 and ced-6 mutants relied on the functions of PIKI-1 and VPS-34 . Zou et al . 's work identified the antagonizing genetic interaction between MTM-1 and the two PI3Ks during cell-corpse removal [46] . However , the dynamic PtdIns ( 3 ) P appearance pattern and the timing of PIKI-1 and VPS-34 functions do not support a role of PtdIn ( 3 ) P in engulfment . In both mammalian cells and C . elegans , the production of PtdIns ( 3 ) P is initiated on nascent phagosomes after the completion of engulfment , and PtdIns ( 3 ) P has not been detected on the extending pseudopods [7] , [8] , [10]–[13] . In fact , the experimental results reported here revealed that a precisely regulated system , composed of MTM-1 , PIKI-1 , and VPS-34 , promotes the degradation but not the engulfment of apoptotic cells , through an accurate temporal regulation of phagosomal PtdIns ( 3 ) P . On the other hand , Neukomm et al . ( 2011 ) , who independently identified the function of MTM-1 as a negative regulator of cell-corpse engulfment , proposed that during engulfment , MTM-1 performs its function through dephosphorylating another substrate , PtdIns ( 3 , 5 ) P2 [31] , [40] . C . elegans MTM-1 might play opposite roles in engulfment and degradation , two consecutive steps of apoptotic-cell removal , by down-regulating different phosphoinositide species . This study , together with a few recent reports , have highlighted the physiological importance of the antagonizing activities of PI kinases and phosphatases in the regulation of PtdIns ( 3 ) P dynamics on various intracellular membranes [29] , [51] . Like the cycling of small GTPases between GDP- and GTP-bound states and the reversible phosphorylation of key protein factors , the reversible phosphorylation of various phophoinositide species on defined membrane domains is likely to be utilized as a common strategy for driving the progression of multi-step biological events .
C . elegans strains were grown at 20°C as previously described [52] . The N2 Bristol strain was used as the reference wild-type strain . Mutations are described in [53] , the Worm Base ( www . wormbase . org ) , and this work , except when noted otherwise: LGI , vps-34 ( h510 ) [35] , unc-108/rab-2 ( n3263 ) [7] , mtm-1 ( op309 ) [40]; LGII , rab-7 ( ok511 ) [8]; LGIV , ced-5 ( n1812 ) ; LGV , unc-76 ( e911 ) ; LGX , piki-1 ( tm3171 ) , piki-1 ( ok2346 ) , dyn-1 ( n4039 ) [20] . Strains carrying the piki-1 ( tm3171 ) and piki-1 ( ok2346 ) alleles were characterized after being out-crossed for four times . Transgenic worms were generated by microinjection as previously described [54] . Plasmids were co-injected with a marker pUNC-76 [unc-76 ( + ) ] [55] into unc-76 ( e911 ) mutant adult hermaphrodites and transgenic animals were identified as non-Unc animals . Transgenes are maintained in animals as extra-chromosomal arrays and , when necessary , introduced into different genetic backgrounds by crosses . The piki-1 cDNA was amplified from a mixed-stage C . elegans cDNA library ( Z . Zhou and H . R . Horvitz , unpublished results ) using polymerase chain reaction ( PCR ) . Pced-1piki-1::gfp was constructed by cloning the piki-1 cDNA into the multi-cloning sites of pZZ829 , a plasmid carrying Pced-1 , a 3′ gfp tag , and the unc-54 3′ UTR [20] . The overlap extension PCR method [56] was used to delete the DNA sequence encoding the PI kinase domain from Pced-1piki-1::gfp and generate Pced-1piki-1 ( Δkinase ) ::gfp . To generate Pced-1piki-1 ( K1059A ) ::gfp , the K1059A mutation was introduced into Pced-1piki-1::gfp using the QuickChange Site-directed Mutagenesis Kit ( Stratagene ) . To make Pced-1gfp::mtm-1 , mtm-1 cDNA was amplified from the mixed-stage C . elegans cDNA library using PCR and subsequently cloned into pZZ956 , a plasmid carrying Pced-1 , a 5′ gfp tag , and the unc-54 3′ UTR . DIC microscopy was performed with an Axionplan 2 compound microscope ( Carl Zeiss , Inc . ) equipped with Nomarski DIC optics , a digital camera ( AxioCam MRm; Carl Zeiss ) and imaging software ( AxioVision; Carl Zeiss ) . Worms were immobilized with 25 mM NaN3 , mounted on 4% agarose pads , and observed under DIC microscopy . Somatic embryonic cell corpses were scored in the embryos at different developmental stages as described by [33] . Germ cell corpses were scored in one of the two gonadal arms of synchronized adult hermaphrodites at indicated time-points post-L4 stages [20] . An Olympus IX70-Applied Precision DeltaVision microscope equipped with standard epifluorescent filter sets and Photometris Coolsnap digital camera was used to capture fluorescence images , which were deconvolved and processed by the Applied Precision SoftWoRx software . To score the numbers of somatic or germ phagosomes labeled by GFP or mRFP reporters on their surfaces , z-sections of DIC and fluorescent images spanning the entire depth of embryos or adult gonads , respectively , were captured and analyzed [8] . To monitor the dynamic subcellular localization pattern of various GFP reporters during the engulfment and degradation of cell corpses C1 , C2 , and C3 , the procedure followed an established protocol [33] . Briefly , time-lapse recording started at 310–320 min post-first cleavage and lasted for 60–120 min at 1 or 2 min intervals . At each time point , 10–15 serial Z-sections at a thickness of 0 . 5 µm were recorded , starting from the ventral surface of embryos . Signs such as embryo elongation and movement were closely monitored to ensure that the embryo being recorded developed normally . Whenever necessary , fluorescence signal intensity was measured and images were analyzed using the ImageJ software [33] . To compare fluorescence signal intensity on phagosomes in wild-type or mutant embryos , the images were captured using the same microscopic parameters , processed , and analyzed using the same procedures . RNAi experiments were performed using feeding protocol as previously described [57] . In brief , mid-L4-stage hermaphrodites ( for vps-34 ( RNAi ) experiments in Figures 1 and S1 ) or synchronized L1-stage hermaphrodites ( for mtm-1 ( RNAi ) and vps-34 ( RNAi ) experiments in Figure 8 ) were transferred to RNAi plates that contained E . coli strain HT115 transformed with RNAi constructs . For RNAi experiments starting at L1 stage , mid-L4-stage animals were retransferred to fresh RNAi plates . 48 h after L4 stage , the number of germ cell corpses per gonad arm were scored under DIC and fluorescent microscopy . The RNAi-by-feeding vector pPD129 . 36 [58] was used as a negative control for RNAi . The RNAi feeding constructs for vps-34 were generated by cloning PCR-amplified cDNA fragments ( vps-34 ( N ) and vps-34 ( C ) ) corresponding to the N- and C-terminal halves of VPS-34 into vector pPD129 . 36 using the following primers: vps-34 ( N ) ( 5′-ttgggaacacgaggatgatg-3′ and 5′-gttcaggatcagctacacag-3′ ) , vps-34 ( C ) ( 5′-taaaggagtccatc-3′ and 5′-ttgtcaagatgacgatcacc-3′ ) . The RNAi feeding constructs for mtm-1 were generated by cloning PCR-amplified cDNA fragments corresponding to the N- and C-terminal halves of MTM-1 into vector pPD129 . 36 using the following primers: mtm-1 ( N ) ( 5′-gcgccccgggatggattcacaatttattg-3′ and 5′-atatcccgggctcgccgagcttctttac-3′ ) , mtm-1 ( C ) ( 5′-aaaggaaattttcagccaatgtt-3′ and 5′-tgcacataaagaaagcaaaatga-3′ ) . Electron microscopy was performed as previously described [59] . Briefly , 2-d-old adult worms were fixed in 0 . 67% gluteraldehyde and 0 . 67% Osmium tetroxide in 10 mM HEPES buffer with microwave fixation 2 times at 90 W for 2 min/ON , 2 min/OFF , 2 min/ON . Worms were cut at the vulva and fixed for 1 h on ice . Worms were then fixed in 2% osmium tetroxide in 10 mM HEPES by microwaving two times at 90 W or 2 min/ON , 2 min/OFF , 2 min/ON and were subsequently incubated on ice for 3 h . Further processing was performed as previously described [59] . Standard procedures were used to generate and stain 50–60 nm sections . Electron microscopy was performed and images were captured with a digital camera . | During animal development and in adulthood many cells are programmed to die by an active process called apoptosis . These dead or dying apoptotic cells are swiftly taken up by scavenger cells into membrane-bound compartments—phagosomes—where they are subsequently degraded when other intracellular organelles containing digestive enzymes fuse with phagosomes—a process called phagosome maturation . Phagocytosis of apoptotic cells is important for tissue remodeling in development and to prevent harmful inflammatory and autoimmune responses . In nematode worms—a model organism in which to study apoptosis—phagosome maturation is accompanied by two waves of the signaling molecule phosphatidylinositol 3-phosphate ( PtdIns ( 3 ) P ) in this compartment: one that forms soon after the formation of the phagosome and lasts for 10–15 minutes , and a second , weaker one 10 minutes later that lasts until the apoptotic cell is fully digested . In this study , we investigated the mechanism that regulates the timing and length of these two waves . We found that they are established by the sequential and combined action of three enzymes: two phosphoinositide 3-kinases , which add a phosphate group to the 3′ site of PtdIns , and one phosphoinositide 3-phosphatase , which removes it . We showed that inactivation of both kinases depleted phagosomes of PtdIns ( 3 ) P and resulted in the arrest of phagosome maturation and degradation of apoptotic cells . In addition , the timely turnover of PtdIns ( 3 ) P catalyzed by the phosphatase was critical for the step-wise progress of phagosome maturation . Our findings suggest that reversible phosphorylation of phophoinositides , catalyzed by distinct sets of kinases and phosphatases , might be a general mechanism to drive multi-step intracellular membrane trafficking events . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"models",
"developmental",
"biology",
"cellular",
"structures",
"model",
"organisms",
"molecular",
"cell",
"biology",
"cell",
"biology",
"genetics",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Two PI 3-Kinases and One PI 3-Phosphatase Together Establish the Cyclic Waves of Phagosomal PtdIns(3)P Critical for the Degradation of Apoptotic Cells |
Identifying the major routes of disease transmission and reservoirs of infection are needed to increase our understanding of disease dynamics and improve disease control . Despite this , transmission events are rarely observed directly . Here we had the unique opportunity to study natural transmission of Bordetella bronchiseptica – a directly transmitted respiratory pathogen with a wide mammalian host range , including sporadic infection of humans – within a commercial rabbitry to evaluate the relative effects of sex and age on the transmission dynamics therein . We did this by developing an a priori set of hypotheses outlining how natural B . bronchiseptica infections may be transmitted between rabbits . We discriminated between these hypotheses by using force-of-infection estimates coupled with random effects binomial regression analysis of B . bronchiseptica age-prevalence data from within our rabbit population . Force-of-infection analysis allowed us to quantify the apparent prevalence of B . bronchiseptica while correcting for age structure . To determine whether transmission is largely within social groups ( in this case litter ) , or from an external group , we used random-effect binomial regression to evaluate the importance of social mixing in disease spread . Between these two approaches our results support young weanlings – as opposed to , for example , breeder or maternal cohorts – as the age cohort primarily responsible for B . bronchiseptica transmission . Thus age-prevalence data , which is relatively easy to gather in clinical or agricultural settings , can be used to evaluate contact patterns and infer the likely age-cohort responsible for transmission of directly transmitted infections . These insights shed light on the dynamics of disease spread and allow an assessment to be made of the best methods for effective long-term disease control .
Containing and ultimately eliminating infectious disease remains a central goal for many animal and public health officials . Dissecting disease transmission – in terms of identifying the routes and potentially heterogeneous rates of disease spread [1] – is an essential step in devising or optimizing intervention strategies aimed at pathogen eradication [1] , [2] . This is because heterogeneities in transmission that arise due to for example age- or sex-specific differences among individuals [2] , [3] can greatly affect invasion and eradication criteria [1] . Unfortunately , precise measurements of transmission remain elusive due to the immense difficulties associated with identifying the nature of a potential contact , the probability of infection given a contact [4] and important drivers of heterogeneities in transmission [2] , [3] , [5] . A key reason for these difficulties is that transmission events are rarely observed directly , with some notable exceptions [6] . One useful approach that can shed partial light on the transmission process is to measure the force-of-infection ( FOI: λ ) , or the per capita conversion rate of susceptible hosts [7] . The simplest way to think about the FOI , is that over a short interval of time – say from time t to t+Δ – the probability that a disease negative individual becomes disease positive is λΔ . The most popular way to estimate λ is through use of the observed age-specific prevalence ( or the proportion of individuals that are disease positive in a cross-sectional sample ) , due to the ease with which it is measured in most populations [8] . Indeed , FOI estimates have been calculated from age-prevalence data for human , and to a lesser extent , wildlife infections [3] , [9] , [10] , [11] , [12] , [13] , [14] , [15] , [16] , [17] , [18] , [19] . Estimating the FOI can help identify those age-classes responsible for transmission and evaluate the relative effects of each group on overall transmission . Here we evaluate the relative effects of sex , age and social structure on the transmission dynamics of the respiratory pathogen Bordetella bronchiseptica within a commercial rabbitry of New Zealand White ( NZW ) rabbits . In doing so we illustrate how analysis of age-prevalence data can be used to estimate the age-specific FOI . The importance of social organization in B . bronchiseptica transmission is also considered . To test for litter-based transmission events – for example , from mother to offspring or between siblings – we checked for significant correlation among the fate of siblings by using a litter-based random-effects binomial regression to estimate the importance of horizontal versus pseudovertical transmission [20] , [21] . The statistical tools we employ here are general and can be applied to a range of directly transmitted medical and veterinary diseases to help shed light on the dynamics of disease spread and allow an assessment to be made of the best methods for effective long-term disease control . The Bordetella genus contains three closely related gram-negative bacteria that cause respiratory infections in humans and other mammals [22] . Whereas B . pertussis and B . parapertussis largely infect humans and cause the acute respiratory disease known as whooping cough [23] , B . bronchiseptica typically causes chronic infections in a wide range of mammals [24] . Indeed B . bronchiseptica infection is often endemic in agricultural settings – including commercial rabbitries [25] , [26] – where rapid spread and persistent infection make it difficult to control [23] . Despite its widespread nature , there is a paucity of data describing the epidemiology of B . bronchiseptica in terms of both the main route ( s ) of , and likely cohort ( s ) responsible for disease transmission . As a respiratory infection , the major physical route of transmission is oral-nasal via direct aerosol droplets [27] , [28] . Based on the published literature [26] , [27] , [29] , we propose a set of plausible routes of transmission within a commercial rabbitry would include: These ( not mutually exclusive ) possible routes of transmission may result in the prevalence of infection changing with age in different ways which can be related to different underlying hazard models/FOI patterns ( Figure 1 ) . As we will explore below , each of these transmission possibilities translates into a specific prediction which can be tested using our statistical framework . To identify the parsimonious hypothesis , we applied a piece-wise constant model for the age-specific FOI [9] , [11] , [15] . Since B . bronchiseptica is an endemic persistent infection [23] , we used a catalytic framework ( which assumes a one-way flow from susceptible to infected ) . The importance of sex and location ( facility building ) and time of sampling on FOI estimates was also determined . We considered the importance of social mixing and organization in B . bronchiseptica transmission using random effect logistic regression estimates to control for litter as confounding variable in transmission models . In parallel , we took a molecular epidemiological approach to investigate whether strain-specific differences existed in the epidemiological pattern of infection [30] , [31] .
All protocols involving rabbits were approved by the Institutional Animal Care and Use Committee ( IACUC ) at the Pennsylvania State University according to the guidelines of the American Association for Laboratory Animal Science . This study was conducted at a commercial rabbitry which breeds NZW rabbits . The rabbitry comprised of three separate animal breeding buildings ( Table 1; buildings A - C ) . Background health checks – in the form of comprehensive monthly pathology reports testing for >17 pathogens – have been carried out since January 2003 ( n = 2 to 4 rabbits/month/building ) . These reports show that B . bronchiseptica has been endemic in the rabbitry since testing began and that of the other pathogens screened , only non-pathogenic Eimeria species ( intestinal coccidia ) are occasionally isolated . Importantly , our rabbitry is Pasteurella multocida free – infection with this respiratory pathogen has long been associated with upper respiratory disease ( URD ) in rabbits [32] – with no URD reported in the last 30 years . Kits were weaned at 4–5 weeks of age and ‘weanlings’ segregated by sex and co-housed in sibling pairs . Rabbits of good breeding stock were selected as ‘future breeders’ and housed in pairs . The remaining ‘stock’ rabbits were housed singly and sold at 8–10 weeks old . ‘Breeder’ rabbits were initially bred at 5 or 6 months of age for females and males respectively . Breeders were , housed individually and rebred when litters were weaned . For all rabbits included in this study , the date of sampling , building and rabbit identification number were ascertained along with nasal swab ( BD sterile swab , product # 220518 ) . A total of eight sampling efforts were carried out from November 2006 to September 2008 , culminating in the collective nasal swabbing of 602 rabbits total . All nasal swabs were streaked onto Bordet-Gengou ( BG ) agar ( Difco ) containing 10% sheep's blood ( Hema Resources ) with 20 µg/mL streptomycin ( Sigma ) as soon as possible after collection and incubated at 35°C for ∼3–5 days . The sampling strategies were as follows: We used Multi-Locus Sequence Typing ( MLST ) analysis [33] , [34] to determine the phylogenetic relationships among 90 B . bronchiseptica isolates from 4 sampling efforts across three rabbitry buildings , as previously described for Bordetella . Briefly , genomic DNA from each isolate was obtained using a DNAeasy Tissue Kit ( Qiagen ) and nucleotide sequences were determined for internal regions of seven housekeeping genes for all 90 isolates ( see the Bordetella MLST database at http://pubmlst . org/bordetella ) . All alleles were double stranded sequenced at The Pennsylvania State University Genomic Sequencing Center and an allele number was assigned to each unique allele sequence . The combination of the allele numbers at the seven loci defines the sequence types ( ST ) or allelic profile of each strain [33] , [35] .
Two different catalytic models were fitted to the data: a piecewise constant FOI with 3 age-intervals ( Figure 2b ) , corresponding to the hypothesized routes of transmission outlined in the Introduction and Figure 1 and a constant FOI corresponding to the null hypothesis . The 3 age-interval model fits the data better ( having the lowest AIC value of 219 . 39 ) than the null hypothesis of a constant FOI ( ΔAIC of model ( c ) versus ( a ) = 39 . 6 ) . The model results show close correspondence between the observed and expected prevalence data ( Figure 2a ) . Both data and model fits exhibit a rapid increase in prevalence during the first and second age-classes ( i . e . in rabbits up to 5 months; Figure 2a ) . During the first month of life , the estimated FOI is substantial ( Figure 2b; FOI = 0 . 16 month−1 ) and peaks in the second age-class ( Figure 2b; FOI = 0 . 20 month−1 ) , with the older age classes ( from 5–30 months of age ) having the lowest FOI estimates ( Figure 2b; from 3 . 3×10−4 to virtually zero ) . Next we examined whether gender differences existed for FOI estimates . No differences between sexes were found in the FOI estimates . In the younger age-classes , the prevalence data and model estimates peaked in the second age-class resulting in positive FOI estimates in young weaned kits ( 1 to 4 month olds: Figure 3a–d ) . In the older age-classes , B . bronchiseptica prevalence asymptoted and subsequently fell for both sexes , with a concomitant decline in the FOI estimates ( Figure 3a–d ) . To examine the likelihood of becoming infected from an infected sibling , we ran a binomial regression on the experimental sibling-to-sibling transmission experiment ( see Sampling Strategy Two in M&M for details ) . Being co-housed with an infected sibling increased the probability of becoming B . bronchiseptica positive ( Figure 4a: co-housed with infected sibling: Z = 2 . 42 , p = 0 . 016 ) , such that uninfected kits were 3 . 85 times more likely to become infected when they were co-housed with an infected- compared to an uninfected-kit ( Figure 4a: 95% C . I . for odds ratio 3 . 85: 1 . 29–11 . 46 ) . None of the solitary Bordetella-free rabbits ( housed alone in isolation ) converted to disease-positive during this time . Using the maternal transmission data ( Sampling Strategy Three in M&M for details ) , the importance of sibling-to-sibling versus mother-offspring routes of transmission was investigated . First , the data revealed substantial correlation ( 0 . 53 ) among the infection fate of siblings and a highly significant litter-random effect ( litter variance = 4 . 2±0 . 7 ) , demonstrating the importance of within-litter transmission . Although the prevalence of B . bronchiseptica was significantly higher in does compared to kits ( Figure 4b; Z = 5 . 03 p<0 . 0001 ) , having an infected mother did not significantly increase the probability of kits being infected ( infected mother: Z = 1 . 74 , p = 0 . 09 ) . Nor was there any significant relationship between the litter random effect and the mothers prevalence status ( Z = −1 . 05×10−15 , p = 1 . 0 ) . We used MLST analysis to characterize the relationship between 90 isolates collected from four sampling efforts across rabbitry buildings . All isolates were of sequence type ( ST ) -14 , which is a member of the B . bronchiseptica complex I [35] . Thus , one circulating strain appears to dominate in our rabbit population .
This study demonstrates how FOI estimates coupled with random effects binomial regression analyses represent powerful tools for discerning between alternative modes of transmission for a directly transmitted pathogen . Specifically , our results support a role for sibling-to-sibling transmission among young weaned kits as a major route of B . bronchiseptica spread in the rabbit population studied ( Figure 2 ) . That the FOI reached a maximum value between 1 to 4 months of age – a time period when kits are re-housed in sibling pairs – followed by a sharp decline in the older age-classes , is consistent with high between-sibling transmission in young weanlings ( Figure 2 ) , regardless of host sex ( Figure 3 ) . Results from the binomial regression analyses further support a major role for sibling-to-sibling transmission in driving B . bronchiseptica dynamics in the rabbitry; being co-housed with an infected sibling increased the risk of infection almost 4-fold ( Figure 4a ) . In comparison , the data did not support all other potential transmission routes; namely maternal , breeder or environmental routes . These insights shed light on the dynamics of disease spread and allow an assessment to be made of the best method ( s ) for effective long-term disease control , discussed more fully below . A basic motivation for this study was to demonstrate how robust statistical tools can be used to disentangle routes and modes of transmission in humans and social animals from infection-at-age data ( within family groups ) , which is of broad medical , ecological and veterinary interest . The FOI analyses we present may have greatest application for analyzing disease dynamics in medical and agricultural settings because here one often has direct access to date-of-birth information , knowledge of the distinct mixing patterns over the lifetime of the host , as well as host infection status ( for example , by detecting a serological response in the live animal , by the polymerase chain reaction ( PCR ) or by pathogen isolation ) . One complexity which often arises in analyses of medical and agricultural diseases is clustering in the data; hosts live in families , litters or herds and once an infection is introduced , hosts within that cluster have a higher instantaneous rate of becoming infected than those outside the cluster . Our use of random effect binomial regression analysis allows us to estimate the subject-specific measure of the effect [20] and evaluate the importance of social mixing in disease spread . Thus , using the following protocol , the transmission dynamics of a range of directly transmitted infections can be analyzed by: ( 1 ) using the catalytic model and associated FOI analysis to determine the core susceptible age-class ( es ) ; ( 2 ) using random-effect binomial regression to inform on whether transmission is largely within the social group ( family/litter/herd etc ) or from an external social group; ( 3 ) carefully constructing transmission experiments , whenever possible , to test whether within-group versus between-group individuals are the dominant source of infection . What might explain heterogeneities in rabbit susceptibility to B . bronchiseptica infection; for example , the decline in B . bronchiseptica prevalence in older-age classes ( in rabbits ∼20 months of age ) ? Between-rabbit variation in protective anti-B . bronchiseptica immunity – and hence resistance to infection – is likely to at least partly explain differences in host susceptibility to infection . Indeed , recent work has shown that the protective immune response against B . bronchiseptica varies between individual rabbits , with robust serum IgG detected in some hosts for up to 5 months post infection , which correlated with clearance from the respiratory tract [28] . Given the persistent nature of B . bronchiseptica infections in rabbits – infections of 5 months were routinely recorded [28] – and other mammals [23] , the decline in prevalence we observe is unlikely to be driven by bacteria clearance and recovery . Rather , some level of enhanced immune protection in older age-classes may be responsible for conferring some level of anti-bordetella resistance . Thus , the low attack rates ( or number of reported cases per unit time in a given age-class , divided by the number in that age class ) in older-age classes likely reflect low proportions of rabbits susceptible to infection – i . e . immune , disease-negative hosts – rather than a real decline in the rate at which susceptible rabbits acquire infection . In addition , between-rabbit heterogeneities in protective anti- B . bronchiseptica immunity might also help explain differences in rabbit susceptibility to infection in the maternal- and co-housed sibling- transmission studies reported here . Is there any epidemiological support for the major route of B . bronchiseptica spread ( sibling-to-sibling ) identified using our statistical framework ? B . bronchiseptica is known to pass efficiently and spread rapidly between populations of young weaned pigs [40] , consistent with a sibling-to-sibling route for B . bronchiseptica transmission amongst young farmed animals . This would be particularly true in agricultural systems where an all-in/all-out ( the facility is completely emptied and cleaned between groups of age-matched animals which move together between phases of production ) policy of animal breeding is not practised , as is the case in the rabbitry under study . However , that our FOI estimates were above 0 . 1 before 1 month of age suggests some maternal or environmental transmission is occurring in young weanlings and may be key to initiating the sibling-to-sibling transmission which follows . Indeed , a maternal route of transmission is thought initiate B . bronchiseptica infections in swine and rabbits [26] , [27] , but that infection only becomes endemic when passed horizontally between different batches of susceptible young [27] . Interestingly , the time when FOI values peaked in young weanlings , coincided with a period where maternal protection wanes in kits – antibodies against B . bronchiseptica decreased between 2- 6 weeks of age in rabbits [25] – and could also contribute to increased susceptibility to infection observed in this age class . Thus , based on our findings and the published literature , we propose that the cycle of B . bronchiseptica infection in our rabbitry is maintained by a proportion of chronically infected breeder females and males ( the infectious reservoir ) with the majority of transmission occurring between young weaned siblings . One important application for the analytical tools presented here is in the implementation of targeted disease control programs . Given that targeting those high-risk subgroups identified as playing key roles in transmission – rather than applying disease control measures randomly – is one efficient strategy to control disease [2] , [6] , a precautionary management approach might rely on the selective removal of infected weanlings to reduce sibling-to-sibling transmission . Selective removal of breeder animals – which may represent potential maintenance hosts for B . bronchiseptica – may also improve disease control by eliminating the infectious reservoir . Indeed , pre-emptive culling based on pre-determined patterns of disease spread has been successfully used to combat the spread of foot-and-mouth disease in cattle [41] , [42] . The relationship between culling intensity and the resulting disease prevalence can be estimated when knowledge on population density and disease prevalence is available [43] . This allows estimates to be made regarding the level of culling needed to produce significant reductions in disease prevalence . The analyses presented here can be applied to a range of medical and veterinary diseases to better understand the dynamics and mechanisms of disease spread , provided they are directly transmitted and induce lifelong immunity to re-infection . For example , the disease caused by mycobacterium – the etiological agent of tuberculosis in animals including bovine and humans – is largely directly transmitted , causes a sub-acute or chronic disease state which is irreversible [14] , [15] , [18] and can be routinely confirmed via culture , making it a tractable disease for application of FOI analyses . Indeed , the tools of infectious disease quantitative epidemiology have successfully been applied to further understand Mycobacterium bovis infection dynamics in wildlife population of badgers [44] , ferrets [14] and bison [15] , [45] and has shed light on likely patterns of mycobacteria transmission in the wild . However , these tools have not been used to the same effect in agricultural settings despite the debilitating effects of this disease and the potential to improve disease control therein . Other veterinary diseases which are tractable for this type of analyses include brucellosis , bovine herpes infection , classical swine fever , bovine mastitis and atrophic rhinitis in swine , to name but a few . Finally , the FOI model presented here can be extended to include diseases with reversion to non-diseased state or non-benign diseases ( i . e . associated with increasing death rate ) , or indeed to include a period where hosts are not exposed to infection ( for example , when maternal antibodies are known to provide protection against specific diseases early in life ) similar to a guarantee time in survival analysis ( see Caley & Hone 2002 for examples of such extensions ) . Our study has some limitations . Although the method we outline can clearly reveal the age-class for which most of the new infection occurs , it cannot easily discern whether that infection is mainly within an age-class versus from a different age-class . However , once the high FOI age-class is identified , careful design of transmission experiments could confirm the likely source of infection , and such studies are underway in our University . To control and possibly eradicate infectious diseases we need a better understanding of pathogen population dynamics and structure . Indeed , only when HIV population structure was understood did the requirement for a three-cocktail HIV drug therapy become clear [46] . Knowledge of pathogen population structure is also needed to determine which disease-associated genes are under directional selection change . To this end we used MLST analysis to investigate whether strain-specific differences existed in the epidemiological pattern of infection [30] , [31] . However , only one major circulating sequence type – ST14 – was identified in our rabbits regardless of rabbit age , sex or facility building . The dominance of ST14 across our facility may be due to the successful expansion of this single serotype over time . Alternatively , a limitation in sampling could have potentially biased our results; the sequence type of only 1 colony per swabbed plate ( i . e . per rabbit ) was determined at each sampling round . Therefore if the rabbit was colonized with multiple strains we most likely detected the dominant type ( ST14 ) . More intensive sequencing typing is required to test whether the lack of genetic variation we report is real and such studies are ongoing . This study demonstrates the ease with which potential routes and reservoirs of infection can be discriminated amongst from age-prevalence data in medical , agricultural , and wildlife setting when we have access to fundamental age-prevalence data . Much remains to be done to achieve a better understanding of the complex dynamics of chronic infections and to extend this model to incorporate factors such as host immunity and parasite genetic variation . | A lack of understanding regarding determinants of infectious disease transmission has hindered improved disease control efforts . Here we had the unique opportunity to study the natural transmission of the respiratory pathogen Bordetella bronchiseptica within a commercial rabbitry . B . bronchiseptica is a directly transmitted gram-negative bacterium belonging to the genus Bordetella , which also comprises B . pertussis and B . parapertussis , the etiological agents of whooping cough in humans . In this study we estimated the importance of rabbit sex , age and social group on disease spread . To do this we first outlined a set of hypotheses about how natural B . bronchiseptica infections may be transmitted between rabbits . We then discriminated between these hypotheses by estimating the rate at which susceptible individuals acquire infection ( or force-of-infection ) using B . bronchiseptica age-prevalence data . The importance of social structure in disease spread was then evaluated using random-effect binomial regression . Our results support young weanlings as the age cohort primarily responsible for B . bronchiseptica transmission and demonstrate that easy to collect age-prevalence data can be used to infer the likely age-cohort responsible for disease transmission . Such insights shed light on the dynamics of disease spread and allow an assessment to be made of the best methods for effective disease control . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"ecology/evolutionary",
"ecology",
"respiratory",
"medicine/respiratory",
"infections",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"infectious",
"diseases/bacterial",
"infections",
"ecolog... | 2010 | Identifying the Age Cohort Responsible for Transmission in a Natural Outbreak of Bordetella bronchiseptica |
The dengue virus has a single-stranded positive-sense RNA genome of ∼10 . 700 nucleotides with a single open reading frame that encodes three structural ( C , prM , and E ) and seven nonstructural ( NS1 , NS2A , NS2B , NS3 , NS4A , NS4B , and NS5 ) proteins . It possesses four antigenically distinct serotypes ( DENV 1–4 ) . Many phylogenetic studies address particularities of the different serotypes using convenience samples that are not conducive to a spatio-temporal analysis in a single urban setting . We describe the pattern of spread of distinct lineages of DENV-3 circulating in São José do Rio Preto , Brazil , during 2006 . Blood samples from patients presenting dengue-like symptoms were collected for DENV testing . We performed M-N-PCR using primers based on NS5 for virus detection and identification . The fragments were purified from PCR mixtures and sequenced . The positive dengue cases were geo-coded . To type the sequenced samples , 52 reference sequences were aligned . The dataset generated was used for iterative phylogenetic reconstruction with the maximum likelihood criterion . The best demographic model , the rate of growth , rate of evolutionary change , and Time to Most Recent Common Ancestor ( TMRCA ) were estimated . The basic reproductive rate during the epidemics was estimated . We obtained sequences from 82 patients among 174 blood samples . We were able to geo-code 46 sequences . The alignment generated a 399-nucleotide-long dataset with 134 taxa . The phylogenetic analysis indicated that all samples were of DENV-3 and related to strains circulating on the isle of Martinique in 2000–2001 . Sixty DENV-3 from São José do Rio Preto formed a monophyletic group ( lineage 1 ) , closely related to the remaining 22 isolates ( lineage 2 ) . We assumed that these lineages appeared before 2006 in different occasions . By transforming the inferred exponential growth rates into the basic reproductive rate , we obtained values for lineage 1 of R0 = 1 . 53 and values for lineage 2 of R0 = 1 . 13 . Under the exponential model , TMRCA of lineage 1 dated 1 year and lineage 2 dated 3 . 4 years before the last sampling . The possibility of inferring the spatio-temporal dynamics from genetic data has been generally little explored , and it may shed light on DENV circulation . The use of both geographic and temporally structured phylogenetic data provided a detailed view on the spread of at least two dengue viral strains in a populated urban area .
The genus Flavivirus includes 53 arthropod borne viruses that can cause severe encephalitis , hemorrhagic fever and febrile illness in humans [1] . Dengue viruses ( DENV ) , Saint Louis Encephalitis virus ( SLEV ) , and Yellow Fever virus ( YFV ) belong to this genus and are important public health issues in most tropical and subtropical countries [2] . Dengue is the most common arboviral infection all over the world [3] . Like other flaviviruses , dengue virus has a single-stranded positive-sense RNA genome of ∼10 , 700 nucleotides that is surrounded by a nucleocapsid and covered by a lipid envelope with viral glycoproteins . The RNA genome contains a single open reading frame ( ORF ) flanked by two untranslated regions ( UTRs 3′ and 5′ ) . The single ORF encodes a precursor polyprotein , which is co- and post-translationally cleaved into three structural ( C , prM and E ) and seven nonstructural ( NS1 , NS2A , NS2B , NS3 , NS4A , NS4B , NS5 ) proteins [4] . The disease is caused by four antigenically distinct virus serotypes ( DENV 1–4 ) and each serotype harbors phylogenetically defined genotypes [5] that have been experiencing massive bursts of genetic diversity as a consequence of the exponentially increasing human population during the last 200 years [5] , [6] , [7] . Dengue infection may be asymptomatic and lead to undifferentiated fever , dengue fever ( DF ) or evolve to more serious conditions ( dengue hemorrhagic fever ( DHF ) or dengue shock syndrome ( DSS ) ) [3] , [8] . DF is an acute febrile viral disease that is characterized by headaches , biphasic fever , skin rash , retro orbital pain , leukopenia , thrombocytopenia and lymphadenopathy [3] . DHF is characterized by high fever , hemorrhagic manifestations and signs of circulatory failure . Patients presenting such symptoms may develop hypovolemic shock , leading to DSS , which can be fatal [8] . Outside Africa , the disease is transmitted mainly by the Aedes aegypti mosquito , which is widely distributed and established in all tropical countries and subtropical countries . Nearly three billion people are at risk of infection by DENV [9] . Brazil was responsible for approximately 94 . 5% of the reported dengue cases in Central and South America and 60% all over the world in 2007 . Moreover , until the 39th epidemiological week , which started in September 23rd 2007 and finished in 29th 2007 , 481 . 316 cases of DF ( out of a population of approximately 186 million people , www . ibge . gov . br/english/ ) were reported along with 1076 DHF manifestations [10] . At the same period , São Paulo State with 21% of the Brazilian population was responsible for 17% of the cases ( 82 . 684 ) . The impact of the disease is very heterogeneous in the State: the city of São José do Rio Preto – included in our study - reported 12% ( 9 . 331 ) of the occurrences in the State having only 1% of its population [11] . Even before the 2006 outbreak , dengue was endemic in São José do Rio Preto [12] . Many molecular phylogeny studies addressed particularities of the dynamics of the different dengue serotypes [6] , [7] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] . However , there is still a need to study particular outbreaks in single urban settings at a fine-grained spatio-temporal scale . In the present work we describe the pattern of spread of distinct lineages of DENV-3 virus circulating in São José do Rio Preto , São Paulo , Brazil during the 2006 outbreak and analyze the dynamics and microevolution during the outbreak .
The city of São José do Rio Preto ( SJRP ) is on the northwestern region of São Paulo State , Brazil ( 20°49′11″ S e 49°22′46″ W ) , with a total area of 434 , 10 Km2 and an urban area of 96 , 81 Km2 . The estimated population in 2007 was 424 , 114 . SJRP has a tropical climate with a mean annual temperature of 25°C and mean rainfall of 1410 mm concentrated in the summer months . The city has development indexes comparable to those of developed countries and its economy encompasses industry , services , commerce and agro-business . The urban area of the municipality is divided in 432 census tracts . The census tracts comprise 300 homes in areas defined by the Instituto Brasileiro de Geografia e Estatística — IBGE ( Brazilian Institute of Geography and Statistics ) to optimize the collection of data sets during census . Although SJRP was infested by the Aedes aegypti in 1985 , only imported dengue cases where reported until 1989 . Human to human DENV-1 transmission was first observed in 1990 . From that date , dengue cases have been reported every year , with the exception of 1992 [12] and DENV-2 and DENV-3 were introduced in 1998 and 2006 . Geo-coding of autochthonous dengue cases was done using ArcGIS 9 . 0 ( Environmental Systems Research Institute , Inc . ) . The geographic position of each patient was assumed to be the latitude and longitude of their postal code ( zip code ) obtained from sampled patient address records provided by the municipality of São José do Rio Preto . Blood samples from patients presenting acute febrile illness , with or without hemorrhagic manifestations , infection with sudden start , nausea , vomit , diarrhea , symptoms of DF and DHF were collected for Flavivirus testing in the municipal health units and hospitals , upon informed consent . This study was approved by the Ethical Review Board of the Faculdade de Medicina de São José do Rio Preto and blood collection was performed upon Written Informed Consent . The blood samples were centrifuged and the viral RNA was extracted from the serum with the QIAmp viral RNA mini kit ( Qiagen ) according to the manufacturer's instructions . The first RT-PCR was performed using Flavivirus generic primers based on the non-structural protein 5 ( NS5 ) , which is a conserved region in dengue viruses and would detect most of the circulating dengue virus in Brazil in a single PCR reaction . In the second PCR , nested assays based on multiplex or conventional systems were used with species-specific primers for virus identification [26] . The forward FG1 ( 5′TCAAGGAACTCCACACATGAGATGTACT3′ ) and reverse FG2 ( 5′GTGTCCCATCCTGCTGTGTCATCAGCATACA3′ ) primer set anneals to the NS5 gene , producing amplicons of approximately 958 bp [27] . A specific inner primer for DENV-3 ( 5′TTCCTCGTCCTCAACAGCAGCTCTCGCACT3′ ) produced amplicons with 659 bp [26] . The fragments were purified from PCR mixtures and sequenced with the BigDye v3 . 1 Terminator ( Applied Biosystens , Foster City , CA , USA ) using the forward FG1 primer and the reverse DENV-3 primer in an ABI377 automated sequencer ( Applied Biosystenss , Foster City , CA , USA ) . The products were aligned with Accelrys Gene 2 . 0 ( Accelrys Software Inc . 2006 ) . In order to type the sequenced samples , 52 reference sequences including representatives of the 4 serotypes were hand-aligned in Se-Al version 2 . 0a11 program ( data available from authors upon request ) . The dataset generated was used for phylogenetic reconstruction with the maximum likelihood criterion using a genetic algorithm method implemented in the program GARLI version 0 . 95 [28] that estimates simultaneously the best topology , branch lengths and the best values for the parameters for the General Time Reversible ( GTR ) model of nucleotide evolution with Gamma-distributed variable rates and invariant sites ( GTR+Γ+I ) . One hundred independent random runs were conducted with GARLI and the tree with highest likelihood was subsequently used as input for further topological optimization with PAUP v . 4 . 0b10 [29] , since both GARLI and PAUP calculate the same likelihood score for a tree under the same model . Support for the topology was sought after 100 bootstrap replicates with GARLI . The best demographic model among: ( i ) constant population size , ( ii ) exponential population growth and , ( iii ) logistic population growth for the data , the rate of growth ( r = Ne . g ) ( i . e . , the effective number of transmission events times the generation time of the pathogen ) , rate of evolutionary change ( μ ) ( subs/site/year ) and , the Time to Most Recent Common Ancestor ( TMRCA ) were estimated from the Coalescent using Bayesian inference with a Markov Chain Monte Carlo ( MCMC ) search method available in BEAST v1 . 6 [30] . Sequences were dated according to the day of sampling and the MCMC were run until the effective sampling size ( ESS ) for each parameter converged at values above 100 . The confidence intervals for each parameter were given by the 95% highest probability density ( HPD ) . The data was analyzed using a relaxed molecular clock ( uncorrelated lognormal ) under the constant , exponential and logistic demographic models . Because the priors were not necessarily the same , demographic models were compared by calculating the Log 10 of the Bayes factor using the harmonic mean without smoothing of the sum of the likelihoods for the coalescent and tree obtained during each MCMC for each model with TRACER v1 . 6 program [30] . First , to examine whether the samples were overly spatially or temporally structured we generated matrices of distances between the samples and compared those matrices . The matrices of distances show how each sample is relatively close to all others , considering separately their date of collection , their place of collection and their genetics . Corrected genetic distances were obtained with PAUP using the values for the GTR+Γ+I model found for the maximum likelihood tree inferred with GARLI . Geographic distances were measured along straight lines ( using geographic information system ) between samples and temporal distances were the interval between sample collections . The null hypothesis of no association between the genetic and geographic and temporal distances was assessed using partial Mantel tests [31] . Statistical distributions were generated by a Monte Carlo method randomising rows ( and corresponding columns ) in the matrix of phylogenetic distances 1 , 000 times , and calculating for each one of these permutations the partial correlation coefficient between the two matrices , controlling for the third ( geographic ) matrix . The one-tailed probability of a type I error ( i . e . , rejection of a true null hypothesis ) was taken as the proportion of correlation coefficients sorted in ascending order that were higher than or equal to the obtained correlation coefficient . We accepted probabilities below α = 0 . 05 as statistically significant . Second , to obtain the putative spatio-temporal pattern of spread of dengue in SJRP we applied a simple algorithm that checks all samples ( but the first ) in a temporal sequence , and would attribute the linkage of ancestry to the ( temporally ) previous sample with higher genetic proximity ( hence , let's call it “nepotistic algorithm” ) . No obvious important geographic barriers for the circulation of the vector or host were identified within the area studied; hence we did not add any friction/cost to the movement of viruses into the algorithm . We did apply instead a limit of 0 . 00739 substitutions/site/year that we considered as the genetic distance that could have been generated by a virus replicating during 100 days at a rate of 10−4 substitutions/site/year . Therefore , if one sample was not closer in distance than this value to any of its chronologically previous samples , it was assumed that this sample resulted from another virus introduction in the locality . Because each viral sequence was obtained from a distinct patient , nodes in the virus gene genealogy can be assumed as transmission events in the human population . Therefore the basic reproductive rate of a pathogen ( R0 ) during an epidemics can be estimated as R0 = 1+D ( ln2/td ) [32] , where D is the mean time of infectiousness ( i . e . , 7 days for virus shedding in humans ) [33] and td is the doubling time of the epidemics and , since the growth rate ( r ) obtained from the viral phylodynamics equals ln2/td [30] , we estimated R0 as 1+Dr . However , the monotonic-increasing population models ( exponential en logistic ) available in BEAST do not account properly for fluctuating dynamics , such as that observed during the studied outbreak , possibly affecting the growth rate estimates using these models . This was further substantiated by the fluctuating nature of the Bayesian skyline ( BSL ) plot for dengue that showed rapid increase at the onset of the epidemic phase , followed in time by a reduction of cases at the end of the outbreak . Therefore , we also used for comparison an alternative way of estimating R0 , based on the growth phase of the epidemics alone , by deriving the force of infection at the increasing phase of the BSL as follows . The normalized median of the Bayesian estimates from the sequences analyzed , y ( t ) , was fitted to a continuous logistic curve according to the following model: ( 1 ) From equation ( 1 ) it is possible to estimate the force of infection for the data [34] , [35]: ( 2 ) where , ( 3 ) The Basic Reproduction Number , R0 , was estimated from the average force of infection , calculated from equation ( 2 ) by its equality to the number of new cases per time unit per susceptibles , according to a previous study [33]: ( 4 ) where , μ is the mosquito mortality rate and γ is the recovery rate of viraemia in humans . The mosquitoes mortality rate is the rate by which moquitoes die , on average , in each unit of time and is the inverse of the average life expectancy of each specific mosquito population . It varies from place to place and in the same place it also varies with environmental conditions , like temperature , rain pattern and othe climatic variables . We used the mosquito mortality rate of to 2 . 23×10-2/day previously estimated for SJRP [33] .
We obtained 399 bp-long sequences of a portion of the NS5 gene from viral genomic RNA amplified directly from the blood of 82 patients from 198 samples collected in the city of São José do Rio Preto for a period of 174 days , from January 12 to June 5th of 2006 , which covered the zenith of the outbreak ( i . e . , above 1000 cases per 100 , 000 inhabitants ) in April 2006 ( GenBank accession numbers from EU715692 to EU715773 ) . We were able to geo-locate 46 patients based on the addresses of their residences , from the cohort of 82 patients . The reason for not geo-coding the other 36 patients was the lack of their complete home address . By aligning the 82 sequences with the 52 references we generated a 399-nucleotide-long dataset ( without gaps ) with 134 taxa . Preliminary phylogenetic analyzes , including the 4 serotypes indicated that all samples were of serotype 3 ( data not shown ) . Furthermore , all of our samples nested within DENV-3 and were closely related to strains circulating in the isle of Martinique in the Caribbean in 2000 and 2001 and , to the DENV-3 strain Den3_BR74886 circulating in Brazil in 2002 ( Figure 1 ) . Moreover , Figure 1 indicated that , given the reference samples included in our analyses , the South American lineages were more related to lineages circulating in South East Asia . Another important finding was that 60 DENV-3 from SJRP formed a monophyletic group ( lineage 1 , shown in blue ) with 90% posterior probability , which were closely related to the remaining 22 isolates that did not form a clear monophyletic cluster ( lineage 2 and 3 , shown in orange ) that had a more basal position in the tree and that intermingled with the other South American references available from Martinique and Brazil . Our demographic analysis using a Bayesian skyline prior with BEAST generated a maximum clade credibility ( MCC ) tree with dated tips and internal nodes that indicated that both lineages split 1 to 3 years before the last sample was collected in SJRP ( Figure 1 ) . Moreover , since there was no sustained DENV- 3 epidemic transmission during this entire period in SJRP , we assumed that these lineages appeared before the 2006 season and that distinct lineages were possibly introduced into the city on different occasions . As a consequence , the major lineages were treated as distinct viral populations during subsequent demographic analyses . Since , most lineages shown in orange in Figure 1 had a common ancestor with 77% posterior probability , they were grouped into a single group ( lineage 2 ) for the sake of demographic analyses . For both groups of samples there was significant population growth initiating 6 months before the last sampling , which matches quite well with reports increasing above 10 cases per 100 , 000 in December 2005 ( Figure 2 ) . Moreover , for both DENV-3 lineages , the constant population size model was rejected ( for lineage 1 , Log 10 Bayes factor>250; lineage 2 , Log 10 Bayes factor>146 ) . Although the signature of BSL in Figure 2 is clearly logistic until around the zenith of the outbreak , the logistic model did not out-perform the exponential growth model for both lineage 1 ( Log 10 Bayes factor = 23 . 9 ) and lineage 2 ( Log 10 Bayes factor = −4 . 436 ) , in the latter case even if the logistic model had a higher Bayes factor ( 146 . 251 ) when compared to the exponential ( 150 . 686 ) , there was no significant improvement by including additional logistic parameters to describe the data . Critically , as indicated by the HPD , the growth rate ( r ) for both lineages were significantly above zero for both , lineage 1 , r = 0 . 0752 ( with 95% HPD from 6 . 96E-5 to 0 . 258 ) with an ESS of 107 . 75 for 150 , 400 , 000 states , and lineage 2 , r = 0 . 0182 ( with 95% HPD from 1 . 627E-4 to 0 . 0391 ) with an ESS of 1239 . 444 for 325 , 600 , 000 states . By transforming the inferred exponential growth rates into the basic reproductive rate , we obtained values for lineage 1 of R0 = 1 . 53 ( with 95% HPD ranging from above 1 to 2 . 8 ) and values for lineage 2 of R0 = 1 . 13 ( with 95% HPD ranging from above 1 to 1 . 3 ) . Although we rejected the logistic model , the basic reproductive rate obtained from the logistic growth rate values ( data not shown ) for lineage 1 was R0 = 3 . 765 ( with 95% HPD ranging from above 1 to 9 . 554 ) and for lineage 2 was R0 = 3 . 093 ( with 95% HPD ranging from above 1 to 8 . 896 ) . These data meant that the rate of growth was almost 50% higher for lineage 1 at exponential growth but only 17% higher under the logistic model . Under the exponential model the most recent common ancestor MRCA of lineage 1 dated 2 years before the last sampling ( with 95% HPD ranging from 6 months to 3 years ) and lineage 2 also dated 2 years before the last sampling ( with 95% HPD ranging from 6 months to 5 years ) . In sum both lineages appear to have similar growth patterns with a trend of increased rate of growth ( and higher R0 ) for lineage 1 strains . One serious limitation of the former approach was that the monotonic-increasing models used in BEAST ( logistic and exponential ) may have not captured the true fluctuating dynamics of the epidemics , since both fail to detect the decrease in numbers of new infections after the Zenith of the outbreak . Therefore we also used other methods [33] , [34] , [35] , [36] . The logistic fitting of the Bayesian skyline plot inferred by MCMC from viral genealogies for y ( t ) is shown in Figure 3A . The Basic Reproduction Number , R0 , was estimated from the average force of infection ( Figure 3B ) , calculated from equation ( 2 ) and equal to 0 . 17 new cases per time unit per susceptibles , according to the method previously proposed in a study [33] . From equation ( 4 ) the basic reproduction number ( R0 ) obtained was 2 . 45 . The data suggested that there was a good match among values obtained directly from the growth rate estimated with BEAST , the one found using the force of infection and , the epidemiological estimates of 3 . 36 previously estimated [37] from the cases' doubling time . By visually inspecting the temporally sorted distance matrices shown in Figure 4 , we noticed some genetically similar ‘blocks’ ( bright patches in the second matrix ) following the temporal gradient ( first matrix ) , but intercalated with “dark” lines representing lower genetic proximity . This intercalation of genetically distant samples seems to be responsible for preventing an overall statistical association and was due to the distinct lineages co-circulating in SJRP during the outbreak . There was also no spatial association between samples ( third matrix ) in any noticeable way when compared with the other matrices . These visual observations were confirmed by the statistical analyses , Spearman correlation between the other ones was very low ( r = 0 . 06 with geographic , and r = 0 . 01 with temporal ) and non-significant ( <0 . 05 as obtained by Mantel method with 1 , 000 interactions , [31] ) . The analyses with the “nepotistic algorithm” suggested the existence of at least three possibly independent introductions of strains from lineages 1 and 2 into SJRP . The three lineages correspond well to the tree shown in Figure 1 . The proposed spatial-temporal associations were represented in three dimensions in Figure 5 , which explained the lack of overall statistical association between the correlation matrices . The three main virus introductions appeared as genetically-similar blocks intercalated with more distant rows in the genetic distance matrix of Figure 4 . In fact , we can see that the “predominantly darker lines” coincide with the lineages names shown in red in Figures 4 and 5 . A closer inspection also revealed that , similarly to what was observed for to the “blue” samples of lineage 1 , the genetic distances between the “red samples” were low among themselves . The branching pattern of the spatio-temporal tree ( Figure 5 ) generated with the “nepotistic algorithm” also explained why there was no correlation between geographic distance matrix and the other distances . This was because the spatial dispersion of the virus starts in three different points and does not appear to follow a single centrifugal pattern . The “North-western” component ( lineage 1 ) shown in blue in the spatio-temporal trees in Figures 5 and 6 , included the earliest samples from January 12th . It was also the most prevalent ( 36 samples ) , and was the more long-lasting , encompassing also the 5 more recent samples: ‘138-173’ , 146-153’ , ‘116-160’ , ‘133-173’ , ‘135-173’ , and ‘138-173’ , collected between June 14th and July 4th . The connections between lineage 1 samples averaged 3 . 25 kilometres ( ranging from 15 meters to 7 . 23 kilometres ) . The average of speed of the propagation was around 67 . 3 meters per day , ranging from 18 centimeters/day ( since two samples ( ‘95-119’ and ‘42-34’ ) were collected only 15 meters apart ) to a maximum speed of 428 . 8 meters/day . Moreover , some lineage 1 samples seem to be the possible source nodes of many other samples ( the names of these samples and the number of generated links are: ‘01-00’: 16; ‘42-34’: 12; ‘64-67’: 2; ‘90-116’:2; ‘98-119’:2 ) . The “South-eastern” samples bundled into at least lineage 2 shown in red in the spatio-temporal tree ( Figures 5 and 7 ) , had lesser components ( 8 samples ) . Its recorded activity ranged from January 30th ( ‘22-18’ ) to June 14th ( ‘132-153’ ) . The average length of its connections was 4 . 7 kilometres ( min: 0 . 2 , max: 8 . 7 kilometres ) , at an average speed of 152 meters per day ( min: 28 meters/day; max: 311 meters/day ) . The ‘122-145’ ( Figures 5 , 6 , and 7 ) would constitute another entry of dengue into SJRP . Figure 8 represents the autochthonous dengue cases confirmed by the Surveillance Service from September 2005 to February 2006 . In September 2005 , there were dengue cases within an irregular residential area ( Santa Clara ) in the northern area and outside the urban perimeter , which lacks proper basic sanitation . During the following months , there were several additional cases in the neighborhood and the disease spread into other areas of the urban perimeter . The molecular characterization of the circulating strains identified at least three different viral introductions 01-00 , 22-18 and122-145 . The first event was at Eldorado , a neighborhood with low socioeconomic levels . Other cases occurred in areas with different socioeconomic levels ( Figures 5 , 6 , and 9 ) . There was a cluster at Gonzaga de Campos ( cases 42 , 85 , 95 and 141 ) , a working-class neighborhood with low socioeconomic levels , flanking the main industrial zone of the municipality . A series of cases linked to case 42 spread among areas with different socioeconomic levels ( Figures 5 , 6 , and 9 ) . Case 22-18 , which led to other cases in lineage 2 , with the exception of case 122-145 , that also occurred in São Deocleciano , an area with low socioeconomic level . There was a relation among the socioeconomic level of the census tracts and the incidence coefficients based on the cases reported by the Surveillance System ( Table 1 ) , despite an even distribution of the RT-PCR positive cases among different census tracts ( Table 1 and Figure 9 ) . Approximately 44% of the census tracts of the lowest socioeconomic level were in the highest quartile of incidences , but only 5 . 5% of the census tracts of the highest socioeconomic level were in the same situation .
Inferences based on sequence data of spatio-temporal dynamics , including the speed and direction of virus propagation , have been little explored until recently [38] . Nevertheless , they can help to better understand dengue outbreaks , providing useful information for public-health systems . In this study we showcased the use of both geographic and temporally structured phylogenetic data providing a relatively detailed view on the spread of at least two dengue viral lineages in the urban area of SJRP . These two groups suggested a pattern of dispersion that was consistent with the dispersal rate of Aedes aegypti , but in some instances , seemed to entail human accidental transport . It also showed that most of the cases until July traced directly to two viral sequences , sampled in January and February , respectively . The connections among our samples averaged 3 . 75 kilometers , and the speed of propagation ranged from 0 to 428 . 8 meters per day . Aedes aegypti lays its eggs at many different sites ( i . e . , skip ovoposition ) [39] , maximizing spread potential and the chance of survival . Nevertheless , our data stressed the notion that both , mosquito and human circulation , do play an important role in the dispersal of the virus and were in good agreement with previous estimates of dispersion distances ranging from 15 to 800 meters [40] , [41] , [42] . The co-circulation of distinct dengue lineages in SJRP could be explained by independent introductions experiencing different dynamic outcomes or accretion of genetic diversity generated locally . Lineages can vanish from a given locality due to unfavourable conditions such as temperature , low mosquito population and changes in immunity status of the population during the outbreak . However , Aedes aegypti eggs are resistant to desiccation and this characteristic ( i . e . , overwintering ) may have important implications for the cryptic maintenance of viral strain . Therefore , viral lineages not detected in one season can re-emerge in the next rainy season due to transovarial transmission that is believed to be the most important factor for the maintenance of the virus in the nature [43] . Therefore it is not possible at this time to determine precisely the order of introduction events and the proper time of introduction of the distinct lineages that we have detected . The region of SJRP experiences an increase in rainfall starting in December reaching a peak in January/February . Concurrently , the number of dengue cases began to increase following the infestation by Aedes aegypti . In SJRP , the two main lineages were present at the zenith of dengue transmission , which coincided with the highest values for temperature and humidity . Another lineage ( represented in green in Figures 4 , 5 , and 6 ) appeared later in May , when temperature , rainfall and humidity were decreasing . However , it is not known if this lineage became established ( we could not identify putative links of this sample with other samples ) , given that the data collection finished in July 2006 . Possibly viruses included in lineage 2 did not succeed in getting established in the city or faded away , given its appearance later in scene ( June 6th ) when the dryer and colder weather did not favour the development of the larvae and/or the survival and activity of the mosquito [44] , [45] . Alternatively , its establishment may have been hampered by the decreasing availability of susceptible hosts in the later stages of the outbreak . The matrices of distances displayed in Figure 4 show how each sample is relatively close to all other samples , considering separately their date of collection , their place of collection and their genetics . As interesting they are in their own , each one cannot suggest a hypothetical scenario for the dynamics of the dispersion of dengue in São José do Rio Preto . Only when we combine all this information ( time , space and genetics ) based on parsimonious assumptions ( each sample should be connected - by descended or siblinghood - to the closest genetic sample of the past ) we can suggest a plausible dispersion scenario and , from this , infer other very useful information - like the speed or direction of the events . DENV-3 was first isolated in São José do Rio Preto in January 2006 . According to our results , both lineages split around one to three years before the collection of the last sample . In the four-year period at Figure 2 , we noticed that during the three years after the introduction of DENV-2 , in 1998 , the incidence started to decrease every year , possibly due to the lack of susceptible hosts , but DENV-2 was still circulating in June 2005 . The introduction of DENV-3 into the naive population to this serotype in SJRP may have occurred during 2005 , because in September of this year , a neighborhood in the northern part of the city presented a significant increase in dengue incidences . The outbreak continued during October and November and culminated in April of 2006 . The notion of a probable start of DENV-3 transmission at the Santa Clara neighborhood in September 2005 and its subsequent spread to the rest of the municipality ( Figure 8 ) is in accordance with the results presented in Figures 2A and 3A . The outset of DENV-3 transmission at this underdeveloped urban area was to be expected , because Aedes aegypti larval infestation studies done in January 2005 [46] indicated that areas in SJRP with lower socioeconomic levels , with deficient sanitation infrastructure , presented higher infestation levels when compared to affluent neighborhoods inside the urban perimeter . The spread of dengue transmission through the entire municipality is in accordance with the high R0 values we found , possibly because the population was susceptible to serotype 3 . An interesting result was that the BSL ( Figure 2A ) , based on viral sequences recovered with great precision the dynamics of the epidemics obtained from case reports ( Figure 2B ) . Nevertheless , the zenith determined by the BSL took place around February , two months before the maxima determined by case report in April ( Figure 2 ) . Interestingly , these results could be explained by the fact that up until March of 2006 clinical differential diagnostics was used in conjunction with serology , which increases the accuracy of the dengue diagnostics . On the other hand , after the number of cases exceeded 300 per 100 , 000 inhabitants in April 2006 , only clinical criteria were used , which may have caused an increase in false positives , due to the lack of further serological confirmation . These results further validate the use of viral gene genealogies to infer epidemiological parameters of DENV in particular and , of fast-evolving viruses in general . The use of home addresses for geo-positioning our patients for spatial analysis was justified because individuals spend considerable time at home , which constitutes a highly probable site of transmission . Nevertheless , transmission might also occur at other places and this fact certainly may have had some impact on our data , which would be hard to account for . Nevertheless , our exercise was valid , since it indicated coherent patterns of transmission , which may be relevant for implementing control measures . Our molecular data indicated that the viral spread was not dependent entirely on vector dispersal . The exponential growth phase associated with linked transmission events beyond the usual flight range of the mosquito may have been caused by under-sampling and movement of viremic humans , but is certainly indicative of fast transmission among susceptible individuals . Therefore , surveillance systems need to be capable of monitor proactively the occurrence of initial low levels of transmission , identify early cryptic circulation of new serotypes and , be able to map where infected patients are circulating , preferably at the lag phase of the outbreak . Although we have not found clear relationships between dengue cases with molecular analysis and socioeconomic levels ( Table 1 and Figure 9A ) , the first samples associated with the beginning of the outbreak , which were possible source nodes for many other samples , were found in regions with low socioeconomic level . Moreover , a higher transmission in poor areas of SJRP has been shown , especially in the north zones of the city [47] . Therefore , it is relevant to further evaluate if the occurrence of transmission in poor areas facilitates a higher dispersion of the virus to other areas of the city . However , the association between higher dengue transmission and low socioeconomic levels is controversial . Some studies have demonstrated the association of poverty and high incidences of dengue [48] , [49] , others have not [50] , [51] , [52] and others have indicated an inverse relation [53] . A study [48] demonstrated that dengue occurred at higher levels in poorer areas of SJRP in 1995 but , from 1998 to 2002 , after the introduction of DENV-2 , the variable that best explained dengue cases was the proportion of one-story homes . The socioeconomic features lost its explanatory power as the years passed and the spatial characteristic of the areas was more relevant [54] . A higher transmission of dengue in poor areas of the north zone was observed again in 2005–2006 [47] . Two hypotheses might explain the controversial pattern of dengue transmission . In both 1995 and 2006 epidemic season , dengue transmission started in poor northern zones of the city in the previous years ( 1994 and 2005 ) with a subsequent spread to other areas [12] , [47] , [48] . Therefore , the highest initial incidence of dengue in the north and the lowest in the other areas might be related to the usual delay in adopting of control measures at the beginning of the outbreak . Another hypothesis is that dengue transmission occurred initially in poor areas and spread to the rest of the city due to the reduction of susceptible individuals in the areas that were primarily affected and , as the years passed , the distribution of the disease became similar in the whole city . This pattern was confirmed previously [55] , for the period of 1994 to 1998 , when only DENV-1 was circulating in the city , but not for the period of 1998 to 2002 , when DENV-1 and DENV-2 were circulating simultaneously . The introduction of new DENV serotypes and genotypes constitutes a major risk factor for severe dengue manifestations [56] . But it is still controversial whether DENV strains that cause severe disease out-compete less virulent strains , which is a cause of major concern [22] . Therefore it is paramount to address in greater detail whether differences in viral dispersion patterns are associated with viral fitness , strain competition and , ultimately , whether it has any association with increase in disease severity . We have shown that spatial analysis using Geographic Information System could provide valuable information on dengue transmission and the spread of the disease in a defined but heterogeneous urban setting , typical of the developing world . We believe that the current study helped determining with greater precision areas where the infection took place , to understand particularities of an outbreak , clarifying the mechanisms of dengue transmission in SJRP . Ultimately , the association of molecular epidemiology with spatial analysis and the understanding of some biological and reproductive characteristics of Aedes aegypti mosquitoes may shed light on the dynamics and distribution of different dengue viral strains . | Most of the molecular phylogeny studies of dengue fever , an important public health problem , use convenience samples for their analysis , and they do not evaluate the spatial and temporal features involved in the spread of the different serotypes ( and genotypes ) circulating in urban settings during an outbreak . Our study describes the patterns of spread of different lineages of dengue 3 virus circulating in a medium-sized city from Brazil , and we also analyzed the dynamics and microevolution of the disease during the 2006 outbreak . We used both geographic and temporally structured phylogenetic data , which provided a relatively detailed view on the spread of at least two dengue viral lineages circulating in an urban area . The pattern of dengue virus circulation might be similar to many other settings all over the world , and the information provided by our study can help a better understanding of dengue outbreaks , providing important information for public-health systems . We could identify at least two lineages , which were introduced in different occasions . They circulated and spread at different rates within the city , and this differential spread and the role of socioeconomic features in this phenomenon are discussed . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"virology/virus",
"evolution",
"and",
"symbiosis",
"virology",
"virology/emerging",
"viral",
"diseases",
"genetics",
"and",
"genomics/population",
"genetics"
] | 2009 | Spatio-Temporal Tracking and Phylodynamics of an Urban Dengue 3 Outbreak in São Paulo, Brazil |
Vascular calcification is an advanced feature of atherosclerosis for which no effective therapy is available . To investigate the modulation or reversal of calcification , we identified calcifying progenitor cells and investigated their calcifying/decalcifying potentials . Cells from the aortas of mice were sorted into four groups using Sca-1 and PDGFRα markers . Sca-1+ ( Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− ) progenitor cells exhibited greater osteoblastic differentiation potentials than Sca-1− ( Sca-1−/PDGFRα+ and Sca-1−/PDGFRα− ) progenitor cells . Among Sca-1+ progenitor populations , Sca-1+/PDGFRα− cells possessed bidirectional differentiation potentials towards both osteoblastic and osteoclastic lineages , whereas Sca-1+/PDGFRα+ cells differentiated into an osteoblastic lineage unidirectionally . When treated with a peroxisome proliferator activated receptor γ ( PPARγ ) agonist , Sca-1+/PDGFRα− cells preferentially differentiated into osteoclast-like cells . Sca-1+ progenitor cells in the artery originated from the bone marrow ( BM ) and could be clonally expanded . Vessel-resident BM-derived Sca-1+ calcifying progenitor cells displayed nonhematopoietic , mesenchymal characteristics . To evaluate the modulation of in vivo calcification , we established models of ectopic and atherosclerotic calcification . Computed tomography indicated that Sca-1+ progenitor cells increased the volume and calcium scores of ectopic calcification . However , Sca-1+/PDGFRα− cells treated with a PPARγ agonist decreased bone formation 2-fold compared with untreated cells . Systemic infusion of Sca-1+/PDGFRα− cells into Apoe−/− mice increased the severity of calcified atherosclerotic plaques . However , Sca-1+/PDGFRα− cells in which PPARγ was activated displayed markedly decreased plaque severity . Immunofluorescent staining indicated that Sca-1+/PDGFRα− cells mainly expressed osteocalcin; however , activation of PPARγ triggered receptor activator for nuclear factor-κB ( RANK ) expression , indicating their bidirectional fate in vivo . These findings suggest that a subtype of BM-derived and vessel-resident progenitor cells offer a therapeutic target for the prevention of vascular calcification and that PPARγ activation may be an option to reverse calcification .
Vascular calcification ( VC ) is a feature of progressive and advanced atherosclerosis that is regarded as a prognostic marker of adverse cardiovascular events [1] , [2] . No therapies are available to ameliorate VC [3] . The pathophysiology of VC involves a strict and active regulatory process that resembles bone formation [4] and functions to maintain a balance between osteoblastic and osteoclastic cells [5] . The origin of osteoblastic cells in the vasculature remains an issue of active debate [6] . Resident vascular smooth muscle cells ( VSMCs ) and calcifying vascular cells have been examined to elucidate the cellular origins of VC . Pericytes , mesenchymal stem cells ( MSCs ) , myofibroblasts , and circulating osteoprogenitor cells have been isolated from the vasculature and shown to have osteoblastic potential [7]–[10] . However , few studies have addressed the origins , features , and roles of osteoclastic and decalcifying cells in the vasculature or the balance between osteoblastic and osteoclastic cells during VC . In this study , we aimed to identify vascular calcifying progenitor cells and to modulate or reverse VC . We first isolated vessel-resident calcifying progenitor cells using stem cell antigen-1 ( Sca-1 ) and platelet-derived growth factor receptor alpha ( PDGFRα ) antibodies in the vasculature . We then identified a population of nonhematopoietic mesenchymal Sca-1+ cells ( Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− cells ) that originated from the bone marrow ( BM ) and could be clonally expanded . Among the Sca-1+ populations , Sca-1+/PDGFRα+ cells possessed unidirectional osteoblastic potential . In contrast , Sca-1+/PDGFRα− cells possessed bidirectional osteoblastic and osteoclastic differentiation potentials . Both calcifying progenitor Sca-1+/PDGFRα+ cells and Sca-1+/PDGFRα− cells induced ectopic mineralization and atherosclerotic calcification in vivo . When PPARγ was activated in bidirectional Sca-1+/PDGFRα− cells , calcium accumulation was reduced , and plaque severity was decreased . This cell population may offer new therapeutic targets and modalities for ameliorating VC .
To identify putative calcifying progenitor cells , we stained tissue sections with stem/progenitor markers [11] . We detected marker-positive cells , particularly Sca-1+ cells , in the artery ( Figure S1 ) . Sca-1 is a marker of hematopoietic stem cells [12] and MSCs [13] in mice . To distinguish among progenitor cells in the vasculature , we also stained for PDGFRα [14] . Both Sca-1+ and PDGFRα+ cells were detected in the artery ( Figure 1A ) . Our double-immunostaining enabled us to categorize cells into the following four groups: Sca-1+/PDGFRα+ , Sca-1+/PDGFRα− , Sca-1−/PDGFRα+ , and Sca-1−/PDGFR− . We subsequently isolated and propagated aortic cells ( Figure S2 ) and performed fluorescence-activated cell sorting ( FACS ) . No difference in sorting was detected between cells sorted immediately ( Figure S3 ) or after 2 wk of cell expansion . We confirmed the purities of the sorted cell populations by immunostaining ( Figure S2C ) . We next assessed the osteoblastic differentiation potentials of the four cell groups over time ( Figure 1B ) . Under three osteoblastic differentiation conditions [15] , Sca-1+ cells ( Sca-1+/PDGFRα+ , Sca-1+/PDGFRα− ) showed significantly higher numbers and activities of alkaline phosphatase ( ALP ) –positive cells than Sca-1− cells ( Sca-1−/PDGFRα+ , Sca-1−/PDGFR− ) . We also confirmed the level of mRNA expression of several osteoblast-related genes in the four groups of cells ( Figure 1C and Figure S4A , B ) . Our results indicate that Sca-1+ cells are superior to Sca-1− cells in terms of osteoblastic differentiation potentials and osteoblastic fate adaptation capacities . We also determined the osteoclastic potentials of the four cell groups . The cells were cultured in osteoclast differentiation media ( Figure 1D ) . Osteoclastic differentiation potentials were then assessed by tartrate-resistant acidic phosphatase ( TRAP ) staining [16] , [17] . The Sca-1+/PDGFRα− cell population was comprised of significantly more TRAP-positive cells than the other three cell groups ( Figure 1E , F ) . Under moderate osteoblastic differentiation conditions , Sca-1+/PDGFRα− cells highly expressed the osteoclast-related gene , receptor activator for nuclear factor-κB ( RANK ) , and the osteoblast-related genes , osteoprotegerin ( OPG ) , and RANK ligand ( RANKL ) . However , in serum containing tumor necrosis factor-α ( TNF-α ) , RANK expression was suppressed ( Figure 1C ) . We detected the expression of other osteoclast-related genes in Sca-1+/PDGFRα− cells , but not in Sca-1+/PDGFRα+ cells ( Figure 1G and ) . To examine whether differentiated Sca-1+/PDGFRα− cells function as osteoclast-like cells , we measured calcium resorption ability . Cells were cultured on a calcium phosphate-coated disc treated with RANKL and macrophage-colony stimulating factor ( M-CSF ) . Observation of the discs by scanning electronic microscopy ( SEM ) indicated that Sca-1+/PDGFRα− cells generated wider areas of calcium resorption and greater pore sizes than Sca-1+/PDGFRα+ cells ( Figure 1H ) . We assayed for the formation of dual actin ring sealing zones [18] . Under osteoclastic differentiation conditions , Sca-1+/PDGFRα− cells cultured on dentine slices formed actin ring of the sealing zones characteristic of active osteoclasts ( Figure 1I ) . These structures were absent in Sca-1+/PDGFRα+ cells cultured under the same conditions . These results confirm that Sca-1+/PDGFRα− cells have the potential to differentiate into functioning osteoclast-like cells . Sca-1+/PDGFRα− cells possess bidirectional differentiation potentials toward both the osteoblastic and osteoclastic lineages , whereas Sca-1+/PDGFRα+ cells have a unidirectional differentiation potential toward the osteoblastic lineage . We tested the fates of Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− cells in mixed medium containing FBS and TNF-α ( osteoblastic differentiation stimulators ) with RANKL and M-CSF ( osteoclastic differentiation stimulators ) . We examined differentiation potentials by immunofluorescent staining of osteocalcin and cathepsin K . Seven days after incubation in mixed medium , both Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− cells dominantly expressed osteocalcin . Interestingly , a few Sca-1+/PDGFRα− cells expressed cathepsin K but did not form multinucleated cells . Sca-1+/PDGFRα+ cells did not express cathepsin K . These data suggest that the fates of bidirectional Sca-1+/PDGFRα− cells could be primarily osteoblastic in the mixed in vivo environment ( Figure S5 ) . Because Sca-1+/PDGFRα+ cells do not possess bidirectional differentiation capacities , we examined whether PDGFRα directly influences the differentiation phenotypes of these cells . We treated Sca-1+/PDGFRα+ cells with PDGF-BB ( 10 ng/ml or 25 ng/ml ) and 10% charcoal-stripped FBS for 0 , 3 , 7 , or 14 d . TNF-α treatment was used as a positive stimulation control in flow cytometry . We analyzed osteoblast-related gene expression by real-time PCR and ALP expression by flow cytometry . The mRNA expression levels of osteoblast-related genes were mildly increased in the PDGF-BB treatment group compared with the FBS only treatment group . Similarly , FACS indicated that PDGF-BB treatment marginally increased ALP expression compared with FBS only . suggesting that PDGFRα is not a functional receptor for osteoblastic differentiation . In contrast , TNF-α treatment markedly increased ALP expression . In PDGFRα+ cells , PDGF-BB mildly stimulated osteoblast differentiation and promoted cells to migrate into artificial bone structures [19] , [20] . These researchers as well as we concluded that PDGFRα was a marker but not a functional receptor for osteoblastogenesis ( Figure S6 ) . To test the capacity of calcifying progenitor cells from single clones , we performed a clonal expansion assay ( Figure 2A ) . Sca-1+ cells generated compact and abundant colonies and exhibited a greater colony-forming efficiency than Sca-1− cells , which barely generated any colonies ( Figure 2B , C ) . We also investigated whether calcifying progenitor cells derived from a single colony could differentiate into either osteoblastic or osteoclastic cells ( Figure 2D ) . Single colony-derived Sca-1+ cells differentiated into osteoblasts ( Figure 2E ) . Sca-1+/PDGFRα− clonally expanded cells differentiated into multinucleated osteoclast-like cells under osteoclastic differentiation conditions ( Figure 2F ) , suggesting that this cell population was comprised of bidirectional calcifying progenitor cells with both osteoblastic and osteoclastic differentiation potentials . Osteoblasts and osteoblastic progenitor cells have been well-characterized in the bone and BM [21] , [22] . To determine whether calcifying progenitor cells originate in BM , we performed a BM transplantation ( BMT ) experiment using GFP mice as a marker ( Figure 3A ) . Five days after cell infusion , to rule out the possibility that cells migrated from the intravascular space to vessels , we examined the presence of GFP+ cells in the arteries and blood using FACS and immunofluorescent staining . A few GFP+ cells from the donor were detected in the peripheral blood ( 1 . 5% ) , but these were rarely detected in the artery ( 0 . 2% ) , indicationg that GFP+ cells from donor marrow did not exhibit diapedesis directly into the arterial wall . Twelve weeks after transplantiation ( Figure 3B ) , GFP+ cells from donor BM reconstituted blood cells in C57 background mice comprised up to 90% of peripheral blood cells . At that point , 13% of arterial resident cells were GFP+ . Takgen together , these data indicate that the majority of GFP+ cells were gradually incorporated into the artery in a considerable amount of time . We then determined vessel infiltration of BM-derived GFP+Sca-1+ cells by immunostaining in the artery ( Figure 3C ) . We also assessed the possibility of the fusion between BM-derived GFP+ cells and non-BM cells using propidium iodide ( PI ) staining . GFP+ cells with DNA contents beyond 4n were not detected ( Figure 3D , E ) . We performed a GFP+ clonal expansion assay in BM-derived GFP+Sca-1+/PDGFRα+ and GFP+Sca-1+/PDGFRα− cell populations from the vessel walls of chimeric mice . GFP+Sca-1+ cells ( GFP+Sca-1+/PDGFRα+ or GFP+Sca-1+/PDGFRα− cells ) were capable of colony generation ( Figure 3F , G ) . Single colony derived GFP+Sca-1+/PDGFRα− cells expanded from a single colony possessed both osteoblastic and osteoclastic differentiation potentials ( Figure 3H ) . We subsequently characterized BM-derived and vessel-resident calcifying progenitor cells . Aortas were harvested from Apoe−/− mice that underwent GFP–BMT . BM-derived GFP+ cells were negative for a hematopoietic lineage antibody cocktail ( Lin− ) containing antibodies targeting CD3 , CD11b ( monophage/macrophage marker ) , CD45R/B220 , TER-11 , and Ly-6G . Calcifying progenitor cells then were isolated from GFP+Lin− cells by detecting Sca-1 and PDGFRα expression . GFP+Lin−Sca-1+/PDGFRα+ cells highly expressed the MSC markers , CD29 and CD106 , and GFP+Lin−Sca-1+/PDGFRα− cells weakly expressed CD29 and CD106 . These results suggest that BM-derived calcifying progenitor cells have characteristics of MSCs but not of hematopoietic cells or of contaminating monocytes/macrophages ( Figure 4A ) . We next isolated vessel-resident calcifying progenitor cells and assessed their differentiation potentials as MSCs . Sca-1+ cells were capable of differentiating into osteoblasts , adipocytes , and chondrocytes ( Figure 4B ) . We measured the expression levels of the following adipocyte- and chondrocyte-specific genes by real-time PCR . Under each differentiation condition , Sca-1+ cells ( Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− ) showed significantly higher adipocyte- and chondrocyte-specific gene expression levels , as compared with Sca-1− cells ( Sca-1−/PDGFRα+ and Sca-1−/PDGFR− ) . These results indicate that Sca-1+ cells have a nonhematopoietic , MSC-like nature ( Figure 4C ) . To confirm the ex vivo osteoblastic and osteoclastic differentiation abilities of BM-derived calcifying progenitor cells , we harvested and cultured cells from the aortas of C57 mice that had undergone GFP–BMT ( Figure S7A ) . Cultured aortic cells were divided into four groups of GFP+ cells with respect to their Sca-1/PDGFRα statuses and were used in osteoblastic and osteoclastic differentiation experiments . Under osteoblastic differentiation conditions , GFP+Sca-1+ cells differentiated more readily into osteoblasts than GFP+Sca-1− cells . Under osteoclastic differentiation conditions , GFP+Sca-1+/PDGFRα− cells exclusively differentiated into osteoclasts . These findings suggest that BM-derived vessel-resident Sca-1+/PDGFRα− cells possess osteoblastic/osteoclastic differentiation potentials . In contrast , GFP+Sca-1+/PDGFRα+ cells displayed only osteoblastic differentiation ( Figure S7 ) . PPARγ activation has been suggested to repress osteoblastogenesis and to activate osteoclastogenesis [16] , [23] . These observations led us to hypothesize that PPARγ activation in calcifying progenitor cells might reverse the process of VC . We first confirmed endogenous PPARγ expression in Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− calcifying progenitor cells ( Figure S8A ) . BM-derived vessel-resident GFP+ cells then were cultured in FBS and TNF-α to strongly induce osteoblast differentiation and the influence of an added PPARγ agonist was assessed ( Figure 5A ) . In the absence of PPARγ activation , osteoblastic differentiation was markedly induced ( Figure 5B ) . Under the same conditions , PPARγ activation of Sca-1+/PDGFRα− cells suppressed osteoblastogenesis ( Figure 5B ) and promoted osteoclastic differentiation ( Figure 5C ) . PPARγ activation suppressed the expression of osteoblast-related genes and enhanced the expression of osteoclast-related genes , facilitating the emergence of TRAP-positive cells ( Figure 5B , C ) as confirmed by real-time PCR ( Figure S8B , C ) . These results indicated that PPARγ activation not only suppressed the osteoblastic differentiation of vascular cells , but also promoted the osteoclastic differentiation of bidirectional calcifying progenitor cells . Hence , the detrimental process of VC may be preventable as well as reversible . To verify the in vivo calcifying ability of progenitor cells and the efficacy of PPARγ activation , we isolated GFP+ calcifying progenitor cells from the arteries of C57 mice that underwent GFP–BMT . GFP+ progenitor cells were combined with bone matrix , implanted subcutaneously into WT C57 mice , and a PPARγ agonist was injected ( Figure 6A ) . After 8 wk of PPARγ agonist treatment , X-ray and three-dimensional computed tomography ( CT ) of the implanted mice indicated a higher mineralization density of a bone-like structure in the Sca-1+ cell groups compared with mice administered phosphate-buffered saline ( PBS ) . This mineralization of mass was remarkably inhibited in Sca-1+ cells treated with a PPARγ agonist compared with cells without PPARγ activation . Specifically , the bone-like structure of Sca-1+/PDGFRα− cells was dramatically reduced by PPARγ activation ( Figure 6B , C ) . The volume and calcium scores of Sca-1+ cells were higher than those of either PBS-treated or Sca-1− cells quantified using the Agatston method ( Figure 6D ) [24] . These scores were significantly increased in Sca-1+/PDGFRα− cells treated with a PPARγ agonist . However , no difference was detected between mice injected with PBS or PBS and a PPARγ agonist ( Figure 6E ) . Masson's trichrome ( MT ) staining indicated that Sca-1+ cell injection enhanced blue staining , indicating bone-like tissues , whereas PPARγ treatment decreased blue staining ( Figure 6F ) . Therefore , we sought to determine the effect of PPARγ activation in Sca-1+ cells by double-staining with osteocalcin and cathepsin K . Injected GFP+Sca-1+/PDGFRα+ cells were osteocalcin-positive and cathepsin K-negative . Injected GFP+Sca-1+/PDGFRα− cells were primarily osteocalcin-positive , but some GFP+Sca-1+/PDGFRα− cells were doubly positive for osteocalcin and cathepsin K . Exclusively cathepsin K-positive cells were observed only in the mass of GFP+Sca-1+/PDGFRα− cells treated with a PPARγ agonist ( Figure 6G–I ) . These results demonstrate that BM-derived Sca-1+/PDGFRα− cells are bidirectional , and PPARγ may act as a regulator of VC in these cells . We assessed the accumulation of aortic calcium in response to diet . Mice fed a high cholesterol/calcium diet showed a significantly higher level of calcium accumulation than mice fed a normal diet ( Figure 7A ) . To understand the function of calcifying progenitor cells in calcified atherosclerotic plaques , we isolated BM cells from a GFP mouse . BM-GFP+Sca-1+/PDGFRα+ and GFP+Sca-1+/PDGFRα− cells were injected with or without a PPARγ agonist into the tail veins of Apoe−/− mice every 2 wk over a period of 8 wk ( Figure 7B ) . The cell properties of injected BM cells , BM-GFP+Sca-1+/PDGFRα+ cells , and GFP+Sca-1+/PDGFRα− cells were compared between normal ( Figure S7 ) and high-cholesterol ( Figure S9 ) diets , and no differences were identified . The calcium accumulation levels of each cell group were measured from harvested arteries . The groups injected with BM-GFP+Sca-1+ cells , especially BM-GFP+Sca-1+/PDGFRα− cells , showed greater calcium accumulation in tissues than animals injected with PBS . This calcium accumulation was significantly avoided by concurrent treatment with a PPARγ agonist . However , no difference in calcium accumulation was measured between mice injected with PBS versus PBS and a PPARγ agonist . Therefore , the preventative effects of the GFP+Sca-1+/PDGFRα− + PPARγ condition should be mainly attributed to the BM-GFP+Sca-1+/PDGFRα− cells , not the PPARγ agonist ( Figure 7C ) . Mice injected with GFP+Sca-1+/PDGFRα− cells showed significantly more severe atherosclerotic calcified plaques identified by MT and von Kossa staining . These plaques were prevented by the addition of GFP+Sca-1+/PDGFRα− cells and a PPARγ agonist ( Figure 7D , E ) . Additional studies were conducted to determine the characteristics of the injected GFP+Sca-1+ cells that infiltrated the calcified atherosclerotic plaques . GFP+ cells were double-stained with osteocalcin and RANK . In the calcified atherosclerotic plaques of PBS injected mice , osteocalcin-positive osteoblasts and a few RANK+ osteoclasts were observed . Mice injected with GFP+Sca-1+/PDGFRα+ cells harbored plaques infiltrated primarily by osteocalcin-positive cells . Mice injected with GFP+Sca-1+/PDGFRα− cells contained plaques that were infiltrated with both osteocalcin and RANK double-positive cells as well as osteocalcin-positive osteoblasts . These data suggest that the Sca-1+/PDGFRα− cells we observed in calcified atherosclerotic plaques possessed both osteoblastic and osteoclastic differentiation potentials . PPARγ agonist treatment significantly decreased the infiltration of osteoblasts and increased the infiltration of osteoclast or double-positive cells into the plaques ( Figure 7F–H ) . We conclude that PPARγ activation can decrease atherosclerotic calcification by modulating the fate of bidirectional Sca-1+/PDGFRα− cells .
VC occurs by an actively regulated process , but the origin of vascular calcifying cells has not been established . In this study , we demonstrated the origin , in vitro and ex vivo characteristics , and differentiation potentials of a population of vascular calcifying progenitor cells . These cells were confirmed to modulate calcification/decalcification through a series of in vivo experiments . A schematic illustration of calcifying/decalcifying progenitor cells and their proposed actions is presented in Figure 8 . Recent studies have shown that circulating BM-derived cells could infiltrate via the vasa vasorum ( micro vessel ) [25] . However , it remained unclear how cells could infiltrate the intima . Others have suggested that the vasa vasorum is increased in the adventitia during inflammation . The vasa vasorum may function as a hallway for cell migration into the adventitia toward atherosclerotic plaques [26] , [27] . However , circulating BM-derived cells could directly infiltrate the atherosclerotic plaques . Shimizu et al . demonstrated that cells infiltrating the intima originated as circulating BM-derived cells [28] , [29] . Cho et al . reported that cells infiltrating the adventitia did not migrate into the intima and only circulating cells infiltrated the plaque lesion [30] . Our results suggest that under homeostatic noncalcified conditions , circulating BM-derived calcifying cells ( Sca-1+ cells ) infiltrated the adventitia through the vasa vasorum . However , under pathologic atherosclerotic calcification or therapeutic intravascular cell delivery , these cells might infiltrate into the intima directly from the blood . Osteoblasts originate from MSCs [6] , whereas osteoclasts are cell–cell fused multinucleated cells derived from granulocyte/macrophage hematopoietic progenitor cells in the bone [31] . Recently , Eghbali-Fatourechi et al . reported that circulating osteoblastic cells are elevated in adolescents and in patients with fractures , reflecting increased bone formation [32] . However , the roles of circulating osteoblastic cells during VC have not been elucidated . Osteoclast-like cells in the vasculature may originate from circulating hematopoietic cells that were recruited to blood vessels [33] , but this possibility has not been confirmed . Here , we demonstrate that a subtype of calcifying progenitor cells , Sca-1+/PDGFRα− cells , originate in BM and infiltrate arteries . Sca-1+/PDGFRα− cells are mesenchymal and possess bidirectional osteoblastic/osteoclastic potentials . The bidirectionality of Sca-1+/PDGFRα− cells was supported by Masuda et al . ( 2001 ) , who identified ALP and TRAP double-positive cells in tissues during endochondral ossification [34] . Others have described the mesenchymal characteristics of CD45−TER119−Sca-1+/PDGFRα+ cells and Sca-1+/PDGFRα− cells , and injected Sca-1+/PDGFRα+ cells primarily differentiate into osteoblasts in vivo [35] . We examined the in vivo modulation of calcifying progenitor cell differentiation by PPARγ activation using an ectopic model and an atherosclerotic calcification model . When tissue was double-stained with osteoblast ( osteocalcin ) and osteoclast ( cathepsin K or RANK ) markers , Sca-1+/PDGFRα− cells mainly differentiated into osteocalcin-positive cells . However , when cells were cultured with a PPARγ agonist , their differentiation shifted from osteoblastic to osteoclastic or double-positive cells in tissue [16] , [36] . These results suggest that PPARγ activation shifts Sca-1+/PDGFRα− cells from osteoblastic to osteoclastic differentiation in vivo [37] , [38] . The concentration of the PPARγ agonist , rosiglitazone ( 5 mg/kg/d ) , used in the mice study was higher than the human AUC ( area under curve ) of the FDA-recommended clinical daily dose . Thus , additional study is warranted to find an efficacious novel PPARγ agonist . We also investigated the mechanism by which PPARγ facilitates osteoblastic differentiation . PPARγ activation suppresses osteoblastogenesis by favoring adipogenesis and improving osteoclastogenesis [23] , [39] . Wei et al . reported that the transcriptional co-activators of PPARγ , PGC1β and ERRα , enable cell differentiation into osteoclasts and adipocytes [40] . These researchers demonstrated that PPARγ activation upregulated osteoclast differentiation by inducing GATA2 , which is required to generate osteoclast progenitors , and by activating PPARγ ligand via c-fos induction , thereby stimulating osteoclast differentiation [40] , [41] . Sca-1+/PDGFRα− cells cultured in TNF-α with PPARγ activation suppressed osteoblastogenesis and enhanced the expression levels of osteoclast-related genes , especially c-fos expression , compared with cells in the absence of PPARγ stimulation . Sca-1+/PDGFRα− cells displayed both osteoblastic and osteoclastic differentiation potentials in this study . However , the hierarchy between Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− cells remains unclear . Sca-1+/PDGFRα− likely are hierarchically above Sca-1+/PDGFRα+ cells because the former are capable of differentiating into osteoblasts or osteoclast-like cells , whereas the latter differentiate into osteoblasts only . Sca-1+/PDGFRα− cells have multi-potentiality with relatively weak MSC marker expression and scanty hematopoietic marker expression . These cells can differentiate into osteoclast-like cells that usually are derived from hematopoietic stem cells , implying that they might be positioned hierarchically between hematopoietic stem cells and MSCs . When they are committed into a more specific lineage , Sca-1+/PDGFRα− cells probably lose their bidirectional potential , becoming unidirectional Sca-1+/PDGFRα+ cells . Notably , calcifying progenitor cells in the artery and BM are not abundant . The present study has several limitations . We have characterized calcifying progenitor cells in C57 wild-type mice , but not in C57 background Apoe−/− mice . Thus , those cells may be different or behave differently when added to atherosclerotic environment . The numbers of Sca-1+/PDGFRα− cells in the BM were less than 1% of the total BM cells . However , the infiltration of Sca-1+/PDGFRα− cells into the vessels was associated with marked aggravation of atherosclerotic calcification . Based on the percentage of cells measured by flow cytometry [42] , BM-derived Sca-1+/PDGFRα− osteocalcin-positive cells comprised 2% of the total cells infiltrating the artery , and Sca-1+/PDGFRα− osteocalcin+ RANK+ ( or cathepsin K+ ) cells comprised 1% . Under PPARγ-activated conditions , the number of Sca-1+/PDGFRα− osteocalcin+ RANK+ ( or cathepsin K+ ) cells increased by 2% , and Sca-1+/PDGFRα− RANK+ cells increased by 0 . 5% . These cells are expected to substantially impact VC homeostasis even if their absolute numbers are low . It is obvious that VSMCs play an important role in calcification of the vasculature [43] , [44] . An analysis of calcification induction in these cells ( Sca-1−/PDGFRα+ ) is beyond the scope of the present study . Our focus is on the characteristics and significance of Sca-1+/PDGFRα+ cells and Sca-1+/PDGFRα− cells . In conclusion , our data demonstrate that BM-derived , MSC-like Sca-1+/PDGFRα− cells reside in the arterial adventitia and possess differentiation potentials toward both osteoblastic and osteoclastic lineages . Even under calcifying conditions , PPARγ activation promoted osteoclastic differentiation of bidirectional cells , both in vitro and ex vivo . Finally , we confirmed the in vivo relevance of bidirectional progenitor cells that are modulated by PPARγ activation . This subtype of BM-derived circulating and vessel-resident calcifying progenitor cells offers a new therapeutic target for VC . PPARγ activation in these cells has the potential for VC management .
Wild-type C57BL/6J mice ( KBT Oriental Co Ltd . , Charles River Grade , Tosu , Saga , Japan ) and ubiquitous enhanced GFP-expressing transgenic mice with a C57 background ( The Jackson Laboratory , Bar Harbor , Maine , USA ) were used in this study . Apoe−/− ( B6 . 129P2-Apoetm1Unc/J; The Jackson Laboratory ) mice were used as a model of atherosclerotic calcification . All mice used in this study were males [45] . All procedures were performed in accordance with the Institutional Animal Care and Use Committee of Seoul National University Hospital . Primary cells cultured from aorta , including the media and adventitia , were prepared by enzyme digestion using collagenase type II ( GIBCO ) as previously described [46] . Cells were cultured in Mesencult media ( Stem Cell Technologies ) . To examine calcifying progenitor cells , cells were cultured for 2 wk and harvested . They then were stained with the surface markers , Sca-1 ( BD Pharmingen ) and PDGFRα ( Cell Signaling ) . Sorting was performed using a FACSAria ( Becton Dickinson ) . Sorted cells were divided into the following four groups: Sca-1+/PDGFRα+ , Sca-1+/PDGFRα− , Sca-1−/PDGFRα+ , and Sca-1−/PDGFRα− . To confirm the characteristics of calcifying progenitor cells , cells were stained with using a hematopoietic lineage antibody cocktail ( Lin− ) containing antibodies targeting CD3 , CD11b ( monophage/macrophage marker ) , CD45R/B220 , TER-11 , and Ly-6G ( BD Pharmingen ) or with antibodies targeting CD29 or CD106 ( R&D Systems ) . Sorted cells were cultured for 14 d under three osteoblastic differentiation culture conditions in MEM alpha media ( GIBCO ) : 10% charcoal-stripped FBS only , 10% charcoal-stripped FBS with 10 ng/ml TNF-α [15] , and 10% charcoal-stripped FBS containing 6 mM CaCl2 , 10 mM sodium pyruvate , and 10 mM β-glycerophosphate [47] . Cells were harvested on days 1 , 3 , 7 , and 14 for reverse transcriptase ( RT ) –PCR or real-time PCR , ALP staining , and ALP activity . To measure ALP activities , cells were grown under osteoblastic differentiation conditions . Harvested cells were washed twice with PBS , lysed in 0 . 05% Triton-X100 in PBS , and subjected to three freeze/thaw cycles . Cell lysate supernatants were transferred to 96-well plates and were incubated with 50 µl alkaline buffer ( Sigma ) for 10 min and 50 µl phosphatase substrate capsules ( Sigma ) until yellow color was observed . P-nitrophenol standard solution ( Sigma ) was used to generate a standard curve . Absorbance at 405 nm was measured using a plate reader ( Bio-Rad ) . ALP activities were normalized to the protein concentrations of the samples . ALP staining was performed using the BCIP/NBT substrate system ( Dako ) . To evaluate gene expression under osteoblastic differentiation conditions , RT–PCR was performed as previously described [48] . Complementary DNA was PCR-amplified using the osteoblastic differentiation markers CBFA-1 , OPG , and RANKL ( Tables S1 and S2 ) [49] . Sorted cells were cultured in MEM alpha medium with 10% charcoal-stripped FBS ( GIBCO ) , 10 ng/ml M-CSF ( R&D Systems ) , and 100 ng/ml RANKL ( R&D Systems ) for 7 d [17] . Cells were harvested weekly for RT–PCR . After 7 d of osteoclast differentiation induction , TRAP staining ( Sigma ) was performed as a measure of osteoclastic activities . TRAP-positive multinucleated cells ( >3 nuclei ) were counted . RT–PCR or real-time PCR was performed to amplify the following osteoclastic differentiation markers: nuclear factor of activated T-cells-1 ( NFATc1 ) , phospholipase C , gamma-1 ( PLCγ1 ) , TNF receptor-associated factor-6 ( TRAF6 ) , RANK , RANKL , c-Fos [50] , and GAPDH ( Table S1 and S2 ) . Sca-1+/PDGFRα+ and Sca-1+/PDGFRα− cells were seeded onto calcium phosphate-coated discs ( BD bioscience ) in 12-well culture plates and were cultured in osteoclastic differentiation media . After 21 d , cells were removed from the discs using a bleach solution . The discs were washed three times with distilled water , air dried , and examined by SEM [51] . Alternatively , after 21 d , cells were stained with FITC-conjugated phalloidin ( Sigma ) and observed by confocal microscopy [52] . To detect MSC-like properties of vessel-resident calcifying progenitor cells , the cells were induced to differentiate into osteoblasts , adipocytes , or chondrocytes by varying the culture conditions . The osteoblast differentiation media consisted of 10% charcoal-stripped FBS with 6 mM CaCl2 , 10 mM sodium pyruvate , and 10 mM β-glycerophosphate [47] . After 14 d , osteoblastic cells were stained with Alizarin Red S ( Sigma ) . After 28 d , differentiated cells , cultured in adipocyte differentiation media ( Stem Cell Technologies ) or chondrocyte differentiation media ( R&D Systems ) , were stained with Oil Red O ( Sigma ) or Safranin O solution ( Sigma ) , respectively . To detect single clonal expansion , sorted arterial cells from C57 mice were seeded into 6-well plates ( 1×104 cells/well ) or into 96-well plates ( 1 cell/well ) . After 14 d of culture , colonies were Giemsa-stained ( Sigma ) . To examine whether calcifying progenitor cells were derived from single cells that could differentiate into osteoblasts or osteoclasts , cells were seeded into 12-well plates ( 200 cells/well ) and were cultured for 7 d . Cells then were induced to differentiate into osteoblasts or osteoclasts during the following 14 d . Differentiated cells were stained with Alizarin Red S or TRAP and measured the mRNA expression of the following Adipocyte and chondrocyte differentiation markers: C/EBP ( CCAAT-enhancer-binding proteins ) α and β , PPARγ [53] , collagen 1a1 , 2a1 , aggrecan [54] , and GAPDH ( Table S3 ) . To measure calcium levels in aortas of mice fed a normal diet , a high cholesterol diet , or a high cholesterol/calcium diet ( 7% fat , 3% cholesterol , 200 , 000 IE/kg vitamin D , 3 , 000 mg/kg calcium , 1 . 700 mg/kg phosphate ) . Isolated aorta tissues were dried at 70°C overnight and weighed . Dried tissues were dissolved in 2 M HCl at 70°C for 24 h . Pellets were obtained and HCl was removed using a vacuum dryer . Dried pellets then were added to 1 ml distilled water . Calcium and phosphorous levels were measured relative to total protein using an autoanalyzer ( Hitachi 7070 , Tokyo , Japan ) . Calcifying progenitor cells were collected and sorted from the arteries of GFP–BMT mice that had been fed a high cholesterol diet . Cells were mixed with triosite ( Zimmer ) and were incubated for 1 h . Transplanted cells then were mixed with fibrinogen and thrombin before being implanted into a subcutaneous pocket of an 8-wk-old C57 mouse . Eight weeks later , ectopic mineralization was assessed by X-ray ( TU-3000DR , Hitachi ) and CT ( Somatom Definition; Siemens Medical Solutions , Forchheim , Germany ) and was quantified using the Agatston score [32] , [55] , [56] . Ectopically mineralized tissues were harvested and embedded in paraffin . BM-derived cells were isolated and sorted from GFP mice . BM-GFP+Sca-1+/PDGFRα+ and GFP+Sca-1+/PDGFRα− cells ( 1×106/100 µl ) were injected into the tail veins of Apoe−/− mice 4 times at 2-wk intervals . Apoe−/− mice were fed a high cholesterol diet for 10 wk prior to cell injection . Mice continued the high cholesterol/calcium diet for another 8 wk following injection . To test the effect of PPARγ activation in vivo , the PPARγ agonist ( rosiglitazone , 10 mg/kg ) was injected intraperitoneally into mice 3 times per week for 8 wk . Arteries were harvested , and calcium and phosphorus levels were measured . Atherosclerotic plaque formation and calcium deposits were evaluated by MT and von Kossa staining . All data were presented as means ± SEM . Intergroup comparisons were performed using the Student's t test or one-way analysis of variance ( ANOVA ) . Data obtained at different time points were analyzed by repeated measures ANOVA . SPSS v16 . 0 was used for all statistical analyses , and P<0 . 05 was considered statistically significant . | Atherosclerosis involves hardening of the arteries and can lead to heart disease . Calcium accumulation in blood vessels contributes to this process , and this process is regulated by cells that promote calcium accumulation ( osteoblasts ) and cells that reverse the accumulation ( osteoclasts ) . In this study , we show that vascular calcifying progenitor cells in the blood vessel have the potential to become either osteoblasts or osteoclasts , and that a drug can push these cells towards becoming osteoclasts instead of osteoblasts . Progenitor cells that express both Sca-1 and PDGFRα cell surface proteins were more committed to differentiate into osteoblasts , while cells that only expressed Sca-1 could differentiate into osteoblasts or osteoclasts in a bidirectional manner . Moreover , treatment with a PPARγ agonist could shift the direction of differentiation of Sca-1+/PDGFRα− progenitor cells toward osteoclast-like cells , whereas it cannot influence the fates of Sca-1+/PDGFRα+ progenitors . These results offer new therapeutic targets for reversing calcium accumulation in blood vessels . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"cardiovascular",
"stem",
"cells",
"atherosclerosis",
"mesenchymal",
"stem",
"cells",
"adult",
"stem",
"cells",
"stem",
"cell",
"niche",
"biology",
"hematopoietic",
"stem",
"cells",
"cell",
"biology",
"vascular",
"biology",
"cellular",
"types",
"molecular"... | 2013 | Vascular Calcifying Progenitor Cells Possess Bidirectional Differentiation Potentials |
Several viruses encode factors that promote host mRNA degradation to silence gene expression . It is unclear , however , whether cellular mRNA turnover pathways are engaged to assist in this process . In Kaposi's sarcoma-associated herpesvirus this phenotype is enacted by the host shutoff factor SOX . Here we show that SOX-induced mRNA turnover is a two-step process , in which mRNAs are first cleaved internally by SOX itself then degraded by the cellular exonuclease Xrn1 . SOX therefore bypasses the regulatory steps of deadenylation and decapping normally required for Xrn1 activation . SOX is likely recruited to translating mRNAs , as it cosediments with translation initiation complexes and depletes polysomes . Cleaved mRNA intermediates accumulate in the 40S fraction , indicating that recognition occurs at an early stage of translation . This is the first example of a viral protein commandeering cellular mRNA turnover pathways to destroy host mRNAs , and suggests that Xrn1 is poised to deplete messages undergoing translation in mammalian cells .
Tight control of gene expression is achieved not only at the level of transcription , but also by modulating post-transcriptional events such as mRNA turnover . Indeed , recent studies show that changes in mRNA stability account for as much as 40% to 60% of the changes in steady-state mRNA levels in basic cellular processes such as signaling pathways [1] , stress responses [2] and cell differentiation [3] . A core set of conserved enzymes is responsible for both basal and regulated mRNA degradation in eukaryotes , including two potent exoribonucleases , the 5′-3′ exonuclease Xrn1 and a 3′-5′ exonucleolytic complex called the exosome ( reviewed in [4] , [5] ) . The activity of these exonucleases on eukaryotic mRNAs is kept in check by the presence of a 5′ 7-methylguanosine cap and a 3′ poly ( A ) tail , which prevent access to the message body . Poly ( A ) removal by one of the cellular deadenylases is likely the rate limiting step of basal mRNA degradation [6] and is in turn required for message decapping [7] by the Dcp2 enzyme in complex with the activator Dcp1A and other cofactors [8] , [9] . Endonucleases are also emerging as important contributors to eukaryotic mRNA decay , particularly in the context of quality control pathways or other situations requiring rapid inactivation of select messages [10] . Because of the central role of mRNA degradation in gene expression regulation , viruses presumably interact with cellular RNA degradation machinery in the process of taking control of the cell for their own replication . Viruses often interface with degradation pathways to prevent turnover of their genomic or messenger RNAs [11] , [12] , and can downregulate specific messages using both viral and cellular microRNAs ( miRNAs ) , often to modulate immune evasion or other steps in their lifecycle [13] . Some viruses can even co-opt core mRNA degradation components for their own purposes , as flaviviruses do to generate noncoding subgenomic viral RNAs [14] . However , no example has been found of viruses that use the mRNA turnover machinery to broadly target cellular messages for destruction . Several virus-encoded factors can cause widespread degradation of cellular mRNAs and block host gene expression ( host shutoff ) , but it is unclear what role host pathways play in this context [15] , [16] , [17] , [18] . One of the viruses causing extensive host mRNA degradation is Kaposi's sarcoma-associated herpesvirus ( KSHV ) , the etiologic agent of Kaposi's sarcoma , primary effusion lymphoma , and B-cell type multicentric Castleman's disease [19] , [20] , [21] . During lytic infection with KSHV or other human and murine gammaherpesviruses , expression of the viral protein SOX ( ORF37; BGLF5 in Epstein-Barr virus ) induces degradation of the majority of the cellular messenger RNAs [15] , [18] , [22] , [23] , [24] . This depletion of cytoplasmic mRNAs leads to nuclear relocalization of poly ( A ) binding protein ( PABPC ) , which subsequently drives mRNA hyperadenylation in the nucleus and a concomitant mRNA export block [25] , [26] , [27] . Thus , SOX activity in the cytoplasm triggers a cascade of events that further restricts gene expression . SOX belongs to a family of DNA alkaline exonucleases ( DNases ) that are conserved in all herpesviral subfamilies and have roles in processing and maturation of the newly replicated viral DNA genomes . In vitro , several herpesviral SOX homologs exhibit robust 5′-3′ DNase activity and weaker exonuclease activity towards RNA substrates [28] , [29] , [30] , [31] , [32] . It is notable , however , that only the homologs from viruses of the gammaherpesvirus subfamily can promote RNA turnover in cells , and in these proteins the DNase and host shutoff activities can be genetically separated [15] , [33] . It is therefore unclear to what extent the in vitro RNase activity contributes to the host shutoff activity in cells , and what cellular cofactors may participate in the specific targeting and efficient destruction of mRNAs . Here , we show that the mechanism of KSHV-induced mRNA turnover involves the coordinated activities of SOX and the cellular 5′-3′ ribonuclease Xrn1 . Unlike canonical cellular mRNA decay , in which Xrn1 generally gains access to mRNAs only after deadenylation and decapping , SOX generates substrates for Xrn1 that have not undergone these rate-limiting events . Our data suggest that this occurs via a SOX-induced site-specific endonucleolytic cleavage on each mRNA , thus providing an accessible 5′ end for Xrn1-mediated degradation of the mRNA body . Furthermore , we show that SOX co-sediments with 40S translation initiation complexes , and causes mRNA cleavage at an early stage of translation . SOX specifically targets polymerase ( Pol ) II-generated mRNAs , but not RNAs transcribed by Pol I or Pol III . This leads to a global depletion of cellular mRNAs in polysomes , and explains the preferential targeting of messenger RNAs during host shutoff . Our data suggest a model in which the virus co-opts host mRNA decay pathways to rapidly liberate cellular translation machinery , presumably to create an optimal host environment for viral gene expression and replication .
The KSHV protein SOX and its homologs in other gammaherpesviruses are potent inducers of cellular mRNA degradation . While they have also recently been shown to exhibit RNase activity in vitro , mutational analyses indicated that this activity cannot fully account for SOX-induced mRNA degradation in cells [28] . We therefore hypothesized that SOX co-opts cellular mRNA degradation machinery to enact host transcriptome destruction . We reasoned that the 5′-3′ exoribonuclease Xrn1 would be a likely candidate co-factor , as it plays a major role in basal mRNA turnover in the cytoplasm . To test the involvement of Xrn1 in SOX-mediated mRNA degradation , we monitored SOX-induced turnover of a GFP reporter mRNA in 293T cells upon transfection of Xrn1-specific versus control small interfering RNAs ( siRNAs ) . Unexpectedly , depletion of Xrn1 did not protect full-length GFP mRNA , but it resulted instead in accumulation of a shorter GFP RNA intermediate ( Figure 1A ) . The intermediate could be specifically detected by Northern blotting with probes against the 3′ but not the 5′ end of the mRNA ( Figure 1A ) , indicating a 5′ end truncation . The reliance on Xrn1 to assist in the degradation of mRNA in SOX-expressing cells was also observed for three additional reporter mRNAs , including DsRed2 ( Figure 1B and S1A-B ) and for the endogenous cellular transcript glyceraldehyde 3-phosphate dehydrogenase ( GAPDH , Figure 1C ) . Notably , although the full-length GFP and DsRed2 reporter mRNAs are of similar size ( 1 . 2–1 . 5 kb ) , the lengths of their degradation intermediates were different: the GFP fragment was approximately 1 . 1 kb whereas the DsRed2 fragment was ∼600 bp ( data not shown ) . In addition , the GAPDH mRNA generated multiple cleavage intermediates . This indicates that the generation of the intermediates is not controlled by a positional cue , such as distance from the mRNA ends . We further confirmed that Xrn1 is involved in SOX-induced RNA degradation in an siRNA-independent manner by use of a flaviviral structural element ( SLII ) that can block 5′-3′ RNA degradation by Xrn1 [14] . In SOX expressing cells , insertion of the SLII element within the 3′ UTR of the DsRed2 reporter resulted in the appearance of a protected fragment of a size consistent with the portion of the mRNA downstream of the SLII ( Figure 1D ) . In contrast , RNA fragments were not observed upon insertion of the SLII in the DsRed2 coding region upstream of the predicted 5′ end of the intermediate or in the 5′ UTR . This agrees with the observation that the 5′ portion of the mRNA is removed via an Xrn1-independent mechanism . Results consistent with this interpretation were also obtained for the GFP and β-globin reporters ( Figure S1C–E ) . Collectively , our data suggest that SOX-mediated RNA depletion is a two-step process: an initiating event that removes a portion of the 5′ end of the mRNA , which is then followed by exonucleolytic degradation of the remaining fragment by the host Xrn1 enzyme . In mammalian cells , sequential removal of the protective 3′ poly ( A ) tail and 5′ cap is generally required for Xrn1 to gain access to the RNA , as it can only degrade RNAs with a free 5′ monophosphate . Moreover , deadenylation is thought to be the rate-limiting step regulating mRNA degradation . Several results , however , indicate that SOX promotes mRNA degradation by Xrn1 while bypassing the requirement for deadenylation . Whereas the effects of a single copy of the Xrn1-blocking SLII element were only apparent in cells transfected with SOX ( as these undergo enhanced mRNA turnover ) ( Figure 1D and S1D-E ) , insertion of two copies of the SLII ( GFP-2xSLII , Figure S2A ) led to a stronger Xrn1 block that can be visualized during basal GFP mRNA turnover even in cells lacking SOX ( Figure 2A ) . In the absence of SOX , in vitro removal of the poly ( A ) tails by incubation with oligo ( dT ) and RNase H had little effect on the mobility of the degradation fragment resulting from the GFP-2xSLII construct , confirming that this fragment was generated after deadenylation . In contrast , the fragment produced in SOX-expressing cells was significantly larger prior to RNase H treatment , indicating that the cleaved fragment retained its poly ( A ) tail ( Figure 2A ) . Similarly , we used RNAse H assays to evaluate the polyadenylation status of the SOX degradation intermediate present in Xrn1-depleted cells . Poly ( A ) tail removal decreased the mobility of both the full-length GFP and DsRed2 mRNAs , as well as the Xrn1-targeted degradation intermediates ( Figure 2B and 2C ) . In addition , we were able to purify the GFP intermediate after poly ( A ) selection of the mRNA over oligo ( dT ) coupled beads ( data not shown ) . Together , these data indicate that SOX can bypass the main regulatory step of normal mRNA degradation and render RNAs directly accessible to Xrn1 . Given that deadenylation generally precedes mRNA cap removal , we predicted that mRNA degradation in SOX-expressing cells would occur in a decapping-independent manner . Consistent with our hypothesis , we observed SOX-induced mRNA turnover in cells overexpressing a dominant negative mutant of the Dcp2 decapping enzyme , Dcp2 E148Q ( J . Lykke-Andersen , personal communication; [34] ) , that had also been subjected to siRNA-mediated depletion of the decapping co-activator Dcp1A ( Figure 2D and S2C ) . Although at least one additional decapping enzyme exists in mammalian cells [35] , overexpression of the Dcp2 mutant was sufficient to reduce basal mRNA turnover of the GFP-2xSLII reporter , consistent with an inhibition of decapping activity ( Figure S2B ) . We also depleted levels of the hRrp41 subunit of the 3′-5′ exosome and saw no difference between control and exosome siRNA-treated cells in our assay , in agreement with previous observations [25] ( Figure S2D ) . Neither 5′ nor 3′ degradation fragments were detected upon hRrp41 knockdown ( Figure S2E ) . A caveat with siRNA knockdown experiments in general is that negative results could be due to insufficient depletion of the protein . However , taken together our data suggest that the viral SOX protein interfaces specifically with the Xrn1 enzyme to accomplish host shutoff . In addition , these results further argue that removal of the 5′ portion of the RNA occurs independently of the canonical 5′-3′ decay pathway . One way that mRNAs can be made accessible to Xrn1 without prior deadenylation or decapping is by internal endonucleolytic cleavage , a mechanism used by host quality control pathways to rapidly eliminate flawed mRNAs . Indeed , the defined length of the intermediates that we observe following Xrn1 depletion is strikingly reminiscent of the Smg6-cleaved intermediates seen during nonsense mediated decay ( NMD ) [36] , [37] . A defined-length intermediate also suggests that cleavage is directed to a specific location or sequence within the mRNA . The different size of the intermediate observed in GFP and DsRed2 ( compare Figures 1A and 1B ) further indicated that a specific element , rather than positional information , is likely to direct the truncation . To test whether there is a specific sequence that leads to an endonucleolytic cut in the presence of SOX , we constructed a modified GFP reporter that contained an internal in-frame repeat of a 201 nt region encompassing the predicted cleavage site ( Figure 3A , GFPrep ) . If this region contained an element targeted by an endonuclease , we expected that in the presence of SOX , GFPrep would give rise to two intermediates , assuming that the two cleavage events were equally likely and mutually exclusive . Indeed , we found that the GFPrep reporter generated two intermediates of sizes corresponding to two independent cleavage events ( Figure 3B , arrowheads ) . These data suggest that an element within the initial 201 nucleotides of the GFP coding region directs an endonucleolytic cleavage in SOX-expressing cells . To identify sequences involved in directing the internal cleavage , we used 5′ rapid amplification of cDNA ends ( RACE ) to map the 5′ end of the degradation intermediates found in GFP , DsRed2 and β-globin . In all cases our RACE results confirmed what we had previously seen with the Northern blot analysis , in that a single predominant intermediate was amplified in Xrn1-depleted cells in a SOX-dependent manner ( Figure 3D and data not shown ) . The majority of the sequences within this band terminated at a single nucleotide or within few nucleotides of each other ( Figure S3 ) . Analysis of the sequences surrounding the cleavage site for each mRNA revealed a conserved stretch of five bases ( TGAAG ) just upstream of the cleavage site . Deletion of the TGAAG element from the GFP construct abolished production of the cleavage fragment , confirming its role in SOX targeting ( Figure 3E ) . However , insertion of this sequence alone ( nt 131–138 ) or of a 25-nucleotide stretch ( nt 126–150 ) surrounding the GFP cleavage site was not sufficient to elicit generation of a second fragment ( Figure 3F ) . We therefore hypothesize that the TGAAG sequence is an essential component of a larger element , perhaps structural , involved in directing endonucleolytic cleavage of mRNAs in SOX-expressing cells . In vitro , several herpesviral homologs of SOX exhibit weak DNA endonuclease activity as well as RNase activity on linear RNA substrates [28] , [29] , [30] , [38] , [39] , [40] . Thus , it is possible that SOX itself carries out the initial truncation of mRNAs in cells . Recently , the structure of SOX and its homolog BGLF5 in Epstein-Barr virus were solved , leading to the identification of catalytic residues responsible for the DNase activity of these proteins , as well as putative DNA binding residues [28] , [30] , [41] . Mutation of one of the catalytic residues in BGLF5 ( D203S ) was shown to abolish all in vitro nuclease activity without misfolding the protein [30] . To test whether the SOX catalytic core is required for host shutoff in vivo , we generated individual SOX mutants for each of the catalytic residues identified in the crystal structures: E184A , D221A or D221S ( equivalent to D203S in BGLF5 ) , E244A and K246I . Indeed , mutation of any one of the catalytic residues abolished the ability of SOX to deplete GFP mRNA in cells ( Figure 4A ) . As a control , we mutated the putative DNA-binding residues ( W135V , R139I , S144I , S146I , S219A , Q376G ) , which are also located in the active site cleft ( Figure S4C ) , and found that only residues R139 , S144 and Q376 were required for host shutoff ( Figure 4B ) . This suggests that SOX may bind RNA and DNA using partially overlapping residues . In contrast with our host shutoff data , all the catalytic and putative DNA binding residues tested were required for in vitro DNase activity ( Figure S4A and S4B ) . Thus , our mutational analysis suggests that the catalytic activity of SOX- presumably the RNase activity- is required for host shutoff . We next tested directly whether the RNA turnover activity of SOX was responsible for the generation of the degradation intermediate . We compared production of the degradation intermediate when expressing a SOX mutant lacking only DNase activity associated with viral genome processing ( Q129H ) [33] , only the mRNA turnover activity ( P176S ) [33] , or a catalytic mutant lacking both activities ( D221S ) ( Figure 4C ) . The degradation intermediate was not produced in either of the two mutants lacking RNA turnover activity , but was readily detectable in cells expressing Q129H , which selectively lacks DNase function . These results indicate that generation of the intermediate by SOX is closely linked to its ability to degrade mRNAs , consistent with the two-step model for SOX-mediated mRNA degradation . We next sought to determine how SOX targets cytoplasmic mRNAs for the initial cleavage event . In vitro , SOX exhibits relatively weak affinity for RNA ( Kd of 87 µM ) , suggesting that in cells it is recruited to mRNAs via associations with host factors [28] . In cellular quality control pathways such as NMD , translation is required for error recognition and the primary endonucleolytic cleavage [42] , [43] . In addition , it has recently been reported that the vast majority of mRNAs in the cytoplasm are polysome-associated [44] , suggesting that targeting mRNAs engaged in translation would be an efficient mechanism to clear host messages during host shutoff . To examine the effects of host shutoff on translating mRNAs , we performed polysome profiling of a B cell line ( BCBL-1 ) derived from a patient with primary effusion lymphoma , which harbors KSHV in a latent state . We used a line of BCBL-1 cells bearing an inducible version of the KSHV major lytic transactivator RTA ( TREx BCBL-1-RTA ) [45] to allow efficient lytic reactivation following RTA induction . Upon chemical stimulation of lytic KSHV replication in these cells ( when SOX is expressed from the viral genome ) , we observed a significant decrease in the polysome population and a corresponding increase in 80S monosomes , consistent with degradation of actively translating mRNAs ( Figure 5A ) . It should be noted that the level of polysome depletion during viral infection is likely underestimated , as ∼20% of induced cells generally fail to enter the lytic cycle . The depletion of polyribosomes was not due to chemical treatment alone , because treatment of the KSHV negative BJAB B cell line did not result in a similar decrease in polysomes ( data not shown ) . We next looked for the presence of SOX in gradient fractions from the corresponding polysome profiles of lytically reactivated BCBL-1 cells . As controls , we also blotted for PAIP2A , a protein not found in polysomes [46] , as well as the ribosomal protein RPS3 ( Figure 5B ) . SOX appeared to sediment primarily with the ribonucleoprotein ( RNP ) , 40S , and monosome fractions , and exhibited partial overlap with RPS3 ( Figure 5B ) . Puromycin treatment disrupted polysomes but failed to alter the SOX sedimentation profile , arguing against a specific association with the 80S and polysome fractions ( data not shown ) . To more accurately determine its sedimentation profile , we increased the resolution of the lighter molecular weight complexes using lower density sucrose gradients . These experiments revealed that SOX indeed sediments in both the RNP and 40S fractions , similar to Xrn1 ( Figure 5C ) . Its sedimentation profile also overlapped with the eIF3j and eIF2α components of the 40S pre-initiation complex , which is recruited to the 5′ cap prior to ribosomal scanning ( Figure 5C ) . This is in contrast to the sedimentation of PABPC , which remains bound to the mRNA throughout the polysome fractions as well ( Figure 5C; data not shown ) . Similar data were obtained upon transient expression of SOX in 293T cells ( Figure S5A ) . We also tested the sedimentation profile of the SOX D221S catalytic mutant , and found it to mimic that of wild-type SOX ( Figure S5B ) , indicating that the catalytic site of the protein is unlikely to be involved in its recruitment to target RNAs . Given that SOX is the dominant effector of host shutoff during KSHV infection , we hypothesized that polysome depletion in BCBL-1 cells was a consequence of SOX activity . We therefore tested the effects of SOX on the endogenous translating mRNA pool through polysome profiling of 293T cells expressing either wild-type SOX , the SOX catalytic mutant D221S , or the single function SOX P176S mutant lacking host shutoff but not DNase activity . Indeed , expression of wild-type SOX alone caused a significant depletion of polysomes relative to vector-transfected cells , whereas neither of the host shutoff defective SOX mutants had this effect ( Figure 5D ) . These results additionally confirm that the catalytic activity of SOX is required to initiate widespread turnover of endogenous host messages . We had previously observed that a translationally incompetent version of the GFP mRNA , which terminates by ribozyme cleavage and lacks a poly ( A ) tail ( GFP-HR ) ( Figure S6A ) , escaped SOX-mediated turnover [25] . Although this mRNA is inefficiently exported , subcellular fractionation experiments confirmed that even the exported cytoplasmic population of GFP-HR was not subject to SOX-induced turnover ( Figure S6B ) . Thus , the failure of SOX to degrade this mRNA in the cytoplasm could be due to its translational incompetence . To further explore a role for translation in SOX-induced mRNA turnover , we examined SOX turnover of RNAs transcribed by Pol I and Pol III . Pol I and Pol III transcription does not result in the addition of the cap and poly ( A ) tail , mRNA modifications critical to translation initiation specifically associated with Pol II transcription . Using a pure population of cells expressing GFP-SOX or GFP alone , we found that neither the endogenous Pol III-generated Y3 and 7SL cytoplasmic non-coding RNAs , nor the 18S rRNA transcribed by Pol I undergo turnover in the presence of SOX ( Figure 6A ) . In contrast , we observed significant depletion of endogenous mRNAs transcribed by Pol II , including GAPDH , β-actin , and stearoyl-CoA desaturase ( SCD ) in SOX-expressing cells ( Figure 6A ) . In principle , the inability of SOX to degrade the non-Pol II transcripts could be due to the absence of an ORF , or the presence of RNP complexes occluding SOX cleavage sites . To exclude these possibilities , we expressed the GFP reporter under the control of Pol I or Pol III promoters and found that in both cases , the GFP RNA was not targeted by SOX ( Figure 6B ) . We confirmed using subcellular fractionation experiments that the inability of SOX to promote degradation of these RNAs was not due to a failure of the RNAs to be exported ( Figure S6E and S6F ) . Collectively , these data indicate that RNAs must be translationally competent to be targeted by SOX . To determine whether mRNA degradation by SOX required ribosomal passage over or near the putative initial cleavage site , we designed a reporter dsRed2 mRNA harboring a termination codon upstream of the putative SOX cleavage site ( dsRed2-100stop ) . This prevented production of full-length dsRed2 protein ( Figure 6C ) . Interestingly , in cells depleted of Xrn1 , dsRed2-100stop was cleaved similarly to wild-type dsRed2 upon SOX expression ( Figure 6D ) , indicating that ribosomal passage over the cleavage site is not necessary for its recognition . To determine whether 60S joining is required for cleavage , we made use of a modified version of the encephalomyocarditis virus internal ribosomal entry site ( IRES ) , termed ΔEMCV , which cannot promote cap-independent translation but is highly structured and decreases translation initiation in vitro [47] . Indeed , insertion of ΔEMCV into the GFP 5′ UTR 30 nt from the cap significantly reduced GFP protein accumulation ( Figure 6E ) . However , in Xrn1-depleted cells , SOX still induced cleavage of this mRNA with similar efficiency as wild-type GFP , and at the same site ( Figure 6F ) . In agreement with the ΔEMCV-GFP data , we observed that SOX could also promote turnover of a GFP mRNA with a 60 nt adenylate tract inserted in the 5′ UTR ( 5′A60-GFP ) ( Figure S6D ) , which similarly reduced GFP protein production ( Figure S6C ) . A 50–70 nt adenylate-rich tract in the 5′ UTR of PABPC has likewise been shown to repress translation of its message in an autoregulatory manner as a consequence of PABPC protein binding to this region [48] . Consistent with the sedimentation profile of SOX , these data suggest that in SOX-expressing cells , mRNAs are targeted for cleavage at an early step during translation , perhaps prior to AUG recognition . To monitor directly where cleavage occurs , we performed sucrose density gradient centrifugation on Xrn1-depleted cells expressing SOX and the GFP reporter . In agreement with our above findings , the cleaved intermediate preferentially accumulates in the 40S fraction ( Figure 7A ) . The degradation intermediate is absent in all fractions from cells expressing the SOX catalytic mutant D221S ( Figure 7B ) . These results lead us to favor a model in which mRNAs are recognized by SOX during formation of the 40S preinitiation complex , at which point they undergo SOX-induced endonucleolytic cleavage and subsequent destruction by Xrn1 .
This study shows that the global destruction of cellular mRNA during KSHV-induced host shutoff is enacted through the coordinated activities of both the viral SOX protein and cellular Xrn1 . Our findings suggest that , in cells , SOX directs endonucleolytic cleavage of mRNAs at an early stage during translation , and subsequently employs host nucleases such as Xrn1 to execute degradation of the mRNA body . Thus , the virus is able to make use of the available host mRNA turnover machinery , yet bypass the rate limiting events that normally precede activation of Xrn1 . Collectively , these data and our previous studies reconcile the seemingly disparate observations that the gammaherpesvirus SOX homologs promote potent host shutoff , yet exhibit relatively weak in vitro RNase activity . We hypothesize that while SOX can catalyze RNA cleavage , it does so in a site-specific manner in cells and requires one or more cellular factors to recruit and/or position it on the mRNA substrate , as well as host enzymes to complete the degradation process . Endonucleolytic cleavages enable rapid mRNA turnover , because they generate unprotected RNA ends while bypassing the requirement for deadenylation and decapping . Cellular endonucleases are therefore generally subject to tight regulation . In eukaryotic cells , endonuclease-mediated decay is associated with mRNA quality control pathways , such as NMD , which targets messages containing premature stop codons [36] , [37] , and No-Go decay ( NGD ) , which destroys mRNAs that experience stalls in translation [49] . In addition , recent studies have described previously unappreciated contributions of eukaryotic endonucleases to other basic processes such as miRNA-mediated mRNA decay [50] , ER stress responses [51] , [52] , and exosome-mediated decay [53] , [54] , [55] . A few endonucleases that regulate decay of specific sets of otherwise normal mRNAs have also been described [10] . Several observations suggest that a primary role of SOX during host shutoff is to mediate endonucleolytic cleavage of target mRNAs , thereby enabling direct access by Xrn1 . A key finding is that Xrn1-depleted cells undergoing host shutoff accumulate defined size degradation intermediates , similar to those derived from transcripts undergoing NMD or NGD under conditions of limiting Xrn1 [36] , [37] , [42] , [49] , [56] . Additionally , the observation that deadenylation and Dcp1/2-mediated decapping are not required to generate the degradation intermediate suggest that the 5′ and 3′ ends of the mRNA remain intact at the time of cleavage . We predict that SOX induces a sequence-specific rather than a position-dependent cut , as duplicating a 201 nt region surrounding the cut site yielded a second degradation intermediate , and addition of sequences upstream of the cleavage site did not alter the cleavage efficiency or location . We identified a conserved stretch of five nucleotides just upstream of the cut site in three different reporters . This consensus site was necessary , but not sufficient to elicit cleavage , suggesting that it represents a portion of a larger element required for SOX targeting . We hypothesize that the full element is a complex sequence or structure , perhaps degenerate , enabling it to be widespread among cellular mRNAs . It is likely that additional enzymes besides Xrn1 participate in clearance of the fragments created by SOX , particularly in removal of the 5′ degradation intermediate . However , we have not observed a role for the canonical 3′-5′ exosome exonuclease complex . One possibility is that another host 3′-5′exonuclease instead mediates degradation of the upstream fragment , or plays a compensatory role upon exosome depletion . This might explain the limited protection of the 5′ fragment afforded by exosome knockdown in studies of another eukaryotic endonuclease [36] . Alternatively , more robust depletion of the exosome might be required to inhibit its activity in the context of cytoplasmic mRNA degradation . SOX bears significant structural similarity to the PD- ( D/E ) XK type II restriction endonuclease superfamily [41] , which includes several proteins that have been experimentally demonstrated to have endonuclease activity on RNA [57] , [58] . The catalytic residues of SOX are essential for mRNA turnover in cells , further supporting a direct role for SOX in the primary cleavage event . Interestingly , in vitro RNase activity has recently been described for SOX and its EBV homolog BGLF5 , although these studies concluded that it functioned as an exonuclease [28] , [30] . However , RNases can harbor both endonuclease and exonuclease activity in the same active site [59] , [60] . In addition , our data indicate that the endonucleolytic activity of SOX occurs in a site-specific manner , whereas the published assays used generic RNAs . Our results support the involvement of an early stage of translation in the targeting of mRNA by SOX . This would allow the virus to selectively eliminate competing host mRNAs , while sparing regulatory RNAs that may be necessary for its own gene expression . Eukaryotic translation initiates with the recruitment of the 40S ribosomal subunit to the cap via interactions with translation initiation factors , followed by scanning to the AUG codon and recruitment of the 60S subunit . Two observations indicate that recognition and turnover in SOX-expressing cells are initiated prior to 60S joining . Specifically , insertion in the 5′ UTR of elements that presumably inhibit or prematurely terminate 40S scanning , and thus significantly reduce protein production , does not affect SOX activity . Moreover , the degradation intermediate accumulates preferentially in the 40S fraction . Our finding that SOX cosediments with the 40S initiation complex suggests that an association with translation initiation machinery directs it to mRNAs . Supporting this prediction is the failure of SOX to target translationally incompetent reporter and endogenous RNAs transcribed by RNA polymerase I or III . We hypothesize that SOX targeting requires recognition of the mRNA 5′ end , likely via associated translation initiation factors , and that this recruitment somehow allows SOX access to the cleavage site ( s ) within the mRNA . Interestingly , our data show that cleavage can occur hundreds of nucleotides away from the site of translation initiation . One interesting parallel to these seemingly discrepant observations is that of the E . coli endonuclease RNase E , which can cleave anywhere along the length of the RNA ( preferably within AU-rich sequences ) , but requires a monophosphate at the mRNA 5′ end . Thus , similar to our proposed model for SOX activity , RNase E can simultaneously recognize two non-adjacent regions of the primary RNA sequence [61] . These observations can be reconciled by the fact that RNAs adopt secondary and tertiary structures within a cell that could presumably enable such sequences to be juxtaposed . Alternatively , this distance could be bridged if there was an additional interaction of SOX with a host protein co-factor bound to the cleavage element . While this is the first example of a viral protein working in concert with host RNA turnover components to broadly target cellular messages , it is likely that several other viruses use similar mechanisms to enact host shutoff . Alphaherpesviruses and SARS coronavirus also encode host shutoff factors ( vhs and nsp1 , respectively ) that promote mRNA degradation , but bear no homology to SOX [16] , [17] . In vitro data indicate that they similarly induce endonucleolytic cleavage of RNAs , and it has been proposed that degradation of the mRNA body in cells may be assisted by host exonucleases , although this has yet to be shown [62] , [63] , [64] , [65] , [66] . In addition , all three viral proteins appear to use components of the translation machinery as a means to access mRNA [62] , [65] , [67] , [68] , [69] . Such parallels suggest that these viruses have adopted remarkably similar strategies to execute mRNA decay , resembling those of host pathways like NMD . In addition to the enhanced rate of decay afforded by an endonucleolytic mechanism , the advantage for the virus may be that this strategy of host shutoff generates intermediates that look indistinguishable from products of quality control pathways . Such intermediates would then be readily degraded by the core degradation machinery and might not be recognized as aberrant , perhaps avoiding activation of stress or innate immune responses . Thus , understanding how viral endonucleases interface with host pathways may provide insight into how manipulation of these pathways contributes to infectious disease , as well as into events that regulate cellular RNA decay .
Protein lysates were either prepared in RIPA buffer [50 mM Tris-HCl ( pH8 . 0 ) , 150 mM NaCl , 1% ( v/v ) Nonidet P-40 , 0 . 5% ( w/v ) sodium deoxycholate , 0 . 1% ( w/v ) sodium dodecyl sulfate ( SDS ) ] , or fractionated using the NE-PER kit ( ThermoScientific ) . Western blots were performed with either mouse monoclonal anti-GFP ( 1∶2000 , BD Biosciences ) , mouse monoclonal anti-dsRed ( 1∶500 , Clontech ) , mouse monoclonal anti-HA ( 1∶2000 , Invitrogen ) , mouse monoclonal anti-RPS3 ( 1∶1000 , Abcam ) , mouse monoclonal anti-Flag ( 1∶1000 , Sigma ) , rabbit polyclonal anti-PAIP2A ( 1:2000 , kindly provided by N . Sonenberg [70] ) , rabbit polyclonal anti-hXrn1 ( 1∶5000 , kindly provided by J . Lykke-Andersen ) , rabbit polyclonal anti-hDcp1a ( 1∶5000 , kindly provided by J . Lykke-Andersen ) , rabbit polyclonal anti-hDcp2 ( 1∶400 , kindly provided by M . Kiledjan ) , rabbit polyclonal anti-hRrp41 ( 1∶1000 , see below ) , rabbit polyclonal anti-SOX J5803 ( 1∶5000 , [33] ) , mouse monoclonal hnRNPC1/C2 ( 1∶2000 , Abcam ) , mouse monoclonal anti-Hsp90 ( 1∶3000 , Stressgen Bioreagents ) , mouse monoclonal anti-tubulin ( 1∶3000 , Sigma Aldrich ) , goat polyclonal anti-actin ( 1∶5000 , Santa Cruz Biotechnology ) , mouse monoclonal 10e10 anti-PABPC ( 1∶2000 , Santa Cruz Biotechnology ) , rabbit polyclonal anti-eIF2α ( 1∶1000 , Cell Signaling ) or rabbit polyclonal anti-eIF3j ( 1∶1000 , Cell Signaling ) . Rabbit polyclonal antibodies were raised against recombinant maltose binding protein ( MPB ) -fused hRrp41 purified from E . coli . Total cellular RNA was isolated for Northern blotting using RNA-Bee ( Tel-Test ) . Where indicated , the NE-PER kit ( ThermoScientific ) was used for cellular fractionation prior to RNA extraction . Northern blots were probed with 32P-labeled DNA probes made using either RediPrime II ( GE Healthcare ) or Decaprime II ( Ambion ) or , for β-globin , an SP6 transcribed 32P-labeled riboprobe . RNase H assays were carried out as previously described [25] . Results in each figure are a representative of at least 3 independent replicates of each experiment . Image J ( http://rsbweb . nih . gov/ij/ ) was used for quantification of Northern and Western blots . Profiles were obtained from uninduced or lytically reactivated TREx BCBL-1-RTA cells , or from 293T cells transfected with the indicated plasmids . Polysome profiles were carried out as described in Jackson and Larkins [71] , except that cells were treated with 100 µg/ml CHX for 30 minutes prior to harvesting . BCBL-1 and 293T extracts were pelleted through 60% sucrose before layering or were directly layered on a 15–60% sucrose gradient containing 100 µg/ml CHX . Additional details , including the procedure for resolution of RNP/40S fractions are described in supplemental procedures ( Text S1 ) . Additional experimental procedures , primer and siRNA/shRNA sequences used in the study are detailed in Supplemental Materials and Methods ( Text S1 ) . | Viruses use a number of strategies to commandeer host machinery and create an optimal environment for their replication . One strategy employed by oncogenic gammaherpesviruses such as Kaposi's sarcoma-associated herpesvirus ( KSHV ) is to block cellular gene expression through extensive destruction of mRNAs . A single viral protein called SOX is sufficient to drive this phenotype , but the mechanism by which it does so has remained unclear . Here we show that host mRNA destruction is the result of the coordinated action of SOX and a cellular RNA degrading enzyme , Xrn1 . By cleaving mRNAs internally , SOX recruits the activity of Xrn1 while bypassing the regulatory mechanisms that normally prevent this enzyme from prematurely degrading mRNAs . We also find that SOX co-sediments with translation complexes , and specifically targets mRNAs for cleavage at an early stage of translation . We hypothesize this allows the virus to selectively target mRNAs , thereby liberating host gene expression machinery . Collectively , these findings describe a novel interplay between the gammaherpesvirus SOX protein and cellular degradation machinery , and shed light on how a single viral component can hijack cellular machinery to efficiently destroy messages . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"rna",
"molecular",
"cell",
"biology",
"nucleic",
"acids",
"virology",
"viral",
"enzymes",
"biology",
"microbiology",
"rna",
"stability",
"molecular",
"biology"
] | 2011 | Coordinated Destruction of Cellular Messages in Translation Complexes by the Gammaherpesvirus Host Shutoff Factor and the Mammalian Exonuclease Xrn1 |
The intimate synapsis of homologous chromosome pairs ( homologs ) by synaptonemal complexes ( SCs ) is an essential feature of meiosis . In many organisms , synapsis and homologous recombination are interdependent: recombination promotes SC formation and SCs are required for crossing-over . Moreover , several studies indicate that initiation of SC assembly occurs at sites where crossovers will subsequently form . However , recent analyses in budding yeast and fruit fly imply a special role for centromeres in the initiation of SC formation . In addition , in budding yeast , persistent SC–dependent centromere-association facilitates the disjunction of chromosomes that have failed to become connected by crossovers . Here , we examine the interplay between SCs , recombination , and centromeres in a mammal . In mouse spermatocytes , centromeres do not serve as SC initiation sites and are invariably the last regions to synapse . However , centromeres are refractory to de-synapsis during diplonema and remain associated by short SC fragments . Since SC–dependent centromere association is lost before diakinesis , a direct role in homolog segregation seems unlikely . However , post–SC disassembly , we find evidence of inter-centromeric connections that could play a more direct role in promoting homolog biorientation and disjunction . A second class of persistent SC fragments is shown to be crossover-dependent . Super-resolution structured-illumination microscopy ( SIM ) reveals that these structures initially connect separate homolog axes and progressively diminish as chiasmata form . Thus , DNA crossing-over ( which occurs during pachynema ) and axis remodeling appear to be temporally distinct aspects of chiasma formation . SIM analysis of the synapsis and crossover-defective mutant Sycp1−/− implies that SCs prevent unregulated fusion of homolog axes . We propose that SC fragments retained during diplonema stabilize nascent bivalents and help orchestrate local chromosome reorganization that promotes centromere and chiasma function .
The formation of gametes typically involves halving of the cellular chromosome complement from diploid to haploid . This is achieved via two consecutive rounds of chromosome segregation during the process of meiosis [1] . Prior to the first meiotic division , replicated chromosomes associate into homologous pairs and become connected along their lengths by synaptonemal complexes ( SCs ) [2] , [3] . SCs are proteinaceous structures with a zipper-like morphology [4]–[8] . The tripartite SC structure comprises two lateral elements , inferred to be elaborations of cohesin-based homolog axes , and a central element consisting of transverse filaments that interconnect the two lateral elements [7]–[10] . SC components show tendencies for self-assembly into ordered arrays and SC formation is believed to occur via polymerization from specific nucleation sites where the homolog axes have been brought into close proximity [6] , [11] . In many organisms , including plants , fungi and mammals , the template-dependent DNA-repair process called homologous recombination is coopted during meiosis to facilitate homolog pairing and synapsis [12] . In these cases , SC formation often nucleates at points where recombination brings the homolog axes together [11] . However , organisms such as Drosophila and C . elegans do not require recombination for homolog pairing and SC formation and instead have evolved dedicated chromosome pairing sites [3] , [13] . In addition to promoting chromosome pairing and synapsis , recombination plays a critical function in directing the disjunction of homologs at the first meiotic division . Specifically , crossover recombination in conjunction with sister-chromatid cohesion results in structures called chiasmata that tether homolog pairs and thereby facilitate their stable biorientation on the spindle [14]–[16] . The interdependence of recombination and SCs is further highlighted by the fact that synapsis promotes crossing-over , at least in part by recruiting crossover-specific recombination factors [17] , [18] . Furthermore , studies in a number of organisms imply a functional relationship between SC nucleation sites and crossovers ( reviewed in [11] ) . Specifically , SC formation often initiates at sites where crossovers will subsequently form . Recent studies in Saccharomyces cerevisiae and Drosophila suggest that centromeres play special roles in meiotic chromosome pairing and the initiation of SC formation [19] , [20] . During early prophase in S . cerevisiae , centromeres undergo homology-independent “coupling” , which depends on the SC central element component , Zip1 [21] , [22] . Centromere coupling is proposed to be a driving force for two-by-two chromosome association that facilitates recombination-dependent homolog pairing [22] , [23] . However , analysis of recombination patterns in the zip1 mutant support an alternative proposal that coupling helps to suppress centromere-proximal crossing-over , which is associated with chromosome nondisjunction [24] [25]–[27] . Following initial coupling , centromeres appear to act as nucleation sites for SC polymerization , although it is clear that recombination sites within the chromosome arms are also utilized [11] , [28] . Consistent with a role for centromeres in nucleating SC formation , in a mutant situation where SCs are assembled independently of recombination , centromeres appear to be the exclusive sites of initiation [29] . This same study indicated that Zip3 , a putative E3-ligase for conjugation of the small protein modifier SUMO , negatively regulates SC initiation between centromeres . A role for centromeres in SC formation is further supported by recent studies in Drosophila females , which showed that centromeres undergo SC-dependent clustering and function as SC initiation sites [19] , [30] , [31] . Although crossing-over is a highly efficient process , achiasmate homologs do occasionally arise . In budding yeast , Zip1-mediated centromere coupling plays an additional late role to facilitate the disjunction of achiasmate chromosomes [21] , [32] . In conjunction with the spindle assembly checkpoint , this process promotes the accurate disjunction of a single pair of achiasmate chromosomes in about 90% of meioses ( random segregation predicts disjunction in only 50% of cells ) [32] , [33] . Efficient achiasmate segregation is also observed in Drosophila , although a role for SC components has not been demonstrated [34] . In some mammals , SC components are also inferred to promote the disjunction of achiasmate chromosomes , specifically the sex chromosomes [35]–[37] . Typically , mammalian X and Y chromosomes synapse at short regions of homology , termed pseudoautosomal regions , where crossing-over occurs to form X-Y chiasmata [38]–[40] . However , in the Elegant Fat-tailed Mouse Opossum ( a marsupial ) and the Mongolian gerbil ( a eutherian mammal ) , X-Y chiasmata are not formed , but persistent structures composed of SC proteins appear to tether the X and Y to facilitate their disjunction during anaphase [35]–[37] . Finally , in some insects , SCs are retained until anaphase I and appear to completely supersede the function of chiasmata in directing disjunction [6] . In this study we analyze the interplay between SCs , recombination and centromeres in the mouse . Immunocytological analysis of prophase spermatocytes from wild-type and mutant mice indicates that centromeres do not undergo early-stage coupling and SC assembly never initiates from centromeres . However , centromeres remain associated throughout much of the diplotene stage , connected by short SC fragments . While this general , SC-dependent centromere association appears to be lost prior to diakinesis , we detect a distinct class of inter-centromeric bridges at this stage . These structures could play a more direct role in biorienting homologs on the spindle and raise the possibility of an achiasmate segregation system in mouse . A second distinct class of retained SC fragment is also observed during diplotene and shown to be crossover-dependent . Structured illumination microscopy reveals that these structures mark sites of developing chiasmata and are lost as the homolog axes fuse . Analysis of the SC-defective Sycp1 mutant suggests a novel role for the SC central element in preventing inappropriate interactions between homolog axes . We discuss the idea that SC fragments retained during diplonema function to locally stabilize homolog associations and coordinate important morphological and compositional changes in preparation for chromosome segregation .
Figure 1 shows representative nuclei from each stage of meiotic prophase . Quantification of centromere association and centromere synapsis throughout these stages is presented in panels 1P and 1Q , respectively . This analysis reveals several key features of prophase centromere behavior in mouse . Analogous observations are made by Bisig et al . in the accompanying study [48] . In budding yeast spo11 mutants , which fail to initiate recombination , persistent Zip1-mediated centromere coupling is observed [22] . Moreover , in the absence of recombination , centromeres become the primary sites of SC initiation in budding yeast [29] . Therefore , we analyzed the relationship between centromeres and SC in Spo11−/− knock-out mice ( Figure S1 ) . Neither early centromere-associations nor preferential initiation of SC formation from centromeres were observed in Spo11−/− spermatocytes . Budding yeast Zip3 is a RING-domain protein and putative E3-ligase for the ubiquitin-like modifier , SUMO [55] , [56] ( also see [57] ) . While yeast zip3 mutants show a general synapsis defect , the vast majority of SCs that do form are initiated between centromeres , leading to the proposal that Zip3 specifically inhibits SC initiation between centromeres [28] , [29] . We recently constructed a knock-out mutation of the mouse Zip3 homolog , Rnf212 ( A . R . , H . Q . , J . K . C . and N . H . , unpublished data ) , allowing us to test the idea that RNF212 has an analogous inhibitory function in mammals [58] . Analysis of initial SC stretches in zygotene-stage Rnf212−/− spermatocytes shows that , as in wild type , SC formation does not initiate between centromeres ( Figure S2 ) . Thus , absence of the Zip3 homolog , RNF212 , does not permit SC to initiate between centromeres in mammals . The SYCP1 staining associated with centromeres and nascent chiasmata during diplonema could reflect the retention of fragments of normal SC , or modified structures peculiar to these sites . To begin to distinguish these possibilities , we examined the localization of three SC central element proteins , SYCE1 , SYCE2 and TEX12 , in addition to SYCP1 [59]–[61] . As shown in Figure 2 , all four central element components localize to sites of paired centromeres and chiasmata during diplonema , consistent with the idea that normal SC fragments are retained at these sites . Previous studies have correlated non-centromeric SC fragments at diplonema with crossover sites ( identified as silver-staining recombination nodules [6] , [50] , [52] , [62] ) but , to our knowledge , dependency of these structures on crossing-over has not been directly demonstrated . On the other hand , we predict that persistent centromere-associated SC fragments should occur independently of crossing-over . These inferences were tested by analyzing Rnf212−/− mutant mice , which have normal homolog synapsis , but show a ≥95% reduction of crossing-over ( A . R . , H . Q . , J . K . C . and N . H . , unpublished data ) ( Figure 3 ) . As diplonema progresses and desynapsis ensues in Rnf212−/− spermatocytes , chromosome arms completely dissociate , but SYCP1-associated centromeres frequently remain connected ( Figure 3C–3E ) . Persistent interstitial SC fragments were not observed in Rnf212−/− nuclei , indicating dependence on crossing over . During diplonema , it is not unusual to detect foci of central element components that remain associated with separated homolog axes ( e . g . Figure 2 ) . Thus , the association of central element components with centromeres and crossover sites might not represent true tripartite SC , but merely SC remnants associated with only one homolog axis . These two possibilities were discriminated using structured illumination microscopy ( SIM ) , which has sufficient resolving power to distinguish the two SYCP3-stained SC lateral elements from the SYCP1-stained central element ( Figure 4 ) [63] , [64] . SIM imaging reveals that centromere and crossover associated SYCP1 is sandwiched between the two homolog axes , as expected for true tripartite SC ( Figure 4A–4D ) . SIM analysis of late diplotene nuclei indicates that crossover-associated SC fragments can be very short , comprising on average only 0 . 24 µm of SYCP1 ( Figure 4 ) , less than 3% of the length of an average late pachytene SC ( 8 . 6 µm ) . Moreover , in 41% of these structures ( 29/71 from 6 nuclei ) , the SYCP3-staining homolog axes remain clearly separate implying that they have yet to be exchanged to form chiasmata ( Figure 4E–4G ) . Notably , in mouse , DNA exchange to form crossovers has been shown occur during pachynema [65] , [66] . Thus , DNA crossing-over and axis-remodeling appear to be temporally distinct aspects of chiasma formation . Intriguingly , the 31% of nascent chiasmata sites in which SYCP3 axes converge and begin to fuse ( 22/71 ) are associated with smaller and less intense SYCP1 foci , which are typically localized to one side of the presumed axis-exchange point ( Figure 4E–4G ) . Finally , 28% of nascent chiasmata ( 20/71 ) comprised SYCP3 fusion-points without associated SYCP1 . The analysis above suggests that SC fragments are retained at crossover sites to regulate the exchange of homolog axes . To further explore this idea , we performed SIM analysis of diplotene-like nuclei from the Sycp1−/− mouse , which fails to form SC central element ( Figure 5A , 5B ) . Although synapsis fails in Sycp1−/− meiocytes , the early steps of recombination occur normally , homologs pair and axes closely associate at sites of recombination ( so called , “axial associations” ) [17] . In nuclei with late-stage chromosome morphology , previously defined to be in diplonema [17] , axial associations appeared as a mixture of separate and conjoined/fused SYCP3 axes ( Figure 5C ) , similar to the nascent chiasmata observed in wild-type diplotene nuclei . However , chiasma-like structures in Sycp1−/− nuclei are distinct from those in wild type . First , they are more numerous , averaging 1 . 8 ( ±1 . 1 SD , n = 89 ) per homolog pair compared to 0 . 96 ( ±0 . 63 SD , n = 74 ) in wild type ( P<0 . 0001 , Mann-Whitney test ) . Second , 87 . 3% ( 137/157 ) of chiasma-like structures in Sycp1−/− cells have conjoined or fused axes compared to 59 . 1% in wild type ( 42/71; P<0 . 0002 , z-test ) . This observation suggests that nascent chiasma sites with separate SYCP3 axes are stabilized by SC central elements . Moreover , given that crossovers are almost completely abolished by Sycp1 mutation [17] ( as shown by the absence of both crossover-specific MLH1 foci and chiasmata ) , the observed chiasma-like structures are forming independently of interhomolog crossing over . In addition , in Scyp1−/− diplotene cells , we repeatedly observed chromosomes in which the non-centromeric ends of the homolog axes had fused to form contiguous terminal loops ( Figure 5D ) ; 15 . 2% of homolog pairs ( 14/89 ) had such structures . Analogous terminal loops were never observed in wild-type diplotene-stage nuclei imaged by SIM . Both the chiasma-like structures and terminal fusions detected in Sycp1−/− spermatocytes suggest a tendency for unregulated interactions between homolog axes in the absence of SC central element . Several characteristics of SC-associated centromeres in diplonema were also refined by SIM analysis ( Figure 6 ) . First , the well-characterized accumulations of SYCP3 at the centromeric termini [47] form paddle-like structures that can be more than three times broader than the homolog axes ( Figure 6A and 6C , highlighted by an arrowhead , and 6D ) . Second , 7% ( 12/172 ) of these terminal SYCP3 structures show a dual morphology suggestive of sister-chromatid individualization; in fact , clear examples of associated centromeres with split sister-axes were observed ( Figure 6A and 6C , highlighted by an arrowhead , and 6E ) . Third , dissociated or even widely separated centromeres can retain SYCP1 staining ( Figure 6F ) . This observation raises the possibility that final desynapsis of centromeres may not occur by simple dissociation of central element proteins from the axes , but by separation of central element transverse filaments ( comprising SYCP1 and other proteins ) that connect homolog axes via a head-to-head configuration of overlapping homodimers [6] . Alternatively , centromeric connections may have become sufficiently weak that they are mechanically disrupted during the spreading procedure . Fourth , we observed several examples of ostensibly achiasmate homologs , without an internal SYCP3 connection , that remain connected by synapsed centromeres ( 4/74 homolog pairs; Figure 6G ) . Whether these homologs are truly achiasmate and their ultimate segregation fate remains unclear , but their detection is consistent with our observation that persistent centromere synapsis occurs independently of crossing-over ( above ) . The possibility that persistent centromere synapsis facilitates the segregation of achiasmate chromosomes is also raised by this observation . Finally , the centromeric ends of the X and Y chromosomes do not show the dramatic SYCP3 accumulation and morphological changes seen for autosomes at this stage . Instead , a general accumulation of SYCP3 signal along the lengths of the X-Y pair is observed ( Figure 6C; highlight by and arrow ) . The analysis above implies that both the association of centromeres in pachynema and their continued connection during diplonema are mediated by SC central element . To test these inferences , we examined centromere association in spermatocytes from the synapsis-defective Sycp1−/− mutant ( Figure 7 ) . For individual pachytene-like Sycp1−/− spermatocyte nuclei , we determined the frequency of associated centromeres as well as the extent of homolog coalignment or “pseudo-synapsis” , which was defined as the fraction of homolog axes that were separated by ≤0 . 8 µm ( Figure 7A–7D; 0 . 8 µm was determined to be the maximum distance measured between regions of coaligned axes in Sycp1−/− nuclei ) . This analysis indicates that centromere association is not absolutely dependent on the SC central element ( Figure 7D ) . In fact , we observed several examples in which centromeric ends of homologous SYCP3 axes appear fused with one another to form a contiguous loop , even though adjacent regions are clearly separated ( e . g . Figure 7B and 7C ) . These terminal fusions are distinct from those observed in diplotene-like nuclei , which involve the non-centromeric chromosome ends ( above and Figure 5D ) . Overall , however , centromere pairing in Sycp1−/− spermatocytes never reaches wild-type levels and centromere regions remain the last to pair . Even in nuclei with >70% pseudo-synapsis , ≤50% centromere pairing is observed . In contrast to the pachytene-like nuclei analyzed above , centromeres are not associated in diplotene-like Sycp1−/− nuclei even though homologs remain stably tethered by one or more axial association ( Figure 7E–7G; 0/89 homolog pairs analyzed by SIM had associated centromeres ) . Thus , persistent centromere association during diplonema is SYCP1 dependent . A number of possible functions can be imagined for centromeric SC fragments during diplonema . For example , continued synapsis of centromeres could resist or maybe even promote the splitting of sister-axes that is detected at this stage ( described above; Figure 6E ) ; or it could promote the reorganization of centromere regions , such as the accumulation of SYCP3 , modification of cohesion and assembly of kinetochore components; or persistent centromere synapsis could indirectly facilitate homolog biorientation , for example by helping establish connections between centromeric heterochromatin similar to those described in Drosophila [67] . To begin to explore these possibilities , SYCP3-stained diplotene-like nuclei from Sycp1−/− spermatocytes were imaged by SIM ( Figure 7H; also see Figure 5 ) . Thickening of the SYCP3-stained centromeric termini was still clearly apparent in Sycp1−/− mutants ( Figure 7H ) . However , duality and splitting , indicative of sister-chromatid individualization , was exaggerated in Sycp1−/− cells ( Figure 7H ) , being observed at 19 . 2% ( 34/177 ) of centromeric termini compared to 7% ( 12/172 ) of wild-type ends ( P = 0 . 0007 , z-test ) . Moreover , Sycp1−/− centromeric termini had a more fragile and fractured appearance ( Figure 7H , arrows highlight gaps or fractures ) ; 13 . 6% ( 24/177 ) of centromeric ends had clear gaps or breaks , a morphology that was never observed in wild-type cells . Thus , diplotene-stage centromeric SYCP3 structures appear to be stabilized by continued synapsis . SYCP1 staining is lost and homologous centromeres desynapse in late diplonema ( above ) . Therefore , unlike Zip1-mediated coupling in budding yeast , persistent centromeres synapsis seems unlikely to play a direct role in promoting the stable biorientation of homologs on the meiosis I spindle , which doesn't assemble until diakinesis when the nuclear membrane breaks down . Although homologous centromeres desynapse during late diplonema , we noted that they often appear closely associated and oriented towards one another , even in the absence of chiasmata ( in Rnf212−/− spermatocytes ) . Moreover , we routinely detected inter-centromeric CREST-staining structures at this stage ( Figure 8 ) giving the impression of interconnecting chromatin bridges . However , distinct DAPI-staining bridges cannot be discerned at this stage because the chromatin is very diffuse ( not shown ) . As described previously , axial SYCP3 mostly disappears from diakinesis/metaphase-I chromosome axes to leave only faint interchromatid foci that define the chiasmata ( e . g . Figure 1N ) [47] . In contrast , centromeric SYCP3 becomes more abundant and remains closely associated with CREST-staining kinetochores [47] , [68] , [69] . Intriguingly , in diakinesis/metaphase I nuclei , we regularly detected closely apposed CREST signals associated with apparently contiguous , bi-lobed SYCP3 structures , or structures that are connected by thin SYCP3-staining strands ( Figure 9A–9C ) . The close apposition of the centromeres in these structures could , in theory , be caused by proximal chiasmata ( although crossover-specific MLH1 foci are rarely found close to centromeres ) . However , analysis of crossover-defective Rnf212−/− mutant nuclei indicates that they arise independently of chiasmata ( Figure 9B ) . On average , around two SYCP3-linked centromere pairs were observed in both wild-type and Rnf212−/− spermatocytes ( 1 . 9±1 . 8 SD , n = 13 and 2 . 2±1 . 4 SD , n = 18 ) . SIM analysis of diakinesis/metaphase-I nuclei supports the inference that these linked centromere-pairs are associated with contiguous SYCP3-staining structures ( Figure 9D–9L ) . In one example , discontinuous CREST staining also appears to bridge between the two homolog-kinetochores ( Figure 9G ) . Taken together , our observations support the possibility that the centromere regions of homologs can remain interconnected long after they have desynapsed during late diplonema .
In contrast to budding yeast and Drosophila , mouse centromeres appear to be refractory to SC formation and are the last sites to synapse . The same conclusions are drawn by Bisig et al . in the accompanying study [48] . The possibility that this is a general feature of mammalian meiosis is supported by the study of Hassold and colleagues , which showed that centromeres of human spermatocyte chromosomes constitute a barrier to the polymerization of SC [11] , [53] . In fact , the early synapsis of budding yeast and Drosophila centromeres may be more the exception than the rule , as late pericentric synapsis is typical of many organisms including fungi , plants and mammals [11] , [49]–[53] . We conclude that centromeres do not drive SC formation in mammals , but are in fact refractory to synapsis . This is true despite the fact that during zygonema , mouse centromeres cluster into a single large chromocenter , which might have been expected to facilitate centromere synapsis [70] . Indeed , recent studies in Drosophila females show that synapsis initiates within the chromocenter , indicating a fundamental difference with mammals [19] , [30] , [31] . We suggest that mammalian centromeres are synapsed only after SC polymerization switches from the initial homology-dependent , recombination-driven phase to the well-characterized ( but poorly understood ) homology-indifferent “synaptic adjustment” mode [6] , [71] . How and why mammalian centromeres resist synapsis remains unclear . Random ( non-homologous ) association of centromeric major satellite DNA within chromocenters could oppose forces that attempt to drive homologous pairing and synapsis of centromeres . Also , suppression of recombination initiation close to centromeres could limit not only local SC initiation , but also the extension of SC polymerization from adjacent sites . Another possibility is that chromatin and/or axial structures associated with centromeres are modified in ways that impede early synapsis . Notably , Roig et al . [51] showed that centromere synapsis is unusually dependent on the AAA+ ATPase , TRIP13 , which facilitates the removal of HORMA-domain proteins from synapsing homolog axes [72] . SC initiation sites have been correlated with crossing-over in a number of organisms . However , it remains unclear whether crossover-designation triggers SC formation , or SC initiation sites trigger crossing-over [2] , [6] , [11] , [73] . Under the latter scenario , absence of SC initiation between centromeres could function to suppress crossing-over within the multi-megabase , repetitive DNA elements that constitute mammalian centromeres [74] . This in turn will minimize the risk of chromosome rearrangements that could result from non-allelic centromere exchanges . Suppression of allelic crossing-over near centromeres will also help minimize the incidence of nondisjunction , which has been associated with such events in yeast , flies and humans [27] , [75] . By contrast , the tiny size of budding yeast centromeres ( 125 bp ) makes them highly unlikely to engage in non-allelic crossing-over . We also note that although Drosophila centromeres serve as SC initiation sites [30] , [31] , they may be protected from meiotic instability by the fact that SC initiation sites do not correlate with crossing-over in this organism . Moreover , homolog pairing and synapsis are not driven by recombination in Drosophila . Our analysis implies that DNA crossing-over and axis remodeling ( at least with respect to SYCP3 ) are temporally distinct aspects of chiasma formation . Analysis of Sycp1−/− spermatocytes suggests that this temporal separation may be mediated by the SC central element . Furthermore , the chiasma-like structures and terminal fusions observed in Sycp1−/− mutants suggest a novel role for SC central element in preventing the unregulated fusion and exchange of homolog axes . After diplonema , chromatin condenses and sister-chromatids individualize to become located on opposite sides of their cohesin axis . As bivalents further condense , chromatids also bend sharply at sites of crossing-over [14] . The requisite local flexibility appears to be reflected by two morphological features of crossover sites: relaxation of sister-chromatid cohesion and reduced chromatin condensation [6] . We suggest that SC fragments could help implement these features by triggering local loss of cohesin and/or differential loading of condensin ( the loading of which may be coupled to SC disassembly [76] ) . A role for crossovers in bivalent remodeling has been clearly demonstrated in C . elegans . In this organism , crossover sites trigger asymmetric loss of SC components and , subsequently , cohesion from bivalent arms [77] , [78] . This global remodeling of bivalents may be peculiar to organisms with holocentric chromosomes . In organisms with conventional centromeres , in which all arm cohesion is lost at anaphase I , we suggest that crossovers only trigger local changes in cohesion and chromatin condensation , as described above . SIM analysis has revealed a tendency for local separation of sister-chromatid axes at synpased centromeres during diplonema . Kleckner et al . have proposed that cycles or chromatin expansion and contraction drive such transient individualization of sister-chromatids in order to facilitate chromosome remodeling and installation of components required for subsequent stages [79] . The enhanced splitting of centromeric SYCP3 structures seen in Sycp1−/− mutants supports the idea that SC fragments retained at diplonema are part of a supporting framework that constrains and targets local expansion to help coordinate remodeling at centromere regions . The chromosomal passenger complex ( CPC ) regulates and orchestrates several key processes during chromosome segregation and cell division . These include sister-chromatid cohesion , kinetochore-microtubule attachments , spindle stability and cell division [80] . In addition , during meiosis the CPC regulates the timing of SC disassembly [76] , [81] , [82] . Cytological analyses of mouse spermatocytes have shown that CPC components , INCENP and Aurora-B , relocalize from centromeric heterochromatin to the inner centromere domain during diplonema , i . e . concurrent with the retention of SC at centromeres [83] , [84] . In addition , INCENP associates with the SC central element [84] . These observations raise the intriguing possibility that centromere-associated SC fragments in diplonema facilitate CPC relocalization and initial stages of kinetochore maturation . Centromere pairing in budding yeast requires PP4-dependent dephosphorylation of the SC component , Zip1 ( which is phosphorylated in response to DSB formation; [23] ) . In addition , the budding yeast SC central element component , Zip1 , can bind SUMO , which is a prominent modification at centromeric heterochromatin and kinetochores [56] , [85] . In Drosophila females , the CPC stabilizes SCs presumably by antagonizing kinases that promote SC disassembly ( see above [76] , [81] ) . Thus , the high concentration of CPC at spermatocyte centromeres could promote local resistance to SC disassembly . How crossovers signal local retention of SC remains mysterious . In rat , the crossover marker , CDK2 , remains at crossover sites until diplonema and could signal SC retention [86] . However , in male mice , CDK2 does not obviously persist at crossover sites beyond pachynema [87] . Persistent association of centromeres throughout diplonema appears to be a conserved feature of meiosis in many organisms , including budding yeast , Drosophila and mouse ( [4] , [21] , [30] , [32] , [50] , [52] , [62] and this study ) . In budding yeast , late centromere coupling promotes the correct , bipolar ( syntelic ) attachment of chiasmate bivalents to the spindle and thereby limits engagement of the spindle assembly checkpoint to correct misalignments . Coupling also serves as a backup mechanism for the disjunction of occasional achiasmate chromosomes [21] , [32] , [33] . The role of late centromere synapsis in Drosophila remains unclear , but association of centromeric heterochromatin is important for achiasmate segregation in this organism [88]–[90] . It seems unlikely that the persistent centromere synapsis observed in mouse is directly analogous to centromere coupling in budding yeast . Notably , centromere synapsis does not persist beyond diplonema so that a direct role in homolog biorientation and achiasmate disjunction is not envisioned . However , coupling could theoretically function indirectly in these processes by promoting centromere association , orientation and/or the organization of kinetochores prior to nuclear envelope breakdown and spindle assembly . The inter-centromeric CREST-staining bridges we detect in late-diplotene/early diakinesis cells are reminiscent of the heterochromatin threads that connect achiasmate ( and perhaps chiasmate ) chromosomes during meiosis in Drosophila females [67] . These structures are proposed to facilitate the congression of achiasmate chromosomes during prometaphase and promote their disjunction at anaphase I . The achiasmate X-Y disjunction systems found in some mammals appear to use specialized structures , derived from SC components , to tether the X and Y chromosomes [35]–[37] . The inter-centromeric CREST bridges and SYCP3 structures that we detect in diakinesis/metaphase-I spermatocytes might reflect the existence of related processes in mouse that can favor the biorientation of homologous centromeres and/or facilitate the disjunction of chromosomes that have failed to crossover .
All experiments conformed to relevant regulatory standards and were approved by the U . C Davis Institutional Animal Care and Use Committee . All mice were congenic with the C57BL/6J background . The Sycp1 and Spo11 knock-out lines were previously described [17] , [91] . Generation of the Rnf212 knock-out line will be described elsewhere ( Reynolds et al . , submitted ) . PCR genotyping of Rnf212 mice was performed using primers exon forward ( 5′-CGCTGGAATGAACGCAGGCGC-3′ ) , exon reverse ( 5′-CAGGGGAGTGAAGCCACGGTC-3′ ) , pH530 ( 5′-TCCATGGGCTTAAACCAGTGC-3′ ) , and VM3 ( 5′-GCGCATGCTCCAGACTGCCTTG-3′ ) . Primers , exon forward and exon reverse , generate a 290-bp fragment diagnostic of the Rnf212 wild-type allele; pH530 and VM3 detect the Rnf212 mutant allele as a 383-bp fragment . PCR conditions were 30 seconds at 94°C , 30 seconds at 60°C , and 1 minute at 72°C for 30 cycles . Testes were removed from 2–4 month old mice and processed for surface spreading as described [92] . Immunofluorescence staining was performed as described [93] using the following primary antibodies overnight at room temperature ( dilutions in parentheses ) : rabbit anti-SYCP3 ( sc-33195 Santa Cruz , 1∶300 ) ; mouse anti-SYCP3 ( sc-74568 Santa Cruz , 1∶200 ) ; mouse anti-rat SYCP1 monoclonal antibody [94] ( 1∶400 ) ; CREST antiserum ( generously provided by Shelby White , ARUP Laboratories; 1∶10000 ) ; mouse monoclonal anti-γH2AX ( 05-636 Millipore , 1∶500 ) , rabbit anti-mouse RAD21L ( a generously gift of K . Ishiguro and Y . Watanabe , University of Tokyo [45] ( 1∶200 ) ; guinea pig anti-SYCE1 ( 1∶2000 ) , guinea pig anti-SYCE2 ( 1∶400 ) and guinea pig anti-TEX12 ( 1∶200 ) [95] , [96] . Slides were subsequently incubated with the following goat secondary antibodies for 1 hour at 37°C: anti-rabbit 488 ( A11070 Molecular Probes , diluted 1∶10000 ) , anti-rabbit 568 ( A11036 Molecular Probes , diluted 1∶2000 ) , anti-human 488 ( A11013 Molecular Probes , 1∶2000 ) , anti-mouse 594 ( A11020 Molecular Probes , 1∶10000 ) , anti-human DyLight 649 ( 109-495-088 Jackson Labs , 1∶200 ) , and anti-guinea pig fluorescein isothiocyanate ( 106-096-006 FITC , Jackson Labs , 1∶200 ) . Coverslips were mounted with ProLong Gold antifade reagent ( Molecular Probes ) . Immunolabeled chromosome spreads were imaged using a Zeiss AxioPlan II microscope with 63× Plan Apochromat 1 . 4 objective and EXFO X-Cite metal halide light source . Images were captured by a Hamamatsu ORCA-ER CCD camera . Image processing and measurements were performed using Volocity ( Perkin Elmer ) and Photoshop ( Adobe ) software packages . Any pair of CREST foci that was ≤0 . 6 µm apart was classified as associated; convergent SYCP1 staining defined synapsed centromeres . To account for overlapping CREST foci , total numbers of CREST foci were counted for all nuclei . In nearly all cases , overlapping pairs of CREST foci could be discerned as larger , more intense , bi-lobed staining structures . Only nuclei for which all centromeres could be accounted for were used to determine levels of centromere association/synapsis . SIM analysis was performed using a Nikon N-SIM super-resolution microscope system and NIS-Elements 2 image processing software . | Gamete cells , such as sperm and eggs , form via the specialized cell division called meiosis . Essential and interdependent features of meiosis include the pairing , recombination , and segregation of maternal and paternal chromosomes . Chromosome pairing culminates with formation of synaptonemal complexes ( SCs ) , zipper-like structures that connect the structural cores or axes of homologous chromosomes . Although SC is known to be important for crossover recombination , details of its function remain enigmatic . In this study , we analyze mouse spermatocytes to investigate the interplay between SC , recombination , and centromeres ( the structures that direct chromosome segregation ) . We show that SC prevents unregulated interactions between chromosome axes . This function appears to be especially important at chromosome ends and at crossover sites where DNA exchange must be coordinated with structural exchange of chromosome axes . We also show that centromeres remain associated by short fragments of SC after general chromosome desynapsis has occurred . Furthermore , we detect a distinct type of inter-centromeric connection that persists even after centromeres desynapse . Such connections may facilitate the segregation of chromosomes that have failed to crossover . Together , our data provide new insights into the functions of SC and raise the possibility of a back-up chromosome segregation system in mammals analogous to those described in fruit flies and budding yeast . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"meiosis",
"animal",
"genetics",
"chromosome",
"structure",
"and",
"function",
"centromeres",
"gene",
"function",
"animal",
"models",
"telomeres",
"model",
"organisms",
"cell",
"division",
"dna",
"chromosomal",
"inheritance",
"chromosome",
"biology",
"biology",
"mouse",... | 2012 | Interplay between Synaptonemal Complex, Homologous Recombination, and Centromeres during Mammalian Meiosis |
In October 2010 , cholera importation in Haiti triggered an epidemic that rapidly proved to be the world's largest epidemic of the seventh cholera pandemic . To establish effective control and elimination policies , strategies rely on the analysis of cholera dynamics . In this report , we describe the spatio-temporal dynamics of cholera and the associated environmental factors . Cholera-associated morbidity and mortality data were prospectively collected at the commune level according to the World Health Organization standard definition . Attack and mortality rates were estimated and mapped to assess epidemic clusters and trends . The relationships between environmental factors were assessed at the commune level using multivariate analysis . The global attack and mortality rates were 488 . 9 cases/10 , 000 inhabitants and 6 . 24 deaths/10 , 000 inhabitants , respectively . Attack rates displayed a significantly high level of spatial heterogeneity ( varying from 64 . 7 to 3070 . 9 per 10 , 000 inhabitants ) , thereby suggesting disparate outbreak processes . The epidemic course exhibited two principal outbreaks . The first outbreak ( October 16 , 2010–January 30 , 2011 ) displayed a centrifugal spread of a damping wave that suddenly emerged from Mirebalais . The second outbreak began at the end of May 2011 , concomitant with the onset of the rainy season , and displayed a highly fragmented epidemic pattern . Environmental factors ( river and rice fields: p<0 . 003 ) played a role in disease dynamics exclusively during the early phases of the epidemic . Our findings demonstrate that the epidemic is still evolving , with a changing transmission pattern as time passes . Such an evolution could have hardly been anticipated , especially in a country struck by cholera for the first time . These results argue for the need for control measures involving intense efforts in rapid and exhaustive case tracking .
Cholera appeared in Haiti in October 2010 , probably for the first time in the country's history [1] . Importation of the vibrio [2] , [3] triggered an epidemic that rapidly proved to be the world's largest epidemic of the seventh cholera pandemic . In January 2012 , a cholera elimination objective was adopted by Haitian and Dominican authorities , the World Health Organization ( WHO ) , the United Nations International Children's Emergency Fund ( UNICEF ) , and many of their partners [4] . However , to establish effective control and elimination policies , strategies rely on the analysis of the dynamics of cholera dissemination . To bolster control policies , various mathematical models have been established [5]–[8] . They have provided varying results , thereby demonstrating the importance of mathematical assumptions and parameter estimations [9] , [10] . One model , issued in March 2011 , has predicted 779 , 000 cases and 11 , 000 deaths for November 2011 [5] . Another model has predicted that the principal peak of the epidemic would occur in April 2011 in several departments [6] . Other studies acknowledged that this peak occurred in December 2010 but predicted tens of thousands of cases for March and April 2011 [7] , [8] . Among the various causes of inaccurate predictions , all reports have used observed cases at the departmental scale , which hardly exhibit outbreak dynamics . Andews et al . [5] have not explicitly modeled spatial diffusion , while other authors [7] , [8] estimated parameters at the country level , assuming homogeneous dynamics between all locations [9] . In contrast , cholera epidemic curves provided during the year 2011 by the Haitian Ministry of Health and Population showed that the cholera evolution profiles greatly varied from that predicted by models , with a marked and unexpected reduction in cholera incidence during the first months of 2011 followed by a new outbreak in May . This observation reveal how crucial it is to generate a comprehensive description of cholera diffusion and monitor cases daily at a communal scale . In other areas affected by recurrent cholera epidemics , it has been shown that studying the spatio-temporal dynamics of cholera outbreaks helped to define more effective control procedures [11]–[15] . Currently , only one publication [16] describes data at this spatio-temporal scale; however , this report aimed only to understand the dynamics of the cholera epidemic during the initial weeks following the outbreak onset . Since this first phase of the epidemic , many other cases have been reported across Haiti with new peaks and possibly new patterns of transmission . Therefore , the objective of the present study was to describe the spatio-temporal dynamics of the first year of this cholera epidemic in Haiti , identify the principal factors explaining the heterogeneity , and assess the epidemic processes .
Cholera-associated morbidity and mortality data were prospectively and anonymously collected by the Departmental Health Directorates at the commune level . Departmental databases were sent to the Haitian Directorate of Health ( Laboratoire National de Santé Publique , LNSP ) , where data were gathered and analyzed after quality control . According to the WHO standard definition [17] , a probable cholera case was defined as profuse acute watery diarrhea with severe dehydration . Bacteriological confirmation of cases was recurrently performed at the LNSP for samples collected throughout the entire country using standard methods [18] . The in-hospital case fatality rate ( ICFR ) was defined as the ratio of cumulative number of deaths reported at Cholera Treatment Centers ( CTCs ) to cumulative number of hospitalized cases ( severe cases ) . The case fatality rate ( CFR ) was defined as the ratio of cumulative number of in-hospital deaths to cumulative number of cases ( reported at any health structure ) . As some communes lacked proper health facilities , some cholera patients had to travel to health structures of the nearest commune . To avoid overestimating case numbers in such locations and underestimating case numbers in surrounding areas , the data derived from these neighboring communes were aggregated after interviewing local health actors and analyzing local reports . In this study , we did not include personal medical data but included the number of incident cases anonymously reported at each health facility . This study was approved by the Haitian Ministry of Public Health and Population ( Ministère de la Santé Publique et de la Population ) . First , the mapping of global attack rates , mortality rates , ICFRs and CFRs observed between October 16 , 2010 and October 15 , 2011 was performed to assess the spatial distribution of the epidemic . Spatial autocorrelation was estimated using Moran's I statistic for areal data [19] . Second , temporal observations for the entire country were assessed to define epidemic phases and trends . Phases were specified using main slope changes in time series after mobile average ( MA ) smoothing ( order two ) . The accuracy of this phase specification was then assessed by using sensitive analysis of the MA order , concordance with the wavelet analysis ( see above ) , and the field expertise of the Haitian epidemiologists . For each epidemic phase , communal daily incidence rates ( DIRs ) were mapped , and spatial clustering was assessed using Kulldorff statistic [20] . To detect high-risk spatial clusters of cases , this algorithm moves a circular ( or elliptic ) scanning window over the study region , centered on each communal centroid with a radius ranging from 1% to 50% of the population at risk . This algorithm compares observed and expected case numbers inside and outside each window and estimates risk ratios based on the Poisson distribution . Using circular scanning windows , cluster significance ( p-value ) was calculated with a likelihood ratio test using the Monte Carlo approach with 999 random simulations under the null hypothesis of no clustering [19] , [21] . Communal epidemic profiles of the different epidemic phases were compared and classified using hierarchical cluster analysis ( HCA ) based on Euclidean distance [22] , and profile classes were then mapped . HCA is an unsupervised classification method that groups similar observations ( the epidemiological curves for each commune ) into classes depending on a similarity criterion ( the daily case numbers recorded for each commune ) . Furthermore , to address the impact of population immunity , we assessed the impact of the accumulation of cases during the second outbreak . We compared the influence of cumulative incidences ( aggregating phases 1 to 4 ) with the incidences observed during the second epidemic ( phases 5 to 6 ) at the commune level using the Spearman correlation coefficient . Third , to assess the environmental factors associated with outbreak spread , cases and rainfall time series at the country level were analyzed using wavelet spectrum analysis . By reducing the noise and capturing the local behavior of non-stationary time series [23] , this approach detect underlying phenomena [24] , [25] , such as periodic variations , regime shifts or sudden perturbations and jumps . This method provides a multiscale analysis extracting the main evolution and trends of time series at different temporal scales and has been previously utilized to study cholera outbreaks [26] , [27] . The relationship between cases and rainfall time series was assessed via cross-spectrum analysis [28] . Daily accumulated rainfall data were obtained from NASA Goddard Earth Sciences . These observations ( TMPA-RT 3B42RT ) were derived from the Tropical Rainfall Measuring Mission ( see http://disc . sci . gsfc . nasa . gov/giovanni/overview/index . html for details ) . For each epidemic phase at the commune level , we also examined the relationship between cases and the following land cover surface factors: plains , mountains and hills , urban zones , rice fields , length of perennial rivers ( 10 km ) , area ( km2 ) , and number of watersheds , which were obtained from the MULTI-MENACES-HA team report [29] . These environmental factors were assessed via multivariate analysis using the Generalized Additive Model ( GAM ) derived from linear regression models [30] , [31] . Standardized incidence ratios ( SIRs ) were estimated using log-transformed population density ( as an offset variable ) and were adjusted on the spatial distribution of communes modeled by thin plate splines following Wood's approach [30] . Because of the over-dispersion of cholera incidences , several models of the Negative Binomial and the Poisson families [32] were first graphically verified to meet the conditions of use and then compared using the Generalized Cross-Validation ( GCV ) score and the Un-Biased Risk Estimator ( UBRE ) score [30] . The stepwise selection of variables was performed using the GCV and UBRE scores . The explained deviance was also verified for model goodness-of-fit assessment . The SIRs and corresponding 95% confidence intervals ( 95CI ) were estimated using the final selected model and tested . Spatial cluster analyses were performed using SaTScan® v8 . 2 . 1 ( Martin Kulldorff , Harvard Medical School , Boston , MA , USA and Information Management Services Inc , Silver Spring , MD , USA ) . Wavelet spectrum analyses were performed using Matlab® v7 . 1 ( The Mathworks Inc . , Natick , MA , USA ) . The other statistical analyses were performed using R® v2 . 13 . 0 ( The R Foundation for Statistical Computing , Vienna , Austria ) with mgcv package ( GAM modeling ) , the DCluster package ( spatial analysis ) , and the cluster package ( HCA ) . The p-values were compared with the probability threshold α = 0 . 05 . The maps were generated using Quantum-GIS® v1 . 7 . 3 ( Open Source Geospatial Foundation Project , Beaverton , OR , USA ) .
One year after October 16 , 2010 , 493 , 069 cases and 6 , 293 deaths associated with cholera had already been reported in Haiti . The global attack rate was 488 . 9 cases per 10 , 000 inhabitants , and the global mortality rate was 6 . 24/10 , 000 inhabitants . The global ICFR and CFR were 1 . 76% and 0 . 83% , respectively . During this first year , 852 of the 1 , 437 stool specimens collected in the ten departments of Haiti were positive for Vibrio cholerae O1 Biotype El Tor , serotype Ogawa . No switch to the Inaba serotype was observed until the second year of the epidemic . The mapping of yearly attack rates ( Figure 1a ) showed that communes were disparately affected , as the rates ranged from 3 , 070 . 9 cases/10 , 000 inhabitants in Mirebalais ( Department of Centre ) to 64 . 7 cases/10 , 000 inhabitants in the western tip of the north peninsula ( communes of Baie de Henne , Bombardopolis , Jean Rabel , and Mole St Nicolas ) . The Moran's I coefficient was particularly low ( I = 0 . 02 , p = 0 . 5 ) , thereby indicating no significant spatial autocorrelation and confirming the highly fragmented pattern at this scale . The mapping of mortality ( Figure 1b ) displayed high yearly mortality rates in the western tip of the south peninsula with 58 . 5 and 45 . 1 deaths/10 , 000 inhabitants in Chambellan and Pestel , respectively . The low spatial autocorrelation together with the high degree of spatial heterogeneity of incident cases showed that outbreak dynamics in Haiti varied from location to location . This fragmented spatial pattern drew attention to the need for separate analyzes at each phase of the outbreak , both at the country and local levels . At the country level ( Figure 2 ) , the epidemic course exhibited two principal outbreaks . Studying the slope changes , the time series were divided into two periods separated by the main peak of the cholera epidemic ( 12/16/2011 , 4 , 289 cases ) . The period preceding this peak was split into two parts separated by the nadir in cholera cases occurring on October 31 , 2010 ( 1 , 053 cases ) , which was observed just before the violent increase in cholera cases reported in early November . The period succeeding the principal peak was split into four parts . During the first part ( phase 3 ) , the attack rate dramatically decreased ( from 3 , 972 to 1 , 131 cases daily ) . Phase 3 ended on January 30 , 2011 and was followed by a lull period characterized by a reciprocation of small increases and decreases ( phase 4 ) , with an average of 835 daily cases ( standard deviation SD = 188 cases ) until May 22 , 2011 . On May 22 , the slope of the epidemic curve changed to a marked increase , thereby signalling the onset of a new epidemic wave and the beginning of phase 5 . After a high tray above 2 , 000 cases per day ( until 06/12/2011 ) , the final recorded decrease characterized phase 6 . The first principal outbreak started during a period with very little rainfall ( ∼2 mm/day during the last 15 days of October 2010 ) . Outbreak onset lasted from mid- to late October ( phase 1 ) and was associated with the introduction of Vibrio cholerae in Meille ( commune of Mirebalais , Department of Centre ) and the abrupt contamination of the Artibonite River [2] , [16] . During this first phase , 23 , 587 cases were reported ( DIR = 1 . 46 cases/10 , 000 inhabitants/day , 95CI[1 . 44–1 . 48] ) ( Figure 3 ) . Spatial cluster analysis displayed only one significant high-risk cluster centered at the Artibonite Valley , with a significantly elevated relative risk ( RR ) of 42 . 72 , 95CI[41 . 1–44 . 4] compared with the other regions of the country ( p = 0 . 001 ) , thereby confirming the link between cholera and proximity to the Artibonite River during the beginning of the epidemic . During the second phase ( November 1–December 15 , 2010 ) , cholera diffused out of the Artibonite Valley concomitant with Hurricane Tomas , and 119 , 347 cases were reported ( DIR = 2 . 63 cases/10 , 000 inhabitants/day [2 . 62–2 . 65] ) . Among the five significant high-risk clusters , the largest cluster encompassed a large portion of the country including Port-au-Prince but spared the South Peninsula ( RR = 3 . 49 [3 . 45–3 . 54] , p = 0 . 001 ) . The hierarchical cluster analysis ( HCA ) of these first two phases of the epidemic profiles ( Figure 4 ) identified the outbreak origin in Mirebalais ( Class A ) , where cases occurred primarily during the first month ( DIR = 19 . 81 cases/10 , 000 inhabitants/day [19 . 44–20 . 18] ) . The class B profile was primarily located in the low Artibonite valley , where the outbreak began a few days later ( DIR = 6 . 22 cases/10 , 000 inhabitants/day [6 . 18–6 . 26] ) . In this area , most cases occurred at the beginning of the wave , and then the number of daily cases decreased . Classes C and D ( DIRs = 3 . 41 cases/10 , 000 inhabitants/day [3 . 0–3 . 46] and 3 . 05 cases/10 , 000 inhabitants/day [3 . 01–3 . 09] , respectively ) displayed a smoother pattern after a delay of approximately one week . Finally , the communes of classes E and F ( DIRs = 1 . 25 cases/10 , 000 inhabitants/day [1 . 24–1 . 27] and 0 . 83 cases/10 , 000 inhabitants/day [0 . 82–0 . 85] , respectively ) were the last affected areas . Overall , the mapping of epidemic profile classes exhibited a centrifugal spread from the Artibonite Valley: distant communes displayed delayed outbreak onsets , lower daily incidence rates , and delayed and smaller outbreak peaks . With 104 , 784 reported cases ( DIR = 2 . 26 cases/10 , 000 inhabitants/day [2 . 25–2 . 27] ) from December 16 , 2010 to January 30 , 2011 , phase 3 was characterized by a marked decrease that was observed in all communes but was more marked in urban communes . Six significant high-risk spatial clusters were identified; the main cluster was centered at the mountains of the Department of Centre with a RR of 2 . 36 [2 . 32–2 . 39] ( p = 0 . 001 ) . The subsequent forth phase was a lull period ending on May 22 , 2011 with 93 , 474 reported cases ( DIR = 0 . 83 cases/10 , 000 inhabitants/day [0 . 82–0 . 83] ) . During this lull phase , four significant clusters of elevated incidence rates persisted . The main cluster was again localized to the mountains of the Department of Centre with a RR of 3 . 27 [3 . 32–3 . 32] ( p = 0 . 001 ) . The remaining clusters displayed particularly local and brief outbreaks . The second principal outbreak began at the end of May ( phase 5 ) , concomitant with the onset of the rainy season , which started late in 2011 and was associated with 35 , 356 cases ( DIR = 1 . 67 cases/10 , 000 inhabitants/day [1 . 65–1 . 69] ) . This outbreak peaked on June 12 . Five significant high incidence clusters were observed , the main cluster still remained localized to the mountains of the Department of Centre with a RR of 4 . 01 [3 . 92–4 . 1] ( p = 0 . 001 ) . The subsequent decrease ( phase 6 ) included 116 , 306 cases ( DIR = 0 . 92 cases/10 , 000 inhabitants/day [0 . 86–0 . 98] ) . Five significant high incidence clusters were identified; the main cluster encompassed approximately five departments ( North-East , North , Centre , Artibonite , and portion of the West department ) , with a RR of 2 . 81 [2 . 78–2 . 85] ( p = 0 . 001 ) . The remaining clusters were located at communes of the south peninsula with local outbreaks . The positive correlation ( 0 . 35 , p = 0 . 001 ) between the two principal outbreaks suggests that population immunity did not play a major role in the cholera epidemic dynamics during the first year . The effect of immunity during this period may be concealed by spatial aggregation of the data at the communal level , population movement during the first weeks of the outbreak , and most notably environmental or social intra-communal determinants . The patterns of these various phases were confirmed by spectral analysis of case time series ( Figure 5a ) , which highlights the elevated velocity and intensity of the first phase in the Artibonite valley ( phase 1 ) and the high ( but less abrupt ) intensity of phases 2 and 5 . Spectral analysis of rainfall series ( Figure 5b ) highlighted the importance of rainfall during both Hurricane Tomas in November 2010 and the 2011 rainy season that began in May , which were the only two heavy rainfall periods associated with incident cases based on cross-spectrum analysis ( Figure 5c ) . Local environmental factors were assessed by quantifying their association with the spread of cholera at teach phase ( Table 1 ) . For phase 1 , the results highlighted the role of the Artibonite River ( Standardized Incidence Ratio for each 10 km portion of perennial rivers , SIR = 2 . 28 [1 . 86–2 . 79] , p<0 . 001 ) and rice fields ( SIR = 16 . 7 [3 . 0–93 . 7] , p = 0 . 002 ) . Conversely , urban zones ( SIR = 0 . 034 [0 . 007–0 . 18] , p<0 . 001 ) and mountainous zones ( SIR = 0 . 113 [0 . 05–0 . 27] , p<0 . 001 ) displayed a protective role . During phases 2 , 4 and 6 , no specific environmental factor was associated with outbreak spread . During phase 3 , urban zones ( SIR = 0 . 68 [0 . 47–0 . 97] , p = 0 . 03 ) experienced a more rapid decrease in case numbers , thereby showing an apparent protective role . Other factors were no more significant during this phase . With the exception of phase 3 , the spatial distribution of communes remained significant ( p<0 . 006 ) , thereby showing that environmental factors did not fully explain the spatial clustering of cases during each phase .
With 493 , 069 cases after one year , the cholera epidemic in Haiti appear to be the largest ever recorded in a single country during the past 50 years . Although it began during the last trimester of 2010 , the cases reported in Haiti accounted for more than 56% of the total cholera burden in 2010 . Yearly attack rates were higher than 20% in several Haitian communes , such as Mirebalais ( 30 . 7% ) , L'Estere ( 29 . 2% ) , Grande Saline ( 22 . 1% ) , and Cabaret ( 26 . 6% ) . To make a comparison , the yearly attack rate during the 2008–2009 epidemic in Harare ( Zimbabwe ) was 1 . 29% , reaching a maximum in the Hopley suburb with 541 cholera cases per 5 , 994 inhabitants ( 9% ) [33] . Before the Haitian epidemic , the largest cholera epidemic ever recorded during the seventh pandemic was the 1991 epidemic in Peru , which accounted for approximately 300 , 000 cumulative cases during the first year [34] . However , the yearly attack rate ( approximately1 . 4% ) of the Peruvian epidemic was approximately 3 . 5 times lower than that of the epidemic in Haiti ( 4 . 9% ) . Due to its exceptional amplitude , the cholera epidemic in Haiti led to a large number of fatalities . Because of the difficulties of identifying all cases and deaths in remote rural areas , it is likely that the recorded 6 , 293 deaths represent only a portion of the actual cholera death toll . The analysis of epidemic profiles at different time phases reveals evidence of different spatio-temporal patterns . The first two phases of the epidemic ( October 16–December 15 , 2010 ) display a clear centrifugal expansion of cholera , with a damping wave centered at the location of the explosive outbreak onset in Mirebalais and the Artibonite Valley . Even if low rainfall had been recorded before early October , no heavy rain was associated with the outbreak onset ( phase 1 ) , and flooding cannot be incriminated . However , several environmental factors ( rice fields , plains , rural zones , and rivers ) were associated with a higher risk of contracting the disease during this early phase . These findings correlate with the results of previously published reports and studies that attribute the onset of the epidemic to massive contamination of the Artibonite River and downstream irrigation canals by an imported pathogenic strain of cholera [2] , [16] , [35] . Conversely , the particularly rapid diffusion of cholera out of the Artibonite Valley ( November - mid-December 2010 , phase 2 ) was not associated with any environmental factors but might be linked to other phenomena . Human-driven dissemination was favored by the massive contamination of the population living in the Artibonite Delta [16] , the lack of immunity among Haitian population , and deficiencies in water , sanitation , and health care systems [35] . The explosive spread of the disease overwhelmed the humanitarian response and the initial attempts to broadcast awareness and hygiene messages . People who fled from the Artibonite Delta to neighboring communes [36] also played an aggravating role in cholera diffusion , thereby favoring the spread of cholera even in remote rural areas . The violent nature of this outbreak spread may also have been promoted by Hurricane Tomas , which reached Haiti on November 5 with rapid flooding in some areas already affected by cholera , such as Gonaives . Finally , riots in Port-au-Prince following the first round of presidential elections in early December 2010 may have also reinforced this explosive epidemic . The relationship between rainfall and cholera spread in Haiti was attested by the association of phase 2 with Hurricane Tomas , the lull transmission period with the dry season ( phases 3 and 4 ) , and the second outbreak ( phase 5 ) with the heavy rainfall during late May2011 . This booster effect of rainfall on cholera outbreaks has been observed in many other countries [13] , [14] , [36]–[38] , where rainfall has caused latrine overflow or the washing up of waste with subsequent contamination of wells and surface waters . However , the relationship between rainfall and cholera likely involves other mechanisms , such as the seasonal modification of human water sources or human behavior such as rice culture activity . Overall , many phenomena affecting environment-to-human and human-to-human transmission may affect this relationship between rainfall and cholera outbreaks , which therefore should not be regarded in Haiti as in Bangladesh , where cholera onset has been associated with vibrio blooms in aquatic reservoirs [39] , [40] . During phases 3 , 4 and 5 , cholera incidence was poorly associated with environmental factors . Cholera attack rates decreased more rapidly in the main towns than in rural areas during phase 3 . Sequential identification of spatial clusters during the successive phases of the epidemic shows that mountainous rural areas located in the northern and eastern portions of the country likely functioned as a reservoir for cholera during the dry season until more favorable climatic factors triggered the second outbreak of late May 2011 . During a field assessment in April 2011 , we found that cholera persisted during the lull period in rural Haiti , circulating from one village to another , and provoking outbreaks linked with the transient local contamination of springs and streams . Due to the difficulty in reaching these mountainous remote areas , the fight against cholera was less efficacious than in towns and plains . Unlike observations made in Asia , where cholera outbreak patterns largely depend on human exposure to the aquatic reservoirs of V . cholerae [41] , or eastern Democratic Republic of the Congo , where lakes play an important role in outbreaks [13] , [15] , our results do not suggest any environmental persistence of cholera . However , this has to be confirmed with environmental sample studies . Currently , cholera presence in the environment has been reported in two cross-sectional studies [42] , [43] , but no environmental spatio-temporal monitoring system has been developed . In contrast , rice fields tended to be protective during the second outbreak ( phase 5 ) . This may be partly due to population immunity acquired during the initial phases of the epidemics , particularly in the Artibonite Delta , which was heavily stricken during the first phase of the epidemic . The protection was likely also due to the action of nongovernmental organizations ( NGOs ) and local actors as well as the reinforcement of a population sensitization program that was implemented in the Artibonite plain [44] . Overall , our findings clearly show that the epidemic is still evolving . Such diversity in transmission patterns could hardly have been anticipated , especially in a country struck by cholera for the first time , which highlights the need for comprehensive studies such as the current investigation . Therefore , we believe it is too early to predict the future pattern of this epidemic , and especially to affirm that cholera will become endemic in Haiti . Notably , the presence of estuaries in an area hit by cholera does not necessarily mean that V . cholerae will perennially settle in the brackish waters and that seasonal outbreaks will recurrently occur in the future . Madagascar , another island with deficient sanitation , a susceptible hydro-geologic environment , a widespread rice culture , political tension , and a lack of resources , was hit by successive cholera waves from 1999 to 2001 [45] . Since this time , the country has not experienced new outbreaks . Like Madagascar , Haiti may benefit from its insular position far from usual endemic foci . The current spatio-temporal analysis shows that dynamics of the cholera epidemic varied from location to location as time passed , following no clearly predictable scheme . Excluding the first phase , no recurrent environmental factor was implicated , except rainfall involved in the exacerbation of the epidemic . After the first phases of the outbreak , the absence of constant spatial clusters and the changing pattern of cholera distribution in Haiti argue for the need for control measures that should include intense efforts in rapid and exhaustive case tracking . | Cholera is the prototypical “waterborne” disease that can provoke deadly acute watery diarrhea epidemics in settings deprived of clean water and proper sanitation . In spite chronic deprivation , Haiti had been spared cholera for a century until the vibrio was imported in October 2010 , which triggered the largest national epidemic ever recorded . To better understand the progression of the epidemic and adapt control measures , we describe and analyze the spatio-temporal dynamics and underlying factors associated with the first year of this cholera epidemic in Haiti . Attack rates reached highly heterogeneous levels between communes ( from 64 . 7 to 3070 . 9 cases per 10 , 000 inhabitants ) , thereby suggesting disparate outbreak processes . While the first principal outbreak spread centrifugally like a damping wave that suddenly emerged from Mirebalais and Lower Artibonite , a second principal outbreak erupted at the end of May 2011 , concomitant with the rainy season , and displayed a highly fragmented epidemic pattern . Environmental factors , such as rivers and rice fields , appeared to play a role in disease dynamics exclusively during the beginning of the epidemic . The dynamics of the cholera epidemic varied from place to place as time passed , following no clearly predictable scheme . Therefore , cholera control measures in Haiti should include rapid and exhaustive case tracking . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"disease",
"mapping",
"infectious",
"diseases",
"cholera",
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"neglected",
"tropical",
"diseases",
"spatial",
"epidemiology"
] | 2013 | Spatio-Temporal Dynamics of Cholera during the First Year of the Epidemic in Haiti |
Few genetic differences between human populations conform to the classic model of positive selection , in which a newly arisen mutation rapidly approaches fixation in one lineage , suggesting that adaptation more commonly occurs via moderate changes in standing variation at many loci . Detecting and characterizing this type of complex selection requires integrating individually ambiguous signatures across genomically and geographically extensive data . Here , we develop a novel approach to test the hypothesis that selection has favored modest divergence at particular loci multiple times in independent human populations . We find an excess of SNPs showing non-neutral parallel divergence , enriched for genic and nonsynonymous polymorphisms in genes encompassing diverse and often disease related functions . Repeated parallel evolution in the same direction suggests common selective pressures in disparate habitats . We test our method with extensive coalescent simulations and show that it is robust to a wide range of demographic events . Our results demonstrate phylogenetically orthogonal patterns of local adaptation caused by subtle shifts at many widespread polymorphisms that likely underlie substantial phenotypic diversity .
Although the predominant population genetics model of adaptation assumes novel advantageous alleles sweep to fixation [1] , [2] , most putative examples of adaptive divergence between human populations lack the full signature of a classic hard sweep [2]–[10] . In fact , classic sweeps may have played a negligible role in the evolutionary changes that have occurred since the most recent common human ancestor [10] , prompting the question of whether the moderately large allele frequency differences observed among modern human populations at a small proportion of loci are indeed adaptive . These divergent loci may indicate more complex and subtle modes of selection , but it is often difficult to demonstrate with statistical confidence that they are not merely the tail end of a stochastic neutral distribution . However , if selection acts independently on the same loci in different geographical locations , data from multiple populations can be leveraged to provide strong evidence for non-neutral evolution . Such parallel adaptation among populations of the same species has been identified in sticklebacks [11] , [12] , whitefish [13] , Drosophila [14]–[16] , and other taxa [17] , [18] . The extent to which the same genes underlie repeated adaptive events is unclear , but a growing number of observations suggest that parallel evolution at the molecular level may be quite prevalent [19]–[21] , and from first principles it seems especially likely when selection acts on the same standing genetic variation in closely related populations [2] , [22] . Demonstration of parallel evolution among populations provides strong support for the hypothesis that repeated selection of the same alleles in distinct environments is an important mechanism of local adaptation . Studying this evolutionary process can also generate a list of candidate sites that likely have functional phenotypic consequences and provide insight into which environments present similar selective pressures . In this study , we develop a novel approach to test the hypothesis that parallel adaptive evolution has shaped extant patterns of human genomic variation . Our method evaluates a set of independent SNPs genotyped in populations that can be clustered into at least four groups forming two or more phylogenetically distinct , allopatric group pairs ( Figure 1 ) . The goal is to test whether the same SNPs show high divergence in phylogenetically independent contrasts [23] between pairs of groups more often than expected under neutrality . To this end , we calculate pairwise FST for all SNPs between all pairs of groups . For each pair of groups , we identify divergent SNPs that exceed the 95th percentile of FST values , and define parallel divergent SNPs as those that are divergent in two independent group pairs . A significant excess of parallel divergent SNPs is interpreted as evidence for parallel adaptive divergence . We demonstrate substantial parallel adaptation in human populations and we characterize the genomic and geographical patterns of the parallel divergent SNPs .
We applied our method to the Human Genome Diversity Project ( HGDP ) data , which consists of approximately 1 , 000 individuals genotyped for over 650 , 000 SNPs [24] . We only analyzed a subset of these data that met our assumptions of independence , chosen as follows . We combined 19 of the HGDP populations , consisting of 343 individuals , into six ecologically and genetically distinct groups ( Figure 1A ) . Fifteen divergence comparisons are possible in this dataset , where each divergence comparison contrasts FST values calculated in two phylogenetically independent pairs of groups . Although it is difficult to completely rule out cryptic gene flow , we carefully examined these groups for any evidence of admixture . Previous population structure analysis indicates that the particular samples that constitute these groups are distinct lineages [25] . Possible gene flow between Pygmies and West Africans is irrelevant to this analysis as these sister populations are never assigned to different groups in a divergence comparison . The only other detectable gene flow signature among these populations is low-level introgression from East Asia into Oceania [25]; therefore the three divergence comparisons that could potentially be confound by this introgression , all comparing East Asia–South America FST against FST between Oceania and another population , are evaluated with this caveat in mind . To supplement these prior population structure results , we confirmed that the established phylogenetic topology has the significantly highest likelihood ( Shimodaira and Hasegawa tests; p<0 . 001 ) . Furthermore , after removing divergent SNPs exceeding the 95th percentile , we regressed the ranks of FST values at remaining SNPs for each group pair against the corresponding ranks for other group pairs and found that non-divergent FST values are uncorrelated ( Bonferroni-corrected p>0 . 05 for all divergence comparisons ) . This result suggests that admixture has had a negligible effect on these populations , since high levels of gene flow would cause SNPs to show relatively similar FST values in multiple group pairs . To address potential confounding factors in our analysis , we only considered a subset of all HGDP SNPs to meet the assumptions of our neutral model ( Figure 1B ) . One confounding factor is variation among SNP frequencies in the common human ancestor population , since the expected divergence due to genetic drift depends on the ancestral minor allele frequency; therefore SNPs with skewed global frequency , a proxy for the ancestral frequency , were excluded . Another confounding factor is mutation rate; therefore only globally polymorphic SNPs showing variation within all group pairs were used , to maximize the probability that alleles are identical by descent and not new mutations . We identified 111 , 724 HGDP SNPs that were intermediate frequency and globally polymorphic . Within this set , linked SNPs are not independent and ascertainment bias may be more pronounced in some populations [26]; therefore only SNPs showing low linkage disequilibrium with each other and collectively similar site frequency spectra in all groups were analyzed together ( Figure S1 ) . We randomly selected 1000 subsets of SNPs ( mean = 26 , 864 SNPs; range = 25 , 371 to 27 , 770 SNPs ) meeting these criteria . Thus , the theoretically expected number of parallel divergent SNPs in a divergence comparison is 67 . 16 ( = 26 , 864*0 . 052; Text S1 ) . An ideal dataset for this analysis would employ full sequence data rather than SNPs and potentially larger sample sizes than the 28 to 116 individuals per population that we used . In the absence of such a dataset , it is unlikely that any sampling scheme will produce a truly independent set of SNPs or precise estimates of allele frequencies , but our method represents a way to eliminate the most egregious potential errors while still retaining a larger number of SNPs for analysis . In the HGDP dataset , fourteen of the fifteen divergence comparisons showed more parallel divergent SNPs than the expected value of 67 . 2 ( Figure 2 ) . The mean was 82 . 0 parallel divergent SNPs ( range = 58 to 96 ) , a 22% increase over the expected value . Eight divergence comparisons showed a mean of 81 . 0 or more parallel divergent SNPs , a significant excess according to our simulation results ( see below ) ( Table S1; Figure 2; Figure 3 ) . Four divergence comparisons showed a mean of 91 or more parallel divergent SNPs and therefore were significant even after a conservative Bonferroni correction ( Fisher's exact tests , p<0 . 05 ) . The highest levels of parallel divergence were observed when divergence between Pygmies and Europeans was compared to divergence between Oceanians and South Americans or East Asians . As the threshold for divergence was dropped to more stringent values below 5% , the excess of parallel divergent SNPs increased ( Figure 4 ) . There is no comparable excess of parallel conserved SNPs with extremely low divergence in multiple group pairs , suggesting that purifying selection is not the cause of correlations among lineages ( Figure 3 ) . The only divergence comparison showing fewer than expected parallel divergent SNPs compares Europe-Oceania FST to East Asia-South America FST . This is the only divergence comparison that does not include an African population , suggesting that similar selection pressures in African and non-African populations may be driving most of the observed patterns . In addition , as extensive migration between Asia and Oceania could lead to false positives in this divergence comparison , the relative dearth of parallel divergent SNPs suggests that such migration does not have a large effect on our results . Many SNPs fell into narrow regions that were represented in a large number of the 1000 independently generated subsets ( Table S2 ) . The mean percentages of both genic and nonsynonymous parallel divergent SNPs were higher than mean for the set of all SNPs in a replicate , a relative excess that increased as the threshold was dropped to more stringent values below 5% , suggesting that positive selection is driving parallel divergence ( Figure 4 ) . Genes overlapping parallel divergent SNPs were modestly enriched for diverse functional categories associated with various cell types including neurons , lymphocytes , cancer , and epithelium ( Table S1; Table S3; Figure 5 ) . Among the most extreme parallel divergent genes ( observed at a threshold of 0 . 5% ) were the skin keratinization gene ABCA12 [27] ( Figure 6 ) ; SH2B1 , which controls serum letpin levels and body weight [28]; GRM5 , a glutamate receptor associated with schizophrenia [29] and with pigmentation via the closely linked TYR [30]; ATP2A2 , which causes a neuropsychiatric/keratinization disorder [31]; F13A1 , a coagulation factor linked to numerous cardiovascular diseases and to Alzheimer's [32]; and IFIH1 , associated with antiviral defense , type 1 diabetes , and psoriasis [33]–[35] ( Figure S2 ) . The pleiotropic nature of many of these genes suggests that selection on one trait may have affected the evolution of other traits . For all divergence comparisons , we calculated the orientation skew , defined as follows . For any divergence comparison , there are two possible allele frequency orientations for a given SNP . In the first pair of groups , one allele is at relatively higher frequency in the first group compared to the second . In the second pair , this allele can either be at relatively higher frequency in the first group or the second , defining the two orientations . The skew is simply the frequency of the more common , or major , orientation for all parallel divergent SNPs , which should be near 0 . 5 under the neutral expectation that both orientations are equally probable ( expected skew is 0 . 55 for 67 parallel divergent SNPs , following the binomial distribution ) . Eleven divergence comparisons showed a significant skew in the mean number of parallel divergent SNPs displaying each orientation across all replicates ( major orientation frequency >60%; p<0 . 05 , one-sided comparison to simulation results; Figure 2; Table S1 ) . These results suggest similar selection pressures acting on disparate groups . For example , the most extreme skew ( major orientation frequency = 85% ) has Europeans diverge from Pygmies in the same direction as East Asians diverge from Oceanians , and there are a myriad of climatological , dietary , and disease variables that could exert similar selective pressures on the two temperate groups relative to the two tropical groups . In fact , four of the five divergence comparisons showing the most parallel divergence show this same general pattern , if Pygmies are considered interchangeable with West Africans and East Asians are considered interchangeable with South Americans ( Table S1; Figure 2; Figure 3 ) . This pattern suggests similar selective pressures on Africans and Oceanians relative to Europeans , East Asians , and South Americans . South Americans may carry alleles adapted to temperate climates due to their ancestral migration across Beringia , and they may have lacked adequate time and/or genetic variation to completely re-adapt to a tropical environment . One SNP that fits this hypothesis lies in DDB1 , which protects the skin from solar UV exposure [36] , and is one of the strongest examples of this parallel divergence pattern , with one allele fixed in South America , over 90% in Europe and East Asia , and less than 40% in African and Oceania . Another possible selective pressure driving this pattern may be parasites historically restricted to the Old World tropics; indeed , several immune-related genes like LRBA and LAT show this parallel divergence pattern even at thresholds well below 5% ( Table S2 ) . To calculate the probability that neutral evolution would generate the observed number of parallel divergent SNPs , we performed nine sets of coalescent simulations , each comprised of 100 full SNP datasets equivalent to our empirical data , under various demographic scenarios including bottlenecks , migration , and growth ( Table S4 ) . Some simulations used realistic sets of parameters while others were not realistic for these human populations but tested the effects of diverse demographic events . For each simulated divergence comparison , we calculated the number of parallel divergent SNPs and the orientation skew . Our primary model ( standard ) employed realistic , previously calibrated demographic parameters [37] featuring recent growth in all populations , no migration among populations , and bottlenecks in all populations , including several population sizes under 1000 individuals over 50 generations . Under standard , the mean number of parallel divergent SNPs among all simulated divergence comparisons was 67 . 22 ( standard deviation among 1500 divergence comparisons = 7 . 85; standard deviation among 100 dataset means = 2 . 37; range of dataset means = 61 . 27 to 72 . 28 ) , nearly identical to the theoretical expectation of 67 . 16 ( standard deviation = 7 . 79; standard deviation of means = 2 . 01 ) ( Figure 2; Figure 7; Text S1; Figure S3 ) . Out of the 1 , 500 standard simulated divergence comparisons , replicated 10 times each , a mean of 81 . 0 or more parallel divergent SNPs was observed 2 . 47% of the time , suggesting that empirical divergence comparisons with at least 81 parallel divergent SNPs exceed the 95% confidence interval . A Bonferroni-corrected significant excess of parallel divergent SNPs ( 91 or more , theoretically expected 5% of the time if divergence comparisons are independent ) was never observed in standard ( p<0 . 01 ) . The mean orientation skew was 0 . 54 , and 4 . 7% of divergence comparisons showed an orientation skew greater than 0 . 60 . Eight variations on standard revealed that our method is in general robust to demographic perturbations , but that extreme demographic events can cause slight excesses of false positives , defined here as values at least 10% greater than the theoretical expectation ( Table S4; Figure 7 ) . Eliminating bottlenecks in some or all populations does not cause an excess of false positives , nor does moderate introgression between populations of up to 10 migrants per generation . False positives are produced by extreme bottlenecks of 150 individuals over 50 generations ( 3tightestbottle ) , extreme migration of 100 migrants per generation ( OC-EAmig100 ) , or low long-term effective size ( constantlow ) . However , no simulation model produced enough false positives to match the empirical results , as mean simulated values for both parallel divergent SNP counts and orientation skew were universally significantly lower than the empirically observed mean results ( Figure 7 ) . In the results from two models implemented to represent unrealistically extreme demographic events , 3tightestbottle and OC-EAmig100 , a minority of individual simulated divergence comparisons ( 2 or 3 , respectively , out of 15 ) were comparable to our most significant empirical divergence comparisons ( >90 parallel divergent SNPs and/or >0 . 70 orientation skew ) in more than 5% of simulated datasets ( Table S4 ) . Presumably , bottlenecks or migration events would have to be even more extreme or prevalent than in these simulations in order to be solely responsible for the mean empirical values across all divergence comparisons . Thus , demographic events are unlikely to be the cause of our empirical results .
The geographic distribution of human genetic diversity shows a pattern of parallel divergence , such that the same variants have undergone exceptionally high divergence repeatedly ( Figure 2; Table S2 ) . Both the counts of parallel divergent SNPs and the orientation skews are higher than expected under a neutral framework , which suggests that spatially varying selective pressures are partially responsible . Because observed parallel divergent SNPs did not exceed the expected value at a 5% threshold by more than twofold , the majority of parallel divergent SNPs are probably neutral false positives and not adaptive variants . The relative excess of parallel divergent SNPs increases as the threshold is lowered ( Figure 4 ) , so candidates at lower thresholds are more likely to be real adaptive variants , but even at these lower thresholds the evidence for selection on any given SNP reported here ( Table S2 ) is relatively weak in the absence of additional experimental data . Rather than pinpointing specific candidates with high confidence , the main strength of our results is instead to emphasize the broad genomic trend of parallel adaptive divergence . Our simulations indicate that particular demographic models can cause patterns of parallel divergence similar to those that could be caused by selection . Although this caveat means that our results must be interpreted with caution , demography alone is unlikely to responsible for the empirically observed patterns . None of our models produced mean counts of parallel divergent SNPs or mean orientation skews as high as we observed in the empirical data ( Figure 7 ) , and those that did produce unusually high counts of parallel divergent SNPs represented very extreme demographic events ( Table S4; Figure 7 ) . Specifically , we only observed an increase in false positives if gene flow occurred on the order of one hundred migrants per generation ( OC-EAmig100 ) , if bottlenecks occurred with an effective size of 150 individuals over 50 generations in multiple populations ( 3tightestbottle ) , or if effective size was consistently low with no growth ( constantlow ) . None of these three models is particularly realistic for the human populations examined here . While previous analysis suggests that there has been a small amount of admixture between East Asia and Oceania , which could potentially confound three of our divergence comparisons [25] , it is unlikely to have been as high as one hundred migrants per generation [38] . Migration between Pygmies and West Africans , or between East Asians and South Americans , would not confound our assumptions of population independence because these populations never appear in different group pairs within the same divergence comparison . Other potential instances of intercontinental gene flow , such as introgression from Europe to South America , do not appear to have contributed to the ancestry of these particular samples , which show negligible evidence of such admixture [25] . Bottlenecks have certainly played an important role in human history , and their magnitude is difficult to estimate with precision , but multiple bottlenecks as extreme as those simulated in 3tightestbottle would have been unlikely . For example , the Pygmy bottleneck effective size is estimated to have been from ∼500 to several thousand individuals [39] , the effective female population size during the colonization of America from Asia has been estimated as ∼1000 individuals [40] , and although Oceania may have been founded by very few individuals [38] , its cumulative effective size as suggested by modern genetic diversity has been slightly higher than for American populations [25] . As recent human population growth is readily apparent , the constant low effective size simulated by constantlow is overly simplistic . Finally , patterns such as an excess of genic and nonsynonymous SNPs ( Figure 4 ) cannot be attributed to demography and imply a prominent contribution by natural selection . Thus , even with the demographic caveats , the highly significant signal of parallel divergence across multiple group pairs suggests that parallel adaptation is an important feature of at least some of these SNPs in some of these populations . The patterns of parallel adaptive divergence we identified do not reflect classic selective sweeps and therefore suggest more complex modes of selection have shaped human genomic diversity [10] . Our results are consistent with fluctuations in the allele frequencies of standing variation , as in soft sweeps [2] , [22] , although alleles have rarely actually swept to fixation . The tempered changes at most loci suggest that they encode quantitative polygenic traits that have reached new optima [2] , that fitness landscapes fluctuate too rapidly , that the sweeps are still ongoing , and/or that gene flow prevents fixation . Our approach is based on the hypothesis that adaptive variants , or variants closely linked to them , were present in the ancestral human population . Parallel adaptation may also occur via independent adaptive mutations [5] , [41] . It is also conceivable that a favorable mutation arising in one population could spread via adaptive introgression to other populations even if gene flow were too low to be detectable at neutral markers . Our method does not preclude the possibility of such newly arisen alleles , although they are less likely to be globally polymorphic . These scenarios still require similar selective pressures acting in multiple distinct populations and thus represent alternate forms of parallel adaptive divergence . Our method complements existing strategies for detecting intraspecies non-neutral divergence . Most previous studies have focused on identifying loci that differentiate individuals on one continent from all others [4] , [6] , [8] , and thus contribute to the same patterns of population structure generated by neutral processes [2] , [7] . While these unique dramatic adaptive events have undoubtedly been important in human evolution , we have shown that phylogenetically orthogonal patterns are also a major component of geographically varying selection . One promising approach for detecting local adaptation is to compare allele frequencies to environmental variables , while controlling for population structure [9] . Our method differs in that we allow for parallel divergence among any group pair regardless of any obvious environmental similarities , and thus we can detect the effects of more cryptic selection pressures . We anticipate that both approaches will be fruitful in uncovering fundamental patterns of local adaptation at the molecular level . Finally , in contrast many other genomic scans for selection [6] , a strength of our method is that we do not merely identify the most extreme outliers; rather , we test whether outliers showing parallel divergence are significantly more frequent than expected under neutrality . Future studies on parallel divergence could infer the haplotype backgrounds of selected SNPs via sequencing or denser SNP genotyping , in order to estimate the lengths of the chromosomal regions affected by selection , to further pursue evidence of cryptic gene flow , and to pinpoint causal adaptive SNPs . In summary , we have demonstrated a statistically significant excess of parallel divergent SNPs in a set of human populations , relative to both the theoretical expectation under neutrality and the values observed in neutral simulations . Although it is difficult to completely rule out the effects of demography on genomic patterns , our simulations and the inferred histories of these populations indicate that a non-adaptive explanation is unlikely . Thus , our results provide statistical support for a major feature of the human evolutionary process: that the same genes are selected independently in multiple environments . Feasible adaptive solutions to selective pressures are therefore limited and are reused in separate lineages [20] , [42] . Our approach may gain additional power and lead to new insights with the coming availability of full sequence data from numerous human populations , as well as data from non-human species .
We united HGDP populations into six phylogenetically independent groups of at least twenty-five unrelated individuals based on climatic and subsistence designations [9] and population substructure analysis [25]: Pygmy tropical hunter-gather ( PY; Biaka and Mbuti ) , West African tropical horticultural ( WA; Mandenka and Yoruba ) , European temperate agricultural ( EU; French , Basque , North Italian , Orcadian , Sardinian , Tuscan ) , East Asian temperate agricultural ( EA; Han , Japanese , Miaozu , Tujia ) , Oceanian tropical horticultural ( OC; Papuan and Melenesian ) , and South American tropical horticultural ( SA; Colombian , Karitiana , Surui ) . The remaining HGDP populations were excluded either because they showed evidence of admixture between disparate sections of the phylogeny ( e . g . Middle Eastern and Central Asian populations ) , or because they were ecologically or genetically distant from the other populations and the sample size was insufficient for a new group ( e . g . San , Pima , Tu ) . We used an adjusted autosomal SNP dataset for which SNPs with missing data were either abandoned or had missing values estimated based on Hardy-Weinberg equilibrium [43] . We conducted all analyses using only SNPs showing polymorphism in all eleven pairs of analyzed groups ( “globally polymorphic” ) and showing a global minor allele frequency greater than 0 . 4 ( “intermediate frequency” ) . In order to confirm the previously published phylogeny [24] , we used the contml package in PHYLIP [44] to determine the evolutionary relationships among our six groups based on allele frequencies at globally polymorphic , intermediate-frequency SNPs . We tested whether the topology with the highest likelihood was significantly better than alternate topologies using Shimodaira and Hasegawa tests [45] with α set at 0 . 001 . Furthermore , for each divergence comparison , we tested whether FST values were correlated among SNPs that were not divergent; such a correlation would suggest that the pairs were not independent due to gene flow . After ranking each SNP in each divergence comparison according to FST values , we removed all divergent SNPs ( exceeding the 95th FST percentile ) in order to eliminate SNPs under divergent selection . We then regressed the remaining ranks for each pair against each other . We tested whether any divergence comparison showed a significant correlation after a Bonferroni correction ( α = 0 . 05/15 = 0 . 0033 ) . For the purpose of calculating linkage disequilibrium in large , panmictic demes , we chose the one population from each group with the largest sample size ( Biaka , Mandenka , Sardinian , Karitiana , Papuan , and Han ) . In each population , we calculated composite pairwise linkage disequilibrium ( D ) among all globally polymorphic , intermediate frequency SNPs within 10 Mb of each other . We then calculated global D as the average among these populations . For each of 1000 replicates , we randomly selected a combination of globally polymorphic , intermediate frequency SNPs showing pairwise D less than 0 . 1 ( “unlinked” ) . We did this by randomly selecting unlinked SNPs one by one , each time recalculating the site frequency spectra , defined as the count of SNPs with a minor allele frequency in each of ten equal bins ( intervals of 5% ) , in each group . In order to minimize the population-specific effects of ascertainment and demography , we did not allow the size of any of the ten bins in the site frequency spectrum in any group to deviate from the corresponding bin in another group by more than 5% , unless the difference was under 100; thus , in our final random sample all groups had very similar site frequency spectra ( “similar-spectrum”; Figure S1 ) . For each pair of groups , we calculated pairwise FST [46] between the two groups at all chosen SNPs and we ranked these FST values . We defined a SNP to be divergent for a particular group pair if its rank exceeded the designated threshold ( top 5% for most analyses ) . If a SNP was divergent between two phylogenetically independent pairs of groups ( a divergence comparison ) , we considered it to be a parallel divergent SNP . For each divergence comparison , we designated all SNPs as either parallel divergent , divergent in only one pair of groups , or not divergent , and tested the significance of these categories using both Fisher's exact test and comparisons to our simulated results ( Text S1 ) . We calculated the mean number per replicate of genic SNPs , nonsynonymous SNPs , and SNPs in genes associated with Gene Ontology ( GO ) or KEGG pathway terms [47] . We tested whether these categories were enriched for parallel divergent SNPs by using the random sample of globally polymorphic , intermediate-frequency , unlinked , similar-spectrum SNPs as the background , and the parallel divergent SNPs themselves , rather than the genes or regions they overlap , as the test group; this approach should preclude any effect of variation in SNPs per gene or in the probability that a SNP is included for analysis . We tested for a bias in allele frequency orientation , a binary variable defined by which groups in each divergence comparison have relatively similar allele frequencies at parallel divergent SNPs . In order to examine whether observed orientations strayed from the neutral expectation of equal probability for both orientations , we calculated the mean number of parallel divergent SNPs with each orientation in each divergence comparison , across all replicates . We assessed the deviation from the binomial expectation using both Fisher's exact test and comparisons to our simulated results . For each of nine demographic models , we used ms [48] to generate 100 independent coalescent-simulated datasets equivalent in size to the empirical HGDP dataset . Each simulation consisted of at least 12 , 000 unlinked regions representing 100 kb of DNA each , with a per-site mutation rate of 2×10−9 and a per-site recombination rate between 0 and 11 . 25×10−8 , drawn from a distribution based on the empirical distribution of recombination rates in humans . Our standard model was a modified cosi demographic model calibrated using African , European , and Asian HapMap populations [37] , splitting the African and Asian populations into two ( representing PY and WA ) , and three ( representing EA , OC , and SA ) populations , respectively . Additional models were variations on this standard model ( Table S4 ) . We excluded all SNPs that were not globally polymorphic and intermediate frequency . For each of ten replicates for each of the 900 independent simulations , we randomly selected 26 , 864 SNPs to analyze , to match the empirical mean , controlling for LD and site frequency spectra in each random sample as we did with the empirical SNPs . Thus , we analyzed 9000 unique sets of 26 , 864 simulated SNPs using the same methods we used on the empirical dataset , including the analysis of fifteen distinct divergence comparisons , for a total of 135 , 000 simulated divergence comparisons . | Identifying regions of the human genome that differ among populations because of natural selection is both essential for understanding evolutionary history and a powerful method for finding functionally important variants that contribute to phenotypic diversity and disease . Adaptive events on timescales corresponding to the human diaspora may often manifest as relatively small changes in allele frequencies at numerous loci that are difficult to distinguish from stochastic changes due to genetic drift , rather than the more dramatic selective sweeps described by classic models of natural selection . In order to test whether a substantial proportion of interpopulation genetic differences are indeed adaptive , we identify loci that have undergone moderate allele frequency changes in multiple independent human lineages , and we test whether these parallel divergence events are more frequent than expected by chance . We report a significant excess of polymorphisms showing parallel divergence , especially within genes , a pattern that is best explained by geographically varying natural selection . Our results indicate that local adaptation in humans has occurred by subtle , repeated changes at particular genes that are likely to be associated with important morphological and physiological differences among human populations . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"population",
"genetics",
"parallel",
"evolution",
"population",
"biology",
"forms",
"of",
"evolution",
"genetic",
"polymorphism",
"evolutionary",
"modeling",
"biology",
"evolutionary",
"genetics",
"adaptation",
"natural",
"selection",
"genetics",
"computational",
"biology"... | 2011 | Parallel Adaptive Divergence among Geographically Diverse Human Populations |
Lymphatic filariasis ( LF ) , a morbid disease caused by the tissue-invasive nematodes Wuchereria bancrofti , Brugia malayi , and Brugia timori , affects millions of people worldwide . Global eradication efforts have significantly reduced worldwide prevalence , but complete elimination has been hampered by limitations of current anti-filarial drugs and the lack of a vaccine . The goal of this study was to evaluate B . malayi intestinal UDP-glucuronosyltransferase ( Bm-UGT ) as a potential therapeutic target . To evaluate whether Bm-UGT is essential for adult filarial worms , we inhibited its expression using siRNA . This resulted in a 75% knockdown of Bm-ugt mRNA for 6 days and almost complete suppression of detectable Bm-UGT by immunoblot . Reduction in Bm-UGT expression resulted in decreased worm motility for 6 days , 70% reduction in microfilaria release from adult worms , and significant reduction in adult worm metabolism as detected by MTT assays . Because prior allergic-sensitization to a filarial antigen would be a contraindication for its use as a vaccine candidate , we tested plasma from infected and endemic normal populations for Bm-UGT-specific IgE using a luciferase immunoprecipitation assay . All samples ( n = 35 ) tested negative . We then tested two commercially available medicines known to be broad inhibitors of UGTs , sulfinpyrazone and probenecid , for in vitro activity against B . malayi . There were marked macrofilaricidal effects at concentrations achievable in humans and very little effect on microfilariae . In addition , we observed that probenecid and sulfinpyrazone exhibit a synergistic macrofilaricidal effect when used in combination with albendazole . The results of this study demonstrate that Bm-UGT is an essential protein for adult worm survival . Lack of prior IgE sensitization in infected and endemic populations suggest it may be a feasible vaccine candidate . The finding that sulfinpyrazone and probenecid have in vitro effects against adult B . malayi worms suggests that these medications have promise as potential macrofilaricides in humans .
Lymphatic filariasis ( LF ) is a debilitating disease caused by the tissue-invasive nematodes Wuchereria bancrofti , Brugia malayi , and Brugia timori . Currently , there are ~ 70 million people infected worldwide and over a billion people at risk for infection [1] . Since 2000 , the Global Programme to Eliminate Lymphatic Filariasis has substantially reduced the number of people infected or at risk for infection [1] . However , it has become apparent that new strategies must be implemented in order to attain global eradication of LF [2 , 3] . Development of new therapeutics that target adult filarial worms would greatly enhance our ability to eliminate lymphatic filariasis . When given individually , the antifilarial drugs diethylcarbamazine ( DEC ) , ivermectin ( IVM ) , and albendazole are effective against the microfilaria ( Mf ) stage but exhibit little activity against adult filarial worms [4] . Use of all three medications together appears to have a macrofilaricidal effect [5] , but due to the adverse effects caused by their potent microfilaricidal activity DEC and ivermectin cannot be used for mass drug administration ( MDA ) in areas endemic for Loa loa or Onchocerca volvulus . Therefore , development of a short-course macrofilaricidal agent or a vaccine would be very valuable for eradication efforts . Unlike cestodes and trematodes , nematodes have a complete intestinal tract . Over the past 15 years , intestinal proteins of Necator americanus ( hookworm ) and Haemonchus contortus ( barber pole worms ) have been shown to be effective vaccine candidates in animal models [6–11] . Considering this work , our group performed a proteomic analysis of the intestine , body wall , and reproductive tract of adult B . malayi worms to potentially identify novel drug and vaccine targets for lymphatic filariasis [12] . We identified 396 proteins that were specific to the intestinal tract of the adult worms . Of these intestinal proteins , we selected a subset for evaluation as drug and vaccine candidates based on high homology with other filarial species , extracellular domains with accessibility to drugs and antibody , and predicted function . In this study , an adult B . malayi intestinal protein , UDP-glucuronosyltransferase ( Bm-UGT ) , was identified as a potential therapeutic target . The protein was predicted to have an enzymatic function that could be inhibited . Furthermore , structural analysis of Bm-UGT by InterPro revealed a large extracellular domain that could be targeted by therapeutics . We determined that this protein was essential for worm survival using small interfering RNA ( siRNA ) to knockdown expression . Importantly , we identified two FDA-approved commercially available UGT inhibitors that exhibit macrofilaricidal activity and display synergy with albendazole in vitro . Finally , we analyzed the antibody response against Bm-UGT in filarial patients and found that neither infected individuals nor endemic normals develop detectable levels of IgE against Bm-UGT , suggesting it would not induce allergic reactions if used in a vaccine .
Previously , we reported that Bm-UGT ( Bm17378 ) was a specific intestinal protein of B . malayi adult worms [12] . Sequence analyses indicated the presence of homologues in human filarial worms ( Brugia sp . , W . bancrofti , L . loa ) with significant homology ( >75% identity ) , and to a lesser extent ( ~35–40% identity ) in other nematodes such as Dirofilaria immitis , Haemonchus contortus , Ancylostoma sp . , Strongyloides sp . , Oesophagostomum dentatum and Toxocara canis . The most similar human proteins were UDP-glucuronosyltransferases as expected , but with low sequence identity <27% . Given the high predicted homology of Bm-UGT between B . malayi and D . immitis , we also evaluated orthologs in cat and dogs . Results of the sequence analyses revealed little homology to Bm-UGT . We then generated a phylogenetic tree by first aligning the Bm-UGT cDNA-derived peptide sequences using MUltiple Sequence Comparison by Log-Expectation ( MUSCLE ) and then creating a tree based on efficient maximum-likelihood estimation method by the LG model . As seen in Fig 1 , there is a high level of relatedness between Bm-UGT and several filarial orthologs , including other Brugia species , W . bancrofti , L . loa , and D . immitis . Interestingly , we could find no UGT ortholog in O . volvulus , and relatedness to the ortholog in Litomosoides sigmodontis , a common murine model of filarial infection , is low . Importantly , there is significant evolutionary distance between the Bm-UGT and orthologs in humans , cats , and dogs . Evaluation of data available from prior transcriptomic and proteomic studies of B . malayi demonstrates that Bm-UGT is not expressed in all the lifecycle stages ( Table 1 ) . A study by Li et al . shows that Bm-UGT transcript is only expressed in third stage larvae ( L3s ) and adult female and male worms [13] . In addition , an RNAseq study by Choi et al . on various lifecycle stages of B . malayi found that Bm-UGT was preferentially expressed during later larval stages . Consistent with these findings , Bm-UGT protein expression was found to be specific to these stages as well [14] . Predictive analysis using the InterPro database revealed that the protein contains a large luminally-expressed domain likely accessible to small molecules or ingested antibodies . Sequence analysis of the Bm-UGT ( Fig 2A ) indicates that residues 1–20 encode a signal peptide [15] followed by a two-domain UGT ( residues 21–278: N-terminal domain; 278–445: co-factor binding domain ) , a linker region , a transmembrane domain , and a short intra-cellular domain . The structure of the Bm-UGT was modeled using SWISS-MODEL with 43 template structures utilized . The final model was based on Protein Data Bank ( PDB ) : 5NLM ( the structure of the Polygonum tinctorium UGT ) with a sequence identity of 19 . 7% [PMID: 29309053] ( Fig 2B ) . UGTs add a glucuronic acid moiety to a substrate by transfer of the glucuronosyl group from uridine 5’-diphospho-glucuronic acid ( UDPGA ) . Further structure-based searches using the PDB identified the 2B7 UGT as a model for the co-factor binding domain , and based on the co-factor interacting residues , the Bm-UGT has a potential UDPGA-binding site ( Fig 2C and 2D ) . The enzyme nucleotide-sugar binding sites utilize a common structural scaffold; while specific interactions with the donor ligands vary between enzymes , analysis of the proposed binding site demonstrates it contains significant sequence homology to the active site of other UGTs . To evaluate whether siRNA is taken up by adult B . malayi worms , Cy3-conjugated Bm-UGT siRNA was added to the culture media of adult B . malayi worms for 24 hrs . Visualization of the adult worms shows clear uptake of siRNA throughout the intestinal tract ( Fig 3A and 3B ) . In contrast , imaging at the same exposure time reveals no apparent signal in the intestine of adult B . malayi cultured in media alone ( Fig 3C and 3D ) . In addition , we also confirmed that antibodies could access the lumen of the intestine via ingestion by the adult filaria ( S1 ) . Adult filaria worms were incubated with Cy3-labeled mouse IgG for 24 hours and then viewed under a fluorescent microscope . The labeled antibodies emitted a positive signal in the gut of the adult female worms while no signal was detected in worms incubated in media alone . We next evaluated reduction in target transcript and protein levels . We selected timepoints of 1 , 3 , and 6 days post-siRNA treatment based on a protein half-life of approximately 10 hrs for UDP-glucuronosyltransferases ( UGT ) [16] . After a 24-hr incubation with target-specific siRNA or scrambled siRNA , we compared Bm-ugt mRNA expression between the treated worms relative to the media control by RT-qPCR ( Fig 4A ) . Transcript expression was normalized employing the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase ( Bma-gapdh , Bm5699 ) . Bm-UGT siRNA treatment resulted in a 77 . 1% reduction in target mRNA compared to the controls 1-day post-siRNA treatment ( p = 0 . 0056 ) . Transcript knockdown was sustained throughout the experiment with a 76 . 2% decrease in Bm-ugt transcription 6 days post-siRNA incubation ( p = 0 . 0003 compared to controls ) . As expected , there was no significant difference in target transcription between the media control and scrambled siRNA groups . In addition , we did observe several worms ( n = 3 at day 2 post-siRNA incubation ) in the UGT siRNA group have no movement early into the experiment without any recovery throughout the course of the experiment . Bm-UGT knockdown was further substantiated by immunoblotting ( Fig 4B ) . Target protein expression was evaluated 24 hrs post-siRNA incubation using anti-Bm-UGT peptide antibodies . There was a robust reduction in UGT expression with the Bm-UGT siRNA treated worms compared to controls normalized to β-actin ( 68 . 1% reduction in Bm-UGT/β-actin , Fig 4C ) . After successfully demonstrating that siRNA reduces Bm-UGT expression , we evaluated for any resultant changes in worm motility , Mf release , and metabolism . Adult worm motility was scored on a scale from 0 to 4 , with 0 indicating no movement and 4 indicating active movement . At day 1 post-siRNA incubation , we observed a 77 . 1% reduction in motility with the Bm-UGT siRNA-treated group compared to the scrambled control ( p = 0 . 0006 , Fig 5A ) . This dramatic reduction in motility was maintained through day 6 ( 78 . 94% reduction , p = 0 . 0004 ) . We also observed a dramatic decrease in Mf release per adult worm per 24-hr period after Bm-UGT knockdown . At day 1 , there was a 62 . 5% reduction from the specific siRNA-treated worms compared to the scrambled siRNA-treated group ( p = . 0048 , Fig 5B ) . The greatest reduction in Mf release occurred at day 3 and was marked by a 95 . 3% difference between the Bm-UGT siRNA group and the scrambled control ( mean number of Mf release in 24h = 14 . 0 vs 294 . 8 , p = 0 . 0096 ) . Finally , we evaluated metabolism using a ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide ( MTT ) reduction assay . Decreased MTT reduction by the Bm-UGT siRNA treated B . malayi was observed at all three timepoints ( Fig 5C ) . The values observed were similar over the course of the experiment with day 6 post-siRNA incubation showing the greatest difference between the specific and scrambled siRNA groups ( 64 . 2% lower values in the Bm-UGT treated group , p = 0 . 0243 ) . The dramatic decrease observed in worm viability and fecundity after siRNA inhibition demonstrated that Bm-UGT is an essential protein for adult B . malayi survival . We next sought to evaluate the effects that UGT inhibitors had on adult worm survival . We tested two non-specific UGT inhibitors that act on multiple UGT isoforms for activity against B . malayi adult worms in vitro . Both of these agents , sulfinpyrazone and probenecid , are FDA-approved medications that are used to treat gout [17–20] . We incubated adult worms at various concentrations and found that both drugs killed filariae in vitro using worm motility as a metric for viability . For sulfinpyrazone , we observed a dose-response relationship for macrofilaricidal activity in vitro ( Fig 6A ) . The most rapid decline in adult worm motility occurred at 2500 μM . At this concentration , the area under the curve ( AUC ) was 3 . 6 and significantly different than the AUC of 31 . 6 for the control worms incubated with media alone ( p<0 . 0001 ) . Macrofilaricidal activity was also seen at 200 μM . The AUC at this concentration was 19 . 7 and significantly different ( p<0 . 0001 ) than the AUC for media control . For probenecid , we also observed a dose-response curve for macrofilaricidal activity in vitro ( Fig 6B ) . The greatest reduction in worm motility occurred at 5000 μM , which had an AUC of 8 . 1 ( p<0 . 0001 ) compared to the media control AUC of 24 . 2 . The lowest concentration that exhibited a significant effect on motility was 500 μM ( p = 0 . 0004 ) . Though worm motility at 250 μM was not significantly different over the course of the experiment , at day 7 there was a significant difference ( p = 0 . 0105 ) between the treatment group compared to the control . For both UGT inhibitors , there were doses ( >1000 μM sulfinpyrazone and 5000 μM for probenecid ) that resulted in worms exhibiting no movement early in the experiment . Drug treatment was stopped for these worms and there was no recovery throughout the course of the experiment . After testing the drugs on adult filariae , we tested for microfilaricidal effect . We observed modest microfilaricidal effects at the highest concentrations for both drugs ( Fig 6C and 6D ) . However , no clear microfilaricidal effect was demonstrated for probenecid at 500 μM and very little for sulfinpyrazone at 200 μM . While these drugs are FDA-approved , we wanted to determine whether the concentrations used were cytotoxic . Employing a lactate dehydrogenase ( LDH ) cytotoxicity assay with human embryonic kidney ( HEK ) cells , we did not detect any cytotoxicity at the concentrations that exhibited macrofilaricidal activity ( S1 Table ) . After observing the macrofilaricidal effect of sulfinpyrazone and probenecid , we decided to investigate whether synergy existed between these inhibitors and albendazole . We hypothesized that because UGTs are involved in drug detoxification [21] , inhibition of these enzymes may potentiate the effect of albendazole on filaria . In fact , one study showed that C . elegans treated with albendazole had upregulation of UGTs and metabolism of albendazole to albendazole-glucosides [22] . Because of this previous finding , we decided to incubate adult filaria with a sub-macrofilaricidal concentration of sulfinpyrazone ( 40 μM ) or probenecid ( 100 μM ) in combination with albendazole ( 10 μM ) . Mild macrofilaricidal activity was observed with the 10 μM of albendazole treatment . Treatment of adult filariae with sulfinpyrazone in combination with albendazole ( Fig 7A ) produced an AUC of 22 . 5 based on worm motility . This was significantly lower than the resultant AUCs from treatment with sulfinpyrazone ( 30 . 63 , p = 0 . 001 ) or albendazole ( 29 . 63 , p = 0 . 0007 ) . With the probenecid/albendazole combination ( Fig 7B ) , we observed a significant decrease in worm motility resulting in an AUC of 24 . 0 . This result was significantly different from a single treatment with either probenecid ( AUC = 30 . 88 , p = 0 . 0007 ) or albendazole ( AUC = 29 . 63 , p = 0 . 0022 ) . Further , no synergistic microfilaricidal effects were observed with albendazole and sub-optimal concentrations of sulfinpyrazone ( 40μM ) or probenecid ( 100 μM ) ( S2 Fig ) . A major obstacle for helminth vaccine development is the potential for individuals living in endemic countries to have pre-existing antigen-specific IgE and thus be at risk for developing allergic reactions when vaccinated with helminth antigens [23] . In this study we employed a luciferase immunoprecipitation system ( LIPS ) assay to determine whether individuals infected with filariae developed Bm-UGT-specific antibodies . Lysate containing Bm-UGT-luciferase fusion protein was incubated with serum from W . bancrofti infected individuals that were categorized as having asymptomatic microfilaremia ( n = 13 ) , chronic pathology ( lymphedema ) ( n = 9 ) , or tropical pulmonary eosinophilia ( n = 8 ) . All patients were untreated at the time of the blood draw . We also tested serum from individuals living in endemic areas that had no evidence of infection ( endemic normal , n = 5 ) . Healthy normal blood bank donor sera were used as negative controls ( n = 5 ) , while anti-Bm-UGT peptide antibodies raised in New Zealand rabbits served as positive controls . In addition , the naïve rabbit sera served as a negative control . The LIPS assay did not detect Bm-UGT-specific IgG ( Fig 8A ) or IgE ( Fig 8B ) in any of the serum samples from filaria infected or exposed patients . As expected , there was no specific IgG or IgE in serum from U . S . blood bank donors while the Bm-UGT peptide IgG antibodies recognized our fusion protein .
Infection studies conducted at FR3 and TRS were approved by their respective Animal Care and Use Committees , and protocols that enable receipt of filarial worms from FR3 and TRS for use at the Uniformed Services University of the Health Sciences ( USUHS ) were approved by the USUHS Animal Care and Use Committee . The generation of peptide antibodies by Genscript was on a protocol approved by the Genscript institutional animal care and use committee . Blood samples were obtained from patients and healthy volunteers who provided written consent under protocols approved by the NIAID’s Institutional Review board . All human subjects were adults . Female B . malayi adults used in this study were obtained from the NIH/NIAID Filariasis Research Reagent Resource Center ( FR3 ) and TRS Laboratories in Athens , Georgia , USA . The worms were cultured in Dulbecco’s Modified Eagle’s Medium ( Corning cellgro ) supplemented with 10% heat-inactivated fetal bovine serum ( Atlanta Biologicals ) , 100 units/mL of penicillin , 100 ug/mL of streptomycin , and 1% L-glutamine ( Sigma ) for 24 hrs at 37°C in 5% CO2 prior to siRNA treatment . Microfilariae were obtained from adult female worms cultured in vitro . Orthologs in other nematode species were identified in WormBase Parasite based on a BLAST query [24] against the Bm-UGT protein sequence ( Bm17378 ) . The following are the accession numbers of each ortholog as identified in WormBase Parasite: Brugia timori ( BTMF_0001026401 ) , Wuchereria bancrofti ( WBA_0000030501 ) , Brugia pahangi ( BPAG_0000208101 ) , Loa loa ( LOAG_03428 ) , Dirofilaria immitis ( nDi . 2 . 2 . 2 . t06727 ) , Litomosoides sigmodontis ( nLs . 2 . 1 . 2 . t00666-RA ) , Ancylostoma caninum ( ANCCAN_05977 ) , Anyclostoma duodenale ( ANCDUO_14383 ) , Dictyocaulus viviparous ( NDV . 1 . 0 . 1 . g111112 ) , Haemonchus contortus ( HCON_00121250 ) , Heligmosomoides polygyrus ( HPOL_0001615101 ) , Nippostrongylus brasiliensis ( NBR_0001252501 ) , Caenorhabditis elegans ( Y37E11AR ) , Strongyloides ratti ( SRAE_2000477000 ) , Strongyloides stercoralis ( SSTP_0001129400 ) , and Oesophagostomum dentatum ( OESDEN_03545 ) . Orthologs in selected mammals were identified in the National Center of Biotechnology Information ( NCBI ) databases based on a BLAST query against the Bm-UGT peptide sequence . The following are the orthologs selected for analyses: Homo sapiens ( NP_066307 ) , Canis lupus familiaris ( XP_005635657 ) , and Felis catus ( BAA2492 ) . The Bm-UGT sequence was initially analyzed for properties including signal peptide sequence , and potential transmembrane sequence using InterPro , SignalP4 . 1 ( PMID: 28451972 ) and TM servers [15] . Using the SWISS-MODEL homology modelling server [PMID: 29788355] , the iUGT sequence was used to search against the SWISS-MODEL template library using BLAST and HHBlits for structures that matched the target sequence . The model was visualized using PyMOL [25] and COOT [PMID: 20383002] , with residues 468–492 manually built using COOT . Based on sequence and structure identity , the co-factor binding domain was further analyzed , utilizing the inferred UDPGA-binding site of Bm-UGT mapped using the UGT 2B7 structure [PMID: 17442341] as a model and visualized using COOT [PMID: 20383002] . Using the BLOCK-iT™ RNAi Designer , we selected the top three Bm-UGT siRNA duplexes for gene silencing activity and specificity . The Bm-UGT siRNA and corresponding scrambled siRNA were synthesized by Life Technologies and purified by standard desalting methods . The 5’-3’ sequences of the Bm-UGT siRNA strands were as follows: For demonstration of siRNA uptake , adult female worms were incubated in 5 μM of 5’ Cy3-labeled Bm-UGT siRNA 3 ( Sigma Aldrich ) for 24 hrs . Adult worms incubated in media alone were used as a negative control . Both groups of worms were then stained with 10 μg/mL of DAPI ( Sigma-Aldrich ) in PBS . Fluorescent images were captured by a Nikon Eclipse E600 fluorescent microscope and converged using NIS-Elements software . For experiments to test ingestion of antibody , adult female worms were incubated with 100 μg of mouse Cy3-labeled IgG isotype control in 2 mL of culture media . The worms were imaged 24 hrs later using the TRITC filter on a Zeiss Axio Observer . A1 . siRNA inhibition of Bm-UGT in B . malayi adult female worms followed a protocol established by Aboobaker et al . with minor modifications [26] . siRNA inhibition in filarial worms has well-known variability and difficulty [27 , 28] . For this study , we analyzed data for experiments that received greater than 60% knockdown . For each timepoint , 5 adult female worms were soaked in an equal mixture of the Bm-UGT siRNAs at a final concentration of 5 μM in 850 μL of culture media in a 5000 MWCO Pur-A-Lyzer™ dialysis tube ( Sigma-Aldrich ) . This concentration of siRNA was shown in multiple studies to be sufficient at silencing gene expression [26 , 29–31] . The dialysis tubes were placed in 1 L beakers with 500 mL of culture media for 24 hrs at 37°C in 5% CO2 . Similarly , 5 adult female worms were soaked in media alone or scrambled siRNA ( 5 μM ) in dialysis tubes for each timepoint as experimental controls . After the 24-hr incubation , the worms for each group were carefully extracted from the dialysis tubes and individually placed into wells with 1 mL of media . The worms were evaluated at timepoints 1 , 3 , and 6 days post-incubation for transcript knockdown , worm motility , MTT reduction , and microfilariae release . Worms were visualized with a dissecting microscope by an observer blinded to treatment category . Motility of the adult female B . malayi worms was rated based on the following scale 4 = active movement , 3 = modest reduction in movement , 2 = severe reduction in movement , 1 = twitching , and 0 = no movement . Metabolic function of the adult female worms was assessed by reduction of ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide ( MTT ) from Sigma using a protocol established by Comley et al [32] . For each group per timepoint , 2 worms were incubated in 0 . 5 mL of phosphate buffered solution ( PBS ) pH 7 . 4 with 0 . 5 mg/mL of MTT for 30 minutes at 37°C in 5% CO2 . The worms were then transferred into separate wells of a 96-well plate containing 200 μL of DMSO and incubated at room temperature for 1 hr . MTT reduction was quantified by absorbance relative to a DMSO blank at 570 nm using a Synergy HTX multi-mode plate reader ( BioTek ) . For each timepoint , adult worms were placed in new culture media 24 hrs prior to enumeration of microfilariae . After the overnight incubation , the worms were then removed for processing by the MTT reduction assay and RT-qPCR . The Mf in the well containing expended culture media ( 1 mL ) were counted under a light microscope at high magnification . For each condition , adult B . malayi female worms ( n = 3 ) were homogenized in TRIzol ( Thermo Fischer Scientific ) after three freeze/thaw cycles using Matrix D lysis tubes ( MP Biomedicals ) agitated by a FastPrep™-24 Biopulverizer ( MP Biomedicals ) for 7 minutes at 6 m/s . Chloroform was added to the homogenate , transferred to Phase Lock Gel tubes ( 5Prime ) , and phase separated at 11 , 900 g for 15 minutes at 4°C . The aqueous phase was collected and cold isopropanol was added to precipitate the RNA , which was then pelleted at 12 , 000 g for 1 hr and washed twice using 75% ethanol . The RNA pellet was resuspended in nuclease-free water and quantified using a NanoDrop 1000 ( Thermo Fischer Scientific ) . cDNA was prepared using Superscript IV ( Thermo Fischer Scientific ) as per the manufacturer’s protocol . The cDNA levels of Bm-UGT and B . malayi house-keeping gene gapdh were assessed in duplicate 20 uL reactions using 1 μL of 20X TaqMan™ gene expression assay ( Thermo Fischer ) , 1 μL of cDNA , and 18 μL of TaqMan™ gene expression master mix ( Applied Biosystems ) . PCR conditions were 2 min at 50°C , 10 min at 95°C , 40 cycles of 15 sec at 95°C , and 1 min at 60°C cycle of 50°C with a 7500 Real-Time PCR System ( Applied Biosystems ) . The primers used were as follows: Polyclonal anti-Bm-UGT peptide antibodies were generated in New Zealand rabbits by Genscript using Bm-UGT peptide sequences conjugated to keyhole limpet hemocyanin ( KLH ) . The peptide sequences used were as follows: CYEKDEHLIAEGRPN , DSTGSKLAKTVKIDC , and CGQIANFDPYGRKMS . Cysteines were added at either the N- or C-terminus to facilitate KLH conjugation . B . malayi adult worms ( n = 5 ) were incubated in 5 μM combination of Bm-UGT siRNA for 24 hrs using the previously mentioned method and transferred into individual wells with 1 mL of media . The adult worms were cultured for an additional 24 hrs and then homogenized in PBS ( pH 7 . 4 ) and 4 μL Halt™ Protease Inhibitor Cocktail ( Thermo Scientific ) using Matrix D lysis tubes ( MP Biomedicals ) agitated by a FastPrep™-24 Biopulverizer ( MP Biomedicals ) for 3 minutes at 4 m/s . Protein levels were quantified by the Bradford protein assay ( Bio-Rad ) . For immunoblot analysis , 10 μg of protein was separated on 10% Bis-Tris NuPAGE gel ( Invitrogen ) and blotted onto 0 . 2 μm nitrocellulose filter paper ( Bio-Rad ) . After blocking overnight in 5% bovine serum albumin ( BSA ) in tris-buffered saline with 0 . 1% Tween 20 ( TBS-T ) , the membrane was incubated with 1:4000 anti-UGT peptide antibodies ( Genscript ) and 1:1000 rabbit anti-β actin antibodies ( Abcam ) for 1 hr . Following this , the filter paper was washed with TBS-T and then incubated with 1:2000 horseradish peroxidase conjugated goat anti-rabbit IgG for 1 hr . The membrane subsequently washed and incubated in Chemiluminescent reagent , SuperSignal™ West Pico PLUS ( Thermo Scientific ) , to visual the bands . Sulfinpyrazone ( ChemCruz ) and probenecid ( Invitrogen , water soluble formulation ) , broadly acting UGT inhibitors , were evaluated for macrofilaricidal activity in vitro . Sulfinpyrazone was resuspended in 1X PBS ( pH 7 . 4 ) and 1% dimethylsulfoxide ( DMSO , v/v ) while probenecid was resuspended in deionized water . When testing sulfinpyrazone , adult B . malayi female worms were incubated in culture media with the drug for 8 days at concentrations of 2500 μM , 1000 μM , 200 μM , 40 μM , and 8 μM . For probenecid , adult female worms were incubated in culture media with the drug for 7 days at concentrations of 5000 μM , 500 μM , 250 μM , and 100 μM . Worms were transferred into new media with corresponding drug concentrations every day except day 4 . As a negative control , worms were incubated in culture media alone with a similar volume of vehicle . Worm motility was scored using the previously mentioned scale for the course of the experiment . Worms that were scored a zero stopped receiving UGT inhibitor treatment . For the albendazole synergy experiments , adult filariae were incubated in culture media with 40 μM of sulfinpyrazone or 100 μM of probenecid in combination with 10 μM of albendazole , which was resuspended in 1X PBS ( pH 7 . 4 ) and 1% DMSO ( v/v ) . The worms were scored for motility for 8 days and were transferred into new media with corresponding drug concentrations every day except day 4 . The above UGT inhibitors were evaluated for microfilaricidal activity in vitro . For each drug , experiments were performed in triplicate at a concentration of 2 x 104 Mf/mL in culture media . Viability was determined by quantifying the number of motile larvae from 100 randomly selected Mf per well . The concentrations used for sulfinpyrazone were 2500 μM and 200 μM while the concentrations used for probenecid were 5000 μM and 500 μM . As a negative control , larvae were incubated in culture media alone . We measured cytotoxicity of the UGT inhibitors using a Pierce LDH Cytotoxicity Assay Kit ( Thermo Scientific ) . HEK cells were seeded at 5 x 104 per well in DMEM ( Quality Biological ) with 10% Hyclone Cosmic Calf Serum ( Thermo Fischer ) , 200 μM of L-glutamine ( Quality Biological ) , and 50 μg/mL of gentamicin ( Quality Biological ) at 37°C in 5% CO2 . We then incubated the cells with various concentrations of the UGT inhibitors overnight . Following this , we transferred 50 μL of media from each well to a new 96-well plate and then added 50 μL of reaction buffer . We incubated the mixture for 30 minutes and then added 50 μL of stop solution . We measured absorbance at 490 nm and 680 nm . We employed the following controls: a spontaneous LDH activity control which was incubated with the vehicle only and a maximum LDH activity control which was incubated with nothing but later lysed prior to incubation with the reaction buffer . We calculated absorbance for each well by subtracting the 680 nm absorbance value ( background ) from the 490 nm absorbance value . We then calculated percent cytotoxicity using the following equation: %Cytotoxicity= ( UGTinhibitorLDHactivity−SpontaneousLDHactvity ) ( MaximumLDHactivity−SpontaneousLDHactivity ) Bm-UGT was expressed as a Renilla reniformis luciferase ( Ruc ) fusion protein by cloning the Bm-UGT coding sequence in pREN2 ( Genscript ) . The Bm-UGT signal sequence as predicted by signalP was removed prior to synthesis . Plasmid encoding the fusion protein was used to transformed TOP10 cells ( Thermo Fischer ) and plasmid DNA was obtained from colonies selected on kanamycin ( 50 μg/ml ) as per the manufacturer’s guidelines ( Qiagen Midi-Prep ) . 293F cells grown in 293 Freestyle Medium as suspension cultures were transfected with 30 μg of Bm-UGT-Ruc plasmid , at a final concentration of 1 μg per 1 x 106 cells ( Thermo Fischer Sceintific ) per mL , and cultured at 37°C with 8%CO2 on a rotary shaker at 125 rpm . After 72hrs , the cells were pelleted and sonicated in LIPS lysis buffer ( 20 mM Tris pH 7 . 5 , 100 mM NaCl , 5 mM MgCl2 , 1% TritonX-100 , 50% glycerol , protease inhibitors ( Mini from Roche ) ) . The lysate was centrifuged to remove cellular debris and supernatant containing the Bm-UGT-Ruc fusion proteins were stored at -80°C for later use . Antibody titers were measured using a luciferase immunoprecipitation system ( LIPS ) assay [33–35] . For IgG and IgE quantification , serum was diluted to 1:100 and 1:10 respectively in 50 μL of LIPS master mix ( 20 mM Tris pH 7 . 5 , 100 mM NaCl , 5mM MgCl2 , 1% Triton X-100 ) and mixed with 50 μL of the UGT-Ruc fusion containing 1 x 106 light units ( LU ) of protein in PBST ( PBS with 0 . 05% Tween-20 ) . The reaction mixture was incubated in a 96-well polypropylene plate for 10 minutes at room temperature and transferred to a 96-well high throughput screening filter plate ( Milipore ) containing 5 μL of a 50% suspension of Ultralink protein A/G ( Pierce ) or Ultralink anti-human IgE beads in PBS and incubated for an additional 15 minutes at room temperature . The plates were washed under vacuum with 200 μL of LIPS master 3x followed by PBS once . The relative light units ( RLU ) were measured with a Berthold LB 960 Centro microplate luminometer with 50 μL of coelenterzine solution ( Promega ) . For these experiments , samples were run in duplicate , and the calculated RLU was adjusted for the measured RLU of UGT fusion protein without serum . All human serum samples were obtained following written informed consent from all subjects using Institutional Review Board-approved protocols that have been registered ( NCT00001345 , NCT00090662 , NCT00342576 ) . Patients were grouped into clinical categories as previously detailed [36] . The siRNA and UGT inhibitor experiments were repeated two times under similar conditions . Data shown is from a single representative experiment . For the siRNA experiments , data was analyzed using one-way analysis of variance ( ANOVA ) or T-test by PRISM 7 . 0 . Following ANOVA , individual comparisons of mean values were performed using Tukey’s multiple comparisons test . For the UGT inhibitor experiments , we performed AUC analysis followed by one-way ANOVA to determine significance . Statistical significance between the experimental and control groups was designated as follows: * for p values <0 . 05 , ** for p values <0 . 01 , and *** for p values <0 . 001 .
UDP-glucuronosyltransferases ( UGT ) are enzymes important for detoxification of xenobiotics and homeostasis of endogenous molecules [21] . Specifically , these phase II enzymes increase the solubility of hydrophobic molecules by attaching sugar moieties such as glucuronic acid . Since this sugar molecule is negatively charged at physiological pH , anion efflux pumps are able to transport these molecules outside the cell [37] . In C . elegans , studies demonstrated that RNAi of detoxification enzymes results in lethality , sluggish movement , or impaired growth [38–40] . In addition , one study showed that glycosylation by phase II enzymes was important for the detoxification of albendazole in C . elegans [22] . While UGTs in helminths have not been studied extensively , there is evidence from intestinal helminths to suggest that these enzymes play a critical role in drug resistance [41–43] . In this study , we showed that B . malayi intestinal UGT expression could be silenced by specific siRNA . This knockdown caused a significant reduction in worm motility , fecundity , and metabolism . While these metrics alone do not determine worm survival , we are fairly confident these are appropriate surrogates . We observed several worms with a scored motility of zero early into the siRNA experiments that did not show any recovery over the course of the experiment . Thus , Bm-UGT appears essential for adult B . malayi worm survival . It is possible that the observed phenotypic changes occurred due to an imbalance in endogenous molecules . Past studies in mice and rats demonstrated that UGTs were critical for protection against free radicals , which if left unchecked could mediate damage to DNA , lipid membranes , and amino acids [44–47] . The reduction in microfilaria release was most likely due to decreased adult worm viability because Bm-UGT is not expressed in the Mf stage [13 , 14] . We suspect that targeting Bm-UGT in vivo would leave the worms not only susceptible to endogenous free radicals but also to those released by host immune cells . Interestingly , it should be noted that the Brugia intestinal UGT was predicted by InterPro to be localized to the plasma membrane , which would be a novel location for this family of enzymes that are typically found in the endoplasmic reticulum . The prediction software also determined that this protein has a large extracellular domain , which potentially makes it readily accessible to drugs or antibodies . Development of short-course macrofilaricidal agents would greatly enhance LF eradication efforts . Because the aim of this study was to evaluate Bm-UGT as a potential drug and vaccine target , we investigated the effects of non-specific UGT inhibitors on adult filariae . We found two UGT inhibitors that are also FDA-approved to treat gout [17–20] . Both drugs , sulfinpyrazone and probenecid , exhibited macrofilaricidal activity in vitro . For worms that were scored a zero for motility , we did not see any recovery after cessation of the UGT inhibitor treatment . The lowest effective concentration for sulfinpyrazone was 200 μM . A previous study investigating sulfinpyrazone showed a maximum concentration ( Cmax ) in humans of 79 . 9 μM for a 400 mg dose [48] . Given that the daily maximum recommended dose for humans is 800 mg , we believe a Cmax similar to our lowest effective sulfinpyrazone concentration may be achievable in humans [49] . Therefore , we speculate that this drug could serve as a novel therapeutic against adult filariae . Likewise , we believe that the same applies to probenecid , which demonstrated robust macrofilaricidal activity in vitro at 500 μM . We also observed a significant reduction in motility by day 7 at 250 μM suggesting that this concentration may be effective at killing adult worms if given over a longer time course . In the context of physiological relevance , one pharmacokinetic study showed the peak concentration in humans given a single 2 g oral dose of probenecid to be 148 . 6 μg/mL ( 520 . 7 μM ) with minimal adverse events [50] . Based on our in vitro data , this level could rapidly kill adult filarial worms . While the UGT inhibitors displayed macrofilaricidal activity , it is possible that probenecid and sulfinpyrazone may act on adult B . malayi worms through mechanisms independent of effects on Bm-UGT . In addition to inhibiting UGTs , probenecid and sulfinpyrazone also inhibit organic anion transporters ( OATs ) [51 , 52] and pannexins [53] . OATs function to transport negatively charged molecules and are likely important for adult filarial worm survival . Innexins , which are present in invertebrate organisms , are structurally very similar to pannexins and primarily function as membrane channels that communicate with the extracellular space [54] . Interestingly , probenecid has been shown to impair touch responses in C . elegans by inhibiting mechanosensitive innexin channels [55] . Elucidating the mechanisms by which probenecid and sulfinpyrazone kill adult filarial worms will be the focus of future studies . While we are interested in developing a drug that kills adult filarial worms , the absence of microfilaricidal activity also presents advantages . Current antifilarial therapeutics such as diethylcarbamazine ( DEC ) and ivermectin ( IVM ) are extremely effective at clearing microfilariae . However , their use is contraindicated in areas co-endemic for loiasis and onchocerciasis because rapid killing of microfilariae in these infections can lead to severe adverse outcomes [56] . Individuals with loiasis when treated with IVM or DEC have a significantly higher risk of experiencing severe neurologic events such as encephalopathy due to rapid Mf death in the vasculature [57 , 58] . Similarly , DEC can induce adverse systemic reactions such as skins lesions , fever , polyarthritis , and ocular reactions in patients with onchocerciasis as determined by Mf load [59 , 60] . Therefore , probenecid , which demonstrated macrofilaricidal but not microfilaricidal activity at 500 μM in vitro , may be an attractive LF treatment candidate in these co-endemic areas . Because UGTs are involved in drug metabolism , we suspected that there may be synergy between the UGT inhibitors and albendazole . A recent study demonstrated that overexpression of UGT-22 in C . elegans resulted in albendazole resistance [61] . This , coupled with an earlier study by Laing et al showing that upregulation of UGTs is associated with metabolism of albendazole into various glucuronide products [22] , suggests that UGTs may play a critical role in nematode metabolism of albendazole . In support of this hypothesis , our data demonstrate a synergistic effect of sub-macrofilaricidal concentrations with sulfinpyrazone or probenecid in combination with albendazole against adult filaria in vitro . These results suggest that combination therapy with albendazole and either probenecid or sulfinpyrazone may be highly effective in treating filarial infections in people . Future studies will determine whether Bm-UGT functions to metabolize albendazole in filarial worms and whether this metabolism is inhibited by UGT inhibitors . We also plan to test the efficacy of combination therapy in animal models of infection . As previously mentioned , these probenecid and sulfinpyrazone inhibitors are FDA-approved and have been shown to be safe in humans [62] , with probenecid characterized as a pregnancy category B drug . If these medicines demonstrate macrofilaricidal activity in vivo , translation into human use could occur quickly . In addition to being a potential drug target for filariasis , we postulate that Bm-UGT could serve as a vaccine candidate as well . One of the challenges in helminth vaccine development is the risk that the vaccine may induce an allergic response in endemic populations . Indeed , generalized urticaria was seen in several Brazilian patients immunized against Ancylostoma-secreted protein 2 during hookworm vaccine trials [23] . A solution to this obstacle is to identify “hidden antigens” which are not exposed to the immune system during natural infection yet are essential to worm survival and present in an anatomical location accessible to host antibodies [63 , 64] . In theory , these proteins would not elicit an IgE-mediated response , and , as postulated by Munn , these antigens may be especially vulnerable to the immune system due to a lack of evolutionary pressure to evade it [64] . There is evidence to support that the intestinal tract of nematodes contains hidden antigens . Studies have demonstrated the absence of pre-existing IgE in serum from endemic populations against hookworm intestinal antigens APR-1 and GST [8 , 9 , 65] . Furthermore , studies have shown that these antigens are protective in animal models [8 , 10 , 11] . There is also evidence of hidden antigens in H . contortus seen with the lack of an antibody response against H11 , a glycosylated intestinal protein [66] . In this study , we observed no detectable Bm-UGT-specific IgE in the serum from individuals infected with or exposed to lymphatic filariae , which suggests this antigen may be safe to administer as a vaccine candidate in endemic populations . However , due to the low numbers of patient samples tested , if vaccine work using Bm-UGT progresses , then future studies would need to evaluate potential for allergic responses by testing far larger numbers of individuals . After demonstrating that recombinant Bm-UGT would not induce an allergic response in endemic populations , we investigated whether adult B . malayi worms could ingest antibody as there is a degree of uncertainty about whether filarial worms use their intestine to feed . Only the adult and L4 stages of filariae have a fully formed intestinal tract , while the Mf , L2 , and L3 stages have an immature intestine inaccessible to nutrients [67] . Furthermore , past studies have already shown that Brugia worms are able to absorb nutrients such as nucleotides , amino acids , sugars , and vitamins through their cuticle , which calls into question the purpose of the intestinal tract [68] . Notwithstanding these findings , Attout et al . investigated the intestinal tract of L . sigmodontis and demonstrated that young adult worms ( 25–56 weeks post-infection ) ingested red blood cells [69] . This suggests that intestinal feeding may occur but only at the early adult stage . Another study showed that D . immitis can ingest labeled serum [70] . With our study , we were able to show that adult B . malayi worms can ingest Cy3-labeled IgG ( S1 Fig ) . This serves as clear evidence that circulating antibody can potentially access the intestine of adult filarial worms . However , not all the worms ingested the labeled antibody , which indicates that , at least during in vitro conditions , adult worms do not feed through the intestine continuously . Future studies will work towards recombinant expression of Bm-UGT in order to test its ability to induce protective immune responses in animal models of filariasis . On the basis of the phylogenetics analyses we conducted , we expect that Bm-UGT may also be essential in W . bancrofti and B . timori given the overall high homology shared between these species and B . malayi . Additionally , there is a high level of sequence homology ( > 70% ) between the B . malayi intestinal UGT and the orthologs found in D . immitis and L . loa and thus the UGT orthologs in these filaria species may also serve as novel therapeutic targets . Interestingly , we did not find an ortholog of Bm-UGT in O . volvulus and speculate this could be the result of evolutionary pressure to lose this gene . In summary , we believe that Bm-UGT is an essential intestinal protein in B . malayi adult worms that does not induce IgE antibodies in endemic populations . Importantly , we found that sulfinpyrazone and probenecid , two commercially available , FDA-approved medications in use for gout , exhibit strong macrofilaricidal activity in vitro at concentrations that are achievable in humans . This promising data warrants future investigation in animal models of Brugia infection as well as assessment of whether these UGT inhibitors exhibit macrofilaricidal activity in vitro against other filarial species . Furthermore , we demonstrated that probenecid and sulfinpyrazone may potentiate the effect of albendazole on filariae , suggesting that combination therapy may be an ideal approach to obtain macrofilaricidal effect . In terms of vaccine development , future studies will focus on recombinantly expressing Bm-UGT and then testing whether it is protective in animal models of filariasis . Finally , the results of this study suggest that the intestinal tract of filarial nematodes may serve as a rich source of essential proteins that can serve as important therapeutic targets . | Brugia malayi is a parasitic nematode and one of the causative agents of lymphatic filariasis , a disease that affects 70 million people worldwide . Currently , there are no effective therapeutics that kill adult filarial parasites when given as a short course . This limitation has hampered global eradication efforts . Studies have shown that the intestinal tract in nematodes can be effectively targeted by drugs and antibodies . Given this potential , we decided to investigate B . malayi intestinal UDP-glucuronosyltransferase as a potential therapeutic target . We determined that this protein is essential for B . malayi adult worm survival , as gene-expression knockdown rapidly decreased motility , fecundity , and microfilarial release . We also identified two FDA-approved UGT inhibitors that cause death of adult filariae in vitro . This is a critical finding due to the need for effective macrofilaricides and the potentially rapid translatability of these drugs for use in filaria-infected people . Finally , we showed that serum from filarial patients does not contain specific IgE to Bm-UGT and thus this protein would likely not induce allergic reaction if given as a vaccine antigen to endemic populations . | [
"Abstract",
"Introduction",
"Results",
"Methods",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"laboratory",
"equipment",
"engineering",
"and",
"technology",
"gene",
"regulation",
"immunology",
"biological",
"cultures",
"animals",
"vaccines",
"culture",
"media",
"brugia",
"infectious"... | 2019 | Intestinal UDP-glucuronosyltransferase as a potential target for the treatment and prevention of lymphatic filariasis |
Decisions involve two fundamental problems , selecting goals and generating actions to pursue those goals . While simple decisions involve choosing a goal and pursuing it , humans evolved to survive in hostile dynamic environments where goal availability and value can change with time and previous actions , entangling goal decisions with action selection . Recent studies suggest the brain generates concurrent action-plans for competing goals , using online information to bias the competition until a single goal is pursued . This creates a challenging problem of integrating information across diverse types , including both the dynamic value of the goal and the costs of action . We model the computations underlying dynamic decision-making with disparate value types , using the probability of getting the highest pay-off with the least effort as a common currency that supports goal competition . This framework predicts many aspects of decision behavior that have eluded a common explanation .
A soccer player moves the ball down the field , looking for an open teammate or a chance to score a goal . Abstractly , the soccer player faces a ubiquitous but challenging decision problem . He/she must select between many competing goals while acting , whose costs and benefits can change dynamically during ongoing actions . In this game scenario , the attacker has options to pass the ball to one of his/her teammates . An undefended player is preferred , but this opportunity will soon be lost if the ball is not quickly passed . If all teammates are marked by opposing players , other alternatives like holding the ball and delaying the decision may be better . Critically , the best option is not immediately evident before acting . To decide which strategy to follow at a given moment requires dynamically integrating value information from disparate sources . This information is diverse relating to both the dynamic value of the goal ( i . e . , relative reward of the goal , probability that reward is available for that goal ) and the dynamic action cost ( i . e . , cost of actions to pursue that goal , precision required ) , creating a challenging problem in integrating information across these diverse types in real time . Despite intense research in decision neuroscience , dynamic value integration into a common currency remains poorly understood . Previous explanations fall into two categories . The goods-based theory [1–7] proposes that all the decision factors associated with an option are integrated into a subjective economic value independently computed for each alternative . This view is consistent with evidence suggesting convergence of value information in the prefrontal cortex [3–5 , 7] . Critically , action planning starts only after a decision is made . While this view is sufficient for decisions like buying or renting a house , modifications are needed for decisions while acting . Alternatively , an action-based theory proposes that options have associated action-plans . According to this theory , when the brain is faced with multiple potential goals , it generates concurrent action-plans that compete for selection and uses value information to bias this competition until a single option is selected [8–14] . This theory has been received apparent support from neurophysiological [8–10 , 15–19] and behavioral [11 , 12 , 20–23] studies . Although the action-based theory explains competition , it leaves mysterious how action cost is integrated with good value ( also referred as stimulus value in some decision-making studies [13] ) that have different currencies and how goods-based decisions that do not involve action competition are made . To solve complex decision problems , the brain must dynamically integrate all the factors that influence the desirability of engaging in an action-plan directed towards a goal . We propose a theory of dynamic value integration that subsumes both goods-based and action-based theories . We provide a simple , computationally feasible way to integrate online information about the cost of actions and the value of goods into an evolving assessment of the desirability of each goal . By integrating value information into a common currency , our approach models many key results in decision tasks with competing goals that have eluded a common explanation , including trajectory averaging in rapid reaching tasks with multiple potential goals , a common explanation for errors due to competition including the global-effect paradigm in express saccadic movements [24] , and a unified explanation for the pattern of errors due to competition in sequential decisions [25] .
Stochastic optimal control has proven a powerful tool at modeling goal-directed movements , such as reaching [26] , grasping [27] and walking [28] ( for review see [29] ) . It involves solving for a policy π that maps states into actions ut = π ( xt ) by minimizing a cost function penalizing actions and deviations from a goal . Despite the growing popularity of optimal control models , most of them are limited to tasks with single goals , because policies are easily defined towards a single goal . On the other hand , it is unclear how to define policies in the presence of multiple goals , each of which may provide different reward and may require different effort . The core difficulty is to develop a single policy that selects actions that pursue many targets but ultimately arrives at only one . One of the simplest solutions is to carefully construct a composite cost function that incorporates all targets . However , naive applications of this approach can produce quite poor results . For instance , an additive mixture of quadratic cost functions is a new cost function with a minimum that does not lie at any of the competing targets . The difficulty is that quadratic cost functions do not capture the winner-take-all implicit reward structure , since mixtures of quadratics reward best for terminal positions in between targets . Even when such a cost function can be constructed , it can be very difficult to solve the policy , since these types of decision problems are P-SPACE complete—a class of problems more intractable than NP-complete . Any dynamic change in targets configuration requires a full re-computation , which makes the approach difficult to implement as a real-time control strategy [30] . To preserve simplicity , we propose to decompose the problem into policy solutions for the individual targets . The overall solution should involve following the best policy at each moment , given incoming information . We can construct a simple cost function that has this property using indicator variables ν ( xt ) . The indicator variables encode the policy that has the lowest future expected value from each state—in other words , it categorizes the state space into regions where following one of the policies to a goal i is the best option . In essence , a goal i “owns” these regions of the state space . We can write the cost function that describes this problem as a ν-weighted mixture of individual cost functions J j ′ s: J = ∑ j = 1 N ν j ( x t ) J j ( x t , π j ) J = ∑ j = 1 N ν j ( x t ) ( ( x T j - S p j ) T Q T j ( x T j - S p j ) + ∑ t = 1 T j π j ( x t ) T R π j ( x t ) ︸ J j ( x t , π j ) ) ( 1 ) where N is the total number of targets and νj is the indicator variable associated with the target j . The cost function Jj ( xt , πj ) describes the individual goal for reaching the target j starting from the current state xt and following the policy πj for time instances t = [t1 , ⋯ , tTj] . Tj is the time-to-contact that target j and S is a matrix that picks out the hand and target positions from the state vector . The first term of the cost Jj is the accuracy cost that penalizes actions that drive the end-point of the reaching trajectory away from the target position pj . The second term is the motor command cost that penalizes the effort required to reach the target . Both the accuracy cost and the motor command cost characterize the “action cost” Vπj ( xt ) for implementing the policy πj at the state xt . Matrices QTj and R define the precision- and the control- dependent costs , respectively ( see S1 Text for more details ) . When there is no uncertainty as to which policy to implement at a given time and state ( e . g . , actual target location is known ) , the ν-weighted cost function in Eq ( 1 ) is equivalent to the classical optimal control problem . The best policy is given by the minimization of the cost function in Eq ( 1 ) with νj = 1 for the actual target j and νi ≠ j = 0 for the rest of the non-targets . However , when there is more than one competing target in the field , there is uncertainty about which policy to follow at each time and state . In this case , the best policy is given by minimizing the expected cost function with expectation across the probability distribution of the indicator variable ν . This minimization can be approximated by the weighted average of the minimization of the expected individual cost functions , Eq ( 2 ) . π m i x ( x t ) = ∑ j = 1 N ⟨ ν j ( x t ) ⟩ ν arg min π j J j ( x t , π j ) = ∑ j = 1 N ⟨ ν j ( x t ) ⟩ ν π j * ( x t ) ( 2 ) where ⟨ . ⟩ν is the expected value across the probability distribution of the indicator variable ν , and π j * ( x t ) is the optimal policy to reach goal j starting from the current state xt . For notational simplicity , we omit the * sign from the policy π , and from now on πj ( xt ) will indicate the optimal policy to achieve the goal j at state xt . The first problem is to compute the weighting factor ⟨νj ( xt ) ⟩ν , which determines the contribution of each individual policy πj ( xt ) to the weighted average πmix ( xt ) . Let’s consider for now that all the alternative targets have the same good values and hence the behavior is determined solely by the action costs . Recall that Vπj ( xt ) represents the value function—i . e . , cost that is expected to accumulate from the current state xt to target j including the accuracy penalty at the end of the movement , under the policy πj ( xt ) . This cost partially characterizes the probability of achieving at least Vπj ( xt ) starting from state x ( t ) at time t and adopting the policy πj ( xt ) to reach the target j , Eq ( 3 ) : P ( V π j ( x t ) | π j ( x t ) , x t , Δ t ) = λ e - 1 λ V π j ( x t ) ( 3 ) where λ is the free “inverse temperature” parameter ( S3 Text ) . This assumption can be taken as is , or justified from the path integral approach in [31] and [32] . The probability that the value function of the policy πj at the current state xt is lower than the rest of the alternatives P ( Vπj ( xt ) < Vπi ≠ j ( xt ) ) can be approximated by the softmax-type equation in Eq ( 4 ) , which gives an estimate of the probability of νj at xt: P ( V π j ( x t ) < V π i ≠ j ( x t ) ) ≈ P ( ν j | x t ) = λ e - 1 λ V π j ( x t ) ∑ i = 1 N λ e - 1 λ V π i ( x t ) ( 4 ) where N is the total number of targets ( i . e . , and total number of policies ) that are available at the current state . Given that all targets have the same good values , the probability P ( νj|xt ) characterizes the “relative desirability” rD ( πj ( xt ) ) of the policy πj to pursue the goal j at a given state xt . It reflects how desirable is to follow the policy πj at that state with respect to the alternatives . Therefore , we can write that: r D ( π j ( x t ) ) = P ( V π j ( x t ) < V π i ≠ j ( x t ) ) ( 5 ) However , in a natural environment the alternative goals are usually attached with different values that we should take into account before making a decision . We integrate the good values into the relative desirability by computing the probability that pursing the goal j will result in overall higher pay-off rj than the alternatives , P ( rj > ri ≠ j ) : r D ( π j ( x t ) ) = P ( V π j ( x t ) < V π i ≠ j ( x t ) ) P ( r j > r i ≠ j ) ( 6 ) To integrate the goods-related component on the relative desirability , we consider two cases: The reward magnitude is fixed and equal for all targets , but the receipt of reward is probabilistic . In this case , the probability that the value of the target j is higher than the rest of the alternatives is given by the reward probability of this target P ( target = j|xt ) = pj: P ( r j > r i ≠ j ) = p j ( 7 ) The target provides a reward with probability pj , but the reward magnitude is not fixed . Instead , we assume that it follows a distribution r j ∼ ( 1 - p j ) δ ( r j ) + p j N ( μ j , σ j 2 ) , where δ ( rj ) is the Delta dirac function , and μj and σj are the mean and the standard deviation of the reward attached to the target j . For simplicity reasons , we focus on the case with two potential targets , in which the goal is to achieve the highest pay-off after N trials . In this case , the goods-related component of the desirability function is P ( r ¯ 1 > r ¯ 2 ) , where r ¯ j = 1 N ∑ k = 1 N r j ( k ) , j = 1 , 2 is the net reward attached to the target j—i . e . , the average reward received from the target j across N trials . To compute the probability P ( r ¯ 1 > r ¯ 2 ) , we need the probability distribution of P ( r ¯ j ) , j = 1 , 2 . Given p ( n ) = Binomial ( n , pj , N ) is the probability of receiving n-times reward after N trials , the probability distribution of r ¯ j is: P ( r ¯ j ) = ∑ n = 0 N p ( n ) ( 1 N ∑ k = 1 n r j ( i ) ) ( 8 ) We can show that a mean based on n samples has a Normal distribution N ( n N μ j , σ j 2 n ) . Therefore , the distribution of r ¯ j can be written as: P ( r ¯ j ) = ∑ n = 0 N p ( n ) N ( r ¯ j ; n N μ j , σ j 2 n ) ( 9 ) For a large number of trials N > > 0 , p ( n ) is concentrated around n = pj N and r ¯ j ∼ N ( p j μ j , σ j 2 p j N ) , j = 1 , 2 . To compute P ( r ¯ 1 > r ¯ 2 ) = P ( r ¯ 1 - r ¯ 2 > 0 ) , we define a new random variable , Z = r ¯ 1 - r ¯ 2 , which has Normal distribution with mean p1 μ1 − p2 μ2 and variance σ 1 2 p 1 N + σ 2 2 p 2 N . We can show that P ( r ¯ 1 > r ¯ 2 ) = P ( Z > 0 ) is given as: P ( r ¯ 1 > r ¯ 2 ) = 1 2 e r f c ( p 2 μ 2 - p 1 μ 1 2 ( σ 1 2 p 1 N + σ 2 2 p 2 N ) ) ( 10 ) where erfc is the complementary error function . Using that erfc ( x ) = 1 − erf ( x ) , where erf is the error function , we can write that: P ( r ¯ 1 > r ¯ 2 ) = 1 2 + 1 2 e r f ( p 1 μ 1 - p 2 μ 2 2 ( σ 1 2 p 1 N + σ 2 2 p 2 N ) ) = C u m N o r m ( Z ; p 1 μ 1 - p 2 μ 2 , σ 1 2 p 1 N + σ 2 2 p 2 N ) ( 11 ) This result is consistent with the common practice of modeling choice probabilities as a softmax function between options . For example , the cumulative normal distribution can be approximated by the following logistic function [33]: P ( r 1 > r 2 ) = l ( r ¯ 1 - r ¯ 2 ; p 2 μ 2 - p 1 μ 1 , S 1 . 6 ) ( 12 ) where S = σ 1 2 p 1 N + σ 2 2 p 2 N . In the preceding sections we developed a theory for the case that targets are presented simultaneously and the expected reward depends only on successfully reaching the target—i . e . reward availability is not state- and time- dependent . However , decisions are not limited only to this case but often involve goals with time-dependent values . In this section , we extend our approach to model visuomotor tasks with sequential goals , focusing on a pentagon copying task . The theory precedes as before , with a set of control schemes that instantiate policies πj ( xt ) —where ( j = 1 , ⋯ 5 ) —that drive the hand from the current state to the vertex j . However , to draw the shape in a proper spatial order , we cannot use the same policy mixing as with simultaneously presented goals . Instead , we have to take into account the sequential constraints that induce a temporal order across the vertices . We can conceive the vertices as potential goals that provide the same amount of reward , but with different probabilities ( i . e . , similar to scenario 2 in the reaching task ) with the exception that we design the target probability to be time- and state- dependent , so that it encodes the order of policies for copying the pentagon . The target probability P ( vertex = j|xt ) describes the probability that the vertex j is the current goal of the task at the state xt after departing from the vertex j − 1 , or in other words , it describes the probability that we copy the segment defined by the two successive vertices j − 1 and j . We define an indicator function ej that is 1 if we arrive at vertex j and 0 otherwise . P ( v e r t e x = j | x t ) = P ( e j = 0 , e j - 1 = 1 | x t ) = P ( e j = 0 | x t ) p ( e j = 1 | x t ) = ( 13 ) = ( 1 - P ( τ a r r i v e j < t ) ) P ( τ a r r i v e j - 1 < t ) ( 14 ) where τ a r r i v e j is the time to arrive at vertex j -i . e , time to complete drawing the segment defined by the vertices j − 1 and j . Let’s assume that we are copying the shape counterclockwise starting from the purple vertex ( see right inset in Fig 1A ) , at the initial state at time t = 0 . The probability distribution of time to arrive at vertex j , τ a r r i v e j , is given by Eq ( 15 ) . P ( τ a r r i v e j ) = ∑ k = 1 j P ( τ a r r i v e k | τ a r r i v e k - 1 ) P ( τ a r r i v e k - 1 ) ( 15 ) where P ( τ a r r i v e k | τ a r r i v e k - 1 ) is the probability distribution of time to arrive at vertex k given that we started from vertex k − 1 . We generated 100 trajectories between two successive vertices and found that P ( τ a r r i v e k | τ a r r i v e k - 1 ) can be approximated by a Normal distribution N ( μ τ a r r i v e , σ τ a r r i v e 2 ) . Using Eq ( 15 ) , we show that P ( τ a r r i v e j ) is also Gaussian distribution , but with j times the mean and variance—N ( j μ τ a r r i v e , j σ τ a r r i v e 2 ) as shown in Fig 1A . Considering that , we estimate that target probability P ( vertex = j|xt ) , Fig 1B . Each time that we arrive at a vertex , we condition on completion , and P ( vertex = j|xt ) is re-evaluated for the next vertices .
The basic architecture of the model is a set of control schemes , associated with individual goals , Fig 2 . Each scheme is a stochastic optimal control system that generates both a goal-specific policy πj , which is a mapping between states and best-actions , and an action-cost function that computes the expected control costs to achieve the goal j from any state ( see S1 Text for more details ) . It is important to note that a policy is not particular a sequence of actions—rather it is a controller that tells you what action-plan uj ( i . e . , sequence of actions ui ) to take from a state xt to the goal ( i . e . , πj ( xt ) = uj = [ut , ut+1 , ⋯ utend] ) . In addition , the action-cost function is a map cost ( j ) = Vπj ( xt ) that gives the expected action cost from each state to the goal . Let’s reconsider the soccer game scenario and assume a situation in which the player has 3 alternative options to pass the ball ( i . e . , 3 unmarked teammates ) at different distances from the current state xt . In such a situation , the control schemes related to these options become active and suggest 3 action-plans ( u1 = π1 ( xt ) , u2 = π2 ( xt ) and u3 = π3 ( xt ) ) to pursue the individual options . Each of the alternative action-plans is assigned with value related to the option itself ( e . g . , teammates’ performance , distance of the teammates to the goalie ) and with cost required to implement this plan ( e . g . , effort ) . For instance , it requires less effort to pass the ball to the nearby teammate No . 1 , but the distant teammate No . 2 is considered a better option , because he/she is closer to the opponent goalie . While the game progresses , the cost of the action-plans and the estimates of the values of the alternative options change continuously . To make a correct choice , the player should integrate the incoming information online and while acting . However , the value of the options and the cost of the actions have different “currencies” , making the value integration a challenging procedure . The proposed theory uses a probabilistic approach to dynamically integrate value information from disparate sources into a common currency that we call the relative desirability function w ( xt ) . While common currency usually refers to integration in value space , relative desirability combines in the space of policy weights . Using relative desirability , integration of disparate values is accomplished by combining each different type of value in its own space , then computing the relative impact of that value on the set of available policies . The crux of our approach is that to make a decision , we only need to know what is the current best option and whether we can achieve it . This changes the complex problem of converting action costs to good values into a simple problem of maximizing the chances of getting the best of the alternatives that are currently available . To integrate value information with different “currencies” , we compute the probability of achieving the most rewarding option from a given time and state . This probability has both action-related and goods-related components with an intuitive interpretation: the probability of getting the highest reward with the least effort . We call this value relative desirability ( rD ) because it quantifies the attractiveness of the policy π for each goal i from state xt relative to the alternative options: r D ( π i ( x t ) ) = P ( c o s t ( i ) < c o s t ( j ≠ i ) | x t ) P ( r e w a r d ( i ) > r e w a r d ( j ≠ i ) | x t ) ( 16 ) The first term is the “action-related” component of the relative desirability and describes the probability that pursuing the goal i has lowest cost relative to alternatives , at the given state xt . The second term refers to the “goods-related” component and describes the probability that selecting the goal i will result in highest reward compared to the alternatives , at the current state xt . Note that the relative desirability values of the alternative options are normalized so that they all sum to 1 . To illustrate the relative desirability function , consider a reaching task with two potential targets presented in left ( target L ) and right ( target R ) visual fields ( gray circles in Fig 3A ) . For any state xt where the policy to the right target is more “desirable” than to the left target , we have the following inequality: r D ( π R ( x t ) ) > r D ( π L ( x t ) ) ( 17 ) This inequality predicts two extreme reaching behaviors—a direct movement to the target R ( i . e . , winner-take-all ) when rD ( πR ( xt ) ) > > rD ( πL ( xt ) ) , and a spatial averaging movement towards an intermediate position between the two targets when rD ( πR ( xt ) ) ≈ rD ( πL ( xt ) ) . Rearranging this equation , using P ( reward ( L ) > reward ( R ) ) = 1 − P ( reward ( R ) > reward ( L ) ) we see that the relative desirability to pursue the target R increases with the odds that target L has more reward and lower cost: r D ( π R ( x t ) ) > r D ( π L | x t ) ) → ( P ( r e w a r d ( R ) > r e w a r d ( L ) ) 1 - P ( r e w a r d ( R ) > r e w a r d ( L ) ) ) ( P ( c o s t ( R ) < c o s t ( L ) | x t ) 1 - P ( c o s t ( R ) < c o s t ( L ) | x t ) ) > 1 ( 18 ) To gain more insight on how action cost and good value influence the reaching behavior , we visualize the relative desirability to reach the right target in 3 scenarios ( the desirability related to the left target is a mirror image of the right one ) : For this case , P ( r e w a r d ( R ) > r e w a r d ( L ) ) = 1 - P ( r e w a r d ( R ) > r e w a r d ( L ) ) which means target R is more desirable when P ( c o s t ( R ) > c o s t ( L ) | x t ) < 0 . 5 Now the action cost ( and hence relative desirability ) is a function of the hand-state , making them difficult to illustrate . For a point-mass hand in 2D , the hand state is captured by the 4D position-velocity . To visualize this 4D relative desirability map in two dimensions , we “slice” through the 4D position-velocity space by making velocity a function of position in the following way . All trajectories are constrained to start at position ( 0 , 0 ) with zero velocity . We then allow the trajectory to arrive at one of a set of spatial positions ( 100 total ) around a circle of radius 85% of the distance between the start point and the midpoint ( black star ) of the two potential targets . For each of these points , we constrain the hand velocity to have direction ( red arrows in Fig 3A ) in line with the start point ( gray square in Fig 3A ) and the hand position on the circle ( black dots in Fig 3A ) . We set the magnitude of the velocities to match the speed of the optimal reaching movement at 85% of completion ( blue trace for left target in Fig 3A ) . From each position-velocity pair on the circle , we sample 100 optimal movements to each of the two targets ( solid and discontinuous traces are illustrated examples for reaching the left and the right target , respectively ) . We discretize the space and compute the action cost to reach the targets from each state—the expected cost from each state to the goal following the policy for that goal , including an accuracy penalty at the end of the movement . Fig 3B and 3C depict these action costs , where blue indicates low cost and red indicates high cost , respectively . Fig 3D illustrates the action costs converted into relative desirability values to reach the right target ( indicted by a solid gray circle ) , where blue and red regions correspond to states with low and high desirability , respectively . Notice desirability increases rapidly as the reach approaches a target , resulting in winner-take-all selection of an action-plan once moving definitely towards a target . However , when the hand position is about the same distance from both targets ( greenish areas ) there is no dominant policy , leading to strong competition and spatial averaging of the competing policies . In this case , desirability also depends on the probability of reward . Since both targets provide the same amount of reward , but with different probabilities , the goods-related term simplifies: P ( r e w a r d ( R ) > r e w a r d ( L ) ) = P ( t a r g e t = R ) = p R where pR describes the probability of earning reward by pursuing the right target . Hence , the target R is more desirable in a state xt when P ( c o s t ( R ) > c o s t ( L ) | x t ) < p R The relative desirability function for the right target is illustrated in Fig 3E , when pR is 4 times higher than the probability of the left target ( pR = 0 . 8 , pL = 0 . 2 ) . The right target is more desirable for most states ( reddish areas ) , unless the hand position is already nearby the left target ( blue areas ) , predicting frequent winner-take-all behavior -i . e . , direct reaches to the right target . More generally , the reward magnitude attached to each target is not fixed , but both the reward magnitude and reward probability vary . We assume that target j provides a reward with probability pj , and that the magnitude follows a Normal distribution with mean μj and standard deviation σj . Hence , the distribution of the rewards attached to the left target ( L ) and right target ( R ) is a mixture of distributions: r e w a r d ( L ) ∼ ( 1 - p L ) δ ( r e w a r d ( L ) ) + p L N ( μ L , σ L 2 ) ( 19 ) r e w a r d ( R ) ∼ ( 1 - p R ) δ ( r e w a r d ( R ) ) + p R N ( μ R , σ R 2 ) ( 20 ) where δ is the Dirac function . In visuomotor decision tasks , the ultimate goal is usually to achieve the highest reward after N trials . In this case , the probability that the right target provides overall higher reward than the left one over N trials can be approximated by a logistic function l with argument pR μR − pL μL ( see Materials and Methods section for more details ) . When the reward values are precisely encoded , this simplifies to: P ( r e w a r d ( R ) > r e w a r d ( L ) ) ≈ l ( p R μ R - p L μ L ) ( 21 ) Hence , pursuing the target R is more desirable in a state xt when P ( c o s t ( R ) > c o s t ( L ) | x t ) < l ( p R μ R - p L μ L ) ( 22 ) Fig 3F illustrates the heat map of the relative desirability values at different states of the policy to reach the right target ( solid gray circle ) , when both targets have the same reward probability pL = pR = 0 . 5 , but μR = 4μL , i . e . reward ( R ) ∼ 0 . 5δ ( reward ( R ) ) + 0 . 5N ( 2 , 1 ) and reward ( L ) ∼ 0 . 5δ ( reward ( L ) ) + 0 . 5N ( 0 . 5 , 1 ) . Similar to the previous scenario , reaching behavior is dominated mostly by the goods-related component and consequently reaching the right target is more desirable than reaching the left target for most states ( reddish areas ) , leading frequently to “winner-take-all” behavior . Several studies have shown that reaching decisions made while acting follow a “delay-and-mix” policy , with the mixing affected by target configuration and task properties [11 , 12 , 22 , 23] . Subjects were trained to perform rapid reaching movements either to a single target or to two equidistant , equiprobable targets ( i . e . , actual target location is unknown prior to movement onset in two-target trials ) . Black and green traces in Fig 4A show single-target trials , characterized by trajectories straight to the target location . Red and blue traces show the delay-and-mix policy for reaches in two-target trials—an initial reaching movement towards an intermediate position between the two stimuli followed by corrective movements after the target was revealed . Relative desirability predicts this behavior ( Fig 3D ) , for equiprobable reward ( scenario 1 ) . In this case , the relative desirability is determined solely by the distance from the current hand position to the targets . Since targets are equidistant , the reaching costs are comparable and hence the two competing policies have about the same desirability values for states between the origin and the target locations ( see the greenish areas in Fig 3D ) . Hence , the weighted mixture of policies produces spatial averaging trajectories ( red and blue traces in Fig 4E ) . Note that each controller i , which is associated with the potential target i , generates an optimal policy πi ( xt ) to reach that target starting from the current state xt . On single-target trials , the actual location of the target is known prior to movement onset and hence the desirability is 1 for the cued target . Consequently the simulated reaches are made directly to the actual target location ( green and black traces in Fig 4E ) . The competition between policies is also modulated by spatial location of the targets [12] . When one of the targets was shifted , reaching trajectories shifted towards a new intermediate position Fig 4B . This behavior is also captured by our framework—perturbing the spatial distribution of the potential targets , the weighted policy is also perturbed in the same direction Fig 4F . This finding is somehow counterintuitive , since the targets are no longer equidistant from the origin and it would be expected that the simulated reach responses would be biased towards the closer target . However , the magnitude of the perturbation is too small to change the action costs enough to significantly bias the competition . More significant are the action costs required to change direction once the target is revealed , and these costs are symmetric between targets . Reaching behavior is also influenced by goods-related decision variables , like target probability . When subjects were informed that the potential targets were not equiprobable , the reach responses were biased towards the target with the highest reward probability [11] . This finding is consistent with relative desirability predictions in scenario 2—targets with higher reward probabilities are more desirable than the alternative options for most of the states . Reward probabilities learned via feedback can also be modeled in the same framework . Instead of informing subjects directly about target probabilities , the experimenters generated a block of trials in which one of the targets was consecutively cued for action [22] . Subjects showed a bias towards the cued target that accumulated across trials ( Fig 4C ) consistent with probability learning . We modeled this paradigm by updating the reward probability using a simple reinforcement learning algorithm ( see S4 Text for more details ) . In line with the experimental findings , the simulated reach responses were increasingly biased to the target location that was consecutively cued for action on the past trials , Fig 4G . Unlike most value computation methods , our approach can make strong predictions for what happens when additional targets are introduced . A previous study showed that by varying the number of potential targets , reaching movements were biased towards the side of space that contains more targets [12] , Fig 4D . Our approach predicts this effect due to normalization across policies . When there are more targets in one hemifield than the other , there are more alternative reaching policies towards this space biasing the competition to that side , Fig 4H . Overall , these findings show that weighting individual policies with the relative desirability values can explain many aspects of human behavior in reaching decisions with competing goals . A good theory should predict not only successful decisions , but also decisions that result in errors in behavior . Experimental studies provide fairly clear evidence that humans and animals follow a “delay-and-mix” behavior even when it appears pathological . A typical example is the “global effect” paradigm that occurs frequently in oculomotor decisions with competing goals . When two equally rewarded targets are placed in close proximity—less than 30° angular distance—and the subject is free to choose between them , saccade trajectories usually end on intermediate locations between targets [24 , 34 , 35] . To test whether our theory can capture this phenomenon , we modeled the saccadic movements to individual targets using optimal control theory ( see S2 Text for more details ) and ran a series of simulated oculomotor decision tasks . Consistent with the experimental findings , the simulated eye movements land primarily in a position between the two targets for 30° target separation ( gray traces in Fig 5A ) , whereas they aim directly to one of them for 90° target separation ( black traces in Fig 5A ) . We visualize the relative desirability of the left target ( i . e . , desirability to saccade to the left target ) at different states , both for 30° and 90° target separation . We followed a similar procedure as for the reaching case but used an ellipse . Particularly , individual saccadic movements are constrained to start at ( 0 , 0 ) and arrive at one of the sequence ( 100 total ) of spatial positions with zero velocity around an ellipse with center intermediate between the two targets ( black star ) , with minor axis twice the distance between the origin and the center of the ellipse , and major axis double the length of the minor axis ( Fig 5B ) . For each position on the ellipse , we generate 100 optimal saccadic movements and evaluate the relative desirability to saccade to the left target ( solid gray circle ) at different states . Fig 5C depicts the heat-map of the relative desirability for 30° target separation . The black traces represent the average trajectories for direct saccadic movements , when only a single target is presented . Notice that regions defined by the starting position ( 0 , 0 ) ( gray square ) and the locations of the targets is characterized by states with strong competition between the two saccadic policies ( greenish areas ) . Consequently the weighted mixture of policies results frequently in spatial averaging movements that land between the two targets . On the other hand , when the targets are placed in distance , such as the 90° case presented in Fig 5D , the targets are located in areas in which one of the policies clearly dominates the other , and therefore the competition is easily resolved . Fig 5E shows examples of saccadic movements ( left column ) with the corresponding time course of relative desirability values to saccade to the left and the right target ( right column ) . The first two rows show trials from the 30° target separation task , where the competition between the two saccadic policies results in global effect ( upper panels ) and saccadic movement to the right target ( middle panels ) . The two policies have about the same relative desirability values at different states resulting in a strong competition . Because saccades are ballistic with little opportunity for correction during the trajectory , competition produces the global effect paradigm . However , if the competition is resolved shortly after saccade onset , the trajectory ends up to one of the targets . On the other hand , when the two targets are placed in distance , the competition is easily resolved and the mixture of the policies generates direct movements to one of the targets ( lower panel ) . These findings suggest that the competition between alternative policies depends on the geometrical configuration of the targets . We quantified the effects of the targets’ spatial distribution to eye movements by computing the percentage of averaging saccades against the target separation . The results presented in Fig 5F ( red line ) indicate that averaging saccades were more frequent for 30° target separation and fell off gradually as the distance between the targets increases ( see the Discussion section for more details on how competition leads to errors in behavior ) . This finding is also in line with experimental results from an oculomotor decision study with express saccadic movements in non-human primates ( green , blue and cyan lines in Fig 5F describe the performance of 3 monkeys [24] ) . In previous sections we considered decisions between multiple competing goals . However , ecological decisions are not limited only to simultaneous goals , but often involve choices between goals with time-dependent values . Time-dependent values mean that some of the goals may spoil or have limited period of worth such that they must be reached within a time window or temporal order . A characteristic example is sequential decision tasks that require a chain of decisions between successive goals . Substantial evidence suggests that the production of sequential movements involves concurrent representation of individual policies associated with the sequential goals that are internally activated before the order is imposed upon them [25 , 36–39] . To model these tasks using our approach , the critical issue is how to mix the individual control policies . State-dependent policy mixing as described previously will dramatically fail , since the desirability values do not take into account the temporal constraints . However , it is relatively easy to incorporate the sequential constraints and time-dependence into the goods-related component of the relative desirability function . We illustrate how sequential decision tasks can be modeled using a simulated copying task used in neurophysiological [25 , 40] and brain imaging studies [41 , 42] . Copying geometrical shapes can be conceived as sequential decisions with goal-directed movements from one vertex ( i . e . , target ) of the shape to another in a proper spatial order . To model this , each controller j provides a policy πj to reach the vertex j starting from the current state . We encode the order of the policies using a time-dependent target reward probability p ( vertex = j|xt ) that describes the probability that vertex j is the current goal of the task at state xt ( see Materials and Methods section for more details ) . In fact , it describes the probability to copy the segment defined by the successive vertices j − 1 and j at a given state xt . We evaluated the theory in a simulated copying task with 3 geometrical shapes ( i . e . , equilateral triangle , square and pentagon ) . Examples of movement trajectories from the pentagon task is shown in Fig 6A . Fig 6B depicts the time course of the relative desirability values of the segments from a successful trial . The desirability of each segment peaks once the model starts copying that segment and falls down gradually , whereas the desirability of the following segment starts rising while copying the current segment . Notice that the competition is stronger for middle segments than the first or the last segment in the sequence . Consequently , errors , such as rounding of corners and transposition errors ( i . e . , copying other segments than the current one in the sequence ) are more frequent when copying the middle segments of the shape , than during the execution of the early or late segments . These simulation results are congruent with studies showing that human/animal accuracy in serial order tasks is better during early or late elements in the sequence [25 , 43] . A characteristic example is illustrated in Fig 6C , in which the competition between copying the “blue” and the “green” segments resulted in an error trial . Notice also that the temporal pattern of desirability values is congruent with populations of neural activity in prefrontal cortex during the copying task that encode each of the segments [25] . The strength of the neuronal population corresponding to a segment predicted the serial position of the segment in the motor sequence , providing a neural basis for Lashley’s hypothesis . Interestingly , the temporal evolution of the population activities resembles the temporal evolution of the relative desirabilities of policies in our theory . This finding provides a direct neural correlate of relative desirability suggesting that the computations in our model are biologically plausible . Finally , Fig 6D and 6E illustrate examples of movement trajectories for copying an equilateral triangle and a square .
We developed our model for cases where the competing options are similar . These are also cases where the relative effort and reward desirabilities are similar . For two options , it means the relative desirabilities would be far from zero or one . Here we consider extreme situations where one option requires much more effort or supplies much less reward . For extreme cases , the relative desirability calculation appears to break down and produces an “indeterminate” form for each alternative option . Here we explain why that happens , and how the indeterminacy is avoided by adding even a tiny amount of noise in implementing the calculation . To illustrate the indeterminacy , consider selecting between an “extremely hard but very rewarding” and an “extremely easy but unrewarding” option . The hard option offers significantly higher reward than the easy option reward ( Hard ) > > reward ( Easy ) , but it requires significantly higher effort to get it than the easy one cost ( Hard ) > > cost ( Easy ) . According to the definition of the relative desirability , the reward-related component of the desirability will approach 1 for the hard option and 0 for the easy option , since P ( reward ( Hard ) > reward ( Easy ) ) = 1 . On the other hand , the effort-related component of the desirability will be 0 for the hard option and 1 for the easy option , since P ( cost ( Hard ) > cost ( Easy ) ) = 1 . The relative reliability multiplies these values and renormalizes , leading to the indeterminate form rD ( option ( 1 ) ) = 0*1/ ( 0*1+1*0 ) = 0/0 and rD ( option ( 2 ) ) = 1*0/ ( 0*1+1*0 ) = 0/0 . In this case the model apparently fails to make a coherent choice . As long as the probability formula for reward and effort are continuous mappings , this indeterminacy will only be experienced in the limit that one option is infinitely harder to get ( inaccessible ) while the accessible option is comparably worthless . However , the indeterminacy is an extreme example of an important class of problems where effort and reward values for the two options are in conflict with each other . Because there is a trade-off associated with reward vs effort neither option is clearly better than the other . While none of the decisions modeled here have extreme conflict , we nevertheless believe that the indeterminacy described above will never occur in a biological decision-making system due to the effects of even tiny amounts of noise on the relative desirability computation . If we assume that desirability values are the brain’s estimate of how “desirable” one option is with respect to alternatives in terms of expected outcome and effort cost , then it is reasonable to assume these estimates are not always precise . In other words , biological estimates of desirability should manifest stochastic errors , which we model by including noise in the estimates . In the S6 Text we show the effect of this noise is profound . For the extreme scenario in which P ( reward ( Hard ) > reward ( Easy ) ) = 1 and P ( cost ( Hard ) > cost ( Easy ) ) = 1 , in the presence of noise the relative desirability of each option is 0 . 5 . Thus , indeterminacy produces a lack of preference—since the “easy” option dominates the “hard” option in terms of effort , but the “hard” option is better than the “easy” option in terms of reward . In general , cases with extreme conflict will produce lack of preference , but these cases are also unstable—small changes in factors affecting the valuation such as the internal states of the subject ( e . g . , hunger level , fatigue level ) can produce large shifts in preference . In the S6 Text , we further discuss the effects of noise in decisions with multiple options . One of the key assumptions in our study is that the brain continuously evaluates the relative desirability—i . e . , the probability that a given policy will result in the highest pay-off with the least effort—in decisions with competing options . Although this idea is novel , experimental studies provide evidence that the brain maintains an explicit representation of “probability of choice” when selecting among competing options ( for a review see [9] ) . For binary perceptual decisions , this probability describes the likelihood of one or another operant response , whereas for value-based decisions it describes the probability that selecting a particular option will result in the highest reward . Classic experimental studies reported a smooth relationship between stimulus parameters and the probability of choice suggesting that the brain translates value information to probabilities when making decisions [57 , 58] . Additionally , neurophysiological recordings in non-human primates revealed activity related to the probability of choice in the lateral intraparietal area ( LIP ) both in “two-alternative force-choice eye movement decisions” and in “value-based oculomotor decisions” . In the first case , the animals performed the random-dot motion ( RDM ) direction discrimination task while neuronal activity was recorded from the LIP [59] . The activity of the LIP neurons reflects a general decision variable that is monotonically related to the logarithm of the likelihood ratio that the animals will select one direction of motion versus the other . In classic value-based decisions , the animals had to select between two targets presented simultaneously in both hemifields [15] . The activity of the LIP neurons is modulated by a number of decision-related variables including the expected reward and the outcome probability . These experimental findings have inspired previous computational theories to model perceptual- and value-based decisions [9] . According to these studies , when the brain is faced with competing alternatives , it implements a series of computations to transform sensory and value information into a probability of choice . The proposed idea of the relative desirability value can be conceived as an extension of these theories taking into account both the expected reward and the expected effort related to a choice . One of the novelties of this theory is that it predicts not only successful decisions , but decisions that result in poor or incorrect actions . A typical example is the “global effect” paradigm that occurs frequently in short latency saccadic movements . When the goal elements are located in close proximity and subjects are free to choose between them , erroneous eye movements usually land at intermediate locations between the goals [24 , 35] . Although the neural mechanisms underlying the global effect paradigm have not been understood fully yet , the prevailing view suggests that it occurs due to unresolved competition between the populations of neurons that encode the movements towards the two targets . Any target in the field is represented by a population of neurons that encodes the movement direction towards its location as a vector . The strength of the population is proportional to the saliency ( e . g . , size , luminance ) and the expected pay-off of the target . When two similar targets are placed in close proximity , the populations corresponding to them will be combined to one mean population with the direction of the vector towards an intermediate location . If one of the targets is more salient or provide more reward than the other , the vector is biased to this target location . Since subjects have to perform saccadic movements to one of the targets , the competition between the two populations has to be resolved in time by inhibiting one of them . The time to suppress the neuronal activity that encodes one of the alternatives may be insufficient for short latency saccades resulting in averaging eye movements . Our findings are consistent with this theory . The strength of the neuronal population is consistent with relative desirability of the policy that drives the effector directed to the target . When the two equally rewarded targets are placed in close proximity , the two policies generate similar actions . Given that both targets are attached with the same goods-related values , the relative desirability of the two policies are about the same at different states , resulting in a strong competition . Because saccades are ballistic with little opportunity for correction during movement , the competition produces averaging saccades . On the other hand , placing the two targets in distance , the two saccadic policies generate dissimilar actions and consequently the competition is easier to be resolved in time . Competition between policies in closely aligned goals can also explain errors in sequential decision tasks that involve serial order movements as described by Lashley [36] . The key idea in Lashley’s pioneer work ( 1951 ) is that the generation of serial order behavior involves the parallel activation of sequence of actions that are internally activated before each of the actions are executed . The main line of evidence of this hypothesis was the errors that occur frequently in serial order tasks , such as speech [37] , typing [38] , reaching [39] and copying of geometrical shapes [25] . For instance , a common error in typing and speaking is to swap or transpose nearby letters , even words . Lashley suggested that errors in sequential tasks would be most likely to occur when executing nearby elements within a sequence . Recent neurophysiological studies provide the neural basis of the Lashley’s hypothesis showing that the serial characteristics of a sequence of movements are represented in an orderly fashion in the prefrontal cortex , in time before the start of drawing [25 , 40] . Training monkeys to copy geometrical shapes and recording the activity of individual neurons in the prefrontal cortex , the experimenters were able to identify populations of neurons that encode each of the segments [25] . The strength of the neuronal population corresponding to a segment predicted the serial position of the segment in the motor sequence . Interestingly , the temporal evolution of the strength of the segment representation during the execution of the trajectories for copying the shapes resembles the temporal evolution of the relative desirabilities of policies in our theory . This finding suggests that the strength of the neuronal population of a particular segment may encode the relative desirability ( or components of the desirability ) of copying that segment at a given time with respect to the alternatives . This hypothesis is also supported by error analysis in the serial order tasks , which showed that errors more frequently occurred when executing elements with nearly equal strength of representation . In a similar manner , our theory predicts that when two policies have about equal relative desirabilities over extended periods of the movement , the competition between them may lead to errors in behavior . Finally , our theory provides a conceptual alternative in understanding important aspects of neurological disorders that cause deficits in choice behavior , such as the spatial extinction syndrome . This syndrome is a subtle form of hemispatial neglect that occurs frequently after brain injury . It is characterized by the inability to respond to stimuli in the contralesional hemifield , but only when a simultaneous ipsilesional stimulus is also presented [60] . Recent studies reported contralesional bias that reminiscent the extinction syndrome , in oculomotor decision tasks after reversible pharmacological inactivation of the LIP [48] and the Pulvinar [61] in monkeys . According to our theory , this effect could be related to a deficit in value integration after inactivation , rather than simply sensory attention deficit . In sum , decisions require integrating both good values and action costs , which are often time and state dependent such that simple approaches pre-selection of goals or fixed weighted mixture of policies cannot account for the complexities of natural behavior . By focusing on a fundamental probabilistic computation , we provide a principled way to dynamically integrate these values that can merge work on decision making with motor control . | Choosing between alternative options requires assigning and integrating values along a multitude of dimensions . For instance , when buying a car , different cars may vary for their price , quality , fuel economy and more . Solving this problem requires finding a common currency to allow integration of disparate value dimensions . In dynamic decisions , in which the environment changes continuously , this multi-dimensional integration must be updated over time . Despite many years of research , it is still unclear how the brain integrates value information and makes decisions in the presence of competing alternatives . In the current study , we propose a probabilistic theory that allows dynamically integrating value information into a common currency . It builds on successful models in motor control and decision-making . It is comprised of a series of control schemes with each of them attached to an individual goal , generating an optimal action-plan to achieve that goal starting from the current state . The key novelty is the relative desirability computation that integrates good- and action- values to a single dynamic variable that weighs the individual action-plans as a function of state and time . By dynamically integrating value information , our theory models many key results in movement decisions that have previously eluded a common explanation . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Dynamic Integration of Value Information into a Common Probability Currency as a Theory for Flexible Decision Making |
A fundamental challenge in neuroscience is to understand what structure in the world is represented in spatially distributed patterns of neural activity from multiple single-trial measurements . This is often accomplished by learning a simple , linear transformations between neural features and features of the sensory stimuli or motor task . While successful in some early sensory processing areas , linear mappings are unlikely to be ideal tools for elucidating nonlinear , hierarchical representations of higher-order brain areas during complex tasks , such as the production of speech by humans . Here , we apply deep networks to predict produced speech syllables from a dataset of high gamma cortical surface electric potentials recorded from human sensorimotor cortex . We find that deep networks had higher decoding prediction accuracy compared to baseline models . Having established that deep networks extract more task relevant information from neural data sets relative to linear models ( i . e . , higher predictive accuracy ) , we next sought to demonstrate their utility as a data analysis tool for neuroscience . We first show that deep network’s confusions revealed hierarchical latent structure in the neural data , which recapitulated the underlying articulatory nature of speech motor control . We next broadened the frequency features beyond high-gamma and identified a novel high-gamma-to-beta coupling during speech production . Finally , we used deep networks to compare task-relevant information in different neural frequency bands , and found that the high-gamma band contains the vast majority of information relevant for the speech prediction task , with little-to-no additional contribution from lower-frequency amplitudes . Together , these results demonstrate the utility of deep networks as a data analysis tool for basic and applied neuroscience .
A central goal of neuroscience is to understand what and how information about the external world ( e . g . , sensory stimuli or behaviors ) is present in spatially distributed , dynamic patterns of brain activity . At the same time , neuroscience has been on an inexorable march away from the periphery ( e . g . , the retina , spinal cord ) , seeking to understand higher-order brain function ( such as speech ) . The methods used by neuroscientists are typically based on simple linear transformations , which have been successful predictors in early processing stages of the nervous system for simple tasks [1–3] . However , linear methods are limited in their ability to represent complex , hierarchical , nonlinear relationships [4] , which are likely present in the neural activity of higher-order brain areas . This linear restriction may not only limit the predictive accuracy of models but may also limit our ability to uncover structure in neural datasets . Multilayer deep networks can combine features in nonlinear ways when making predictions . This gives them more expressive power in terms of the types of mappings they can learn at the cost of more model hyperparameters , more model parameters to train , and more difficult training dynamics [5] . Together with the recent success of deep learning in a number of fields including computer vision , text translation , and speech recognition [6–8] , the ability of deep networks to learn nonlinear function from data motivates their use for understanding neural signals . The success of deep learning in classic machine learning tasks has spurred a growth of applications into new scientific fields . Other nonlinear methods such as random trees/forests can also be used on nonlinear neural data but often require more feature selection/reduction and are not typically used on data with thousands or tens of thousands of features [9] . Deep networks have recently been applied as classifiers for diverse types of physiological data including electromyographic ( EMG ) , electroencephalographic ( EEG ) , and spike rate signals [10–13] , on stimulus reconstruction in sensory regions using electrocorticography ( ECoG ) [14] , as models for sensory and motor systems [15–19] . Compared to datasets used in traditional machine learning , neuroscientif datasets tend to be very small . As such , models in neuroscience tend to be smaller ( fewer layers , units per layers ) and in this sense more similar to neural networks from previous decades . However , modern deep learning techniques such as ReLUs , Dropout , and optimization algorithms like Nesterov momentum are crucial to train them to high held-out performance . While these studies have demonstrated the superior performance of deep networks as black-box predictors , the utilization of deep networks to gain understanding into brain computations is rare . Whether deep networks can be used to elucidate the latent structure of scientific and neuroscientific datasets is still an open question . Vocal articulation is a complex task requiring the coordinated orchestration of several parts of the vocal tract ( e . g . , the larynx , tongue , jaw , and lips ) . To study the neural basis of speech requires monitoring cortical activity at high spatio-temporal resolution ( on the order of tens of milliseconds ) over large areas of sensorimotor cortex ( ∼1300mm2 ) [20] . Electrocorticography ( ECoG ) is an ideal method to achieve the simultaneous high-resolution and broad coverage requirements in humans . Using such recordings , there has been a surge of recent efforts to understand the cortical basis of speech production [20–25] . For example , analyzing mean activity , Bouchard et . al . [20] demonstrated , much in the spirit of Penfield’s earlier work [26] , that the ventral sensorimotor cortex ( vSMC ) has a spatial map of articulator representations ( i . e . lips , jaw , tongue , and larynx ) that are engaged during speech production . Additionally , it was found that spatial patterns of activity across the vSMC network ( extracted from trial average activity with principal components analysis at specific time points ) organized phonemes along phonetic features emphasizing the articulatory requirements of production . Understanding how well cortical surface electrical potentials ( CSEPs ) capture the underlying neural processing involved in speech production is important for revealing the neural basis of speech and improving speech decoding for brain-computer interfaces [27 , 28] . Previous studies have used CSEPs and linear or single layer models to predict speech categories [23 , 29–32] , or continuous aspects of speech production ( e . g . , vowel acoustics or vocal tract configurations ) [22 , 25] , with some success . However , given the challenge of collecting large number of samples across diverse speech categories , it is not clear that we should expect high performance from deep networks for speech classification . Exploring the use of deep networks to maximally extract information for speech prediction is not only important for brain machine interfaces which restore communication capabilities to people who are “locked-in” , but also for identifying cortical computations which are the underlying basis for speech production . In general , understanding information content across neural signals , such as different frequency components of CSEPs , is an area of ongoing research [33–37] . A number of studies have found relationships between different frequency components in the brains electrical potentials . These can take the form of phase and amplitude structure of beta ( β ) waves [38 , 39] or correlations between lower frequency oscillations and spiking activity or high gamma ( Hγ ) activity [37 , 40] . One observation is that β band ( 14-30Hz ) amplitude and coherence [33 , 41] often decreases during behavior , when the state is changing [42] . This has lead to the interpretation that β may be serving a “maintenance of state” function . However , often these effects are not differentiated between functional areas that are active versus inactive during behavior . Indeed , in other contexts , aggregation has been shown to mask structure in neural signals [43] . The somatotopic organization of speech articulator control in human vSMC , and the differential engagement of these articulators by different speech sounds , potentially provides the opportunity to disentangle these issues . Furthermore , classifying behaviors , such as speech , from CSEPs can be used as a proxy for information content in a signal , obfuscating the interpretation of the results . However , this is often done using linear methods , which may not be able to take full advantage of the information in a signal . Since deep networks are able to maximize classification performance , they are an ideal candidate for comparing information content across neural signals . In this work , we investigated deep networks as a data analytics framework for systems neuroscience , with a specific focus on the uniquely human capacity to produce spoken language . First , we show that deep networks achieve superior classification accuracy compared to linear models , with increased gains for increasing task complexity . We then “opened the black box” and used the deep network confusions to reveal the latent structure learned from single trials , which revealed a rich , hierarchical organization of linguistic features . Since deep networks classified speech production from Hγ activity with higher accuracy that other methods , they are also candidates for determining the relative information content across neural signals . We explored the cross-frequency amplitude-amplitude structure in the CSEPs and discovered a novel signature of motor coordination in β-Hγ coupling . Using deep networks , we then show that although there is information relevant to speech production in the lower frequency bands , it is small compared to Hγ . Critically , the lower frequency bands do not add significant additional information about speech production about and beyond Hγ . Furthermore , the correlations are not tightly related to overall information content and improvements in accuracy . Together , these results demonstrate the utilization of deep networks not only as an optimal black-box predictor , but as a powerful data analytics tool to reveal the latent structure of neural representations , and understanding the information content of different neural signals .
The experimental protocol , collection , and processing of the data examined here have been described in detail previously [20–22] . The experimental protocol was approved by the Human Research Protection Program at the University of California , San Francisco . Briefly , four native English speaking human subjects underwent chronic implantation of a subdural electrocortigraphic ( ECoG ) array over the left hemisphere as part of their clinical treatment of epilepsy . The subjects gave their written informed consent before the day of surgery . The subjects read aloud consonant-vowel ( CV ) syllables composed of 19 consonants followed by one of three vowels ( /a/ , /i/ or /u/ ) , for a total of 57 potential consonant-vowel syllables . Subjects did not produce each CV in an equal number of trials or produce all possible CVs . Across subjects , the number of repetitions per CV varied from 10 to 105 , and the total number of usable trials per subject was S1: 2572 , S2: 1563 , S3: 5207 , and S4: 1422 . CVs for which there was not enough data to do cross-validation ( fewer than 10 examples ) were excluded per-subject . Cortical surface electrical potentials ( CSEPs ) were recorded directly from the cortical surface with a high-density ( 4mm pitch ) , 256-channel ECoG array and a multi-channel amplifier optically connected to a digital signal processor ( Tucker-Davis Technologies [TDT] , Alachua , FL ) . The time series from each channel was visually and quantitatively inspected for artifacts or excessive noise ( typically 60 Hz line noise ) . These channels were excluded from all subsequent analysis and the raw CSEP signal from the remaining channels were downsampled to 400 Hz in the frequency domain and then common-average referenced and used for spectro-temporal analysis . For each useable channel , the time-varying analytic amplitude was extracted from 40 frequency domain , bandpass filters ( Gaussian filters , logarithmically increasing center frequencies and semi-logarithmically increasing band-widths , equivalent to a frequency domain Morlet wavelet ) . The amplitude for each filter band was z-scored to a baseline window defined as a period of time in which the subject was silent , the room was silent , and the subject was resting . Finally , the amplitudes were downsampled to 200 Hz . For each of the bands defined as: theta [4-7 Hz] , alpha [8-14 Hz] , beta [15-29 Hz] , gamma [30-59 Hz] , and high gamma [70-150 Hz] , individual bands from the 40 Gaussian bandpassed amplitudes were grouped and averaged according to their center frequencies . The lower frequency features are all highly oversampled at the Hγ rate of 200 Hz . To make comparisons across frequency bands more interpretable , control for potential overfitting from training on oversampled signals , and to reduce the computational complexity of training deep networks with concatenated input features , we downsampled each of the lower frequency bands in time so that the center frequency-to-sampling rate ratio was constant ( ratio = 112 . 5/200 ) for each band . Given limited data , deep networks are tasked with deciding whether a change across input features is relevant or irrelevant for prediction . The lower frequency bands are highly oversampled at 200 Hz , however , the higher frequencies will not have exactly zero amplitude do to numerical noise even though these are irrelevant signals . Downsampling the bands to a fixed ratio makes comparing CV decoding accuracy across frequency bands more interpretable . Based on previous results [20–22] , we focused on the electrodes in the ventral sensorimotor cortex ( vSMC ) . The activity for each of the examples in our data set was aligned to the acoustic onset of the consonant-to-vowel transition . For each example , a window 0 . 5 seconds preceding and 0 . 79 seconds following the acoustic onset of the consonant-to-vowel transition was extracted . The mean of the first and last ∼ 4% time samples was subtracted from the data per electrode and trial ( another form of amplitude normalization that is very local in time ) . This defined the z-scored amplitude that is used for subsequent analyses . Supervised classification models often find their model parameters , Θ ^ , which minimize the negative log-likelihood of the training data and labels , {x ( i ) , y ( i ) } , under a model which gives the conditional probability of the labels given the input data Θ ^ = arg min Θ − log P ( Y | X ; Θ ) , { x ( i ) , y ( i ) } . ( 1 ) Deep networks typically parametrize this conditional probability with a sequence of linear-nonlinear operations . Each layer in a fully-connected network consists of an affine transform followed by a nonlinearity: h 1 = f ( w 1 · x + b 1 ) , h i = f ( w i · h i − 1 + b i ) , with Θ = { w 1 , … , w n , b 1 , … , b n } ( 2 ) where x is a batch of input vectors , wi and bi are trainable parameters ( weights and biases , respectively ) for the ith layer , hi is the ith hidden representation , and f ( ⋅ ) is a nonlinearity which can be chosen during hyperparameter selection . Single layer classification methods , such as multinomial logistic regression , are a special case of deep networks with no hidden representations and their corresponding hyperparameters . For the fully-connected deep networks used here , the CSEP features were rasterized into a large feature vector per-trial in a window around CV production . These feature vectors are the input into the first layer of the fully connected network . The feature dimensionality is the number of electrodes by 258 time points which corresponds to Subject 1: 22 , 188 , Subject 2: 20 , 124 , Subject 3: 21 , 414 , and Subject 4: 25 , 542 features . The final layer non-linearity is chosen to be the softmax function: P ( y ^ i ) = softmax ( h i ) = exp ( h i ) ∑ j exp ( h j ) ( 3 ) where hi is the ith element of the hidden representation . This nonlinearity transforms a vector of real numbers into a vector which represents a one-draw multinomial distribution . It is the negative log-likelihood of this distribution over the training data which is minimized during training . To train and evaluate the networks , the trials were organized into 10 groupings ( folds ) with mutually exclusive validation and test sets and 80-10-10% splits ( training , validation , testing ) . Since some classes may have as few as 10 examples , it was important to split each class proportionally so that all classes were equally distributed . Training terminated when the validation accuracy did not improve for 10 epochs and typically lasted about 25 epochs . Theano , Pylearn2 , and Scikit-learn [44–46] were used to train all deep and linear models . As baseline models , we used multinomial logistic regression . Logistic regression required no additional dimensionality reduction and had the highest classification accuracy compared to other linear classifiers , i . e . linear support vector machines and linear discriminant analysis on the Hγ features ( 10 . 4 ± 6 . 7% and 16 . 0 ± 10 . 0% respectively compared to 28 . 0 ± 12 . 9% for logistic regression ) . Additionally , the conditional class distribution used in logistic regression ( multinomial ) is the same as the one used for deep networks , which facilitated comparison of confusions . Neuroscientists commonly study the model/confusions of linear analysis methods to gain insight into the structure of neural data . Deep networks can learn high dimensional , nonlinear features from data . Here , these features are learned by training the networks to perform classification , i . e . maximize P ( Y ^ i | X i ; Θ ) where the subscript i indicates true class membership . It has been shown that these features contain more information than the thresholded multinomial classification prediction [50 , 51] . The off-diagonal values: P ( Y ^ i | X j ; Θ ) , i ≠ j , in this learned distribution represent prediction uncertainty for a given CSEP measurement . Uncertainty is learned during the training process and larger pairwise uncertainty between class labels means that the model has a harder time distinguishing those classes . Since the uncertainty ( similarity ) is not encoded in the supervised labels , this means that the neural data for those class labels is more similar . To gain insight into the nature of the vSMC neural activity , we analyzed the structure of deep network predictions . The mean network prediction probabilities on the test set are used as features for each CV . A dendrogram was computed from the hierarchical clustering ( Ward’s method ) of these features . To aid visualization of these results , a threshold in the cluster distance was chosen by visual inspection of when the number of clusters as a function of distance rapidly increased , and the linguistic features were labeled by hand . The CV order from this clustering was used to order the features in the soft-confusion matrix and accuracy per CV . The soft confusion matrix shows mean network prediction probabilities on the test set rather than the aggregated thresholded predictions often shown in confusion matrices . To compare the articulatory features and the deep network features quantitatively across subjects , pairwise distances between CVs were computed in both the articulatory and deep network spaces ( see S1 Fig for articulatory features ) . These pairwise distances were then correlated per for each CV and subject and articulatory grouping . To examine the relationship between the amplitudes of different frequency components of recorded CSEPs , we first performed a correlation analysis . For this analysis , the data was trial-averaged per CV then organized into a data-tensor , DCV , frequency , electrode , time . The frequency bands were then either used individually or aggregated into canonical frequency components , such as Hγ D ¯ ( H γ ) CV , electrode , time = ⟨ D ¯ CV , frequency , electrode , time ⟩ frequency ∈ H γ . ( 7 ) D ¯ ( H γ ) CV , electrode , time was correlated across time at 0 ms lag with each of the 40 Gaussian bandpassed amplitudes averaged across CVs and electrodes . The correlation between D ¯ ( H γ ) CV , electrode , time and D ¯ ( β ) CV , electrode , time was computed and histogrammed across CVs and electrodes . The average Hγ power was averaged in a window 70 ms before and 140 ms after the CV acoustic transition and histogrammed across CVs and electrodes . This window was chosen as it is the most active and informative time period for consonants and vowels . Since the ECoG grid covers a large functional area of vSMC and the CV task differentially engages articulators for different consonant and vowels , the correlations can be computed independently for “active” versus “inactive” electrodes for each CV ( averaged across trials ) . To define active and inactive electrode groups for a band , B and Hγ , first , the B -H γ amplitude-amplitude correlation , C ( B , H γ ) and average Hγ amplitude , A ( Hγ ) , with positive average amplitude ( greater than baseline ) are used to fit one linear model with ordinary least-squares regression m ^ , b ^ = arg max m , b ∑ i j ( C i j ( B , H γ ) − ( m A i j ( H γ ) + b ) ) 2 ( 8 ) for all electrodes , i , and CVs , j . The electrodes were then divided into “active” and “inactive” per CV by thresholding the average Hγ activity where the linear fit predicted 0 correlation . A thresh ( H γ ) = − b ^ m ^ . ( 9 ) Electrodes with average Hγ activity above threshold were active , and those with lower average Hγ activity were inactive . The active and inactive electrodes per CV were separated and D ¯ ( H γ ) CV , electrode , time was correlated across time at 0 ms lag with each of the 40 Gaussian bandpassed amplitudes averaged across CVs and electrodes independently for the active and inactive electrodes and for each subject . An extended sets of lower frequency features per trial were used in addition to the Hγ features for each of the theta , alpha , low beta , high beta , and gamma bands . The lower frequency amplitudes are highly oversampled at 200 Hz ( the Hγ sampling frequency ) , and overfitting due to this mismatch will confound the interpretations of signal content . To minimize overfitting , the lower frequency amplitudes were downsampled as described in the Signal processing subsection . For each frequency band , fully-connected deep networks were trained first on the individual bands’s features and then with the band’s features concatenated with the Hγ features . Deep network training was done in the same manner at the networks trained solely on Hγ features . The resulting classification accuracies were then compared with the baseline Hγ classification accuracy and then with the band’s features concatenated with the Hγ features .
Despite being able to mathematically specify the computations happening everywhere in the model , deep networks are often described as “black boxes” . What deep networks learn and how it depends on the structure of the dataset is not generally understood . This means that deep networks currently have limited value for scientific data analysis because their learned latent structure cannot be mapped back onto the structure of the data . Many current uses of deep networks in scientific applications rely on their high accuracy and do not inspect the network computations [19 , 53 , 54] , although there are results in low dimensional networks [15] and early sensory areas [18] . Nevertheless , deep networks’ ability to consume huge datasets without saturating performance means that expanding their use in science is limited by our understanding of their ability to learn about the structure of data . For the dataset consider in this work , previous studies have shown that an articulatory hierarchy can be derived from the trial-averaged Hγ amplitude using principal components analysis at hand-selected points in time [20] . Note that the articulatory structure of the consonants and vowels are not contained in the CV labels nor are the individual consonant or individual vowel labels due to the CVs being encoding in a one-hot fashion , i . e . , /ba/ ( label = 0 ) is as different from /bi/ ( label = 1 ) as it is from /gu/ ( label = 8 ) according to the CV labels even though they share a consonant ( likewise for shared vowels ) . To explore whether deep networks can infer this latent structure from the training data , we examined the structure of network output to better understand the organization of deep network syllable representations extracted from vSMC . Deep networks used for classification predict an entire distribution over class labels for each data sample . This learned distribution has been shown to be a useful training target in addition to the thresholded class labels [50 , 51] . We clustered these learned representations and compared them to articulatory representations of the CVs . The dendrogram resulting from agglomerative hierarchical clustering on the trial averaged output of the softmax of the deep network ( i . e . , before thresholding for classification ) averaged across subjects shows clusters spread across scales ( Fig 4A ) . A threshold was chosen by inspection of when the number of clusters as a function of cutoff distance rapidly increased ( Fig 4B ) and used to color the highest levels of the hierarchy . At the highest level , syllables are confused only within the major articulator involved ( lips , back tongue , or front tongue ) in the syllable . This is followed by a characterization of the place of articulation within each articulator ( bilabial , labio-dental , etc . ) . At the lowest level there seems to be a clustering across the consonant constriction degree and vowel categories that capture the general shape of the vocal tract in producing the syllable . When ordered by this clustering , the soft confusion matrix ( Fig 4C ) resulting from the average output of the final layer softmax shows block-diagonal structure corresponding to the articulatory hierarchy . In contrast , deep networks trained on the mel-cepstral coefficients and their time-differences ( similar dimensionality to the Hγ amplitude ) show a largely inverted hierarchy , results which mirror those found in more general studies of deep network processing of spoken acoustics [55] ( See S5 Fig for this analysis on the data presented here ) . There is a large amount of variation in the per-CV accuracies ( Fig 4D ) . This hierarchy can be quantified by comparing the space of deep network prediction probabilities and the space of articulatory features associated with each CV . This comparison was made by correlating pairwise CV distances in these two features spaces across all pairs of CVs . The resulting structure of correlations is consistent with an articulatory organization in vSMC ( Fig 4C ) . The major articulators feature distances are most correlated with the distances between CVs in deep network space , then consonant constriction location , and finally consonant constriction degree and vowel . Together , these results show that deep networks trained to classify speech from Hγ activity are learning an articulatory latent structure from the neural data . Qualitatively similar hierarchies can be derived using PCA and logistic regression . Indeed , this structure is in agreement with previous analyses of mean spatial patterns of activity at separate consonant and vowel time points [20] while allowing the consonants and vowels to be part of the same hierarchy . However , the deep network hierarchy has larger correlations and more separation between levels than the hierarchy derived from the Logistic regression model ( shown in S6 Fig ) . Together , these results demonstrate the capacity of deep networks to reveal underlying structure in single-trial neural recordings . Complex behaviors , such as speech , involve the coordination of multiple articulators on fast timescales . These articulators are controlled by spatially distributed functional areas of cortex . Lower frequency oscillations have been proposed as a coordinating signal in cortex . Previous studies have reported movement- or event-related beta ( β ) -Hγ desynchronization or decorrelation [33 , 34 , 42] . The differential structure of these correlations across tasks and functions areas is not commonly analyzed . Since cortex often shows sparse and spatially-differentiated activity across tasks [22] , averaging over electrodes and tasks may obscure structure in the cross-frequency relationships . The CV task and grid coverage allow average neural spectrograms ( zscored amplitude as a function of frequency and time ) to be measured at two electrodes during the production of the syllable \ga\ ( Fig 5A and 5B , median acoustic spectrogram is shown above ) . In order to investigate this , we measured cross frequency amplitude-amplitude coupling ( correlation ) for individual lower frequency bands and Hγ . We also examine the aggregate β band . Some previous studies attempt to distinguish band-limited and broadband signals in lower frequencies , e . g . , β [56–58] . However , methods for distinguishing these signals are generally not applied to high sampling rate signals and often require hand-tuning and is an ongoing area of research . As such , here , we are simply looking at correlations between bandpassed signals and not estimating and removing any broadband components . Further modeling would be needed to interpret these signals as correlations between biophysical sources ( see Discussion for discussion ) . Initially , we pool results across all electrodes and CVs in order to replicate methods from previous studies . The Hγ and β amplitudes show a diverse set of temporal relationships in these regions ( Fig 5C and 5D ) . Across frequencies , Hγ correlation is positive for low frequencies ( < 15Hz ) , then we see negative and near-zero correlations between Hγ and the β range across subjects , and finally the correlation rises for the γ range ( 30–59 Hz ) as the frequencies approach Hγ ( Fig 5E ) . However , these mean correlations mask a broad range of Hγ-β correlations ( Fig 5F ) across Hγ activity ( across CVs and electrodes ) . This includes a large number of positive correlations . Similarly , although most of the amplitudes measured are smaller than baseline ( Fig 5G ) , there is a long tail to amplitudes larger than baseline ( above 0 ) . This diversity of correlations and amplitudes across CVs and electrodes indicates there is potentially substructure in the data that is being averaged over . This motivates a disaggregated analysis of the amplitude-amplitude correlations . Naïvely , one might expect to see different cross-frequency relationships in areas that are actively engaged in a task compared to area which are not engaged . The broad coverage of the ECoG grid and the diversity of articulatory movements across the consonants and vowels in the task allow us to investigate whether there is substructure in the amplitude-amplitude cross frequency correlations . In order to investigate this , we grouped the Hγ activity for each electrode and CV into “active” and “inactive” groups based on the average Hγ power and computed correlations for these two groups . For the two subjects with high accuracy , we observe a positive correlation between Hγ power and Hγ-β correlation ( Fig 6A ) . For the two subjects with low CV classification accuracy , we observe a generally negative correlation between Hγ power and Hγ-β amplitude ( Fig 6B ) . The Hγ correlation can be recomputed separately for active and inactive electrodes per CV . For the subjects with high CV classification accuracy ( Subjects 1 and 4 ) , we find a novel signature of motor coordination in the active electrodes: a positive correlation in the β frequency band ( Fig 6C , lines with white centers ) . This is in contrast to the inactive electrodes , which show small or negative correlation ( Fig 6C , solid lines ) which is similar to the aggregated results ( Fig 5E ) . For the two subjects with low CV classification accuracy ( Subjects 2 and 3 ) , the disaggregated results ( Fig 6D ) show less dichotomous structure . Overall , we find that there is structure across bands in addition to cross-frequency relationship with the Hγ band which has been used in the preceding classification analysis . As far as we are aware , this is the first observation of dichotomous amplitude-amplitude cross frequency correlation during behavior . This observation was only possible because of the broad functional coverage of the ECoG grid and the diverse behaviors represented in the CV task . The gamma and Hγ band-passed CSEP amplitudes are commonly used both on their own and in conjunction with other frequency bands for decoding produced and perceived speech in humans due to their observed relation to motor and sensory tasks [14 , 22 , 23 , 31 , 59 , 60] . Other frequency bands have been shown to have amplitude or phase activity which is correlated with Hγ amplitude or spiking activity [37–40] . Indeed , in the data used in this study , we find amplitude-amplitude correlation structure between Hγ and lower frequency bands . Although these correlations imply that information is shared between Hγ and other CSEP frequency bands , it is not known whether the other bands contain additional information about motor tasks beyond Hγ or whether the information is redundant . In order to understand the relative information content in CSEP frequency bands , we classified CVs from two different sets of features . Linear classification methods would not give a satisfactory answer to this question since they are limited to simple hyper-plane segmentation of the data which may trivially lead to the result of no information . Indeed , since we have shown that deep networks can outperform linear methods when classifying from Hγ , they are also candidates for showing whether there is any relevant information in these bands . For the theta , alpha , beta , high beta , and gamma bands , each band’s features were first used for classification and then concatenated with the Hγ features and used for classification . The raw classification accuracy and improvement beyond Hγ are two measures that give insight into information content in the other bands . Fig 7 shows the accuracies , normalized to chance , across the four subjects . Fig 7A shows the classification accuracies across subjects for single band features . Across subjects , all single bands have CV classification accuracies greater than chance , although subject-to-subject variation is observed . Although this is significantly above chance , the ranges of improvements for the subject means range between 1 . 5x to 2x chance , a small accuracies compared to Hγ accuracies which ranged from 6x to 21x chance . For the single band features , accuracy above chance implies that there is relevant information about the task in the bands . Fig 7B shows the chance in classification accuracy relative to Hγ accuracy , normalized to chance . No bands see a significant improvement in accuracy over the baseline accuracy obtained by classifying from Hγ . Indeed , all measured mean changes in accuracy are smaller than the cross-validation standard deviations for the Hγ accuracy . Together , these results show that there is task-relevant information in lower frequency bands , but the information is largely redundant to the information contained in the Hγ amplitude . The correlations observed in Figs 5 and 6 imply that there is some shared information between the lower frequency bands and the Hγ band . However , the classification accuracies from Hγ alone ( Fig 3 ) are much higher than any other individual frequency band and are not improved by the addition of extra features from lower frequency bands . This shows that the high frequency CSEPs ( Hγ band ) , which is commonly used in motor decoding , are highly informative signals .
The structure or information content of neural data is often estimated by regressing neural features against known features in the stimulus or behavior . Traditionally , this has been done with linear models , which are often poorly matched to the structure of this relationship . Here , we have shown that deep networks trained on high gamma ( Hγ ) cortical surface electrical potentials ( CSEPs ) can classify produced speech with significantly higher accuracy than traditional linear or single layer models . When classifying syllables , deep networks achieved state-of-the-art accuracy and channel capacity: for the subject with higher accuracy , this was 55 . 1% and 3 . 92 bits per syllable . At word durations from Mugler et al . [23] and one CV syllable per word duration , 3 . 92 bits per syllable corresponds to 7 . 5 bits per second or 75 words per minute [61] . This could also be combined with a language model to improve accuracy in clinical applications [31] towards the eventual goal of natural spoken speech rates ( 250-600 words per minute ) . Generally , we expect that as neuroscientific datasets grow , more modern deep learning techniques and architectures will be used for higher precision such as Residual layers [6] , variational auto-encoders [62] , and recurrent models for timeseries [7] . Together , these results show that deep networks are a promising analytic platform for brain-computer interface ( BCI ) for speech prosthetics , an application where high accuracy and high training sample efficiency are crucial . Since deep networks are highly parameterized nonlinear models , their online interactions with learning may be more complex than typical methods [63] . Studying how deep networks behave in an online BCI will be important future step in integrating them into clinical settings . Training the deep networks described here to high accuracy required an extensive hyperparameter search over model architectures including layer numbers , layer dimensions , and nonlinearity , along with optimization hyperparameters like learning rate , momentum decay , and dropout fraction . In general , the optimal hyperparameters may depend on the latent structure of the dataset being used , however , they may also depend strongly on the size of the dataset in terms of the number of samples or the dimensionality of each sample . Understanding the relationship between the optimal structure of deep networks and the structure of datasets is a future direction of research . We observed classification accuracies were highest , both relative to chance and linear models , for consonant-vowel syllables compared to the consonants or vowels individually . This is consistent with previous reports on the presence of both anticipatory and perseverative coarticulation effects in vSMC ( see also Fig 1E ) [21 , 23] . Coarticulation refers to the fact that , at a behavioral level , the production of speech phonemes is systematically influenced by the surrounding phonemic context . For communication prosthetics , one might hope to decode the most atomic units , phonemes , and then express the combinatorial complexity of language through combinations of the small number of phonemes . Combined with other studies , the results presented here indicate that coarticulation is a feature of speech motor control that must be accounted for in BCIs . In contrast to many commercial applications of deep learning , where optimizing prediction accuracy is often the primary goal , in science , it is also desirable to extract latent structure from the data to advance understanding . In the context of the current study , we used deep networks to determine which features of speech production were extracted from the neural activity to solve the classification task . Examination of the consonant-vowel confusions made by the deep networks reveal the underlying articulatory organization of speech production in the vSMC . At the highest level , the deep networks cluster the CVs into the major articulator involved in forming the consonant , i . e . lips , front tongue , or back tongue . The consonant constriction location , e . g . teeth-to-lips versus lips , is in the intermediate level of the hierarchy . Finally , consonant constriction degree and vowel are clustered at the lowest level of the hierarchy . Crucially , the consonant articulatory hierarchy is not present in the CV labels which means that the deep network is extracting this hierarchy from noisy , single-trial CSEPs during training . The articulatory organization we find is consistent with previous studies , which used PCA on the trial-averaged data at specific points in time [20] . However , we note that , while consistent with previous findings , the hierarchy observed here reflects structure across consonants and vowels together . This could not have been examined with the previous methodology , which required analyses at separate time points . In this way , deep networks were able to extract novel , more general structure from the data , and did so with much less human supervision . As with many studies of human ECoG , there was substantial variability across subjects . Subjects 1 and 4 had the highest CV classification accuracy from Hγ and also showed similar patterns of Hγ correlations with lower frequencies ( Fig 5 ) as well as Hγ-β correlation distinctions in active versus inactive electrodes ( Fig 6 ) . Subjects 2 and 3 had lower accuracy and had less consistently structured Hγ correlations . While the precise nature of cross-subject variability is unknown , likely extrinsic contributors are uncontrollable variation in the degree of contact of electrodes with cortex which could impact frequency specific SNR , differences in variance across recording sessions blocks , or degree of subject engagement in task . Further intrinsic sources of variability could include the lack or presence of particular articulator representations in the recorded activity or differing levels of broadband signals in lower frequency bands . However , the frequency specific bump in the β range observed in Subjects 1 and 4 is unlikely to be explained by a change in broadband power in active electrodes . This would require the power of the broadband signal to be mainly found in the β range and not in the frequencies on either side , which is not consistent with broadband power fluctuations . Interestingly , we found no clear relationship between CV decoding accuracy and the number of trials , suggesting that the variance is due to differences with the underlying signal and not overfitting . Developing machine learning techniques for training networks on CSEPs that generalize across subjects ( ECoG grid placement , underlying functional organization , differences in spectral strucure , etc . ) is an important direction of future research with broad applications for BCIs [64] . Previous studies of motor cortex have claimed the existence of “beta-desynchronization” ( most commonly a decrease in beta amplitude ) during motor production [33 , 41] . This has led to a variety of hypothesized functions of beta ( β ) band in motor preparation and control , with little consensus across studies . A common methodology in many of these previous studies ( especially those done in humans , where the number of samples is small and function of cortex is often sub-sampled ) is to aggregate data across all electrodes and tasks . For the two subjects for which there was high-quality decoding accuracy , and thus , likely higher quality CSEP recordings , we found a novel positive coupling , i . e . , correlation , between the β band and the Hγ band amplitudes . The positive correlation was band-limited , occurring in the β range with a peak near 23Hz , and present at electrode-syllable combinations in which the electrode was active . Thus , uncovering this correlation required that we disaggregate the relation between β and Hγ according to whether an electrode , i . e . , articulator , was engaged in the production of a given speech sound . The presence of this coupling is correlated with the classification accuracy from the Hγ amplitude across subjects . The coupling in engaged functional areas is an example of the possible pitfalls of aggregation across functional areas and specific behaviors or stimuli when the combination of spatial specialization of function and task structure gives rise to sparse activation patterns . The structure and biophysical origin of broadband and band-limited signals in cortex is an area of active research [35 , 56 , 57 , 59 , 65 , 66] . In neural power spectra , it has been reported that there are broadband fluctuations in power that can be considered as a separate signal from band-limited signals , e . g . , β power [56 , 57 , 67] . Since this signal is broadband , it may mask or enhance cross-frequency correlations between underlying band-limited signals . Several methods have been proposed for estimating broadband signals and separating them from band-limited signals [56–58] . However , these methods are typically applied to ∼1 second windows with a step size that corresponds to ∼1-2 Hz . In this study , the cross-frequency analysis was performed at 200 Hz ( although the lowest frequency bands have autocorrelations due to the choice of bandwidths that correspond to ∼4 Hz ) . At 200 Hz , across the 4 subjects analyzed here , and across all electrodes in vSMC , there are approximately 250 million points at which a broadband signal would need to be extracted . To our knowledge , methods for estimating broadband signals at this scale that are computationally efficient and do not require per-fit hand tuning have not been developed . Developing methods for estimating high sampling rate , continuous broadband signals is an important direction of future research . Frequency bands besides Hγ are known to contain information about stimuli , behavioral , and state variables [33 , 34 , 36 , 38 , 40 , 41 , 59] . However , comparisons of task-relevant information across neural signals are rarely made . Information theory provides a way of measuring the amount of information about a task in a neural signal , the mutual information , but measuring mutual information across continuous , high dimensional signals is notoriously difficult . In the context of classifying discrete speech tokens , this information can be approximated through the information transfer rate . Being able to compare information across features is particularly useful for CSEPs which results from a variety of electrical processes in the brain [35] . Since they achieved higher accuracy then linear or single layer methods , deep networks optimized for accuracy can put a tighter bound on the task-relevant information in a set of neural features . We found that , for the amplitudes of frequency bands lower than Hγ , it is possible to decode speech syllables with above chance accuracy , though at relatively modest levels . Furthermore , when combined with Hγ features , the relative improvement in accuracy above Hγ accuracy is small compared to the cross-validation variance . Thus , for BCIs , these results imply that , for the CV task examined here , only Hγ activity ( or higher frequency signals ) need be acquired and analyzed: the other parts of the signal may profitably not be acquired to minimize data acquisition hardware and signal-processing in the decoder . Although deep networks have shown the ability to maximize task performance across scientific and engineering fields , they are still largely black boxes [68] . While there has been some initial investigations [69–73] , theoretical and empirical studies have not yet shown how deep networks disentangle the structure of a dataset during training . Currently , deep networks are most commonly used in science in cases where understanding of the deep network’s hidden representation is not needed . While we have taken some initial steps in that direction by examining the networks confusions , revealing how the deep networks disentangled articulatory features from the neural data will be an important extension of this work . An unresoved question is when we can expect deep networks to solve tasks through interpretable latent variables ( like phonetic features in the context of speech ) and how we can extract these variables from all layers of the learned deep network features ( here we only use the learned output probabilities ) . In general , understanding the interaction between dataset structure and deep network training will make deep networks more broadly useful as a tool for data analytics in science . Neuroscientists continue to create devices that measure more features in the brain while the stimuli or behavior during data collection become more complex and naturalistic . As the complexity of datasets increase , the tools needed to disentangle and understand these datasets must also evolve . Recently , deep networks have shown promise in analyzing and modeling neural responses in this work and others [17–19] . Moving beyond their utility as high-accuracy regression methods will require a more profound understanding of how deep networks learn to represent complex structure from data sets , and tools to extract that structure so as to provide insights to humans . Indeed , many of the open theoretical and analytical challenges facing deep networks are also core to understanding the brain . | It has been demonstrated that deep networks can achieve state of the art results on a number of classic machine learning tasks , but it is not currently clear whether deep networks can enjoy the same success in science , where not only accuracy but scientific understanding is desired . For example , the relationship between neural features and features of speech is often examined through the use of single-layer statistical models . However , single-layer models are unlikely to be able to describe the complex representations in higher-order brain areas during speech production . In this study , we show that deep networks achieve state of the art accuracy when classifying speech syllables from the amplitude of cortical surface electrical potentials . Furthermore , deep networks reveal an articulatory speech hierarchy consistent with previously studies which used hand-designed features . We also report a novel positive coupling between the beta and high-gamma bands during speech production in “active” cortical areas . However , using deep networks we show that , compared to lower frequency bands , the high gamma amplitude is by far the most informative signal for classifying speech . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"syllables",
"acoustics",
"linguistics",
"neural",
"networks",
"engineering",
"and",
"technology",
"signal",
"processing",
"social",
"sciences",
"neuroscience",
"consonants",
"computer",
"and",
"information",
"sciences",
"vowels",
"speech",
"grammar",
"phonetics",
"physic... | 2019 | Deep learning as a tool for neural data analysis: Speech classification and cross-frequency coupling in human sensorimotor cortex |
Increased susceptibility to influenza virus infection during pregnancy has been attributed to immunological changes occurring before and during gestation in order to “tolerate” the developing fetus . These systemic changes are most often characterized by a suppression of cell-mediated immunity and elevation of humoral immune responses referred to as the Th1-Th2 shift . However , the underlying mechanisms which increase pregnant mothers’ risk following influenza virus infection have not been fully elucidated . We used pregnant BALB/c mice during mid- to late gestation to determine the impact of a sub-lethal infection with A/Brisbane/59/07 H1N1 seasonal influenza virus on completion of gestation . Maternal and fetal health status was closely monitored and compared to infected non-pregnant mice . Severity of infection during pregnancy was correlated with premature rupture of amniotic membranes ( PROM ) , fetal survival and body weight at birth , lung viral load and degree of systemic and tissue inflammation mediated by innate and adaptive immune responses . Here we report that influenza virus infection resulted in dysregulation of inflammatory responses that led to pre-term labor , impairment of fetal growth , increased fetal mortality and maternal morbidity . We observed significant compartment-specific immune responses correlated with changes in hormonal synthesis and regulation . Dysregulation of progesterone , COX-2 , PGE2 and PGF2α expression in infected pregnant mice was accompanied by significant remodeling of placental architecture and upregulation of MMP-9 early after infection . Collectively these findings demonstrate the potential of a seasonal influenza virus to initiate a powerful pro-abortive mechanism with adverse outcomes in fetal health .
Influenza virus has been responsible for four pandemics in the past century , with an additional global health burden of seasonal influenza-related illness estimated at five million cases of severe illness and nearly 500 , 000 deaths annually [1] . Pregnant women are among the high-risk groups who are more susceptible to seasonal and pandemic influenza viral infections , with pronounced lung immunopathology [2] and increased incidence of complications , such as pre-eclampsia , pneumonia or heart failure during all 3 trimesters of gestation resulting in high hospitalization rates and mortality [1 , 3–8] . Notably , during the 2009 H1N1 pandemic , pregnant women in the United States showed a disproportionately high mortality rate , accounting for 5% of deaths while representing only 1% of the total population [9] . The majority of pregnant women who died of influenza-related illness during the pandemic were infected in the second and third trimesters of pregnancy [10] . Both seasonal and pandemic influenza virus have a substantial impact on the developing fetus . Infection during late-second or third trimester of pregnancy is associated with significant increases in miscarriages , stillbirths , and early neonatal diseases and death [1 , 11] . Several studies indicate that infants born to influenza virus-infected mothers have an increased risk of developing health problems later in life ranging from chronic immune diseases to schizophrenia [12 , 13] . Though the mechanisms predisposing pregnant women and their fetuses to an increased rate and severity of influenza virus infection-related complications are not fully elucidated , it is generally accepted that the feto-placental tolerance developed during pregnancy to support life of a semi-allogeneic fetus is a contributing factor to adverse outcomes [6 , 7 , 14 , 15] . Dynamic immunological processes that occur throughout pregnancy are regulated by pregnancy hormones , whose receptors are found on most immune cells [7 , 16–20] . Researchers have demonstrated a shift away from type Th1 cell-mediated responses toward enhanced type Th2 humoral-mediated responses during pregnancy , suggesting a protective role for the developing fetus , although these alterations could predispose to increased susceptibility to infections from respiratory viruses such as influenza [18 , 21] . Inflammation plays a major role in tissue pathogenesis and it is more pronounced during the last 2 trimesters of pregnancy due to changes in cytokine levels that regulate fetal-placental tolerance , that suppress some cytokines ( e . g . IFN-γ and VEGF ) , while elevating the levels of others ( e . g . TNF-α and G-CSF ) [14 , 22 , 23] . Presently there are very few studies that have examined the adverse effects of influenza virus infection in pregnancy mainly using the pregnant mouse model due to histological similarities of rodent to human placenta [24] . These reports demonstrated adverse effects of the 2009 pandemic influenza A ( H1N1 ) virus on lung immunopathology [25–27] , and more recently Kim et al reported severe pathogenesis of influenza B virus in the pregnant mouse model . However , the impact of seasonal H1N1 influenza A virus infection on maternal and fetal health during pregnancy , and the exact mechanisms leading to premature rupture of membranes ( PROM ) and preterm birth after influenza virus infection during pregnancy have not been fully elucidated [6 , 11] . The objective of this study was to recapitulate the adverse outcomes of seasonal influenza virus infection in maternal , placental and fetal compartment using the mouse model . We aimed to correlate disease burden with progesterone and prostaglandin levels since their role is critical in lung and placental health; and assess the magnitude of endocrine and immunological changes that through tissue inflammation and pathology lead to maternal and fetal distress . To achieve this objective , we chose as a model influenza virus the seasonal H1N1 A/Brisbane/59/07 strain to sub-lethally infect BALB/c mice at the second third of gestation period and monitor maternal-fetal health . All correlations between progesterone and prostaglandins with compartment-specific inflammation and immune responses as well as tissue remodeling took place at 4 days post-infection , when weight loss was first exhibited in our mouse model and at which time point pregnant women are most susceptible to ICU admission and death in previous pandemics [10] .
In order to assess the impact of seasonal influenza virus infection in maternal and fetal health , we first infected pregnant mice at day 12 of gestation and non-pregnant controls with a low dose of H1N1 A/Brisbane/59/07 . With this approach we aimed to recapitulate human infection at the end of 2nd to 3d trimester of pregnancy with a low pathogenicity seasonal influenza virus . Uninfected pregnant controls reached a peak bodyweight that was 56 . 2% greater than that of their pre-pregnancy bodyweight ( Fig 1A ) . The infected pregnant animals reached a peak bodyweight of 34% of their original bodyweight and the average gestation period decreased by 6 . 7% to 19 . 6 days ( p<0 . 001 ) ( Fig 1B ) . Additionally , pregnant mice infected during gestation dropped to 89% of their initial body weight following pre-term delivery , while uninfected pregnant mice remained at 116% of their initial body weight following delivery . We found that infection with a low pathogenicity influenza virus interrupted the normal progression of pregnancy to completion inducing pre-term labor in our mouse model . The initial experiment recapitulated the clinical phenotype of pre-term birth observed in human women infected mid-gestation with seasonal influenza virus [28] . Next , we examined the impact of infection on offspring viability and health status of pregnant mice . Numbers of viable and non-viable pups were recorded , bodyweights were taken at birth and classified as non-viable ( ≤1 g ) , SGA ( 1 . 1–1 . 25 g ) and healthy ( >1 . 25 g ) and monitored daily for growth curves . Litter size averaged 4 . 5 pups per uninfected healthy pregnant controls while infected mothers gave birth to an average of 2 . 7 pups per litter; however , because infected mothers often consume sick or stillborn offspring , these results are not necessarily reflective of number of offspring carried to parturition . The average weight of offspring from uninfected pregnant mice was 1 . 46g . When mothers were infected with the virus , the body weight of newborns was 19% lower , averaging 1 . 18g ( p<0 . 0001 ) ( Fig 1C ) . Pups born from infected mothers at days 18 and 19 of gestation had lower body weight compared to the pups born at day 20; however , 10 mice born on day 21 were all stillborn ( Fig 1D ) . Thus , the length of pregnancy did not necessarily correlate with the health outcome of the offspring; that is , SGA condition was not dependent on the length of gestation . In uninfected pregnant mothers , 85% of pups were born with a healthy weight , while 13% were SGA pups and 2% were stillborn ( Fig 1E ) . Infection with seasonal influenza virus reduced the number of healthy offspring to 25%; the remaining pups were 25% SGA , and 50% stillborn ( Fig 1E ) . Overall , offspring born from infected mothers were approximately 20% smaller in terms of body weight compared to offspring of uninfected mothers . Infection nearly doubled the incidence of SGA pups from 13% to 25% , and the likelihood for stillbirth increased by 35-fold from 2% to 70% following infection ( Fig 1C , 1D and 1E ) . Low birthweight or SGA offspring are a common outcome of pregnant mothers infected with pandemic H1N1 influenza virus during the second and third trimester and has been observed in cohorts of pregnant women infected with seasonal H1N1 influenza virus in Nova Scotia [6] . Thus our model replicates more serious clinical birth complications associated with influenza A viral infection during mid- to late-gestation pregnancy . The severity of viral infection in pregnancy was initially assessed by examination of various organs for systemic spread of pathogen , viral load and tissue inflammation at 4 d . p . i . The results from lungs of infected pregnant mice were compared to those of non-pregnant infected controls . Although both pregnant and non-pregnant infected lungs were heavily infiltrated by neutrophils showing intense inflammation ( Fig 2A ) , viral loads were 8-fold higher ( p<0 . 05 ) in pregnant animals compared to non-pregnant ones , suggesting hampered virus clearance during pregnancy ( Fig 2B ) . Interestingly , lungs of mock-infected pregnant mice showed increased airway inflammation over mock-infected non-pregnant mice , indicating that there is increased physiological stress in the lungs during pregnancy ( Fig 2a: a , d ) . Viral titers were undetectable in the placentas and fetuses of infected mice . This was further confirmed with qPCR measuring viral RNA ( M gene ) levels ( Fig 2C ) . Lack of detectable viral RNA in placental and fetal tissues indicates that transplacental transmission of influenza virus from mother to child is unlikely and that retardation in fetal development was not due to in utero infection . These findings correspond with clinical reports on pregnant mothers infected with influenza A virus during pregnancy; while several H1N1 strains can infect fetal trophoblast cells in vitro , clinical evidence of vertical transmission is uncommon and inconclusive [29–31] . Despite the presence of influenza virus receptors α-2 , 6 sialic acids on human placental membranes , only H5N1 viruses have been reported to transmit via maternal vertical transmission [32 , 33] . This indicates that the negative effects on offspring born to mothers infected with influenza virus during pregnancy are mediated not by direct infection of placenta and fetus , but indirect causes such as dysregulated hormonal signaling , increased inflammation , or immune system activation against placental and/or fetal tissue . Progesterone promotes endometrium and uterine changes to support embryo implantation and feto-placental development . Insufficient progesterone concentrations can lead to preterm delivery and miscarriage in humans and rodents [34 , 35] . To assess the effect of influenza virus infection on systemic progesterone levels and subsequently in pre-term labor , we first measured cytokine expression and progesterone levels in serum of uninfected pregnant and non-pregnant mice ( Fig 3 ) . Serum progesterone was 7-fold higher ( p<0 . 01 ) in uninfected pregnant mice compared to non-pregnant controls as is expected during healthy gestation ( Fig 3A ) . These levels inversely correlated with the levels of pro-inflammatory cytokines IL-1β , IL-12p40 and GM-CSF showing a reduction of 41% , 50% and 63% respectively when compared to their levels in non-pregnant mice ( p= . 01 ) ( Fig 3B , S1 Table ) . None of the Th2 ( IL-3 , IL-4 , IL-5 , IL-10 , IL-13 ) or Th1 ( IL-2 , IL-12p70 , IFN-γ ) cytokines showed any differences between pregnant and non-pregnant mice ( S1 Table ) . To further establish the threshold of inflammation in the pregnant mouse model , we measured the levels of serum prostaglandin F2α ( PGF2α ) and prostaglandin E2 ( PGE2 ) in uninfected pregnant mice and compared them to uninfected non-pregnant controls . PGE2 selectively suppresses effector functions of macrophages and neutrophils and the Th1 , CTL and NK cell-mediated type-1 immunity , but promotes Th2 , Th17 , and Treg responses [36] . In contrast to increased levels of progesterone during pregnancy , PGF2α and PGE2 were reduced by 50% and 33% respectively in serum of pregnant mice compared to non-pregnant controls ( Fig 3C and 3D ) . Thus we established in our mouse model that pregnancy hormones or hormone-like mediators of homeostasis are subjected to systemic changes to prepare an immune environment trending toward anti-inflammatory responses . Hormonal and cytokine responses were further confounded during pregnancy following influenza virus infection . Infection caused a 5-fold reduction of serum progesterone in pregnant mice 4 d . p . i . to levels similar to those seen in non-pregnant mice ( Fig 3A ) . There was a significant inverse correlation between serum progesterone levels and lung viral load ( r2 = 0 . 8 , p = 0 . 04 ) ( S1C Fig ) . Similarly , serum PGF2α levels were approximately 5 times lower than pre-infection in both pregnant and non-pregnant cohorts ( p = 0 . 02 ) ( Fig 3C ) whereas PGE2 did not show any significant differences ( Fig 3D ) . Infection in pregnancy elevated the serum inflammatory cytokines IL-1β , IL-12p40 , GM-CSF and eotaxin from 2 to 10 times whereas it significantly reduced RANTES ( Fig 3E; S1 Table ) . When compared to infected non-pregnant controls , pregnant mice had less than half the number of upregulated inflammatory cytokines ( IL-1β , IL-6 , IL-12p40 , eotaxin , G-CSF , GM-CSF , MIP-1α , TNF-α , RANTES , ) ( S1 Table ) . The majority of these cytokines had approximately 50% lower levels in pregnant mice ( Fig 3F ) . Interestingly , IL-10 increased significantly only in infected pregnant mice ( p = 0 . 02 ) suggesting a robust regulatory response to infection ( Fig 3E ) . Pro-inflammatory IL-17 which is detected mainly in serum and placenta [37] was elevated 2-fold during pregnancy ( Fig 3B and 3F ) . Although IL-17 production was not affected by virus infection , its levels were significantly higher than those observed in infected non-pregnant animals ( Fig 3F , S1 Table ) . These data show that pregnancy results in a unique signature of systemic antiviral cytokine expression . Progesterone and PGF2α reduction led to an inflammatory response although its magnitude was lower than the response seen in the control group . Progesterone increase in pregnancy results in relaxation of smooth muscles in the airways , dilating the bronchial tissue [38] . PGE2 and PGF2α are key players in lung’s physiological response to infection and airway reactivity [38] . PGE2 is a vasodilator that increases the permeability of the lung vasculature [39 , 40] and works with PGF2α , a bronchoconstrictive hormone to finely tune lung function [41] . Influenza virus infection damages lung epithelial cells , releasing free reactive oxygen species ( ROS ) that activate cyclooxeganse-2 ( COX-2 ) via the arachidonic acid pathway and upregulate synthesis of PGE2 [42] . In order to assess the effect of these hormones on inflammatory responses in infected lungs of pregnant mice , we determined their levels along with a panel of 23 cytokines and chemokines in tissue lysates of pregnant and non-pregnant mice 4 d . p . i . Pregnancy increased expression of lung progesterone 11 . 6-fold ( p<0 . 001 ) and PGF2α 2 . 7-fold ( p = 0 . 03 ) ( Fig 4A and 4B ) , while no detectable changes were found in PGE2 levels between pregnant and non-pregnant mice ( Fig 4C ) . Notably , in the absence of infection , pregnant uninfected mice overexpressed lung neutrophil-recruiting chemokines ( KC , 1 . 37 times higher ) and cytokines ( IL-1β , IL-6 and G-CSF; 1 . 64- , 7 . 6- and 3-times higher respectively ) relative to non-pregnant controls ( Fig 4D , S2 Table ) [43] . These data suggest that these changes may prepare the lung environment during pregnancy to fight microbial insults , more so than during a non-pregnant state . Infection reduced progesterone expression in the lungs of pregnant animals by 5-fold ( p<0 . 0001 ) when compared to non-pregnant mice that showed a 2-fold reduction ( p = 0 . 01 ) ( Fig 4A ) . Moreover we observed a strong correlation between lung progesterone with viral load ( r2 = 0 . 79 , p = 0<0 . 0001 ) ( S1D Fig ) . Following the same trend , PGE2 levels when significantly decreased ( p = 0 . 008 ) ( Fig 4C ) . However , PGF2α lung expression was not altered after infection in either pregnant or non-pregnant mice ( Fig 4B ) . While influenza virus infection caused dramatic increases in expression of inflammatory cytokines and chemokines in lungs of both pregnant and non-pregnant mice ( S2 Table ) , we detected a few differences between these groups . The levels of IL-1α , IL-1β and KC were 50% , 200% and 40% higher than those estimated in infected non-pregnant controls , while IL-6 levels were 50% lower in the pregnant group ( Fig 4E , S2 Table ) . The results on IL-1α and IL-6 are in agreement with previous reports on the role of IL-1α as a negative regulator of IL-6 expression [44] . Influenza virus infection upregulated MIP-1α expression in both pregnant and non-pregnant cohorts reaching similar levels ( S2 Table ) whereas MIP-1β expression was 40% lower in pregnant animals ( Fig 4E , S2 Table ) . While both cytokines are involved in cellular recruitment and inflammation at the site of infection , they signal through different receptors . MIP-1α signals through CCR1 , CCR4 , and CCR5 and MIP-1β signals directly through the CCR5 [45] , suggesting that pregnancy may induce different pathways signaling in response to infection . Our findings point to a model for the excessive morbidity seen in infected pregnant mice . Pregnancy induces a combination of physiological changes , including vasoconstriction in the lungs via PGF2α , the negative impact of infection on progesterone and PGE2 expression , and robust inflammatory responses in the lungs . Decreased bronchodilation and intense neutrophil infiltration in the alveolar spaces leads to respiratory failure , which is evident also in lung histopathology ( Fig 2A ) [46–48] . In addition to their effects on lung architecture and immune responses , progesterone , PGE2 and PGF2α were measured following infection of pregnant mice because of their respective roles in uterine contractility . Progesterone elevation in pregnancy supports fetal development by thickening the endometrium to allow for embryo implantation and suppressing inflammation in the uterus . PGE2 increases vasodilation , induces uterine contractions , and decreases T-cell proliferation and lymphocyte migration [49] while PGF2α initiates vasoconstriction of uterine and endometrial blood vessels [50] , resulting in induction of labor [6 , 12 , 13 , 51–53] . At the same time , PGF2α stimulates the production of pro-inflammatory cytokines and may enhance uterine production of leukotriene B4 ( LTB4 ) that in turn activates various neutrophil functions [54] . We found that progesterone levels in the placentas of infected pregnant mice were decreased by 40% compared to uninfected pregnant mice ( p= . 04 ) ( Fig 5A ) . In contrast , PGF2α levels were almost 5-fold higher in the same group compared to uninfected controls ( p= . 0002 ) ( Fig 5B ) . Placental lysates did not show a difference in PGE2 expression ( Fig 5C ) after infection . Progesterone and PGF2α expression were inversely correlated with viral load following infection although significant association between measured parameters was only observed in progesterone ( r2 = 0 . 46 , p = 0 . 007 and r2 = 0 . 44 , p = 0 . 22 respectively ) ( S1B and S1E Fig ) . The association of serum , lung or placental progesterone with viral load supports the notion that seasonal influenza virus infection during gestation disrupts the production of progesterone , and provides evidence an explanation for early termination of pregnancy and spontaneous abortions . Fold changes of cytokine expression in the placentas and fetuses of pregnant mice were also compared before and after infection . Eleven placental cytokines showed suppressed expression ranging from 29–72% after infection ( S2 Fig , S3 Table ) . G-CSF and RANTES however were enhanced by 4 . 8- and 1 . 5-fold respectively . Similarly , fetuses showed a dramatic decrease in immune responses , although pre-infection cytokine and chemokine levels were very low compared to the placental compartment . With the exception of IL-1α , which was increased following infection , cytokine expression was decreased in the fetus post-infection by at least 60% ( S2 Fig , S3 Table ) . This suppression indicates that the placenta and fetus are effectively shielded from the inflammatory cytokine signature of the mother’s circulatory system and that the physiological changes induced by progesterone and prostaglandins are likely causes of poor fetal and offspring outcome . Placenta is the sole source of nutrients and oxygen to the developing fetus . The placenta in both humans and rodents is derived from the maternal endometrial decidua and fetal trophoblasts [55] . The development of this critical organ is carefully negotiated through regulation by pregnancy hormones , initiated by prolactin and progesterone , and tightly regulated immune tolerance by uterine regulatory T cells [56] . Inflammation in the uterus and placenta due to infection or autoimmune responses has been linked to preeclampsia , endometriosis , and spontaneous abortion [57 , 58] yet influenza virus has not been definitively shown to cross the maternal-fetal barrier in mice or humans . Viral infection in the placenta was examined and yet no viral RNA ( M gene ) was detected via qPCR ( Fig 2C ) . Hence , the effect of influenza infection on placental function and health was examined via histology and molecular assays . Placentas from uninfected mothers at E16 ( 16 days of gestation ) ( Fig 6A: a , b , c , g ) and mothers 4 d . p . i . ( Fig 6A: d , e , f , h ) were compared . Placentas from uninfected mothers maintained structural integrity between the maternal decidual layer , the fetal spongiotrophoblast layer , and the placental labyrinth ( Fig 6A: a ) while placentas from infected mothers ( Fig 6A: d ) showed increased regions of fetal endothelial cellular death ( stars ) with dark-blue nuclei suspended within the tissue , indicating that infection increases cellular death within the placenta . Infection increased gaps within the spongiotrophoblast layer ( ▲ ) , a region through which maternal spiral arteries cross to deliver oxygen to fetal blood within the labyrinth ( Fig 6A: b , e , f ) . Infection also increased the incidence of fibrinoid necrosis in-between the maternal decidual layer and the fetal spongiotrophoblast ( ■ ) region . ( Fig 6A: h ) [59] . Histology slides were scored for the incidence of fetal endothelial cellular death ( FED ) , degradation of the spongiotrophoblast layer , and fibrinoid necrosis ( FN ) ; infection increased the occurrence of these indicators of poor placental health ( Fig 6B ) . To determine underlying molecular causes of the placental spongiotrophoblast degradation , matrix metalloproteinase ( MMP ) expression was quantified via Western Blot in placental lysates from infected and uninfected pregnant mice at E16 ( Fig 6C ) . MMPs are proteolytic enzymes , localized in the placenta , that remodel tissues throughout the body via endopeptidase activity towards extracellular membranes ( ECM ) , breaking down tightly adherent cells within the tissue [60–62] . MMP-2 and MMP-9 are gelatinases that have been widely implicated in placental pathology and preterm labor , both in rodent models and in humans [57 , 63–66] . MMP-9 can be induced by high levels of IL-1β via the p38 MAPK signaling pathway [61 , 67 , 68] , consistent with the increased levels of IL-1β observed in placentas of infected mice ( S3 Table ) . Infection increased placental expression of the pro-enzyme and activated enzyme form of MMP-2 and MMP-9 ( Fig 6C ) , thus exposing both the placenta and the amniotic membranes surrounding the fetus and placenta to degradation . Sex hormone synthesis and feedback pathways are intrinsically linked with innate immune signaling pathways of TNF-α , IL-1β , IL-4 , IL-6 , IL-10 , IL-17 , and IFN-γ [42 , 69 , 70] . Hence , we quantified hormone regulators , progesterone-induced blocking factor ( PIBF ) and COX-2 in order to examine the interdependent relationships in endocrine and innate signaling in our pregnancy model . Successful outcome of gestation requires expression of PIBF , a progesterone-responsive immunomodulatory protein with multiple active isoforms: 90kD , 66kD , and 55kD [71 , 72] . PIBF promotes Th2 cytokine production [73] and inhibition of natural killer ( NK ) cell activity [34 , 51] . While this function is crucial to limiting cytolytic activity in the uterus and placenta reduction of NK activity in lungs leaves the pregnant mother vulnerable to infection . Expression of PIBF isoform 66kD in the placenta was nearly abrogated following infection while 90 kD and 55 kD isoforms were undetectable ( Fig 7A and 7B ) . In non-infected non-pregnant and pregnant mouse lungs , these isoforms were differentially expressed . The isoforms 90 kD and 55 kD showed 3-fold and 9-fold higher expression respectively in pregnant mice as compared to non-pregnant controls . In contrast , the levels of 66 kD isoform were 2-fold lower when comparing the pregnant to the non-pregnant group . Following infection of pregnant mice , isoform 66 kD was reduced 3-fold and isoform 55 kD was increased 2-fold . In infected pregnant mice , all isoforms were uniformly suppressed by 2-fold ( Fig 7C , 7D and 7E ) . The results of this study demonstrate that different isoforms are highly expressed in placenta and lungs of pregnant mice , that infection down-regulates all isoforms and that there is a differential expression of isoforms in lungs during pregnancy . Notably the regulation of isoform 55 kD in pregnant and non-pregnant population follows opposite trends , suggesting a role in immunomodulation of placental-fetal interface . COX-2 is a key regulator in the arachidonic acid and prostaglandin synthesis pathway . Influenza virus infection induces COX-2 expression , thus increasing secretion of PGF2α in bronchial epithelial cells [42] . Previous studies demonstrated that COX-2 deficiencies reduce the recruitment of macrophages and neutrophils to the site of influenza viral infection , and COX-2-/- mice infected with influenza virus had reduced lung viral clearance compared to wild-type infected mice [69] [74] . COX-2 placental expression was decreased upon infection ( Fig 8A and 8B ) which is consistent with lack of upregulation of PGE2 in the placenta following infection ( Fig 5C ) . Thus , preterm birth in our model is correlated with the physiological changes mediated by the withdrawal of pregnancy-supportive progesterone , the increased expression of labor-inducing PGF2α , and upregulation of structure-damaging degradative proteinases in the placenta rather than increased inflammation mediated by COX-2 , PGE2 , or cytokines . Pregnancy elevated expression of active COX-2 45kD in the lungs ( Fig 8C and 8E ) . Infection did not exert a significant effect on active 45 kD COX-2 expression in the lungs of non-pregnant mice but resulted in a 2-fold increase during pregnancy ( Fig 8E ) . Enhanced COX-2 expression during pregnancy may explain increased expression of neutrophil-recruiting chemokines in the lungs ( Fig 4E ) , providing a regulatory mechanism for the enhanced immunopathology seen following infection .
Influenza viral infection disrupts the supportive molecular environment for fetal development and offspring health . The objective of this study was to investigate how seasonal H1N1 influenza virus infection affects the outcomes of pregnancy and how pregnancy alters the innate immune response during infection in a BALB/c mouse model . This study is unique in that it ties together hormonal changes induced by infection with cytokine dysregulation across three compartments; maternal , placental and fetal . We demonstrate that influenza virus infection alters key hormone levels required to maintain healthy pregnancy , and results in increased immunopathology thus compromising maternal and fetal health . Our study expands upon previously published work from Kim HM et al[27] , Kim JC et al [75] , Klein and Robinson [17 , 19 , 76 , 77] , and Moran et al [18 , 78] by examining compartmental changes in hormone and cytokine expression following influenza virus infection . We linked seasonal influenza virus infection to clinical observations of adverse outcomes in pregnancy ( preterm birth , SGA , stillbirth , increased maternal morbidity ) , enhanced lung and placental histopathology and reduced control of viral replication in lungs of infected pregnant mothers . We provide a model for how influenza virus infection , while contained in the lung , results in global dysregulation of the hormonal signaling required to sustain healthy gestation . In our study , the state of pregnancy dampens innate immune responses compared to non-pregnant controls by at least 50%; in some compartments , the maternal body maintains some form of systemic tolerance , while in the respiratory compartment , the lungs remain primed to fight invading pathogens . Consistent with previous clinical reports and animal studies , we found that pregnancy increases the severity of seasonal influenza virus infection and thus impairs maternal and offspring health and recovery [25–27 , 75 , 79–81] . Pregnant mice exhibited increased viral replication in the lungs by 8-fold compared to infected , non-pregnant controls , a trend that was also found in similar studies looking at pandemic H1N1 influenza virus in 2012 and influenza B virus in 2014 [27 , 75] . We found a significant correlation between lung viral load and progesterone levels in lung , placenta and serum suggesting that seasonal influenza virus infection during gestation disrupts the production of progesterone leading to preterm labor , SGA and increased fetal mortality . We also observed a dramatic increase of placental PGF2α after infection pointing to the role of this prostaglandin in the intense local inflammatory response . We also documented that the immune responses are compartment-specific in the mouse model , resulting in distinct signatures of inflammation within each compartment ( S2 Fig ) . Similar to findings by Kim et al with pandemic 2009 H1N1 influenza virus , we demonstrated that seasonal H1N1 influenza virus enhanced lung pathology in pregnant mice via an increase in IL-6 , IL-1α , and G-CSF expression and that infection reduced serum concentrations of progesterone during pregnancy [27] . The differences in systemic responses between pregnant and non-pregnant groups in our mouse model suggest that the unique endocrine environment supporting gestation compromises most of the innate immune responses to infection . These shifts in response to infection during pregnancy may have evolved to protect the fetus from activation of inappropriate activation of T cell responses , at the expense of increased inflammation in the lungs . There are a few caveats to our pregnant mouse model of infection . Most obvious is that the average length of pregnancy in humans is 40 weeks while gestational length in mice is around 19–21 days depending on the mouse strain . However , both mice and humans possess a hemochorial placenta where the fetal chorion is in direct contact with maternal blood [82 , 83] . Common transcription factors involving the expression of various placental genes have been identified in both human and mouse placenta [84] . Another caveat is the different mechanisms for maintaining progesterone secretions during gestation in mice and humans . In mice , progesterone is produced by the placental corpus luteum , which is regulated by prolactin for the first half of pregnancy and then by trophoblastic giant cells for the remainder of gestation [84] . In humans , prolactin is not required for the maintenance of pregnancy and the corpus luteum depends on the human chorionic gonadotropin ( hCG ) produced by trophoblasts to stimulate progesterone . After 8 weeks of gestation , the progesterone produced by the placenta is sufficient to maintain pregnancy in humans [84] . These differences should be taken into account when interpreting the role of progesterone in pregnancy following influenza virus infection . Lastly , Periolo et al . studied nasopharyngeal swab samples obtained in the second and third trimester of 41 pregnant women with confirmed pandemic H1N1 influenza A virus infection [2] . The authors reported increased expression of inflammatory cytokines IL-8 , IL-6 , and TNF-α and significantly lower levels of TGF-β and IFN-β compared to the pregnant women who survived or non-pregnant controls , with particular emphasis in the role of IL-6 in severity of disease . We have not seen these differences in seasonal H1N1 influenza virus-infected pregnant mouse model , and to our knowledge , there have been no studies examining the extent of cytokine expression in response to seasonal influenza virus infection in human pregnant women . Influenza virus infection in our mouse model interfered with progesterone-mediated anti-inflammatory effects during pregnancy . Elevated levels of progesterone in the lungs during pregnancy suppress the activation of COX-2 by blocking IL-1β activation of NF-κB thus creating an anti-inflammatory environment at the expense of a quick response to foreign antigen [85] . However , when slow immune response results in uncontrolled viral growth , NF-κB initiates transcription of a host of immunomodulatory genes , including COX-2 [86] . It has been reported that influenza virus infection results in epithelial cell damage , which releases free oxygen radicals and activates COX-2 [42] . In our pregnant model we demonstrate that influenza virus infection induced COX-2 activation and upregulation of PGF2α in the lungs resulting in vasoconstriction and inflammation . These changes reduce the amount of oxygen available to the mother and developing fetus that may lead to respiratory distress , increased morbidity and retardation of fetal growth . Previous studies have indicated that progesterone treatment at levels equivalent to hormonal birth control are sufficient to limit immunopathology of H1N1 influenza virus infection in non-pregnant mice [87 , 88] . In our study , influenza A virus infection resulted in a decrease of progesterone expression in pregnant mice , and progesterone therapy may be a promising treatment to mitigate the effects of viral infection on lung pathology and to prevent PROM and early delivery of the fetus . COX-2 inhibitors have been proposed as antivirals for treating inflammation caused by influenza virus infection [89 , 90] . This study demonstrates that pregnancy has a unique relationship with COX-2 upregulation in the lungs and that pregnant women may benefit from these molecular inhibitors following influenza virus infection . Vertical transplacental transmission of influenza virus has been debated [29 , 30 , 91] . In our study , influenza virus was not detectable in the placenta and fetus . Instead , poor offspring health was likely due to imbalances in the maternal endocrine and immune physiology responsible for proper fetal development . Influenza virus infection resulted in a breakdown of the placental architecture , likely caused by inflammation , reduced progesterone expression and activation of structure-remodeling MMP-2 and MMP-9 . Programmed cell death in amnion membranes is routine to normal parturition in rodents; membranes rupture in part due to tissue weakening rather than solely dependent on mechanical stress of labor [92] . Weakened strength of the amniotic membrane combined with reduced placental health creates an impetus for premature rupture of membranes ( PROM ) and thus , preterm birth . Increases in the placental concentrations of vasoconstrictor PGF2α and immune cellular activators G-CSF and RANTES following infection create an environment where uterine contractility is triggered prematurely and immune cells can be activated against fetal cells at the maternal-fetal interface ( Figs 5B , S2 ) . While we have shown disruption of key hormonal regulators in the lungs following infection during pregnancy , COX-2 does not seem to be involved in the phenotype of pre-term labor following viral infection during pregnancy . COX-2 is directly correlated with PGE2 levels , which do not change in the placenta following influenza virus infection . Rather , major changes in progesterone and PGF2α production have been shown to induce pre-term labor , while synthesis of PGF2α does not depend on COX-2 cellular expression [74] . Compromise of placental structure and function may lead to reduced oxygen and nutritional supply and buildup of gas and waste in the fetus , resulting in retardation of fetal growth and neural development or stillbirth . Clinical studies indicate that maternal influenza A virus infection during pregnancy may predispose offspring to psychosis and schizophrenia due to fetal neurodevelopment being dysregulated by hypoxia and inflammation [93 , 94] . In future studies , offspring of influenza virus-infected mothers will be followed to adulthood and examined for neurological disorders as well as variation in the timing of viral dose at early , mid , and late gestation . These data support a model where influenza virus infection “breaks through” the balance of maternal systemic tolerance towards developing fetuses , enhancing pathogenesis in the lungs and triggering pre-term birth thus affecting the health of both mother and offspring . We hypothesize that fetal health is impaired by the spillover of inflammatory cytokines , whose expression is influenced by pregnancy hormones , and structural remodeling proteins intended to repair maternal airways have long ranging effects outside the compartments they were intended to work in , such as the placenta . Previous studies have examined the effect of individual hormones on male and female innate immune responses to influenza virus infection . Klein et al have demonstrated that 17β-estradiol and progesterone treatments reduce influenza virus infection immunopathology , promoting expression of TGFβ , IL-22 , and IL-6 and suppressing inflammatory cytokine production [77 , 87 , 95] . We found that reduced expression of progesterone in the lungs and placenta might be a causative factor for fetal and maternal respiratory distress and early termination of pregnancy . Infection in the maternal lungs resulted in the upregulation of cellular activation markers in the placenta and inflammatory signaling in the fetus , indicating that while in utero offspring are not in contact with influenza virus; they are nevertheless negatively impacted by maternal infection . Our findings also suggest a breach of feto-placental tolerance , which relies on suppression of uterine and placental immune cells to ensure an inflammation-free environment for the developing fetus . This study further explains the interconnected nature of hormonal and cytokine signaling during pregnancy and the unique predicament of maternal tolerance during respiratory viral infection . Recent work has shown that immunosuppression in the lungs during pregnancy creates an ideal environment for adaptation of influenza viruses to more virulent strains [81] . Thus , understanding how pregnancy hormones modulate immune responses to influenza virus infection is not just necessary for developing clinical interventions for a high-risk population but also as part of a global strategy to reduce the incidence of highly pathogenic influenza viruses with pandemic potential . Pregnant women are a target population for improving vaccination efficacy in order to reduce their risk of pregnancy complications due to infection and to increase protective strength of passive immunity to their offspring [96 , 97] . We demonstrate that increased progesterone expression required to support pregnancy results in an immunosuppressive lung environment and that pregnancy increases susceptibility to prostaglandin-induced inflammation as a result of infection . These conclusions are important for understanding respiratory viral pathogenesis in a pregnant mother .
Madin-Darby canine kidney ( MDCK ) cells ( ATCC CCL 34 , American Type Culture Collection , Manassas , VA ) were maintained in Dulbecco's Modified Eagle's Medium ( DMEM ) ( Mediatech , Herndon , VA ) containing 10% fetal bovine serum ( Hyclone , Thermo Scientific , Rockford , IL ) . A/Brisbane/59/07 ( H1N1 ) virus stock was propagated in MDCK cells . The hemagglutination ( HA ) activity was determined using turkey blood cells ( LAMPIRE Biological Laboratories , Pipersville , PA ) [98] . The mouse-adapted A/Brisbane/59/07 ( H1N1 ) strain was obtained by serially passaging the virus in lungs of BALB/c mice . The LD50 was determined by Reed-Munch formula [99] and the viral titers were determined by plaque assay [100]; 2xLD50 mouse-adapted virus is approximately 155 plaque forming units ( p . f . u ) per infection . 8-week-old female BALB/c mice ( Harlan Laboratories , Dublin , VA ) were bred and housed in a biosafety level 1 facility for breeding; infections were conducted in a biosafety level 2 at Emory University Whitehead animal facility . All animal studies were approved by the IACUC at Emory University . The mouse estrous cycle is divided into four stages: estrous , metestrous , diestrous , and proestrous . Each stage can be determined by visual examination of vaginal opening [101] . Tagged female mice were used for breeding . Cages were set up with three to four female mice in proestrous or estrus and one a male for 3 days [102] . Females were monitored for the presence of a copulation plug , indicative of mating and body weight changes were monitored daily . In uninfected BALB/c pregnant mice , gestation lasts approximately 21 days , and pregnancy can be determined when mice gain 20% of their initial body weight . Pregnant mice were placed in separate cages , two females per cage , and the remaining mice were mated again . Pregnant females that had noticeable body weight increases ( in the range of 20–25% ) of their initial bodyweights between days 12–14 after mating were infected with 2xLD50 mouse-adapted A/Brisbane/59/07 ( dose calculated for non-pregnant , healthy 8-week-old female mice ) . Intranasal infections were performed under light isoflurane anesthetization , and euthanasia was performed via CO2 asphyxiation . Severity of infection , completion of gestation , and health outcome of offspring ( i . e . healthy , small for gestational age , or stillborn ) were all closely monitored and compared to uninfected pregnant mice . In another cohort of mice , sera , lungs , placenta , and fetuses were collected 4 d . p . i . in order to assess viral presence and extend of inflammation in various compartments , host innate immune responses , and hormone expression . Serum was stored at -20°C in the presence of Halt Protease Inhibitor ( Thermo Fisher ) until further use . Lung and placental lysates were harvested from infected and uninfected pregnant mice and homogenized , filtered through 70μm filters and stored in 1x DMEM , 1x Halt Protease Inhibitor and/or RNAse Later ( Ambion ) . Viral RNA was isolated using a QIAamp Viral RNA Mini Kit ( Qiagen ) and probed for the presence of H1N1 HA and M genes via qPCR . Viral RNA was converted to cDNA iTaq Universal SYBR Green Supermix ( Bio-Rad ) and amplified using 3 nmol of commercially available or synthesized primers in a CFX96 Real-Time PCR Detection System ( Bio-Rad ) . PCR mixtures were heated at 50°C for 10 minutes to convert RNA to cDNA , 95°C for 1 minute , and cycled 40x ( 95°C for 10 seconds , annealing temperatures were varied dependent on primer melting curve for 15 seconds , and 72°C for 30 seconds ) . A melting curve from 65°C to 95°C was performed to assess quality of primer binding . H1 HA was amplified using Influenza A Virus H1 Primers ( BEI Resources , NR-12316 , with an annealing temperature of 61°C . H1 M was amplified using synthesized primers ( FLUAM-1F: AAGACCAATCCTGTCACCTCTGA; FLUAM-1R: CAAAGCGTCTACGCTGCAGTCC; Operon ) with an annealing temperature of 61°C , and mouse GAPDH ( realtimeprimers . com , VMPS-7317 ) with an annealing temperature of 50°C [103] . Analysis of threshold values and normalization to GAPDH were performed in CFX Manager Software ( Bio-Rad ) . PCR products were electrophoresed on a 1% agarose/TAE gel and imaged using ethidium bromide . Viral titers in the lung tissue samples were quantified via plaque assay in MDCK cells as described previously [104] . Viral titers were assessed per gram of tissue . Progesterone , prostaglandin F2α ( PGF2α ) and prostaglandin E2 ( PGE2 ) were quantified via ELISA kits from ALPCO ( Salem , NH ) . Experimental values for lungs and placentas were normalized by the mass of the tissue and the volume in which samples were homogenized . For cytokine expression , a 23-plex Luminex assay from Bio-Rad ( Hercules , CA ) was used . Experimental values for lungs , placentas , and fetuses were normalized by volume and mass of the tissues . Fold changes were expressed relative to uninfected controls . A heat map was constructed to visualize fold changes within the tissues in pregnant and non-pregnant mice . Molecular analysis of protein expression was performed on 5 mg of placental tissue lysate via Western blot . Purified anti-murine MMP-9 ( BioLegend 819701 , 1:500 ) and anti-murine MMP-2 ( BioLegend 680002 , 1:500 ) were used to detect degradative proteins and secondary antibodies ( goat anti-mouse , 1:10 , 000 ) . Rabbit anti-murine COX-2 ( Abcam , ab52237 ) and rabbit anti-murine C13orf24 ( PIBF ) ( Abcam , ab156267 ) were used to detect markers for oxidative and hormonal stress in the lungs and placenta , and goat anti-rabbit ( 1:10 , 000 ) antibodies were used as secondary antibodies . Rabbit anti-murine β-actin ( 1:2000 ) was used as a loading control with secondary antibodies ( goat anti-rabbit , 1:10 , 000 ) . Blots were developed with Super Signal Femto Maximum Sensitivity Substrate ( Thermo Scientific 34096 ) and imaged using a Bio-Rad ChemiDoc Touch . Volume intensity of MMP-2 and MMP-9 was normalized to β-actin loading controls . A two-way ANOVA analyzing factors of protein maturation ( p<0 . 0001 , *** ) and infection ( P<0 . 0296 , * ) was performed in GraphPad Prism 7 . Placentas were isolated 4 d . p . i . ( day 16 of gestation ) and submerged in histology cassettes in 4% paraformaldehyde overnight at 4°C . Tissues were embedded in paraffin , sectioned in 4 μm , and fixed to glass microscopy slides . Hematoxylin and eosin staining was performed by Yerkes National Primate Research Center Pathology Core , and slides were imaged on a Zeiss Akioskop with SpotFlex 15 . 2 camera with Spot Advanced 4 . 7 software . Pathology scores were determined by randomizing histology slides prior to imaging and blindly scoring the incidence of exposed fetal endothelial nuclei ( FEN ) , degradation in the spongiotrophoblast layer , and fibrinoid necrosis ( FN ) . All statistical analysis used a student’s t-test , linear regression , or one-way ANOVA using GraphPad Prism statistical software with a significance level ( α ) of 0 . 05 . Emory University Division of Animal Resources veterinary staff ascertained welfare of animals in addition to research scientists , and DAR staff performed regular care and wellness assessments . Animal work was conducted according to Emory University Institutional Animal Care and Use Committee ( IACUC ) guidelines according to an approved protocol ( DAR2002950-122617BN ) in accordance with the United States federal Animal Welfare Act ( PL 89–544 ) and subsequent amendments . Emory University is registered with the United States Department of Agriculture ( 57-R-003 ) and has filed an Assurance of Compliance statement with the Office of Laboratory Animal Welfare of the National Institutes of Health ( D16-00113 ) . Emory University has been fully and continuously accredited by AAALAC International since 1992 ( Unit 000781 ) . The Georgia Fee-Exempt Wild Animal Permit Customer Number for animals maintained by the Division of Animal Resources is 22257 . The sequence for each genomic segment of the mouse-adapted virus used in this study may be found at NCBI GenBank MG460793 ( HA ) , MG460794 ( M1 ) , MG460795 ( NA ) , MG460796 ( NP ) , MG460797 ( NS1 ) , MG460798 ( PA ) , MG460799 ( PB1 ) , MG460800 ( PB2 ) . | Maternal immunology is finely balanced to maintain a tolerant and supportive molecular environment for the developing fetus while continuing surveillance against foreign microbial threats . Influenza viral infection during pregnancy is a significant clinical risk for mothers and their newborns , increasing hospitalization , preterm birth , low birth weight , and maternal and neonatal deaths worldwide . In a mouse pregnancy model , we show how influenza virus infection disrupts the delicate and interconnected cytokine and hormonal signaling pathways that respond to respiratory pathogens . The health of mothers and offspring was impacted in our study , after pregnant mothers’ lung and placental architecture was compromised by infection . Influenza virus infection increased the stress on the mother’s body already present due to pregnancy , or reversed the hormonal environment required to establish and maintain healthy pregnancy . By dissecting the effects of inflammation post-infection throughout the mother’s anatomy , we can tailor anti-inflammatory treatments for the pregnant population . Also , thorough knowledge of immune responses will assist in tailoring vaccine design and dosage for this delicate period of women’s immunological and reproductive health . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"immune",
"physiology",
"cytokines",
"pathology",
"and",
"laboratory",
"medicine",
"respiratory",
"infections",
"pathogens",
"immunology",
"microbiology",
"orthomyxoviruses",
... | 2017 | H1N1 influenza virus infection results in adverse pregnancy outcomes by disrupting tissue-specific hormonal regulation |
Tumor Necrosis Factor receptor-associated factor-3 ( TRAF3 ) is a central mediator important for inducing type I interferon ( IFN ) production in response to intracellular double-stranded RNA ( dsRNA ) . Here , we report the identification of Sec16A and p115 , two proteins of the ER-to-Golgi vesicular transport system , as novel components of the TRAF3 interactome network . Notably , in non-infected cells , TRAF3 was found associated with markers of the ER-Exit-Sites ( ERES ) , ER-to-Golgi intermediate compartment ( ERGIC ) and the cis-Golgi apparatus . Upon dsRNA and dsDNA sensing however , the Golgi apparatus fragmented into cytoplasmic punctated structures containing TRAF3 allowing its colocalization and interaction with Mitochondrial AntiViral Signaling ( MAVS ) , the essential mitochondria-bound RIG-I-like Helicase ( RLH ) adaptor . In contrast , retention of TRAF3 at the ER-to-Golgi vesicular transport system blunted the ability of TRAF3 to interact with MAVS upon viral infection and consequently decreased type I IFN response . Moreover , depletion of Sec16A and p115 led to a drastic disorganization of the Golgi paralleled by the relocalization of TRAF3 , which under these conditions was unable to associate with MAVS . Consequently , upon dsRNA and dsDNA sensing , ablation of Sec16A and p115 was found to inhibit IRF3 activation and anti-viral gene expression . Reciprocally , mild overexpression of Sec16A or p115 in Hec1B cells increased the activation of IFNβ , ISG56 and NF-κB -dependent promoters following viral infection and ectopic expression of MAVS and Tank-binding kinase-1 ( TBK1 ) . In line with these results , TRAF3 was found enriched in immunocomplexes composed of p115 , Sec16A and TBK1 upon infection . Hence , we propose a model where dsDNA and dsRNA sensing induces the formation of membrane-bound compartments originating from the Golgi , which mediate the dynamic association of TRAF3 with MAVS leading to an optimal induction of innate immune responses .
Following exposure to pathogen-associated molecular patterns ( PAMPs ) , the innate immune response and the subsequent inflammatory reaction rely on evolutionarily conserved receptors termed pattern-recognition receptors ( PRRs ) [1] . These signalling receptors can be expressed at the cellular membrane ( Toll-like receptors ( TLRs ) 1 , 2 , 4 , 5 , and 6 ) , in acidic endosomes ( TLRs 3 , 7 , 8 , and 9 ) , or in the cytoplasmic compartment ( the double-stranded RNA ( dsRNA ) -activated kinase ( PKR ) ; the RIG-I-like helicases ( RLH ) : retinoic-acid-inducible gene I ( RIG-I ) , melanoma differentiation antigen 5 ( MDA5 ) , and LGP2; the HIN-200 family members: Absent In Melanoma 2 ( AIM2 ) and interferon ( IFN ) -inducible IFI16 protein [2]; the DNA-dependent activator of interferon regulatory factors ( IRFs ) ( DAI ) and the nucleotide-binding oligomerization domain ( NOD ) receptors ) . RIG-I and MDA5 have been characterized as important cytoplasmic sensors for viral RNA [3]–[6] . Once activated by dsRNA molecules , RIG-I and MDA5 are recruited to the mitochondrial adaptor protein know as Mitochondrial AntiViral Signaling ( MAVS ) ( also called IPS-1 , Cardif and VISA ) in order to trigger signalling cascades leading to IRF-3 and NF-κB activation , two essential players involved in the establishment of a cellular antiviral state [7]–[10] . Tumor Necrosis Factor ( TNF ) receptor-associated factors ( TRAFs ) are part of a family of adaptor proteins that bridge the intracellular domains of multiple receptors , such as TNFR , IL1R , and TLRs , to downstream effectors involved in the inflammatory and innate immune signalling pathways . The TRAF family is composed of seven members , TRAF1 through TRAF7 . They all share a C-terminal TRAF domain , which is composed of a coiled-coil domain followed by a conserved receptor-interacting domain . This domain mediates self-association and interaction with receptors or signalling proteins . Their N-terminal regions are composed of one or more zinc-finger motifs and , with the exception of TRAF1 , a RING-finger domain that mediates E3 ubiquitin ligase activity and signalling [11] . All mammalian TRAFs localize to the cytoplasm except TRAF4 , which is found in the nucleus . Importantly , gene deletion studies have identified TRAF3 as a critical mediator involved in the induction of the type I interferons ( IFNs ) by the RLH pathway [12] , [13] . TRAF3 has originally been shown to associate with TNF receptors ( e . g . BAFFR , CD40 , LTβR , RANK , CD30 , and Fn14 ) , which are activators of the non-canonical NF-κB pathway [14]–[17] . TRAF3 acts as a negative regulator in this pathway by promoting the recruitment of the TRAF2-cIAP1-cIAP2 E3 ligase complex to NF-κB-inducing kinase ( NIK ) in order to control its rapid turnover in resting cells [18] , [19] . However , in the RLH pathway , the adaptor protein TRAF3 acts as a positive regulator . Its interaction with MAVS and TRADD is important to trigger IRF-3 phosphorylation through the adaptor molecule TANK and the IKK-related kinases TBK1 and IKKi [20] , [21] . The TRADD-mediated recruitment of FADD and RIP1 to MAVS also enhances the interaction between TANK and TRAF3 . A model was then proposed in which TRADD simultaneously organizes FADD- and RIP1-mediated NF-κB signalling on one hand and TRAF3- and TANK-mediated IRF-3 signalling on the other [21] , [22] . However , this possible mechanism of action requires further investigation to determine how TRAF3 is recruited to the mitochondrial adaptor protein MAVS upon viral infection . Here , we have used a proteomics-based strategy to identify novel TRAF3 interacting proteins that are implicated in the induction of type I IFN . Using this approach , we have identified two novel TRAF3 interactors , Sec16A ( also known as KIAA0310 ) and p115 ( also known as USO1 ) , which have characterized roles in the Endoplasmic Reticulum ( ER ) -to-Golgi vesicular transport system . Both proteins were shown to play a primary role in the anterograde trafficking at the ER-Golgi interface by influencing the assembly and transport of coat protein complex II ( COPII ) vesicles . Sec16A assembles on the ER membrane and forms organized scaffold defining ER exit sites ( ERES ) where COPII assembly occurs [23]–[25] . The coiled-coil myosin-shaped molecule p115 was demonstrated to be an important tethering adaptor , which mediates vesicle tethering at the ER [26] , Endoplasmatic Reticulum-Golgi Intermediate Compartment ( ERGIC ) [27] , and in conjunction with tether proteins giantin and GM130 at the cis-Golgi [28] , [29] . Since novel essential mediators of the type I IFN response were recently found to be associated with the ER or the exocyst pathway , such as STING ( also called MITA , ERIS , and MPYS ) , Sec61β and Sec5 [30]–[33] , we postulated that Sec16A and p115 may exert a similar function through the ER-to-Golgi transport compartments . Co-immunoprecipitation experiments and confocal microscopy confirmed the association and co-localization of TRAF3 with p115 and Sec16A . Importantly , overexpression of p115 or Sec16A increased the type I IFN response , whereas their knockdown impaired the induction of antiviral genes . Interestingly , the Golgi apparatus fragmented into cytoplasmic punctate structures following both RLH and cytoplasmic DNA sensor pathway activation , allowing TRAF3 to colocalize and associate with MAVS . Our study identifies p115 and Sec16A as new scaffold proteins involved in the establishment of the antiviral state .
In order to find novel players involved in the type I IFN pathway , we have used a functional proteomics approach based on FLAG affinity purification and mass spectrometry analysis ( AP/MS ) . HEK293 cells stably expressing FLAG-TRAF3 were harvested , subjected to IP with an anti-FLAG antibody under native conditions and FLAG-TRAF3 complexes were analyzed by liquid chromatography coupled to tandem mass spectrometry ( LC-MS/MS ) . In parallel , multiple AP/MS analyses were performed from cells expressing the FLAG alone . Following standard database searches , stringent statistical filtering was performed using SAINT ( see Methods ) . Proteins detected with AvgP≥0 . 7 were manually inspected for frequency of detection across a database of ∼1000 AP/MS analyses , and proteins frequently detected in AP-MS experiments were removed . This resulted in the identification of 12 interaction partners for TRAF3 , including TBK1 , a well-known TRAF3 interactor [12] , [13] . Surprisingly , Sec16A and p115 , two proteins involved in ER-to-Golgi vesicular trafficking , were found to associate with TRAF3 immunocomplexes with a high confidence ( Figure 1A and Figure S1 ) . To confirm these interactions , we performed conventional co-immunoprecipitation experiments by overexpressing the candidate tagged-proteins with FLAG-TRAF3 in 293T cells . The interactions between FLAG-TRAF3 and Myc-p115 or EGFP-Sec16A were clearly detected ( Figure 1B–C ) . To further substantiate the interaction network between TRAF3 , Sec16A and p115 , we additionally established a pool of HEK293 cells stably expressing FLAG-p115 and analyzed its physiological interactors by FLAG affinity purification and LC-MS/MS , followed by analysis with SAINT . FLAG-p115 was found to be associated with nine proteins ( after filtering ) , including Sec16A and GM130 ( also known as GOLGA2 ) , an established physical partner of p115 [34] ( Figure 1D and Figure S1 ) . The interaction of EGFP-Sec16A with Myc-p115 was further confirmed by co-immunoprecipitation experiments ( Figure 1E ) . However , endogenous TRAF3 was not recovered in our FLAG-p115 analysis . This result may be explained by the fact that p115 has many higher-abundance interactors and/or is part of alternative complexes independent from TRAF3 . However , overexpressed Myc-TRAF3 was recovered from FLAG-p115 complexes when the latter were immunoprecipitated from 293T cells co-expressing both constructs ( Figure 1F ) . The TRAF3 interactome network identified by functional proteomics ( Figure S1B ) suggests the presence of at least a fraction of TRAF3 in close proximity to the Golgi network . Members of the TRAF family often share common interacting partners . For example , TRADD and RIP1 strongly bind to TRAF1 , TRAF2 and TRAF3 [21] , [35] , whereas the mitochondrial anti-viral signaling protein MAVS interacts with TRAF2 , TRAF3 and TRAF6 [10] , [20] . To verify the binding selectivity of the newly identified TRAF3 interactors , we next performed co-immunoprecipitation experiments in 293T cells overexpressing FLAG-tagged TRAF2 , TRAF3 or TRAF6 , TRAF molecules involved in type I IFN and inflammatory responses , along with Myc-p115 or EGFP-Sec16A . Only Myc-p115 was found to be enriched in FLAG-TRAF3 immunocomplexes ( Figure S1C ) . A similar result was obtained with EGFP-Sec16A , except that a weak enrichment was observed with TRAF2 when compared to TRAF3 ( Figure S1D ) . To further validate the interaction between TRAF3 and these new interactors , we next analyzed their subcellular localization by confocal microscopy . The vesicle-tethering protein p115 is known to colocalize and interact specifically with the NH2 terminus of the cis-Golgi protein GM130 ( [36] and see Figure 2A , panel 2 ) . Upon ectopic expression of Myc-p115 and FLAG-TRAF3 , we observed a co-localization of these two proteins ( Figure 2A , panel 1 ) . FLAG-TRAF3 was also observed to localize to the Golgi apparatus where it exhibits a high degree of overlap with the cis-Golgi marker GM130 ( Figure 2A , panel 3 ) . p115 was previously reported to be present in the ERGIC , through an interaction involving activated Rab1 [37] , [38] . This cellular localization of FLAG-p115 can be visualized with the conventional ERGIC marker , ERGIC53 ( Figure 2A panel 4 ) . Notably , FLAG-TRAF3 ( or Myc-TRAF3 ( unpublished data ) ) was also present in the ERGIC ( Figure 2A , panel 5 ) . No significant colocalization was detected between TRAF3 and the ER marker calnexin ( Figure 2A , panel 6 ) , the lysosomal compartments ( Figure 2A , panel 7 ) and the mitochondrial network ( Figure 2A , panel 8 ) . In HeLa cells , Sec16A was demonstrated to define ERES [24] , [25] , localizing to punctate structures on the ER membrane . This pattern was reproduced in this study ( Figure 2B , panel 1 ) . Since Rab1 recruitment of p115 to ERES [26] represents an essential step for the subsequent docking of ER-derived vesicles to the ERGIC [39] , we next examined the colocalization of EGFP-Sec16A and FLAG-p115 . The two proteins clearly colocalized at the perinuclear region ( Figure 2B , panel 2 ) . FLAG-TRAF3 and EGFP-Sec16A also mainly colocalized at the perinuclear region in HeLa cells ( Figure 2B , panel 3 ) . Moreover , a colocalization of FLAG-TRAF3 with endogenous Sec16A at ERES distributed in the cytoplasm could be observed . However , some FLAG-TRAF3 punctae also appeared in close proximity to those containing Sec16A ( Figure 2B , panel 4 , compare arrows ) . Ectopic expression of FLAG-TRAF2 and FLAG-TRAF6 revealed that only TRAF3 exhibits a cellular Golgi-like distribution ( Figure S2A ) and colocalizes with endogenous Sec16A ( Figure S2B ) or Myc-p115 ( Figure S2C ) . Importantly , endogenous staining of TRAF3 revealed that the majority of TRAF3 proteins localized to the juxtanuclear region containing both the cis-Golgi marker GM130 and the ERGIC marker ERGIC53 ( Figure 2C ) . To further confirm the localization of TRAF3 to the Golgi apparatus , we next treated the cells with nocodazole . Microtubule depolymerization is known to result in the reorganization of the Golgi complex into characteristic mini-stacks , which appear as punctate structures throughout the cell [40] . Nevertheless , FLAG-TRAF3 was detected to colocalize with GM130 and Myc-p115 in cells treated with nocodazole ( Figure S3 , panels 1 , 2 and 3 ) . Treatment with brefeldin A ( BFA ) , leads to relocalization of the components of the cis-Golgi matrix to cytoplasmic punctate structures ( also called remnants ) that appear close to ERES [41] , [42] . GM130 and p115 are cis-Golgi proteins , which are known to be relocalized to these remnants [41] . FLAG-TRAF3 was also relocated to cytoplasmic remnants upon BFA treatment , where it co-localized with GM130 and Myc-p115 ( Figure S3 , panels 4 and 5 ) . Altogether , results from our pharmacological experiments and confocal microscopy strongly suggest that TRAF3 localized to ER-to-Golgi transport compartments , where it tightly associates [43] , [44] . It has been proposed that a structurally intact TRAF3 molecule is required for its biological function . Indeed , TRAF3 lacking its N-terminal RING or the C-terminal TRAF domain lacks antiviral activity [20] . We therefore examined the subcellular localization of TRAF3 deletion mutants in reconstituted TRAF3 knockout MEF cells . Removal of the N-terminal Ring Finger domain ( Figure 3A , panel 1 ) , the N-terminal Ring and Zinc finger domains ( Figure 3A , panel 2 ) or the C-terminal TRAF domain ( Figure 3A , panel 3 ) resulted in TRAF3 molecules that no longer colocalize with the Golgi marker GM130 . Furthermore , coimmunoprecipitation experiments in 293T cells revealed that immunocomplexes containing p115 are detected only with full length TRAF3 and that Sec16A-containing immunocomplexes required at least the isoleucine zipper and the TRAF domain ( Figure 3B ) . Moreover , TRAF3 is known to interact with several substrates containing a particular motif ( PxQxS/T ) called the TRAF interaction motif ( TIM ) [45] . The mutation of two amino acids located in the TIM-binding pocket of TRAF3 , Y440 and Q442 , abrogates these interactions [20] . Interestingly , a strong interaction was detected between FLAG-TRAF3 Y440/Q442A and Myc-p115 or EGFP-Sec16A ( Figures 3C and 3D ) , implying that this interaction is independent of the TIM motif . Thus , it is not clear yet whether TRAF3 interacts directly with Sec16A or p115 or requires other components such as TFG ( [46] and see Figure S1 ) . Collectively , these data suggest that an intact TRAF3 molecule is required for its proper localization and interaction with components of the ER-to-Golgi vesicular pathway . Our data demonstrate that TRAF3 does not associate with the mitochondrial network in resting cells ( Figure 2A , panel 8 ) . However , TRAF3 was demonstrated to link the mitochondrial membrane-bound protein MAVS to the activation of TBK1 , which is required for IRF3/7 phosphorylation and type I IFN induction in response to viral infection [20] , [47] . Therefore , we next addressed the subcellular localization of endogenous TRAF3 upon viral infection and RNA/DNA sensor pathway activation . Intracellular delivery of the double-stranded RNA mimicry molecule , poly I:C , or the dsDNA mimicry agent poly dA:dT resulted in disorganization of the ribbon-like structure of the Golgi apparatus , giving rise to the formation of Golgi ministacks containing GM130 ( Figure 4A , arrows in panel 2 and 3 ) . Importantly , the localization of endogenous TRAF3 followed these Golgi fragments . Similar observations were made in cells infected with RIG-I inducers , Sendai virus ( SeV ) , Respiratory Syncytial Virus ( RSV ) and Influenza virus ( Figure 4B ) . Additionally , we addressed the association of TRAF3 with p115- and Sec16A-containing complexes upon PAMP exposure . In unstimulated cells , a weak but constitutive association of endogenous TRAF3 with endogenous Sec16A and p115 was detected ( Figure 5 A and B ) . However , upon viral infection or transfection with poly I:C or poly dA:dT , immunocomplexes containing endogenous TRAF3 were enriched with p115 and Sec16A . Importantly , the induced association of TBK1 with TRAF3 closely mirrored the presence of p115 and Sec16A . ( Figure 5B ) . From these results we hypothesized that the localization of TRAF3 to the ER-to-Golgi compartment and the Golgi fragmentation of the latter into punctate structures might be required for the proper positioning of TRAF3 with MAVS . To verify this hypothesis , loss-of-function experiments were conducted using HeLa cells exposed to siRNA duplexes targeting Sec16A and p115 . As previously observed for p115 and Sec16A [24] , [25] , [48] , [49] , reducing the expression level of Sec16A or p115 led to a drastic disorganisation of the Golgi paralleled by a relocalization of TRAF3 as observed by the formation of small punctate structures ( Figure S4A , panels 2 and 4; Figure S4B , panel 2 ) . However , the majority of these GM130 positive punctae do not colocalize with TRAF3 and thus appear to be different from those observed following dsRNA and dsDNA sensing ( compare Figure S4A , panel 4 with Figure 4 ) . Next we examined the effect of reducing the expression level of Sec16A or p115 on the ability of TRAF3 to colocalize with MAVS upon SeV infection . As expected , TRAF3 localization reorganized into punctate structures following SeV infection , allowing a significant proportion of TRAF3 to colocalize with MAVS ( Figure 6A , panel 2 ) . This effect was severely compromised by reducing the expression of Sec16A or p115 ( Figure 6A , compare panels 4 and 6 with panel 2 ) . Additionally , co-immunoprecipitation experiments revealed that TRAF3 formed an immunocomplex with MAVS upon SeV infection . Interestingly , silencing the expression level of p115 or Sec16A clearly blunted the ability of TRAF3 to bind to MAVS upon SeV infection ( Figures 6B and 6C ) . Thus loss-of-function experiments targeting p115 and Sec16A led to a mislocalization of TRAF3 and its subsequent incapacity to associate with MAVS upon RLH pathway activation . This prompted us to ask whether enforced retention of TRAF3 at the ER-to-Golgi compartment could negatively influence the type I IFN response . In order to verify this , a TRAF3 mutant containing a COPI and COPII sorting signal peptide [50] , namely “AKKFF” [51] , at its C-terminal end was generated and used in confocal microscopy and reporter gene assays . Confocal microscopy experiments revealed that addition of dilysine and dihydrophobic residues to the C-terminal end of TRAF3 resulted in the formation of large TRAF3 aggregates which failed to colocalize with the Golgi marker GM130 upon infection with SeV ( Figure 7A ) . Consequently , the ability of the TRAF3-AKKFF mutant to mediate TRAF3-dependent synergistic activation of the IFNβ promoter was drastically reduced ( Figure 7B ) , which is likely due to less binding to MAVS ( Figure S5 ) . Altogether , these results indicate that the localization of TRAF3 to the ER-to-Golgi compartment is involved in the proper positioning of TRAF3 within the mitochondrial network and the induction of type I IFN innate immune response . The results presented above suggest a role for the ER-to-Golgi compartment in TRAF3-dependent innate immune response . To investigate whether Sec16A or p115 play a role in the type I IFN response , we overexpressed both proteins in Hec1B cells and assessed NF-κB and IRF-3 transcription factor activation using reporter gene assays . Without any stimulation , overexpression of either protein did not significantly activate the IFNβ promoter . However , following viral infection , the response was increased in cells overexpressing Sec16A or p115 ( Figure 8A ) . Overexpression of Sec16A and p115 also increased the activation of the ISG56 promoter ( IRF3-dependent promoter ) ( Figure 8B ) and the NF-κB-dependent promoter ( Figure 8C ) following SeV infection . Moreover , we observed a synergistic effect on IFNβ promoter activity when Sec16A or p115 were co-expressed with MAVS ( Figure 8D ) , TBK1 ( Figure 8E ) and , interestingly , the TLR3 essential effector TRIF ( Figure 8F ) . Similar results were also obtained for the ISG56 promoter and the NF-κB-dependent promoter ( Figure S6 ) . To further substantiate that the positive transcriptional effect of Sec16A and p115 is dependent on TRAF3 , TRAF3-knockout MEF cells were transfected with p115 and Sec16A in the presence or absence of TRAF3 and used in the IFNβ promoter reporter assay . As suspected , the enhanced promoter activation , induced by ectopically expressed p115 and Sec16A , was entirely dependent on the presence of TRAF3 ( Figure 8G ) . Thus , when expressed in relatively low amounts in Hec1B and MEF cells ( not shown ) , p115 and Sec16A positively participate in a TRAF3-dependent type I IFN response , probably reflecting the ability of a subpopulation of cytoplasmic TRAF3 to further associate with the ER-to-Golgi components under these conditions of mild ectopic expression . Interestingly , several recent studies have demonstrated that overexpression of p115 or Sec16A in highly transfectable cell lines and depletion of Sec16 or p115 resulted in identical cellular outcomes ( i . e . Golgi fragmentation ( see Figure S4 and Figure 6 ) and delayed ER-to-Golgi transport ) , thereby suggesting that they are required in stoichiometric amounts [24] , [25] , [52] . Thus , when ectopically expressed in high amounts in 293T cells , p115 and Sec16A were expected to blunt TRAF3-dependent transcriptional activation . Indeed , transfection of increasing amounts of p115 or Sec16A efficiently blunted TRIF- , RIG-I- , and MAVS-induced IFNβ promoter activation ( Figure S7A–C ) as well as NF-κB promoter activation ( data not shown ) . Importantly , adding increasing amounts of TRAF3 in this specific reporter gene assay dose-dependently reversed the inhibitory effect of p115 and Sec16A , once more substantiating the relationship that exist between Sec16A , p115 and TRAF3 ( Figure S7D ) . Moreover , transfection of these plasmids also blunted TBK1-induced ISRE promoter activation ( Figure S7E ) , but did not affect the transactivation response induced by the use of a constitutively active form of IRF-3 ( IRF3-5D ) ( Figure S7F ) , suggesting that the ER-to-Golgi compartment plays upstream of IRF-3 in type I IFN signalling . To further confirm the implication of p115 and Sec16A in the type I IFN response , loss-of-function experiments were conducted next . As suspected , an RNAi approach targeting Sec16A and p115 , which leads to Golgi fragmentation ( see Figure S4 and Figure 6 ) significantly diminished Ifnb , ifit1 ( ISG56 ) , and oas1 mRNA induction following poly I:C and poly dA:dT transfection and SeV infection ( Figure 9 ) . To verify whether this approach affected IRF-3 activation and the induction of an IRF-3-dependent antiviral protein [53] , we next verified the phosphorylation state of IRF-3 and the induction of ISG54 in HeLa cells expressing either shRNA duplexes targeting p115 and Sec16A or cells expressing a non-targeting ( NT ) shRNA duplex . The phosphorylation of IRF-3 and the expression of ISG54 were readily observed upon SeV infection , poly I:C and poly dA:dT transfection in HeLa cells expressing the NT shRNA duplex but was clearly reduced in cells expressing different shRNA duplexes targeting p115 ( Figure 10A ) and Sec16A ( Figure 10B ) . Altogether , these data indicate that TRAF3 localization to the ER-to-Golgi vesicular pathway is necessary for a proper type I IFN response .
Gene disruption strategies have revealed that TRAF3 plays a major role in the type I IFN response [12] , [13] . However , how TRAF3 assembles into functional signalling complexes is still not fully understood . In general , TRAF3 is thought to reside in the cytosol and translocate to surface membrane receptors upon engagement of CD40 or other TNFR family members [54] . Akin to its role in MyD88-dependent cytokine production and TRIF-dependent type I IFN production [55] , TRAF3 conceivably also has the capacity to associate with endosomal compartments enriched in TLR3 , TLR4 , TLR7 , TLR8 and TLR9 receptors [56] . Additionally , upon RLH activation TRAF3 interacts with MAVS and TRADD to trigger IRF-3 phosphorylation through the adaptor molecule TANK and the IKK-related kinases TBK1 and IKKi [20] , [21] . However , how TRAF3 associates with MAVS upon RLH activation remains unanswered . Herein , we report that TRAF3 localizes to the ER-to-Golgi compartments through its ability to interact with p115- and Sec16A-containing complexes . A pharmacological approach using the microtubule depolarizing agent nocodazole led to the redistribution of TRAF3 into small punctate cytoplasmic structures discrete from the ER along with both Golgi matrix proteins p115 and GM130 . Both the structure and positioning of the Golgi apparatus have been shown to be highly dependent on the microtubule cytoskeleton [57] . Interestingly , a link between TRAF3 and the microtubule network has been already established in a previous study through its interaction with Microtubule-Interacting Protein that associates with TRAF3 ( MIP-T3 ) [58] . TRAF3 was dissociated from this complex upon CD40L stimulation and , consequently , it was suggested that microtubule association of TRAF3 could be responsible for directing TRAF3 to defined membrane microdomains in the cell . A similar scenario is proposed here where , in response to viral infection , the association of TRAF3 with complexes containing p115 and Sec16A at the ER-to-Golgi vesicular pathway may play an important role in positioning TRAF3 with MAVS ( see Figure 11 ) . Indeed , the following findings suggests a role for Sec16A and p115 in the TRAF3-mediated RLH type I IFN response: ( 1 ) Sec16A and p115 are found in immunocomplexes containing TRAF3 , but not TRAF2 or TRAF6; ( 2 ) inactivation of TRAF3 by deletion of its N-terminal RING finger domain and the C-terminal TRAF domain displaces TRAF3 from the ER-Golgi transport compartments; ( 3 ) in non-treated cells , TRAF3 colocalizes and tightly associates with p115 , Sec16A , ERGIC53 and GM130 , markers of the ER-to-Golgi vesicular compartment; ( 4 ) activation of the RLH pathway leads to reorganization of the Golgi apparatus into punctate structures containing TRAF3 and GM130; ( 5 ) an increased association between TRAF3 , Sec16A , p115 and TBK1 is observed in virally-infected , dsRNA- and dsDNA-transfected cells; ( 6 ) mild overexpression of both proteins enhances SeV- , TBK1- and MAVS-stimulated IFNβ , ISG56 and NF-κB promoter induction; ( 7 ) knocking down the expression level of p115 or Sec16A affects the cellular distribution of TRAF3 , impairs its capacity to associate with MAVS and diminishes the type I IFN response following poly I:C or polydA:dT transfection and SeV infection; and ( 8 ) enforced retention of TRAF3 at the ER-to-Golgi compartment by the addition of a COPI and COP II sorting signal peptide impairs TRAF3 recruitment to the cis-Golgi and diminishes the type I IFN response . Thus , we propose that these two trafficking proteins , Sec16A and p115 , form a complex with TRAF3 at ER-to-Golgi transport compartments in order to ensure its proper recruitment to the mitochondrial network during a viral infection . Interestingly , enforced expression of Sec16A or p115 also increases TRIF-mediated IFNβ promoter activation , reinforcing the role for the ER-to-Golgi vesicular compartment in TLR3 and TLR4 signalling , as recently reviewed [59] . In support of our findings , the ER-to-Golgi transport compartment seems to also host several proteins involved in type I IFN signalling such as TRADD [21] , the translocon [31] and potentially the exocyst [32] ( Clement and Servant , unpublished observations ) . How these proteins cooperate with TRAF3 at the ER-to-Golgi transport compartments is currently unclear and will be the objective of future studies . Nevertheless , all these data suggest a model where vesicles and/or membranes originating from reorganized ER-to-Golgi compartments come in close proximity with the mitochondrial network in order to facilitate the assembly of a functional MAVS signalling complex . In addition to its role in the RNA sensing pathways , STING is now considered an important effector of innate immune signalling in response to DNA pathogens [60] . Interestingly , STING is an ER-resident protein , which in response to dsDNA treatment , was recently demonstrated to traffic from the ER to the Golgi [61] , [62] giving rise to punctate structure formation [62] . It is likely that the use of dsDNA ( polydA:dT ) used in our study might activate both the RNA-dependent pathway ( through RNA polymerase III [63] ) and the recently described DNA-dependent pathway ( through IFI16 [2] ) , allowing TRAF3-loaded punctae to interact with both MAVS and STING respectively for proper innate immune signalling ( Figure 11 ) . Even though this needs to be investigated further , we speculate that the membranous network composed of the ER , Golgi and mitochondria provides a convenient platform on which antiviral cell-signalling complexes are arranged and optimally activated . It is noteworthy that , as a common feature , plus-stranded RNA viruses have the ability to induce cytoplasmic membrane rearrangements that facilitate their replication . Consequently , the formation of these RNA replication complexes results in dramatic reorganization of the secretory pathway of host cells [64] . For example , poliovirus-infected cells accumulate membranous vesicles derived from COPII vesicles [65] whereas Kunjin virus induces “convoluted membranes” that contain markers from the ERGIC [66] . The precise role for this internal membrane rearrangement in the virus propagation and virus-host interaction requires further investigation . Nevertheless , localization of TRAF3 and TRADD to these vesicular transport compartments could represent a cellular strategy to increase the rate of RNA detection and the formation of an effective signalling complex at the mitochondrial membrane . The observation that TRADD translocates from the cytoplasm to the mitochondria during Influenza A virus infection supports this model [67] . Additionally , recent observations highlight the fact that viruses have evolved a variety of mechanisms involving the Golgi apparatus to specifically block TRAF3 recruitment into a functional signalling complex . Notably , the SARS Coronavirus M protein , a Golgi localized protein , was recently found to impede the formation of a TRAF3-TANK-TBK1/IKKi complex at the Golgi apparatus [68] . The NY-1 strain Hantavirus glycoprotein ( Gn ) was also shown to disrupt TRAF3-TBK1 interaction by interacting with TRAF3 through its cytoplasmic tail [69] . The notion of cellular proximity to favor exchanges and signalling events between organelles has been an intense field of interest for many years . Recently , mitofusin 2 present on the ER was shown to tether the ER to mitochondria in order to promote efficient Ca2+ uptake into the mitochondria for oxidative phosphorylation purposes . Interestingly , mitofusin 2 was also shown to inhibit RLH pathway signalling by interacting with the C-terminal of MAVS [70] . Furthermore , the Golgi localization of the glycolipid GD3 is important for its transport to the mitochondria after TNF-α stimulation [71] , [72] . Membrane scrambling between Golgi and mitochondria following Fas stimulation is another example pointing to the connection between different cellular organelles [73] . Moreover , signalling at the Golgi apparatus and endosomes has been observed for different types of membrane-bound receptors [56] , [74] and protein kinase cascades [75] . Although Bouwmeester and colleagues reported an NF-κB-inducing kinase-dependent interaction between Sec16A and NF-κB 2/p100 in an exhaustive study mapping the human TNF-α/NF-κB signal transduction network [76] , a role for the ER-to-Golgi vesicular pathway in RLH-induced innate immune response was still unknown until now . Future characterization of the TRAF3 interactome will undoubtedly help to understand the molecular relevance of the specific subcellular localization of TRAF3 for an optimal type I IFN response .
Commercial anti-GM130 antibody was purchased from BD Transduction ( San Jose , CA ) . The monoclonal anti-FLAG epitope ( M2 ) , the polyclonal anti-FLAG and the anti-β-actin ( clone AC-74 ) were obtained from Sigma ( Oakville , Ontario , Canada ) . The c-Myc ( 9E10 ) monoclonal antibodies , as well as the polyclonal p115 ( H-300 ) and TRAF3 ( C-20 , H-20 , and G-6 ) antibodies were purchased from Santa Cruz ( Santa Cruz , CA ) . The anti-GFP ( monoclonal 1218 ) antibody and the polyclonal goat anti-GFP antibody were obtained from ABCAM ( Cambridge , MA ) and US Biological ( Swampscott , MA ) respectively . The anti-ERGIC53 and anti-calnexin antibodies were from Enzo Life Sciences , anti-p-IRF3 Ser398 was from Millipore ( Billerica , MA ) and anti-ISG54 was from Novus Biologicals ( Littleton , CO ) . The polyclonal anti-Sec16A and p115 antibodies were obtained from Bethyl Laboratories and Santa Cruz . The plasmid encoding for EGFP-Sec16A was a kind gift of Dr . David Stephens ( University of Bristol , UK ) . Human TRAF3 and p115 cDNAs were amplified from the MGC bank collection and respectively subcloned in pcDNA3 and pTag2B ( FLAG ) or pTag3B ( Myc ) vectors ( Invitrogen , Burlington , ON , Canada ) . Human TRAF6 cDNAs were purchased from Origene ( Rockville , MD ) and subcloned in the pTag2B/3B vectors . The pFLAG-CMV2-TBK1 and pFLAG-TRAF3 Y440/Q442A were gifts from Drs . John Hiscott ( McGill University ) . pcDNA3 . 1-FLAG-MAVS construct was from Rongtuan Lin ( McGill University ) . The pcDNA3-His-TRIF construct was from Dr . Daniel Lamarre ( Université de Montréal ) . The pRK5-TRAF2-FLAG was obtained from Dr . Nathalie Grandvaux ( Université de Montréal ) . pFLAG-CMV2 TRAF3 deletion mutants ( 1–117 , 1–381 , 114–568 , 259–568 and 389–568 ) were from Dr . Carl Ware ( La Jolla Institute for Allergy and Immunology ) . The IFNß reporter plasmid , pGL3-IFN-ß-LUC was described previously [77] as well as the ISG56-luciferase [78] and the NF-κB p2 ( 2 ) TK reporter plasmids [77] . The pFLAG-TRAF3 mutant with C-terminal retention motif AKKFF was generated by PCR and subcloned in pCDNA3 . 1 ( + ) and pMRX-ires-puro ( a kind gift from Dr . Shoji Yamaoka , Tokyo Medical and Dental University , Japan ) . Poly I:C was purchased from GE HealthCare ( Waukesha , WI ) and transfected with Lipofectamine2000 ( Invitrogen ) at final concentrations of 1 . 0 to 2 . 5 µg/ml . Poly dA:dT was from InvivoGen and used at 1 µg/ml . BFA and nocodazole were obtained from Calbiochem and used at a final concentration of 5 µg/ml . HeLa , Hec1B , HEK 293 , HEK 293T , HEK 293 QBI cell lines and TRAF3 knockout MEF cells ( a kind gift from Dr . John Hiscott , McGill University ) were maintained in Dulbecco's modified Eagle Medium supplemented with 10% fetal bovine serum . All DNA transfections in human cell lines were performed with Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's protocol . Transient transfection of immortalized MEF cells was performed by microporation with the Microporator Apparatus ( Montreal Biotech ) according to the manufacturer's instructions . Sendai virus ( SeV ) was obtained from Specific Pathogen-Free Avian Supply ( North Franklin , CT ) and used at 200 HAU/ml . Respiratory Syncytial Virus ( RSV . A2 ) ( a kind gift from Nathalie Grandvaux , Université de Montréal ) was used at a MOI of 3 . Influenza A ( PR8 ) virus was a kind gift from Dr . Rongtuan Lin ( McGill University ) . Preparation of whole cell extracts , co-immunoprecipitation studies , Native-PAGE and immunoblot analysis were performed as described previously [79] . A RIPA buffer ( 50 mM Tris-HCl , pH 7 . 4 , 100 mM NaCl , 5 mM EDTA , 50 mM sodium fluoride , 40 mM β-glycerophosphate , 1 mM sodium orthovanadate , 1% Triton X-100 , 0 . 1% SDS , 0 . 5% sodium deoxycholate , and protease inhibitors mixture ( Sigma ) ) was used for the extraction of the TRAF3 AKKFF mutant . Antibodies were used as recommended by the manufacturers . For immunofluorescence , cells were fixed with 4% paraformaldehyde ( PFA ) in PBS for 20 min followed by permeabilization with 0 . 1% Triton X-100 for 5 min . Cells were washed with PBS ( pH 7 . 2 ) and blocked with 0 . 5% BSA in PBS . Anti-FLAG antibody ( M2 , Sigma ) was used at 1∶1000 , anti-GM130; 1∶100 , anti-ERGIC53; 1∶100 , anti-FLAG polyclonal antibody; 1∶400 , anti-GFP ( ABCAM ) ; 1∶100 , anti-Myc 9E10; 1∶100 , anti-TRAF3; 1∶200 , anti-P115; 1∶100 , anti-Sec16A; 1∶200 , and anti-MAVS; 1∶100 . Secondary fluorophore-conjugated antiserum ( Alexa Fluor 488 and 564 ) was obtained from Molecular Probes ( Eugene , OR ) and used at 1∶500 in PBS 0 . 5% BSA . The nucleus was revealed by 4′ , 6-diamidino-2-phenylindole ( DAPI ) staining . The confocal micrographs represent a single optical section through the plane of the cell . Images were acquired with LSM v3 . 2 software ( Zeiss ) on a LSM 510 inverted microscope ( Zeiss , Germany ) with a plan-apochromat 63×/1 . 4 oil disc lens using 405 nm in conjunction with a LP 505 for DAPI , 488 nm in conjunction with a BP 505–530 for Alexa 488 , and 543 nm in conjunction with BP 560–615 for Alexa 568 . Images were assembled in Adobe Photoshop CS 3 . 0 . FLAG-affinity purification was performed as described previously [80] with the following modifications . Detergent concentration in the lysis buffer was 0 . 5% NP-40; the lysis buffer was added at 4 ml/g wet cell pellet , and cells were subjected to passive lysis ( 30 minutes ) followed by one freeze-thaw cycle and centrifugation . Immunoprecipitation was performed on the cleared lysate by adding 25 µl packed FLAG M2 beads ( Sigma ) and incubating for two hours . Beads were washed three times in lysis buffer , and three times in 50 mM ammonium bicarbonate . Samples were eluted with ammonium hydroxide , lyophilized in a speed-vac , resuspended in 50 mM ammonium bicarbonate ( pH 8–8 . 3 ) , and incubated at 37°C with trypsin overnight . The ammonium bicarbonate was evaporated , and the samples were resuspended in HPLC buffer A2 ( 2% acetonitrile , 0 . 1% formic acid ) , then directly loaded onto capillary columns packed in-house with Magic 5 µm , 100A , C18AQ . MS/MS data was acquired in data-dependent mode ( over a 2 hr acetonitrile 2–40% gradient ) on a ThermoFinnigan LTQ equipped with a Proxeon NanoSource and an Agilent 1100 capillary pump . Acquired RAW files were converted to mgf format using ProteoWizard . The searched database was human RefSeq ( version 45 ) . * . mgf files were searched with the Mascot search engine ( version 2 . 3 ) using the following variable parameters: semi trypsin digestion , one missed cleavage allowed , asparagine deamidation and methionine oxidation . The fragment mass tolerance was 0 . 6 Da ( monoisotopic mass ) , and the mass window for the precursor was +/−3 Da ( only +2 and +3 charge ions were processed ) . Mascot results were parsed for further analysis into a LIMS system developed at the Samuel Lunenfeld Research Institute [81] . Scoring of specific interactors for FLAG-TRAF3 and FLAG-p115 was performed using the statistical tool SAINT ( Significance Analysis of INTeracome ) . SAINT converts label free quantification , such as spectral counts , for each prey protein identified in a purification of a bait into the probability of true interaction between the two proteins [82] , [83] . SAINT can calculate a probability of interaction even for proteins proteins frequently detected in AP-MS experiments , providing that a quantitative enrichment is detected in the purification of the sample [84] . For each bait , two biological replicates were used . Twelve negative control runs ( consisting of cells expressing the FLAG tag alone ) were processed in parallel and combined into 5 virtual controls for SAINT modeling . SAINT calculates scores differently depending on the availability of negative control purifications , and thus the implementation for spectral count data incorporating control purification data was used ( details are described in [82] ) . The probability score was first computed for each prey in independent biological replicates separately ( iProb ) . Then the final probability score for a pair of bait and prey proteins was calculated by taking the average of the probabilities in individual replicates ( AvgP ) ; final results with AvgP≥0 . 5 were further inspected . A manual cross-reference against a database containing >1000 independent FLAG AP-MS runs was finally performed to identify potential proteins frequently detected in AP-MS experiments and were removed from the final dataset . HeLa cells were transfected with 40 nM siRNA using Lipofectamine2000 ( Invitrogen ) . siRNA p115 , Sec16A and the non-targeting pool siRNA duplexes were purchased from Dharmacon ( Lafayette , CO ) . Sequences are as follow: Sec16A ( #3: 5′-ggagagcuuucgcgcugua-3′; #4: 5′-ccucaguccucuagcgugu-3′ ) and p115 ( #3: 5′-guuauuauguggagguuug-3′; #4: 5′-ugauggagguauaguaguu-3′ ) . shRNA vectors targeting p115 ( TRCN0000065070 , TRCN0000065071 , TRCN0000065072 ) and Sec16A ( TRCN0000246015 , TRCN0000246016 ) and non-Targeting control shRNA were purchased form Sigma ( St . Louis , MO ) . Lentiviral vector production and transduction was conducted as described previously [85] . After stimulation , total RNA was extracted from HeLa cells using Trizol reagent ( Invitrogen ) . 2 µg of RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription Kit with random primers ( Applied Biosystems ) as described by the manufacturer . SYBR green PCR reactions were performed using 2 µl of cDNA samples ( 25–50 ng ) , 5 µl of the Fast SYBR qPCR Master Mix ( Applied Biosystems ) and 10 pmol of each primer in a total volume of 10 µl . The IFN qRT-Primer set for real-time quantification of the IFN response ( IFNβ , ISG56 ( ifit1 ) and OAS1 ) was purchased from InvivoGen ( San Diego , CA ) . The ABI PRISM 7900HT Sequence Detection System ( Applied Biosystems ) was used to measure the amplification level . All reactions were run in triplicate and the average Cts were used for quantification . TBP ( TATA binding protein ) was used as endogenous control . Subconfluent Hec1B , HEK 293 QBI and 293T cells in 24 well-plates were transfected with 25 ng of pRL-TK reporter ( renilla luciferase for internal control ) and 125 ng of pGL3-IFN-β-LUC , pGL3-ISG56-LUC , pGL3-ISRE or pGL3-NF-κB-LUC using the conventional CaPO4 transfection protocol ( for Hec1B , HEK 293 QBI cells and 293T cells ) or Lipofectamine 2000 ( for TRAF3 knockout MEF cells ) . Cells were harvested 24 h post-transfection , lysed in passive lysis buffer ( Promega , Madison , WI ) , and assayed for dual-luciferase activity using 10 µl of lysate according to the manufacturer's instructions . All firefly luciferase values were normalized to renilla luciferase to control for transfection efficiency . Statistical analyses were performed using GraphPad Prism version 5 . 0 for Mac ( GraphPad Software , San Diego , CA ) . Comparison of two groups was carried out using a two-tailed unpaired t-test , and comparison of more than two groups was carried out with one-way ANOVA and a Bonferroni posttest . Statistical significance was accepted at a P-value below 0 . 05 . TRAF2; 7186 , TRAF3; 7187 , TRAF6; 7189 , Sec16A; 9919 , p115; 8615 , MAVS; 57506 , STING; 340061 , ISG56; 3434 , ISG54; 3433 , TRIF; 148022 , TBK1; 29110 , RIG-I; 23586 , MDA5; 64135 , LGP2; 79132 , AIM2; 9447 , IFI16; 3428 , IRF3; 3661 , NIK; 9020 , TRADD; 8717 , TANK; 10010 , IKKi; 9641 , GM130; 2801 , RIP1; 8737 , IFNβ: 3456 , NFκB: 4790 , OAS1; 4938 , ERGIC53; 3998 , calnexin; 821 , Sec61β; 10952; Sec5; 55770 . | In response to pathogens , such as viruses and bacteria , infected cells defend themselves by generating a set of cytokines called type I interferon ( IFN ) . Since Type I IFN ( namely IFN alpha and beta ) are potent antiviral agents , understanding the cellular mechanisms by which infected cells produce type I IFN is required to identify novel cellular targets for future antiviral therapies . Notably , a protein called Tumor Necrosis Factor receptor-associated factor-3 ( TRAF3 ) was demonstrated to be an essential mediator of this antiviral response . However , how TRAF3 reacts in response to a viral infection is still not totally understood . We now demonstrate that , through its capacity to interact with other proteins ( namely Sec16A and p115 ) that normally control protein secretion , TRAF3 resides close to the nucleus in uninfected cells , in a region called the ER-to-Golgi Intermediate Compartment ( ERGIC ) . Following viral infection , the ERGIC reorganizes into small punctate structures allowing TRAF3 to associate with Mitochondrial AntiViral Signaling ( MAVS ) , an essential adaptor of the anti-viral type I IFN response . Thus , our study reveals an unpredicted role of the protein secretion system for the proper localization of TRAF3 and the antiviral response . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunity",
"innate",
"immunity",
"immunity",
"to",
"infections",
"immunology",
"biology",
"immune",
"response"
] | 2012 | Proteomic Profiling of the TRAF3 Interactome Network Reveals a New Role for the ER-to-Golgi Transport Compartments in Innate Immunity |
Cell-cycle progression and cell division in eukaryotes are governed in part by the cyclin family and their regulation of cyclin-dependent kinases ( CDKs ) . Cyclins are very well characterised in model systems such as yeast and human cells , but surprisingly little is known about their number and role in Plasmodium , the unicellular protozoan parasite that causes malaria . Malaria parasite cell division and proliferation differs from that of many eukaryotes . During its life cycle it undergoes two types of mitosis: endomitosis in asexual stages and an extremely rapid mitotic process during male gametogenesis . Both schizogony ( producing merozoites ) in host liver and red blood cells , and sporogony ( producing sporozoites ) in the mosquito vector , are endomitotic with repeated nuclear replication , without chromosome condensation , before cell division . The role of specific cyclins during Plasmodium cell proliferation was unknown . We show here that the Plasmodium genome contains only three cyclin genes , representing an unusual repertoire of cyclin classes . Expression and reverse genetic analyses of the single Plant ( P ) -type cyclin , CYC3 , in the rodent malaria parasite , Plasmodium berghei , revealed a cytoplasmic and nuclear location of the GFP-tagged protein throughout the lifecycle . Deletion of cyc3 resulted in defects in size , number and growth of oocysts , with abnormalities in budding and sporozoite formation . Furthermore , global transcript analysis of the cyc3-deleted and wild type parasites at gametocyte and ookinete stages identified differentially expressed genes required for signalling , invasion and oocyst development . Collectively these data suggest that cyc3 modulates oocyst endomitotic development in Plasmodium berghei .
The mechanisms of mitotic cell division and the various molecules involved are well studied in many model systems including yeast , plants and human cells . Progression through mitosis is controlled by a range of factors , including cyclins , protein kinases ( PKs ) and phosphatases ( PPs ) , and the anaphase-promoting complex ( APC ) components [1–4] . Cyclins play active roles at distinct stages of the cell cycle [4] via regulation of cyclin-dependent kinases ( CDKs ) . Cyclins possess a conserved ~100-residue sequence known as the cyclin box that mediates CDK binding and activation [5] . Certain cyclins are capable of binding several CDKs , which are themselves able to associate with multiple cyclins [6 , 7] . These distinct but overlapping functions orchestrate cell cycle progression . In most model systems , the synthesis and level of cyclins are tightly regulated during the cell cycle , with each cyclin being degraded via ubiquitination once its function is complete [8 , 9] . Key cell cycle transitions are regulated by specific cyclins: G1/S cyclins , which are essential for cell cycle entry at G1/S ( start ) , and G2/M cyclins , which are essential at the G2/M ( mitosis ) transition . In some species , there are multiple forms of G1 and G2 cyclins . For example , in vertebrates there are at least three G1 cyclins ( C , D , and E ) and two G2 cyclins ( A , which is also active in S phase , and B ) . Many other cyclins with additional functions have also been described [10 , 11] . Recent bioinformatic analyses have identified 3 distinct classes of cyclin: Group I ( including cyclin A , B , D , E , F , G , I , J , and O families ) , Group II ( P/Pho80-like , and Y ) , and Group III ( C , H , K , L , and T ) [12 , 13] . Group I includes known essential cell cycle cyclins including some involved in mitosis and meiosis [10 , 14 , 15] , while Group III contains cyclins associated with transcription and RNA processing [11] . Group II contains both cell cycle cyclins and those of other function , such as cell metabolism ( e . g . P-type cyclins ( CYCPs ) , which were originally believed to be unique to plants ) and appear to link cell cycle regulation to nutritional status [16–18] . The cyclin box of CYCP in plants shows high similarity to the corresponding domain of trypanosome cyclins CYC2 , CYC4 , CYC7 , CYC10 and CYC11 [13 , 19] and S . cerevisiae PHO80-like cyclins [20] . Malaria parasites , Plasmodium , which belong to the phylum Apicomplexa , divide and proliferate in an unusual way compared to other eukaryotes . During its life cycle , Plasmodium exhibits two types of cell division; one in asexual stages that resembles endomitosis and one in sexual stages . The endomitotic-like asexual stage is characterized by multiple nuclear divisions preceding cytokinesis , with maintenance of the nuclear membrane wherein the microtubule organising centre ( MTOC ) or spindle body is embedded [21–23] . This process is observed at three developmental stages of the parasite life cycle: during schizogony , replication and multiplication within liver and red blood cells in the vertebrate host [21] , and at the sporogonic stage of parasite development in the mosquito vector [24] . The other type of cell division occurs during male gametogenesis , where three rounds of rapid DNA replication are followed by cell division and chromosome condensation giving rise to eight microgametes [23 , 25 , 26] . In the human malaria parasite , Plasmodium falciparum , the stages of development during asexual multiplication inside red blood cells have been described as ( a ) ring stage ( early trophozoite ) with a single ( haploid ) interphase nucleus in G0 , ( b ) mature trophozoites ready for chromosome replication ( G1 ) and undergoing DNA synthesis ( S phase ) , and ( c ) the schizont stage when asynchronous nuclear division begins ( M phase ) and repeated S and M phases continue resulting in a multinucleate syncytium [21 , 27] . At the end of the growth phase , cell division occurs in the late stage schizont or segmenter to form merozoites that are released to invade new red blood cells . This process resembles that of endomitosis observed in Drosophila cells [28] . The molecular mechanisms that regulate nuclear and cell division in the malaria parasite remain largely unknown . We have previously shown that the single putative homologue of cell division cycle protein 20 ( CDC20 ) , a well-characterised activator of the anaphase promoting complex/cyclosome ( APC/C ) , is crucial for karyokinesis and cytokinetic control of male gametogenesis [29] . Furthermore , systematic genome-wide functional analysis of the protein kinome and phosphatome has identified molecules crucial for both male gametogenesis and asexual multiplication during sporogony , including CDKs , CDPK4 , MAP2 and PTPLA [30–32] . Plasmodium has a number of genes that code for calcium dependent protein kinases ( CDPKs ) that are implicated in control of growth and cell division [31 , 32] . Other kinases , such as ARK1 , an aurora-like kinase associated with the spindle pole body , have been identified and implicated in cell division [33] . Knowledge of cyclin function in Plasmodium is limited . Previously , four cyclin genes were described in P . falciparum , Pfcyc1 to -4 [34 , 35] . Biochemical studies showed that PfCYC1 , PfCYC3 and PfCYC4 associate with histone H1 kinase activity present in the parasite extract [35] , and PfCYC1 and PfCYC3 bind and activate the Plasmodium CDK1 homologue ( PK5 ) in vitro [34 , 35] . PfPK5 has also been shown to be activated by mammalian proteins ( p25 and RINGO ) that have no detectable primary sequence similarity to cyclin [34 , 35] . Recently , PbCYC3 has been shown to be a target gene of the AP2-O transcription factor , and is involved in oocyst development [36] . Here , we describe an in-depth phylogenetic analysis of the cyclin repertoire in Plasmodium . We then focus on the single P-type cyclin ( CYC3 ) and examine its role during parasite cell division , using the rodent malaria model Plasmodium berghei . We show that whilst CYC3 is dispensable for asexual cell cycle progression in the mammalian host , it modulates oocyst development and the subsequent differentiation of sporozoites , the endomitotic process of sporogony within the mosquito vector .
Cyclins are a diverse superfamily of proteins ( for example see [13] ) . A number of PFam models incorporate parts of the conserved domain ( including PF08613 , PF00134 , PF02984 ) , but none covers the entire conserved region and each is biased towards particular classes . To investigate the Plasmodium cyclin repertoire , we built a pan-cyclin hidden Markov model ( HMM ) of cyclins from a range of model species and used this to identify putative cyclins in a wide range of eukaryotes ( see Methods ) . Our HMM showed good sensitivity , identifying full cyclin repertoires in diverse organisms not included in the seed alignment , and the resultant proteins sequences were aligned , trimmed to conserved regions and classified by constructing phylogenetic trees ( Fig 1A and S1 Fig ) . Plasmodium species encode only three identifiable cyclins–CYC1 , CYC3 and CYC4 . A P . falciparum gene ( encoded by PF3D7_1227500 ) annotated as PfCYC2 did not show a significant match to any of the cyclin-specific HMMs built during these analyses ( even at extremely liberal thresholds ) , nor to the domains built by PFam . This protein was originally identified as a cyclin on the basis of a very limited similarity to a cyclin A from the sea star Patiria ( Asterina ) pectinifera ( 13% identity across alignable length ) [35] , but it lacks key alignable residues across most of the cyclin box and has no detectable cyclin-like function in biochemical assays [35] . These data strongly suggest that CYC2 is not a true cyclin and it is not included in the repertoire described here . The Plasmodium cyclin repertoire is highly unusual in that it entirely lacks Group I , the largest group of cyclins . This group contains most canonical cyclin families that regulate specific cell cycle transitions with their CDK partners: Cyclin D-CDK4/6 for G1 progression , Cyclin E-CDK2 for the G1/S transition , Cyclin A-CDK2 for S phase progression and CyclinA/B-CDK1 for mitosis , although in fission yeast , all cell cycle transitions are driven by a single Cyclin B/CDK complex ( CDC13/CDC2 ) [3 , 37] . In keeping with this key role and previous analyses [13] , Group I cyclins were found in all non-apicomplexan species examined here , including the alveolate Tetrahymena thermophila [13] . However , none of the apicomplexan species examined contained Group I cyclins , except Cryptosporidium , which was found to encode three Group I cyclins of indeterminate family , suggesting that there has been a loss of Group I cyclins during the evolution of apicomplexan lineages ( Fig 1B ) . Plasmodium species encode only one cyclin from the P family ( Group II ) and two Group III cyclins from families H and L . The cyclin P family is not found in animals , but includes many plant cyclins and Pho80 in budding yeast , which link nutritional sensing to cell cycle progression . In contrast , both H and L families are associated with transcription: the CDK7/Cyclin H/MAT1 complex functions as a Cdk-activating kinase in cell cycle regulation [38] and as a modulator of the general transcription factor TFIIH [39 , 40] . Similarly , CDK11-Cyclin L complex in fission yeast regulates the formation of the Mediator complex , a coactivator of RNA polymerase II transcription [41] . We used quantitative RT-PCR ( qRT-PCR ) to investigate the RNA levels of cyc1 , 3 and 4 in six stages of the wild-type parasite life cycle . The transcription profiles showed expression of the cyclins throughout parasite development with the highest RNA levels for all three cyclins found in gametocytes ( both non-activated and activated ) and schizonts ( particularly cyc4 ) . The cyc3 RNA level was highest in non-activated gametocytes , whereas for cyc1 and cyc4 , the levels were highest in activated gametocytes ( Fig 2A ) . These results are similar to those obtained previously for P . berghei by RNA-seq analysis [42] and for P . falciparum using both RT-PCR [35] and RNA-seq [43] . To investigate the localisation of CYC3 , the only P-type cyclin in Plasmodium , throughout the P . berghei life cycle , we generated a C-terminal GFP fusion protein using single crossover recombination at the endogenous cyc3 gene locus ( PBANKA_123320; S2A Fig ) . Correct integration was confirmed by PCR , pulsed-field gel electrophoresis ( PFGE ) and Southern blot ( S2B–S2D Fig ) . CYC3-GFP parasites were able to complete the full life cycle with no detectable phenotype observed from tagging with GFP including oocyst development at 14 days post infection ( dpi ) in mosquitoes ( S2F and S2G Fig ) . Western blot with an anti-GFP antibody confirmed the expression of CYC3-GFP . A 54 kDa protein was detected in lysates from three stages of parasite development ( schizonts , activated gametocytes and ookinetes ) with the highest expression observed in activated gametocyte and ookinete stages compared with the 29 kDa GFP control extracted from a parasite line constitutively expressing GFP ( GFPcon 507 cl1 ) [44] ( S2E Fig ) . Live imaging of parasites revealed CYC3-GFP presence throughout the parasite cell body with a predominantly cytosolic localisation at most of the key Plasmodium life cycle stages examined ( trophozoite , male and female gametocyte , zygote , ookinete , oocyst and salivary gland sporozoite ) ( Fig 2B ) . However , we could not exclude the presence of CYC3-GFP also in the nucleus of trophozoites , gametocytes and ookinetes ( Fig 2B ) . Therefore , for a more detailed analysis of the CYC3 localisation at trophozoite , schizont , gametocyte and ookinete stages , we used deconvolution fluorescence imaging ( Fig 2C ) . Although no expression was detected in schizonts , two dimensional optical slices from 3D stacks showed that CYC3-GFP was uniformly present throughout both the cytoplasm and the nucleus in trophozoites , gametocytes and ookinetes ( Fig 2C ) and noticeably enriched in the nucleus of ookinetes ( Fig 2C , line profile ) . To assess the function of CYC3 in the Plasmodium life cycle , we used a double crossover homologous recombination strategy to delete the cyc3 gene ( S3A Fig ) . Successful integration of the targeting construct at the cyc3 locus was confirmed by diagnostic PCR across the junction of the expected integration site , as well as by Southern blot and PFGE ( S3B–S3D Fig ) . Analysis of two independent cyc3 deletion clones , cyc3 cl1 and cyc3 cl3 ( hence forward called Δcyc3 ) showed no overt phenotype during blood stage asexual proliferation , microgamete exflagellation or ookinete conversion in vitro when compared with control parasites which constitutively express untagged GFP ( WTGFPcon 507 cl1 line , henceforth known as WT ) [44] ( Fig 3A–3C ) . To determine whether CYC3 is essential for parasite development in the mosquito vector , we fed female Anopheles stephensi mosquitoes with the Δcyc3 mutant or WT parasites . There was no significant reduction in the number of Δcyc3 oocysts compared to WT controls at 5 dpi , however a significant reduction was observed at 7 dpi , which became even more evident at 10 , 14 and 21 dpi ( Fig 3D and 3E ) . Furthermore , oocysts at 14 and 21 dpi appeared substantially smaller in Δcyc3 compared with WT lines ( Fig 3F ) . To quantify this reduction in oocyst size , the diameter of WT and Δcyc3 oocysts was measured at multiple time points during development in three independent mosquito infections ( Fig 3G ) . Even at early stages during oocyst development ( 5 and 7 dpi ) , we already detected a difference in mean oocyst diameter between Δcyc3 and WT parasites ( 8 μm and 11 μm for 5 dpi; 10 μm and 14 μm for 7 dpi , respectively ) and by 10 dpi , the majority of Δcyc3 oocysts were substantially smaller than WT ( 12 μm and 25 μm mean diameter , respectively ) ( Fig 3F and 3G ) . This difference in oocyst size increased dramatically at 14 and 21 dpi as WT oocysts reached maturity ( mean diameter of 31 μm at 14 dpi compared to 15 μm for Δcyc3 oocysts , Fig 3G ) . The majority of Δcyc3 oocysts were smaller in size even at 5 dpi . After this , although a small number of the Δcyc3 oocysts continued to develop normally , the majority appeared to arrest and remained the same size until 21 dpi ( Fig 3G ) . In addition to the decrease in size , the number of Δcyc3 sporozoites in infected mosquitoes was significantly reduced compared to WT in midguts at 14 and 21 dpi and in salivary glands at 21 dpi ( Fig 3H , S4A and S4B Fig ) . However , when infected mosquitoes were allowed to feed on mice in bite-back experiments , we found that not only did transmission occur successfully but the pre-patent period for Δcyc3 was the same as for WT parasites ( S4A and S4B Fig ) . These data show that sporozoites produced from the few normal oocysts in the Δcyc3 mutant are not affected in their efficacy of host infectivity . Next we investigated whether the oocyst growth phenotype was related to a defect in ookinete structure , motility or DNA content . Ultrastructural analysis of ookinetes by transmission electron-microscopy revealed no morphological differences between Δcyc3 and WT ( S4C Fig ) . Similarly , both DNA content and the gliding motility of Δcyc3 ookinetes were similar to WT ( S4D and S4E Fig ) suggesting that ookinetes are not affected by cyc3 deletion . Finally , we wanted to investigate whether Plasmodium CYC3 might function as a G1 or G2/M cyclin , which would explain the Δcyc3 phenotype during sporogony . To this end , we examined the ability of a codon-optimized version of the P . berghei cyc3 gene to complement a triple cln ( G1 cyclin ) mutant of the budding yeast Saccharomyces cerevisiae or a temperature sensitive growth phenotype of a cdc13-117 B-type cyclin mutant of the fission yeast Schizosaccharomyces pombe ( G2/M transition ) , respectively ( S5A and S5B Fig ) . In both cases , P . berghei cyc3 failed to rescue yeast cyclin mutant strains , suggesting that P . berghei CYC3 does not function as a classical G1 cyclin or G2/M cyclin under these conditions . Since CYC3-GFP was expressed in gametocytes ( Fig 2B ) , we examined whether the defect in oocyst formation was sex-specific . To do this we performed genetic crosses between Δcyc3 parasites and lines deficient in either male ( Δp48/45 and Δhap2 ) or female ( Δdozi and Δnek4 ) gametes [45–48] . As scored by an increase in normal size oocysts at 14 dpi , we found that all mutants could only partially rescue the phenotype of the cyc3 knockout , which affected both male and female lines equally ( Fig 3I ) . These results reveal that the functional cyc3 is inherited through both male and female lines and its function is independent of sexual commitment at the gametocyte stage ( Fig 3I ) . As the Δcyc3 phenotype is observed during early oocyst development , we next examined CYC3-GFP expression during oocyst and sporozoite development at the same time points ( 5 , 7 , 10 , 14 and 21 dpi ) as described for the Δcyc3 phenotypic analysis within mosquitoes using fluorescence microscopy ( Fig 4 ) . Fluorescent imaging showed no detectable CYC3-GFP expression in oocysts at 5 and 7 dpi ( representative images in Fig 4 ) . Expression of CYC3-GFP was first observed at low levels in the majority of oocysts at 10 dpi with the highest expression detected at 14 dpi . Oocysts that had formed fully mature sporozoites showed the highest protein expression . After 14 dpi , we observed a decrease in CYC3-GFP expression in oocysts up to day 21pi ( Fig 4 ) . To define further the defect in oocyst growth during different developmental stages of sporogony , Δcyc3 and WT parasite-infected mosquito midguts at 7 , 10 , 14 , and 21 dpi were examined by transmission electron microscopy . Marked differences were observed at the later time points in the ultrastructure of the majority of Δcyc3 compared to WT oocysts , although some Δcyc3 oocysts appeared similar to WT ( Fig 5 ) . At every time point there were significantly fewer oocysts in the guts of mosquitoes infected with Δcyc3 compared to the WT parasites ( 5 oocysts/gut Δcyc3 compared to 60 oocyst/gut WT at 7 dpi ) . At 7 and 10 dpi the oocysts of both the WT and Δcyc3 had similar structural appearance being spherical with the cytoplasm completely filling the cyst ( Fig 5Ai , 5Aiv and 5Avii ) . The cytoplasm contains numerous nuclear profiles and homogenous appearing cytoplasm containing mitochondria and apicoplasts . However , a proportion ( approximately 10% ) of the Δcyc3 oocysts also showed centrally located nuclei and exhibited some vacuolation of the cytoplasm ( Fig 5Avii ) . By 14 dpi , the majority ( >98% ) of WT oocysts exhibited various stages of sporozoite formation ( Fig 5Aii , 5Aiii and 5Bi–5Biii ) or oocysts with mature sporozoites ( Fig 5Aiii ) . In contrast , there were many fewer Δcyc3 oocysts ( 10 Δcyc3 compared to 60 WT oocysts ) and only a proportion ( <40% ) of these showed similar features of sporulation ( Fig 5Av and 5Avi ) . Many ( >60% ) of the Δcyc3 oocysts showed no evidence of sporulation ( Fig 5Aviii and 5Aix ) . These oocysts showed no retraction of the plasmalemma to form the sporoblasts and there was no evidence of the initiation of sporozoite formation ( Fig 5Aviii and 5Aix ) . However , cytoplasmic changes were observed including abnormal membrane reduplication ( Fig 5Biv ) and nuclei containing large numbers of microtubules , suggesting mitosis and cytokinesis were affected ( Fig 5Bv ) . There was evidence of nuclei with apoptotic-like chromatin changes and dilated endoplasmic reticulum ( Fig 5Bvi ) . There appeared to be continued cytoplasmic vacuolation consistent with progressive cell death ( Fig 5Aviii , 5Aix and 5Bvi ) . These results are consistent with the idea that DNA synthesis , endomitosis and cytokinesis are severely defective in most but not all oocysts . Moreover , structural abnormalities in some Δcyc3 oocysts suggest that they are incapable of forming viable sporozoites . The marked changes in Δcyc3 oocyst morphology and growth led us to analyse the regulation of mRNA in Δcyc3 parasites compared with WT . We first used strand-specific RNA sequencing ( RNA-seq ) to investigate the global transcript levels in Δcyc3 and WT activated gametocytes and ookinetes . The deletion of cyc3 was confirmed by RNAseq , with no reads mapping to the region of the gene targeted for disruption ( S6A Fig ) . Generally , most transcript levels were very similar in activated gametocytes , with strong linkage between levels in Δcyc3 and WT lines . However , the ookinete transcriptome was greatly altered by loss of cyc3 , with many genes down-regulated and a smaller number up-regulated . In total , 813 and 2 , 069 genes showed modulated expression in activated gametocytes and ookinetes , respectively ( p <0 . 05 and fold change >2; 702 and 1 , 891 genes in activated gametocytes and ookinetes , respectively , with p <0 . 01 and fold change >2 ) ( Fig 6A , S6B Fig and S2 Table ) , including those with roles in reversible phosphorylation , transcription , cell signalling and inner membrane complex function ( Fig 6B , S6C Fig and S3 Table ) . We identified several functional clusters that were significantly differentially expressed in Δcyc3 ( Fig 6B , S6C Fig and S3 Table ) which may be collectively responsible for the observed phenotype . Global Gene Ontology ( GO ) analysis showed enrichment of genes associated with GO terms ‘kinase activity’ , ‘protein phosphorylation’ and’ inner membrane complex’ ( S6D Fig ) . To validate our RNA-seq data , we performed qRT-PCR analysis of specific sets of genes known to have roles in cell cycle , signalling and transcription , or invasion and oocyst development ( Fig 6C and S2 Table ) . These data showed very good correlation with the RNA-seq results ( Fig 6D ) validating the results for activated gametocytes and ookinetes . In addition , qRT-PCR data were collected for schizonts and/or day 14 oocysts ( Fig 6C ) ; these data showed that most of the changes in transcript levels for these sets occur at the ookinete stage , with some additional effects in oocysts . Importantly , both RNA-seq and qRT-PCR analyses showed that neither of the other two Plasmodium cyclin genes were differentially expressed in response to deletion of cyc3 , suggesting that there is no compensation provided by up-regulation of other cyclins . In contrast , several of the cdks were differentially expressed at different stages ( Fig 6C ) : of particular interest was crk1 and crk5 , which were significantly up-regulated in ookinetes and are known to have roles in gene transcription [49] and cell cycle [50] , respectively . Recently , it has been shown that a family of transcription factors ( apiAP2 ) is responsible for gametocyte commitment , as well as ookinete and sporozoite development [36 , 51–55] . This family is also a feature of chromerid genomes and their evolutionary history suggests they have mediated evolutionary changes during lineage differentiation [56] . Our analysis showed that several apiAP2 family members were affected by cyc3 deletion , notably a significant down-regulation of ap2-sp and ap2-l in Δcyc3 ookinetes and oocysts , consistent with a reduction in normal oocyst formation ( Fig 6C , S6C Fig and S2 Table ) . Due to the important role of reversible phosphorylation in cell cycle progression in other systems [3] , we next looked at the expression of protein kinases , phosphatases and cell cycle proteins involved in parasite development , especially those that either display a similar phenotype to that of Δcyc3 when deleted ( pk7 , ppm5 ) or are involved in sexual development ( nek4 ) . The expression of genes involved in male sexual development ( cdpk4 and cdc20 ) [29 , 57 , 58] was not significantly affected . Furthermore , the transcription of ppm7 , ppm5 and pk7 , which have similar protein localisation and/or similar phenotypes in gene knock-out lines [30 , 31] , were also unaffected ( Fig 6C ) . However , two genes important for zygote differentiation ( nek2 and nek4 ) [47 , 59] were significantly mis-regulated in activated gametocytes and ookinetes . Based on the phenotype observed in Δcyc3 parasites and the fact that several known invasion genes were significantly affected , as detected in the RNA-seq analysis , we next focused on genes involved in ookinete invasion and oocyst development . Interestingly , there was a marked down-regulation of genes required for ookinete motility and invasion ( Fig 6C ) , such as mtip , ctrp and soap [60 , 61] , despite the fact that ookinete development and motility appeared to be unaffected ( S4 Fig ) . Genes known to be required during oocyst development , such as cap380 , coding for a capsule protein necessary for healthy oocyst formation and subsequent transmission [62] , and csp , important for infectivity of the host [63] , were mis-regulated in multiple stages ( Fig 6C ) . These data suggest that the oocyst development phenotype in Δcyc3 may be a downstream consequence of the mis-regulation of these genes , well before sporogony commences . Overall , deletion of cyc3 caused the modulated expression of a number of genes related to the observed phenotype , especially in ookinetes , including transcription factors , putative cyclin binding partners , and genes for signalling and subsequent oocyst development .
The molecular mechanisms controlling the cell cycle and mitosis are regulated by key molecules including CDKs and cyclins . Cyclins are key regulators of CDKs and are synthesised and degraded at specific phases of the cell cycle [5 , 64] . Although cdk/cyclins have been well studied in many eukaryotic model systems , their role in the unicellular parasite Plasmodium is not well understood . Here , we show that despite its complex and atypical pattern of cell cycle and division , Plasmodium only has three cyclin genes , a very small complement compared to other organisms including human ( 30 annotated genes in the Refseq dataset ) , yeast ( 22 genes ) and Drosophila ( 12 genes ) . Plasmodium has no Group I cyclins but has homologues belonging to Group II ( CYC3 ) and Group III ( CYC1 and CYC4 ) . CYC3 is a P-type cyclin , a family which contains many cyclins of plants ( CYCPs ) [17] . It also shows similarity to PHO80 cyclin from Saccharomyces cerevisiae [20] and a G1 cyclin present in kinetoplastids [65 , 66] . Both Pho80 and plant CYCPs appear to link cell cycle regulation to nutritional status: in yeast , Pho80 and other P-like cyclins associate with CDK Pho85 to phosphorylate a variety of substrates involved in phosphate , glycogen and carbon source metabolism ( reviewed in [16] ) , whereas CYCP2;1 in Arabidopsis is required for G2/M transition in meristem activation in a nutrition-responsive manner [18] . Trypanosomes encode at least three P-type cyclins , CYC2 , 4 , and 7 , with CYC10 and 11 being likely divergent members of the family ( “Pho80-like” cyclin TbCYC5 [66] is a cyclin Y , see S1 Fig ) . In contrast to CYCP , Trypanosoma brucei CYC2 is required for progression from G1 phase in the insect stage of the parasite [67] , which would put its cell cycle activity in a position more akin to the Pho80-like cyclins ( Pcls ) in yeast [16] . In keeping with this , TbCYC2 is able to rescue a yeast CLN mutant [19] , in which 3 Group I G1 cyclins , CLN1 , 2 and 3 , were inactivated [68] . There are only three clearly identifiable CDKs ( PfPK5 , Pfcrk1 and Pfcrk3 ) in the Plasmodium genome [31 , 69] . The lack of cyclin D in Plasmodium here mirrors the absence of CDK4/6 orthologues previously reported [49] , and Pfcrk1 is a CDK10/11-related protein [69] , which matches well with the presence of cyclin L family members identified here . Moreover , a fourth protein , Pfmrk , was initially described as a CDK7 homologue and was shown to be stimulated by binding to human cyclin H in vitro [70] , but subsequent work showed that its affinities are less clear [69] and functional work suggests it might be a RNA polymerase II carboxyl-terminal domain kinase rather than a CDK [71] . There is no homologue of Pho85 identified in Plasmodium and the essential kinase PfPK5 is a relative of CDK1/2 , which would normally be activated by a Group I cyclin in animals or fungi . Transcription of P . berghei cyclins varies during different phases of the life cycle with the highest transcription seen in gametocytes and schizonts [35] . Our localisation studies of CYC3-GFP by fluorescence and deconvolution microscopy showed both cytoplasmic and nuclear localisation , which is expected given the localisation and role of most cyclins in model systems [10] . The subcellular location of P-type cyclins ( for example , CYCP2;1 ) in plants is mainly nuclear [18] , however , TbCYC2 is also expressed in the cytosol of procyclic forms and controls posterior morphogenesis of the parasite during the G1/S and G2/M transition [19 , 65 , 66 , 72] . A recent study on the AP2-O transcription factor in Plasmodium showed that cyc3 is a target of AP2-O and used deletion of cyc3 as a validation tool to show that it had a role in ookinete to oocyst development [36] . Our gene deletion mutant is consistent with this analysis in terms of number and size of oocysts , but further dissects in depth the function of cyc3 throughout the lifecycle using transcriptome analysis and high resolution microscopy . Maturation of most of the ∆cyc3 oocysts , based on their abnormal shape and size , was impaired in infected mosquito guts , and detailed studies revealed that this defect in oocyst growth and differentiation began during early sporogony . Plasmodium oocysts contain a highly lobed syncytial nucleus that divides at the time of sporozoite budding into a number of lobes , which undergo subsequent mitotic division resembling endomitosis [23 , 25] . The ∆cyc3 phenotype suggests that although initial oocyst formation occurs , in many of the oocysts the division of nuclear lobes with oocyst growth is drastically affected and further budding of sporozoites from these lobes does not occur . This defect leads to abnormalities in membrane formation , vacuolation and subsequent cell death during the later stages of sporogony . The characteristics of ∆cyc3 oocyst development suggest that maturation and differentiation are arrested , with cells unable to progress further to form additional lobes and start sporozoite budding and mitotic division . A number of gene deletion mutants have been described in Plasmodium that affect oocyst maturation for example; PbCAP380 , affecting oocyst capsule formation [62]; PbGEX and PbDMC1 , affecting oocyst size and sporozoite formation [73 , 74]; PbLAPs and PbCDLK , affecting sporogony but not the size of the oocyst [31 , 75 , 76]; PbMISFIT , causing small sized oocysts [77]; and G actin sequestering protein ( C-CAP ) affecting oocyst development and showing similar features to Δcyc3 during early oocyst development [78] . However , the phenotype of Δcyc3 lines differs from that of all these mutants due to the fact that some normal oocysts are produced , which are able to form invasive salivary gland sporozoites and initiate liver stage infection . Moreover , none of the other mutated genes was implicated in cell cycle control or cell division . The presence of a small number of normal oocysts/sporozoites in the Δcyc3 mutant does suggest that CYC3 has a subtle role in oocyst/sporozoite development . The reduction in sporozoite number ( in addition to the reduction in oocyst number ) is drastic at 14 and 21 dpi compared to the WT parasites and there is clear expression of CYC3-GFP in 14 dpi oocysts , both indicative of a role in sporozoite development . This might suggest that there are unknown regulatory mechanisms involved in the control of the oocyst development that can by-pass a default CYC3-dependent pathway to some degree in these parasites during early oocyst development and sporogony . Specific protein kinase and phosphatase mutants ( PbPK7 and PbPPM5 ) show some resemblance to the Δcyc3 phenotype , with a reduced number of oocysts , which were abnormally small and did not complete sporogony before arresting [30 , 31] . However , only minor changes in the transcript level of protein kinase ( pk7 ) or phosphatase ( ppm5 ) were detected . Conversely , cyc3 transcripts are down-regulated in Δppm5 parasites [30] suggesting that PPM5 acts upstream of CYC3 . Gene deletion revealed that CYC3 is dispensable during asexual multiplication in erythrocytes and in the liver ( erythrocytic and exo-erythrocytic schizogony ) and sexual development during male gametogenesis . However , CYC3 clearly modulates sporogony via endomitotic multiplication during oocyst development in the mosquito . Oocysts are the only replicative extracellular stage during the parasite life cycle , and therefore it is possible that CYC3 modulates this extracellular replication in response to metabolic sensing within the mosquito . For example , Arabidopsis CYCP2;1 is transcribed in response to sugar signals by a specific transcriptional factor , allowing cell cycle progression [18] . CYC3 may have a direct or indirect role in regulation of transcription in the ookinete , of specific genes involved in oocyst growth ( defined as a G1 phase ) and sporogony ( defined as S/mitotic phases ) , and that could explain why we see expression of CYC3-GFP in ookinetes and 10 to 14 dpi oocysts . Rescue experiments in yeast suggest that Plasmodium CYC3 does not behave as a classical G1 or G2/M cyclin , or at least cannot substitute for the function of yeast cyclins . Nevertheless the peak in expression of CYC3-GFP around 10 to 14 dpi of oocyst development , just before sporogony , and the defect in sporozoite production in Δcyc3 mutants , are both consistent with a cyclin-B-like role in cell cycle progression and a role in G1/S progression . Global transcript analysis of the Δcyc3 mutant suggested that neither cyc1 nor cyc4 compensates for the cyc3 deletion . Analysis of all the CDKs showed that several , including crk1 , with a role in transcription [49] , were mis-regulated at several stages in Δcyc3 parasites; however it is unlikely that transcription of a putative CDK partner would be controlled by CYC3 . It has been reported previously that PfCYC3 binds and activates the CDK1 homologue , PfPK5 , in vitro [35] , and a similar result was reported for the homologues from Eimeria tenella [79] , although we saw no significant mis-regulation of pk5 in Δcyc3 parasites and no study has shown a Plasmodium cyclin-CDK interaction in vivo . Other cell cycle genes were affected including predicted members of the mini chromosome maintenance ( MCM ) family . While no MCM has been functionally characterised in Plasmodium , the MCM family has been bioinformatically well classified in Apicomplexa [80 , 81] and MCMs are known to be important for G1/S phase progression and initiation of DNA replication [82] . Misregulation of genes involved in ookinete invasion and structure ( such as mtip , ctrp and soap ) had no observable effect on ookinete motility or oocyst formation at 5 dpi and we observed no detectable phenotype with electron microscopy , however , a subtle delay in oocyst initiation ( due to ookinete invasion or motility ) may be enough to initiate deleterious consequences downstream during oocyst development . In addition to this , we cannot rule out the possibility that CYC3 is indirectly involved in nutrient/environmental sensing in the mosquito gut , a known function of cyclins in other systems [18] . The mis-regulation of these , as yet undefined , genes may be responsible for the mixed population of small , defective oocysts versus normal , healthy oocysts . Thus , the absence of cyc3 may make these parasites more sensitive to environmental stimuli such as nutrient abundance , or the impact of developmental mis-timing ( via mis-regulation of invasion genes ) . Oocyst growth may further deteriorate following the mis-regulation of genes required for oocyst development ( such as cap380 ) . Transcription analysis also showed that several members of the apiAP2 transcription factor family , which have been shown to regulate various stages of development , were affected in ∆cyc3 oocysts . This is perhaps unsurprising considering that normal regulation of these genes is required for successful oocyst development [52 , 53 , 55] . Global transcriptomics is a useful tool for the identification of possible dysregulation and compensatory mechanisms in Δcyc3 parasites however the measurement of mRNA levels is not necessarily indicative of the corresponding protein levels . Future proteomic work may provide more information on protein-protein interactions betweenCYC3 and putative CDK partners or on the effects of cyc3 deletion on global ( or cell cycle specific ) protein levels in the Δcyc3 mutant . In conclusion , this is the first study to classify phylogenetically the cyclins in Plasmodium and uncover , in depth , an important function for CYC3 , a novel P-type cyclin . We describe a key role for this cyclin during the early stages of ookinete to oocyst development , specifically the G1/S phase , which subsequently affects differentiation and sporogony , suggesting it is a modulator of transcription and oocyst endomitotic development in Plasmodium .
All animal work at Nottingham has passed an ethical review process and was approved by the United Kingdom Home Office . Work was carried out in accordance with the United Kingdom ‘Animals ( Scientific Procedures ) Act 1986’ and in compliance with ‘European Directive 86/609/EEC’ for the protection of animals used for experimental purposes under UK Home Office Project Licenses 40/3344 and 30/3248 . Sodium pentabarbitol was used for terminal anaesthesia and a combination of ketamine followed by antisedan was used for general anaesthesia . All efforts were made to minimise animal suffering . Six-to-eight week old female Tuck-Ordinary ( TO ) ( Harlan ) outbred mice were used for all experiments . To identify cyclin-like proteins , a pan-cyclin hidden Markov model was used to perform a similarity search using HMMER3 [83] . Briefly , all annotated cyclins were taken from the predicted proteomes of human , Arabidopsis thaliana , Caenorhabditis elegans , Saccharomyces cerevisiae , Schizosaccharomyces pombe , and Toxoplasma gondii . These cyclins were aligned using MAFFTv6 . 925b [84] with the L-INS-i strategy [85] , trimmed to conserved regions and used to create a HMM , which was used to search the predicted proteomes of 12 apicomplexan parasites ( EuPathDB; http://eupathdb . org; see Fig 1C ) as well as 20 other eukaryotes from diverse groups . The conserved cyclin domains were excised from these sequences based on HMM hits with e-values ≤ 10−18 , realigned and used to create a refined HMM , which was then used to re-search the apicomplexan proteomes at a threshold of e-value ≤ 10−12 . Alignments of cyclins were trimmed to conserved regions and used to infer maximum likelihood phylogenies as implemented by the program PhyML3 . 0 [86] using the LG substitution matrix with a gamma-distributed variation in substitution rate approximated to 4 discrete categories ( shape parameter estimated from the data ) . Trees shown are a majority-rule consensus of 100 bootstrap replicates for each alignment . For GFP-tagging by single homologous recombination [30] , a 986 bp region of cyc3 starting 332 bp upstream of the ATG start codon and omitting the stop codon was amplified using primers T0891 and T0892 . This was inserted upstream of the gfp sequence in the p277 vector using KpnI and ApaI restriction sites . The p277 vector contains the human dhfr cassette , conveying resistance to pyrimethamine . Before transfection , the sequence was linearised using HindIII , The gene targeting vector for ∆cyc3 lines was constructed using the pBS-DHFR plasmid , which contains polylinker sites flanking a T . gondii dhfr/ts expression cassette conveying resistance to pyrimethamine , as described previously [31] . PCR primers N0451 and N0452 were used to generate a 411 bp fragment of 5′ upstream sequence of cyc3 from genomic DNA , which was inserted into ApaI and HindIII restriction sites upstream of the dhfr/ts cassette of pBS-DHFR . A 663 bp fragment generated with primers N0453 and N0454 from the 3′ flanking region of cyc3 was then inserted downstream of the dhfr/ts cassette using EcoRI and XbaI restriction sites . The linear targeting sequence was released using ApaI/XbaI . All of the oligonucleotides used to make these constructs can be found in S1 Table . P . berghei ANKA line 2 . 34 ( for GFP-tagging ) or ANKA line 507cl1 ( for gene deletion [44] ) were then transfected by electroporation [44] . Briefly , electroporated parasites were mixed immediately with 100 μl of reticulocyte-rich blood from a phenylhydrazine ( 6mg/ml , Sigma ) treated , naïve mouse , incubated at 37°C for 20 min and then injected intraperitoneally . From day 1 post infection pyrimethamine ( 70 μg/ml , Sigma ) was supplied in the drinking water for four days . Mice were monitored for 15 days and drug selection was repeated after passage to a second mouse . Resistant parasites were then used for cloning by limiting dilution and subsequent genotyping . For the C-terminal fusion GFP-tagged parasites , a diagnostic PCR reaction was used as illustrated in S2 Fig . Primer 1 ( INT T89 ) and Primer 2 ( ol492 ) were used to determine correct integration of the gfp sequence at the targeted locus ( S1 Table ) . After confirmation of correct integration , genomic DNA from wild type and transgenic parasites was digested with BsmI and the fragments were separated on a 0 . 8% agarose gel , blotted onto a nylon membrane , and probed with a PCR fragment homologous to the P . berghei genomic cyc3 sequence cloned in the p277 vector , using the Amersham ECL Direct Nucleic Acid Labelling and Detection kit ( GE Healthcare ) . Parasites were also visualised on a Zeiss AxioImager M2 ( Carl Zeiss , Inc ) microscope fitted with an AxioCam ICc1 digital camera ( Carl Zeiss , Inc ) and analysed by Western blot to confirm GFP expression and the correct protein size . For the gene knockout parasites , two diagnostic PCR reactions were used as shown in S3 Fig . Primer 1 ( INT N45 ) and primer 2 ( ol248 ) were used to determine successful integration of the selectable marker at the targeted locus . Primers 3 ( N45 KO1 ) and 4 ( N45 KO2 ) were used to verify deletion of the gene . After confirmation of integration , genomic DNA from wild type and mutant parasites was digested with HindIII and the fragments were separated on a 0 . 8% agarose gel , blotted onto a nylon membrane ( GE Healthcare ) , and probed with a PCR fragment made with primers N0453 and N0454 which is homologous to the P . berghei genomic DNA just outside of the targeted region . All of the oligonucleotides used to genetically confirm these mutant parasite lines can be found in S1 Table . Chromosomes of wild type , gene knockout and GFP-tagged parasites were separated by pulsed field gel electrophoresis ( PFGE ) on a CHEF DR III ( BioRad ) using a linear ramp of 60–500 s for 72 h at 4 V/cm . Gels were blotted and hybridized with a probe recognizing both the resistance cassette in the targeting vector and , more weakly , the 3′UTR of the P . berghei dhfr/ts locus on chromosome 7 . To record the nuclei number per schizont , merozoites in late schizonts were counted 18–24 hours after schizont cultures were made . Exflagellation was examined on day 4 to 5 post-infection . Ten μl of gametocyte-infected blood was obtained from the tail with a heparinised pipette tip and mixed immediately with 40 μl of ookinete culture medium ( RPMI1640 containing 25 mM HEPES , 20% fetal bovine serum , 10 mM sodium bicarbonate , 50 μM xanthurenic acid at pH 7 . 6 ) . The mixture was placed under a Vaseline-coated cover slip and 15 min later exflagellation centres were counted by phase contrast microscopy in 12–15 fields of view using a 40× objective and 10× ocular lens . Ookinete formation was assessed the next day . Ten μl of infected tail blood was obtained as above , mixed immediately with 40 μl ookinete culture medium , and incubated for 2 hours at 20°C to allow completion of gametogenesis and fertilisation . Each culture was then diluted with 0 . 45 ml of ookinete medium and incubated at 20°C for a further 21–24 hours to allow ookinete differentiation . Cultures were pelleted for 2 min at 5000 rpm and then incubated with 50 μl of ookinete medium containing Hoechst 33342 DNA dye to a final concentration of 5 μg/ml and a Cy3-conjugated mouse monoclonal antibody 13 . 1 [58] recognizing the P28 protein on the surface of ookinetes and any undifferentiated macrogametes or zygotes . P28-positive cells were counted with a Zeiss AxioImager M2 microscope ( Carl Zeiss , Inc ) fitted with an AxioCam ICc1 digital camera . Ookinete conversion was expressed as the percentage of P28 positive parasites that had differentiated into ookinetes [45] . For mosquito transmission experiments 20–50 Anopheles stephensi SD500 female mosquitoes were allowed to feed for 20 min on anaesthetised infected mice whose asexual parasitaemia had reached ~12–15% and were carrying comparable numbers of gametocytes as determined by Giemsa stained blood films . Days 5 , 7 , 10 , 14 , and 21 days post-infection ( dpi ) 20 mosquitoes were dissected and oocysts on their midguts counted . Oocyst formation was examined following Hoechst 33342 staining in PBS for 10–15 min and guts were mounted under Vaseline-rimmed cover slips . Counting and images were recorded using 10x and 63x oil immersion objectives on a Zeiss AxioImager M2 microscope fitted with an AxioCam ICc1 digital camera . At 14 and 21 dpi , the same mosquito midguts used to record the oocyst number were homogenised in a loosely fitting homogeniser to release sporozoites , which were then quantified using a haemocytometer . Only for 21 dpi mosquitoes , salivary glands were dissected and homogenised in a loosely fitting homogeniser to release sporozoites , which were then quantified using a haemocytometer . Due to day-to-day variations in transmission levels , all data were normalised to a matching number of wild type controls analysed on the same day . Mosquitoes infected with wild type or Δcyc3 parasites were used to perform bite back experiments with a TO mouse each in three independent experiments . Blood stage parasitaemia was measured for wild-type and Δcyc3 by Giemsa staining at 4 dpi . Oocyst diameter was measured with the AxioVision Imager software from images of 50–60 oocysts , in triplicate for 5 , 7 , 10 , 14 and 21 dpi using a 63x oil immersion objective on a Zeiss AxioImager M2 microscope fitted with an AxioCam ICc1 digital camera . For genetic complementation , we used either male ( ∆p48/45 and ∆hap2 ) or female ( ∆dozi and ∆nek4 ) parasites using a method described previously [30] . Briefly , mice were infected with combinations of the different parasite lines mentioned above and subsequently fed to 3–6 day old mosquitoes . These were dissected 12–14 dpi and the diameter of oocysts was determined as mentioned above . Statistical analyses were performed using GraphPad Prism ( GraphPad Software ) . For comparison between Δcyc3 and WT , an unpaired Student’s t-test was used . For the fluorescence pictures of CYC3-GFP oocysts , mosquito midguts have been dissected at 5 , 7 , 10 , 14 and 21 dpi and images were recorded using 63x oil immersion objectives on a Zeiss AxioImager M2 microscope fitted with an AxioCam ICc1 digital camera . Schizont , gametocyte and ookinete samples were isolated as described below . WT-GFP or CYC3-GFP samples were then purified using a GFP-Trap kit to immunoprecipitate GFP-fusion protein ( Chromotek ) . After the addition of Laemmli sample buffer , the samples were boiled and an equal concentration of total protein was loaded on a 4–12% SDS-polyacrylamide gel . Samples were subsequently transferred to nitrocellulose membranes ( Amersham Biosciences ) and immunoblotting performed using the Western Breeze Chemiluminescent Anti-Rabbit kit ( Invitrogen ) and anti-GFP polyclonal antibody ( Invitrogen ) at a concentration of 1:1250 , according to the manufacturer's instructions . Ookinete samples ( described below ) and mosquito midguts at 7 , 10 , 14 and 21 dpi were fixed in 4% glutaraldehyde in 0 . 1 M phosphate buffer and processed for routine electron microscopy as previously described [87] . Briefly , samples were post fixed in osmium tetroxide , treated en bloc with uranyl acetate , dehydrated and embedded in Spurr's epoxy resin . Thin sections were stained with uranyl acetate and lead citrate prior to examination in a JEOL1200EX electron microscope ( Jeol UK Ltd ) . Purification of gametocytes was achieved using a protocol modified from [88] . Mice were treated by intra-peritoneal injection of 0 . 1 ml of phenylhydrazine ( 6 mg/ml , Sigma ) in PBS to encourage reticulocyte formation four days prior to infection with parasites . Four days after parasites injection in mice , mice were treated with sulfadiazine ( Sigma ) at 20 mg/L in their drinking water for two days to eliminate asexual blood stage parasites . On day six post-injection ( p . i ) , mice were bled by cardiac puncture into heparin and gametocytes separated from uninfected erythrocytes on a 48% NycoDenz gradient ( 27 . 6% w/v NycoDenz in 5 mM Tris-HCl , pH 7 . 20 , 3 mM KCl , 0 . 3 mM EDTA ) in coelenterazine loading buffer ( CLB ) , containing PBS , 20 mM HEPES , 20 mM Glucose , 4 mM sodium bicarbonate , 1 mM EGTA , 0 . 1% w/v bovine serum albumin , pH 7 . 25 . Gametocytes were harvested from the interface and washed twice in RPMI 1640 ready for activation of gamete formation . Blood cells from day 5 p . i mice were placed in culture ( 40 ml RPMI 1640 , 8 ml foetal bovine serum , 0 . 5 ml penicillin and streptomycin; per 0 . 5 ml blood ) for 24 h at 37°C for schizont- ( with rotation at 100 rpm ) and at 20°C for ookinete production , as described above . Schizonts and ookinetes were purified on 60% and 63% NycoDenz gradients , respectively and harvested from the interface and washed . For 14 dpi oocysts , 20 mosquito midguts were collected and homogenised with PBS in a loosely fitting homogeniser to release sporozoites as described above . Parasites ( activated gametocytes and ookinetes ) were collected from ∆cyc3 or GFP-expressing lines . Samples were passed through a plasmodipur column to remove host DNA contamination prior to RNA isolation . Total RNA was isolated from purified parasites using an RNeasy purification kit ( Qiagen ) . RNA was vacuum concentrated ( SpeedVac ) and transported using RNA stable tubes ( Biomatrica ) . Strand-specific mRNA sequencing was performed from total RNA using TruSeq Stranded mRNA Sample Prep Kit LT ( Illumina ) according to the manufacturer's instructions . Briefly , polyA+ mRNA was purified from total RNA using oligo-dT dynabead selection . First strand cDNA was synthesised primed with random oligos followed by second strand synthesis where dUTPs were incorporated to achieve strand-specificity . The cDNA was adapter-ligated and the libraries amplified by PCR . Libraries were sequenced in an Illumina Hiseq machine with paired-end 100bp read chemistry . RNA-seq read alignments were mapped onto the P . berghei ANKA genome ( May 2015 release in GeneDB—http://www . genedb . org/ ) using Tophat2 ( version 2 . 0 . 13 ) [89] with parameters “—library-type fr-firststrand–no-novel-juncs–r 60” . Transcript abundances ( reported as FPKM- fragments per kilobase of exon per million fragments ) were quantified and differential expression analysis was performed using Cuffdiff2 version 2 . 2 . 1 [90] . Genes with fold change greater than 2 and p-value less than 0 . 05 were considered as significantly differentially expressed . As a form of independent validation of the differentially expressed genes , transcript abundances were extracted as raw read counts using htseq-count [91] and differential expression analysis performed using DESeq2 [92] in R version 3 . 2 . 1 . P . berghei GO terms ( Gene Ontology ) were downloaded from GeneDB ( http://www . genedb . org/; May 2015 release ) and gene ontology enrichment analysis was performed for the DEG lists using GOstats R package [93] . All analyses and visualizations were done with R packages- cummeRbund [94] and ggplot2 [95] . Total RNA was isolated from purified parasites using an RNeasy purification kit ( Qiagen ) . For qRT-PCR , cDNA was synthesised using an RNA-to-cDNA kit ( Applied Biosystems ) allowing quantification from 250 ng of total RNA . qRT-PCR reactions consisted of 2 μl cDNA , 5 μl SYBR green fast master mix ( Applied Biosystems ) , 0 . 5 μl ( 500 nM ) each of the forward and reverse primers , and 2 μl DEPC-treated water . Where possible , one of the primer pairs was placed over an intron , primers had melting temperatures of 60–62°C and together amplified a region 70–200 bp long . Analysis was conducted using an Applied Biosystems 7500 fast machine with the following cycling conditions: 95°C for 20 sec followed by 40 cycles of 95°C for 3 sec; 60°C for 30 sec . Wild-type expression was determined using the Pfaffl method [96] . Relative quantification in the mutant line was normalised against wild-type expression using the ∆∆Ct method . Both methods used hsp70 ( and arginine-tRNA synthetase for wild-type expression ) as a reference gene to provide a baseline of transcription levels between replicates to allow normalization of the data [29] . Three biological replicates were used for each stage ( each with two technical replicates ) . See S1 Table for a full list of the primers used for qRT-PCR . Statistical analyses were performed using Excel and GraphPad Prism ( GraphPad Software ) , with graphs showing normalised expression in ∆cyc3 compared to a transcription baseline derived from WT . For relative gene expression , a Student’s unpaired t-test was used . For RNA-seq and qRT-PCR comparison , we used ≥30 genes at each stage and used GraphPad Prism ( GraphPad Software ) to calculate fit and coefficient of determination . Standard protocols of handling Schizosaccharomyces pombe and Saccharomyces cerevisiae were followed [97 , 98] . For yeast complementation experiments , the triple-cln mutant ( cln1∆ cln2∆ cln3∆ TRP::GAL1-CLN3 ade1 leu2-3 his2 trp1-1 ura3∆ bar1∆ pep4∆::LEU2 ) or cdc13 ts mutant strain ( h- cdc13-117 leu1-32 ) was used . High resolution live cell imaging was performed using an Olympus-based Delta Vision Elite work station fitted with a 100x objective ( numerical NA 1 . 4 , oil ) . Post-acquisition analysis was carried out using Applied Precision software . Images presented are 2D projections of deconvolved Z-stacks of 0 . 3 μm optical sections . The ookinete motility assay was performed as previously described [99] . Twenty five microliters of the ookinete cultures were added to an equal volume of Matrigel ( BD ) on ice , mixed thoroughly , added to a slide , covered with a Vaseline-rimmed cover slip , and sealed with nail varnish . The Matrigel was then allowed to set at room temperature for at least 30 minutes . After identifying a field containing ookinetes , time-lapse videos ( 1 frame every 5 sec , for 10 min ) were taken of ookinetes using the differential interference contrast ( DIC ) settings with a 63× objective lens on a Zeiss AxioImager M2 microscope fitted with an AxioCam ICc1 digital camera controlled by the Axiovision ( Zeiss ) software package . Speed of motility of individual ookinetes was measured by multiplying the number of body lengths moved by the length of the ookinete during the 10 min video , divided by 10 . Multiple independent slides and cultures were used for each parasite line . The nuclear content of the Δcyc3 ookinete was measured by the formula as previously described [58] . Briefly , to measure nuclear DNA content of microgametocytes , digital images of Hoechst-stained fixed or unfixed cells were obtained using a Zeiss AxioImager M2 microscope fitted with an AxioCam ICc1 digital camera and analysed using ImageJ software version 1 . 33u ( National Institutes of Health , USA ) . The relative nuclear fluorescence intensity was determined with the following formula: Area ( pixel ) × ( average intensity ( relative units ) − average background intensity ( relative units ) ) . The nuclear fluorescence intensity was standardized to the haploid DNA content by measuring the fluorescence intensity of ring-stage parasites in parallel on the same slide and with the same microscope and camera settings . | The malaria parasite is a single-celled organism that multiplies asexually in a non-canonical way in both vertebrate host and mosquito vector . In the mosquito midgut , atypical cell division occurs in oocysts , where repeated nuclear division ( endomitosis ) precedes cell division , which then gives rise to many sporozoites in a process known as sporogony . The molecular mechanisms controlling this process are poorly understood . In many model organisms including mouse and yeast cells the cell cycle is regulated by members of the cyclin protein family , but the role of this family in the malaria parasite is unknown . Here , we show that there are only three cyclin genes and investigate the function of the single P-type cyclin ( CYC3 ) in the rodent malaria parasite , Plasmodium berghei . We show that CYC3 has a cytoplasmic and nuclear localisation throughout most of the parasite lifecycle and by gene deletion we demonstrate that CYC3 is important for normal oocyst development , maturation and sporozoite formation . Moreover , we show that deletion of cyc3 affects the transcription of genes required for cell signalling and oocyst development . The data suggest that CYC3 modulates asexual multiplication in oocysts and plays a vital role in parasite development in the mosquito . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Plasmodium P-Type Cyclin CYC3 Modulates Endomitotic Growth during Oocyst Development in Mosquitoes |
Many redundancies play functional roles in motor control and motor learning . For example , kinematic and muscle redundancies contribute to stabilizing posture and impedance control , respectively . Another redundancy is the number of neurons themselves; there are overwhelmingly more neurons than muscles , and many combinations of neural activation can generate identical muscle activity . The functional roles of this neuronal redundancy remains unknown . Analysis of a redundant neural network model makes it possible to investigate these functional roles while varying the number of model neurons and holding constant the number of output units . Our analysis reveals that learning speed reaches its maximum value if and only if the model includes sufficient neuronal redundancy . This analytical result does not depend on whether the distribution of the preferred direction is uniform or a skewed bimodal , both of which have been reported in neurophysiological studies . Neuronal redundancy maximizes learning speed , even if the neural network model includes recurrent connections , a nonlinear activation function , or nonlinear muscle units . Furthermore , our results do not rely on the shape of the generalization function . The results of this study suggest that one of the functional roles of neuronal redundancy is to maximize learning speed .
In the human brain , numerous neurons encode information about external stimuli , e . g . , visual or auditory stimuli , and internal stimuli , e . g . , attention or motor planning . Each neuron exhibits different responses to stimuli , but neural encoding , especially in the visual and auditory cortices , can be explained by the maximization of stimulus information [1]–[3] . This maximization framework can also explain learning that occurs when the same stimuli are repeatedly presented; previous neurophysiological experiments have suggested that perceptual learning causes changes in neural encoding to enhance the Fisher information of a visual stimulus [4] . However , a recent study has suggested that information maximization alone is insufficient to explain neural encoding . Salinas has suggested that “how encoded information is used” needs to be taken into account: neural encoding is influenced by the downstream circuits and output units to which neurons project , and it is ultimately influenced by animal behavior [5] . In the motor cortex , neural encoding is influenced by the characteristics of muscles ( output units ) because motor cortex neurons send motor commands to muscles through the spinal cord . In adaptation experiments , some motor cortex neurons exhibit rotations in their preferred directions ( PDs ) , and these rotations result in a population vector that is directed toward a planned target [6] . Neural encoding therefore changes to minimize errors between planning and behavior , suggesting that neural encoding is influenced by behavior and properties of output units . A critical problem exists in the relationship between motor cortex neurons and output units: the neuronal redundancy problem , or overcompleteness , which refers to the fact that the number of motor cortex neurons far exceeds the number of output units . Many different combinations of neural activities can therefore generate identical outputs . Neurophysiological and computational studies have revealed that the motor cortex exhibits neuronal redundancy [7] , [8] . However , it remains unknown how neuronal redundancy influences neural encoding . In other words , we do not yet understand the functional roles of neuronal redundancy in motor control and learning , though other types of redundancies are known to play various functional roles [9] . One of these types of redundancy is muscle redundancy: many combinations of muscle activities can generate identical movements . The functional roles of this muscle redundancy include impedance control to achieve accurate movements [10] , reduction of motor variance by constructing muscle synergies [11] , and learning internal models by changing muscle activities [12] . Another redundancy is kinematic redundancy: many combinations of joint angles result in identical hand positions . This redundancy ensures the stability of posture even if one joint is perturbed [13] , and it facilitates of motor learning by increasing motor variance in a dimension irrelevant to the desired movements [14] . Redundancies therefore play important functional roles in motor control and learning . Similar to the muscle and kinematic redundancies , neuronal redundancy likely has functional roles in motor control and learning . However , the functional roles of this redundancy are unclear . Here , using a redundant neural network , we investigate these functional roles by varying the number of model neurons while holding the number of output units constant . This manipulation allows us to control the degree of neuronal redundancy because , if a neural network includes a large number of neurons and a small number of output units , many different combinations of neural activities can generate identical outputs . It should be noted that we used a redundant neural network model that can explain neurophysiological motor cortex data [7] . The key conclusion arising from our study is that one of the functional roles of neuronal redundancy is the maximization of learning speed . Initially , a linear model with a fixed decoder was used . Analytical calculations revealed that neuronal redundancy is a necessary and sufficient condition to maximize learning speed . This maximization is invariant whether the distribution of PDs is unimodal [6] or bimodal [15]–[17]; both distributions have been reported in neurophysiological investigations . Second , numerical simulations confirmed the invariance of our results , even when the neural network included an adaptable decoder , a nonlinear activation function , recurrent connections , or nonlinear muscle units . Third , we show that our results do not depend on learning rules by using weight and node perturbation , both of which are representative stochastic gradient methods [18] . Finally , we demonstrate that our hypothesis does not depend on the shape of the generalization function which shape depends on the task ( broad or sharp in force field [19] , [20] or visuomotor rotation adaptation [21] , respectively ) . Our results strongly support our hypothesis that neuronal redundancy maximizes learning speed .
Although we have revealed that neuronal redundancy maximizes learning speed when , it is important to verify that the effect is caused by the neuronal redundancy , i . e . , the dimensional gap between and , and not simply the number of neurons . In this section , we investigate this question by varying both and while assuming that each component of is randomly sampled from a Gaussian distribution . Figures 5A and 5B show the learning speed and the learning curve produced when , and with a fixed non-homogeneous decoder . If alone were important for maximizing learning speed , learning speed would be faster when than when or . However , the results shown in these figures support the opposite conclusion , i . e . , learning speed becomes slower when compared to the other cases . This result suggests that the number of neurons alone is not important for maximizing learning speed . Figures 5C and 5D show the learning speed and learning curve produced when , or with and a fixed non-homogeneous decoder . If neuronal redundancy were important , the learning speed would be faster when than when or . These figures support this hypothesis; learning speed increased when compared to the other cases . Taken together , these results indicate that the important factor for maximizing learning speed is in fact neuronal redundancy and not simply the number of neurons . In addition , we investigated whether neuronal redundancy or neuron number is important when is adaptable . In this case , we only show learning curves because learning speed cannot be exponentially fitted , which makes it impossible to calculate learning speed . Figures 5E and 5F show the learning curves calculated when , or and , or with . These figures show the same results as the case when is fixed; even when is adaptable , the important factor for maximizing learning speed is neuronal redundancy , not simply the number of neurons . The generality of our results should be investigated because we analyzed only linear and feed-forward networks , but neurophysiological experiments have suggested the existence of recurrent connections [25] and nonlinear neural activation functions [26] . Also , only a linear rotational perturbation task was considered , so we need to investigate whether our results hold when the constrained tasks are nonlinear because , in fact , motor cortex neurons solve nonlinear tasks . The neurons send motor commands and control muscles whose activities are nonlinearly determined: muscles can pull but cannot push . Using numerical simulations , we show that neuronal redundancy maximizes learning speed , even when the neural network includes recurrent connections ( Figure S1 ) , when it includes nonlinear activation functions ( Figure S2 ) , and when the task is nonlinear ( Figure S3 ) . In addition , we used only deterministic gradient descent , so the generality regarding the learning rule needs to be investigated . In fact , previous studies have suggested that stochastic gradient methods are more biologically relevant than deterministic ones [27] , [28] . Analytical and numerical calculations confirm that our results are invariant even when the learning rule is stochastic ( Figure S4 ) . Our results therefore have strong generality .
We have quantitatively demonstrated that neuronal redundancy maximizes learning speed . The larger the dimensional gap grows between the number of neurons and the number of constrained tasks , the faster learning speed becomes . This maximization does not depend on whether the PD distribution is unimodal or bimodal , the decoder is fixed or adaptable , the network is linear or nonlinear , the task is linear or nonlinear , or the learning rule is stochastic or non-stochastic . Additionally , we have shown that neuronal redundancy has another important functional role: it provides robustness in response to neural noise . Furthermore , neuronal redundancy maximizes learning speed in a manner independent of the shape of the generalization function . These results strongly support the generality of our results . Neuronal redundancy maximizes learning speed because only equalities , , need to be satisfied , and -dimensional neural activity is adaptable ( ) . This dimensional gap yields the large dimensional subspace of in which the equalities are satisfied . The more increases , the greater the fraction of the subspace becomes: . Neuronal redundancy , rather than the number of neurons , thus enables to rapidly reach a single point in the subspace . This interpretation likely applies even in the cases of an adaptable decoder , recurrent connections , a nonlinear network , a nonlinear task , and a stochastic learning rule . Furthermore , this interpretation is supported by the results shown in Figure 5; the bigger grows , the faster learning speed becomes . At first glance , our results may seem inconsistent with the results of Werfel et al . [18] , who concluded that learning speed is inversely proportional to . In their model , because they considered the single-layer linear model , is the same as the number of input units , which is defined as ( = ) in the present study . A similar tendency can be observed in Figure 5; the more increases , the slower learning speed becomes . We calculated the optimal learning rate and speed as shown in Text S1 , and confirmed that learning speed is inversely proportional to . Thus , our results are consistent with Werfel's study and additionally suggest that neuronal redundancy maximizes learning speed . Neuronal redundancy plays another important role: generating robustness in response to neural noise ( Figure 6 ) . Because neuronal redundancy has the same meaning as overcompleteness , its functional role is the same as the robustness of overcompleteness in the face of perturbations in signals [32] . This additional functional role further supports our hypothesis that neuronal redundancy is a special neural basis on which to maximize learning speed . For example , if we increase the learning rate in a non-redundant network , the learning speed approaches the maximal speed in a redundant network in which the learning rate is fixed to . As shown in Figure 6 , however , a non-redundant network is not robust with respect to neural noise . Furthermore , neuronal redundancy minimizes residual errors when the neural network includes synaptic decay [7] ( see the Methods section and Figure S5 ) . Thus , neuronal redundancy represents a special neural basis for maximizing learning speed while minimizing residual error and maintaining robustness in response to neural noise .
Our study assumed the following task: participants move their arms towards one of radially distributed targets . If the th target is presented in the th trial , the neural network model receives the input ( , ) , where . The input units project to neurons ( hidden units ) , the activities of which are determined by ( 10 ) where is synaptic weight in the th trial , is the standard deviation of neural activity noise , denotes independent normal Gaussian random variables , and is the number of neurons ( Figure 1 ) . The th neuron has a PD given by and a modulation depth , where , this cosine tuning having been reported by many neurophysiological studies . The neural population generates a force of through a decoder matrix : ( 11 ) where is the number of outputs , which , in most cases , is set to 2 . When is fixed and homogeneous , the th and th components of are defined as and , respectively , where division by is used for scaling and FD is randomly sampled from a uniform distribution ( ) . When is fixed and non-homogeneous , is randomly sampled from a probability distribution and divided by . As a result , the neural network generates a final hand coordinate : ( 12 ) which means that is perturbed by a rotation which assumes a visuomotor rotation or curl force field . Rotational perturbations are assumed because many behavioral studies have used them . Because we discuss only the endpoint of the movement , we refer to as the motor command . The constrained tasks are those that the neural network generates toward , i . e . , , which means the number of constrained tasks is the same as . We used instead of in the following sections . If the error occurs between and , synaptic weights are adapted to reduce the squared error , which is defined as , based on a gradient descent method ( 13 ) where is the synaptic decay rate , is the learning rate ( is set to 0 . 2 in most parts of the present study ) , is the strength of synaptic drift , and denotes normal Gaussian random variables . Since each component of is , multiplying by allows trial-by-trial variation of both and to be . As shown in Text S1 , the optimal learning rate is ( ) , suggesting that we consider a quasi-optimal learning rate . It should be noted that our results hold whether the learning rate is optimal or quasi-optimal because the results do not depend on . It should also be noted that the amount of variation in does not explicitly depend on . Equation ( 13 ) yields the following update rule of squared error: ( 14 ) where , and denotes the identity matrix . At first , we assume a case in which for simplicity . Because is symmetric , can be decomposed as , where each row of is one of the eigenvectors ( ) and each diagonal component of a diagonal matrix is one of the eigenvalues of . This decomposition transforms equation ( 14 ) into the simple form ( 15 ) where and . This recurrence formula yields the analytical form of the learning curve: ( 16 ) Equation ( 16 ) requires that the larger the eigenvalues become , the faster the learning speed becomes and the smaller the residual error becomes ( Figure S5 ) . Because ( 17 ) whose component is , simple algebra gives the analytical form of the eigenvalues , ( 18 ) which are also , suggesting that learning speed does not depend explicitly on . Because the learning speed is determined by the smaller eigenvalue , the necessary and sufficient conditions to maximize learning speed , or to maximize the smaller eigenvalue , are ( 19 ) and ( 20 ) What kind of conditions can simultaneously satisfy equations ( 19 ) and ( 20 ) ? The only answer is sufficient neuronal redundancy , i . e . , , because sufficient neuronal redundancy enables self-averaging properties to exist in a neural network , i . e . , ( 21 ) ( 22 ) and ( 23 ) where is the probability distribution in which FDs are randomly sampled . Conversely , if equations ( 21 ) , ( 22 ) , and ( 23 ) are satisfied in all of the sets of randomly sampled FDs , the number of neurons needs to satisfy because the fluctuation of Monte Carlo integrals is [24] . Therefore , to maximize learning speed , the necessary and sufficient condition is sufficient neuronal redundancy . The above analytical calculations hold even when . Equation ( 13 ) yields the recurrence equation of the squared error: ( 24 ) where is set to 1 for simplicity . Using , this equation can be written as ( 25 ) The larger the eigenvalue becomes , the faster learning speed becomes if and have the same sign , or if . This inequality is appropriate if the equality can be proved , where is a positive constant . To prove this equality , let us assume that in the 1st trial after the rotational perturbation is applied , output can be written as because the neural network can generate accurate outputs if there is no perturbation . In this case , ( 26 ) where is a positive constant . Thus , the larger becomes , the faster learning speed becomes even when ; analytical calculations show that neuronal redundancy maximizes learning speed even when . When is fixed and non-homogeneous , i . e . , , , , , and , the necessary and sufficient conditions for maximizing learning speed are given by the following equations: ( 27 ) ( 28 ) with neuronal redundancy assumed . Equations ( 27 ) and ( 28 ) can be satisfied when , for example , ( 29 ) ( shown in Figure 3A with and ) , or ( 30 ) ( shown in Figure 3D with , , and ) , where . Because the learning rate of motor commands is determined by ( see the following section ) , is determined based on the results of behavioral studies [33] . We cannot analytically calculate the general class of and the distributions of PDs satisfying equations ( 27 ) and ( 28 ) , but broad classes of those distributions can satisfy these equations because the classes include even asymmetric distributions , e . g . , when , . When is adaptable , this is also adapted to minimize the squared error: ( 31 ) where is set to , is a normal Gaussian random variable , and is set to 0 . 1 in the Adaptable Decoder section and 0 . 05 in the Importance of Neuronal Redundancy section . This learning rule corresponds to back-propagation [34] . In the Importance of Neuronal Redundancy section , the neural network generates the output , which is determined by ( 32 ) for the th trial . An initial value of is randomly sampled from the normal Gaussian distribution and divided by for scaling . The input is randomly sampled from the normal Gaussian distribution and is normalized to satisfy to avoid the effect of this value on learning speed . In addition , we used a fixed value of because the generalization function ( see the following section ) strongly depends on , i . e . , . It should be noted that learning speed does not explicitly depend on because learning speed is determined only by the minimum eigenvalue of . Equation ( 13 ) yields the following update rule for motor commands: ( 33 ) If equations ( 27 ) and ( 28 ) ( or ( 22 ) and ( 23 ) ) are satisfied , equation ( 33 ) can be written as ( 34 ) where the cross term of and determines the generalization function , e . g . , , if we define . We set and to satisfy . It should be noted that equation ( 34 ) corresponds to a model for sensorimotor learning that can explain the results of behavioral experiments [35] , suggesting that our hypothesis is consistent with the results of behavioral experiments . Because the shape of the generalization function depends on the task , we need to confirm the generality of our results with regard to the shape of the generalization function . To simulate various shapes of generalization functions , we used the von-Mises function ( 35 ) where , , and are the precision parameter , the preferred direction of the th input unit , and the number of input units , respectively . The normalization factor is determined to make to avoid the influence of this value on the learning speed , where . This normalization permits us to investigate the influence of the shape of the generalization function alone on learning speed . The larger the value of , the sharper the shape of the generalization function becomes . We set to 100 throughout this study . We conducted 100 baseline trials with and to identify the baseline values of . The initial value of , , was set to . After these trials , 100 learning trials were conducted using and . Learning speed was calculated by fitting the exponential function to . All the figures denote which was obtained only in learning trials . The present study calculated learning speed and learning curves by averaging the results of 1000 sets of baseline and learning trials , each set including an identical target sequence that was randomly sampled , and each set using different FD values . For all of the statistical tests , we used the Wilcoxon sign rank test . It should be noted that the -value was indicated only if the value was significantly different from 0; no statistically significant differences were detected . | There are overwhelmingly more neurons than muscles in the motor system . The functional roles of this neuronal redundancy remains unknown . Our analysis , which uses a redundant neural network model , reveals that learning speed reaches its maximum value if and only if the model includes sufficient neuronal redundancy . This result does not depend on whether the distribution of the preferred direction is uniform or a skewed bimodal , both of which have been reported in neurophysiological studies . We have confirmed that our results are consistent , regardless of whether the model includes recurrent connections , a nonlinear activation function , or nonlinear muscle units . Additionally , our results are the same when using either a broad or a narrow generalization function . These results suggest that one of the functional roles of neuronal redundancy is to maximize learning speed . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biology",
"neuroscience"
] | 2012 | Maximization of Learning Speed in the Motor Cortex Due to Neuronal Redundancy |
Although bone has a unique restorative capacity , i . e . , it has the potential to heal scarlessly , the conditions for spontaneous bone healing are not always present , leading to a delayed union or a non-union . In this work , we use an integrative in vivo - in silico approach to investigate the occurrence of non-unions , as well as to design possible treatment strategies thereof . The gap size of the domain geometry of a previously published mathematical model was enlarged in order to study the complex interplay of blood vessel formation , oxygen supply , growth factors and cell proliferation on the final healing outcome in large bone defects . The multiscale oxygen model was not only able to capture the essential aspects of in vivo non-unions , it also assisted in understanding the underlying mechanisms of action , i . e . , the delayed vascularization of the central callus region resulted in harsh hypoxic conditions , cell death and finally disrupted bone healing . Inspired by the importance of a timely vascularization , as well as by the limited biological potential of the fracture hematoma , the influence of the host environment on the bone healing process in critical size defects was explored further . Moreover , dependent on the host environment , several treatment strategies were designed and tested for effectiveness . A qualitative correspondence between the predicted outcomes of certain treatment strategies and experimental observations was obtained , clearly illustrating the model's potential . In conclusion , the results of this study demonstrate that due to the complex non-linear dynamics of blood vessel formation , oxygen supply , growth factor production and cell proliferation and the interactions thereof with the host environment , an integrative in silico-in vivo approach is a crucial tool to further unravel the occurrence and treatments of challenging critical sized bone defects .
Although bone has a unique restorative capacity , i . e . it has the potential to heal scarlessly , the conditions for spontaneous bone healing are not always present , leading to a delayed union or a non-union . The orthopedic literature does not specify a universally accepted definition of a fracture non-union [1] , [2] . The eventual bony union after an atypical long period of healing , in comparison to the normal healing period , is called a delayed union . The absence of healing during at least three to six months defines a fracture non-union in humans . Fracture non-unions ( hypertrophic , atrophic or oligotrophic ) are classified based on their radiographic and histological appearance [1] , [3] . Hypertrophic non-unions are characterized by an abnormal vascularity and abundant callus formation . They are typically caused by excessive motion at the fracture site , which prevents bony bridging although the essential biological factors are present [1] . Atrophic non-unions , however , are the result of inadequate biological conditions and typically appear on radiographs as blunted bony ends . They show little callus formation around the fracture gap , filled with mostly fibrous tissue and little or no evidence of mineral deposition [1] . Oligotrophic non-unions have some radiographic and biological characteristics of both hypertrophic and atrophic non-unions , i . e . they possess the required biological activity but show little or no callus formation [4] . Excess motion , a large interfragmentary gap [5] , open fracture [5]–[7] , the particular bone [8] , location of the trauma within the bone [8] , loss of blood supply [9] , severe periosteal and soft-tissue trauma [6] , [7] are some of the mechanical and biological risk factors for the development of a non-union . Preexisting patient risk factors such as old age [10] , cachexia and malnutrition [11] , immune compromise [12] , genetic disorders ( e . g . type 1 neurofibromatosis ) , osteoporosis [13] , anticoagulants [14] , anti-inflammatory agents [15] , etc . may also affect the fracture healing outcome but are not the primary causes [16] . Besides an extensive amount of experimental research , several computational models have also been developed to further unravel the occurrence of fracture non-unions . For comprehensive reviews on mathematical models of fracture healing , we refer the reader to Geris et al . [17] , Isaksson et al . [18] and Pivonka et al . [19] . Despite the large amount of ( often phenomenological ) information existing in the literature , additional in vivo , in vitro and in silico research is still required to address the key mechanisms that lead to fracture non-unions , determine the factors predictive of fracture complications and establish the optimal therapeutic strategies for each type of fracture non-union . In this work we propose an integrative in vivo - in silico approach to investigate the occurrence of oligotrophic and atrophic non-unions as well as to design possible treatment strategies thereof . The gap size of the domain geometry of a previously published mathematical model has been enlarged in order to study the complex interplay of blood vessel formation , oxygen supply , growth factors and cell proliferation on the final healing outcome in large bone defects . The simulation results are corroborated by comparison with dedicated experimental data . Next , the mathematical model is used to explain the underlying mechanisms that lead to the experimental observations as well as design different treatment strategies . Finally , the potential of the combined in silico - in vivo approach is demonstrated by applying it to the case of BMP-treated fracture healing .
All animal experiments were conducted according to the regulations and with approval of the Animal Ethics Committee of the KU Leuven . C57BL/6 mice were purchased from the R . Janvier Breeding Center ( France ) . A segmental defect was created in the right tibia of 14 week-old male mice as described elsewhere [20] . Briefly , animals were anaesthetized with a ketamine-xylazine mixture ( 100 mg/kg ketamine and 15 mg/kg xylazine ) and the right lower leg was shaved . A custom-made external fixator , based on the Ilizarov external fixation device , was fixed to the tibia using 27 G steel needles . Subsequently , the tibia was exposed and a 4 . 5–5 mm mid-diaphyseal segment was excised with a 6 . 5 mm diamond saw disk ( Codema n . v . , Kortrijk , Belgium ) . A demineralized CopiOs scaffold ( 2 . 5×2 . 5×5 mm3; Zimmer b . v . b . a . , Wemmel , Belgium ) seeded with 1×106 mouse periosteal cells ( passage 4 ) was implanted , the skin was sutured and animals received postoperative analgesia ( buprenorphine , 60 µg/kg body weight ) . The demineralized CopiOs- scaffold was used to minimize the soft tissue collapse within the critical size defect . After 3 , 14 or 56 days animals were sacrificed , the tibia was excised and samples were analyzed by μCT and then processed for histology . Murine periosteum-derived cells ( mPDC ) were isolated from the long bones of 8 week-old male mice as previously described [21] . In short , the femurs and tibias isolated from 8 week old male C57BL/6J mice were dissected and digested with collagenase-dispase after protecting the epiphyses with low melting point agarose . After a filtration and washing step , the cells were plated at 1×104 cells per square centimeter and replated when reaching 80–90% confluency . After isolation , cells were pooled per 2–3 mice and cultured in a humidified incubator at 37°C with 5% CO2 in α-minimal essential medium ( α-MEM ) supplemented with 2 mM glutaMAX-I , 1% penicillin/streptomycin ( 100 units/ml and 100 µg/ml respectively ) and 10% fetal bovine serum ( all from Gibco , Life Technologies , Gent , Belgium ) . When reaching 80–90% confluency , cells were trypsinized and reseeded at 7500 cells/cm2 . To deliver BMP2 at the defect site , mPDCs were transduced 72 hours prior to implantation with an adenoviral vector encoding human BMP2 ( a generous gift from Dr . Frank Luyten , KU Leuven , Belgium ) at a multiplicity of infection of 50 . Bone formation in large bone defects was followed by radiographic images at different time points after surgery using the Skyscan 1076 high resolution in vivo micro-computed tomography ( μCT ) scanner ( Bruker-μCT , Kontich , Belgium ) . For bone quantification , samples retrieved at day 56 were scanned using the high resolution SkyScan 1172 μCT system ( Bruker-μCT ) at a pixel size of 10 µm with 50 kV tube voltage and 0 . 5 mm aluminum filter . Projection data was reconstructed using the NRecon software and quantification of mineralized tissue was performed using the CTAn software ( both from Bruker-μCT ) . Isolated bones were fixed in 2% paraformaldehyde overnight and decalcified in EDTA for 14 days at 4°C prior to dehydration , embedded in paraffin and sectioned at 4 µm . Histochemical staining with hematoxylin and eosin ( H&E ) and immunohistochemical staining for mouse CD31 is routinely performed in our lab and has been described previously [21] . Images were taken on a Zeiss Axioplan 2 light microscope using the Zeiss AxioVision software . Data are presented as means ± standard error of the means . Data were analyzed by one-way ANOVA using the NCSS statistical software . Differences were considered statistically significant at p<0 . 05 . The multiscale computational framework for the mathematical modelling of bone fracture healing and its relation to angiogenesis was established earlier and has been described in detail in [22] . The framework consists of ( 1 ) a tissue level describing the various key processes of bone fracture healing with 10 continuous variables , ( 2 ) a cellular level representing the developing vasculature with discrete endothelial cells and ( 3 ) an intracellular level that defines the internal dynamics of the Dll4-Notch signaling pathway in every endothelial cell ( Figure 1 ) . The model accounts for the various key processes that occur during the soft and hard callus phase of bone fracture healing ( see [22] for a more detailed description ) . While the model described in [22] already partially accounted for the role of oxygen , we have recently extended the model to capture the various effects of oxygen on cellular processes in a much more complete and refined way [23] . A brief description of the oxygen model is found below and more details are given in Supporting Text S1 . After the initial inflammation phase ( which is not included in the current mathematical model ) , the fracture callus is filled with a cocktail of granulation matrix , stem cells and growth factors . In regions where oxygen is abundantly available ( i . e . close to the cortex in the case of normal fracture healing ) , the mesenchymal stem cells will directly differentiate into osteoblasts and form bone through the intramembranous pathway . In regions where the oxygen tension is lower ( i . e . the central fracture callus in the case of normal fracture healing ) , the mesenchymal stem cells will differentiate to chondrocytes that will form a cartilage template to mechanically stabilize the fracture . This is followed by endochondral ossification during which blood vessels and osteoblasts are attracted to the central fracture callus , resulting in degradation of the cartilage template and bone formation . Finally , the newly formed bone is remodeled ( not included in the current mathematical model ) . At the tissue level , the fracture healing process is described by calculating the spatiotemporal evolution of the density of mesenchymal stem cells ( cm ) , osteoblasts ( cb ) , chondrocytes ( cc ) , fibroblasts ( cf ) , bone ( mb ) , cartilage ( mc ) , fibrous matrix ( mf ) , osteochondrogenic growth factor ( gbc ) , angiogenic growth factor ( gv ) and oxygen ( n ) using 10 non-linear , coupled partial differential equations of the taxis-diffusion-reaction type . At the cellular level , the evolution of the discrete vasculature is determined by sprouting , vascular growth and anastomosis and is modeled by a lattice-free method . At the intracellular level , an agent-based model is used to implement the rules that capture the intracellular dynamics of the Dll4-Notch signaling pathway which determines tip cell selection during sprouting angiogenesis . The oxygen model includes an accurate description of the oxygen dependency of a number of cellular processes , namely osteogenic and chondrogenic differentiation , cell proliferation , cell death , oxygen consumption and the hypoxia-dependent production of an angiogenic growth factor . The cellular consumption of oxygen was described using a Michaelis-Menten kinetic law where the cell-specific maximal oxygen consumption rate has the following relative cellular order: chondrocytes<MSCs<osteoblasts<fibroblasts . The oxygen values at which the considered cell-specific oxygen-dependent processes occur at maximal rate or at which their rate changes are based on a rigorous literature screening of the state-of-the-art experimental knowledge ( Figure 2 ) . More specifically , the relative order of the oxygen dependent processes was determined as accurately as possible since it is crucial to the behavior of the oxygen model . The complete description of the set of equations , the boundary and initial conditions , the parameter values , implementation details as well as some underlying assumptions and simplifications can be found in Supporting Text S1 as well as in previous publications [22] , [24] , [25] . The geometrical domain of the fracture callus , as well as the boundary conditions and initial positions of the endothelial cells ( cv ) are shown in Figure 3-B . Note that the periosteum near the bone ends is considered to be well vascularized such that a muscular contribution to the vasculature ( i . e . the initial position of the endothelial cells ) is unnecessary . To simulate the bone regeneration process in a large bone defect , the domain was extended over a distance equal to half the gap size of a murine critical sized defect ( 5 mm ) . The effect of the host environment on the fracture healing process is explored with several combinations of boundary conditions , however in the standard compromised condition the influence of the host environment is neglected thereby representing the worst-case scenario ( Figure 3-B ) . It has been shown experimentally that the amount of cells and growth factors is significantly reduced in a large fracture gap [26] , [27] . Therefore , in order to simulate this effect , the initial conditions for the MSCs and osteochondrogenic growth factors were decreased tenfold to 2 . 103 cells/ml and 10 ng/ml respectively in the central callus area ( indicated with dots in Figure 3-B ) . The initial oxygen tension ( ninit ) in the central callus area is equal to 3 . 7% . All other model parameters as well as initial and boundary conditions were left unchanged with respect to the normal healing case [23] and can be found in Supporting Text S1 ( Figure 3 ) . Note that the computational model does not simulate the presence of the demineralized CopiOs scaffold , which was used to minimize the soft tissue collapse within the critical size defect . Previous results have however shown that the demineralized carrier structure does not contribute nor enhance the bone formation process . The results of the mathematical model are quantified in terms of tissue fractions , specified for each part of the fracture callus ( i . e . endosteal , periosteal and intercortical ) . The tissue fractions are calculated by the following procedure: first the spatial images are binarized using tissue-specific thresholds ( 0 means that the tissue is not present , 1 means that the tissue is present in a grid cell ) . Subsequently , an equal weight is assigned to the different tissues , i . e . if a grid cell contains three tissues , the area of that grid cell is divided by three in the final calculations of the tissue ( area ) fractions [23] .
A qualitatively similar healing progression is predicted by the simulation results as observed experimentally ( Figures 4 and 5 ) . At early time points a periosteal reaction , characterized by a thickening of the periosteal layer ( Figures 5-B1 , B1′ and C1 ) as well as the presence of a hematoma , a fibrous-like tissue associated with the presence of numerous ( red ) blood cells ( Figures 5-B1 , B1″ ) , are observed at the cortical host bone site , both supporting the initial and boundary conditions that were applied in the multiscale model ( Figure 3-B ) . In the center of the large bone defect no signs of tissues or infiltration of blood vessels are detected , only scaffold material together with a low cellularity is observed ( Figures 5-B1-center , C1′ ) , corresponding to the predictions of the in silico model ( Figure 4-G ) . On day 14 , a periosteal endochondral ossification reaction is seen , evidenced by the presence of cartilage ( large round cells staining grey-blue with H&E ) and trabecular-like bone ( dense matrix , staining bright pink with H&E , with the clear presence of embedded osteocytes ) ( Figures 5-B2 , B2′; arrow indicates cartilage ) , while direct bone formation occurs endosteally ( Figures 5-B2 , B2″ ) . The mathematical model predicts a similar distribution of tissue formation , i . e . direct bone formation near the bony ends and endochondral ossification further away in the fracture callus ( Figure 4-B , C ) . In the center of the defect only a highly dense fibrous tissue is observed in both the experimental , the scaffold remains stained pink-blue with H&E but lack the presence of embedded cells ( Figure 5-B2-center ) , as well as the mathematical model ( Figure 4-A ) . In contrast to the experimental model , the mathematical model does not predict any blood vessels in the central callus area ( indicated with dots in Figure 3 ) . These vessels , however , appear to be small and immature whereas the blood vessels that are associated with the sites of bone formation are large and mature ( compare Figures 5-C2 and C2′ ) . This discrepancy might be explained by the fact that the mathematical framework only models angiogenesis , i . e . blood vessel growth through the creation of new vessel branches from existing ones , whereas vasculogenesis , i . e . de novo network formation from scattered endothelial cells , is not included here . Indeed , after bone fracture the hematoma will be filled with blood , containing amongst others endothelial precursor cells , which could explain the small , immature blood vessels observed experimentally . We would like to stress , however , that this is a first hypothesis that is currently being explored further . Notice the closure of the bone marrow canal by new bone on day 56 , separating the bone marrow ( right ) from the scaffold region ( left ) ( Figure 5-B3 , B3′ , B3″ ) . As such , capping of the bone ends has occurred both in the experimental and the mathematical model ( Figure 4-C ) . The blood vessels in the center are still much smaller compared to those near the edges of the defect ( compare Figure 5-C3 and C3′ ) . In the center of the defect no signs of bone formation are detected , only fibrous tissue is seen , at this time point associated with a very low cellular content ( Figure 5-B3-center ) . Also in the mathematical model no additional bone formation is predicted between post fracture day ( PFD ) 60 and 90 , thereby classifying this fracture as a non-union [1] , [2] . After this qualitative validation of the model predictions with the experimental observations of bone healing in a large defect , the model was used to understand the mechanisms underlying the occurrence of fracture non-unions . It appears that in the mathematical model , chondrogenic differentiation and cell survival are severely impaired in the central callus area ( indicated with dots in Figure 3 ) due to the harsh hypoxic conditions ( optimal oxygen tension for chondrogenic differentiation is 3% , minimal oxygen tension for MSC and chondrocyte survival is 0 . 5% , see Figure 2 ) ( Figure 4-D , F ) . Consequently , the angiogenic growth factor ( gv ) , which is the major stimulus for vascular growth and as such endochondral ossification , is not produced in the central callus area ( Figure 4-E ) . As a result , the bone healing stops after capping of the bony ends , resulting in an atrophic non-union ( Figure 4-C ) . Note that the predicted bone front extents further into the callus than observed in the in vivo model . This might be due to some limitations of the computational model . Firstly , in the current model all the progenitor cells can differentiate towards both the chondrogenic and osteogenic lineage , depending on the local growth factor concentrations and oxygen tensions . In reality , however , it has been shown that the progenitors from the endosteal callus can only differentiate towards the osteogenic lineage , resulting in the absence of cartilage in the endosteal callus [28] . Progenitor cells from the periosteum do have the capability to differentiate to both lineages , explaining why endochondral ossification mainly occurs in the periosteal callus [28] . As such , the current simplification of the model leads to an overestimation of the amount and the location of the cartilage matrix , resulting in an overestimation of the predicted bone formation . Secondly , the current model does not account for changes in callus size and shape during the healing process which may also influence the bone formation process . After establishing the in silico and in vivo non-union model , the in silico model was further used to explore the influence of the gap size on the healing outcome ( Figure 6 ) . By increasing the gap size , the bone tissue fraction at PFD 90 is reduced whereas the cartilage fraction remains similar ( close to zero ) and the fibrous tissue fraction is greatly increased ( Figure 6 ) . Although the bone tissue fraction reaches 84% in a 3 mm defect , there is no cortical bridging which indicates the formation of a non-union . The simulation therefore predicts that a murine bone defect becomes critical at 3 mm . In the remaining part of this study we will focus on the bone regeneration process in 5 mm defects , in correspondence with the in vivo set-up described above . Since for all the different gap sizes explored in Figure 6 , the same set of initial and boundary conditions was employed , the occurrence of fracture non-unions might be attributed to an inadequate vascularization of the central callus region . More specifically , the ingrowing vasculature which originates from the bony ends , needs to cover a larger distance in larger defects , resulting in a too late vascularization of the central fracture area ( Figure 4-G ) and consequently harsh hypoxic conditions ( Figure 4-F ) . As was explained above , these hypoxic conditions lead to cell death thereby arresting the production of angiogenic growth factors and ultimately the bone healing process ( Figure 4-C ) . Clearly , the spatiotemporal patterns of oxygen tension are an important determinant of successful bone repair which prompted us to investigate the complex interplay between oxygen delivery , diffusion and consumption in a critical size defect ( 5 mm ) . An extensive sensitivity analysis was performed on the parameter values describing the delivery of oxygen ( Gn ) , the diffusion of oxygen ( Dn ) and the oxygen consumption by osteoblasts ( Qb ) , chondrocytes ( Qc ) , MSCs ( Qm ) and fibroblasts ( Qf ) . Moreover , since experimental evidence has shown that the biological potential ( e . g . the amount of osteoprogenitor cells and growth factors present ) might be greatly reduced in critical size defects [26] , [27] , we also explored the influence of the initial conditions ( cm , init , gbc , init , cf , init , mf , init , ninit ) in the central callus area ( indicated with dots in Figure 3 ) on the fracture healing outcome ( Table S1 in the supplementary material ) . The initial position of the endothelial cells ( see Figure S1 in the supplementary material ) , has a small influence on the final bone tissue fraction ( +/−2% ) . This difference can be attributed to a different spatial filling of the blood vessels in the 2D simulated geometry and is in the same range as the influence of the stochastic component in the description of blood vessel migration on the simulation outcome ( +/−3% ) [24] . Based on these findings , we consider deviations of more than 2% with respect to 50% of bone tissue fraction to be sufficient to warrant further analysis . In order to gain more understanding in the complex non-linear dynamics of the oxygen model , the mechanisms underlying these significant deviations were investigated and are discussed in more detail below . The sensitivity analysis revealed a non-linear influence of the initial amount of MSCs ( cm , init ) on the bone tissue fraction at PFD 90 . This can be explained by the fact that on the one hand a low initial concentration of MSCs ( cm , init<2 . 104 cells/ml ) reduces the biological potential of the fracture site since less cells can contribute to the bone healing process . On the other hand , a high initial concentration of MSCs ( cm , init>2 . 105 cells/ml ) will worsen the detrimental hypoxic conditions in the central callus region due to the increased amount of oxygen consumption . The initial concentration of fibroblasts ( cf , init ) does not show this non-linear behavior . High initial concentrations of fibroblasts and/or MSCs are detrimental ( cf , init>5 . 105 cells/ml ) since the increased oxygen consumption will lower the average oxygen tension in the central callus area . Contrary to the MSCs , low initial concentrations of fibroblasts do not seem to have a major influence on the final amount of bone formation . This is mainly because fibroblasts do not contribute to the biological potential of the hematoma as they cannot differentiate towards the osteogenic or chondrogenic lineage . The sensitivity analysis also indicates that the amount of osteochondrogenic growth factors present in the fracture hematoma ( gbc , init ) is a critical determinant of the final amount of bone formation . Indeed , increasing the growth factor concentration results in a significant increase in the amount of bone formation measured after 90 days of healing . This result can be attributed to an increased chondrogenic differentiation which limits the oxygen consumption since chondrocytes consume less oxygen than MSCs . As such , the central hypoxic area will be reduced leading to more bone formation . After the inflammation phase , the fracture callus is filled with granulation tissue ( represented here by mf , init ) . It appears that a large amount of granulation tissue negatively influences the fracture healing outcome which is due to its inhibitory effect at large matrix densities on the proliferative capacities of MSCs , fibroblasts , chondrocytes and osteoblasts . Similar to the initial amount of MSCs also the initial oxygen tension ( ninit ) has a non-linear effect on the final amount of bone formation . Very low oxygen tensions ( ninit<0 . 5% ) lead to a larger hypoxic area and less bone formation whereas oxygen tensions above 4% ( ninit>4% ) hamper the proliferation of chondrocytes , thereby disrupting the cartilage production and consequently the endochondral ossification process . Interestingly , in the intermediate range of oxygen tensions ( 0 . 5%<ninit<4% ) , lower initial oxygen tensions appear to result in more bone formation ( Table S1 , 0 . 7% versus 3 . 7% oxygen tension of the standard compromised condition ) . Although intuitively we would expect that these low oxygen tensions would lead to worse hypoxic conditions , model analyses show that the average oxygen tension in the fracture callus remains above 0 . 8% during the entire healing period ( note that the low oxygen tensions of the central callus area are averaged with the high oxygen tensions near the bony ends ) , which is well above the oxygen threshold for chondrocyte and MSC cell death ( i . e . 0 . 5% ) . As such the oxygen tension is low enough to inhibit extensive proliferation ( as the chondrocytes and MSCs preferentially proliferate at 3% and 4% oxygen tension respectively , Figure 2 ) and therefore avoiding too much oxygen consumption , but high enough to keep a small amount of remaining stem cells alive . Moreover , the oxygen consumption is not only reduced due to the smaller amount of consuming cells . The cellular consumption of oxygen is also oxygen dependent , leading to a lower cellular consumption in low oxygen environments . It is the combination of these effects that limits the drop of the average oxygen tension , allowing the MSCs to survive and contribute to the bone healing process for a longer period of time ( 40 days for case ninit = 0 . 7% versus 4 days in the standard condition ) . A similar reasoning can be made for the case where an initial gradient of oxygen tensions was applied to the central callus region ( ninit , gr = 0 . 8%/mm*x ) . In this simulation the oxygen tension varied from 0% in the middle of the callus to 4% at the bony ends . The low oxygen tensions in the central area supported the maintenance of a small population of MSCs for a longer period of time ( 6 days versus 4 days in the standard condition ) . This resulted in a larger amount of cartilage and finally bone . Note that this specific gradient in oxygen tension is less beneficial for the amount of bone formation at PFD 90 than a uniform distribution of 0 . 7% , as in the case of the gradient the oxygen tension in the middle of the callus is too low to sustain cell viability . Besides investigating the influence of the initial conditions , the sensitivity analysis also focused on the complex interplay between oxygen delivery ( Gn ) , diffusion ( Dn ) and consumption ( Qb , Qc , Qm , Qf ) . Altering the oxygen delivery ( Gn ) by the vasculature has a large effect on the final amount of bone formation . Very low values of oxygen delivery increase cell death in the central hypoxic area , resulting in the absence of any bone formation . Increasing the value of oxygen delivery slightly improves the fracture healing outcome . Note that , although the bone tissue fraction is 37% in case of Gn = 22 . 10−12 mol/cell . day and 55% in case of Gn = 3 . 2 . 10−12 mol/cell . day , the spatial extent of bone ingrowth at PFD 90 is very similar ( results not shown ) . This is however masked by the increased proliferation and matrix production of fibroblasts who thrive in the well-oxygenated environment created by Gn = 22 . 10−12 mol/cell . day . As such , the bone tissue fraction for Gn = 22 . 10−12 mol/cell . day is reduced with respect to Gn = 3 . 2 . 10−12 mol/cell . day . The parameter values of the cell-specific oxygen consumption rates ( Qb , Qc , Qm , Qf ) also influence the outcome of the model significantly . For all cell types , it is beneficial to reduce the oxygen consumption rates since this will limit the decrease in oxygen tension in the central fracture area and consequently the amount of cell death . This benefit is greatest for the MSCs and chondrocytes as these cell types mainly populate the central fracture area and contribute to the hypoxic conditions encountered here . Conversely , the amount of bone formation is greatly reduced when the oxygen consumption rate of the MSCs ( Qm ) or chondrocytes ( Qc ) is increased . The model outcome is also negatively affected by a high osteoblastic oxygen consumption rate ( Qb ) whereas a high fibroblastic consumption rate ( Qf ) only slightly reduces the final amount of bone . In the first case , the oxygen tension near the bony ends is reduced , resulting in hampered osteogenic differentiation and limited bone formation . In the latter case , the fibroblasts reduce the oxygen tension in the entire callus area ( the fibroblasts are initially uniformly distributed in the fracture callus ) but this drop is limited due to the small amount of fibroblasts present . Interestingly , a similar reasoning does not hold for the MSCs ( although they are also initially uniformly distributed and limited in cell population ) since they mainly grow in the central fracture zone whereas the fibroblasts optimally proliferate in a well-oxygenated environment such as the tissues surrounding the bony ends . As such , a high oxygen consumption rate of MSCs severely impairs the bone formation process whereas a high oxygen consumption rate of fibroblasts only slightly reduces the amount of bone formed at PFD 90 . It can be noticed from Table S1 and Figure S2 that the diffusion properties of oxygen have a major impact on the simulation outcome . Reducing the diffusion coefficient of oxygen impairs the bone formation due to the creation of a larger hypoxic zone ( Figure S2-A , C ) . Increasing the diffusion coefficient appears to be beneficial although a closer look at these simulation results reveals that the endochondral process is not captured correctly anymore with bone formation largely preceding the ingrowth of new blood vessels ( Figure S2-D , F ) . Note that also in this case a non-union is formed , since there is no cortical bridging , even though a bone tissue fraction of 89% is reached ( Table S1 ) . Increasing the diffusion coefficient even further results in a complete absence of bone formation since the resulting oxygen tensions are too low for any cell type to survive ( Figure S2-H ) ( see Supporting Text S3 ) . In conclusion , we can state that the initial conditions have an important impact on the final amount of bone formation . They are however not sufficient to result in complete healing of critical size defects due to insufficient vascularization of the central callus area , leading to hypoxic conditions and cell death . As such , an adequate and timely restoration of the vasculature appears to be an important determinant of the healing outcome . Inspired by the importance of a timely vascularization as well as by the limited biological potential of the fracture hematoma , we explored the influence of the host environment on the bone healing process in critical size defects . It appears that the fracture healing process is intimately linked to the surrounding muscle envelope since clinical evidence has found that open fractures with significant muscle injury complicate fracture healing and are a risk factor for the development of non-unions [5]–[7] . Moreover , tibial shaft fractures , which are only covered by a thin layer of soft tissue , are prone to a number of complications often resulting in additional surgical interventions [5] , [9] , [29] . There are a number of ways by which the skeletal muscle can contribute to the bone healing process . Firstly , experimental studies have shown that blood vessels originating in the overlying muscle contribute to the vascularization of the fracture callus [30] , [31] . Secondly , muscle cells are a source of growth factors ( e . g . FGF-2 , TGF-β ) [32] as well as progenitor cells [33]–[35] . Thirdly , the muscle envelope might provide the adequate biomechanical stimuli required for successful bone healing [36] , [37] . In order to further unravel the potential mechanisms of interaction that exist between the bone regeneration process and the overlying skeletal muscle , the role of the skeletal muscle as a source for vascularization , progenitor cells and growth factors or a combination thereof was investigated by applying different boundary conditions to the in silico model ( Figure 7–8 ) . More specifically , the contribution of the muscle to the vascularization of the fracture callus was simulated by initializing additional endothelial cells on the border of the central callus area with the muscle , either partially or fully covering the fracture gap . The influence of the muscle as a source of MSCs or growth factors was represented by a Dirichlet boundary condition , applied to the upper border for the entire duration of the simulation ( i . e . 90 days ) and fully covering the fracture gap ( Figure 7 ) . The value of the Dirichlet boundary conditions is equal to the ones applied in the standard case , i . e . 2 . 104 cells/ml for the MSCs and 2 µg/ml for the osteochondrogenic growth factors [25] , [38]–[40] . Since the mathematical model does not take into account any mechanoregulatory stimuli , the influence of mechanoregulatory stimuli generated by the overlying muscle on the bone formation processes cannot be evaluated in this study . The results , summarized in Figure 8 , underline the importance of the host environment for successful fracture healing since all the investigated conditions improve the amount of bone formation with respect to the standard condition or result even in bridging of the critical size defect . Note that the host environment is also more efficient in stimulating the bone regeneration process than the initial conditions tested in Table S1 , since the host environment continuously provides the fracture callus with fresh growth factors , cells and blood vessels ( or a combination thereof ) whereas the initial conditions represent only a single ( initial ) contribution to the bone regeneration process . In order to limit the length of the paper , we will touch upon the most important findings of Figure 8 and refer the reader to Supporting Text S4 for an in depth discussion of the results . Both the contribution of the muscle as a source of vascularization ( case A ) as well as of osteoprogenitor cells ( case D ) results in the formation of a union , whereas a partial supply of blood vessels from the host environment ( case B ) or muscle derived release of osteochondrogenic growth factors fails to result in a complete bridging of the defect ( Figure S3 ) . In most cases , except for cases E and G , the combination of two or more boundary conditions enhances the bone formation process . Indeed , the combined delivery of cells and growth factors results in less bone formation ( case E ) than the delivery of cells alone ( case D ) . Figure 8 also shows that without vascular ingrowth from the muscular environment , the delivery of cells results in the largest amount of bone formation ( case C versus case D ) . However , if the fracture callus is fully or partially vascularized by the overlying muscle , the delivery of growth factors is more beneficial than the delivery of cells for the final healing outcome ( case F versus G , case I versus J ) . In conclusion , we can state that the contribution of the host environment , and more specifically its role as a source of vascularization is critical for successful bone healing . Interestingly , the results indicate that the lack of adequate vascularization can be rescued by a continuous delivery of osteoprogenitor cells , potentially in combination with osteochondrogenic growth factors . Intrigued by the results of the previous section , we wondered if the lack of adequate vascularization could also be rescued by a single contribution of ( more optimal ) initial conditions . Or , from another perspective , whether the initial conditions that were insufficient to result in successful bone healing in a compromised environment ( Table S1 ) , would be able to stimulate the bone regeneration process more in a permissive host environment . In order to answer this question , we use the model in which the fracture callus is partially supplied by blood vessels from the overlying muscle ( case B , Figure S3 ) since this environment is not as compromised as the standard compromised condition ( Table S1 ) but nevertheless results in the formation of a non-union without additional cells or growth factors ( Figure 8 , case B ) . We tested three potential treatment strategies: the injection of growth factors , the injection of cells and the injection of a combination product . All the injections take place at day zero , making them initial conditions ( Figure 9 ) . According to the results of Figure 9 , all the treatment strategies yield at least the same ( within 2% of intrinsic variability ) or more bone formation than a non-treated fracture in a permissive environment . Moreover , the permissive environment is clearly beneficial since the amount of bone formation is increased with respect to the compromised environment for all the treatment conditions ( Table S1 ) . The injection of precursor cells does not significantly improve the bone healing outcome since the vascularization of the central callus area is still delayed , resulting in hypoxic conditions and cell death ( Figure 11 ) . The injection of osteochondrogenic growth factors is able to heal the critical size defect surrounded by a muscular envelope that partially contributes to its vascularization if the concentration is sufficiently high ( Figures 9–10–11 ) . The mechanism of action underlying this result can be explained as follows . The large initial concentration of growth factors will lead to the differentiation of the osteoprogenitor cells into chondrocytes , which consume less oxygen than MSCs . Consequently , the pool of oxygen-consuming MSCs is reduced thereby limiting the oxygen consumption . As such , a large initial concentration of growth factors makes the hypoxic area shrink , finally leading to the successful healing of the critical size bone defect ( Figure 11 ) . The injection of the combination product has improved the amount of bone formation but is not as beneficial as osteochondrogenic growth factor injections alone ( Figures 9–10–11 ) . Indeed , the growth factor concentration is not high enough to commit the entire population of MSCs to the chondrogenic lineage . As such , a small amount of MSCs remains undifferentiated and can continue to proliferate and consume oxygen . Since MSCs consume more oxygen than chondrocytes , the remaining MSC pool increases the drop in oxygen tension and consequently cell death . As a result , the amount of bone formation is lower than in the case of growth factor treatment alone . The predictions of the mathematical model are compared with the results of the in vivo set-up where the influence of BMP-2 overexpressing periosteal cells on bone formation in large defects was explored . Defects , treated with a collagen scaffold containing mPDCs , thereby mimicking the initial conditions of the computational model , show little bone formation ( Figure 12-B ) , which is also predicted by the mathematical model ( Figure 9 , standard permissive condition ) . Interestingly , while defects treated with mPDCs show no presence of bone or cartilage in the center of the defect ( Figure 12-C1 ) , large amounts of bone and the presence of cartilage ( arrow ) and bone marrow are noted in the defects treated with BMP-2 overexpressing mPDCs ( Figure 12-C2 , C3 ) . Clearly , the presence of BMP-2 enhances the bone formation process which results in a clinical union . Similarly , the computational model predicts that the injection of only growth factors is sufficient to heal a large defect in a permissive environment ( Figure 9 , gbc , init ) . Note , however , that in the experimental set-up BMP-2 overexpressing cells are implanted whereas computationally an initial bolus injection of growth factors is simulated . In the experimental model , the sites of bone formation are closely associated with numerous large blood vessels ( indicated in dark brown by the CD31 staining , Figure 12-C5 , C6 ) , in contrast to the small blood vessels observed in the center of the defects treated with mPDCs only ( Figure 12-C4 ) . In the mathematical model , the blood vessel formation is also closely connected to the bone formation process ( Figure 10 ) . In the central callus area we hypothesize that the small blood vessels observed in vivo arise through vasculogenesis , which is not accounted for in the mathematical model . However , since these small blood vessels appear to be immature and not associated with bone formation , the mathematical model does predict the correct tissue distribution in the central callus area even in the predicted absence of small blood vessels . As expected , no blood vessels are observed at the site of cartilage formation ( Figure 12-C5 , arrow ) . While the influence of the amount of seeded cells alone or in combination with BMP-2 overexpression was not explored experimentally , the computational simulations predicted an improvement in the amount of bone formation but not a complete healing of a large bone defect ( Figure 9 , cm , init and cm , init/gbc , init ) . Note that in a compromised environment , a large defect will develop into a non-union , irrespective of a growth factor treatment ( Table S1 , standard compromised condition ) , whereas in an environment with full vascular ingrowth from the overlying muscle , a union will develop , irrespective of a growth factor treatment ( Figure 8 , case A ) . As such , since the computational predictions of the growth factor treatment in a permissive environment reproduce the in vivo observations correctly , one may speculate that the muscle overlying the large defect in the in vivo set-up partially contributed to the vascularization of the fracture callus and consequently the bone healing process . Further characterization of the origin of the vasculature growing towards the defect area would be required to confirm this . From the results discussed above , we can conclude that a single injection of osteochondrogenic growth factors is able to compensate for the lack of adequate vascularization . Since a single injection of cells fails to promote complete bridging of the critical size defect in a permissive environment , a sequence of cellular injections might be more appropriate strategy . Finally , we used the in silico model to optimize the treatment strategy of the previous section for critical size defects surrounded by a compromised host environment . As can be concluded from Table S1 , the lack of muscular contribution to the vascularization of the fracture callus as well as of osteoprogenitor cells or growth factors , greatly hampers the bone regeneration process and results in the formation of a non-union . Furthermore , the initial conditions can be tuned to improve the amount of bone formation but are insufficient to provide complete healing of the critical size defect ( Table S1 ) . This was attributed to the delayed vascularization of the central callus area , leading to hypoxia and cell death . In order to improve the limited biological potential of the fracture callus and host environment , additional progenitor cells or growth factors can be injected in the fracture callus . However , cellular strategies would miss their therapeutic target if injections would take place at day 0 , since cell survival would be very limited in these challenging hypoxic conditions . Therefore , we investigated whether a single injection of MSCs , osteochondrogenic growth factors or a combination thereof at a later time point would improve the bone healing outcome , as in this way the blood vessel network will have restored at least partially ( Figure 13 ) . As can be seen in Figure 13 , the injection of osteochondrogenic growth factors does not improve the bone healing outcome , except at PFD 0 . This can be attributed to the increased chondrogenic differentiation and consequently limited oxygen consumption , as was discussed previously . At the other time points , the delay in vascularization of the central callus area results in hypoxia and cell death . Consequently , the injection of additional growth factors is to no avail since there are no cells present on which they can exert their influence . Interestingly , the time at which the MSCs or the combination product was injected , appears to be a critical determinant for the final amount of bone formation . If the cellular treatment is administered before PFD 35 , the amount of bone is reduced ( compared to no treatment ) since the additional cells increase the oxygen consumption thereby worsening the hypoxic conditions in the central callus area . One can notice a further decrease of the effectiveness of the cellular treatment ( cells only as well as the combination with growth factors ) for injections at PFD 7 and 14 . This can be related to the oxygen tension encountered in the central callus area at the time of injection , and the fact that this oxygen tension evolves with time . More specifically , at PFD 0 the oxygen tension has dropped only slightly and at PFD 28 the vasculature is already growing into the fracture callus so that in both cases the oxygen tension in the central callus area is able to support the injected cells . At intermediate time points , however , the oxygen tension is too low to support the injected cells , explaining why injections at PFD 7 and 14 are the least effective . If cells are administered at PFD 35 , the delay of 35 days between the occurrence of the fracture and the start of the cellular therapy allows for a partial restoration of the blood vessel network , which seems to be optimal for the injection of cells only . The effectiveness of the combination product , however , continues to increase when the treatment is further postponed ( up to day 56 ) . The non-linearities in the predicted bone tissue fractions as a function of time of administration , as well as the discrepancy in optimal timing between the cellular and combination treatment can again be explained by the evolving oxygen tension of the central callus area which gradually increases as a function of time through a combination of oxygen release from the active vasculature and passive diffusion . More specifically , the average oxygen tension in the central callus area of a large non-treated defect surrounded by a compromised environment increases from 2 . 2% at 35 days , to 3% at 42 days , 3 . 9% at 49 days , 5 . 6% at 56 days and 6 . 2% at 63 days . At PFD 35 both the oxygen tension as well as the osteochondrogenic growth factor concentration are low in the central callus area so that only limited chondrogenic differentiation occurs upon injection of MSCs . However , the low oxygen tension inhibits extensive cellular proliferation ( avoiding too much oxygen consumption ) , resulting in a small amount of “quiescent” stem cells ( similar to ninit = 0 . 7% , Table S1 ) . When the oxygen tension in the central callus area subsequently increases to 3% , the remaining MSCs differentiate to chondrocytes and contribute to the bone regeneration process . As such , there are two bursts of chondrogenic differentiation which results in an increased amount of bone formation . If the cellular administration occurs after PFD 35 the increased oxygen tension will enhance the chondrogenic differentiation , thereby reducing or even eliminating the pool of “quiescent” stem cells . As such the proliferation and survival of the newly formed chondrocytes will mainly determine the extent of the bone formation . Since the average oxygen tension in the central callus area is higher at PFD 56 and 63 than at PFD 49 , more bone will be formed for cellular injections in the former two cases , compared to PFD 49 . By combining osteochondrogenic growth factors with stem cells , all the injected MSCs will directly become chondrocytes thereby depleting the pool of osteoprogenitor cells , irrespective of the starting time of the treatment . Similar to the cellular treatments started after PFD 35 the proliferation and survival of the chondrocytes will mainly determine the extent of the bone formation . It appears that the average oxygen tension at PFD 56 results in an optimal proliferation and survival of the chondrocytes and hence subsequent endochondral bone formation . At PFD 63 the average oxygen tension becomes too high for optimal chondrocytic proliferation thereby reducing the amount of bone formation . According to the model results , cellular injections are only effective if delayed until a specific time point ( i . e . day 35 , Figure 13 ) in order to allow for a partial restoration of the blood vessel network . Note that although the cellular as well as the combination treatment lead to an increased amount of bone compared to no treatment ( provided the injection is sufficiently delayed ) , they nevertheless result in the formation of a non-union ( with a maximal amount of bone at 90 days up to 80% and 85% respectively ) . As such , a single injection is insufficient and future research should focus on an optimal sequence of injections in order to heal critical sized defects in a compromised environments .
This study has used an integrative in vivo - in silico approach to investigate the occurrence of oligotrophic and atrophic non-unions as well as to design possible treatment strategies thereof . An extensive sensitivity analysis was performed in order to study the complex interplay of blood vessel formation , oxygen supply , growth factors and ( osteoprogenitor ) cells on the final healing outcome in large bone defects . The results of the sensitivity analysis indicated that the initial conditions ( osteochondrogenic growth factor , MSCs , oxygen ) are necessary for the bone regeneration process but not sufficient for complete bone healing of a critical size defect ( 5 mm ) . They do , however , have an important impact on the final amount of bone formation . Interestingly , simulation results of the same oxygen model in a small defect ( 0 . 5 mm ) were found to be robust to changes in the initial conditions [23] . Although the performed sensitivity analysis yields interesting results , the interpretation thereof should be done carefully due to a number of reasons . For example , the sensitivity analysis that was performed in this study ( Table S1 ) used a one-at-a-time ( OAT ) design , where the effect of one factor is assessed by varying the value of only that factor and keeping all other factors fixed . The main disadvantage of this simple method is its inability to capture interactions between factors . A simple combination of the ‘optimal’ values of initial conditions ( see Table S1: the value that for an OAT design yielded the most bone at PFD 90 ) indicates for instance that a more adequate design is necessary to unravel these ( non-linear ) interactions between the different parameters of the oxygen model . Indeed , combining the ‘optimal’ initial conditions of MSCs ( cm , init ) , fibroblasts ( cf , init ) , osteochondrogenic growth factors ( gbc , init ) and oxygen ( ninit ) results in 38% of bone after 90 days which is less than for the respective ‘optimal’ initial conditions alone . The conclusions of the sensitivity analysis are also only valid for this specific set of parameter values since , for example , the optimal initial oxygen tension will vary depending on the initial stem cell concentration . Despite its limitations , the OAT-design already indicates some interesting non-linear responses of the model with respect to the initial MSC cell density and the oxygen tension as well as their interactions . Future work should focus on more complex designs , including latin hypercube design and uniform design [41] , to calculate quantitative metrics of sensitivity and study these non-linearities and ( higher-order ) parameter interactions further in order to unravel the underlying mechanisms and define new research hypotheses . The dynamics of all cellular variables ( apart from the endothelial cells ) is described by means of continuum equations , meaning amongst others that cell proliferation was captured by means of a logistic growth equation . While this equation accounts for a maximal cell density in the callus area , it does not allow to specify an upper limit to the number of division cycles a cell ( such as an MSC ) can undergo before senescence . Because of this upper limit , in reality the amount of cells that can be obtained through division is dependent on the original pool size whereas in the mathematical model , a single cell can theoretically divide until the entire callus reaches maximal cell density . The main consequence of this limitation is that our predictions might be too optimistic in that fracture healing might be even more challenging in reality , because a sufficient number of cells ( such as MSCs ) cannot be reached to heal the fracture . In the future we will try to implement a description that allows to account for a limited number of population doublings , potentially through the extension of the agent-based description of endothelial cells to the skeletal cell types . Even though the predictions of the current model might be too optimistic , all of the conditions explored in Table S1 nevertheless resulted in the formation of a non-union . Indeed , the simulation predicts that a murine bone defect becomes critical at 3 mm ( Figure 6 ) which corresponds to the experimental observation of Zwingenberger et al . [42] . They report the creation of a persisting femoral bone defect in nude mice when the defect size is 3 mm [42] . The predicted value is also in the same range as other mouse femoral critical defect sizes reported in the literature: 2 mm [43] , 3 . 5 mm [44] and 4 mm [45] . As such , the computational framework is able to model the occurrence of non-unions and can be used to design several treatment strategies depending on the host environment . In our model , a single initial ( i . e . at PFD 0 ) injection of osteochondrogenic growth factors at sufficiently high concentration ( gbc , init = 1 µg/ml ) directly into a callus surrounded by a permissive environment resulted in complete healing of the critical size defect ( Figure 9 ) . The beneficial effect of growth factor delivery was also confirmed by the study of Patel et al . [46] . They report that the BMP-2 release from gelatin microparticles incorporated within the pores of a scaffold that was implanted within a 8 mm rat cranial critical defect resulted in significantly higher bone formation after 12 weeks , i . e . 37 . 4±18 . 8% ( test ) versus 7 . 8±7 . 1% ( control ) bone volume respectively . Similar conclusions were made by Willett et al . who studied the influence of recombinant human BMP-2 ( rhBMP-2 ) delivery on tissue regeneration in a murine composite injury model [47] . The in vivo composite injury model consisted of a critically sized femoral bone defect and an adjacent volumetric muscle injury in the quadriceps ( both 8 mm ) [47] . They have shown that treated bone defects without volumetric muscle loss were consistently bridged whereas the treatment failed to promote the regeneration process in the challenging composite injury [47] . Although care must be taken when directly comparing these findings to our in silico results ( since the exact role of the muscle in the in vivo setting of Willett et al . was not characterized ) , they do predict the same trends . Indeed , the multiscale model predicts a successful healing in the case of growth factor administration to a critical sized defect that is fully or partially supplied by blood vessels from the overlying muscle ( Figure 9 ) . In contrast , in a compromised environment where the role of the muscle as a source of vascularization is lacking , additional injections of growth factors , either at PFD 0 ( Table S1 ) or at later time points ( Figure 13 ) do not induce bony bridging of the large bone defect . In large bone defects not only the initial concentration of growth factors but also the initial amount of osteoprogenitor cells might be reduced [26] , [27] . Consequently , the use of stem cells for the treatment of critical size defects is actively being pursued [48] . The injection of MSCs in the callus area elicited an improved healing response ( although without reaching full bridging ) in silico if the environment is sufficiently vascularized to sustain the cell viability , which according to the model meant that injections were only effective if delayed until a certain time point ( day 35 according to Figure 13 ) . Similar conclusions were drawn by Geris et al . who investigated the occurrence of bone atrophic non-unions by an integrative approach [49] . Based on the recovery of the blood supply to the interfragmentary gap , they predicted with an in silico model that the injection of MSCs at three weeks post-osteotomy would prevent the onset of an atrophic non-union which was also confirmed by experimental results [49] . The necessity of vascularization for successful healing of challenging critical size defects is also substantiated by the results of Table S1 ( no contribution of the overlying muscle to the vasculature ) and Figure 9 ( partial contribution of the overlying muscle to the vasculature ) , where an initial injection ( i . e . at PFD 0 ) of additional cells in a defect that is insufficiently vascularized does not significantly improve the bone formation outcome . As such , the mathematical model retrieves the beneficial effect of cellular injections in some cases , similar to the experimental observations reported in literature [50] , [51] , although the effectiveness is strongly dependent on the available vasculature . Interestingly , the model results indicate that the effectiveness of a therapy ( consisting of the injection of cells , growth factors or a combination thereof ) is dependent on the timing of the treatment as well as the host environment . The former effect is strongly related to the biological potential of the fracture callus at the time the treatment is applied , while the latter potentially constitutes a source of additional osteoprogenitor cells , growth factors or vascularization . For example , growth factor injections at PFD 0 or at later time points in a compromised host environment lead to only 63% and 52% of bone respectively whereas growth factor injections at PFD 0 in a permissive environment result in the formation of a union . In all three cases the main cause underlying the formation of a non-union in a large defect ( without treatment ) is the increased cell death in the central ( hypoxic ) callus area . Since growth factor injections at PFD 0 result in increased chondrogenic differentiation , which in turn limits the oxygen consumption and the decrease of oxygen tension ( severity of hypoxia ) , this treatment increases the amount of bone formation . Note that nevertheless a hypoxic area arises which results in the formation of a non-union . Consequently , a permissive environment that provides additional vascular ingrowth , improves the bone formation outcome even further . Growth factor injections at later time points in a compromised environment are , however , to no avail since there are no cells left in the central callus area . In summary , we can state that a treatment will be most beneficial if it tackles the underlying mechanism of action causing the hampered bone formation . Although this statement seems a logical and intuitive design rule , the underlying mechanisms of actions are a result of the complex non-linear , oxygen-dependent dynamics of blood vessel formation , oxygen supply , angiogenic growth factor production , cell differentiation , cell proliferation and oxygen consumption . The fact that many cellular processes , like survival , proliferation and differentiation are ( non-linearly ) dependent on oxygen tension and that they all have a specific range of oxygen tension at which they are ‘optimized’ ( maximally affected ) ( Figures 1–2 ) , makes it virtually impossible to intuitively predict the resulting bone healing outcomes . Instead , it requires a rigorous computational modelling of the governing mechanisms and dependencies ( Figure 14 ) . Taken all the results together , we can conclude that complete cortical bridging of a challenging critical size defect will only occur if growth factors , osteoprogenitor cells and vasculature are present at the same time and place ( Figure 14 ) . Indeed , the blood vessels will supply the necessary oxygen to ensure cellular survival whereas the growth factors will promote the correct differentiation cascade finally resulting in the continuation and successful completion of the bone regeneration process . Consequently , the most stringent factor that is lacking in a certain area or at a certain time point will be an ideal candidate for potential treatment strategies . For example , bone tissue engineering treatments where a scaffold seeded with cells and osteochondrogenic growth factors is implanted in a bone defect , should focus on a timely vascularization in order to ensure the survival of the implanted cells . Potential strategies of vascularization include the induction of a Masquelet-membrane [52] , [53] , the delivery of angiogenic growth factors [46] as well as the in vitro creation of a pre-vascularized construct by co-culture of osteoprogenitor cells with endothelial cells [54] . Encouraging results were for example obtained by Patel et al . who showed that the dual release of vascular endothelial growth factor ( VEGF ) and bone morphogenetic protein-2 ( BMP-2 ) in a 8 mm rat cranial critical size defect enhanced the bone formation at 4 weeks , suggesting a synergistic effect of these growth factors during early bone regeneration [46] . Note that besides the biological stimuli also mechanoregulatory stimuli influence the bone formation process [36] , [37] . The current multiscale model does not take this into account , meaning amongst others one assumes that the fracture is sufficiently stabilized through external or internal fixation such that excessive loading will not play a role in the formation of a non-union ( Supporting Text S2 ) . In conclusion , the multiscale oxygen model was able to capture the essential aspects of in vivo atrophic and oligotrophic non-unions . Interestingly , thorough model analyses assisted in understanding the underlying mechanisms of action , i . e . the delayed vascularization of the central callus region resulted in harsh hypoxic conditions , cell death and finally disrupted bone healing . Since a timely vascularization was found to be critical for the successful healing of large bone defects , the oxygen model was used to design and test potential treatment strategies for both permissive and compromised host environments . A qualitative correspondence between the predicted outcomes of certain treatment strategies and experimental observations was obtained , clearly illustrating the model's potential . Furthermore , the results of this study demonstrate that due to the complex non-linear , oxygen-dependent dynamics of blood vessel formation , oxygen supply , angiogenic growth factor production , cell differentiation , cell proliferation and oxygen consumption , it becomes virtually impossible to determine the effectiveness of a treatment strategy intuitively thereby underlining the importance computational modelling tools . Moreover , the model predictions also showed that the effectiveness of a therapy is strongly influenced by the host environment since it can serve as a source of additional osteoprogenitor cells , growth factors or vascularization to populate the fracture callus and increase the biological potential thereof . Consequently , future research should focus on extensive experimental characterization as well as computational modelling of the host environment and its interaction with potential treatment strategies . | In 5–10% of fracture patients , the bone fractures do not heal in the normal healing period ( delayed healing ) or do not heal at all ( non-union ) . In order to investigate the causes of impaired healing and design potential treatment strategies , we have used a combined experimental and computational approach . More specifically , large bone defects were analyzed in mouse models and simulated by a previously published computational model . After showing that the predictions of the computational model match the observations of the experimental model , we have used the computational model to investigate the underlying mechanisms of action . In particular , the results indicated that the new blood vessels do not reach the central fracture zone in time due to the large defect size , which leads to insufficient oxygen delivery , increased cell death and disrupted bone healing . The healing , however , could be rescued by adequate blood vessel ingrowth from the overlying soft tissues . Moreover , potential treatment strategies were designed based on the influence of these soft tissues . In conclusion , this study demonstrates the potential of a combined experimental and computational approach to contribute to the understanding of pathological processes like the impaired bone regeneration in large bone defects and design future treatments thereof . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"theoretical",
"biology",
"biology",
"and",
"life",
"sciences",
"computational",
"biology"
] | 2014 | Size Does Matter: An Integrative In Vivo-In Silico Approach for the Treatment of Critical Size Bone Defects |
Q fever is an endemic disease in different parts of Iran . This study aimed to investigate the prevalence of acute Q fever disease among at-risk individuals in northwestern Iran . An etiological study was carried out in 2013 in Tabriz County . A total of 116 individuals who were in contact with livestock and had a nonspecific febrile illness were enrolled in the study . IgG phase II antibodies against Coxiella burnetii were detected using ELISA . The prevalence of acute Q fever was 13 . 8% ( 95% confidence interval [CI]: 8 . 0 , 21 . 0% ) . Headache ( 87 . 5% ) and fatigue and weakness ( 81 . 3% ) were the dominant clinical characteristics among patients whit acute Q fever . Acute lower respiratory tract infection and chills were poorly associated with acute Q fever . Furthermore , 32% ( 95% CI: 24 , 41% ) of participants had a history of previous exposure to Q fever agent ( past infection ) . Consumption of unpasteurized dairy products was a weak risk factor for previous exposure to C . burnetii . This study identified patients with acute Q fever in northwestern of Iran . The evidence from this study and previous studies conducted in different regions of Iran support this fact that Q fever is one of the important endemic zoonotic diseases in Iran and needs due attention by clinical physicians and health care system .
Q fever is a zoonotic contagious disease caused by an intracellular gram-negative bacterium called Coxiella burnetii [1] . Q fever is mostly asymptomatic in livestock and animals , except in some cases , where it causes abortion or stillbirth . Infected animals shed this bacterium in their milk , faeces , urine and especially in birth products [2] . Inhalation of infectious aerosol particles constitutes the major route of acquiring the disease in humans , so inhalation of only one single C . burnetii can cause illness in humans [3] . Nevertheless , other routes of transmission of this infection to human are consumption of contaminated milks and dairy products , tick bites and transmission from a person to person [4] . Domestic ruminants ( including cattle , sheep and goats ) are the most important reservoirs of C . burnetii in the nature . However , transmission of the infection to human by dogs , cats , rabbits , birds , reptiles and arthropods , especially ticks and mites , has also been reported [5 , 6] . Clinical manifestations of Q fever in humans include asymptomatic , acute , and chronic to fatigue syndromes . Almost 60% of the infected people may not show any clinical symptoms . Acute Q fever is defined as a primary infection with C . burnetii [5 , 7] . The most frequent clinical manifestation of acute Q fever is a flu-like and self-limited illness , and the major clinical presentations of these patients are high and prolonged fever , severe headache , coughing , atypical pneumonia , hepatitis , myalgia , arthralgia , cardiac involvement , skin rash and neurologic signs [2 , 8] . The case fatality rate of acute Q fever is reported 1–2% [4 , 5] . Chronic Q fever is a disease occurring in less than 5% of acutely infected patients . It may occur several months , years , or even decades after the onset of the acute infection . This form of the disease can occur after infection with or without symptoms . Chronic Q fever is accompanied by symptoms such as endocarditis , vasculitis , prosthetic joint arthritis , osteoarticular infection and lymphadenitis . [7 , 9] . Endocarditis and vascular infection caused by Q fever are fatal if untreated [10] . Q fever is mainly diagnosed based on serological tests and antibody patterns that are different between acute , convalescent and chronic forms of the disease . There are two distinct antigenic phases to which humans develop antibody responses . Acute or chronic form of Q fever is diagnosed based on the dominant type of antibodies in response to antigens of phase I and II [11] . In acute Q fever infection , antibodies against phase II antigens are predominant , whereas phase I antibody titers are more prevalent in cases of chronic Q fever [10 , 12 , 13] . In Iran , the first clinical cases of acute Q fever were reported in 1952 , including two patients with symptoms of severe fever and neurological signs in Abadan city , southwest Iran [14] . Furthermore , in 1970 , four acute Q fever patients with pneumonic illness were reported from Shiraz , southern Iran [15] . Forty nine patients with acute Q fever were reported from Abadan city during 1970 to 1973 [16] . From 1972 to 1976 , 80 patients with acute Q fever were diagnosed , among them three cases had pleuropericarditis lesions [17] . From 1976 , the disease was neglected in Iran , and no human case was reported . In 2009 , C . burnetii antibodies were reported in febrile patients in Kerman Province , southeastern Iran , and investigation on Q fever was resumed [18] . Afterwards , various seroepidemiological studies were conducted on animal and human population [19–23] . The first patient with Q fever endocarditis was reported in 2013 in Tehran [24] . Studies conducted in Iran emphasize that Q fever is an endemic disease in different parts of the Iran [25] . Since a few studies have been conducted to identify patients with acute Q fever in Iran , present study aimed to investigate the prevalence of acute Q fever among at-risk individuals in northwestern Iran .
Tabriz is one of the major cities in Iran and the capital of East Azerbaijan Province located in northwestern Iran . This city , 237 square kilometers , is the third-largest city in Iran . The population of Tabriz and its suburb is approximately 1 . 8 million . East Azerbaijan Province is among the top five provinces in Iran in terms of production of dairy products with 10 million animal units . Tabriz has a semi-arid climate with hot summers and cold winters . Due to its mountainous climate , keeping and breeding livestock , especially sheep and goats are very common in this province . Dairy sheep and goats have an important role in dairy industry , and their milk is usually used to produce various traditional cheeses such as Lighvan cheese . The study was carried out in 2013 in Tabriz County . Individuals ( 1 ) with high-risk occupation exposed to animals or livestock products or samples of human patients ( veterinarians , farmers , butchers and laboratory personals ) , ( 2 ) people living in areas close to livestock whereabouts , or ( 3 ) people with a history of keeping animals ( including livestock and pets ) in the previous two months , were enrolled in the study provided that they had a nonspecific febrile illness ( fever above 38°C accompanied by symptoms such as fatigue , myalgia , chills , headaches , atypical pneumonia and dyspnea ) . Sampling was done randomly among patients who referred to the Central Laboratory of East Azerbaijan Province ( in Tabriz city ) and had the above criteria . The ethical committee of the Pasteur Institute of Iran approved the consent procedure , the proposal and the protocol of this study , covering all the samples ( blood ) , questionnaire and verbal informed consent as most participants were either illiterate or had a primary education . After obtaining informed consent from participants , researcher-developed questionnaire including demographic characteristics and Q fever risk factors were collected from each person by a researcher-developed questionnaire . Then , 6-ml blood sample was taken from each patient , and a second blood sample was taken after 4 weeks . Blood samples were centrifuged for 10 minutes at 3000 rpm and were kept at -20°C after extraction of their sera . Sera samples were transferred to the national reference laboratory of Plague , Tularemia and Q fever ( Research Centre for Emerging and Reemerging Infectious Diseases , Pasteur Institute of Iran ) . IgG phase II antibodies against C . burnetii were detected using a commercial quantitative enzyme-linked immunosorbent assay ( ELISA ) kit ( Serion ELISA classic , Institut Virion/Serion GmbH , Würzburg , Germany ) and according to the manufacturer's instructions . Paired sera samples of each patient were tested simultaneously . Dilution protocols were used according to the manufacturer's instructions , using a 1:500 dilution for the IgG phase II assay . The plates were read at 405 nm using a microplate reader ( ELx808 , BioTek Instruments Inc . , USA ) . Obtained ODs were analyzed according to the Virion/Serion protocol and IgG phase II was quantitatively reported . IgG phase II extinctions were expressed in U/ml titer using a logistic-log-model calculation and were defined as positive when the titer was >30 U/ml , as borderline when the titer was 20–30 U/ml and as negative when the titer was < 20 U/ml . The laboratory diagnosis of acute infection by C . burnetii was made based on any of the following serological criteria: ( 1 ) seroconversion , defined as the appearance of specific antibodies against the phase II antigens of C . burnetii at a titer of at least 30 U/ml in the convalescent phase ( whereas the serum antibody titer was negative at the initial acute phase ) , and ( 2 ) a fourfold increase in serum antibody titer between the acute phase and the convalescent phase ( in two blood samples obtained 4 weeks apart ) . If the primary and the secondary serum titers were positive , and the fourfold rise was not observed in antibody titers , it was considered as a previous history of exposure to Q fever ( past infection ) . Statistical analysis was performed using the STATA software version 11 . Descriptive data were reported in numbers ( percentage ) . Chi-square test was used to assess the association among the categorical variables . P-values less than 0 . 05 were considered statistically significant , and P-values between 0 . 05 and 0 . 1 were considered as borderline significant .
A total of 140 patients were initially enrolled; the second blood sample was taken from 116 patients ( 82 . 9% ) . Mean ( SD ) age of the patients was 39 . 3 ( 18 . 05 ) years; 61 . 4% of them were male . 28 . 6% were urban residents and 47% had high-risk occupations for Q fever infection; 81 . 2% of patients had a history of domestic animal keeping , and 93 . 2% lived near animal shelters . Histories of unpasteurized dairy product's consumption , abortion ( among females ) , and tick bites were reported in 62 . 3% , 12 . 2% , and 4 . 7% , respectively . The prevalence of acute Q fever was 13 . 8% ( 95% CI: 8 . 0 , 21 . 0% ) . In terms of acute Q fever infection , there was no statistically significant difference between people were exposed to risk factors ( including gender , age , living location , history of residence in nearby animal shelters , previous consumption of unpasteurized dairy products , history of abortion ( in women ) and history of ticks bites ) and those were not ( Table 1 ) . Acute lower respiratory tract infection ( 26 . 3% ) and chills ( 20 . 4% ) were associated with acute Q fever ( borderline significance: P = 0 . 08 ) . There was no significant statistical association between acute Q fever infection and other clinical symptoms ( headache in the past two weeks , cough , fatigue and weakness , diarrhea , myalgia , arthralgia , chest pain , atypical pneumonia , dyspnea and hepatitis ) ( Table 2 ) . Headache ( 87 . 5% ) and fatigue and weakness ( 81 . 3% ) were the dominant clinical characteristics among the acute Q fever patients ( Fig 1 ) . The levels of IgG phase II antibody in patients with acute Q fever are shown in Table 3 . After excluding acute cases from all participants , 32% ( 95% CI: 24 , 41% ) of participants had a history of previous exposure to Q fever ( past infection ) . Consumption of unpasteurized dairy products ( 40 . 6% ) was a risk factor for previous exposure to C . burnetii ( borderline significance: P = 0 . 07 ) . None of other risk factors were associated with previous exposure to C . burnetii ( Table 1 ) .
This study was conducted to identify patients with acute Q fever in northwestern Iran . One hundred and forty suspected patients were enrolled in the study , 116 of whom were assessed for acute Q fever . The prevalence of acute Q fever was 13 . 9% among the suspected febrile patients . It was also shown that 32% of the participants had serological evidence of previous infection ( past infection ) to Q fever . In a similar study in Zahedan city ( southeastern Iran ) in 2011 , 35 . 2% of 105 suspected feverish patients were diagnosed acute Q fever [26] , which is much higher than the current report . Furthermore , in another study conducted among febrile patients suspected to have brucellosis in Kerman Province ( southern Iran ) , 36% had phase II IgG antibody of Q fever [18] . In a study conducted in 2012 in Ardebil Province ( northwestern Iran neighboring East Azerbaijan Province ) , remarkable seroprevalence of Q fever ( 33 . 6% ) was observed among sheep [19] . The evidences from this study and previous studies conducted in different regions of Iran support the fact that Q fever is a prominent endemic zoonotic disease in Iran and needs more attention by physicians and health care system . In similar studies conducted in France , Denmark , Mali and Croatia 2 . 1% ( of 179794 ) , 2 . 3% ( of 1613 ) , 3 . 9% ( of 165 ) and 5 . 8% ( of 552 ) of febrile suspected patients were diagnosed to have acute Q fever , respectively [27–30] . The rate of infection in this study was reported much higher than the so-called as in the current study , patients with epidemiological risk factors ( having high-risk occupation or living in areas close to livestock or having a history of keeping animals ) and clinical risk factors ( having a nonspecific febrile illness ) were enrolled as the cases and which increased the chance of finding acute Q fever patients . In one study , 1985 to 2009 in France , 3723 ( 2 . 1% ) of 179 , 794 suspected patients with Q fever were diagnosed with acute Q fever , and the number of diagnosed Q fever patients was ascending over the years [28] . It could be due to the thereby improvement and development of diagnostic tests of Q fever as well as increasing attention of the physicians and the health care system to Q fever in France . Acute Q fever was not found significantly associated with any of the studied risk factors . In a study in southeastern Iran , contacts with domestic animals and consumption of unpasteurized dairy products were identified as risk factors [26] . In two studies conducted in Australia , France and Croatia , age and gender were reported as risk factors for acute Q fever [28 , 30 , 31] . In Mali , gender , consumptions of unpasteurized dairy products and contact with newborn animals were risk factors [29] . Probably small number of patients and small number of diagnosed acute Q fever cases compared with other study are the causes for lack of finding risk factors . A larger number of suspected patients are recommended to be evaluated in future studies . Furthermore , if there had been a data bank of patients with Q fever , it would have contributed to the achievement of risk factors of Q fever infections as well as the incidence of the disease in Iran . In this study , chills and acute lower respiratory tract infection were poorly associated with acute Q fever infection . In a study from 1983 to 1999 in Spain and in a study from 2004 to 2007 in Taiwan , the most frequent clinical pictures of Q fever patients were fever with chills [32 , 33] . During an outbreak of acute Q fever in the Netherlands , acute lower respiratory tract infection was the most frequent symptoms among the patients [34] . The findings showed that headache ( 87 . 5% ) , fatigue and weakness ( 81 . 3% ) , arthralgia ( 75% ) , myalgia ( 68 . 8% ) , chills ( 62% ) , chest pain ( 56 . 3% ) and dyspnea ( 43 . 8% ) were the most prevalent clinical symptoms in patients with acute Q fever . In the similar study in southeastern Iran , major clinical symptoms in patients with acute Q fever were fever ( 100% ) , myalgia ( 59 . 4% ) , headaches ( 43 . 2% ) , and arthralgia ( 37 . 8 ) [26] . Fever ( 69% ) and headaches ( 52% ) were the most prevalent clinical symptoms in patients with acute Q fever in Mali [29] . In Portugal , the major clinical symptoms were fever ( 100% ) , myalgia ( 68 . 8% ) , headache ( 62 . 5% ) , weakness ( 56 . 3% ) , sweating ( 53 . 1% ) and chills ( 43 . 8% ) [35] . In South Korea , fever ( 89 . 3% ) , myalgia ( 67 . 9% ) , weakness ( 53 . 6% ) and chills ( 50% ) were reported as the most prevalent symptoms [36] . In Spain , the most common symptoms of acute Q fever were headache ( 58 . 5% ) , hepatitis ( 49 . 2% ) , arthromyalgia ( 37 . 7% ) , fever ( 31 . 7% ) and pneumonia ( 19 . 1% ) [37] . In Taiwan , fever ( 99% ) , relative bradycardia ( 73% ) , chills ( 69% ) and headaches ( 45% ) were the most common clinical symptoms [33] . In Tunisia , the highest clinical signs and symptoms in patients with acute Q fever were fever ( 100% ) , fatigue ( 76% ) , hepatitis ( 71 . 5% ) , chills ( 47 . 5% ) , headache ( 42 . 8% ) and sweating ( 33 . 3% ) [38] . Comparing the clinical symptoms obtained in the present study with the so-called studies , the most non-specific symptoms for acute Q fever include fever , headache , chills , myalgia , arthralgia , weakness and fatigue and there are little differences in clinical symptoms . Although , some clinical signs such as hepatitis have a highlighted role in some countries[37 , 38] , they were not observed in the current our study and some other studies . If this disease is a part of the health surveillance system in Iran and also if more clinical cases of Q fever are diagnosed and recorded , a much better view of clinical symptoms of Q fever can be obtained patients data . In this study , 32% of participants showed evidence of previous infection ( past infection ) to Q fever; this rate was less than one in Zahedan city ( 34 . 3% ) [26] and Kerman Province ( 36% ) [18] . In similar studies , 35 . 5% in Mali [29] , 21 . 74% in Croatia [30] and 8 . 7% [27] in Denmark were previously infected with Q fever . In a systematic review conducted in Africa , human seroprevalence was <8% with the exception of studies among children and in Egypt , it was 10–32% [39] . In the seroepidemiological studies of Q fever among various populations in Iran , different rates have been also reported; seroprevalence of Q fever was reported 22 . 5% in the Sistan va Baluchestan Province among butchers and slaughterhouse workers [40] , 27 . 8% in the Kurdistan Province ( western Iran ) among butchers , slaughterhouse workers , hunters , health care workers , and patients who referred to medical laboratory [21] and 68% in Kerman Province among slaughterhouse workers [22] . Seroprevalence of Q fever was reported 12 . 8% in Northern Ireland [41] and 3 . 1% in USA [42] , compared with the lower compared with the studies carried out in Iran . The consumption of unpasteurized dairy products was a weak risk factor for previous exposure to C . burnetii . C . burnetii were isolated from livestock milks ( bovine , ovine , caprine , and camel ) in the different parts of Iran [43–46] . Therefore , the risk factors must be considered seriously . Among various diagnostic methods of Q fever , serological methods are the gold standard method for diagnosis . There are a variety of serological methods for the detection of C . burnetii antibodies such as complement fixation assay ( CFA ) , ELISA and indirect immunofluorescence assay ( IFA ) . Although the reference method for serological diagnosis of Q fever is IFA , ELISA was used for serological diagnosis of Q fever due to lack of access to IFA . In a very comprehensive recent study , different serological methods and commercial kits were assessed in 10 different laboratories in the Netherlands and three international reference laboratories in three countries ( USA , France and Australia ) . It was shown that IFA , ELISA , and CFA are all suitable serodiagnostic assays to diagnose acute Q fever . Sensitivity and specificity of the ELISA methods in the detection and diagnosis of IgG phase II antibody compared with reference method ( IFA ) were 100 and 100 , respectively . Therefore , ELISA can be used as an alternative method for the diagnosis and screening of Q fever particularly in cases of acute Q fever [47] . A larger number of patients are recommended to be studied for better understanding of Q-fever risk factors . Also , molecular diagnostic techniques are recommended to be used along with serological methods . | Q fever is a zoonotic contagious disease caused by a bacterium called Coxiella burnetii . It is mostly asymptomatic in livestock and animals . Clinical manifestations of Q fever in humans includes asymptomatic , acute and chronic to fatigue syndrome . The most frequent clinical manifestation of acute Q fever is a flu-like and self-limited illness , and clinical presentations of these patients are highly variable and extensive . Chronic Q fever is accompanied by symptoms such as endocarditis , vasculitis , prosthetic joint arthritis , osteoarticular infection and lymphadenitis . Studies conducted in Iran emphasize that Q fever is an endemic disease in different parts of Iran . Since few studies have been conducted to identify acute Q fever patients with in Iran , the present study set to investigate the prevalence of acute Q fever among at-risk individuals in northwestern Iran . The prevalence of acute Q fever was 14 . 0% among 116 suspected febrile patients . It was also shown that 32% of the participants had serological evidence of previous infection ( past infection ) with Q fever . The findings showed that most clinical symptoms in patients with acute Q fever were fever , headache , fatigue and weakness , arthralgia , myalgia , chills , chest pain and dyspnea , respectively . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"livestock",
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"fatigue",
"pathogens",
"geographical",
"locations",
"microbiology",
"headaches",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"imm... | 2017 | Acute Q fever in febrile patients in northwestern of Iran |
In a previous work by Alvarez-Martinez et al . ( 2011 ) , the authors pointed out some fallacies in the mainstream interpretation of the prion amyloid formation . It appeared necessary to propose an original hypothesis able to reconcile the in vitro data with the predictions of a mathematical model describing the problem . Here , a model is developed accordingly with the hypothesis that an intermediate on-pathway leads to the conformation of the prion protein into an amyloid competent isoform thanks to a structure , called micelles , formed from hydrodynamic interaction . The authors also compare data to the prediction of their model and propose a new hypothesis for the formation of infectious prion amyloids .
Transmissible spongiform encephalopathies , or prion diseases , are a group of fatal neurodegenerative disorders of humans and animals . The pathogenic process is typically associated with conformational conversion of a cellular protein , called prion or PrPC , to a misfolded isoform , called PrPSc . The “protein-only” model asserts that this rogue PrPSc represents the infectious prion agent , self-propagating by binding PrPC and inducing its conversion to the abnormal PrPSc [1] , [2] . This scenario was quantitatively described as a nucleation-dependent amyloid polymerization [3] , [4] . It is now generally accepted that the prion development process results from an amyloid polymerization after an initial nucleus formation in the very early phase of protein aggregation . Models based on nucleation-dependent polymerization [5]–[7] describe a molecular mechanism at the origin of the formation of large protein aggregates , by involving thermodynamically unfavourable steps that become favourable when the nucleation kinetic barrier is reached . As a striking consequence of these models , the unfavourable first steps can be bypassed by seeding with preformed polymers . However , due to the transient nature of the initial nucleus , our understanding of the interactions that form this initial structure is very sparse and thus the understanding of species allowing the prion proteins to overcome the strong kinetic barrier to form a specific amyloid conformation is highly limited [8] . This may have led to one of the most notable persisting fallacy claiming that the Lag phase of prion proliferation , defined as the required phase for the nucleus formation [3] , [9] , reflects the unfavourable nucleation phase . This idea was challenged by experimental results obtained by numerous authors who revealed a linear dependence of the lag time ( denoted by ) to monomer concentration not exceeding a nucleus size of about 2 monomers [10] , [11] . This result , also found for some other spontaneous amyloid-forming proteins [12] , was generally attributed to an accumulation of large off-pathway species whose formation is competitive with the on-pathway processes that lead to amyloid [11] , [13] . However , in the case of hamster rPrP ( Recombinant Syrian Hamster Prion Protein 90–231 ) polymerized in vitro , we previously found no kinetic evidence for an off-pathway [10] . Consequently , we propose that an additional on-pathway step is necessary to explain the results observed . We hypothesize that this new stage stands very likely for a first step . It would occur before nucleation , because experimentally , we were able to show that seeded polymerization always begins after a time delay that can be interpreted as the time needed to generate active monomers [10] . We show indeed that micelles are formed , which leads to an amyloid competent isoform of the prion protein ( denoted by PrP* ) considered as a necessary step to induce nucleation and amyloid polymerization . This hypothesis is supported by [14] , [15] where it is found that lipid interactions play a key role in the conformation of PrPC into PrPSc by β-sheet enrichment , see Sec . Discussion . We assume here that a similar interaction occurs in our experimental condition by pure PrPC interaction with the help of micelles formation . To analyse the consequences of this hypothesis , we develop a quantitative model with an explicit description of the microscopic processes , and we compare experimental data with the results predicted by the model .
In order to find out what types of structures could be involved as an on-pathway , we have performed a time dependent electron microscopic analysis during polymerization of hamster rPrP . Few minutes after dilution into polymerization buffer , we observed spherical structures looking like rigid micelles ( Fig . 1A and Fig . S6 in Supporting Information ) . The size distribution fits well with a log normal distribution ( Fig . 1B ) . The size of these spheres was heterogeneous with a mean around 30 nm . A sphere of 30 nm diameter reaches a surface area of around 30 nm2 , the rPrP have a diameter ( thanks to Protein Workshop v1 . 0 and PDB ID 1B10 ) of about 1 . 5 to 2 nm thus the sphere contains about 1000 proteins on its surface . In order to ensure that these structures are formed of rPrP , we decided to label them with antibodies . This method clearly identified spheres consisting of rPrP ( details in Fig . S1 ) . In a previous study [10] , we showed that polymerization kinetics could not be explained by the existence of an off-pathway . Thus an important question remains: what exactly is the role of micelles in the polymerization mechanism ? To answer this , we decided to analyse the evolution of the micelle quantity during polymerization kinetics . Qualitative analysis using electron microscopy revealed that the number of micelles is important a few minutes after dilution into polymerization buffer , and then rapidly decreases when fibrils are formed ( Fig . 2 ) . A semi-quantitative analysis of the amount of round shape structures suggested a precursor relationship between the micelles and fibrils ( Fig . 2 ) . This was established in two different buffers exhibiting very different lag phase and thus showing that it is very likely a common feature of in vitro prion polymerization . To quantitatively analyse the consequences of an on-pathway micelle intermediate , we built an a priori model describing the different steps with the microscopic processes involved and their contributions to the whole system . From this microscopic model it is then possible to quantify macroscopic data , such as micelles , polymers and monomers concentration . The model can be detailed into four main parts . A standard approach to formalize particle interactions between monomers , micelles and polymers , is to describe the transition rates between the states of the system through kinetic schemes . Well-known in chemistry , such a modelling is also used for polymerization models [4] and according to this method , it is then possible to write the differential equations of the concentration describing the dynamics of each quantity involved in the system . In order to give the clearest insight as possible of the model elaboration process both for biologists and mathematicians , let us consider the 4 phases listed in the last section and progressively introduce the microscopic processes involved together with their kinetic schemes and the corresponding equations . A qualitative analysis of the dynamics of micelles and polymers given by our model ( see Fig . 6 ) is consistent with the one observed in experiments ( see Fig . 2 ) . As expected , the correlation between polymers formation and the decreasing of micelles concentration is connected with the PrP* formation . For this purpose , we assume in our simulations that PrP* monomers originate from micelles and do not exist before , i . e . their initial concentration is null . Furthermore , this model was built to analyse the lag phase and be compared to data . Several definitions of the lag time exist but they are mostly related to the half-time ( denoted by ) , which is the time when half of the final polymerized mass is reached . One of the definitions links the to the lag time by the relation: , with the maximal slope of the sigmoid [6] describing polymer dynamics with respect to time . This formula makes sense for a genuine sigmoidal equation , but here we cannot have any explicit solution to the model . Since the half-time is better defined and more tractable on our data , we choose to use it to analyse our results . It is the possible to focus ourselves on four main results provided by the model: Taken together , these four points , allow us to conclude that experimental data corroborate our model . Furthermore , it suggests a simple explanation for the weak dependency of the lag time with initial concentration and it proposes a new interpretation of the overcoming kinetic barrier of the prion protein . A posteriori , the microscopic processes involving prion proliferation , built here , describe the observed in vitro macroscopic facts .
Evidences for the existence of micelle as an on-pathway during the formation of amyloid in vitro leads then to the question of the existence of such intermediate in vivo . Indeed , the concentrations of rPrP used to study the in vitro polymerization are far above those observed in vivo and buffers involved are not compatible with life . But , in the view presented here , micelles play the important role to sustain the conformation that is eligible for the amyloid formation . This suggests that the PrPC should reach a specific conformation to be able to polymerize into amyloid . What happens in vivo ? It was recently shown that: Thus our hypothesis is the formation of mixed-micelles containing phospholipids and rPrP reducing the concentration necessary to reach CMC ( Critical Micellar Concentration ) under physiological conditions . We believe that it is consistent with the appearance of nucleation , in vivo , at low concentration of proteins . The main characteristic of the in vivo formed amyloids is infectiosity and this property is related to the amyloid structure [25] . It is important to remind that most of the amyloids produced in vitro are not infectious . However , recently it was shown that addition of phospholipids [21] during in vitro polymerization leads consistently to infectious amyloids [20] , [26] . Furthermore , it was proven that rPrP proteins interact with membrane phospholipids [19] and this interaction precedes conformational changes [18] , a phenomenon also observed for other amyloid forming peptides [14] . Our hypothesis of formation of mixed-micelles containing phospholipids and rPrP , in such mixed-micelles , as in pure prion micelles , allow the protein to reach the PrP* conformation competent to generate infectious amyloids .
This model is used to fit data , that is why we have to give assumptions on rates to obtain a physical and biological relevant model . In the case of micelles ( or 3D spherical structures as it is observed experimentally here ) , we assume as in [27] that assimilation rates and are constant for any and we denote them by and respectively , both of them being positive constants . Indeed , we assume that both types of proteins have the same affinity with respect to micelles of any size . The depletion rate needs to take the spherical structure of a micelle into account , which radius linked to the number of monomers that composes it . In [27] , this term is given under the form ( 18 ) which is derived from chemical potential for one species of monomers . However , for the sake of simplicity we interpret differently this form and adapt it to one model . First , we suppose that in the smallest size , micelles do not transconformed PrPC monomers into amyloid competent isoform . We justify this assumption by thermodynamic constraints , assumed to be stronger in the greatest size . As the term in is dominant for small micelles , we let ( 19 ) The term in is dominant in the greatest size , thus this part is taken into account for the depletion of PrP* , ( 20 ) Now , for polymerization , we consider a constant polymerization rate , such that is equal to for any , and a depolymerization rate equal to a constant for , where is the nucleus size and equal to for longer polymer , , i . e . polymerization becomes an irreversible process after the nucleus is reached . Moreover , a linear splitting rate is taken , that is with and a uniform kernel given by ( 21 ) | Understanding the mechanism of prions is an important issue . Indeed , it involves a mechanism modifying the structure of the proteins that are of high interest in theoretical biology . Knowing the underlying mechanism that leads to prion disease could help further investigations in the world of amyloid disease and for example the so-called Alzheimer's disease . The theory of prion , also known as Protein-Only , has been widely studied . Nevertheless no mathematical models are able to reproduce the phenomena in silico . This suggests a lack of information in the theory . Here we propose a new model , built with a new approach theory that fits experimental data in a very satisfactory way . This model , together with experiments , maintains the idea that an intermediate conformation of the protein helps the disease to spread . Besides , this work is an excellent example of a strong interaction between mathematical modelling and biological approach . Indeed , because of a strong discrepancy between theoretical results of the early original model and biological data on pathological prion formation , the team of biologists decided to investigate more closely their experiments . They came out with a new discovery: the crucial role of micelles in the pathological conformation of the prion protein . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"computer",
"and",
"information",
"sciences",
"mathematical",
"computing",
"mathematics",
"protein",
"interactions",
"molecular",
"complexes",
"theoretical",
"biology",
"proteins",
"applied",
"mathematics",
"biology",
"and",
"life",
"sciences",
"computing",
... | 2014 | A Micellar On-Pathway Intermediate Step Explains the Kinetics of Prion Amyloid Formation |
Although the Factor V Leiden ( FVL ) gene variant is the most prevalent genetic risk factor for venous thrombosis , only 10% of FVL carriers will experience such an event in their lifetime . To identify potential FVL modifier genes contributing to this incomplete penetrance , we took advantage of a perinatal synthetic lethal thrombosis phenotype in mice homozygous for FVL ( F5L/L ) and haploinsufficient for tissue factor pathway inhibitor ( Tfpi+/- ) to perform a sensitized dominant ENU mutagenesis screen . Linkage analysis conducted in the 3 largest pedigrees generated from the surviving F5L/L Tfpi+/- mice ( ‘rescues’ ) using ENU-induced coding variants as genetic markers was unsuccessful in identifying major suppressor loci . Whole exome sequencing was applied to DNA from 107 rescue mice to identify candidate genes enriched for ENU mutations . A total of 3 , 481 potentially deleterious candidate ENU variants were identified in 2 , 984 genes . After correcting for gene size and multiple testing , Arl6ip5 was identified as the most enriched gene , though not reaching genome-wide significance . Evaluation of CRISPR/Cas9 induced loss of function in the top 6 genes failed to demonstrate a clear rescue phenotype . However , a maternally inherited ( not ENU-induced ) de novo mutation ( Plcb4R335Q ) exhibited significant co-segregation with the rescue phenotype ( p = 0 . 003 ) in the corresponding pedigree . Thrombosis suppression by heterozygous Plcb4 loss of function was confirmed through analysis of an independent , CRISPR/Cas9-induced Plcb4 mutation ( p = 0 . 01 ) .
Venous thromboembolism ( VTE ) affects 1:1000 individuals in the US each year and is highly heritable [1 , 2] . A single nucleotide variant ( SNV ) in the F5 gene , referred to as Factor V Leiden ( FVL , p . R506G ) is present in 5–10% of Europeans , conferring a 2–4 fold increased risk for VTE [3] . Although ~25% of VTE patients carry the FVL variant [4] , only ~10% of individuals heterozygous for FVL develop thrombosis in their lifetime . To identify genetic variants that could potentially function as modifiers for FVL-associated VTE risk , we recently reported a dominant ENU screen [5] in mice sensitized for thrombosis . Such sensitized screens have been previously successful in identifying modifier genes for various phenotypes [6–9] . This screen was based on mice homozygous for the FVL mutation ( F5L/L ) and haploinsufficient for tissue factor pathway inhibitor ( Tfpi+/- ) . As previously reported , F5L/L Tfpi+/- mice exhibit normal embryonic development , although nearly all die of widespread , systemic thrombosis in the immediate perinatal period [10] . After ENU mutagenesis , 98 G1 F5L/L Tfpi+/- progeny survived to weaning ( “rescues” ) and 16 progeny exhibited successful transmission of the ENU-induced suppressor mutation . However , subsequent efforts to genetically map the corresponding suppressor loci were confounded by complex strain-specific differences introduced by the required genetic outcross [5] . Similar genetic background effects have complicated previous mapping efforts [11] and have been noted to significantly alter other phenotypes [12 , 13] . Additional challenges of this mapping approach include the requirement for large pedigrees and limited mapping resolution , with candidate intervals typically harboring tens to hundreds of genes and multiple closely linked mutations . High throughput sequencing methods have enabled the direct identification of ENU-induced mutations . Thus , mutation identification in ENU screens is no longer dependent upon an outcross strategy for gene mapping [14–16] . We now report whole exome sequencing ( WES ) of 107 rescue mice ( including 50 mice from the previously reported ENU screen [5] ) . Assuming loss of gene function as the mechanism of rescue , these WES data were analyzed gene-by-gene to identify genes enriched with mutations ( mutation burden analysis ) . The Arl6ip5 gene emerged as the top candidate suppressor locus from this analysis . However , an independent CRISPR/Cas9-generated Arl5ip5 mutant allele failed to demonstrate highly penetrant rescue of the F5L/L Tfpi+/- lethal phenotype . Surprisingly , a maternally inherited ( not ENU-induced ) de novo mutation ( Plcb4R335Q ) exhibited significant co-segregation with the rescue phenotype ( p = 0 . 003 ) in an expanded pedigree .
In the previously reported ENU screen [5] , viable F5L/L Tfpi+/- rescue mice were outcrossed to the 129S1/SvImJ strain to introduce the genetic diversity required for subsequent mapping experiments . However , complex strain modifier gene interactions confounded this analysis and resulted in a large number of “phenocopies” ( defined as viable F5L/L Tfpi+/- mice lacking the original rescue mutation ) . To eliminate confounding effects of these thrombosis strain modifiers , we generated an additional 2 , 834 G1 offspring exclusively maintained on the C57BL/6J background . The 42 G1 F5L/L Tfpi+/- mice alive at 6 weeks of age were mated to F5L/L mice to test the heritability of the survival phenotype . Twenty-one of these 42 mice generated at least one litter and 15 ( 1 female , 14 males ) produced ≥1 offspring with the F5L/L Tfpi+/- genotype . Fifteen new rescue pedigrees were established from this screen ( S1 Table ) . All pedigrees were expanded until all rescues either were infertile or died without producing any progeny with rescue genotype . The frequency , survival , weight , and sex distributions of identified rescues were consistent with our previous report ( S1 Fig ) . Though many of the pedigrees previously generated on the mixed 129S1/SvImJ-C57BL/6J background generated >45 rescue progeny per pedigree ( 8/16 ) [5] , all pedigrees on the pure C57BL/6J background yielded <36 rescue mice ( most generating ≤5 rescues ) ( S1 Table ) . Significantly smaller pedigrees in comparison to the previous screen ( p = 0 . 010 , S2 Fig ) are likely explained by a generally positive effect of the hybrid 129S1/SvImJ-C57BL/6J strain background either directly on rescue fertility ( hybrid vigor ) or indirectly by reducing the severity of the F5L/L phenotype . Although a contribution from nongenetic factors cannot be excluded , the C57BL/6J and 129S1/SvImJ strains have been shown to exhibit significant differences in a number of hemostasis-related parameters , including platelet count and TFPI and tissue factor expression levels [17] , with the genetic variations underlying such strain specific differences likely contributing to the genetic mapping complexity noted in the previous report [5] . As the rescue pedigrees were maintained on a pure C57BL/6J background , the only genetic markers that could be used for mapping were ENU-induced variants . WES of one G1 or G2 member of the three largest pedigrees ( 1 , 6 , and 13 , S2 Table ) , identified a total of 86 candidate ENU variants that were also validated by Sanger sequencing analysis ( S3 Table ) . Of these 86 candidate genes , 69 were present in the G1 rescue but not its parents ( G0 ) , indicating that they were likely ENU-induced variants . These 69 variants were then further genotyped in all other rescue progeny in the respective pedigrees . Given the low number of identified genetic markers ( 20–26 per pedigree ) , these three pedigrees were poorly powered ( 29 . 6% , 21 . 7% and 39 . 4% , respectively ) to identify the rescue variants by linkage analysis ( S3A–S5A Figs ) . None of the 19 ENU variants tested in pedigree 1 ( S3B Fig ) , showed linkage with a LOD-score >1 . 5 ( S3C Fig ) . Similarly , 26 and 24 variants analyzed in pedigrees 6 and 13 , respectively ( S4B and S5B Figs ) also failed to demonstrate a LOD-score >1 . 5 ( S4C and S5C Figs ) . Failure to map the causal loci in any of these pedigrees was likely due to insufficient marker coverage . However , in these analyses , we could not exclude the contribution from a non-ENU-induced variant [18] or an unexpectedly high phenocopy rate . While WES has been successfully applied to identify causal ENU variants within inbred lines [19] and in mixed background lines [20 , 21] , whole genome sequencing ( WGS ) provides much denser and more even coverage of the entire genome ( ~3 , 000 ENU variants/genome expected ) and outperforms WES for mapping [15] . However , a WGS approach requires sequencing multiple pedigree members [16] , or pooled samples at high coverage [15] , resulting in considerably higher expense with current methods . In order to identify exonic ENU mutations , a total of 107 G1 rescues ( 57 from the current ENU screen and an additional 50 rescues with available material from the previous screen [5] ) , were subjected to WES ( S2 Table ) . From ~1 . 5 million initially called variants ( 34 , 000 in exonic regions ) 6 , 735 SNVs and 36 insertions-deletions ( INDELs ) within exonic regions were identified as potential ENU-induced mutations , using an in-house filtering pipeline ( see Materials and methods ) . The most common exonic variants were nonsynonymous SNVs ( 47% ) , followed by mutations in 3’ and 5’ untranslated regions ( 31% ) and synonymous SNVs ( 15% ) . The remaining variants ( 7% ) were classified as splice site altering , stoploss , stopgain , or INDELs ( Fig 1A ) . T/A -> C/G ( 47% ) , and T/A -> A/T ( 24% ) SNVs were overrepresented , while C/G -> G/C ( 0 . 8% ) changes were greatly underrepresented ( Fig 1B ) , consistent with previously reported ENU studies [22 , 23] . Since ENU is administered to the G0 father of G1 rescues , only female progeny are expected to carry induced mutations on the X chromosome , while males inherit their single X chromosome from the unmutagenized mother . Among the called variants , all chromosomes harbored a similar number of mutations in both sexes , with the exception of the X chromosome where a >35 fold increase in SNVs per mouse was observed in females ( Fig 1C ) . The average number of exonic ENU mutations for G1 rescues was ~65 SNVs per mouse ( Fig 1D ) , consistent with expected ENU mutation rates [16 , 23] . These data suggest that most called variants are likely to be of ENU origin . WES data for 107 independent rescue mice were jointly analyzed to identify candidate genes that are enriched for potentially deleterious ENU-induced variants including missense , nonsense , frameshift , and splice site altering mutations ( 3 , 481 out of 6 , 771 variants in 2 , 984 genes , ~32 . 5 potentially deleterious variants per mouse , S4 Table ) . Similar mutation burden analyses have been used to identify genes underlying rare diseases caused by de novo loss-of-function variants in humans [24–27] . In our study , the majority of genes harbored only a single ENU-induced variant , with 15 SNVs identified in Ttn , the largest gene in the mouse genome ( Fig 1E ) . After adjusting for coding region size and multiple testing ( for 2 , 984 genes ) , the ENU-induced mutation burden of potentially deleterious variants was significantly greater than expected by chance for 3 genes ( FDR<0 . 1 , Arl6ip5 , Itgb6 , C6 ) and suggestive for 9 additional genes ( FDR<0 . 25 ) . Sanger sequencing validated 36 of the 37 variants in these 12 candidate genes ( S4 Table ) . While in this study , stringent correction for multiple testing suggested no significant enrichment ( Arl6ip5 FDR = 0 . 68 , Fig 2 ) , the potential power of this burden analysis is highly dependent on the number of possible genes that could result in a viable rescue . If there were 30 such genes in the genome and every one of the 107 rescue mice carried a mutation in one of these 30 genes , each gene would be , on average , represented by ~3 . 5 mutations ( 107/30 ) , with >7 genes expected to carry 5 or more mutations , which should have been sufficient to distinguish from the background mutation rate . However , if 500 genes could rescue the phenotype , sequencing close to a thousand mice would be required to achieve sufficient mapping power . The power could be further compromised by modifier genes with incomplete penetrance , imperfect predictions for potentially harmful mutations , and by the previously reported background survival rate for the rescue mice [10] . Due to the uncertainty of the power of these analyses , we proceeded to experimentally test the thrombosupressive effects of loss of function mutations in the genes identified by mutation burden analysis . Independent null alleles were generated with CRISPR/Cas9 for the top candidate genes ( Arl6ip5 , C6 , Itgb6 , Cpn1 , Sntg1 and Ces3b; Fig 2 ) to test for thrombosuppression . From 294 microinjected zygotes with pooled guide RNAs targeting these 6 genes , we obtained 39 progeny . CRISPR/Cas9 genome editing was assessed by Sanger sequencing of the sgRNA target sites . Approximately 190 independent targeting events were observed across the 6 genes in 36 of the 39 mice including small INDELs , single nucleotide changes , and several large ( >30bp ) deletions or inversions . Targeted alleles were either homozygous , heterozygous , or mosaic , with the number of editing events varying greatly for different sgRNAs ( 2 . 5–85% ) . Two or more different CRISPR/Cas9-induced alleles for each of the candidate genes ( S5 Table ) were bred to isolation but maintained on the F5L background for subsequent test crossing . The progeny of F5L/L Tfpi+/+ mice crossed with F5L/+ Tfpi+/- mice ( one of these parental mice also carrying the CRISPR/Cas9-induced allele ) were monitored for survival of F5L/L Tfpi+/- offspring ( Table 1 , S6 Table ) . Over 100 progeny were generated for each of the candidate genes with no obvious rescue effect . A slight increase in rescues carrying the F5L/L Tfpi+/- Arl6ip5+/- genotype was noted , although it remained non-significant after surveying 205 offspring ( p = 0 . 21 , Table 1 ) . Our sensitized suppressor screening strategy is highly dependent on the underlying thrombosis model . Modifier genes rescuing the F5L/L Tfpi+/- synthetic lethal phenotype are potentially relevant to the common human FVL variant and our previous observations that mutations in F8 and F3 can rescue F5L/L Tfpi+/- demonstrate the sensitivity of this model to genetic alterations in coagulation system balance . However , rescue of F5L/L Tfpi+/- lethality by haploinsufficiency for F3 ( the target of TFPI ) only exhibits penetrance of ~33% [5] , a level of rescue which current observations cannot exclude for Arl6ip5 and Sntg1 . For most of the other candidate genes , the number of observed F5L/L Tfpi+/- mice did not differ from the expected background survival rate for this genotype ( ~2% ) [10] . Though higher numbers of rescues were observed for offspring from the Sntg1 cross , these were equally distributed between mice with and without the Sntg1 loss-of-function allele . The number of G1 rescues produced from each ENU-treated G0 male is shown in Fig 3A . Though most of the 182 G0 males yielded few or no G1 rescue offspring , a single G0 produced 6 rescues out of a total of 39 offspring ( Fig 3A ) , including the founder G1 rescue for the largest pedigree ( number 13 ) . This observation suggested a potential shared rescue variant rather than 6 independent rescue mutations from the same G0 founder . Similarly , another previously reported ENU screen identified 7 independent ENU pedigrees with an identical phenotype mapping to the same genetic locus , also hypothesized to result from a single shared mutation [11] . While rescue siblings could theoretically originate from the same mutagenized spermatogonial stem cell and share ~50% of their induced mutations [28] , such a common stem cell origin was excluded by exome sequence analysis in the rescue G1 sibs identified here ( see Materials and methods ) . Analysis of WES for 3 of the G1 rescues originating from this common G0 founder male ( Fig 3B , S2 Table ) identified 3 protein-altering variants ( Plcb4R335Q , Pyhin1G157T , and Fignl2G82S ) shared among 2 or more of the 6 G1 rescues ( S7 Table ) . Plcb4R335Q was detected as a de novo mutation in one of the non-mutagenized G0 females in phase with the Tfpi null allele ( Fig 3B ) and was present in 3 out of 6 G1 rescue siblings . Plcb4 is located approximately 50 megabases upstream of the Tfpi locus on chromosome 2 ( predicted recombination between Plcb4 and Tfpi ~14 . 1% ) ( Fig 3C ) [29 , 30] . While non-rescue littermates exhibited the expected rate of recombination between the Plcb4R335Q and Tfpi loci ( 20 . 2% ) , all 43 rescue mice ( 3 G1s and their 40 ≥G2 progeny ) were non-recombinant and carried the Plcb4R335Q variant . This co-segregation between the Plcb4R335Q variant and the rescue phenotype is statistically significant ( p = 0 . 003; Fig 3C ) . Plcb4R335Q lies within a highly conserved region of Plcb4 ( Fig 3D ) and is predicted to be deleterious by Polyphen-2 [31] . The other identified non-ENU variants ( Pyhin1G157T and Fignl2G82S ) did not segregate with the rescue phenotype ( S6 Fig ) . Although the estimated de novo mutation rate for inbred mice ( ~5 . 4 x 10−9 bp/generation ) is 200X lower than our ENU mutation rate , other de novo variants have coincidentally been identified in ENU screens [32] . Of note , the Plcb4R335Q variant was originally removed from the candidate list by a filtering step based on the assumption that each ENU-induced mutation should be unique to a single G1 offspring . Although this algorithm was very efficient for removing false positive variants in our screen and others [21] , our findings illustrate the risk for potential false negative results that this approach confers . An independent Plcb4 null allele was generated by CRISPR/Cas9 . Three distinct INDELs were identified by Sanger sequencing in the 25 progeny obtained from the CRISPR/Cas9-injected oocytes . One of these alleles introduced a single nucleotide insertion at amino acid 328 , resulting in a frameshift in the protein coding sequence and a marked decrease in the steady state mRNA level from the mutant allele ( ~2% compared to wildtype ) , consistent with nonsense-mediated decay ( Plcb4ins1 , Fig 4A and 4B ) . A total of 169 progeny from a F5L/L Plcb4+/ins1 X F5L/+ Tfpi+/- cross yielded 11 F5L/L Tfpi+/- rescue progeny surviving to weaning ( Fig 4C , S8 Table ) . Ten of these 11 rescues carried the Plcb4ins1 allele , consistent with significant rescue ( p = 0 . 01 , Fig 4C ) with reduced penetrance ( ~40% ) . Plcb4 encodes phospholipase C , beta 4 and has been recently associated with auriculocondylar syndrome in humans [33] . No role for PLCB4 in the regulation of hemostasis has been previously reported , and the underlying mechanism for suppression of the lethal F5L/L Tfpi+/- phenotype is unknown . The above rescue of the F5L/L Tfpi+/- phenotype by an independent Plcb4 mutant allele , strongly supports the identification of the de novo Plcb4R355Q mutation as the causal suppressor variant for Pedigree 13 . These findings are also most consistent with a loss-of-function mechanism of action for the Plcb4R355Q mutation . The lack of a positive signal from this genomic region by the linkage analysis described above ( S5 Fig ) is likely explained by the absence of a nearby genetically informative ENU variant ( the closest , Abca2 is located >50 Mb downstream from both Tfpi and Plcb4 ( S3 Table , S5 Fig ) ) . Of note , 4 of the 107 rescue mice in the WES mutation burden analysis also carried a Plcb4 mutation consistent with its suppressor function , though below the level of statistical significance . Nonetheless , these findings highlight the feasibility of our approach , given sufficient power . In conclusion , we performed a dominant , sensitized ENU mutagenesis screen for modifiers of thrombosis . Analysis of extended pedigrees identified Plcb4 as a novel thrombosis modifier . Though mutation burden analysis suggested several other potential modifier loci , including Arl6ip5 , incomplete penetrance and the background phenocopy rate significantly limited the power to detect additional thrombosis suppressor genes . Future applications of this approach will likely require significantly larger sample sizes and/or a more stringent sensitized genotype for screening . Nonetheless , our findings demonstrate the power of a sensitized ENU screen and mutation burden analysis to identify novel loci contributing to the regulation of hemostatic balance and candidate modifier genes for thrombosis and bleeding risk in humans .
Mice carrying the murine homolog of the FVL mutation [34] ( F5L; also available from Jackson Laboratories stock #004080 ) or the TFPI Kunitz domain deletion ( Tfpi- ) [35] were genotyped using PCR assays with primers and conditions as previously described [34 , 35] , and maintained on the C57BL/6J background ( Jackson Laboratories stock #000664 ) . All animal care and procedures were performed in accordance with the Principles of Laboratory and Animal Care established by the National Society for Medical Research . The Institutional Animal Care and Use Committee at the University of Michigan has approved protocols PRO00005191 and PRO00007879 used for the current study and conforms to the standards of “The Guide for the Care and Use of Laboratory Animals” ( Revised 2011 ) . ENU mutagenesis was performed as previously described [5] , with all mice on the C57BL/6J genetic background . Briefly , 189 F5L/L male mice ( 6–8 weeks old ) were administered three weekly intraperitoneal injections of 90 mg/kg of ENU ( N-ethyl-N-nitrosourea , Sigma-Aldrich ) . Eight weeks later , 182 surviving males were mated to F5L/+ Tfpi+/- females and their G1 progeny were genotyped at age 2–3 weeks to identify viable F5L/L Tfpi+/- offspring ( ‘rescues’ ) . F5L/L Tfpi+/- G1 rescues were crossed to F5L/L mice on the C57BL/6J genetic background ( backcrossed >20 generations ) and transmission was considered positive with the presence of one or more rescue progeny . Theoretical mapping power in rescue pedigrees was estimated by 10 , 000 simulations using SIMLINK software [36] . Gender , age , WES details , and other characteristics for 108 rescue mice are provided in S2 Table . Genomic DNA ( gDNA ) extracted from tail biopsies of 56 G1 offspring from the current ENU screen and from an additional 50 F5L/L Tfpi+/- mice on the C57BL/6J background from the previous screen [5] were subjected to WES at the Northwest Genomics Center , University of Washington . Sequencing libraries were prepared using the Roche NimbleGen exome capture system . DNA from an additional two rescue offspring was subjected to WES at Beijing Genomics Institute or Centrillion Genomics Technologies , respectively ( S2 Table ) . These two libraries were prepared using the Agilent SureSelect capture system . 100 bp paired-end sequencing was performed for all 108 exome libraries using Illumina HiSeq 2000 or 4000 sequencing instruments . Two WES mice represented rescue pedigree 1: the G1 founder and a G2 rescue offspring . The latter was used for linkage analysis , but excluded from the burden analysis ( S2 Table ) . Average sequencing coverage , estimated by QualiMap software [37] , was 77X , and >96% of the captured area was covered by at least 6 independent reads ( S2 Table ) . All generated fastq files have been deposited to the NCBI Sequence Read Archive ( Project accession number #PRJNA397141 ) . A detailed description of variant calling as well as in-house developed scripts for variant filtration are online as a GitHub repository ( github . com/tombergk/FVL_mod ) . In short , Burrows-Wheeler Aligner [38] was used to align reads to the Mus Musculus GRCm38 reference genome , Picard [39] to remove duplicates , and GATK [40] to call and filter the variants . Annovar software [41] was applied to annotate the variants using the Refseq database . All variants within our mouse cohort present in more than one rescue were declared non-ENU induced and therefore removed . Unique heterozygous variants with a minimum of 6X coverage were considered as potential ENU mutations . Among 107 whole exome sequenced G1 mice , 38 were siblings ( 13 sib-pairs and 4 trios , S2 Table ) . 190 heterozygous variants present in 2 or 3 mice ( representing sibpairs or trios ) out of 107 rescues were examined , with 15 found to be shared by siblings ( S7 Table ) . Of the 7 sibs/trios sharing an otherwise novel variant , none shared >10% of their identified variants–inconsistent with the expected 50% for progeny originating from the same ENU-treated spermatogonial stem cell . All ENU-induced variants predicted to be potentially harmful within protein coding sequences including missense , nonsense , splice site altering SNVs , and out-of-frame insertions-deletions ( INDELs ) , were summed for every gene . The number of potentially damaging variants per gene was compared to a probability distribution of each gene being targeted by chance . Probability distributions were obtained by running 10 million random permutations using probabilities adjusted to the length of the protein coding region . A detailed pipeline for the permutation analysis is available online ( github . com/tombergk/FVL_mod ) . Genes that harbored more potentially damaging ENU-induced variants than expected by chance were considered as candidate modifier genes . FDR statistical correction for multiple testing was applied as previously described [42] . All coding variants in pedigrees 1 , 6 , and 13 as well as all variants in candidate modifier genes from the burden analysis were assessed using Sanger sequencing . Variants were considered ENU-induced if identified in the G1 rescue but not its parents . All primers were designed using Primer3 software [43] and purchased from Integrated DNA Technologies . PCR was performed using GoTaq Green PCR Master Mix ( Promega ) , visualized on 2% agarose gel , and purified using QIAquick Gel Extraction Kit ( Qiagen ) . Sanger sequencing of purified PCR products was performed by the University of Michigan Sequencing Core . Outer primers were used to generate the PCR product which was then sequenced using the internal sequencing primers . All outer PCR primers ( named: gene name+’_OF/OR’ ) and internal sequencing primers ( named: gene name+’_IF/IR’ ) are listed in S9 Table . Guide RNA target sequences were designed with computational tools [44 , 45] ( http://www . broadinstitute . org/rnai/public/analysis-tools/sgrna-design or http://genome-engineering . org ) and top predictions per each candidate gene were selected for functional testing ( S10 Table ) . Single guide RNAs ( sgRNA ) for C6 , Ces3b , Itgb6 , and Sntg1 were in vitro synthesized ( MAXIscript T7 Kit , Thermo Fisher ) from double stranded DNA templates by GeneArt gene synthesis service ( Thermo Fisher ) while the 4 sgRNAs for Arl6ip5 were in vitro synthesized using the Guide-it sgRNA In Vitro Transcription Kit ( Clontech ) . The sgRNAs were purified prior to activity testing ( MEGAclear Transcription Clean-Up Kit , Thermo Fisher ) . Both the Wash and Elution Solutions of the MEGAclear Kit were pre-filtered with 0 . 02 μm size exclusion membrane filters ( Anotop syringe filters , Whatman ) to remove particulates from zygote microinjection solutions , thus preventing microinjection needle blockages . Target DNA for the in vitro cleavage assays was PCR amplified from genomic DNA isolated from JM8 . A3 C57BL/6N mouse embryonic stem ( ES ) cells [46] with candidate gene specific primers ( S10 Table ) . In vitro digestion of target DNA was carried out by complexes of synthetic sgRNA and S . pyogenes Cas9 Nuclease ( New England BioLabs ) according to manufacturer's recommendations . Agarose gel electrophoresis of the reaction products was used to identify sgRNA molecules that mediated template cleavage by Cas9 protein ( S7 Fig ) . Arl6ip5 was assayed separately , with one out-of-four tested sgRNAs successfully cleaving the PCR template . Synthetic sgRNAs that targeted Cpn1 were not identified by the in vitro Cas9 DNA cleavage assay . As an alternative assay , sgRNA target sequences ( Cpn1-g1 , Cpn1-g2 ) were cloned into plasmid pX330-U6-Chimeric_BB-CBh-hSpCas9 ( Addgene . org Plasmid #42230 ) [47] and co-electroporated into JM8 . A3 ES cells as previously described [48] . Briefly , 15 μg of a Cas9 plasmid and 5 μg of a PGK1-puro expression plasmid [49] were co-electroporated into 0 . 8x107 ES cells . On days two and three after electroporation media containing 2 μg/ml puromycin was applied to the cells; then selection free media was applied for four days . Genomic DNA was purified from surviving ES cells . The Cpn1 region targeted by the sgRNA was PCR amplified and tested for the presence of indel formation with a T7 endonuclease I assay according to the manufacturer’s instructions ( New England Biolabs ) . CRISPR/Cas9 gene edited mice were generated in collaboration with the University of Michigan Transgenic Animal Model Core . A premixed solution containing 2 . 5 ng/μl of each sgRNA for Arl6ip5 , C6 , Ces3b , Itgb6 , Sntg1 , and 5 ng/μl of Cas9 mRNA ( GeneArt CRISPR Nuclease mRNA , Thermo Fisher ) was prepared in RNAse free microinjection buffer ( 10 mM Tris-Hcl , pH 7 . 4 , 0 . 25 mM EDTA ) . The mixture also included 2 . 5 ng/μl of pX330-U6-Chimeric_BB-CBh-hSpCas9 plasmid containing guide Cpn1-g1 and a 2 . 5 ng/μl of pX330-U6-Chimeric_BB-CBh-hSpCas9 plasmid containing guide Cpn1-g2 targeting Cpn1 ( S10 Table ) . The mixture of sgRNAs , Cas9 mRNA , and plasmids was microinjected into the male pronucleus of fertilized mouse eggs obtained from the mating of stud males carrying the F5L/+ Tfpi+/- genotype on the C57BL/6J background with superovulated C57BL/6J female mice . Microinjected eggs were transferred to pseudopregnant B6DF1 female mice ( Jackson Laboratories stock #100006 ) . DNA extracted from tail biopsies of offspring was genotyped for the presence of gene editing . The Plcb4 allele was targeted in a separate experiment in collaboration with the University of Michigan Transgenic Animal Model Core using a pX330-U6-Chimeric_BB-CBh-hSpCas9 plasmid that contained guide Plcb4 ( 5 ng/μl ) . Initially , sgRNA targeted loci were tested using PCR and Sanger sequencing ( primer sequences provided in S10 Table ) . Small INDELs were deconvoluted from Sanger sequencing reads using TIDE software [50] . A selection of null alleles from >190 editing events were maintained for validation ( S5 Table ) . Large ( >30 bp ) deletions were genotyped using PCR reactions that resulted in two visibly distinct PCR product sizes for the deletion and wildtype alleles . Expected product sizes and genotyping primers for each deletion are listed in S5 Table . All genotyping strategies were initially validated using Sanger sequencing . A qPCR approach was applied to exclude large on-target CRISPR/Cas9-induced deletions . All DNA samples were quantified using the Quant-iT™ PicoGreen® dsDNA Assay Kit ( Life Technologies ) and analyzed on the Molecular Devices SpectraMax® M3 multi-mode microplate reader using SoftMax Pro software and diluted to 5ng/μl . Primer pairs were designed for each gene using Primer Express 3 . 0 software ( S9 Table ) and samples were measured in triplicate using Power SYBR Green PCR Master Mix ( Thermo Fisher Scientific ) on a 7900 HT Fast Real-Time PCR System ( Applied Biosystems ) with DNA from wildtype C57BL/6J mice used as a reference . While large CRISPR/Cas9 induced deletions extending the borders of the PCR primers have been reported [51 , 52] , qPCR did not detect evidence for a large deletion in any of the CRISPR targeted genes . The ratio of WT to Plcb4+ins1 mRNA levels was determined as previously described [18] . In short , a whole brain tissue sample was snap frozen in liquid nitrogen from a Plcb4+/ins1 mouse . Total RNA was extracted using an RNA extraction kit ( Nucleospin RNA from Macherey-Nagel ) and 250 ng of total RNA was converted to complementary DNA ( cDNA ) using SuperScript IV One-Step RT-PCR ( Invitrogen ) following the manufacturer’s instructions . Genotyping primers spanning the nearest intron ( primers Plcb4_cDNA_OF1 and Plcb4_cDNA_OR , S9 Table ) were used to amplify a segment of Plcb4 containing the +1 insertion from the cDNA samples . PCR products were extracted from agarose gels using a QIAquick Gel Purification Kit ( Qiagen ) and submitted for Sanger sequencing ( primer Plcb4_cDNA_OF2 , S9 Table ) . The differential allelic expression was estimated from the ratio between the wildtype and Plcb4ins1 sequence peak areas in the cDNA sample compared to gDNA using Phred software [53] . This ratio was calculated for 10 consecutive positions within the PCR product where the wildtype and Plcb4ins1 alleles contain a different nucleotide . Kaplan-Meier survival curves and a log-rank test to estimate significant differences in mouse survival were performed using the ‘survival’ package in R [54] . A paired two-tailed Student’s t-test was applied to estimate differences in weights between rescue mice and their littermates . Fisher’s exact tests were applied to estimate deviations from expected proportions in mouse crosses . Mendelian segregation for CRISPR/Cas9-induced alleles among non-rescue littermates was assessed in a subset of mice by Sanger sequencing and then assumed for the rest of the littermates in the Fisher’s exact tests . Benjamini and Hochberg FDR for ENU burden analysis , Student’s t-tests , and Fisher’s exact tests were all performed using the ‘stats’ package in R software [55] . Linkage analysis was performed on the Mendel platform version 14 . 0 [56] and LOD scores ≥3 . 3 were considered genome-wide significant [57] . | Abnormal blood clotting in veins ( venous thrombosis ) or arteries ( arterial thrombosis ) are major health problems , with venous thrombosis affecting approximately 1 in every thousand individuals annually in the United States . Susceptibility to venous thrombosis is governed by both genes and environment , with approximately 60% of the risk attributed to genetic influences . Though several genetic risk factors are known , >50% of genetic risk remains unexplained . Approximately 5% of people carry the most common known risk factor , Factor V Leiden . However , only 10% of these individuals will develop a blood clot in their lifetime . Mice carrying two copies of the Factor V Leiden mutation together with a mutation in a second gene called tissue factor pathway inhibitor develop fatal thrombosis shortly after birth . To identify genes that prevent this fatal thrombosis , we studied a large panel of mice carrying inactivating gene changes randomly distributed throughout the genome . We identified several genes as potential candidates to alter blood clotting balance in mice and humans with predisposition to thrombosis , and confirmed this protective function for DNA changes in one of these genes ( Plcb4 ) . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] | [
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"cardiovascular",
"medicine",
"thrombosis",
"mutation",
"molecular",
"biology",
"techniques",
"dna",
"sequencing",
"mammalian",
"genomics",
"research",
"and",
"analysis",
"methods",
"coagulation",
"disor... | 2018 | Whole exome sequencing of ENU-induced thrombosis modifier mutations in the mouse |
Regular treatment with praziquantel ( PZQ ) is the strategy for human schistosomiasis control aiming to prevent morbidity in later life . With the recent resolution on schistosomiasis elimination by the 65th World Health Assembly , appropriate diagnostic tools to inform interventions are keys to their success . We present a discrete Markov chains modelling framework that deals with the longitudinal study design and the measurement error in the diagnostic methods under study . A longitudinal detailed dataset from Uganda , in which one or two doses of PZQ treatment were provided , was analyzed through Latent Markov Models ( LMMs ) . The aim was to evaluate the diagnostic accuracy of Circulating Cathodic Antigen ( CCA ) and of double Kato-Katz ( KK ) faecal slides over three consecutive days for Schistosoma mansoni infection simultaneously by age group at baseline and at two follow-up times post treatment . Diagnostic test sensitivities and specificities and the true underlying infection prevalence over time as well as the probabilities of transitions between infected and uninfected states are provided . The estimated transition probability matrices provide parsimonious yet important insights into the re-infection and cure rates in the two age groups . We show that the CCA diagnostic performance remained constant after PZQ treatment and that this test was overall more sensitive but less specific than single-day double KK for the diagnosis of S . mansoni infection . The probability of clearing infection from baseline to 9 weeks was higher among those who received two PZQ doses compared to one PZQ dose for both age groups , with much higher re-infection rates among children compared to adolescents and adults . We recommend LMMs as a useful methodology for monitoring and evaluation and treatment decision research as well as CCA for mapping surveys of S . mansoni infection , although additional diagnostic tools should be incorporated in schistosomiasis elimination programs .
Schistosomiasis is a debilitating parasitic disease in tropical and sub-tropical areas which has recently attracted increased focus and funding for control through large scale mass drug administration ( MDA ) of praziquantel ( PZQ ) [1] . However the ability of current control initiatives to operate cost effectively is reduced by technical limitations of currently available schistosomiasis diagnostics [2] , [3] . With the recent resolution on schistosomiasis elimination by the 65th World Health Assembly , schistosomiasis diagnostics research for population-based assessment is critical with careful consideration given to how those tools might be used within disease elimination programmes [4]–[6] . At present the World Health Organization ( WHO ) recommends the Kato-Katz ( KK ) method as the standard tool for the qualitative and quantitative diagnosis of Schistosoma mansoni infection because of its assumed high specificity , relative simplicity in field conditions and attractive price . WHO also recommends MDA with PZQ to be delivered to community populations defined where KK surveys show an estimated prevalence of over 50% in school-aged children and to be delivered to children aged 6–16 years where the estimated prevalence is between 10% and 50% in this age group . However , it is well known that KK method from single stool samples , particularly at low infection endemicities and following PZQ MDA , can underestimate Schistosoma infection prevalence ( and intensities ) and thus overestimate cure rates [7]–[11] , whilst even multiple slides over multiple days of stool sampling can influence specificity and overestimate prevalence [10]–[15] . The necessity for more “field-applicable” , sensitive and cost-effective diagnostics than the KK method , at least for the routine surveillance of human S . mansoni infection such as that inherent within mapping of at-risk populations has also been recently highlighted [16] , [17] . Even more worrying , in some endemic regions microscopic stool samples examination is considered too logistically difficult in terms of personnel available for routine surveillance and therefore S . mansoni infection remains undetected and untreated in control programmes [18] . A promising diagnostic option is a urine strip test for Circulating Cathodic Antigen ( CCA ) which is a genus-specific glycan regurgitated by adult schistosome worms into the blood stream [19] . A number of cross-sectional studies evaluating CCA accuracy pre-treatment have stressed the need for further research assessing the potential role of this diagnostic assay at different stages of schistosomiasis control programs [15] , [17] , [20] . In this study , we assessed for the first time the diagnostic accuracy of CCA and of double KK faecal slides over three consecutive days for S . mansoni infection by age group at baseline and at two follow-up times post treatment which should be viewed as a proxy for low transmission areas . Because no true gold standard diagnoses are available , we developed and fitted latent Markov models ( LMMs ) to estimate diagnostic test sensitivities and specificities , the true underlying infection prevalence and the probabilities of transitions between infected and uninfected states [21] . LMMs - which are sometimes also referred to as latent transition models or regime-switching models – are used to analyze discrete-time longitudinal data where respondent observations contain measurement error . This approach defines the true states as categories ( latent classes ) of a dynamic latent ( unobservable ) variable within a statistical model . The Markov assumption is reflected in the model via transition probabilities which allow for correlation between a respondent's true state at times t−1 and t [22] . We analyzed a detailed dataset from a longitudinal cohort living along the shorelines of Lake Victoria in Uganda who received one or two doses of PZQ treatment at baseline . We demonstrated how the use of LMMs allows estimation of the ‘true’ prevalence of S . mansoni infection over time and the quantification of the additional benefit of a second PZQ dose in reducing re-infection levels by age group .
Ethical clearance was obtained from the Uganda National Council of Science and Technology and the study was also presented to the Danish National Committee on Biomedical Research Ethics in Denmark ( Reference Number: UNCST: HS 59 ) . Informed consent was obtained from individual adult participants but for children the parents or guardians consented on their behalf . Thereafter , each individual signed a consent form before any activity started . All information obtained from participants was kept confidential . Because some of the participants might have potentially received PZQ treatment recently through MDA , field survey assistants asked each participant detailed questions about previous treatment in order to exclude such individuals ever treated through MDA , although no such pre-treated individuals were identified in the current study . We conducted our study in Musoli village , Mayuge district at baseline and nine weeks after treatment during 2005 . Participants of this study were randomly allocated to one or two PZQ ( Shin Poong Pharmaceuticals , Seoul Republic of Korea ) doses at baseline ( 40 mg PZQ per kg body weight; double treatment group: two times PZQ 40 mg per kg body weight administered two weeks apart ) . The field survey assistants who delivered these treatments were not aware of the infection status of any participants at any time . First follow-up data collection was performed at nine weeks , and hence at a time aimed to assess cure rates where the risk of any eggs detected occurring as a consequence of reinfection was minimal . Second follow-up data collection was performed two years later during 2007 . With regards to treatment after two years , this is part of the National programme to treat everyone living in an endemic area and thus a second MDA was offered after two years . The study location was selected as this is an area of Uganda known for perennial S . mansoni transmission , situated on the shore of Lake Victoria [23] where the community population is not targeted by the National Control Program [24] , [25] . The community consists primarily of fishermen and their dependants with the lake being the only source of fresh water for them . Furthermore , infrequent use and/or availability of latrines leads to contamination of the lake especially near the shoreline , where there is underwater vegetation suitable for the aquatic intermediate host snails to thrive . In addition , only individuals >6 years of age were enrolled as very often this is the age from which schistosome-induced morbidity , in general , becomes evident . Before any data collection took place , trained and experienced demographers conducted in 2005 a census of the village population . During this process the inhabitants in each household registered their relation to the head of the household , year of birth , gender , occupation , duration of residency in the village and tribal membership . A stratified random sample for age and sex was then selected from the census data . Sample size calculations included detection of a significant difference of cure rates between the two treatment groups with reference to cure rates obtained along Lake Albert . A significance level of α = 0 . 05 , a power of 90% with a drop-out rate of 40% over the two years of studies contributed to the calculation of 552 individuals but at baseline we managed to recruit 446 with full parasitological and CCA in urine data as described elsewhere [23] , [26] . A further six cases were then eliminated from the analysis presented in this study due to missing data in number of treatments and their age . Stool samples were collected on three consecutive days from each member of the cohort and examined for the presence of S . mansoni ova . Two duplicate slides of each stool sample were examined using the KK technique at each day . Each slide was read by two trained microscopists and any discrepancies resolved before results were recorded as eggs per gram ( EPG ) faeces . The results for all six slides were combined for the descriptive results presented in Table 1 as this is considered the best KK diagnostic performance scenario [13] while results of two slides of each single day are combined and incorporated in the statistical models' derived results ( Tables 2A–3B ) because in many studies only a single stool sample is analysed [25] , [27]–[29] . Single urine samples were kept cold after collection and after return to the laboratory at the end of the day aliquoted and stored frozen . At baseline and nine weeks , CCA urine samples were kept in a freezer in Uganda for nine months before they were transferred frozen to Leiden . At the two year follow-up CCA urine samples were delivered to Leiden within four weeks of collection . In Uganda , all CCA urine samples were kept at −20 Celsius degrees . The freezer in which urine was stored was −18 to −20 degrees C . Results should not be affected as the antigen is very stable and its detection is not influenced by periods of being frozen , freeze thaw cycles , or even storage for weeks at room temperature . They were kept frozen during transport to the Department of Parasitology , in Leiden , The Netherlands , where the CCA urine assays were performed as previously described [19] , [30] . Briefly , for the laboratory-based test , 25 µL of completely thawed and vortexed urine was added to a tube containing dried carbon conjugated antibody , along with 75 µL of buffer and mixed well . Test strips were added , and allowed to develop for 40 minutes . Strips were removed , allowed to dry , and read against a set of five Quality Control ( QC ) standards of 0 , 10 , 100 , 1000 and 10 000 ng of semi-purified worm antigen ( containing CCA ) per ml negative urine . A score of 0 indicated a negative result; the 0 . 5 stands for trace , while 1 , 2 , and 3 , indicated that the intensity of the test line was similar to that of the respective 100 , 1000 , and 10 000 ng/ml spiked QC samples . Strips were scored in a blinded fashion by at least two individuals and in case of discrepancies a third person was consulted to conclude on the score . The classification of the trace result is decided later ( see ‘Model selection’ section ) based on the use of specific latent variable models . All other positive results ( scores 1 , 2 and 3 ) were merged into one positive category . LMM consists of a structural model for the latent infection states ( analogous to the latent classes in Latent Class Analysis ) and a measurement model for the observed indicators ( these are the four diagnostic tools: the average of two KK measurements on three consecutive days and the CCA ) , conditional on latent infection state . Let Y = ( yi1t , … , yiPt ) be a response pattern for the ith individual at time t on P observed binary indicators ( in this study P = 4 binary diagnostic tests ) with values ‘0’ and ‘1’ indicating negative and positive diagnostic test results , respectively . We assume that for each individual the true underlying infection state , at each discrete time point t ( where t = 1 , …T and T = 3: baseline , nine weeks and two years ) is explained by a latent categorical variable denoted by C with two latent infection states ( i . e . those with S . mansoni infection and those without ) . The responses to the ( P×T ) y indicators are assumed to be independent conditional on the latent infection state membership which in our analysis implies that the results from the four diagnostics are assumed to be independent conditional on the true underlying infection state both within and across time points . In addition , the latent categorical variable Ci , t depends on Ci , t-1 but not on earlier latent categorical variables , known as the first-order Markov property . Under these assumptions , the probability of observing a particular response pattern Y for a randomly selected individual i is:where represents the baseline infection state prevalence at the first time point ( i . e . baseline ) , and represents the transition probability to latent infection state jt at time t conditional on membership in latent infection state jt-1 at time t-1 . represents the diagnostic specificity for infection when the probability of the p diagnostic test is negative conditioned on Ci , t representing the ‘Not Infected’ latent state . Similarly , the diagnostic sensitivity for infection of each of the four diagnostics is obtained when the probability of the p diagnostic test is positive conditioned on Ci , t representing the ‘Infected’ latent state . Figure 1 presents our LMM in a path diagram . We fitted several LMMs using MPLUS v . 6 . 1 ( Muthén & Muthén , Los Angeles ) [31] with full information maximum likelihood in which we assumed that missing data were missing at random and making maximum use of data from individuals with incomplete data at the time points under study , for two different age groups: a ) n = 167 children of age 7–16 years old and b ) n = 273 adolescents and adults of age 17–76 years old . It is well known that different contact patterns with infected water bodies and consequently acquired exposure , immunity and susceptibility to infection might be experienced by different age groups [32] . We thus estimated the following sets of parameters for these two age groups: We selected the LMMs that optimally combined goodness of fit and parsimony as measured using the Akaike Information Criterion ( AIC ) and the Bayesian Information Criterion ( BIC ) . [33] Both AIC and BIC are penalized-likelihood criteria . AIC is an estimate of a constant plus the relative distance between the unknown true likelihood function of the data and the fitted likelihood function of the model , so that a lower AIC means a model is considered to be better . BIC is an estimate of a function of the posterior probability of a model being true , under a particular Bayesian framework , so that a lower BIC means that a model is considered to be better . BIC penalizes model complexity more heavily and thus , whenever there is an inconsistency AIC indicates a preference for a more complex model than BIC [34] . Both criteria are based on assumptions and asymptotic approximations . Each , despite its heuristic usefulness , has therefore been criticized as having questionable validity for real-world data [34] , [35] . In the current study , we examined AIC and BIC as well sample-size-adjusted BIC among models that were considered biologically plausible for the epidemiological settings under study ( given the effect of one and two PZQ doses on schistosomiasis cure rates within the studied age groups ) . We first tested whether the interpretation of a ‘trace’ CCA test line should be classified as positive or negative as several studies have reported on the ambiguity of infection status among those classified as ‘trace’ [16] , [17] , [36] , [37] . For this problem we used a one factor analysis model [38] on the baseline data , to explore the interrelationships among the four diagnostic tests for both age groups , treating the three results of CCA ( ‘negative’ , ‘trace’ and ‘positive’ ) as nominal ( unordered ) . We plotted the posterior distribution of the latent variable given the three possible responses to CCA for children ( Figure 2 ) and adolescents and adults ( Figure 3 ) , after estimating the one factor analysis model . Having observed the similarities in the posterior distributions for the ‘trace’ and ‘negative’ categories in Figures 2 and 3 , we decided to treat ‘negative’ and ‘trace’ CCA as a single ‘negative’ category for the rest of this analysis . Similar statistical analysis might be useful for data related to point of care-CCA in order to further validate results and proceed with generalization of recommendations about trace results for the CCA in the field . We subsequently fitted LMMs where we initially tested whether the item response probabilities ( ρ's , i . e . the sensitivities and specificities ) assumed to be the same at baseline and two years but allowed to vary at nine weeks for both KK and CCA tests . In other words we tested the hypothesis of measurement invariance which assumes the equality of the parameters of the measurement model i . e . the conditional item response probabilities for the latent infection states at the different time points . We did not test for different ρ's at each of the three time points ( baseline , nine weeks and two years ) because if they did differ at each time point the meaning of the latent infection states would have changed over time making the transition probabilities τ's uninterpretable . The reason for this is that along with interpreting quantitative change over time in latent infection state membership ( i . e . through the τ's ) , it also becomes necessary to interpret change over time in the meaning of the latent infection state ( i . e . which latent infection state represents the ‘true’ negatives and ‘true’ positives at each time point ) . We also tested whether the provision of one or two PZQ doses affected the transition probabilities τ's in the non-treatment intervals ( nine weeks and two years ) or only in the treatment interval ( baseline to nine weeks ) since there might be long-term benefits of two PZQ doses in the absence of snail control as in this community . Equal transition probability τ's across each of the treatment intervals under study was not tested because this was not consistent with prior knowledge of the MDA impacts . Although there was information identifying the household of each individual in the study , there were not enough data within each of the age groups studied to take into account between-household variability in the considered models .
The observed numbers of “true” and “false positives” and “true” and “false negatives” , and corresponding observed sensitivity and specificity , for the CCA diagnostic test are presented assuming , for illustration , the combination of three duplicate KK measures over the three consecutive days as being both 100% sensitive and 100% specific ( Table 1 ) . When the sensitivities and specificities of single-day double KK results were estimated by fitting LMMs , the information criteria indicated that , for both age groups , the transition probability matrices [for the baseline-to-nine weeks follow-up ( treatment interval ) and for the nine weeks-to-two years follow-up ( non-treatment interval ) ] depended on the provision of one or two PZQ doses . Similarly , the measurement invariance hypothesis for CCA was accepted based on the information criteria; however , the measurement invariance hypothesis was rejected for KK ( i . e . the diagnostic performance of KK was found to vary over time but that of CCA was not ) . For further details on final model selection , see Table S1 . The estimated sensitivities and specificities for CCA and KK tests , from the best LMM for each age group are presented in Table 2 . Based on these estimates and if one took the crude average of the estimated sensitivities and specificities of the two diagnostic measures and time points under examination , CCA overall was found to be more sensitive but less specific than double KK from a single faecal sample for both age groups . The sensitivity of double KK from a single faecal sample was found to be lower at nine weeks than at baseline and two years for both age groups in all three days for adolescents and adults and for two of the three days for children . The estimated transition probabilities between the latent infection states over time are presented in Table 3 . Diagonal elements represent the probabilities of remaining in the same latent infection state at time t as at the previous time ( t-1 ) . The majority of adolescents and adults apparently remained , over the non-treatment interval , in the same latent infection state . Among those infected at baseline in both age groups , two PZQ doses produced a higher probability of clearance of infection at nine weeks than just one PZQ dose; for children these probabilities were 0 . 717 ( 2 doses ) and 0 . 458 ( 1 dose ) ( see Table 3A ) , while for adolescents and adults these probabilities were 0 . 846 ( 2 doses ) and 0 . 561 ( 1 dose ) ( see Table 3B ) . The S . mansoni prevalence estimate for each time point and each age group is presented in Figure 4 with 95% confidence intervals . Following treatment the prevalence of S . mansoni infection decreased dramatically between baseline to nine weeks for both age groups . Among children there is a substantial rebound by two years . These patterns are also reflected in the transition probabilities in Table 3 . Although visual examination of Table 4 superficially suggests that the LMM estimate conflicts with that obtained based on the 6 KK measurements ( assumed positive if one or more of the six measurements was positive ) , further calculations are required to underpin full interpretation of the results . For instance , for children at baseline , the estimated prevalence based on the LMM is 92 . 6% . One would expect the following percentage of true positives:where 0 . 990 , 0 . 938 and 0 . 960 are the estimated sensitivities of two KK measurements from days one , two and three , respectively One would expect the following percentage of false positives:where 0 . 751 , 0 . 641 and 0 . 709 are the estimated specificities of two KK measurements from days one , two and three , respectively . Thus , based on the estimated sensitivities , specificities and the assumption that results from days one , two and three , were independent , conditional on the true status , the LMM predicts that the estimated prevalence based on 6 KK measurements would be 97 . 6% ( 92 . 6% true positive and 5 . 0% false positives ) . This is highly consistent with the estimate obtained from the 6 KK measurements: 97 . 6% . For the remaining time points and age groups please see in Supporting information .
This study analysed a longitudinal detailed dataset from Uganda in which one or two doses of PZQ treatment were provided at baseline using LMM that accounted for the longitudinal study design and the measurement error in the diagnostic methods under study . Our primary objective was to assess the CCA diagnostic accuracy at baseline and at two follow-up times after treatment but we also evaluate double KK faecal slides over three consecutive days for S . mansoni infection . To our knowledge , this is the first study which provides rigorous model-based diagnostic performance of CCA and single-day double KK measurements over three consecutive days for the diagnosis of S . mansoni infection in two different age groups pre- and post- PZQ treatment . CCA's diagnostic performance was found to be constant over time and overall approximately 90% sensitive but less specific than single-day double KK faecal slides for S . mansoni infection in both age groups . Day-to-day variation in faecal egg output among Schistosoma parasites [7]–[9] has been shown to be greater [11] and with lower sensitivity of KK after PZQ treatment [8] . The single-day double KK sensitivity is likely to depend strongly on the observed prevalence [8] , [15] , [39] . Our study confirmed these findings and arguments and showed clearly that sensitivity of single-day double KK was much lower at nine weeks than at baseline and two years for both age groups in all three days for adolescents and adults and for two of the three days for children while its specificity increased after PZQ treatment ( Table 2 ) . These findings bridge existing gaps in schistosomiasis diagnostics research such as for instance the lack of CCA evaluation in adolescents and adults and the lack for evidence for its capacity to determine if a person has been cured after treatment , as previously highlighted [4] , [17] . The current analysis provides model-based estimates of sensitivity and specificity and their uncertainties ( through the provision of 95% confidence intervals ) without assuming any gold standard diagnostic test in the statistical analysis . The exact numbers of false positive and false negative results are almost always unknown and thus in the current study we estimated rather than assumed values for the parameters displayed in Tables 2 and 3 ( sensitivities , specificities and transition probabilities ) [40] , [41] . Without quantification of the uncertainties regarding the performance of the key diagnostic tests , generalization of epidemiological results and development of useful recommendations for which diagnostics to use and at which stages of schistosomiasis control are hampered . This approach evaluates the risk of potential misinterpretation with regards to diagnosis of S . mansoni infection through CCA or KK in this endemic setting pre- and post- PZQ treatment as the numbers and infection intensities are brought down [12] . For instance , results in Table 1 demonstrate that by using 6 KK measurements over three days as the gold standard ( i . e . assuming 100% sensitivity and 100% specificity ) , the resulting empirical estimates of CCA sensitivity and specificity are mistakenly shown to vary over time . We do not expect that the clearance of the antigen is influenced by treatment . Furthermore , the hypothesis of measurement invariance was not rejected based on information criteria for the fitted LMMs ( see Text S1 and Table S1 in Supporting Information ) . Glinz et al . 2010 [42] discussed possible reasons for false positive diagnoses from KK tests . Results from model in Table 2 clearly indicate that the specificity of single-day double KK measurements is lower than 100% . This means that despite highly qualified and skilled co-workers , contamination of stool sieves in the field and data entry errors cannot completely be avoided . Because our estimated specificities of single-day double KK measurements were less than 100% , the estimated ‘true’ S . mansoni prevalence ( Figure 4 ) is lower at each time points than the estimated prevalence obtained assuming that any individual with positive results on one or more of the six KK tests conducted over three consecutive days was infected ( Tables 1 and 4 ) . This is in accordance with work on diagnostic performance of KK for animal schistosomiasis infection [14] . Previous work using stochastic models have demonstrated that the sensitivity of the KK would vary according to the number of stool samples provided [7]–[9] and the ‘true’ S . mansoni prevalence at baseline can be calculated using the De Vlas pocket chart [8] . For the children group , the chart is not applicable since the observed prevalence in this study is beyond the limits where the De Vlas model is valid and thus we cannot compare it with the estimates of ‘true’ prevalence from our model ( Figure 4 ) . For the adolescents and adults group however estimates of the ‘true’ prevalence ( Figure 4 ) were not consistent with this chart because it was based on an assumption of 100% KK specificity . The estimated transition probability matrices ( Table 3 ) provide parsimonious yet important insights into the re-infection and cure rates in the two age groups . The cure rate was higher in adolescents and adults than in children following treatment . This can be explained by the fact that those infected in the older age group had lower burdens than the infected children and would be therefore more likely to become negative . From an immunological perspective view , it can be argued that the older age group are more likely to have developed protective immune response and are therefore more efficient in affecting and killing the worms [43] , [44] . The quantification of the additional benefit of a second PZQ dose in reducing infection levels for both age groups was demonstrated by higher transition probabilities from infected to non-infected among those who received two PZQ doses compared to those who received one PZQ dose within the treatment interval . For the non treatment interval ( i . e . between nine weeks and two years ) there were no differences in reinfection or cure rates among those who received one or two PZQ doses . As the LMMs are estimated using an iterative algorithm , at each step of the algorithm , estimated values very close to 0 or 1 can create estimation instability and therefore are automatically from MPLUS fixed to 0 and 1 respectively in order to avoid non-convergence of the estimation algorithm . Consequently , such values should be treated with caution due to computational limitations in these categories during the model estimation . Finally , we recognize that the results of this study depend upon the assumption of conditional independence assumed by the models fitted here . Once one conditions on the latent infection state , we believe though that there are good reasons to assume that KK and CCA are independent as Schistosoma eggs and antigens are excreted through different routes in the human body , for instance . To conclude , in the absence of a diagnostic gold standard this study has demonstrated that LMMs can be useful for the evaluation of available diagnostic tools for S . mansoni infection . More generally , we recommend LMMs to be used for the evaluation of diagnostic tests of other diseases without gold standard diagnostic tools whenever longitudinal data are available as such modelling permits questions about changes in true infection states and test the measurement invariance hypothesis of the diagnostic tests of interest over time - making them very useful tools indeed for control program M&E research [21] . Further work in evaluating the trace result and the ability of CCA to quantify intensity of infection is also warranted in the M&E of schistosomiasis control programs and dynamic latent factor models ( such models would assume continuous hypothetical constructs or typologies-i . e . intensity of infection ) might be appropriate statistical methods for the analysis of relevant data . Similar studies should be considered at other sites in order to build on our results . We found that the CCA diagnostic performance remained constant after provision of PZQ treatment and that the test is overall more sensitive but less specific than single-day double KK for the diagnosis of S . mansoni infection . In line with the results from our study and those of a recent multi-country cross-sectional study which showed that for lower S . mansoni intensity settings the CCA sensitivity was demonstrated to be higher than KK [17] , we recommend that CCA to be used for mapping surveys of S . mansoni infection . As public health measures are aimed at the elimination of residual foci of schistosomes , data generated using diagnostics with high specificity will be required to avoid unnecessarily prolonging MDA and wasting scarce resources [5] . Detection of parasite-specific DNA [45] , [46] or circulating anodic antigen in serum or urine [47] might present alternative opportunities in schistosomiasis elimination programs and further evaluations of these diagnostics merit attention . | Schistosomiasis remains one of the most prevalent parasitic diseases in developing countries , with Schistosoma mansoni being the most widespread of the human-infecting schistosomes . For the routine surveillance of human S . mansoni infection more “field-applicable , ” sensitive , and cost-effective diagnostics that replicate faecal samples over several consecutive days [the Kato-Katz ( KK ) method] , are needed . We propose a statistical modelling framework in order to evaluate the diagnostic performance of the urine strip test for Circulating Cathodic Antigen ( CCA ) and single-day double KK measurements over three consecutive days for the diagnosis of S . mansoni infection in two different age groups from Uganda pre- and post- praziquantel ( PZQ ) treatment . We demonstrate that CCA is an appropriate tool for mapping surveys of S . mansoni infection . Our findings should allow for evaluation of the risk of potential misinterpretation with regards to diagnosis of S . mansoni infection through CCA or KK in this endemic setting pre- and post- PZQ treatment as the numbers and infection intensities are brought down , bridging existing important gaps in schistosomiasis diagnostics research . More generally , the proposed statistical analysis can reveal important biological insights from other diseases without gold standard diagnostic tools whenever longitudinal data are available . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | A Latent Markov Modelling Approach to the Evaluation of Circulating Cathodic Antigen Strips for Schistosomiasis Diagnosis Pre- and Post-Praziquantel Treatment in Uganda |
Major histocompatibility complex class I ( MHC-I ) molecules present antigenic peptides to CD8+ T cells , and are also important for natural killer ( NK ) cell immune surveillance against infections and cancers . MHC-I molecules are assembled via a complex assembly pathway in the endoplasmic reticulum ( ER ) of cells . Peptides present in the cytosol of cells are transported into the ER via the transporter associated with antigen processing ( TAP ) . In the ER , peptides are assembled with MHC-I molecules via the peptide-loading complex ( PLC ) . Components of the MHC-I assembly pathway are frequently targeted by viruses , in order to evade host immunity . Many viruses encode inhibitors of TAP , which is thought to be a central source of peptides for the assembly of MHC-I molecules . However , human MHC-I ( HLA-I ) genes are highly polymorphic , and it is conceivable that several variants can acquire peptides via TAP-independent pathways , thereby conferring resistance to pathogen-derived inhibitors of TAP . To broadly assess TAP-independent expression within the HLA-B locus , expression levels of 27 frequent HLA-B alleles were tested in cells with deficiencies in TAP . Approximately 15% of tested HLA-B allotypes are expressed at relatively high levels on the surface of TAP1 or TAP2-deficient cells and occur in partially peptide-receptive forms and Endoglycosidase H sensitive forms on the cell surface . Synergy between high peptide loading efficiency , broad specificity for peptides prevalent within unconventional sources and high intrinsic stability of the empty form allows for deviations from the conventional HLA-I assembly pathway for some HLA-B*35 , HLA-B*57 and HLA-B*15 alleles . Allotypes that display higher expression in TAP-deficient cells are more resistant to viral TAP inhibitor-induced HLA-I down-modulation , and HLA-I down-modulation-induced NK cell activation . Conversely , the same allotypes are expected to mediate stronger CD8+ T cell responses under TAP-inhibited conditions . Thus , the degree of resistance to TAP inhibition functionally separates specific HLA-B allotypes .
MHC-I molecules play a pivotal role in immune surveillance of intracellular pathogens by presenting antigenic peptides to cytotoxic T cells ( CTL ) . They also function to regulate natural killer ( NK ) cell activity by engaging NK cell receptors including KIR3DL1 [1] , KIR2DL1/2/3 [2] , CD94-NKG2A [3] and KIR3DS1 [4 , 5] . MHC-I molecules have strong influences on disease progression in a number of infectious diseases and cancers [6 , 7] . In many cases , the peptide-binding characteristics of individual MHC-I proteins are the major factor that determines immune control of diseases , but other characteristics of the MHC-I molecules , such as those relating to variations in the assembly and stability of individual MHC-I molecules , may also have an influence on disease outcomes . Intracellular proteins are generally degraded into peptide fragments by the ubiquitin-proteasome system [8] . Peptides that bind MHC-I molecules are typically translocated into the ER lumen by the transporter associated with antigen processing ( TAP ) and then loaded onto MHC-I molecules with the help of other components of the peptide-loading complex ( PLC ) , including tapasin , calreticulin and ERp57 [9] . Empty forms of MHC-I molecules are less thermostable than peptide-filled versions of MHC-I molecules [10–12] . ER quality control , including interactions with the PLC and calreticulin-mediated retrieval [13] , contributes to the intracellular retention of empty forms of MHC-I molecules . Additionally , tapasin and the tapasin-related protein ( TAPBPR ) edit and proofread the MHC-I peptide repertoire by replacing suboptimal low affinity peptides with optimal high affinity peptides [14–20] that can mediate more durable CD8+ T cell responses . In general , an intact PLC is essential for efficient peptide assembly with MHC-I molecules and successful ER quality control . However , individual MHC-I allotypes are known to have different requirements for each component of the PLC . For example , high cell surface expression of some human MHC-I ( HLA-I ) allotypes is observed in tapasin-deficient cells , whereas other allotypes are poorly expressed [16 , 21 , 22] . There are known differences in steady state binding of HLA-I molecules to TAP [23] . There are also known allomorph-specific differences in proteasome-dependence [24] . TAP is thought to be the major cellular source of peptide for assembly of most MHC-I molecules . In TAP-deficient cells , MHC-I cell surface expression is generally severely compromised [25–27] . Many viruses down-regulate or inhibit TAP to evade CTL responses [28 , 29] . In previous in vitro studies , we found that HLA-B allotypes display a hierarchy of refolding efficiencies and thermostabilities of heavy chains with β2-microglobulin ( β2m ) in the absence of peptide [12 , 22] , suggesting distinct intrinsic stabilities of empty forms of HLA-B . Molecular dynamics stimulations have also indicated that empty forms of some HLA-B molecules are more disordered than others [30–32] . Hein et . al . have shown that increasing intrinsic stability of H2-Kb-β2m complex by connecting the α1 and α2 helices with a disulfide bond close to the F-pocket , allowed suboptimally loaded forms of H2-Kb to bypass all cellular quality control steps in TAP-deficient cells [33] . Thus , in the trafficking process , the stability of empty heavy chain-β2m complexes is a key factor that determines the fate of MHC-I molecules in TAP-deficient cells . These findings raised the question of whether empty and suboptimally loaded forms of the more thermostable HLA-B allotypes can bypass ER quality control , traffic to the cell surface and maintain an increased steady-state presence there . Additionally , there can be influences of MHC-I peptide-binding specificities upon HLA-I cell surface expression levels under different conditions . It is known that peptides containing proline at the P2 or P3 position are poorly transported by TAP [34 , 35] , making it possible that MHC-I allotypes with these binding preferences ( for example , HLA-B allotypes of the B7 supertype [36] ) are more reliant on additional/alternate sources of peptide , and will have reduced sensitivity to TAP inhibition . Based on these observations , we hypothesized therefore that cell surface expression of MHC-I molecules would be differently dependent on TAP ( the major source of MHC-I peptides ) , based on the intrinsic stabilities of their empty forms and peptide-binding specificity differences . As described below , our studies revealed differential expression levels of HLA-B allotypes on the surface of TAP-deficient and TAP-inhibited cells . Intrinsic stability of the empty form as well as peptide-binding preferences determine cell surface expression levels under TAP-deficiency conditions . Furthermore , we showed that cells expressing HLA-B molecules with Bw4 epitopes that are resistant to inhibition of TAP are more resistant to the activation of KIR3DL1+ NK cells under TAP-inhibited conditions . Together , our findings indicate that HLA-I molecules have evolved to assemble via distinct pathways , which are allotype dependent , as a way to counter pathogen evasion strategies that target the conventional assembly pathway .
In TAP-deficient cells , where the majority of peptides are prevented from entering ER , most HLA-I molecules are empty or suboptimally loaded and HLA-I cell surface expression is generally significantly reduced [25–27] . We expected that when peptide supply is highly deficient in the ER , allotypes with higher intrinsic stabilities of their empty forms might have a better chance to bypass the quality control system as empty molecules or after being loaded with suboptimal peptides to become expressed on the cell surface . To examine whether HLA-B allotypes differ in their abilities to become expressed on the surface of TAP-deficient cells , several HLA-B allotypes that occur at the highest frequencies in United States populations were expressed in the TAP1-deficient human melanoma cell line SK-mel-19 ( SK19 ) [37] or in a TAP2-deficient human fibroblast cell line STF1 [38] using the previously described retroviral infection method [22 , 39] . Cell surface expression of HLA-B allotypes was analyzed by flow cytometry after staining with W6/32 , which recognizes different HLA-I allotypes with similar affinities . HLA-B allotypes showed large variations in cell surface expression in SK19 cells and STF1 cells ( Fig 1A and 1B ) . Cell surface expression of HLA-B*57:03 , B*35:03 , B*15:01 , B*35:01 , and B*15:10 was over 10-fold higher than the cell surface expression of the endogenous HLA-I of SK19 cells and over 5-fold higher than the cell surface expression of the endogenous HLA-I of STF1 cells ( Fig 1A and 1B ) . Cell surface expression of B*44:03 , B*58:02 and B*44:02 was very low or undetectable in STF1 cells , and less than two-fold above endogenous HLA-I cell surface expression in SK19 cells ( Fig 1A and 1B ) . Other HLA-B allotypes showed intermediate phenotypes ( Fig 1A and 1B ) . In general , there was poor correlation between exogenous HLA-I cell surface expression assessed by flow cytometry ( Fig 1A and 1B ) and total cellular expression assessed by immunoblotting analyses for HLA-I heavy chains ( S1A and S1B Fig ) . For SK19 cells or STF1 cells with HLA-B that were detectable at low or high levels on the cell surface , overexpression of exogenous HLA-B molecules did not induce any consistent unfolded protein response ( UPR ) compared with vector-infected cells , as assessed by immunoblots for BiP ( S2 Fig ) , induction of which is an UPR indicator [40] . There was a strong correlation between HLA-B cell surface expression levels in STF1 cells and those in SK19 cells ( Fig 1C ) , suggesting that the HLA-B cell surface expression differences were not cell dependent , but rather were TAP-deficiency dependent . Supporting the latter possibility , we have previously shown small differences in the cell surface expression of the HLA-B allotypes in TAP-expressing cells such as a CD4+ T cell line , CEM [22] . To verify that the measured W6/32 signals in SK91 and STF1 cells reflect the intended HLA-B signals rather than any other possible signals , HA-tagged versions of selected HLA-B that were detectable at high or low levels in SK19 and STF1 cells were constructed and expressed in SK19 cells by retroviral infection . An antibody against the HA epitope tag was used to test cell surface or total HA-tagged HLA-B ( HA-HLA-B ) expression . The HA-HLA-B versions maintained the same expression phenotypes as their untagged counterparts ( Fig 1D , S1C Fig ) . To confirm varying TAP-dependencies of HLA-B cell surface expression , we examined TAP1-mediated cell surface induction of HLA-B molecules following further infection of selected SK19-HLA-B cell lines with a TAP1-encoding retrovirus ( S3A Fig ) . There was an inverse correlation between the extent of TAP1-mediated induction ( +TAP1/-TAP1 ) and cell surface expression under TAP1-deficiency conditions ( Fig 2A and 2B ) . TAP1 expression was also reconstituted in SK19 cells expressing the HA-HLA-B ( S3B Fig ) . There was again an inverse correlation between the extent of TAP1-mediated induction ( +TAP1/-TAP1 ) and cell surface expression under TAP1-deficiency conditions ( Fig 2C ) . To validate the TAP-dependency results , TAP1 was knocked-down in a TAP-sufficient easily–transfectable cell line , Hela . TAP1-knock down ( KD ) or parent Hela cells were infected with retroviruses encoding selected HLA-B allotypes that were detectable at high , low or intermediate levels in TAP1 and TAP2-deficient cells ( as shown in Fig 1A and 1B ) . The allotypes expressed at high levels in SK19 and STF1 cells were down-modulated to a lesser extent by TAP1 knockdown compared to the HLA-B allotypes expressed at low levels in SK19 and STF1 cells ( Fig 2D and 2E ) , consistent with the conclusion from TAP induction experiments . Thus , HLA-B allotypes have differential resistance to inhibition of TAP ( RIT ) phenotypes . Higher intrinsic stability of the empty form , measured for many tapasin-independent allotypes [12 , 22] , would also favor a higher efficiency of peptide loading and thus cell surface expression under TAP-deficiency conditions . Zernich et . al . [41] attributed the advantage of B*44:05 cell surface expression under conditions of limiting peptide supply to the high peptide loading efficiency of nascent B*44:05 , which also causes its tapasin independency [16 , 22 , 41] . The structural similarities between the F-pockets of B*44:05 and B*57:03 ( the presence of Y116 ) might confer efficient peptide loading to both allotypes , while residue 116 is a D in B*44:02 and S in B*57:01 . Differences in peptide loading efficiencies between B*57:03 and B*57:01 could explain the differences in tapasin- and TAP-dependencies of these two closely-related allotypes , which differ only in the F-pocket regions , at positions 114 and 116 . While there is a partial positive correlation between TAP-dependence and tapasin-dependence of HLA-B cell surface expression ( Fig 3A and 3B ) , some allotypes are clear outliers . Individual HLA-B allotypes have different dependencies on TAP and tapasin . Some highly tapasin-independent allotypes such as B*18:01 and B*40:01 , both members of the B44 supertype ( pink , favoring peptides containing glutamic acid at position 2 ( P2 ) ) , are more TAP dependent . Some highly tapasin-dependent allotypes such as B*51:01 , a member of the B7 supertype ( blue , similar to B*35:01 and B*35:03 , favoring peptides containing proline at P2 ) , are less TAP dependent ( Fig 3A and 3B ) . These findings indicate that , the underlying mechanisms of TAP-independence and tapasin-independence are not fully overlapping . Recent mass spectrometric studies have identified large numbers of HLA-I peptidomes for different allotypes . Comparisons of the anchor residue preferences based on peptide sequences mined from two recent datasets [42 , 43] revealed that RIT allotypes generally have higher P2 diversity than several other non-RIT HLA-B ( Fig 3C and 3D ) , which would also favor selection of TAP-independent peptide from unconventional sources . It is noteworthy that there is a strict conservation of P2 among members of the B44 supertype ( including B*44:02 , B*44:03 , B*18:01 and B*40:01 ( pink; Fig 3C and 3D ) ) compared to members of the B7 supertype ( including B*35:01 , B*51:01 and B*07:02 ( blue; Fig 3C and 3D ) ) . Glutamic acid is stringently conserved as a P2 anchor among these members of the B44 supertype , whereas proline , alanine , and other residues occurring at lower frequencies , are found as P2 anchors among members of the B7 supertype ( based on data from Ref . 42 ( Fig 3C ) and 43 ( Fig 3D ) ) . B*15:01 , another allotype with high RIT , also displays high sequence diversity at the peptide P2 position ( 53% Q , 15% L , 9% V , 6%I , 5% S , 12% other ) ( based on data from Ref . 42; Fig 3C ) . Although a large peptidome dataset is not available for HLA-B*57:03 , recent B*57:01 peptidome data indicate high diversity at the peptide P2 position ( based on data from Ref . 43; Fig 3D ) . Structural similarities between the B pockets of B*57:01 and B*57:03 ( the P2 binding pocket ) predict a high P2 diversity for peptides that bind B*57:03 , similar to B*57:01 . Based on prior studies [44–47] , signal peptides and hydrophobic peptides are expected to be a TAP-independent source of MHC-I peptides . We first examined the prevalence of anchor residues for TAP-dependent and RIT allotypes within human signal sequence datasets . Within known human signal peptide sequences ( www . signalpeptide . de ) , N-terminal prolines and alanines ( excluding the last 6 residues at the C-terminus , which cannot be a P2 residue for any HLA-I epitope ) , preferred anchor residues for the B7 supertype , are significantly more prevalent than N-terminal glutamic acid , the preferred anchor residue for the B44 supertype ( Fig 3E ) . The low prevalence of glutamic acid within signal sequences could explain why the TAP-dependence phenotypes of B*18:01 does not mirror its high tapasin-independence and stability [12 , 22] . Conversely , the higher prevalence of proline/alanine within signal sequences could explain why the TAP-dependence phenotype of B*51:01 is less stringent than predicted by its strong tapasin-dependence and lower stability [12 , 22] . Preferred P2 residues for other RIT allotypes , such as B*57:03 ( A/S/T ) and B*15:01 ( Q/L ) , are also highly represented within the N-termini of signal peptide sequences ( Fig 3E ) . Further , using the NetMHC algorithm [48 , 49] , epitope predictions were undertaken with the signal peptide sequences from the signal peptide database ( www . signalpeptide . de ) , for epitope estimation for several allotypes ( Fig 3F ) . Significantly more peptides with IC50 < 500 nM ( weak binders ) or < 50 nM ( strong binders ) were identified for B*35:01 , B*57 and B*15:01 compared to several members of the B44 supertype . We also examined the prevalence of anchor residues ( Fig 3G ) and predicted weak and strong binders ( Fig 3H ) for TAP-dependent and RIT allotypes within human transmembrane sequence datasets ( TMbase25 , ftp://ftp . ncbi . nih . gov/repository/TMbase/ ) . Similar trends were noted as with signal sequences . Thus , our data support the model that peptide loading in the ER contributes to ER exit of RIT allotypes , which is favored by the increased prevalence of peptides with an appropriate P2 residue within signal peptides or transmembrane domains . There is prior evidence for TAP-independent presentation of peptides derived from both of these sources [44–47] . Findings from Fig 3 suggest that signal peptides and protein transmembrane domain-derived peptides could contribute to cell surface HLA-B molecules of RIT allotypes . However , limitation in this pool could result in loading with suboptimal sequences or in partial escape of empty molecules to the cell surface . To test the extent of peptide-receptive cell surface HLA-B , brefeldin A ( BFA ) decay assays were further conducted in SK19-HLA-B cells that were pre-incubated in the presence or absence of relevant HLA-B-specific peptides . Since anterograde transport is blocked by BFA , and cell surface HLA-I internalization is expected to be more rapid for empty or suboptimally loaded HLA-I [50] , the peptide-inducible fraction of the cell surface RIT HLA-B provides an estimate of the fraction of empty or suboptimally loaded cell-surface HLA-B . Based on these analyses , about 30–40% of cell surface RIT HLA-B including B*35:01 ( Fig 4A ) , B*57:03 ( Fig 4B ) , B*15:01 ( Fig 4C ) and B*44:05 ( Fig 4D ) are estimated to be expressed in an empty or suboptimally loaded form in TAP1-deficient SK19 cells after overnight culture at 26 oC . Under this condition , empty MHC-I was previously shown to be induced at the cell surface and stabilized by exogenous peptides [25 , 50] . Interestingly , even after overnight culture at 37 oC , a condition under which empty MHC-I are generally labile , significant fractions ( ~20–30% ) of the RIT HLA-B allotypes were peptide-inducible ( Fig 4A–4D ) . In contrast , on the surface of TAP-sufficient cells , only a small percentage ( ~5% ) of HLA-B molecules are peptide receptive ( Fig 4E and 4F ) . Thus , TAP-deficiency induces expression of HLA-B that is partially peptide-receptive . To confirm the presence of suboptimally loaded HLA-B on the cell surface of TAP-deficient cells at 37°C , SK19 cells expressing different RIT HLA-B allotypes were stained with HC10 [51] , which detects empty or open HLA-I conformations [52] . Higher levels of HC10-reactive RIT HLA-B allotypes were detectable on the cell surface compared to other HLA-B allotypes ( Fig 4G ) . TAP1 supplementation generally reduced HC10-reactive RIT HLA-B , while simultaneously enhancing the W6/32-reactive forms , contributing to a net decrease in the HC10 / W6/32 ratios ( Fig 4H ) . In the classical secretion pathway , HLA-I molecules are transported through the Golgi-network to the cell surface . In this pathway , the quality control machinery will prevent suboptimally loaded HLA-I from migration into the medial Golgi apparatus where proteins are modified and become Endoglycosidase H ( Endo-H ) resistant . Since a subset of RIT HLA-I molecules are suboptimally loaded under TAP-deficiency conditions ( Fig 4A–4H ) , alternative non-classical secretion pathway might exist to transport suboptimally loaded HLA-I molecules to the cell surface [53] . To address this model , the Endo-H sensitivities of HLA-I molecules in TAP-sufficient CEM and TAP-deficient SK19 cells were assessed . As shown in Fig 4I , most of the HLA-I molecules from either cell surface or total lysate of CEM-B*35:01 cells are Endo-H resistant , indicating that , in the steady state , most HLA-I molecules in CEM cells are mature and they traffic to the cell surface largely through the conventional pathway ( Fig 4I ) . In contrast , a greater fraction of HLA-I molecules from SK19 cells expressing exogenous HLA-B molecules are Endo-H sensitive , suggesting that a larger fraction is ER-retained in SK19 cells compared to CEM cells . Interestingly , following surface biotinylation , a detectable portion of RIT HLA-B molecules on the surface of SK19 cells were found to be Endo-H sensitive , in contrast to the predominantly Endo-H resistant HLA-I of CEM-B*35:01 cells . On the other hand , consistent with flow cytometry data ( Fig 1A ) , cell surface expression of a highly TAP-dependent HLA-allotype B*44:02 was barely detectable following surface biotinylation and immunoblotting ( Fig 4I , lanes 13 and 14 ) . These findings suggest a non-Golgi route exists for the trafficking of a subset of HLA-I from the ER to the cell surface of SK19 cells . Taken together , the results reported above suggest that under TAP-deficiency conditions , although a fraction of HLA-B molecules are transported to the cell surface through the conventional pathway , a fraction of RIT-HLA-B molecules follow an alternative non-conventional secretory pathway to reach the cell surface . As an important component of the PLC , TAP becomes a target of immune evasion in many virus-infected cells and tumor cells . For example , the Epstein-Barr virus ( EBV ) -encoded lytic phase protein BNLF2a acts as a TAP inhibitor by arresting TAP in a transport-incompetent conformation [54] . We examined the effects of BNLF2a on cell surface down-modulation of HLA-B allotypes . Although BNLF2a was transduced to similar levels into CEM cells expressing different HLA-B allotypes ( Fig 5A ) , variable BNLF2a-induced HLA-B down-modulation was observed ( Fig 5B ) , consistent with the prior expression results in TAP-deficient cells ( Fig 1A and 1B ) . Similar results were obtained in K562 cells , which express no endogenous HLA-I ( Fig 5C and 5D ) . Thus , TAP-inhibition has differential effects on cell-surface expression of HLA-B allotypes . Cell surface HLA-I with Bw4 epitopes function as inhibitory ligands for NK receptor KIR3DL1 [1] . Down-modulation of HLA-I with Bw4 epitopes can induce NK cell activation via the disengagement of KIR3DL1 . We expected that under infection conditions which inhibit TAP function , cells expressing RIT HLA-B would be more resistant to NK cell lysis . For comparisons , we chose K562 cells expressing a highly TAP-dependent allele B*44:03 , and a RIT allele B*57:03 and cells subsequently infected with a retrovirus encoding BNLF2a . Cell surface expression of B*44:03 was more strongly decreased by BNLF2a than B*57:03 ( Fig 6A and 6B ) . After co-incubation with K562 cells , NK cells from PBMCs of three donors , D136 , D187 and D215 , were activated , and expression of IFN-γ was measured ( Fig 6C , Column 1 ) . Expression of B*57:03 and B*44:03 in K562 cells strongly inhibits KIR3DL1+ NK cell activation ( Fig 6C , Columns 2 and 4 ) . In B*44:03 expressing cells ( Fig 6C , Column 5 ) but not B*57:03 expressing cells ( Fig 6C , Column 3 ) , KIR3DL1+ NK cells activation was increased by BNLF2a expression , consistent with the reduced expression of B*44:03 compared to B*57:03 on the cell surface .
Although the specific epitopes presented by HLA-I allotypes are well studied , the influences of folding and assembly variations among HLA-I allotypes on immunity are poorly characterized . Under normal conditions that are suitable for peptide loading , the effect of folding and assembly variations might not be significant . However , their effects could be amplified under pathological conditions whereby the function of PLC is disrupted by viral infection or tumorigenesis . In support of our prediction , we found that HLA-B allotypes are expressed at different levels on the surface of TAP-deficient or TAP-inhibited cells . Our previous findings indicated that , in the absence of peptide , the refolding efficiencies and thermostabilties of HLA-B allotypes are quite variable [12 , 22] . Under a tapasin-deficient condition , the capacity for assembly was generally higher for allotypes that had high refolding efficiencies in the absence of a peptide ligand [22] . HLA-I molecules with higher intrinsic stabilities of their peptide-deficient forms were expected to breach ER quality control mechanisms and more readily survive unfavorable assembly conditions such as low peptide supply ( TAP-deficiency condition ) . However , we found that high stability of the peptide-deficient form alone is insufficient to induce the highest level of expression , as exemplified by the intermediate expression level of B*18:01 , for which the ER peptide supply is predicted to be highly limiting under TAP-deficiency conditions ( Fig 3F and 3H ) . Based on the findings in this study , we propose the following model: in normal cells when peptide is not limited for most allotypes , cell surface HLA-I molecules are generally loaded with optimal peptides as a result of the abundant peptide pool ( Fig 7A ) . Under a suboptimal condition where the assembly factor tapasin is deficient , the observed expression hierarchy is determined by intrinsic stabilities and peptide loading efficiencies ( Fig 7B ) [22] . Under a third condition where peptide is highly limited due to TAP inhibition or deficiency ( Fig 7C ) , surface expression of the majority of HLA-B allotypes is strongly reduced . On the other hand , surface expression of RIT allotypes is less affected , because they have high intrinsic stabilities , high peptide loading efficiencies or broader specificities for peptides prevalent within signal sequences or other unconventional sources . Despite the expected role for peptides from unconventional sources as a determinant of TAP-independent HLA-B expression , many cell surface RIT HLA molecules are suboptimally loaded ( Figs 4 and 7C ) . Suboptimally loaded HLA molecules arise as a result of a limiting supply of peptides in the ER , an imperfect ER quality control system for the retrieval of suboptimally loaded molecules , and alternative ( non-Golgi ) pathways for transport to the cell surface ( Figs 4I and 7C ) [53] . The Endo-H sensitive pool of RIT HLA-B is particularly noteworthy ( Fig 4 ) , and suggestive of models of peptide loading within a non-conventional secretory pathway for nascent HLA-I molecules , previously described within professional antigen presenting cells ( APC ) [53] . Other cell types such as melanoma cell lines also appear to have such pathways ( Fig 7C ) . Although components of the PLC are very important for peptide loading to MHC-I molecules , unconventional antigen processing and peptide loading pathways do appear to widely exist ( Fig 7B and 7C ) . Among the tested HLA-B allotypes , B*35:01 , B*35:03 and B*15:01 are noteworthy for their high expression when either TAP or tapasin are deficient . Since inhibition of TAP and tapasin is a common evasion strategy used by pathogens and tumors [28 , 55] we propose that the folding and assembly characteristics of these allotypes have evolved to allow CD8+ T cell-mediated immune surveillance to persist in the face of pathogenic challenges to the conventional pathway . The B7 supertype is particularly noteworthy for the higher propensity for TAP-independent expression ( Fig 3 ) . Allotypes belonging to this supertype bind peptides with proline at P2 , which are highly disfavored for TAP-mediated transport [35] . In a recent study , we showed that , compared with other HLA-B , those belonging to the B7 supertype tend to be expressed at lower levels in normal human lymphocytes but not monocytes . Taken together with findings in this study , it appears that mismatch between TAP-transporter specificity and HLA-I peptide binding specificity causes suboptimal assembly and expression of allotypes belonging to the B7 supertype in some cell types , but confers an expression advantage under TAP-deficient or TAP-inhibited cells and possibly in professional antigen presenting cells that have specialized antigen acquisition pathways for HLA class I . While previously it was found that empty MHC-I molecules move to the surface of TAP-deficient cells only at sub-physiological temperature [50] , here we show that partially peptide-receptive forms of RIT HLA-B allotypes are expressed on the surface of TAP-deficient cells even at physiological temperature ( Fig 4 ) . Duration of HLA-I molecules on the cell surface is dependent on their stabilities [56] . HLA-I molecules with higher stability of their empty forms are also expected to be more stable on the cell surface in their empty forms . On the other hand , for many allotypes , the empty forms will be rapidly internalized and degraded at physiological temperature due to the relative instability . Empty or open MHC-I conformers have been drawing increasing attention in recent years . They are proposed to be ligands for many receptors , including KIR3DS1 [4 , 5] , KIR3DL2 [57] , KIR2DS4 [57] and LILRB2 [58] . Many of the described interactions with open MHC-I involve in vitro studies with acid-treated classical HLA-I . The natural prevalence of empty forms of classical HLA-I in cells is thus far poorly characterized . Under normal and TAP-deficiency conditions , RIT allotypes provide a natural source of partially empty class I , and might thus also be more efficient in triggering signals through receptors specific for open HLA-I . Our recent studies indicate that empty HLA-B*35:01 molecules on the cell surface can augment CD8+ T cell activation through enhanced engagement with CD8 [12] . Based on those findings , we expect that , under TAP-inhibited conditions , empty forms of all RIT HLA-B can synergize with reduced levels of antigenic peptide-bound versions to facilitate and maintain some level of CD8+ T cell surveillance of infections . Thus , although RIT HLA-I molecules may not show specific advantages under optimal antigen presentation conditions , they are expected to be more efficient in presenting TAP-independent peptides to CD8+ T cells in infection or tumor conditions involving TAP blockade . Nonetheless , it is important to note that viruses and cancers have developed many other strategies to evade immune recognition , such as the direct down-regulation of HLA-I expression and interference with IFN-γ signaling ( for example , [59] ) . Thus , cells expressing RIT HLA-B could still escape immune surveillance under other different pathogenic conditions . T cell epitopes associated with impaired antigen presentation ( TEIPP ) [60 , 61] are known to emerge under conditions of inhibited antigen presentation , including TAP-deficiency conditions . In fact , it is reported that CD8+ T cells responsive specifically to TAP-inhibited cells are widely prevalent in the human blood probably due to the prevalence of viruses that encode TAP inhibitors such as EBV , CMV and HSV [62] . Given the high expression levels and suboptimal peptides , RIT HLA-B molecules may contribute dominantly to the HLA-B-restricted CD8+ T cell repertoire against TEIPP ( including both self-peptides and viral epitopes ) under conditions where TAP expression is inhibited or TAP function is suppressed , an area for further assessment . Moreover , the prevalence of RIT HLA-B molecules might be a reason that there is only mild immunodeficiency in TAP deficient humans [63] , and RIT-HLA-I may be the dominant antigen presenting alleles in these patients . In conclusion , it is well recognized that pathogens have developed strategies to escape cytotoxic T cell surveillance by , for example , disrupting HLA-I assembly pathways [28 , 29] . It is now apparent that HLA-I molecules have also evolved to assemble via distinct pathways , which are allotype dependent , as a way to counter pathogen evasion strategies that target the conventional assembly pathway ( Fig 7 ) . Thus , the textbook-defined HLA-I assembly pathways are not fully applicable to all allotypes . In this study , we demonstrate that 15% of tested HLA-B allotypes are resistant to inhibition or deficiency in TAP , which is considered a central source of peptides for HLA-I assembly . Cell surface expression of several HLA-B allotypes is readily observable under TAP-deficiency conditions , and relates to HLA-B intrinsic stabilities , peptide loading efficiencies , peptide binding preferences and unconventional secretory pathways . Thus , TAP-independent pathways of antigen acquisition are quite broadly prevalent . RIT HLA-B molecules are expected to confer immune recognition advantages for the CTL response under TAP-inhibited conditions , via the mechanisms outlined above . Conversely , when TAP function is blocked , HLA-B allotypes with Bw4 epitopes that are strongly down-modulated confer induced abilities to mediate NK activation , via reduced KIR3DL1+ NK cell binding ( Fig 6 ) . Overall , the findings in this study point to important functional distinctions within the HLA-B locus that relate back to intrinsic structural features of the proteins and their intracellular assembly characteristics .
Blood was collected from consented healthy donors for functional studies in accordance with a University of Michigan IRB approved protocol ( HUM00071750 ) . All donors provided informed written consent . Human melanoma cell line SK-mel-19 ( SK19 ) [37] ( obtained from the laboratory of Dr . Pan Zheng ) , fibroblast cell line STF1 [38] ( obtained from the laboratory of Dr . Henri de la Salle ) , cervical cancer cell line Hela ( obtained from the laboratory of Dr . Oveta Fuller ) and ecotropic virus packaging cell line BOSC ( obtained from the laboratory of Dr . Kathleen Collins ) were grown in DMEM ( Life Technologies ) supplemented with 10% ( v/v ) FBS ( Life Technologies ) and 1× Anti/Anti ( Life Technologies ) ( D10 ) . T4-lymphoblastoid cell line CEM-ss ( CEM ) cells ( obtained from the laboratory of Dr . Kathleen Collins ) and chronic myelogenous leukemia cell line K562 cells ( obtained from ATCC; CCL-243 ) were grown in RPMI 1640 ( Life Technologies ) supplemented with 10% ( v/v ) FBS , 1× Anti/Anti , 2 mM glutamine ( Life Technologies ) and 10 mM HEPES ( Life Technologies ) ( R10 ) . The following monoclonal antibodies were used in this study: Pacific Blue-conjugated anti-human CD3 ( clone UCHT1; BioLegend ) , PE-Cy7-conjugated anti-human CD56 ( clone CMSSB; eBioscience ) , FITC-conjugated anti-human KIR3DL1 ( clone DX9; BioLegend ) , Alexa Fluor 700 conjugated anti-human IFN-γ ( clone B27; BioLegend ) , purified anti-HA . 11 ( Clone 16B12; BioLegend ) , anti-BiP ( Clone C50B12; Cell Signaling Technology ) , anti-GAPDH ( Clone 14C10; Cell Signaling Technology ) and anti-vinculin ( Clone E1E9V; Cell Signaling Technology ) . Dead cells were excluded from flow cytometric analyses with 7-amino-actinomycin D ( 7-AAD; BD Biosciences ) or the amine-reactive dye Aqua ( 405nm , Life Technologies ) . HLA-I antibodies W6/32 , HC10 and 171 . 4 were produced in the University of Michigan Hybridoma Core . The TAP1 antibody 148 . 3 was kindly gifted by Dr . Robert Tampé . All HLA-B alleles in the retroviral vector LIC pMSCVneo were prepared as described previously [22] . HA-tagged versions of HLA-B*35:01 , B*35:03 , B*57:01 , B*44:02 and B*4405 were prepared as described previously [64 , 65] . To prepare HA-tagged versions of HLA-B*15:01 , B*44:03 , B*57:03 and B*58:02 , corresponding clones from pMSCVneo [22] were digested with NaeI and XhoI to prepare the 3′ regions of these HLA-B ( encoding the portion of the protein downstream of the signal sequence ) . The B*35:01 signal sequence plus HA-tag was isolated by EcoRI and NaeI digestion of HA tagged B*35:01 . Finally , the HLA-B*15:01 , B*44:03 , B*57:03 and B*58:02 NaeI–XhoI fragments and the EcoRI-NaeI fragment from HA-B*35:01 were ligated into pMSCVneo ( cut with EcoRI and XhoI ) in a three-way ligation . Retroviruses were generated using BOSC cells and used to infect SK19 , STF1 , Hela , CEM or K562 cells . Cells were infected with retroviruses encoding the HLA-B molecules , selected by treatment with 1 mg/ml G418 ( Life Technologies ) , and maintained in 0 . 5 mg/ml G418 . Exogenous HLA-I expression was verified by immunoblotting analyses of cell lysates using the mouse anti-human monoclonal antibody 171 . 4 or anti-HA and secondary antibodies GαM-HRP ( Jackson ImmunoResearch Laboratories ) or GαM-IRDye 800CW ( LI-COR Biosciences ) . SK19 cells expressing exogenous HLA-B molecules were infected with the human TAP1-encoding retrovirus and selected by treatment with 1 μg/ml puromycin ( Sigma-Aldrich ) , and cells were maintained in 0 . 5 μg/ml puromycin . TAP1 expression in SK19 cells was verified by immunoblotting analysis of cell lysates using mouse anti-human TAP1 monoclonal antibody 148 . 3 [66] and secondary antibodies GαM-HRP or GαM-IRDye 800CW . The Western blots were developed for chemiluminescence using the GE Healthcare ECL Plus kit or scanned for IRDye fluorescence using Odyssey System ( LI-COR Biosciences ) . CEM and K562 cells expressing exogenous HLA-B molecules were infected with the BNLF2a-encoding retrovirus and selected by treatment with 1 μg/ml puromycin ( Sigma-Aldrich ) , and cells were maintained in 0 . 5 μg/ml puromycin . MSCV-N BNLF2a was a gift from Dr . Karl Munger [67] ( Addgene plasmid # 37941 ) . BNLF2a expression was verified by intracellular staining with primary antibody anti-HA and secondary antibody PE-conjugated goat anti-mouse IgG ( GαM-PE , Jackson ImmunoResearch Laboratories ) . TAP1 was knocked-down in Hela cells by using the CRISPR/Cas9 system based TAP1 Double Nickase Plasmid from Santa Cruz Biotechnology according to manufacturer’s protocol . Puromycin selection and limiting dilution was subsequently undertaken to obtain monoclonal TAP1-KD cell lines . TAP1 knockdown was verified by immunoblotting analysis of cell lysates using anti-TAP1 antibody 148 . 3 [66] and secondary antibodies GαM-HRP ( goat anti-mouse horse radish peroxidase ) and by intracellular staining with 148 . 3 [66] and secondary antibody GαM-PE . HLA-B alleles were expressed in Hela or Hela-TAP1-KD cells using the method described above . A total of 1×105−1×106 cells were washed with FACS buffer ( phosphate-buffered saline ( PBS ) , pH 7 . 4 , 1% FBS ) and then incubated with W6/32 or HC10 antibodies at 1:250 dilutions or anti-HA at 1:50 dilution for 30–60 min on ice . Following incubation , the cells were washed three times with FACS buffer and incubated with GαM-PE or GαM-PE-Cy7 at 1:250 dilutions for 30–60 min on ice . The cells were then washed three times with FACS buffer and analyzed using a BD FACSCanto II cytometer . The FACS data were analyzed with FlowJo software version 10 . 0 . 8 ( Tree Star , San Carlos , CA ) . Data are deposited in the Dryad repository: http://dx . doi . org10 . 5061/dryad . m4862mk [68] . The night before the experiment , cells were moved to 26°C or kept at 37°C . The next day , cells were washed with PBS , and the medium ( containing 100 μM peptide where indicated ) was added and cells were incubated at 26°C for 2h . Cells were then incubated at 37°C in the presence of 20 μg/ml brefeldin A ( BFA ) for an additional 2h and then harvested . The HLA-B signals were quantified by flow cytometry after staining with W6/32 and subtracting signals obtained from cells infected with a retrovirus lacking HLA-B . Peptide receptive HLA-I was quantified as ( MFI HLA-I ( +peptide ) –MFI HLA-I ( -peptide ) ) / MFI HLA-I ( +peptide ) *100 and averaged across 3–4 independent measurements for each condition . Peptides used ( S1 Table ) were B*57:03-restricted epitopes TSTLQEQIGW ( TW10 ) and KAFSPEVIPMF ( KF11 ) , B*44:05-restricted epitopes VEITPYKPTW ( VW10 ) and EEFGRAFSF ( EF10 ) , B*15:01-restricted epitopes LEKARGSTY ( LY9 ) and ILKEPVHGVY ( IY10 ) and B*35:01-restricted epitopes FPVRPQVPL ( FL9 ) and LPSSADVEF ( LF9 ) [64] . All peptides were purchased from peptide 2 . 0 ( Chantilly , VA , USA ) . All peptides are in the IEDB database except self-peptide LF9 . Cell surface proteins were biotinylated by incubating cells with 2mM EZ-Link NHS-PEG4-Biotin ( Thermo Scientific ) in PBS for 10 min at room temperature followed by three washes in PBS . After washing , labeled cells were lysed in lysis buffer ( 1× PBS , 1 mM phenylmethylsulfonyl fluoride , and 1% Triton X-100 ) for 1h on ice . The lysates were centrifuged at 13 , 000 g to remove cell debris . Biotinylated proteins were bound to streptavidin conjugated beads for 2 h at 4°C . Beads were washed three times with lysis buffer , and boiled for 10 min in the presence of denaturing buffer . As controls , total cell lysates were directly boiled for 10 min in denaturing buffer . The materials obtained from the beads and total cell lysates were split into two equal aliquots and one of the aliquots was digested with Endo-H ( New England Biolabs ) according to the manufacturer’s protocol . HLA-I molecules were separated by SDS-PAGE and then immunoblotted using the mouse anti-human monoclonal antibody 171 . 4 . Fresh blood collected from donors was subjected to centrifugation over a Ficoll-Paque Plus ( GE Healthcare Life Sciences ) density gradient , washed twice with PBS + 2% FBS and resuspended in R10 . Isolated PBMCs were cryopreserved in Recovery Cell Culture Freezing Medium ( Life Technologies ) . IFN-γ expression in NK cells was detected by intracellular cytokine flow cytometry . Briefly , frozen PBMCs ( 2 × 105 cells/well ) were incubated with K562 cells expressing or lacking HLA-B molecules at 1:1 ( PBMC:K562 ) ratio in 200 μL complete media in 96-well U-bottom plates . GolgiPlug ( containing brefeldin A , BD Biosciences ) was added at 1:1000 1h later . After incubation for an additional five hours , cells were stained with Pacific Blue-conjugated anti-CD3 , PE-Cy7-conjugated anti-CD56 and FITC-conjugated anti-KIR3DL1 mAbs for 30 minutes at 4°C , fixed in 4% paraformaldehyde for 10 minutes at room temperature , and permeabilized with 0 . 2% saponin for 10 minutes . Cells were then stained with Alexa Fluor 700-conjugated anti-IFN-γ for 30 minutes at 4°C and analyzed by flow cytometry . Statistical analyses ( ordinary one-way ANOVA analysis with Fisher’s LSD test ) were performed using GraphPad Prism version 7 . | Human leukocyte antigen ( HLA ) class I molecules present pathogen-derived components ( peptides ) to cytotoxic T cells , thereby inducing the T cells to kill virus-infected cells . A complex cellular pathway involving the transporter associated with antigen processing ( TAP ) is typically required for the loading of peptides onto HLA class I molecules , and for effective anti-viral immunity mediated by cytotoxic T cells . Many viruses encode inhibitors of TAP as a means to evade anti-viral immunity by cytotoxic T cells . In humans , there are three sets of genes encoding HLA class I molecules , which are the HLA-A , HLA-B and HLA-C genes . These genes are highly variable , with thousands of allelic variants in human populations . Most individuals typically express two variants of each gene , one inherited from each parent . We demonstrate that about 15% of tested HLA-B allotypes have higher resistance to viral inhibitors of TAP or deficiency of TAP , compared to other HLA-B variants . HLA-B allotypes that are more resistant to TAP inhibition are expected to induce stronger CD8+ T cell responses against pathogens that inhibit TAP . Thus , unconventional TAP-independent assembly pathways are broadly prevalent among HLA-B variants . Such pathways provide mechanisms to effectively combat viruses that evade the conventional TAP-dependent HLA-B assembly pathway . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"flow",
"cytometry",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"antigen",
"presentation",
"immune",
"cells",
"molecular",
"probe",
"techniques",
"immunology",
"immunoblotting",
"cytotoxic",
"t",
"cells",
"molecular",
"bio... | 2018 | Selected HLA-B allotypes are resistant to inhibition or deficiency of the transporter associated with antigen processing (TAP) |
Goat raising is a growing industry in Lao People’s Democratic Republic , with minimal disease investigation to date , especially zoonoses . This study determined the proportional seropositivity of two zoonotic diseases: Q fever ( causative agent Coxiella burnetii ) and Brucellosis ( Brucella species ) in goats across five provinces ( Vientiane Capital , Xayaboury , Xiengkhuang , Savannakhet and Attapeu ) . A total of 1458 goat serum samples were tested using commercial indirect ELISA for both pathogens , plus Rose Bengal agglutination test for Brucellosis . Overall individual seropositivity of C . burnetii was 4 . 1% and Brucella spp . was 1 . 4% . A multiple logistic regression model identified that province ( Vientiane Capital , p = 0 . 05 ) , breed ( introduced Boer mixed breed , p = 0 . 006 ) and age ( goats ≥3 years old , p = 0 . 014 ) were significant risk factors for C . burnetii seropositivity . The results of the survey indicated that province ( Vientiane Capital , p<0 . 001 ) , breed ( introduced Boer mixed breed , p<0 . 001 ) , production system ( commercial , p<0 . 001 ) , age ( adult , p = 0 . 004 ) , and farm size ( large , 0 . 001 ) were all significant risk factors seropositivity for Brucella spp . It was concluded that Lao goats have been exposed to both C . burnetii and Brucella spp . however the risk of clinical disease has not yet been determined and there is an urgent need to determine human health risks and economic losses caused by Q fever and Brucellosis .
Lao People’s Democratic Republic ( Laos ) is a landlocked country in the Greater Mekong Sub-region with an economy greatly dependent on agriculture [1] . Livestock have become increasingly important for improving rural livelihoods in Laos , providing a source of high quality protein , manure as fertiliser for plant growth , a means of household wealth storage , and income to buy food , education and healthcare [2] . Goats are becoming increasingly important for smallholder food farming in Laos [3 , 4] , providing livestock products that are perceived to require lower inputs than cattle and buffaloes . Furthermore , following regional economic growth there has been an increase in regional demand for goat meat in Vietnam and China , leading to rapidly increasing smallholder goat population and appearance of several commercial farms throughout Laos . Anecdotal reports suggest that there is no commercial milk or cheese production . The 2011 census reported that 45 , 000 farm households raised goats [3] , however , it is difficult to report accurately on diversity of goat farming in Laos as it is the smallest livestock sector and is not always included in demographic reports . There is a need to focus on goat and goat farmer health in Laos as this has largely gone without investigation . Human health is closely linked to livestock health . Healthy livestock can provide food , wealth and financial security , whereas unhealthy or diseased livestock cannot , and may be a reservoir for diseases infectious to humans ( i . e . , zoonoses ) . The close working relationship of farmers and their families with their animals allows for zoonotic disease transmission [1] with Coxiella burnetii ( causing Q fever in humans ) and Brucella melitensis considered potentially important bacterial zoonotic pathogens associated with goats in Laos [5] . Both agents can cause undulant fever and chronic disease in humans [5 , 6] . These pathogens have the ability to cause large-scale outbreaks due to their low infectious dose , resistance in the environment and ability to travel via aerosolisation of the pathogens [5–7] . Q fever and Brucellosis are difficult to diagnose and treat in humans due to their non-specific presentation and intracellular nature [5 , 7 , 8] . Furthermore , C . burnetii and Brucella spp . can economically impact rural livelihoods as they reduce productivity due to reproductive loss in livestock herds [5 , 8 , 9] . C . burnetii and Brucella spp . are considered biothreats and classified as “Select Agents” in the USA [10 , 11] . C . burnetii seroprevalence studies in Laos have revealed the pathogen is not widely distributed in cattle , despite the likelihood of an epidemiological hotspot in the Thailand-bordered province of Xayaboury [12 , 13] . Thailand is a significant trading partner with Laos , and C . burnetii antibodies are reportedly present in 4% of Thai goats [14] . Surveys for Brucella spp . antibodies in Laos has revealed only a limited distribution of this pathogen in the Lao cattle herd [12 , 13] . However , studies have revealed 1 . 4% of goats and 12 . 1% of goat herds in Thailand are seropositive for Brucella spp . [15] . Human Q fever and brucellosis case reporting is increasing in Thailand , with close contact with goat parturition material and consumption of raw goat meat considered as risk factors for contracting both diseases [14 , 16] . One Thai study estimated that 12 . 6% of occupationally-exposed persons are seropositive for C . burnetii antibodies in selected provinces [14] . A report from 2004 outlines two Brucellosis cases , the first report in scientific literature since the 1970s , with one case having a history of drinking raw goat’s milk [17] . Following an outbreak of undulant fever in 2006 in a province northeast of Bangkok , village members were tested for exposure to Brucellosis of which 43 . 5% had antibodies to B . melitensis , with risk factors including recent contact with goats during their parturition and eating raw goat meat [16] . There are no overall human population seroprevalence or incidence estimates for Q fever or Brucellosis in Thailand available in the literature . This increase in reporting could be caused by increased disease awareness rather than an increase in disease incidence . Even though there are no reports of human Q fever or Brucellosis in Laos to date , this may be due to under-reporting rather than lack of disease manifestation , as there have been no specific investigations into these pathogens reported in the literature to date . Considering the high risk of zoonoses due to factors including to the close working nature of farmers and livestock in Laos , porous borders , rapidly increasing smallholder goat population and appearance of commercial enterprises , it is important to investigate the prevalence of C . burnetii and Brucella spp . in Lao goats . This study reports on a serological survey of that aimed to determine the seroprevalence of C . burnetii and Brucella spp . antibodies in Lao goat herds in selected provinces and to identify potential risk factors associated with presence of infection . Furthermore , an understanding of infectious diseases in Lao goats enables development of disease prevention and control strategies and public health policies , supporting smallholder livestock farmers to increase their productivity , and therefore income , whilst minimizing human disease .
This study was conducted in compliance with State Acts and National Codes of Practice for Ethical Standards , with animal and human ethics approval obtained from The University of Sydney Ethics Committee ( project no . 2015/765 and 2014/783 , respectively ) . Verbal consent was obtained from all of the goat owners prior to the collection of the sample . This study was conducted in five provinces in Laos: Vientiane Capital , Xayaboury , Xiengkhuang , Savannakhet and Attapeu ( Fig 1 ) . Samples were collected between October 2016 and May 2017 . Vientiane Capital was selected due to the emergence of commercial goat enterprises and the convenience of proximity to the National Animal Health Laboratory ( NAHL ) . Xayaboury was selected as a follow up to a previous study suggesting there was an epidemiologic “hot spot” for Q fever in cattle located in districts bordering Thailand [12 , 13] . Samples were collected from Xiengkhuang , Savannakhet and Attapeu as part of necropsy training workshops run by the same team . To provide some guidance regarding the number of samples to be collected , sample size was calculated using the formula: n = ( Z2 P ( 1-P ) ) /e2 , where Z is the value from a standard normal distribution corresponding to the desired confidence level , P is the expected true proportion and e is the desired precision . A sample size to estimate individual level seroprevalence with a precision of 0±1% with an expected herd prevalence of 3 . 5% and confidence level of 95% was calculated using “AusVet Epitools” website [18] . As no goat seroprevalence studies for C . burnetii and Brucella spp . had been performed in Laos previously , estimated herd prevalence was extrapolated from caprine C . burnetii seroprevalence studies in neighbouring Thailand [14] . In 2011 , the Lao agriculture census estimated that there were 215600 goats in Laos [3] . This resulted in a target sample size of 1291 goats however this was an estimate for the whole of Laos rather than selected provinces/districts . Using this sample size guidance , a total of 69 villages within 15 districts were selected ( Fig 1 ) . Districts were chosen with close assistance from staff from the Department of Livestock and Fisheries in Laos ( DLF ) . Districts were visited if they had the following criteria: accessible via road vehicle; close working relationship with DLF staff; and households or farms that raised goats . Upon arrival at each district , the sampling team consulted provincial and district staff as to which villages raised goats . The sampling team visited participating different households and farms within each village following discussions between the sampling team , DLF staff , village chiefs and farmers . The team aimed to sample as many goats as possible within each village and household , this included every goat presented to the team . Blood ( 3–5ml ) was collected from the jugular vein into a sterile syringe and allowed to clot . The serum was removed and stored at 4°C for transport back to NAHL where it was centrifuged at 5000 rpm for 5 minutes and stored at -80°C until further analysis . Samples collected in 2017 obtained epidemiological data including date of collection , owner name , age via examination of teeth [19] and gender . Additionally , for the provinces of Vientiane Capital , Xayaboury and parts of Attapeu , breed ( native Kambing-Katjang or introduced Boer/ Boer mixed ) , the number of goats per household owned , and production system ( commercial enterprise with employed persons to raise goats or small holder , family raised household goats ) were recorded . However , no such data was recorded for the 704 samples collected in 2016 . Data was entered into Microsoft Excel and analysed using Stata/SE version 15 . 0 for Macintosh ( StataCorp , College Station , TX ) . For both pathogens , serological prevalence was calculated as the proportion of animals that had detectable antibodies in the same population , with 95% confidence intervals . For both C . burnetii and Brucella spp . , measures of association for categorical data were assessed using either Pearson’s Chi-squared test or Fisher’s exact test as appropriate . For calculations regarding C . burnetii seropositivity only , univariable logistic regression models were fitted to obtain unadjusted estimates of odds ratios ( OR ) . Furthermore , a multivariable logistical regression model was performed to determine factors independently associated with C . burnetii seropositivity . Only data with full epidemiological information ( 754 samples ) were included in this model , as the remaining samples were collected prior to 2017 without full epidemiological and clinical information . Variables with univariable significance ( p≤0 . 05 ) were entered into the multivariable model . All tests of significance were performed at 5% level of significance .
The distribution of provinces , districts and villages sampled are outlined ( Fig 1 , Table 1 ) . The number of serum samples collected per village ranged from 2 to 80 , with a median of 19 . A total of 1458 goats were sampled . Overall , 60/1458 ( 4 . 1%; 95% CI 3 . 0 , 5 . 0 ) of goats sampled were seropositive for C . burnetii antibodies ( Table 2 ) . Notably , the OR of individual goat C . burnetii seropositivity within Vientiane Capital was 33 . 4% ( 95% CI 4 . 6 , 243 . 7 , p = 0 . 001 ) and in Xayaboury was 8 . 4% ( 95% CI 1 . 0 , 67 . 6 , p = 0 . 046 ) respectively , with both statistically significant compared to the chosen reference province of Savannakhet ( Table 2 ) . Within Vientiane Capital , all districts had some seropositive animals ( Table 3 ) although there was a significant difference in seroprevalence between the districts sampled in Vientiane Capital ( p<0 . 001 ) , with the highest seroprevalence in Parkgnum ( 25 . 4% ) ( Table 3 ) . Vientiane Capital had 75% ( 8/12 ) villages with a seropositive result , with significant difference between villages ( p<0 . 001 ) including Ponsavan village with the highest seropositivity of 29 . 6% ( Table 3 ) . Within Xayaboury , Kentao ( 7 . 3% ) was the only district to display any seropositivity . Where full epidemiological information was available ( n = 744; 51% ) , univariable and multivariable logistic regression was performed . The OR for Boer crossbred goat seropositivity for C . burnetii was significantly greater when compared to native Kambing-Katjang goats ( OR 9 . 2; 95% CI 4 . 3 , 19 . 8 , p<0 . 001 ) ( Table 2 ) . The OR for goats sampled from commercial farms for C . burnetii seropositivity was significantly greater than goats sampled from smallholder farms ( OR 5 . 5; 95% CI 2 . 8 , 10 . 6 , p<0 . 001 ) ( Table 2 ) . The OR for goats sampled from large farms indicated C . burnetii seropositivity was significantly greater when compared to other farms ( OR 3 . 5; 95% CI 1 . 4 , 9 . 0 , p<0 . 001 ) ( Table 2 ) . The likelihood of seropositivity increased with age , with antibodies detected in 10 . 5% of goats ≥3 years of age when compared with only 2 . 3% of young adults ( ages 1–2 ) and 2 . 1% of kids ( aged <1 ) . The adult goats ( aged >3 years ) had significantly increased exposure to C . burnetii than kid goats ( OR 5 . 5; 95% CI 2 . 4 , 12 . 5 , p<0 . 001 ) ( Table 2 ) . However , there was no statistical difference between the young adult goats ( aged 1–2 ) and the kid goats , ( OR = 1 . 1; 95% CI 0 . 4 , 2 . 9 , p = 0 . 855 ) . Overall , age was a significant variable in the univariable analyses , p = 0 . 014 ( Likelihood ratio test , not in the table ) . All variables ( province , breed , production system , age category , gender and farm size ) were significantly associated with C . burnetii seropositivity on univariable analysis ( Table 2 ) . There was a significant difference between genders ( p = 0 . 025 ) with 4 . 7% of female goats being seropositive and only 1 . 5% of male goats demonstrating seropositivity ( Table 2 ) . Subsequently , all variables were included in the multivariable analysis . The following variables had multivariable significance: introduced Boer breed goats ( OR = 6 . 9; 95% CI 1 . 7 , 27 . 3 , p<0 . 001 ) ; goats 3 years or older ( OR 4 . 1; 95% CI 1 . 3 , 12 . 5 , p = 0 . 004 ) ; and goats located in Vientiane Capital had marginal significance ( OR 5 . 4; 95% CI 1 . 0 , 29 . 3 , p = 0 . 05 ) ( Table 2 ) . Overall , 20/1458 ( 1 . 4%; 95% CI 0 . 8 , 2 . 2 ) goats tested demonstrated Brucella spp . seropositivity having serial positivity to both ELISA and Rose-Bengal agglutination tests ( Table 4 ) , despite 3 . 0% of goat samples returned seropositive ELISA results alone . Significant differences were noted between provinces ( Pearson’s chi2 p<0 . 001 ) , with the highest seroprevalence noted in Vientiane Capital ( 4 . 0% ) , and in Attapeu ( 1 . 6% ) . Brucella spp . seropositivity was not detected from Xayaboury , Savannakhet or Xiengkhuang ( Table 4 ) . There was a significant difference of seropositivity to Brucella spp . between districts sampled in Vientiane Capital ( p = 0 . 009 ) , with the highest seroprevalence in Naxaythong ( 9 . 5% ) , followed by Xaythany ( 3 . 4% ) ( Table 3 ) . Parkgnum district had no Brucella spp . positive samples . Within Vientiane Capital , only three ( 25 . 0% ) villages were seropositive for Brucella spp: Ponsavan village ( 37 . 0% ) , Hongngua ( 22 . 6% ) , and Nagnang ( 3% ) ( Table 4 ) . This finding was significant compared with other villages within Vientiane Capital ( Fisher’s exact p<0 . 001 ) . Within Attapeu province , there was a significant difference in seropositivity detected between districts ( Fisher’s exact p = 0 . 014 ) , with 2/9 ( 18 . 1% ) animals seropositive in Phouvong and 1/120 ( 0 . 8% ) of animals seropositive in Sahmakisai . The other two districts within Attapeu had no Brucella seropositive animals . Where full epidemiological information was ( n = 754 ) available , there was a significant difference demonstrated between goat breeds ( p<0 . 001 ) , with 5 . 6% of introduced Boer or Boer cross breeds found positive , and no seropositivity detected in native breeds ( Table 4 ) . There was a significant difference demonstrated between production systems ( p<0 . 001 ) , with 5 . 6% of goats on commercial farms seropositive for Brucella spp . and no seropositivity detected in smallholder systems ( Table 3 ) . Animals were more likely to have positive antibodies detected if they were sampled on a large farm with over 40 animals ( 5 . 3% ) , with no seropositivity detected on farms with 40 or fewer goats ( p<0 . 001 ) . There was a significant difference in age of the goats ( p<0 . 001 ) , where adults ≥3 years recorded the highest proportion of seropositivity ( 3 . 3% ) followed by goats 1–2 years old ( 1 . 1% ) ( Table 4 ) . Only 0 . 3% of kids aged <1 year old was found to be seropositive . There was no association detected between genders ( p = 0 . 23 ) ( Table 3 ) .
This study investigated the presence of exposure of Lao goats to zoonotic pathogens , C . burnetii and Brucella spp , and determined their seropositivity . The seroprevalence of both C . burnetii and Brucella spp . was much higher and more widespread in goats compared with previous cattle seroprevalence studies within Laos [12 , 13] . Similar results were found in Thai goats for C . burnetii seroprevalence ( 4% ) [14] and Brucella spp . seropositivity ( 1 . 4% ) , although Brucella spp . seropositivity appears more widespread in Thailand ( 12 . 1% of herds ) [15] . The Coxiella study in Thailand utilised a similarly prepared ELISA from a different company ( IDEXX ) to this study however samples were also taken using convenience methodology so comparability of seroprevalence results between studies is limited [14] . The Brucella study utilised compliment fixation , Rose-Bengal and ELISA and if any test were positive the animal was considered positive [15] which may have artificially increased seropositivity when compared to the study presented here . The results presented here clearly demonstrate a spatial difference of C . burnetii seropositivity with Vientiane Capital having significantly higher individual goat seropositivity than other provinces . Similar to previous studies , a hotspot of C . burnetii seroprevalence was demonstrated in cattle in Xayaboury province , located on Laos-Thailand border [12 , 13] . Furthermore , goats in Vientiane were much more likely to be exposed to Brucella spp . than goats in other areas . This spatial distribution has not previously been reported for either pathogen in Laos , although is probably associated with Vientiane Capital being a major thoroughfare for international trade and the location of emerging commercial goat enterprises . In other global seroprevalence surveys for both pathogens , areas with high international trade are considered high risk for exposure to C . burnetii and Brucella spp . [22 , 23] . It is a known problem that illegal movement of animals occurs throughout South East Asia through “porous borders” as demonstrated through Foot and Mouth disease outbreak studies [24] . It is possible that goats brought into Laos from other counties were already exposed to the pathogens , or that travel and intensification of production has contributed to possible infection . Nevertheless , there appears a significant public health risk to goat farmers and possibly consumers within Vientiane Capital in Laos , as recently identified with Orf virus infection [4] . It is interesting that for both pathogens , introduced Boer mixed bred goats were significantly more likely to be seropositive than native Kambing-Katjang goats , with Brucella spp . seroprevalence only reported in the introduced goats . Reasons for C . burnetii being higher in Boer mixed bred goats may be that these animals tended to be more intensively raised and on commercial farms and were likely to have been or descendants of imported animals . Intriguingly , resistance in native goats to Brucella spp . has been previously suggested with seroprevalence studies in Mexico [25] and Malaysia [22] reporting increased Brucella spp . exposure in imported breed goats compared to native animals . It has been suggested that different breeds of cattle may also be resistant to Brucella spp . infection through genetic innate immunity [26 , 27] . Further studies are necessary to determine the possible role of genetics of goat immunity to a variety of pathogens . Age was independently associated with seropositivity of C . burnetii , and there was a significant difference between age groups as risk for Brucella spp . seropositivity . Adult animals ≥3 years were more likely to be exposed to both pathogens , a finding consistent with literature and likely due to increasing opportunities of pathogen exposure [5 , 28] . Similarly , female goats were more at risk of having antibody titer against C . burnetii than male goats , likely representing to the tropism for both pathogens to the placenta and mammary lymph nodes [6 , 8] . Results demonstrated here indicating risk factors for infection in Laos included commercialization in comparison with smallholder systems and association of larger farms with higher seroprevalence are in contrast with studies in Thailand and elsewhere [15 , 29 , 30] . It is thought that smallholder farmers with free ranging goats were at higher risk for Brucella spp . exposure as the mobility of wandering herds favours spread of infectious disease when allowed to mix with naïve herds [15 , 29 , 30] . Intensification has been reported as a risk factor in seroprevalence studies [25 , 31] , where close contact within herds may also favour pathogen spread . Although herd size and commercialisation were significant risk factors on univariable analysis for C . burnetii seroprevalence , neither variable was significant on multivariable analysis and hence were potential confounding variables , despite the likelihood that these herds were introduced to infection before or after importation . To accurately estimate prevalence of a disease in the population , sensitivity and specificity of a test must be known to approximate the occurrence of false results . The ID-Vet Q fever iELISA has been determined at 100% sensitivity and 100% specificity for Coxiellosis in cattle located in France , however no sensitivity and specificity reports have been performed for small ruminants [25 , 31] so these estimations may not be accurate for small ruminant studies . Furthermore , as the status of a disease ( endemic or not endemic ) in a population can alter the sensitivity and specificity of a test in a region , analysis of sensitivity and specificity is required in Laos , and South East Asia as a whole . Internal company testing of the ID-VET Brucella iELISA found 100% specificity on a herd of 160 goats in France , and 100% sensitivity on 5 goats in Southern Italy [21] . The small sample size brings into question the reliability of the estimations , which additionally might not be applicable to the region of South East Asia as sensitivity and specificity , can differ by regions [32 , 33] . The Rose-Bengal test has been assessed at 94% sensitive and 99% specific [34] yet there is no estimation of sensitivity or specificity for the combined tests specific to the region . Latent class analysis can be used to determine the sensitivity and specificity of tests in the absence of a gold standard test for the region . Furthermore , it can be estimated using latent class analysis which test is most suitable , in what order it should be performed , and also if the tests should be utilised in parallel ( all positives are included ) , or serially ( only positives on all tests included , as in this study ) . For example , studies of cattle serology in Zambia indicated a competitive-ELISA paired with Fluorescent Polarisation test results in the highest sensitivity and specificity [35] , and for sheep in Europe the blocking-ELISA is most accurate [36] , and finally Rose-Bengal plus competitive ELISA test gave the best results for cattle testing in Zimbabwe [37] . However , it is a weakness of this study that this analysis has not been performed to determine the utility of testing in parallel or serially . Brucella spp . and Q fever serology results must be interpreted in the context that Lao goats are not vaccinated for either disease , nor are there active control programs and as such it is highly unlikely that any of the positive serology results were attributable to local vaccine strains . Nevertheless , it is a limitation of this study that the farmers were not queried about vaccination . A positive Brucella spp . serology result can be caused by a cross-reaction with a range of bacterial species including Yersinia spp . giving rise to results that may not be fully accurate [38] although repeating tests can increase the reliability of results [34] . It is possible that the use of serial testing with Rose-Bengal and ELISA for Brucella antibodies reduced the number of sites with Brucella seropositivity than would have been reported by ELISA alone . Furthermore , following discussions with the ELISA kit manufacturers , a decision was made to classify "suggestive" positive samples as negative thereby decreasing the possibility of false-positive results . While the decision to employ serial ELISA and Rose-Bengal in this study was based on methodologies of previous Brucella spp . studies within Laos [12 , 13] , there is potential for latent class analysis to be utilised on this data set to determine the best combination of tests for the most appropriate sensitivity and specificity result in Laos . In a preliminary investigation during this study , vaginal swabs were collected from goats located within Vientiane Capital , with one found to be positive for Brucella spp . DNA with real time PCR and none to have C . burnetii DNA ( personal communication , Dr . Reka Kanitpun ) . This preliminary finding indicated that farmers and their families were at potential risk for contracting Brucellosis from their goats , although further investigations are needed to understand shedding patterns and speciation of Brucella spp . , preferably using molecular diagnostic tools . This study did not differentiate the species of Brucella spp . as both the serological tests utilised were genus specific only . Goats are generally associated with B . melitensis , the species that is most pathogenic to humans , yet infection with B . abortus or B . suis is also common [5] . Although this present seroprevalence data demonstrates previous exposure of animals to these pathogens , serology alone does not provide a complete picture of infection status within an animal population . Seropositivity does not indicate disease manifestation , current shedding of pathogens , or consequently current risk of transmission . Studies have suggested that a significant proportion of animals that shed C . burnetii or Brucella spp . are not seropositive; furthermore animals can be seropositive and not be shedding [5 , 6] . Furthermore , the proportion of animals shedding C . burnetii is independent of abortion history in a herd , and shedders might represent clinically unapparent infections [9 , 39] . Outbreaks of human Q fever and Brucellosis are commonly linked to seasonal parturition in small ruminant production system [5 , 9 , 40 , 41] . In developing nations , many fevers presenting to medical clinics go undiagnosed due to their general “ill-thrift” nature [42 , 43] . It is imperative that medical practitioners in Laos are aware of Q fever and Brucellosis as differential diagnoses , especially for at risk populations , including livestock farmers with recent exposure to animal parturitions , pregnant women and people consuming raw goat products . Control of zoonotic and transboundary disease pathogens proves difficult in developing nations where veterinary support and resource may be limited or unavailable . Vaccination of animals for either C . burnetii or Brucella spp . is not recommended as it must be given according to parturition calendars . Furthermore , vaccination will only reduce but not stop shedding , can cause goat abortions if given at incorrect times and can cause human disease if self-inoculated [5 , 40 , 42] . Test and culling is recommended for Brucella spp . seropositive farms and may have a role in this study where very few villages were considered likely to be affected . It is acknowledged this is currently difficult politically and financially . Disinfecting farms quarterly may reduce disease spread for both pathogens[22] although may not be applicable to rural village settings with free ranging goats . Despite these issues , this study has addressed important knowledge gaps on C . burnetii and Brucella spp . seroprevalence in Lao goats whilst raising a number of other questions . Further investigations of the potential risk factors for transmission of the different species and farming practices are necessary to determine why Boer crossbred goats have higher seroprevalence . Further studies investigating shedding of both C . burnetii and Brucella spp . are required for speciation and potential trace back of disease transmission , plus collection of caprine placentas for PCR and pathology . There is urgent need to determine current Q fever and Brucellosis seroprevalence and occurrence of the diseases in humans , especially in at -risk populations including livestock farmers , others exposed to goat effluent , plus people consuming raw goat meat or milk products [17] . With the increasingly important contribution of goats to Lao and regional food security , the zoonotic issues from Lao production systems will very likely become increasingly important . International aid groups and commercial farms are advised to serologically test goats prior to importing them into Laos , and work closely with Lao veterinary services to ensure limited pathogen spread occurs both between villages and from animals to humans . | Goat raising is a growing industry in Lao People’s Democratic Republic however there is very little information whether or not goat raising poses a disease threat to farmers and the general population through diseases that may be transmitted between animals and humans ( i . e . , zoonotic diseases ) . To determine this , we tested goats for antibodies against two zoonotic diseases: Q fever ( causative agent Coxiella burnetii ) and Brucellosis ( Brucella species ) in Lao goats across five provinces ( Vientiane Capital , Xayaboury , Xiengkhuang , Savannakhet and Attapeu ) . The presence of antibodies does not necessarily indicate active disease but that animals have been previously exposed to Q fever and Brucellosis . A total of 1458 goat serum samples were tested and the overall antibody positivity of the goats for C . burnetii was 4 . 1% and Brucella spp . was 1 . 4% . The highest risk of having Q fever antibodies was the goats being based in Vientiane Capital , of Boer mixed breed and ≥3 years old . The highest risk of having Brucella spp . antibodies was being based in Vientiane Capital , of Boer mixed breed as well as factors related to production system , age , and farm size . There is an urgent need to determine human health risks and economic losses caused by Q fever and Brucellosis . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"livestock",
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"ruminants",
"pathogens",
"tropical",
"diseases",
"microbiology",
"vertebrates",
"animals",
"mammals",
"brucellosis",
"bacterial",
"diseases"... | 2018 | Serosurveillance of Coxiellosis (Q-fever) and Brucellosis in goats in selected provinces of Lao People’s Democratic Republic |
The evolutionary dynamics of HIV during the chronic phase of infection is driven by the host immune response and by selective pressures exerted through drug treatment . To understand and model the evolution of HIV quantitatively , the parameters governing genetic diversification and the strength of selection need to be known . While mutation rates can be measured in single replication cycles , the relevant effective recombination rate depends on the probability of coinfection of a cell with more than one virus and can only be inferred from population data . However , most population genetic estimators for recombination rates assume absence of selection and are hence of limited applicability to HIV , since positive and purifying selection are important in HIV evolution . Yet , little is known about the distribution of selection differentials between individual viruses and the impact of single polymorphisms on viral fitness . Here , we estimate the rate of recombination and the distribution of selection coefficients from time series sequence data tracking the evolution of HIV within single patients . By examining temporal changes in the genetic composition of the population , we estimate the effective recombination to be ρ = 1 . 4±0 . 6×10−5 recombinations per site and generation . Furthermore , we provide evidence that the selection coefficients of at least 15% of the observed non-synonymous polymorphisms exceed 0 . 8% per generation . These results provide a basis for a more detailed understanding of the evolution of HIV . A particularly interesting case is evolution in response to drug treatment , where recombination can facilitate the rapid acquisition of multiple resistance mutations . With the methods developed here , more precise and more detailed studies will be possible as soon as data with higher time resolution and greater sample sizes are available .
The human immunodeficiency virus ( HIV-1 ) ranks among the most rapidly evolving entities known [1] , enabling the virus to continually escape the immune system . After infection with HIV , patients typically enter an asymptomatic period lasting several years during which the virus is present at low to medium levels , typically at a viral load of to copies per ml plasma . Nevertheless , the number of virions produced and removed is estimated to be around per day with a generation time slightly less than two days [2] . Due to this rapid turnover and the high mutation rate of per site and generation , the sequence diversity of HIV within a single patient can rise to % ( env gene ) within a few years and the divergence from the founder strain increases by % per year [3] , although this rate is not constant [4] . The genotypic diversity is subject to positive selection for novel variants that are not recognized by the host immune system or that reduce the sensitivity to anti-retroviral drugs [5]–[7] , as well as to purifying selection by functional constraints [8] . In addition to high substitution rates and strong selection , genomes of different HIV particles within the same host frequently exchange genetic information . This form of viral recombination works as follows: Whenever a cell is coinfected by two or more viruses , the daughter virions can contain two RNA strands from different viruses[9] , [10] . In the next round of infection , recombinant genomes are generated by template switching of the reverse transcriptase while producing cDNA . It has been shown that recombination in HIV contributes significantly to the genetic diversity within a patient [11]–[13] . In cases of super-infection with several HIV-1 subtypes , recombination can give rise to novel forms that become part of the global epidemic [14] . The observation of recombinant viruses after a change in anti-retroviral drug therapy [15] suggests that recombination might play an important role in the evolution of drug resistance , as predicted by theoretical models [16] . The amount by which recombination speeds up the evolution of drug resistance depends on the parameters governing the population dynamics [17] , many of which are not known to sufficient accuracy . In vitro estimates of the recombination rate have shown that the reverse transcriptase switches templates about times while transcribing the entire genome , resulting in a recombination rate of per site and generation [18] , [19] . However , the bare template switching rate is only of secondary importance , since recombination can generate diversity only if the virion contains two RNA strands that originate from different viruses , which requires coinfection of host cells[20] . The effective in vivo recombination rate is therefore a compound quantity , to which the template switching rate and the probability of coinfection of a single host cell contribute . This effective recombination rate has been estimated with coalescent based methods developed in population genetics [13] , [21] . These methods use a single sample of sequences obtained from the diverse population and estimate the recombination rate from topological incongruencies in the phylogenetic tree of the sequence sample . Together with an estimate of the mutation rate , this allows to estimate the recombination rate in real time units . Shriner et al . [13] report an estimate of per site and generation , implying almost ubiquitous coinfection of host cells . Here , we present a different method to estimate recombination rates from longitudinal sequence data , which has been obtained from 11 patients at approximately 6 month intervals [3] , [22] . By comparing sequence samples from successive time points , we can estimate recombination rates from the distance and time dependence of the probability of cross-over between pairs of polymorphic sites . We find that the effective rate of recombination is per site and generation . Furthermore , we estimate the strength of selection on nonsynonymous polymorphisms by measuring the rate at which allele frequencies change . We find that a fraction of about 15% of the observed nonsynonymous polymorphisms are selected stronger than % per generation .
Recombination produces new combinations of alleles from existing genetic variation and randomizes the distribution of genotypes . To illustrate this process and the challenges of estimating recombination rates , consider the pair of polymorphic sites in Figure 1 . Generically such a pair will have arisen by the following sequence of events: ( i ) Site 1 becomes polymorphic by mutation , e . g . AC . ( ii ) A mutation occurs at site 2 on a genome that carries one of the variants of site 1 , e . g . giving rise to the haplotypes A…T , A…G and C…T . ( iii ) The missing haplotype , in this example C…G , can be generated by further mutation ( AC at site 1 or TG at site 2 ) or by crossing over two of the existing haplotypes , as illustrated in Figure 1 . In population genetics , the occurrence of the fourth haplotype is often taken as sufficient condition for recombination ( the four gamete test[24] ) . While this is true for bacteria and eukaryotes because of their low mutation rates , the HIV population within a patient is large and mutates rapidly [25] . Hence , the biggest challenge in estimating recombination rates is to separate recombination from recurrent mutations or homoplasy . A second confounding effect stems from the small number of sequences available per time point , such that the sequences containing the fourth haplotype could have been missed due to undersampling . To disentangle these two effects from recombination , we make use of the fact that only the recombination rate depends strongly on the distance between the two sites , while recurrent mutations and sampling noise should not . For each pair of biallic sites at time that was found in three of the four possible haplotypes , we asked whether the missing haplotype is observed the at time ( comp . Figure 1 ) and calculated the frequency of this event as a function of the separation of the two sites , as shown in Figure 2A . This frequency increases with the separation of the two sites from about 0 . 1 to about 0 . 35 at 500 bp separation , in line with the expectation that recombination is more rapid between distant sites . To corroborate that this distance dependence is indeed due to recombination , we performed the following similar analysis: The curve labelled “other haplotypes” in Figure 2A shows the frequency of observing a haplotype at time , which contains alleles not observed at time , again averaged over all available data . Any such haplotype could have arisen by mutation in the time interval between and , or could have been present at time but not sampled . It cannot , however , be assembled by recombination from the alleles found at time . The important observation is that the frequency of observing such a haplotype does not increase with distance . This is consistent with our expectation that an additional mutation or undersampling should be independent of an additional polymorphism nearby . The clear separation between the two classes of haplotypes suggests that the contribution from homoplasy and sampling can be accounted for by a distance independent constant . The probability of recombination between two sites increases with the product of the time elapsed and the distance between the sites , rather than with distance alone as plotted in Figure 2A . Panel B of Figure 2 shows the probability of appearance of a putative recombinant haplotype as a function of the product of distance and number of generations ( generation time 2 days ) . To estimate the recombination rate , we have to know how the observed saturation behavior is related to recombination . Let and be the alleles at site 1 and 2 and the frequency of the missing haplotype . The probability not to detect haplotype in a sample of size is ( 1 ) Assuming the allele frequencies remain constant , will relax from its initial value to the product of the allele frequencies as through recombination[26] . The frequency of detecting a haplotype at time , given it was not detected in the previous sample at time , is therefore . This quantity , averaged over all pairs of polymorphic sites at distance such that falls into a specified bin , is shown in Figure 2B ( the average extends over all patients and time points ) . To understand how depends on the recombination rate , it is useful to consider the two limiting cases of small and large ( 2 ) where we assumed that is small compared to one ( inspection of Figure 2B shows this is true , we discuss this in more detail in Methods ) . Hence , the recombination rate is proportional to the slope of at . The intercept is simply the probability that we detect at time in absence of recombination , given we missed it at time . We determine the slope and the intercept by fitting the function to the data . The recombination rate is then given by ( 3 ) where is measured directly . The best fit yields recombinations per site and generation ( one standard deviation ) . The uncertainty of was estimated by resampling the patients with replacement 500 times . This bootstrap distribution of the recombination rate estimate is shown in Figure 3 . We have assumed that the allele frequencies and remain constant in the interval . We will see below , however , that some allele frequencies change rapidly . We expect that repeated sweeps will cause our method to overestimate the recombination rate: When the frequencies of the minor alleles increase , the missing haplotype is produced more rapidly then expected . Positive selection on the variable regions in env and purifying selection on the conserved regions have been repeatedly reported in the literature [5] , [23] , [27]–[29] . Most of the these studies compared the rates of synonymous and non-synonymous substitutions ( ratio ) . An observation of indicates selection for novel variants , while indicates that the amino acid sequence changes more slowly than the nucleotide sequence , indicating functional constraint at the protein level . The overall rate of substitutions , however , yields only very limited information about the strength of selection . Selection for a specific variant implies that this variant confers an elevated rate of reproduction compared to the population mean . Ignoring random drift , the strength of selection is related to the rate at which the frequency of the variant changes [26]: ( 4 ) This equation has a straightforward solution and from two measurements of at time and , can be determined to be ( ) [11] ( for a review on selection and drift , see [30] ) . However , when using this formula , emphasis is put on rare alleles , whose frequencies can't be measured accurately in small samples . It proves more robust to estimate the rate of allele frequency change directly as , where is the difference in allele frequency between consecutive samples and is the time interval . The discrete derivative can serve as a proxy for selection which is less sensitive to rare alleles . The observed will be a sum of contributions from selection , noise from random sampling , and genetic drift . The latter can be estimated by measuring for synonymous polymorphisms which are putatively neutral and their observed frequency changes are assumed to be dominated by sampling noise . However , they can be affected by selection on nearby non-synonymous polymorphisms ( hitch-hiking ) or be themselves under selection , e . g . for translation efficiency or RNA secondary structure . Despite the limited resolution of and due to small sample sizes and large time intervals between samples ( 6–10 month ) , we can make a meaningful statement about the strength of selection when averaging over all sites , patients and time points . Figure 4 shows the cumulative distributions of the rates of change of allele frequencies observed during the interval between two consecutive time points for non-synonymous and synonymous polymorphisms . The histograms are shown as insets in the Figure . There are consistently more fast changing non-synonymous polymorphisms than there are synonymous ones , suggesting that a fraction of the non-synonymous polymorphisms is indeed responding to selection . To check whether the fast changing synonymous polymorphisms can be attributed to hitchhiking , we excluded synonymous polymorphisms that are closer than 100bp to a non-synonymous polymorphism that changes faster than per generation . The resulting distribution is much narrower with no allele frequency changes beyond per generation , indicating that the fast changing synonymous polymorphisms are indeed “hitch-hiking” . The cumulative histograms can be compared by the Kolmogorov-Smirnov test , which uses the maximal vertical distance between curves as a test statistics . The test reveals that the non-synonymous distribution is significantly different from both the unconditional synonymous distribution ( p-value ) and the synonymous distribution without hitch-hiking ( p-value ) . Note that not all observations are independent since nearby sites are linked and move coherently . Hence , realistic p-values will be larger . The fastest allele frequency changes detected are about per generation , which is our resolution limit ( 6 month 100 generations ) . This low time resolution results in a tendency to underestimate the rates of change , while the finite sample size will generate spurious frequency changes due to sampling noise . However , the narrow distribution of allele frequency changes of synonymous polymorphisms excluding hitchhiking suggests that the contribution from sampling noise is small ( on the order of per generation ) . Hence , the tail of the histogram of the frequency changes of nonsynonymous polymorphisms contains a rough measure of the distribution of selection coefficients[31]: about 15% of the observed non-synonymous polymorphisms change faster than per generation , compared to almost none of the synonymous polymorphisms . Assuming that this difference is due to selection , we conclude that about 15% of the observed non-synonymous polymorphisms are positively selected with per generation . In the last step , we rearranged Eq . ( 4 ) to obtain with . Most of the strongly deleterious polymorphism are of course never observed in the samples .
The dynamics of HIV within a single patient is characterized by large diversity due to high mutation rates , intense selection , frequent recombination and stochasticity resulting from bursts of viruses originating from a single cell . The simultaneous importance of these four evolutionary forces makes HIV evolution difficult to analyze with traditional population genetic methods , which typically assume that one of the evolutionary forces is predominant . Coalescent based methods , for example , assume that evolution is neutral , i . e . stochasticity dominates over selection . While it is possible to include recombination or selection into a coalescent description [32] , [33] , the analysis becomes difficult and computationally demanding . Estimators based on site frequency spectra , like Tajima's , work best when recombination is strong compared to selection . Phylogenetic analysis , on the other hand , assumes absence of recombination . These methods have been designed to infer parameters of the population dynamics from snapshots of the population , which is a formidable challenge . The lack of time series data requires the assumption of a model of the population dynamics , which can be extrapolated back in time to the most recent common ancestor . In neutral models , the time to the most recent common ancestor is on the order of generations , being the population size . During this long time interval , there is ample opportunity for selection or demography to invalidate the assumption of the model . If time series data is available , the task of estimating parameters is greatly simplified since one can trace the dynamics of alleles and genotypes directly . Such longitudinally sampled data has for example been used to gauge the molecular clock of bacterial evolution [34] . We have used time resolved data from 11 patients to estimate parameters of the population dynamics directly . In accordance with existing studies , we find that recombination in HIV is frequent and contributes significantly to sequence diversity [11]–[13] , [21] . The template switching rate of HIV is known to be about per site[18] , [19] . For template switching to result in a novel genetic variant , the two RNA strands in the infecting virus have to be different , which implies co-infection of the cell the virus originated from . Any estimate of recombination rates from sequence diversity will therefore measure an effective recombination rate being approximately the product of the coinfection probability , the probability of forming a heterozygote and the template switching rate . Our estimate of this effective recombination rate per site and generation is about a factor of 20 lower than the template switching rate , but still implies a probability of coinfection of about 10% . Our estimate is lower than the estimate by Shriner et al . [13] , who estimated per site and generation . However , two other estimators also reported in that paper yielded lower recombination rates comparable to our estimate . In our analysis we have assumed that the rate of recombination is constant across the env gene . However , the breakpoint distribution in circulating recombinant forms show strong variation along the genome , with particulary little recombination in env [35] . The variation of the breakpoint distribution can largely be explained by low sequence homology between different subtypes and dysfunctional recombinants [36] . In the present study , however , all patients were infected with a single subtype and gradually built up diversity which remained much lower than the distance between subtypes . The effects causing recombination rate variation should therefore be of minor importance . By comparing the distribution of the rate of allele frequency changes of synonymous ( putatively neutral ) and nonsynonymous ( possibly selected ) polymorphisms , we estimated the distribution of selection coefficients on single sites . We find that 15% of the observed nonsynonymous polymorphisms are selected with coefficients greater than % per generation . In using the distribution of allele frequency changes to infer selection coefficients , we have assumed that each locus is selected for its own effect on fitness , rather than changing its frequency due to selection on a linked neighboring locus [37] or some epistatic combination of loci [38] , [39] . A sweeping polymorphism “drags” along neutral variation in a region [40] , which using our estimates of , and an effective population size of evaluates to bp . This is consistent with our finding that most of the rapid allele frequency changes of synonymous polymorphisms occur in the vicinity ( ) of rapidly changing non-synonymous polymorphisms ( Figure 4 ) . The sequence diversity in our samples is on the order of 3% [3] , such that two polymorphisms are expected to be on average 30bp apart . Roughly half of the observed polymorphisms are non-synonymous , of which 15% are under strong positive selection . Hence the distance between simultaneously sweeping loci is on the order of 400bp , which is of the same order as . If the rate of sweeps was much higher , sweeps would cease to be independent and interfere . While these estimates have large uncertainty , it is conceivable that the rate of sweeps is limited by recombination [41] . Selection coefficients in HIV have been estimated before by Liu et al . [11] in a patient infected with two HIV-1 subtype B viruses . In this patient , a small number of recombinant forms competed against the ancestral strains and selection differentials were estimated to lie between % and % . These selection coefficients are higher than our estimates , which is plausible since they are associated with entire recombinant genotypes that differ at many sites rather than the single site estimate presented here [39] . The strength of selection for novel epitopes in several HIV genes ( rate of escape from cytotoxic T-lymphocyte mediated killing ) during the asymptomatic phase of HIV infection has been shown to be of similar magnitude as our estimates [42] . The present study is limited by low time resolution and small sequence samples and more accurate and detailed answers could be obtained from larger samples . Larger sample sizes will require generalizations of the method used to estimate recombination rates which depends on pairs of sites where one of the four possible haplotypes is absent . In large samples , pairs of high frequency polymorphisms will most likely be present in all four possible haplotypes . In this case , one can measure linkage disequilibrium directly and observe how it decays from one timepoint to the next , e . g . by measuring its autocorrelation function . The method presented here is an approximation of this more general method suiteable for noisy data .
Sequences of the C2-V5 region of env from 11 patients , which were part of the MACS study [3] , [22] , were obtained from the Los Alamos National Lab HIV data base ( Special interest alignments , accession numbers: AF137629-AF138703 , AF204402-AF204670 , AY449806 - AY450055 & AY450056 - AY450257 ) . The sequences were translated into amino acid sequences and aligned for each patient separately using MUSCLE v3 . 6 with default settings [43] . Aligned nucleotide sequences were then constructed by inserting a 3bp gap for each gap in the amino acid alignment . Scripting and plotting was done in Python using the NumPy and Matplotlib environment [44] , [45] . To estimate recombination rates , we calculated the frequency of generating the missing haplotype as a function of the distance between polymorphic sites . Specifically , the algorithm proceeds as follows: For each pair of biallelic and gapless sites , we constructed a list of haplotypes , i . e . a list of alleles that co-occur in the same sequence . The list typically contains 3 or 4 haplotypes , but can also contain only 2 of the 4 possible pairs due to undersampling or selection against some of the allele combinations . Only those pairs with 3 haplotypes were included in the estimation . Furthermore , we used only those sites where both alleles were observed at least three times , since rare alleles are very sensitive to sampling noise . The analysis was repeated with this cutoff at two or four alleles , yielding similar results . For each pair of biallelic sites for which 3 haplotypes were detected in the sample at time , we determined the 4th haplotype that can be formed from the alleles observed at time , i . e . C…G in the example given in Figure 1 . We then checked whether this missing haplotype is detected at time . By averaging over all time points ( but the last one ) , all patients , and all pairs within a given distance interval , we determined the frequency of finding the missing haplotype as a function of the distance between the sites and as a function of the product of distance and time difference . The error bars indicated in Figure 2 are calculated as the product of the estimated value and . This would be approximately one standard deviation if all counts were independent , which they are not since the observations are pairs of sites and each site contributes to multiple pairs . However , these error bars indicate the relative uncertainties of the different points , which is all that is needed for an unbiased fit . According to Eq . 2 , we can estimate the recombination rate from the axis intercept and the slope at , which are extracted from the data by fitting ( 5 ) to the data . The fit is done by minimizing the squared deviation , weighted by the relative uncertainty of the data points indicated by the error bars in Figure 2 . was averaged over all pairs of sites contributing . To estimate a confidence interval for the estimate of , the estimation was repeated 500 times with a set of 11 patients sampled with replacement from the original 11 patients . The method relies on pairs of biallelic sites in a sample of size , where 3 out of the 4 possible haplotypes are observed . In large samples from a recombining population , one would naively expect to observe all possible haplotypes in most cases . However , due to the very skewed allele frequency distributions , the haplotype formed by the rare alleles is often sufficiently rare that it is missed even in large samples . We denote the alleles at site 1 by and at site 2 by , with the and being the minor alleles . First , observe that the mean frequency of haplotype , , at linkage equilibrium is . Hence for , in accord with observed in the data . We further assumed that the frequency of the unobserved haplotype , , is significantly smaller than . There are several reasons why is typically significantly smaller than and hence smaller than : ( i ) the condition that haplotype is not observed pushes down , ( ii ) minor alleles tend to be in negative linkage equilibrium if they are involved in selective sweeps , ( iii ) allele frequency spectra are even more skewed than due to purifying selection . The degree to which these 3 reasons reduce can be best observed in Figure 2B at small . The Figure shows the probability to observe haplotype at time , given it was not observed at time . At small , linkage disequilibrium is not yet broken down by recombination and the probability to observe haplotype at time gives us a measure of at time , which is indeed much smaller than 1 . Hence we can expand for small in Eq . 2 . The distribution of the rate of change of allele frequencies was estimated from pairs of successive time slices with the following characteristics: ( i ) site biallelic without indels at time with no constraint on timeslice , and ( ii ) monomorphic sites at time that are biallelic at time ( without indels ) . In both cases there are two alleles and whose frequencies can be meaningfully compared . The difference in allele frequencies were calculated as , where is the count of allele at time and is the sample size . Unless a third allele arose between and ( case ( i ) ) , and the common estimate of the rate of change , , was added to the cumulative distribution . In rare cases where a third allele did arise , both and have been added . Sometimes , a nucleotide changed from one state to another between time or with no polymorphisms detected . In such a case , was used . To detect the action of selection , we produced cumulative histograms of the rate of change in allele frequency for nonsynonymous and synonymous ( putatively neutral ) polymorphisms by averaging over all patients and all pairs of consecutive time points and with both and greater than . We then tested for excess of fast changes among the nonsynonymous polymorphisms using the Kolmogorov-Smirnov test . The test statistic is the maximal vertical difference between the cumulative distributions divided by , where and are the number of observations of synonymous and non-synonymous polymorphims . The Kolmogorov-Smirnov test was performed using the statistics module of SciPy [46] . To assess the role of hitch-hiking of synonymous polymorphisms with nearby selected non-synonymous polymorphisms , we produced a list of positions of non-synonymous polymorphisms whose allele freqency changes faster than per generation between time and . Synonymous polymorphism closer than 100bp to any of these positions where excluded from the conditioned histogram . | Evolution , in viruses and other organisms , is the result of random genetic diversification by mutation or recombination and selection for survival . In most organisms , evolution is too slow to be observed directly and the evolutionary past has to be reconstructed from static snapshots of the population . This reconstruction requires simple models of evolution that typically neglect selection or recombination . In vigorously evolving organisms like HIV , such assumptions are questionable . However , HIV evolves rapidly enough that substantial evolution is observable during a chronic HIV infection within single patients . Using such time series data of evolution , we estimate the effective recombination rate of HIV ( the rate of viral sex ) to be similar to the mutation rate , rather than much larger as previously reported . We also study the strength of selection exerted on the virus by the immune system . We find that about 15% of the observed virus variants with mutations in the surface protein are favored and selected at a rate of 0 . 8% to 2% per virus replication cycle . Knowledge of the recombination rate and the strength of selection is essential for quantitative modeling and understanding of HIV evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"infectious",
"diseases/hiv",
"infection",
"and",
"aids",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"genetics",
"and",
"genomics/population",
"genetics"
] | 2010 | Recombination Rate and Selection Strength in HIV Intra-patient Evolution |
The small G-protein Ras is a conserved regulator of cell and tissue growth . These effects of Ras are mediated largely through activation of a canonical RAF-MEK-ERK kinase cascade . An important challenge is to identify how this Ras/ERK pathway alters cellular metabolism to drive growth . Here we report on stimulation of RNA polymerase III ( Pol III ) -mediated tRNA synthesis as a growth effector of Ras/ERK signalling in Drosophila . We find that activation of Ras/ERK signalling promotes tRNA synthesis both in vivo and in cultured Drosophila S2 cells . We also show that Pol III function is required for Ras/ERK signalling to drive proliferation in both epithelial and stem cells in Drosophila tissues . We find that the transcription factor Myc is required but not sufficient for Ras-mediated stimulation of tRNA synthesis . Instead we show that Ras signalling promotes Pol III function and tRNA synthesis by phosphorylating , and inhibiting the nuclear localization and function of the Pol III repressor Maf1 . We propose that inhibition of Maf1 and stimulation of tRNA synthesis is one way by which Ras signalling enhances protein synthesis to promote cell and tissue growth .
The Ras small G-protein is one of the key conserved regulators of cell growth and proliferation . Over three decades of research have defined the textbook model of how Ras is activated by growth factors to stimulate a core RAF kinase , MEK ( Mitogen-activated protein kinase kinase ) and ERK ( Extracellular signal–regulated kinase ) signalling cascade . Work in model organisms such as Drosophila , C elegans and mouse has shown how this Ras/ERK pathway coordinates tissue growth and patterning to control organ size during development and homeostatic growth in adults . Given its central role in development it is not surprising that defects in Ras signalling contribute to disease . Most notably , activating mutations in Ras and RAF occur in a large percentage of cancers , and lead to hyper-activation of ERK , which drives tumour formation in both epithelial and stem cells [1] . Ras pathway mutations are also seen in several genetic developmental disorders–described collectively as RASopathies–often characterized by abnormal growth[2] . Understanding how Ras promotes cell proliferation and tissue growth is therefore an important concern in biology . Drosophila has been a powerful model system to understand the biological roles of Ras signalling . In flies , Ras functions downstream of epidermal growth factor ( EGF ) and activation of its tyrosine kinase receptor ( the EGFR ) . A series of genetic studies initiated over 25 years ago were pivotal in defining the canonical EGFR/Ras/ERK pathway in Drosophila ( for reviews of this early work see:[3 , 4] ) . Extensive studies since then have established when , where and how the pathway is activated during the fly life cycle to control development . This work has emphasized the importance of Ras signalling in the control of cell growth and proliferation ( e . g . [5–9] . Notably , during larval development Ras/ERK promotes EGFR-mediated cell proliferation and tissue growth in epithelial organs such as the imaginal discs , which eventually give rise to adult structures such as the legs , wings and eyes [10–14] . In addition , in the adult the EGFR/Ras/ERK signalling controls proliferation of stem cell populations to maintain homeostasis and promote regenerative growth [15–19] . How does Ras mediate these effects on cell and tissue growth ? Most work on this area has focused on transcriptional effects of Ras signalling . Work in Drosophila has identified several transcription factors that are targeted by ERK such as fos , capicua , and pointed , and that regulate growth [19–22] . Ras signalling has also been shown to crosstalk with other transcriptional regulators of growth such as the hippo/yorkie pathway and dMyc [10 , 23–27] . These transcriptional effects control expression of metabolic and cell cycle genes important for growth[20 , 21] . Less is known , however , about how Ras/ERK may regulate mRNA translation to drive growth . The prevailing view , arising mostly from mammalian tissue culture experiments , is that ERK controls protein synthesis by stimulating the activity of translation initiation factors[28] . In particular , these effects are mediated via two ERK effector families—the MNK ( MAP kinase-interacting serine/threonine-protein kinase ) and RSK ( ribosomal s6 kinase ) kinases [28–30] . These kinases are important for cellular transformation and tumour growth in mammalian cells [31–34] . However , MNK and RSK mutants in mice and Drosophila have little growth or developmental phenotypes , and mouse MNK mutant cells show no alterations in protein synthesis [33–37] . These findings suggest Ras uses additional mechanisms to control translation and growth in vivo during animal development . In this paper , we report that the Ras/ERK pathway can stimulate RNA polymerase III-dependent tRNA synthesis . We find that these effects are required for Ras to drive proliferation in both epithelial and stem cells . Finally , we show that ERK promotes tRNA synthesis by inhibiting the Pol III repressor Maf1 . These findings suggest that stimulation of tRNA synthesis may be one way that Ras promotes mRNA translation to drive cell and tissue growth .
We first examined whether Ras signalling regulates protein synthesis in Drosophila S2 cells using a puromycin-labelling assay [38] . When a constitutively active Ras mutant ( RasV12 ) was expressed in Drosophila S2 cells using an inducible expression vector , we found an increase in protein synthesis , which was blocked by treatment of cells with cycloheximide ( CHX ) , an inhibitor of mRNA translation ( Fig 1A ) . Also , using polysome profiling to measure mRNA translation , we saw an increase in polysome levels in RasV12 overexpressing cells when compared with control cells ( Fig 1B ) . Conversely when we blocked Ras/ERK signalling by treating cells with the MEK inhibitor , U0126 , protein synthesis was decreased ( Fig 1C ) . Finally , we found that total protein content/cell increased after RasV12 was overexpressed in S2 cells ( Fig 1D ) . Our findings suggest that one way that the Ras/ERK signalling pathway may drive growth in Drosophila is by promoting protein synthesis . We previously identified the regulation of RNA Polymerase III and tRNA synthesis as a mechanism for controlling protein synthesis in Drosophila larvae [39 , 40] . We showed that these processes were regulated by TORC1 kinase signalling , and that they were important for driving tissue and body growth[39 , 40] . We were therefore interested in examining Ras signalling could also promote tRNA synthesis . We first used qRT-PCR to examine both pre-tRNA and total tRNA levels in S2 cells . We began by using the pharmacological MEK inhibitor U0126 to examine the effects of blocking Ras signalling . We found that treatment of S2 cells with UO126 lead to a decrease in levels of both pre-tRNAs and total tRNAs ( Fig 2A and 2B ) . Also , using Northern blots , we saw that treatment with UO126 lead to reduced pre-tRNA and tRNA levels in S2 cells ( S1A Fig ) . We also examined the effects of Ras pathway inhibition by using RNAi to knockdown Raf . We found that treatment of cells with dsRNA to Raf lead to reduced levels of both pre-tRNA and total tRNAs ( Fig 2C and 2D ) . In contrast to Ras pathway inhibition , we found that RasV12 ( constitutively active Ras ) overexpression lead to an increase in both pre-tRNA and mature tRNA levels as measured by both qRT-PCR ( Fig 2E and 2F ) and Northern blot ( S1B Fig ) , indicating enhanced tRNA synthesis . In contrast , to these effects on tRNA levels , we found no effect of inhibiting Ras signalling on expression levels of TBP ( which was previously reported[41] ) or on levels of Brf1 or Trf1 –both of which are components of TFIIIB complex , which is required for Pol III recruitment to tRNA genes ( S1C Fig ) . We also found that altering Ras signalling ( either by MEK inhibition or overexpression of RasV12 ) had no effect on levels of 5S or 7SL RNA in S2 cells ( S1C and S1D Fig ) . Both of these genes transcribed by RNA polymerase III , but they are different Pol III gene types ( type I and III respectively ) that use a different set of core promoter factors compared to tRNA genes ( which are type II Pol III genes ) . These finding suggest that the changes in tRNA synthesis we observed upon altering Ras signalling are not due to alterations in the levels of the basal transcriptional machinery required for tRNA transcription . These data also suggest Ras signalling may predominantly affect type II RNA pol III genes . We also examined the effects of Ras signalling on tRNA levels in the developing wing imaginal discs . We used the temperature-sensitive escargot-Gal4 ( esg-Gal4ts ) system , which allows for inducible transgenes expression in all imaginal tissues . When we overexpressed UAS-Rafgof using this system , we found a marked increase in pre-tRNA levels in wing discs as measured by qRT-PCR on dissected wing discs ( Fig 2G ) . Brf1 is a conserved component of TFIIIB complex , which is required for Pol III recruitment to tRNA genes [42] . We previously showed that Brf1 is involved in controlling Pol III-dependent transcription , and tissue and body growth in Drosophila larvae [39] . Here we examined whether Brf1 is required for Ras-induced tRNA synthesis . We found that knocking down Brf1 using dsRNA in S2 cells ( S2A and S2B Fig ) suppressed the RasV12 induced increase in tRNA levels ( Fig 3A ) without altering the strong induction of ERK phosphorylation seen with RasV12 expression ( S2B Fig ) . Hence , our data here suggest that the elevation of tRNA levels upon Ras activation is due to increased Pol III transcription . We then examined whether Brf1 is required for Ras-induced growth in Drosophila wing discs . We expressed UAS-driven transgenes in the dorsal compartment of the wing imaginal disc ( using an apterous-Gal4 driver , ap-GAL4 ) and then , in each case , measured tissue size in wandering stage third instar larvae , Overexpression of UAS-EGFR ( UAS-λtop ) in the dorsal compartment of the wing imaginal disc stimulates Ras/ERK signalling and leads to tissue growth ( Fig 3B and 3C ) . We found that RNAi-mediated knockdown of Brf1 by expression of a UAS-Brf1 inverted repeat line ( UAS-Brf1 RNAi ) in the dorsal compartment had little effect on tissue growth . However , expression of UAS-Brf1 RNAi blocked the overgrowth seen with UAS-EGFR expression . Expression of UAS-Brf1 RNAi with ap-GAL4 had little effect on tissue growth , suggesting we are not knocking down Brf1 to a level that cannot support any growth . We previously showed that Brf1 knockdown had no effect on ribosome synthesis , suggesting that its predominant effect was to block Pol III function [39] . Hence , these data indicate that Brf1 and Pol III transcription is required for EGFR/Ras/ERK-mediated increases in epithelial tissue growth in Drosophila . A major role for the EGFR/Ras/ERK pathway is in the growth and maintenance of the Drosophila intestine . In larvae , activation of the pathway plays a central role in controlling the proliferation of adult midgut progenitor cells ( AMPs ) , which eventually give rise to the adult intestine [9] . In the adult the EGFR/Ras/ERK pathway is required to promote stem cell proliferation and tissue regeneration [15–17 , 19 , 43] . We therefore examined whether Brf1-mediated Pol III transcription was required for these proliferative effects of Ras/ERK signalling . We first examined the larval intestine . During the larval period , AMPs proliferate and give rise to clusters of ~5–10 cells scattered throughout the larval intestine . These cell clusters eventually proliferate and fuse during metamorphosis to give rise to the adult intestinal epithelium . The EGFR/Ras/ERK pathway controls the proliferation of AMPS [9] . Overexpression of either UAS-EGFR or UAS-Rafgof in the AMPs using the temperature-sensitive escargot-Gal4 ( esg-Gal4ts ) system lead to a massive increase AMP proliferation and an increase in the numbers of AMP cells per cluster as previously reported . We found that expression of UAS-Brf1 RNAi ( Fig 3D and 3E and S2C Fig ) lead to a small reduction in the number of cells per cluster . However , we found that when co-expressed along with UAS-EGFR or UAS-Rafgof , UAS-Brf1 RNAi blocked the increase in AMP cell numbers . These data indicate Brf1 is required for EGFR/Ras/ERK mediated cell proliferation . We next examined Brf1 function in homeostatic growth in the adult intestine . Damage to intestinal epithelial cells leads to an increase in expression and release of EGF ligands from both intestinal cells and underlying visceral muscle [15] . These EGF ligands then act on the intestinal stem cells ( ISCs ) to stimulate the Ras/ERK pathway , which triggers stem cell growth and division , and promotes regeneration of the intestinal epithelium . This damage-induced increase in ISC proliferation is dependent on EGFR/Ras/ERK signalling and can be mimicked by genetically activating the pathway specifically in the stem cells [15–17 , 19] . We tested a requirement for Brf1 in this Ras-mediated homeostatic growth response . We first examined the effects of intestinal damage . As previously reported [19] , we found that feeding flies either DSS or bleomycin–two different gut stressors–leads to an increase in ISC proliferation . However , we found that this effect was inhibited when we knocked down Brf1 ( using UAS-Brf1 RNAi expression ) specifically in the ISCs and their transient daughter cells , the enteroblasts ( EBs ) , using the inducible esg-Gal4ts system ( Fig 4A and 4B ) . We next examined the effects of activation of the Ras/ERK pathway . We first overexpressed UAS-RasV12 in the adult intestine using the inducible esg-Gal4ts system , and observed an increase in pre-tRNA levels ( Fig 4C ) . As previously reported , when we overactivated the pathway in stem cells by expressing UAS-Rafgof using esg-Gal4ts , we saw an increase cell proliferation as indicated by a marked increase in GFP labelled ISCs and EBs ( Fig 4D ) . Expression of a UAS-Brf1 RNAi had little effect on GFP labelled cells , but when co-expressed with UAS-Rafgof it blocked the increase in cell proliferation . These results suggest that Brf1 and Pol III-dependent transcription is required for stem cell proliferation in the adult intestine . We next wanted to examine how Ras signalling stimulates Pol III-dependent tRNA transcription . One candidate regulator we tested was dMyc . In both mammalian cells and Drosophila , Myc can interact with Brf1 and stimulate Pol III-dependent transcription [39 , 44 , 45] . Moreover , studies in both mammalian cells and Drosophila suggest Ras signalling can regulate dMyc levels and that Myc is required for Ras-induced growth[10 , 23 , 24 , 46 , 47] . Indeed , we found that the UAS-EGFR- and UAS-RasV12S35-induced proliferation of larval AMPs was blocked when we knocked down dMyc by expression of a UAS-dMyc RNAi construct ( S3A–S3C Fig ) . We therefore examined whether dMyc functions downstream of Ras in the control of Pol III . Using S2 cells we found that the increase in tRNA levels seen following RasV12 expression was blocked when cells were treated with dsRNA to knockdown dMyc ( Fig 5A ) . In contrast , we found that overexpression of dMyc in S2 cells was not able to induce tRNA synthesis when the Ras pathway was inhibited by treatment with the MEK inhibitor UO126 ( Fig 5B ) . Under these conditions of Ras pathway inhibition , however , dMyc mRNA levels were not affected ( Fig 5C ) and overexpressed dMyc was still able to significantly stimulate expression of Nop60B , PPAN and NOP5—three dMyc Pol II target genes ( Fig 5D ) –although the effect on PPAN and NOP5 was somewhat reduced . Nevertheless , these data suggest that U0126 does not simply abrogate dMyc’s ability to stimulate transcription of its target genes , and that dMyc is required , but not sufficient , to mediate the effects of Ras signalling on tRNA synthesis . These data suggest that Ras/ERK signalling can use an additional mechanism to control Pol III transcription . Another candidate that we considered as a mediator of Ras-induced tRNA synthesis was the conserved Pol III repressor , Maf1 . Studies in yeast , Drosophila and mammalian cells have shown that inhibition of Maf1 is the main way that the nutrient-dependent TORC1 kinase pathway stimulates Pol III and tRNA synthesis [39 , 48–51] . Knockdown of Drosophila Maf1 ( dMaf1 ) has been shown to promote tRNA synthesis , and to enhance tissue and body growth [40] . Here , we found that when we expressed UAS-dMaf1 RNAi in the Ras-responsive AMP cells during larval development using esg-GAL4ts , we observed a modest , but significant increase in the number of AMP cells per cluster ( S4A Fig ) . Although considerably weaker than the effect of Ras pathway activation ( e . g . see comparison with effect of UAS-EGFR , S4B Fig ) this effect of dMaf1 knockdown was similar to the increase in AMP numbers seen with overexpression of dMyc , another stimulator of tRNA synthesis and mRNA translation ( S4C Fig ) . We therefore next examined whether the Ras/ERK pathway functions to promote tRNA synthesis by inhibiting dMaf1 . We examined pre-tRNA levels using qRT-PCR in S2 cells , and , as described above , we saw that treatment of cells with the MEK inhibitor UO126 led to reduced tRNA synthesis ( Fig 6A and 6B ) . However , we found that this decrease in tRNA synthesis was reversed when cells were treated with dsRNA to knockdown dMaf1 levels ( Fig 6A and 6B ) . We observed similar effects when we used Northern blotting to measure pre-tRNA and tRNA levels ( S4D Fig ) . We also used treatment of cell with dsRNA to Ras to block Ras signalling , and saw a decrease in tRNA synthesis ( S4E Fig ) . However , as with UO126 treatment , we found that this decrease in tRNA synthesis caused by dsRNA to Ras was reversed by co-treatment of cells with dsRNA to dMaf1 . These data suggest that one main way that Ras/Erk signalling functions to promote tRNA synthesis is by inhibiting the Pol III repressor function of dMaf1 . Studies in both yeast and mammals indicate that Maf1can be regulated by controlling its nuclear localization ( e . g [50 , 52] ) . We first tested this in S2 cells using an antibody to endogenous dMaf1 . Under our normal media culture conditions , we observed that dMaf1 was localized throughout the cell ( Fig 6C ) . When we carried out antibody staining in dMaf1 dsRNA-treated cells ( which leads to a strong knockdown of both dMaf1 mRNA , S5A Fig , and dMaf1 protein , S5B Fig ) we saw minimal background staining , suggesting that the antibody is specific for dMaf1 ( S5C Fig ) . We found that treatment of cells with the MEK inhibitor U0126 lead to a significant increase in nuclear localization of dMaf1 ( Fig 6C and 6D ) , without having any effect on overall dMaf1 protein levels ( S5D Fig ) . We also found that genetic inhibition of Ras signalling in AMPs , by overexpression of dominant-negative Ras ( UAS-RasN17 ) , lead to an increase in nuclear localization of dMaf1 ( Fig 6E and 6F ) . Similar results were seen when we used expression of either UAS-EGFR or UAS-Ras RNAi to block Ras signalling in AMPs ( Supplemental Fig 6A ) . Thus , Ras/ERK signalling functions to prevent nuclear accumulation of dMaf1 , hence blocking its Pol III repressor activity and promoting tRNA synthesis . Previous studies showed that the TORC1 pathway can regulate Maf1 nuclear localization and repressor function via phosphorylation [48–51 , 53] . We therefore explored whether the Ras pathway could also control the phosphorylation status of dMaf1 in S2 cells . We used the phos-tag reagent , which slows the migration of phosphorylated proteins in SDS-PAGE gels , and hence helps resolve phosphorylated vs . non-phosphorylated versions of a protein on a western blot . For example , when we examined total ERK levels by western blotting following SDS-PAGE with Phos-tag , we observed two ERK bands . The relative levels of the upper band were reduced when we treated cells with the MEK inhibitor ( Fig 7A ) , while levels of the upper band were increased in cells overexpressing RasV12 ( Fig 7B ) , thus indicating this method can detect protein phosphorylation changes . We then examined dMaf1 protein levels in western blots following SDS-PAGE with Phos-tag . As with ERK , we observed two dMaf1 bands , and the relative levels of the upper band were reduced when we treated the sample with phosphatase prior to SDS-PAGE ( S6B Fig ) , suggesting this upper band is a phosphorylated version of dMaf1 . Also , like ERK , we found that relative levels of the upper band were reduced when we treated cells with the MEK inhibitor ( Fig 7A ) , while levels of the upper band were increased in cells overexpressing RasV12 ( Fig 7B ) . Together these data suggest that Ras signalling may regulate dMaf1 phosphorylation , and based on previous work with TORC1 signalling , this may be one way that Ras regulates dMaf1 nuclear vs . cytoplasmic localization ( Fig 7C ) .
We propose that stimulation of RNA polymerase III and tRNA synthesis contributes to the ability of the conserved Ras/ERK pathway to promotes mRNA translation and growth . Our data indicate that Ras can control Pol III by inhibiting the Maf1 repressor , in part by preventing its nuclear accumulation . Maf1 is a phospho protein and studies in yeast and mammalian cells have described how phosphorylation can regulate Maf1 nuclear localization . For example , both TORC1 and PKA can phosphorylate Maf1 on several conserved residues[48–53] . This phosphorylation prevents Maf1 nuclear accumulation and allows both kinases to stimulate Pol III . In contrast , dephosphorylation of Maf1 by both PP2A and PP4 protein phosphatases leads to nuclear accumulation of Maf1 and Pol III repression [54–56] . Thus , it is possible that ERK may function by promoting Maf1 phosphorylation–either directly or indirectly–to prevent its function . Other mechanisms may also be important for Ras to simulate tRNA synthesis . For example , one study in mammalian cells showed that ERK could phosphorylate and regulate Brf1 function [57] . Also , Ras was shown to upregulate TBP , which can increase transcription by all three RNA polymerases [41] , although we did not see a similar effect . Interestingly , we found that the decrease in tRNA synthesis caused by inhibiting Ras signalling could be completely reversed by dMaf1 knockdown . This result suggests that while Ras signalling may exert multiple effects to control Pol III transcription , inhibition of dMaf1 seems to be an important effector of Ras in the control of tRNA synthesis . Maf1 function is conserved suggesting that the Ras/ERK-dependent regulation of Maf1 and tRNA synthesis that we describe in Drosophila may operate in other organisms , particularly human cells . Our data using the phos-tag reagent suggest that one way that Ras/ERK signalling may control dMaf1 is via phosphorylation . Previous studies in both yeast and mammalian cells have shown that the TORC1 pathway can control the nuclear localization and repressor function of dMaf1 via phosphorylation of several conserved residues [48–50 , 53] . One can therefore speculate that Ras signalling may work in a similar manner . Although further studies are required to identify if ERK directly phosphorylates dMaf1 and to identify the phosphorylated residues , it is interesting to note that two of the conserved TORC1 phosphorylation sites on dMaf1 are serine residues followed by proline , which are sites that are often phosphorylated by ERK , a proline-directed kinase . We also show that the transcription factor dMyc is required for the effects of Ras on Pol III and tRNA synthesis . Previous work from both mammalian cells and Drosophila has shown that in some cells Ras can promote Myc levels and that Ras-mediated growth requires Myc function[10 , 23 , 24 , 46 , 47] . We previously showed that Drosophila Myc could stimulate expression of the Pol III transcription factor , Brf1 , and also other Pol III subunits [39] . In addition , Myc can directly interact with Brf1 and localize at Pol III to directly stimulate tRNA transcription in Drosophila and mammalian cells [39 , 44 , 45] . We suggest that both these effects are under the upstream control of Ras/ERK signalling and may , in part , explain the requirements for Myc in Ras-induced growth in both animal development and cancer . Given our findings with dMaf1 and dMyc , we attempted to address which of the two mechanisms—inhibition of dMaf1 or activation of Myc—might explain the main effects of Ras/ERK signalling on tRNA synthesis . To do this , we inhibited Ras/ERK signalling in S2 cells and then asked whether knockdown of dMaf1 or overexpression of dMyc could maintain tRNA synthesis . We found that , of these two manipulations , only dMaf1 inhibition could restore tRNA synthesis when ERK signalling was inhibited . We interpret these findings to suggest that , while dMyc is required for tRNA synthesis , it is the inhibition of dMaf1 that explains a substantial part of the mechanism of action of Ras/ERK signalling in the regulation of Pol III and tRNA synthesis . We previously showed that dMaf1 knockdown does not alter expression of dMyc target genes [39] , suggesting that enhancement of dMyc function doesn’t explain why Maf1 knockdown can maintain tRNA synthesis in cells in which Ras/ERK signalling is inhibited . Previous studies in mammalian cells have shown that Ras/ERK signalling can promote protein synthesis by stimulating translation initiation factor function . We suggest that inhibition of Maf1 represents another target of Ras/ERK signalling , and that the subsequent increase in tRNA levels may cooperate with enhanced translation initiation factor activity to promote maximal stimulation of mRNA translation . Most of the work on Ras-mediated gene expression has focused on the effect of several Pol II transcription factors identified downstream of Ras in Drosophila such as fos , pointed , and capicua [19–22] . Stimulation of Pol III transcription to enhance tRNA levels and mRNA translation may provide another layer of control on overall gene expression by Ras signalling . For example , translational control of cell cycle genes has been proposed as one way to couple growth signalling pathways to cellular proliferation [58 , 59] . Furthermore , selective translational regulation of certain mRNAs has been shown to regulate growth and metastatic behaviour of tumour cells [60–62] . It important to note though that we find that simply knocking down dMaf1 alone has only a modest effect on cell proliferation AMPs , compared to the strong hyperproliferative effect of overactivation of Ras signalling . This is likely because increasing Pol III is only one downstream effect of Ras signalling and that the full Ras effect on cell proliferation requires the coordinated increase in the expression of many genes . Indeed , it is likely that Ras stimulates the activity of all three RNA polymerases to drive cell growth and proliferation . Ras is one of the most often overactivated or mutated pathways in cancer , hence our findings may also have implications for processes that contribute to tumour growth and metastasis . Indeed , there is increasing appreciation for potential roles for alterations in tRNA biology in cancer cells [63] . For example , tRNA expression profiling has revealed that levels of many tRNAs are elevated in different cancer types [64 , 65] . Interestingly , these changes in tRNA levels have been shown to correlate with codon usage in mRNAs whose expression also changes in cancer cells [66] . Several studies have reported that increasing the levels of specific tRNAs can promote tumour growth and metastatic behavior [67–70] . Previous work also showed that increasing tRNA levels alone is sufficient to drive growth in Drosophila [40 , 71] . Hence , an increase in tRNA levels caused by oncogenic Ras signalling may be a driver of tumour growth and progression , rather than simply a consequence of increased growth . Ras also controls other process such as cell fate specification , differentiation and cell survival . Many of these effects are mediated through translation and so may also rely on the effects of Ras on tRNA synthesis .
Flies were raised on standard medium ( 150 g agar , 1600 g cornmeal , 770 g Torula yeast , 675 g sucrose , 2340 g D-glucose , 240 ml acid mixture ( propionic acid/phosphoric acid ) per 34 L water ) and maintained at 25°C , unless otherwise indicated . The following fly stocks were used: For all GAL4/UAS experiments , GAL4 lines were crossed to the relevant UAS line ( s ) and the larval or adult progeny were analyzed . Control animals were obtained by crossing the relevant GAL4 line to either w1118 or yw depending on the genetic background of the particular experimental UAS transgene line . For the esg-gal4ts system , larvae and were flies were initially raised at 18°C and then for each experiment they were shifted to 29°C to inactivate the temperature sensitive GAL80 and to allow GAL4-mediated transgene expression . Drosophila Schneider S2 cells were grown at 25°C in Schneider’s medium ( Gibco; 11720–034 ) supplemented with 10% fetal bovine serum ( Gibco; 10082–139 ) , 100 U/ml penicillin and 100 U/ml streptomycin ( Gibco; 15140 ) . Stably transfected inducible RasV12 cells were a gift from the lab of Marc Therrien [73] . Stably transfected inducible dMyc cells were a gift from the lab of Paula Bellosta [74] . Both RasV12 and dMyc expression are under the control of a metallothionein promoter . For all experiments RasV12 or dMyc were induced by addition of copper sulphate to the culture media . dsRNA Treatment of S2 cells: dsRNAs were synthesized with RiboMAX large-scale RNA production system ( Promega ) using PCR products from either cDNAs or genomic DNA ( primer sequences in S2 Table ) . Cells were pretreated with 15 μg of dsRNAs in the absence of serum for 30 mins and then 2 mls of media plus serum was added , and cells were then incubated for 96 to 120 hrs . Control cells were treated with ds RNA to Green Fluorescent Protein ( GFP ) . Cells were harvested by centrifugation at 4 °C and washed with cold PBS and frozen for RNA isolation or protein extraction . MEK inhibitor ( U0126 ) treatment of Drosophila S2 cells: S2 cells were cultured at 25°C in Schneider’s medium ( Gibco; 11720–034 ) supplemented with 10% fetal bovine serum ( Gibco; 10082–139 ) , 100 U/ml penicillin and 100 U/ml streptomycin ( Gibco; 15140 ) . Cells were treated with either 10 μM U0126 ( Promega Cat . No . V1121 ) or DMSO ( Sigma; D2650 ) for 2 hours . Then cells were washed twice with ice-cold PBS . Cells were then used to isolate RNA or make protein extracts as described below . Drosophila S2 cells were lysed with a buffer containing 20 mM Tris-HCl ( pH 8 . 0 ) , 137 mM NaCl , 1 mM EDTA , 25% glycerol , 1% NP-40 and with following inhibitors 50 mM NaF , 1 mM PMSF , 1 mM DTT , 5 mM sodium ortho vanadate ( Na3VO4 ) and Protease Inhibitor cocktail ( Roche Cat . No . 04693124001 ) and Phosphatase inhibitor ( Roche Cat . No . 04906845001 ) according to the manufacturer’s instruction . Drosophila S2 cells were lysed with a buffer containing 20 mM Tris-HCl ( pH 8 . 0 ) , 137 mM NaCl , 25% glycerol , 1% NP-40 and with following inhibitors 1 mM PMSF , 1 mM DTT and Protease Inhibitor cocktail ( Roche Cat . No . 04693124001 ) and Phosphatase inhibitor without EDTA . Phos-tag SDS-PAGE was prepared according to the manufacturer’s instruction ( Wako Chemicals USA , Inc ) . Cell lysates were separated on 12 . 5% SDS-polyacrylamide gel containing 20 uM Phos-tag acrylamide ( AAL-107 Wako Chemicals USA , Inc ) , and transferred onto PVDF membranes ( Bio Rad ) . Protein concentrations were measured using the Bio-Rad Dc Protein Assay kit II ( 5000112 ) . Protein lysates ( 15 μg to 30μg ) were resolved by SDS–PAGE and electrotransferred to a nitrocellulose membrane , subjected to Western blot analysis with specific antibodies , and visualized by chemiluminescence ( enhanced ECL solution ( Perkin Elmer ) . Brf primary antibodies were against a C-terminal fragment of Drosophila Brf , alpha-tubulin ( E7 , Drosophila Studies Hybridoma Bank ) , dMyc [24] , phospho-ERK ( Cell Signalling Technology 4370 ) and ERK ( Cell Signalling Technology 4695 ) . Peptide antiserum against Drosophila Maf1 was raised by immunizing rabbits with synthetic peptide LADFSPNFRC corresponding to residues 65–74 ( GL Biochem ( Shanghai ) Ltd ) . 10 μM puromycin was added to Drosophila S2 cell culture media and the cells were incubated with puromycin for 30 min at 25 °C . Cells were harvested by centrifugation at 4°C and washed with cold PBS . Cells were frozen on dry ice and then lysed according to the Western blot protocol described above and analyzed by SDS-PAGE and western blotting using an anti-puromycin antibody ( 3RH11 ) ( Kerafast , Catalog No . EQ0001 ) at 1:2000 dilution . Total RNA was extracted from Drosophila S2 cells using TRIzol . 5 μg total RNA was separated on a 5% denaturing polyacrylamide/urea gel and northern blotting was carried using alkaline transfer . Hybridization of tRNA probes were carried out as described in Roche DIG Easy Hyb ( Cat . No . 11603558001 ) . Digoxigenin-labelled probes were made by in vitro transcription using either full-length cDNAs or PCR fragments as templates . Primers used for PCR are included in S1 Table . Drosophila S2 cells were fixed in 4% paraformaldehyde at room temperature for 20 mins on cover slips . Cells were then washed with 1x PBS and permeabilized with 0 . 1% Triton X in PBS by washing 2x for 5 mins . Cells were blocked with 5% FBS , 0 . 1% Triton X in PBS for 2 hours . Primary dMaf1 antibody was diluted in 5% BSA in PBS at 1:500 dilution and incubated overnight at 4°C . Then washed 3x with 0 . 1% Triton X in PBS for 5 min each and Alexa 568 ( Molecular probes ) goat-anti rabbit secondary antibody was diluted at 1:400 in 5% BSA in PBS for 2 hours at room temperature . Then , cells were washed 3x with 0 . 1% Triton X in PBS for 5 min each and mounted using VectaShield mounting medium . Drosophila larvae were inverted and fixed in 8% paraformaldehyde/PBS at room temperature for 45 mins . After blocking for 2hrs in 1%BSA in PBS/0 . 1% Triton-X 100 , larval carcasses were incubated overnight in anti-dMaf1 antibody ( 1:1000 ) . Primary antibody staining was detected using Alexa 488 ( Molecular probes ) goat-anti rabbit secondary antibodies . For experiments looking at dMaf1 subcellular localization , we used Image J to measure dMaf1 staining intensity . Nuclear localization was measured was calculated by measuring the total intensity of signal in the nucleus and dividing this by the total intensity in the cytoplasm ( calculated as total overall cellular signal intensity minus total nuclear signal intensity ) . Total RNA was extracted using TRIzol according to manufacturer’s instructions ( Invitrogen; 15596–018 ) . RNA samples were DNase treated according to manufacturer’s instructions ( Ambion; 2238G ) and reverse transcribed using Superscript II ( Invitrogen; 100004925 ) . The generated cDNA was used as a template to perform qRT–PCRs ( ABI 7500 real time PCR system using SyBr Green PCR mix ) using specific primer pairs ( sequences available upon request ) . PCR data were normalized to either actin or Glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) levels . Each experiment was independently repeated a minimum of three times . All primer sequences are in S3 Table . Polysome gradient centrifugation was performed as described [40] . 100 million Drosophila S2 cells were lysed in 1 ml of lysis buffer ( 25 mM Tris pH 7 . 4 , 10 mM MgCl2 , 250 mM NaCl , 1% Triton X-100 , 0 . 5% sodium deoxycholate , 0 . 5 mM DTT , 100 mg/ml cycloheximide , 1 mg/ml heparin , Complete mini Roche protease inhibitor ( Roche ) , 2 . 5 mM PMSF , 5 mM sodium fluoride , 1 mM sodium orthovanadate and 200 U/ml ribolock RNAse inhibitor ( Fermentas ) using a Dounce homogenizer . The lysates were centrifuged at 15 , 000 rpm for 20 minutes and the supernatant was removed carefully . 150 to 250 g μg RNA was layered gently on top of a 15–45% w/w sucrose gradient ( made using 25 mM Tris pH 7 . 4 , 10 mM MgCl2 , 250 mM NaCl , 1 mg/ml heparin , 100 mg/ml cycloheximide in 12 ml polyallomer tube ) and centrifuged at 37 , 000 rpm for 150 minutes in a Beckmann Coulter Optima L-90K ultracentrifuge using a SW-41 rotor . Polysome profiles were obtained by pushing the gradient using 70% w/v Sucrose pumped at 1 . 5 ml/min into a continuous OD254 nm reader ( ISCO UA6 UV detector ) showing the OD corresponding to the RNA present from the top to the bottom of the gradient . All qRT-PCR data and quantification of immunostaining data were analyzed by Students t-test , or two-way ANOVA followed by post-hoc students t-test where appropriate . All statistical analysis and data plots were performed using Prism software . In all figures , statistically significant differences are presented as: * p<0 . 05 , ** p<0 . 005 , *** p<0 . 0005 , **** p<0 . 0001 . | The Ras oncogene is one of the primary drivers of cell and tissue growth in both normal development and in diseases such as cancer . In this report , we identify the stimulation of tRNA synthesis as an important mechanism by which Ras functions . Using fruit fly genetics , we show that Ras promotes tRNA synthesis by inhibiting Maf1 , a protein that normally inhibits RNA polymerase III , the enzyme complex that stimulates tRNA synthesis . We further show that stimulation of tRNA synthesis is required for Ras to promote growth in two important cell types—stem cells and epithelial cells . This work provides new insight into mechanisms that are important for growth and that may contribute to cancer . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"transfer",
"rna",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"rna",
"interference",
"animals",
"animal",
"models",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"experimental",
"organism",
"systems",
"rn... | 2018 | Ras/ERK-signalling promotes tRNA synthesis and growth via the RNA polymerase III repressor Maf1 in Drosophila |
Phenotypic states and evolutionary trajectories available to cell populations are ultimately dictated by complex interactions among DNA , RNA , proteins , and other molecular species . Here we study how evolution of gene regulation in a single-cell eukaryote S . cerevisiae is affected by interactions between transcription factors ( TFs ) and their cognate DNA sites . Our study is informed by a comprehensive collection of genomic binding sites and high-throughput in vitro measurements of TF-DNA binding interactions . Using an evolutionary model for monomorphic populations evolving on a fitness landscape , we infer fitness as a function of TF-DNA binding to show that the shape of the inferred fitness functions is in broad agreement with a simple functional form inspired by a thermodynamic model of two-state TF-DNA binding . However , the effective parameters of the model are not always consistent with physical values , indicating selection pressures beyond the biophysical constraints imposed by TF-DNA interactions . We find little statistical support for the fitness landscape in which each position in the binding site evolves independently , indicating that epistasis is common in the evolution of gene regulation . Finally , by correlating TF-DNA binding energies with biological properties of the sites or the genes they regulate , we are able to rule out several scenarios of site-specific selection , under which binding sites of the same TF would experience different selection pressures depending on their position in the genome . These findings support the existence of universal fitness landscapes which shape evolution of all sites for a given TF , and whose properties are determined in part by the physics of protein-DNA interactions .
A powerful concept in evolution is the fitness landscape: for every possible genotype there is a number , known as the genotypic fitness , that characterizes the evolutionary success of that genotype [1] . Evolutionary success is typically quantified as the probability of surviving to reproduce , number of offspring , growth rate , or a related proxy [2] , [3] . The structure of the fitness landscape is key to understanding the evolutionary fates of populations . Most traditional studies of molecular evolution rely on simplified models of fitness landscapes [3]–[6] or empirical reconstructions of landscapes based on limited experimental data [3] , [7]–[10] . However , fitness landscapes are fundamentally shaped by an intricate network of interactions involving DNA , RNA , proteins , and other molecular species present in the cell . Thus we should be able to cast these landscapes in terms of biophysical properties such as binding affinities , molecular stabilites , and degradation rates . The increasing availability of quantitative high-throughput data on in vitro and in vivo molecular interactions has led to growing efforts aimed at developing models of evolution that explicitly incorporate the underlying biophysics [11]–[25] . These models combine evolutionary theory with physical models of molecular systems , for example focusing on how protein folding stability or specificity of intermolecular interactions shapes the ensemble of accessible evolutionary pathways and steady-state distributions of biophysical phenotypes . Evolution of gene regulation is particularly well-suited to this type of analysis . Gene activation and repression are mediated by binding of transcription factors ( TFs ) to their cognate genomic sites . These binding sites are short nucleotide sequences , typically 5–25 bp in length , in gene promoters that interact specifically with TF DNA-binding domains [26] . In eukaryotes , a given TF can have numerous binding sites in the genome , and many genes are regulated by several TFs [26] , [27] . Understanding TF-mediated regulation is key to understanding complex regulatory networks within eukaryotic cells − one of the main challenges facing molecular biology . Moreover , the availability of high-throughput data on the genomic locations of TF binding sites [28]–[31] and on TF-DNA energetics [32]–[35] make it possible to develop biophysical models of evolution of gene regulation . Here we consider evolution of TF binding sites in the yeast Saccharomyces cerevisiae . We study how energetics of protein-DNA interactions affect the structure of the binding site fitness landscape . In a significant extension of previous work which analyzed a single yeast TF [22] , we consider a collection of 25 S . cerevisiae TFs for which models of TF binding energetics were built using high-throughput in vitro measurements of TF-DNA interactions [35] . We focus on 12 TFs for which sufficient data on genomic sites [31] are also available . We use a model of monomorphic populations undergoing consecutive substitutions [19] , [36]–[38] to infer fitness landscapes , as functions of TF-DNA binding energy , from observed distributions of TF binding sites in the yeast genome [22] . In contrast to the previous work [22] , we rationalize these fitness landscapes in terms of a simple parametric model based on thermodynamics of TF-DNA binding , obtaining explicit values of effective evolutionary parameters . Our analysis sheds light on the genome-wide importance of TF-DNA interactions in regulatory site evolution . Moreover , we investigate the hypothesis that universal biophysical constraints , rather than site-specific selective pressures , dominate evolution of regulatory sites . We test the relationship between TF binding energies and various biological properties , such as the essentiality of the corresponding gene [39] . We find no clear relationship between physical and biological properties of TF sites , which indicates that evolution of site energetics is largely insensitive to site-specific biological functions and is therefore driven by global biophysical constraints .
We now turn to considering the evolution of TF binding sites in S . cerevisiae . How well does S . cerevisiae satisfy the assumptions of our evolutionary model ? First of all , S . cerevisiae is not a purely haploid organism but rather goes through haploid and diploid stages . In S . paradoxus , most of the reproduction is haploid and asexual with 1000 generations spent in the haploid stage for each generation in the diploid stage , and heterozygosity is low [45] . Based on the analysis of yeast genomes , wild yeast populations show limited outcrossing and recombination and are geographically distinct [46] . Thus , S . cerevisiae may be regarded as haploid to a reasonable approximation , with sufficient recombination during the diploid stages to unlink TF binding sites . This is consistent with our model , which assumes a haploid population and independent evolution of binding sites . We next consider whether natural populations of S . cerevisiae are within the mutation rate limits required for monomorphism . The mutation rate for S . cerevisiae has been estimated to be mutations per bp per cell division [45] . Assuming binding site loci of length , the bound on the effective population size is , below which the population will be monomorphic . This is close to the estimated effective population size of S . cerevisiae of individuals [45] , based on the analysis of neutral regions in the yeast genome . Thus it is plausible that S . cerevisiae population sizes are below or near the limit for monomorphism , justifying the use of Eq . 3 . Furthermore , in S . cerevisiae and S . paradoxus the proportion of polymorphic sites in a population has been found to be about 0 . 001 [45] , [47] , [48] , generally with no more than two alleles segregating at any one site [45] . According to this estimate , we expect about 1% of binding sites of length 10 bp to be polymorphic , corresponding to an average polymorphism of 1 . 01 in Fig . 2C . For S . cerevisiae , the equilibration time estimate is generations , or about years with 8 generations per day [49] . This is several times less than the 5–10 million years of divergence time for the most recent speciation event with S . paradoxus [50] . Thus steady state may plausibly be reached over evolutionary times scales for a fast-reproducing organism like S . cerevisiae .
In this work , we have considered how fitness of a single-cell eukaryote S . cerevisiae is affected by interactions between TFs and their cognate genomic sites . Changing the energy of a site or creating new sites in gene promoters may change how genes are activated and repressed , which in turn alters the cell's chances of survival . Under the assumptions of a haploid monomorphic population in which the evolution of binding sites has reached steady state , the fitness landscape as a function of TF binding energy can be inferred from the distribution of TF binding sites observed in the genome , using a biophysical model which assigns binding energies to sites . We use a simple energy matrix model of TF-DNA energetics in which the energy contribution of each position in the site is independent of all the other positions . The energy matrix parameters are inferred from a high-throughput data set in which TF-DNA interactions were studied in vitro using a microfluidics device [35] . We consider two types of fitness functions: Fermi-Dirac , which appears naturally from considering TF binding as a two-state process ( Eq . 1 ) , and exponential , which is motivated by the observation that for many TFs , the logarithm of fitness appears to decrease linearly as energy increases . A single fitness landscape for all genomic binding sites of a given TF can only exist in the absence of site-specific selection . Indeed , it is possible that TF sites experience different selection pressures depending on the genes they regulate: for example , sites in promoters of essential genes may be penalized more for deviating from the consensus sequence . In this case , the fitness function is an average over all sites which evolve under different selection constraints: as an extreme example , consider the case where each site has a Fermi-Dirac fitness function ( Eq . 7 ) with different parameters , , and . The resulting observed distribution of energies would then be the average of the distributions predicted by Eq . 5: ( 18 ) which defines the “average” fitness function with effective parameters , , , . Thus the fit can be carried out even in the presence of site-dependent selection , but the fitted parameters correspond to fitness functions of individual sites only in an average sense . To gauge the importance of site-specific selection in TF binding site evolution , we have performed several statistical tests aimed at discovering correlations between binding site energies and biological properties of the sites and the genes they regulate . These tests considered gene essentiality , growth rates of strains with nonessential genes knocked out , gene expression levels , ratios based on alignments with S . paradoxus , and the distance of the site to the TSS . We find no consistent correlations among these properties , indicating that for a given TF , the evolution of regulatory sites is largely independent of the properties of regulated genes . Previously , low correlations have been observed between essentiality and conservation of protein and coding sequences [56]–[62] , which has fueled considerable speculation as it contradicts the prediction of the neutral theory of evolution that higher selection pressures lead to lower evolutionary rates . It has also been found that the growth rates of strains with nonessential genes knocked out are significantly ( though weakly ) correlated with conservation of those genes [63] . It has therefore been suggested that selection pressures are so strong that only the most nonessential genes experience significant genetic drift [56] . Previous studies have also found that gene expression levels are a more reliable ( though still very weak ) predictor of selection pressures than essentiality [60] , but we do not find this to be the case for TF binding sites , nor do we observe a consistently significant correlation between gene expression levels and TF binding energies . Available data does not rule out the possibility of time-dependent selection in combination with forms of site-dependent selection for which we have not accounted . In this scenario , the variation in site binding affinity is not due to genetic drift , but to variable selection pressures across sites and over time , such that the sites are strongly tuned to particular binding energies which change from locus to locus . Indeed , there is evidence that there is frequent gain and loss of TF binding sites and that the gene regulatory network is highly dynamic [64]–[70] . However , it is possible that rapid turnover of binding sites in eukaryotes may be due to evolution acting on whole promoters rather than individual binding sites . Many promoters contain multiple binding sites for a single TF , and it may be that while individual binding sites are lost and gained frequently , the overall binding affinity of a promoter to a TF may be held constant [71]–[73] . Our evolutionary model can account for this scenario using a promoter-level fitness function , which we intend to consider in future work . Out of 12 TFs with sufficient binding site data , five have , indicating a large fitness penalty for deleting such sites . However , this conclusion is strongly supported by the AIC differences between unconstrained and non-lethal Fermi-Dirac fits for only one TF , RPN4 ( Table 2 ) . RPN4 is classified as nonessential in the Yeast Deletion Database . It may be that this misclassification is due to a mismatch between genomic sites , in which the core motif is preceded by , and the energy matrix in which the binding energies upstream of the core motif are non-specific ( Table S2 ) . We also classify REB1 and MCM1 binding sites as nonessential , although knocking out these TFs is lethal in yeast . This discrepancy may be due to a minority of essential sites being averaged with the majority of nonessential sites to produce a single fitness function , as described above . In addition , although a penalty for deleting any single site may be small , the cumulative penalty for deleting all sites ( or , equivalently , deleting the TF ) may be lethal . We find that in 10 out of 12 cases , fitting an exponential fitness function is less supported by the data than fitting a Fermi-Dirac function ( Table 2 ) . This is interesting since constructing a position-specific weight matrix by aligning genomic sites is a common practice which implicitly assumes factorization of exponential fitness and independence of each position in the binding site . Our results show the limitations of this approximation . It is important to note that a key difference between the Fermi-Dirac fitness landscape and the exponential landscape is that the former contains magnitude epistasis [8] , [25] ( i . e . , the magnitude of a mutation's fitness effect depends on the rest of the site sequence ) , while the latter is non-epistatic . Thus , our results indicate that epistasis is widespread in the evolution of TF binding sites [22] . Finally , we find that depending on the TF , the distribution of TF binding energies may fall on the exponential tail , across the threshold region , or on the saturated plateau where the sites are always occupied ( Table 1 ) . In the first two categories , variation of TF concentration in the cell will lead to graded responses , which may be necessary to achieve precise and coordinated gene regulation . In the third regime , TF binding is robust and not dynamic . We also find that the fitted inverse temperature is typically not close to the value based on room temperature ( Table 1 ) . In particular , our analysis of the variation of selection strength with indicates that selection appears to be stronger for most TFs than expected from pure biophysical considerations , suggesting the presence of additional selection pressures beyond those dictated by the energetics of TF-DNA binding .
In the limit where is the mutation rate per nucleotide , is the number of nucleotides in a locus , and is an effective population size , mutations are sufficiently rare that each new mutation either fixes or goes extinct before the next one arrives [41] . Thus populations evolve by sequential substitutions of new mutations at a locus , which consist of a single new mutant arising and then fixing . The rate at which a given substitution occurs is thus given by the rate of producing a single mutant times the probability that the mutation fixes [36]: ( 19 ) where is the mutation rate from genotype to and is the probability that a single mutant fixes in a population of wild-type . We will assume that is nonzero only for sequences and differing by a single nucleotide . Given an ensemble of populations evolving with these rates , we can define to be the probability that a population has genotype at time . This probability evolves over time according to the master equation ( 20 ) where is the set of all possible genotypes at the locus of interest . This Markov process is finite and irreducible , since there is a nonzero probability of reaching any sequence from any other sequence in finite time . Hence it has a unique steady-state distribution satisfying [74] ( 21 ) For population models obeying time reversibility , we can show that the steady-state distribution must have the form in Eq . 3 [38] . We assume the fixation probability depends only on the ratio of mutant to wild-type fitnesses: . This occurs in most standard population models and is expected whenever only relative fitness matters ( e . g . , when the total population size is constant ) . If the population dynamics are time reversible , the substitution rates and steady state must obey the detailed balance relation . Assuming the neutral dynamics also obey detailed balance , , we can show that ( 22 ) where . Equation 22 implies that , leading to for some exponent . It can be shown that must be proportional to the effective population size [38]; for the Wright-Fisher model , . Now Eq . 3 follows from ( 23 ) This form of the steady state assumes only time reversibility and dependence on fitness ratios; otherwise , any form of the fixation probability must satisfy it . While many population models do not obey time reversibility exactly , it can be shown that even these irreversible models satisfy Eq . 3 to a very good approximation [38] . For a given TF , let be the set of binding site sequences and the parameters of the fitness function ( Eq . 7 ) . The log-likelihood is given by ( 24 ) where is the fitness function , and is the normalization . Because the log-likelihood function has degenerate or nearly-degenerate regions in the parameter space of , instead of maximizing by gradient ascent we obtain a global map of the likelihood by calculating the function over a mesh of points in the following parameter domain: generated from for running from to in steps of ; in steps of ; generated from for running from to in steps of ; and generated from for running from to in steps of . Our predicted maximum is the maximum likelihood point in the mesh , which is sufficiently fine to estimate all fitting parameters . We have made the code for this procedure and for the analysis of site-specific selection available at www . physics . rutgers . edu/~morozov/publications . html . We consider a haploid asexual Wright-Fisher process [43] . The population consists of organisms , each with a single locus of nucleotides . The new generation is created by means of a selection step and a mutation step . In the selection step , sequences from the current population are sampled with replacement , weighted by their fitness , to construct a new population of size . In the mutation step , each position in all sequences is mutated with probability . For simplicity , the mutation rates between all pairs of nucleotides are the same . We characterize the difference between the distribution expected by our model , ( Eq . 3 ) , and the distribution observed in simulations , , using the total variation distance ( TVD ) : ( 25 ) The TVD ranges from zero for identical distributions to unity for completely non-overlapping distributions . We calculate the TVD for the distributions in energy space , where the sum in Eq . 25 is over discrete energy bins ( we bin the observed sequences by energy by dividing the range from the minimum to the maximum sequence energy for a particular energy matrix into 100 bins of equal size ) . We begin by randomly generating the energy matrix parameters Each in the energy matrix is sampled from a uniform distribution and then rescaled such that the distribution of all sequence energies has standard deviation of 1 . 0 . This is achieved by dividing all entries in the energy matrix by a factor : ( 26 ) where is the energy matrix element for base at position , is the binding site length , is the average energy contribution at position , and is the background probability of nucleotide ( for all in our simulations ) . It can be shown that is the standard deviation of the random sequence energy distributution , which is approximately Gaussian [11] . We generate the energy matrix once and use it in all subsequent simulations and maximum likelihood fits . We perform the Wright-Fisher simulations in a range of mutation rates from to with a “non-lethal” Fermi-Dirac fitness function ( Eq . 7 with , , and ) . We run simulations for each mutation rate for steps , enough to reach steady state . Each simulation starts from a monomorphic population with a randomly chosen sequence . We construct the steady state distribution for each mutation rate by randomly choosing a single sequence from the final population of each simulation . Collected across all simulations , these are used to construct a distribution of sequences at each mutation rate . Additionally , we record the average final number of unique sequences at each mutation rate . We perform another set of Wright-Fisher simulations with the same fitness function and energy matrix as above , and . We run simulations , each starting from the same monomorphic population with a specific sequence of . At regular intervals in each simulation , we record a randomly chosen sequence from the population . Collected across all simulations , these are used to construct a distribution of sequences at each point in time . We obtain curated binding site locations for 125 TFs from Ref . [31] , which provides a posterior probability that each site is functional based on cross-species analysis . We only consider sites with a posterior probability above 0 . 9 . For this analysis , we use the Saccharomyces Genome Database R53-1-1 ( April 2006 ) build of the S . cerevisiae genome . We obtain position-specific affinity matrices ( PSAMs ) for a set of 26 TFs from an in vitro microfluidics analysis of TF-DNA interactions [35] . This study provides PSAMs for each TF determined using the MatrixREDUCE package [34] . We convert the elements of the PSAM to energy matrix elements using , where at room temperature . For each of these 26 TFs , genomic sites are available in Ref . [31] , although we neglect PHO4 since it does not have any binding sites above the 0 . 9 threshold of Ref . [31] , leaving us with 25 TFs for which both an energy matrix and a set of genomic binding sites are available . We align the binding site sequences from Ref . [31] to the corresponding energy matrices , choosing the alignment that produces the lowest average binding energy for the sites . The Yeast Deletion Database classifies genes as essential , tested ( nonessential ) , and unavailable , which number 1156 , 6343 , and 529 , respectively [39] , [51] . For each essential or tested gene , we determine all TF binding sites less than 700 bp upstream of the gene's transcription start site ( on either strand ) , which we designate as the sites regulating that gene . Growth rates for nonessential knockout strains are provided under YPD , YPDGE , YPG , YPE , and YPL conditions , relative to wild-type . We choose the lowest of these growth rates to represent the fitness effect of the knockout . To measure the rate of nonsynonymous substitutions , we align the non-mitochondrial , non-retrotransposon ORFs taken from the Saccharomyces Genome Database R64-1-1 ( February 2011 ) build [75] of S . cerevisiae to those of S . paradoxus using ClustalW [76] . We measure the rate of nonsynonymous mutations using PAML [77] . We ran PAML with a runMode of −2 ( pairwise comparisons ) and the CodonFreq parameter ( background codon frequency ) set to 2; we also tested CodonFreq set to zero and obtained very similar results . We find the rate of nonsynonymous substitutions to be 0 . 04 , and a Spearman rank correlation of ( ) between growth rate of knockouts and the nonsynonymous substitution rate of the knocked-out gene . This is consistent with the results of Ref . [58] , which found the rate of substitutions to be 0 . 04 and the rank correlation between growth rate and substitution rate to be ( ) . To compare binding energy to evolutionary conservation , we calculate the mean Hamming distance between S . cerevisiae sites and corresponding sites in S . paradoxus [31] . To test for significance in the difference of mean energies and Hamming distances of sites regulating essential and nonessential genes , we use a null model which assumes that the sites were randomly categorized into essential and nonessential . We randomly choose a subset of the sites in our dataset to be “nonessential , ” equal in size to the number of sites regulating nonessential genes as classified by the Yeast Deletion Database . By repeating this procedure times , we build a probability distribution for the difference in the means of the nonessential and essential groups . The -value is the probability of obtaining a difference in the means greater than or equal in magnitude to the empirically measured value . We construct the neutral probability of a sequence with length as ( 27 ) where is the background probability of a nucleotide , and is the background probability of a dinucleotide . These probabilities are determined from mono- and dinucleotide frequencies in the intergenic regions of the S . cerevisiae genome ( build R61-1-1 , June 2008 ) . We project into energy space using Eq . 2 to obtain , the neutral distribution of binding energies for sequences of length . If intergenic sequences evolve under no selection with respect to their TF-binding energy , the neutral distribution of site energies should closely match the actual distribution of -mer sequences obtained from intergenic regions . Table S2 , column B shows that these two distributions match very well except at the low-energy tail , which is enriched in functional binding sites . Note that accounting for dinucleotide frequencies is important; mononucleotide frequencies alone are insufficient to reproduce the observed distribution [54] . | Specialized proteins called transcription factors turn genes on and off by binding to short stretches of DNA in their regulatory regions . Precise gene regulation is essential for cellular survival and proliferation , and its evolution and maintenance under mutational pressure are central issues in biology . Here we discuss how evolution of gene regulation is shaped by the need to maintain favorable binding energies between transcription factors and their genomic binding sites . We show that , surprisingly , transcription factor binding is not affected by many biological properties , such as the essentiality of the gene it regulates . Rather , all sites for a given factor appear to evolve under a universal set of constraints , which can be rationalized in terms of a simple model inspired by transcription factor – DNA binding thermodynamics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"evolutionary",
"modeling",
"biophysics",
"theory",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"biophysics"
] | 2014 | Biophysical Fitness Landscapes for Transcription Factor Binding Sites |
Members of the COE family of transcription factors are required for central nervous system ( CNS ) development . However , the function of COE in the post-embryonic CNS remains largely unknown . An excellent model for investigating gene function in the adult CNS is the freshwater planarian . This animal is capable of regenerating neurons from an adult pluripotent stem cell population and regaining normal function . We previously showed that planarian coe is expressed in differentiating and mature neurons and that its function is required for proper CNS regeneration . Here , we show that coe is essential to maintain nervous system architecture and patterning in intact ( uninjured ) planarians . We took advantage of the robust phenotype in intact animals to investigate the genetic programs coe regulates in the CNS . We compared the transcriptional profiles of control and coe RNAi planarians using RNA sequencing and identified approximately 900 differentially expressed genes in coe knockdown animals , including 397 downregulated genes that were enriched for nervous system functional annotations . Next , we validated a subset of the downregulated transcripts by analyzing their expression in coe-deficient planarians and testing if the mRNAs could be detected in coe+ cells . These experiments revealed novel candidate targets of coe in the CNS such as ion channel , neuropeptide , and neurotransmitter genes . Finally , to determine if loss of any of the validated transcripts underscores the coe knockdown phenotype , we knocked down their expression by RNAi and uncovered a set of coe-regulated genes implicated in CNS regeneration and patterning , including orthologs of sodium channel alpha-subunit and pou4 . Our study broadens the knowledge of gene expression programs regulated by COE that are required for maintenance of neural subtypes and nervous system architecture in adult animals .
The Collier/Olfactory-1/Early B-cell factor ( COE ) family of transcription factors is necessary for animal development . COE proteins possess an atypical HLH domain and a unique zinc finger DNA binding domain conserved across metazoans [1] . Invertebrates encode a single homolog of COE , with roles in mesoderm and ectoderm development [2] , [3] , whereas vertebrates have four COE paralogs with functions in diverse cell types including B-cells and adipocytes [4] . In the central nervous system ( CNS ) , COE regulates neuronal differentiation , migration , axon guidance , and dendritogenesis during development [2] , [3] , [5]–[13] and maintains neuronal identity throughout adulthood [14] , [15] . COE proteins have also been proposed to function as tumor suppressors [16] and are associated with cancers such as acute lymphoblastic leukemia and glioblastoma [17]–[20] . However , the specific genetic programs regulated by these genes in adult stem cells and mature neurons remain poorly understood . Stem cells can be studied to determine how transcriptional regulators orchestrate developmental processes or cause disease [21] . An excellent animal model to investigate stem cell regulation in vivo is the freshwater planarian Schmidtea mediterranea [22] . S . mediterranea has the ability to regenerate all tissue types from a population of adult stem cells ( called neoblasts ) . These cells constitute approximately 10–20% of all the cells in the animal and include pluripotent [23] and lineage-committed neoblasts [24]–[29] . The planarian CNS is composed of two cephalic ganglia and a pair of ventral nerve cords that run along the length of the animal , which are comprised of molecularly diverse neuronal subtypes that are regenerated within days after injury or amputation [30]–[32] . Functional analysis of transcription factors in planarians using RNA interference ( RNAi ) has begun to identify regulatory molecules required for the generation and maintenance of specific neuronal subpopulations in the CNS such as serotonergic and cholinergic neurons [24]–[27] , [33]–[35] . Thus , planarians are outstanding organisms to study basic mechanisms that underlie stem cell-based maintenance and regeneration of the adult CNS . A previous functional screen for transcription factors encoding a helix-loop-helix domain identified a planarian coe homolog that is expressed in a small population of neural-committed stem cells ( approximately 4–7% of the neoblast pool ) and in neurons [24] . We showed that animals fed dsRNA designed to silence coe expression ( coe ( RNAi ) animals ) regenerated abnormal brains; furthermore , uninjured coe ( RNAi ) planarians displayed behavioral defects and reduced expression of neural subtype-specific genes [24] . In this study , we sought to identify genes regulated by coe with roles in CNS renewal by comparing the transcriptome profiles of uninjured control and coe ( RNAi ) animals , uncovering differentially expressed genes with predicted roles in CNS function . We validated a subset of these genes by testing for loss of expression after coe knockdown and visualizing their expression in coe+ cells . These analyses revealed a set of nine candidate targets of coe in adult neurons , many of which are important for neuronal subtype identity ( e . g . , ion channels , neuropeptides , and neurotransmitters ) . In addition , our findings demonstrate that coe functions to drive gene expression in multiple neuronal classes , including excitatory and inhibitory neurons . To gain insights into the roles candidate COE targets play in CNS turnover and repair , we analyzed the function of downregulated transcripts using RNAi . Our functional screen identified several genes required for CNS regeneration , including homologs of a voltage-gated sodium channel α-subunit ( scna-2 ) and the transcription factor pou4l-1 . Our results suggest that COE is required for the expression of neural-specific genes in differentiating and mature neurons , a function that is essential to maintain CNS architecture and regulate neuronal regeneration .
Using an optimized whole-mount in situ hybridization protocol ( WISH ) ( see Materials and Methods ) , we found that coe mRNA was primarily restricted to neurons in S . mediterranea ( Fig . 1A ) . In agreement with our previous findings [24] , we also observed coe transcripts in a subset of cycling stem cells ( h2b+ ) ( Fig . 1B–C ) . We previously reported that coe ( RNAi ) animals regenerate cephalic ganglia that fail to connect at the anterior commissure and have significantly smaller brains with fewer cpp-1+ , npp-4+ , and npy-2+ neurons when compared to the controls [24] . This defect is not restricted to the anterior portion of the animal . Additional experiments showed coe ( RNAi ) animals do not properly regenerate their ventral nerve cords ( Fig . S1A–B ) . Moreover , analysis of the brain patterning defect using anti-VC-1 , a marker of the photoreceptor neurons and their axons , revealed that the optic chiasm failed to connect at the midline in coe ( RNAi ) animals ( Fig . S1C ) . These data demonstrate that coe is essential for neuronal regeneration at both anterior and posterior facing wounds and that coe regulates genes required for reestablishing midline patterning following brain amputation . In addition , we previously noted that silencing of coe in intact uninjured animals results in a reduction of ChAT+ and pc2+ neurons near the anterior commissure and a loss of cpp-1+ neurons . Following the 6th feeding of coe dsRNA , 100% of the animals exhibited impaired negative phototaxis [24] . To investigate the specificity of the coe knockdown phenotype on the CNS , we examined the effect of coe RNAi on the intestine and muscle as representative endodermal or mesodermal tissues , respectively . We hybridized uninjured control and coe ( RNAi ) animals with riboprobes specific to ChAT ( as a positive control ) , mat [36] , and collagen [37] . As expected , we observed a decrease in ChAT+ neurons in the head [24] and noted a decrease in ChAT expression throughout the animal ( Fig . 2A ) ; by contrast , we did not observe a change in the spatial distribution of mat or collagen following coe knockdown ( Fig . 2B–C ) . To quantify the effect of coe RNAi treatments on the expression of ChAT , mat and collagen , we measured relative mRNA levels by reverse transcription quantitative PCR ( RT-qPCR ) . First , we confirmed coe knockdown led to a significant decrease in the relative expression of coe mRNA ( down 60%±16% compared to the controls; Fig . 2D ) . Measurement of ChAT , mat and collagen from coe ( RNAi ) planarians revealed that ChAT mRNA levels were significantly down ( 45%±15% ) compared to control animals; in contrast to ChAT , the relative mRNA levels of mat or collagen were not affected by coe RNAi treatment ( Fig . 2D ) . Combined with our previous work [24] , these results strongly suggest that coe knockdown specifically affects gene transcription in the nervous system and does not cause obvious defects in other tissues such as the intestine or muscle . Furthermore , our results are consistent with reports demonstrating that COE is required to maintain cholinergic and peptidergic neuronal subtype-specific gene expression in Caenorhabditis elegans and Drosophila melanogaster [14] , [15] . To investigate if the inhibition of coe perturbs nervous system architecture downstream of gene expression changes , we labeled neuronal cell bodies and their projections using anti-CRMP-2 , which labels a subset of neuronal cell bodies and their axon projections , and anti-β-tubulin to visualize nerve projections ( Fig . 3A–C ) . In coe ( RNAi ) animals , we observed a striking decrease in axon projections labeled by anti-CRMP-2 and anti-β-tubulin compared to the controls; however , expression of CRMP-2 was retained in the cell bodies ( Fig . 3C ) . In addition , when we labeled sensory neurons using cintillo [38] , coe ( RNAi ) animals exhibited significantly fewer cintillo+ cells ( Fig . 3D ) . Our results strongly suggest that nervous system architecture is severely reduced or lost in the absence of coe . These structural defects likely underlie the behavioral abnormalities observed in coe-deficient planarians . Although COE has been shown to drive differentiation of several classes of neurons during development [39] , the transcriptional programs controlled by this transcription factor in adult nervous system function are poorly defined . We reasoned that the CNS-specific coe RNAi phenotype in intact planarians represents an excellent opportunity to identify gene expression programs controlled by COE in the post-embryonic nervous system . Thus , we used comparative mRNA sequencing ( RNA-seq; see Materials and Methods ) to sequence mRNAs isolated from uninjured controls and coe ( RNAi ) animals one week after the 6th RNAi treatment , which was the point in time we consistently observed behavioral defects and loss of neural-specific gene expression in 100% of coe-deficient animals and did not detect overt defects in other tissues ( Fig . 2 ) . RNA-seq analysis identified 909 differentially expressed genes; 397 were downregulated , and 512 were upregulated ( Table S1 ) . Functional annotation using DAVID software showed that the set of downregulated genes was significantly enriched for Gene Ontology ( GO ) terms associated with “ion channel , ” “neuronal activities , ” “nerve-nerve synaptic transmission , ” “voltage-gated ion channel , ” and “cell adhesion molecule”; by contrast , the upregulated genes were enriched for GO terms associated with “cytoskeletal protein” and “muscle development” ( Table 1 ) . coe mRNAs were not detected in a muscle pattern ( Fig . 1 ) , nor did we detect overt phenotypes associated with muscle differentiation ( Fig . 2 ) . However , the RNA-seq data raised the possibility that coe might negatively regulate mesoderm specification , which is required for muscle development [3] , [40] . It is possible upregulation of muscle genes is an indirect consequence of a loss of nervous system influence such as cholinergic transmission and/or neuropeptide regulation . Previous studies have demonstrated cholinergic neurotransmission is required for coordinated muscle contractions in planarians [41]–[43] . Thus , we speculate that loss of nervous system modulation disrupts muscle homeostasis and leads to changes in expression of muscle-related genes . Although our experiments do not definitively assign the role of COE in muscle differentiation or maintenance , our data do clearly indicate that coe is required for expression of nervous system-specific genes in adult planarians . Based on the annotation of differentially expressed genes , we hypothesized that genes predicted to play roles in nervous system functions in the downregulated category likely include direct COE targets . To test our hypothesis and validate genes found in our RNA-seq dataset , we selected 65 genes that were dramatically downregulated , associated with neural functions , or annotated as transcription factor homologs . First , we performed WISH to determine the tissue-specific pattern of expression of all 65 genes ( representative examples are shown in Fig . 4 ) . As we expected , the most prominent mRNA expression pattern was in the nervous system ( 26 of 65 genes; see Table S2 ) , similar to ChAT and cpp-1 , which we had previously found to be putative downstream targets of COE [24] . In addition , we observed genes that were expressed broadly in the nervous system ( such as neural cell adhesion molecule-2 ( ncam-2 ) , vesicle-associated membrane protein like-1 ( vamp ) , gamma-aminobutyric acid receptor subunit beta like-1 ( gbrb-1 ) , and voltage-gated sodium channel alpha-1 ( scna-1 ) ) or in discrete neuronal subpopulations ( such as secreted peptide prohormone-19 , -18 , -2 ( spp-19 , -18 , -2 ) , neuropeptide like ( npl ) , voltage-gated sodium channel alpha-2 ( scna-2 ) , and caveolin-1 ( cav-1 ) ) ( Fig . 4A–J ) . Our list also included transcripts that labeled subsets of neurons in the brain ( such as netrin-1 ) ( Fig . 4K ) [44] . In addition , we found that the transcription factors iroquios-1 ( irx-1 ) and pou class 4 transcription factor 4 like-1 ( pou4l-1 ) were expressed at or near the cephalic ganglia ( Fig . 4L–M ) , and their mRNA was detected in ChAT+ neurons by fluorescent in situ hybridization ( FISH ) ( Fig . S2 ) . Next , we tested the effect of coe RNAi on the expression of 33 genes that could be visualized in discrete cell populations by WISH . Knockdown of coe led to a marked reduction in the expression of 31 genes ( Table S2; representative results are shown in Fig . 4A′–H′ , K′–M′ ) ; for two genes , scna-2 and cav-1 , we observed a loss of expression at the midline ( Fig . 4I′–J′ ) . Furthermore , we quantified the number of cells labeled by spp-19 , spp-18 , and npl probes . As expected , we found there was a significant reduction in the number of spp-19+ , spp-18+ , and npl+ cells following coe RNAi ( Fig . 4N ) . As an additional test to validate the in situ hybridization results , we measured the relative expression levels of downregulated genes in control and coe RNAi-treated planarians using RT-qPCR ( Fig . S3A ) . All of the genes we tested showed a decrease in relative expression following coe RNAi ( 9 of 14 genes were significantly downregulated; P<0 . 05 , Student's t-test ) . By contrast , when we measured the relative expression of CNS-expressed genes that were not on our list of differentially expressed genes , none were significantly reduced ( 11 of 11 genes; Fig . S3B–C ) . Although some of the control genes we selected were reduced near levels comparable to some genes downregulated following coe RNAi ( e . g . , ncam2 , vamp , and gbrb1; Fig . S3A ) , we noted that isotig13897 and npp-2 [30] , which are transcripts detected in subsets of neurons or throughout the CNS , respectively , remained unchanged ( Fig . S3B–C ) . It is possible that some changes in gene expression associated with coe RNAi are consequence of a reduction in nervous system tissue . We proceeded to perform double-FISH to coe and validated genes to determine if any were potential genetic targets of COE . Of the 17 genes we were able to reliably detect by FISH ( 33 genes were tested; see Table S2 ) , 11 were expressed in coe+ cells ( representative results are shown in Fig . 5 and Fig . S4 ) , including ChAT and cpp-1 [24] . Together , these results identified nine novel candidate targets of COE in the nervous system , including genes important for maintaining neuronal subtype identity such as ion channels , ion channel receptors , and neuropeptide genes ( Table 2 ) . In addition , our data suggest that COE is essential to maintain genetic programs in multiple classes of adult neuronal subtypes including excitatory ( cholinergic ) and inhibitory ( GABAergic ) neurons . Our RNA-seq dataset revealed that coe is essential to maintain the expression of hundreds of genes in the adult animal . This change in the neuronal gene expression landscape led to abnormal CNS structure and behavior . To identify genes downstream of coe that contribute to CNS differentiation , we took advantage of the experimental ease in examination of gene function in planarian regeneration and analyzed the role of 11 downregulated genes that were expressed in neurons or predicted to encode transcription factors ( Table 3 ) . Following RNAi , animals were amputated pre- and post-pharyngeally and allowed to regenerate for 10 days . We found that 6 out of 11 genes resulted in defective brain regeneration ( see Table 3 ) ; scna-2 , pou4l-1 , and nkx2l caused the strongest phenotypes . Compared to the controls , scna-2 ( RNAi ) animals had less eye pigmentation or developed a single eyespot; nkx2l ( RNAi ) animals exhibited photoreceptor defects; and pou4l-1 ( RNAi ) animals had less photoreceptor pigment ( Fig . 6A–D ) . To examine CNS architecture , we stained scna-2 , nkx2l , and pou4l-1 RNAi treated planarians with anti-SYNAPSIN and the coe-regulated genes ChAT and npl . Although subtle , all three showed abnormalities in brain morphology ( Fig . 6A–D ) . However , when we measured the area of the brain stained by anti-SYNAPSIN , only scna-2 and pou4l-1 RNAi animals had a significant reduction in neuropil density ( Fig . 6E ) . Consistent with this observation , the ChAT+ brain areas were smaller in scna-2 ( RNAi ) and pou4l-1 ( RNAi ) animals ( Fig . 6F ) but not in nkx2l ( RNAi ) animals . The smaller brain phenotype was accompanied by fewer npl+ neurons in scna-2 ( RNAi ) animals; however , despite their smaller brains , pou4l-1 ( RNAi ) animals regenerated significantly more npl+ cells than controls ( Fig . 6G ) . These findings demonstrate that scna-2 is required for CNS regeneration and highlight the importance of ion channels in neurogenesis regulation during CNS development , maintenance , and repair [45]–[47] . Interestingly , these data suggest that pou4l-1 plays a role in the specification of certain neuronal lineages . It is possible that in the absence of pou4l-1 , planarians regenerate the incorrect proportion of neuronal subtypes and have disorganized brains , but this possibility will require further analysis with additional neuronal subtype-specific markers . By contrast , our results suggest nkx2l is not required for CNS regeneration per se . Following coe RNAi , nkx2l expression was reduced by in situ hybridization and RT-qPCR ( Table S2 and Fig . S3A ) , but nkx2l , which is primarily expressed in stem cells and in progeny [48] , was not detected in the nervous system ( Fig . S5A ) . We hypothesize nkx2l functions in early regeneration to establish patterning , which is consistent with the observation that nkx2l ( RNAi ) planarians fail to regenerate properly patterned head ( Fig . 6C ) and tail tissues ( Fig . S5B ) . It is noteworthy that several transcription factors that we identified in our screen are putative COE targets in Xenopus development , including irx-1 , tal , pou4l-1 , and nkx2l [39] . Of these genes , we found that expression of pou4l-1 was important for CNS regeneration and nkx2l was involved in patterning . NKX and POU orthologs play critical roles during CNS development of invertebrate and vertebrate organisms [49]–[51] . These data suggest that regulatory genes downstream of COE are conserved and have roles in CNS regeneration . However , it will be important to experimentally resolve whether these transcription factors are bona fide targets of COE in planarians or other animals such as Xenopus . COE proteins are known to function as terminal selectors of neuronal identity in adult organisms [14] , [15] , [52] , yet the neuronal subtypes and specific genetic programs regulated by COE in the adult CNS are not well understood . In this study , we exploited the high rate of tissue turnover and regenerative capacity of planarians to expand our understanding of how COE may function in the post-embryonic nervous system . We combined RNAi with RNA-seq analysis and identified a set of differentially expressed genes associated with nervous system biological roles . Expression analysis of a subset of these genes revealed novel candidate targets of coe in planarian neurons ( Fig . 7A ) , some of which underscored coe's essential role in maintaining expression of genes vital for neuronal subtype identity and function ( such as neurotransmitter receptors , ion channels , and neuropeptide encoding genes ) ( Fig . 7A–B ) . Decoding which transcriptional changes are direct or indirect consequences of coe loss in the planarian model will be vital to further elucidate how mutations in COE proteins cause or contribute to disease pathologies in the CNS . The next step will be to find direct COE binding sites genome-wide using in silico and chromatin immunoprecipitation ( ChIP ) approaches and combining these findings with our differential expression data . In addition , molecular profiling of coe+ cell populations ( such as stem cells , postmitotic progeny , and neurons ) will be essential to determine how coe function alters in cell type-specific contexts . In conclusion , our study demonstrates the importance of COE family proteins in neuronal turnover and repair of the adult CNS and broadens our understanding of the regulatory programs governed by these factors .
Asexual Schmidtea mediterranea ( CIW4 ) were reared in 1× Instant Ocean Salts ( 0 . 83 mM MgSO4 , 0 . 9 mM CaCl2 , 0 . 04 mM KHCO3 , 0 . 9 mM NaHCO3 , and 0 . 21 g/L Instant Ocean Aquarium Salt diluted in ultra-pure water ) at 20°C . Animals were starved for one week , and those ranging between 2–5 mm in length were used for experimentation . Animals were administered six feedings of bacterially expressed dsRNA complementary to the indicated gene over three weeks as previously described [53]; gfp dsRNA was fed as a control . Unless otherwise indicated , all intact RNAi animals were fixed seven days following the 6th dsRNA treatment . For regeneration experiments , planarians were amputated pre- and post-pharyngeally 24 hours following the 6th dsRNA feeding . Animals were processed for colorimetric whole-mount in situ hybridization using the protocol described in [54] . Fluorescent in situ hybridization experiments were performed as described in [24] , [54] and developed using Tyramide Signal Amplification ( TSA ) as described in [55] . Briefly , animals were incubated for 5 min . in borate buffer ( 100 mM borate pH 8 . 5 , 0 . 1% Tween-20 ) and then developed in TSA Reaction Buffer ( borate buffer , 2% dextran sulfate , 0 . 1% Tween-20 , 0 . 003% H2O2 ) , containing fluor-tyramide and 4-iodophenylboronic acid for 30 min . For double-FISH , animals were quenched in 1% H2O2 for 1 hour . For γ-irradiation experiments , animals were fixed 6 days following a 100 Gy treatment , a time point when both stem cells and postmitotic progenitors are ablated . Accession numbers for the sequences used in this study are listed in Table S3 . For immunostaining with anti-SYNORF1 ( 1∶400 , 3C11 , DSHB ) or anti-VC-1 ( 1∶10 , 000; kindly provided by Hidefumi Orii ) , animals were fixed with Carnoy's solution [56] . For anti-CRMP-2 ( 1∶50 , 9393S , Cell Signaling ) or anti-β-TUBULIN ( 1∶1000; E7 , DSHB ) labeling , animals were fixed with formaldehyde , processed without a reduction step , and labeled using TSA [54] . One week after the final dsRNA treatment , RNA was extracted from three independent control and coe ( RNAi ) animal groups using Trizol ( Life Technologies ) . RNA samples were treated with DNase using the Turbo DNA-free Kit ( Life Technologies ) and purified using the RNeasy MinElute Cleanup kit ( Qiagen ) . Sequencing libraries were synthesized using the TruSeq RNA Sample Prep Kit v2 and sequenced on a HiSeq 2000 System ( Illumina ) . More than 12 million 100-bp single-end reads were generated for each sample . Sequenced reads were submitted to the Sequence Read Archive ( NCBI ) under the accession number PRJNA235907 . Reads were mapped to the planarian genome using TopHat [57]; gene models were predicted using a published transcriptome [58] , [59] . Differentially expressed genes were identified using the R Bioconductor package edgeR [60] with cutoffs of logCPM score ≥0 and FDR≤0 . 05 . Changes in gene expression detected by RNA-seq were represented as linear fold changes over controls . For the differentially expressed Schmidtea mediterranea transcripts , we performed BLASTX against the human UniProt database ( cutoff<1×10−4 ) ; human accession numbers were then used to assign Gene Ontology terms and perform clustering analysis using DAVID software [61] , [62] with the “Panther_BP_all” and “Panther_MF_all” gene annotation settings and an Enrichment Score cutoff >1 . 3 . For validation studies , transcript sequences were analyzed by BLASTX against protein sequences from human , mouse , fly , and nematode and identified as the top BLAST hit ( Table S3 ) . Sequences were obtained from a cDNA collection [63] or cloned into pJC53 . 2 [30] or pPR244 [64] using gene specific primers . GenBank accession numbers and the primers used in this study are listed in Table S3 . Total RNA was extracted and purified as described above . cDNA was synthesized using the iScript cDNA Synthesis Kit ( BioRad ) . Reverse transcription quantitative PCR was performed on a Bio-Rad CFX Connect Real-Time System using SsoAdvanced SYBR Green Supermix ( Bio-Rad ) with a two-step cycling protocol and annealing/extension temperature of 58 . 5°C . At least three biological replicates and two technical replicates were performed for each experiment . The relative amount of each cDNA target was normalized to Smed-β-tubulin ( accession no . DN305397 ) . The normalized relative changes in gene expression , standard deviations , and t-tests were calculated in Bio-Rad CFX Manager Software v3 . 0 . Primers are listed in Table S3 . Images of live animals and whole mount in situ hybridization samples were acquired using a Leica DFC450 camera mounted on a Leica M205 stereomicroscope . Fluorescent images were acquired with a Zeiss Axio Observer . Z1 equipped with an Axiocam MRm camera and ApoTome; images are displayed as maximum image projections from ten 1-µm optical sections . For all experiments , we counted cells by hand using ImageJ Software [65] , and biological replicates ( n≥3 ) were averaged and shown as mean ± standard deviation . The number of cintillo+ , spp-19+ , spp-18+ , and npl+ cells ( Fig . 4N ) was normalized to animal length ( mm ) . We used anti-SYNAPSIN staining and ChAT expression to determine brain area ( Fig . 6E–F ) , normalized to animal length ( µm ) . To quantify npl+ brain-specific neurons following amputation , npl+ cells were counted in the cephalic ganglia and normalized to the average total brain area ( Fig . 6G ) . When comparing two groups , we used a Student's t-test and significance was accepted at P<0 . 05 . | COE transcription factors are conserved across widely divergent animals and are crucial for organismal development . COE genes also play roles in adult animals and have been implicated in central nervous system ( CNS ) diseases; however , the function of COE in the post-embryonic CNS remains poorly understood . Planarian regeneration provides an excellent model to study the function of transcription factors in cell differentiation and in terminally differentiated cells . In planarians , coe is expressed in differentiating and mature neurons , and its function is required for CNS regeneration . In this study , we show that coe is required to maintain structure and function of the CNS in uninjured planarians . We took advantage of this phenotype to identify genes regulated by coe by comparing global gene expression changes between control and coe mRNA-deficient planarians . This approach revealed downregulated genes downstream of coe with biological roles in CNS function . Expression analysis of downregulated genes uncovered previously unknown candidate targets of coe in the CNS . Furthermore , functional analysis of downstream targets identified coe-regulated genes required for CNS regeneration . These results demonstrate that the roles of COE in stem cell specification and neuronal function are active and indispensable during CNS renewal in adult animals . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"neurobiology",
"of",
"disease",
"and",
"regeneration",
"medicine",
"and",
"health",
"sciences",
"stem",
"cells",
"animal",
"cells",
"cell",
"biology",
"regeneration",
"neurology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
... | 2014 | COE Loss-of-Function Analysis Reveals a Genetic Program Underlying Maintenance and Regeneration of the Nervous System in Planarians |
Phylogenetic networks are necessary to represent the tree of life expanded by edges to represent events such as horizontal gene transfers , hybridizations or gene flow . Not all species follow the paradigm of vertical inheritance of their genetic material . While a great deal of research has flourished into the inference of phylogenetic trees , statistical methods to infer phylogenetic networks are still limited and under development . The main disadvantage of existing methods is a lack of scalability . Here , we present a statistical method to infer phylogenetic networks from multi-locus genetic data in a pseudolikelihood framework . Our model accounts for incomplete lineage sorting through the coalescent model , and for horizontal inheritance of genes through reticulation nodes in the network . Computation of the pseudolikelihood is fast and simple , and it avoids the burdensome calculation of the full likelihood which can be intractable with many species . Moreover , estimation at the quartet-level has the added computational benefit that it is easily parallelizable . Simulation studies comparing our method to a full likelihood approach show that our pseudolikelihood approach is much faster without compromising accuracy . We applied our method to reconstruct the evolutionary relationships among swordtails and platyfishes ( Xiphophorus: Poeciliidae ) , which is characterized by widespread hybridizations .
Evolutionary relationships are typically visualized in a tree , which implicitly assumes vertical transfer of genetic material from ancestors to descendants . However , not all species follow this paradigm . If genes can be horizontally transferred between some organisms , a tree is not a good representation of their history . Such reticulate events include hybridization , horizontal gene transfer or migration with gene flow , and require methods to infer phylogenetic networks . While a great deal of research has flourished for the inference of phylogenetic trees from different types of data , methods to infer phylogenetic networks are still limited and under development . There are mainly two kinds of phylogenetic networks: implicit and explicit . Implicit networks–also called split networks–describe the discrepancy in gene trees , or other sources of data , and methods are well developed to reconstruct these networks [1–4] . These methods tend to be fast . However , implicit networks lack biological interpretation as the internal nodes do not represent ancestral species . Explicit networks , on the other hand , represent explicit reticulation events and each node represents an ancestral species . Combinatorial methods to infer explicit networks ( which we call phylogenetic networks here ) are fast but ignore gene tree error and incomplete lineage sorting ( ILS ) as a possible source of gene tree discordance ( e . g . [5] ) . Model-based methods are most accurate but can be computationally challenging . They calculate the likelihood of an observed gene tree given a species network taking into account both reticulation and ILS [6–8] . Their scope was expanded in [9] to search for the most likely phylogenetic network based on multi-locus data ( see also [10] for a different likelihood framework , where sites instead of genes are treated as independent and ILS is ignored ) . The likelihood-based method in [9] , implemented in PhyloNet , provides a solid theoretical framework to estimate the maximum likelihood phylogenetic network from a set of gene trees . It has several advantages: it incorporates uncertainty on the gene trees estimated from sequence data , accounts for a background level of gene tree discordance due to ILS , and controls the complexity of the network with a cross validation step . However , its likelihood computation is heavy and becomes intractable when increasing the number of taxa or the number of hybridizations , making this method practical for small scenarios of up to about 10 species and 4 hybridizations in the network . Here , we provide a fast statistical method to estimate phylogenetic networks from multi-locus data . We first present the theory for the pseudolikelihood of a network . We do so by deriving the proportion of the genome that has each 4-taxon tree ( quartet concordance factors ) as expected under the coalescent model extended by hybridization events , and we prove the generic identifiability of the model . We then use the observed quartet concordance factors as inferred from the multi-locus data to estimate the species network . Our method SNaQ ( Species Networks applying Quartets ) is implemented in our open-source software package PhyloNetworks in Julia and publicly available at https://github . com/crsl4 . Like PhyloNet , our method can incorporate uncertainty in estimated gene trees and gene tree discordance due to ILS . Our pseudolikelihood has computational advantages . It is simpler and more scalable to many species , compared to the full likelihood . It also scales to a large number of loci because estimation of gene trees can be highly parallelized , then summarized by only 3 tree frequencies on each 4-taxon subsets used as input in the pseudolikelihood . In simulations , our method showed good performance and scaled to scenarios for which PhyloNet could not run . We also used SNaQ to infer the evolutionary relationships between Xiphophorus fishes , from 1 , 183 loci across 24 taxa . Our results were congruent with [11] and refined the placement of some hybridizations found in that study . The analyses here presented show that SNaQ can enable scientists to incorporate organisms to the “tree of life” in parts that are more net-like than tree-like , and thus , complete a broader picture of evolution .
Intuitively , a phylogenetic network is a phylogenetic tree with added hybrid edges , causing some nodes to have two parents ( but see [12] ) . Phylogenetic networks can describe various biological processes causing gene flow from one population to another such as hybridization , introgression , or horizontal gene transfer . Hybridization occurs when individuals from 2 genetically distinct populations interbreed , resulting in a new separate population . Introgression , or introgressive hybridization , is the integration of alleles from one population into another existing population , through hybridization and backcrossing . Genes are horizontally transferred when acquired by a population through a process other than reproduction , from a possibly distantly related population . Although these three processes are biologically different , we do not make the distinction when modeling them with a network . In other words , our model takes into account all three biological scenarios , but those scenarios are not distinguishable in the estimated phylogenetic network unless more biological information is provided . Just like phylogenetic trees , networks can be rooted or unrooted . A rooted phylogenetic network on taxon set X is a connected directed acyclic graph with vertices V = {r} ∪ VL ∪ VH ∪ VT , edges E = EH ∪ ET and a bijective leaf-labeling function f: VL → X with the following characteristics . The root r has indegree 0 and outdegree 2 . Any leaf v ∈ VL has indegree 1 and outdegree 0 . Any tree node v ∈ VT has indegree 1 and outdegree 2 . Any hybrid node v ∈ VH has indegree 2 and outdegree 1 . A tree edge e ∈ ET is an edge whose child is a tree node . A hybrid edge e ∈ EH is an edge whose child is a hybrid node . Unrooted phylogenetic networks are typically obtained by suppressing the root node and the direction of all edges . We also consider semi-directed unrooted networks , where the root node is suppressed and we ignore the direction of all tree edges , but we maintain the direction of hybrid edges , thus keeping information on which nodes are hybrids . The placement of the root is then constrained , because the direction of the two hybrid edges to a given hybrid node inform the direction of time at this node: the third edge must be a tree edge directed away from the hybrid node and leading to all the hybrid’s descendants . Therefore the root cannot be placed on any descendant of any hybrid node , although it might be placed on some hybrid edges . We further assume that the true network is of level-1[1] , i . e . any given edge can be part of at most one cycle . This means that there is no overlap between any two cycles ( but see the Discussion ) . Refer to [1] for other types of evolutionary networks . Throughout this work , we denote by For example , in Fig 1 ( center ) n = 7 , h = 2 , k1 = 3 and k2 = 4 . The main parameter of interest is the topology N of the semi-directed network . Like phylogenetic trees , this network can be rooted by a known outgroup . The other parameters of interest are t , the vector of branch lengths in coalescent units ( see below ) , and a vector of inheritance probabilities γ , describing the proportion of genes inherited by a hybrid node from one of its hybrid parent ( see Fig 1 ) . Only identifiable branch lengths are considered in t . For example , with only one sequenced individual per taxon , the lengths of external edges are not identifiable and are not estimated . The input for our method is a table of quartet CFs observed from multi-locus data ( the X values in Eq ( 1 ) ) , across many or all 4-taxon subsets from the n taxa of interest .
We carried out simulations to compare the speed and accuracy of SNaQ and PhyloNet . Given that PhyloNet uses the rooted and full gene trees , SNaQ can only be expected to perform as accurately as PhyloNet at best . Our simulations show that a pseudolikelihood approach does not compromise too much accuracy , but greatly improves speed . We simulated g gene trees with ms [30] under four different networks: ( n , h ) = ( 6 , 1 ) , ( 6 , 2 ) , ( 10 , 1 ) and ( 15 , 3 ) , with γ values set to 0 . 2 or 0 . 3 on each minor hybrid edge ( see S1 Text ) These network topologies were chosen at random by simulating a tree with n taxa under the coalescent , then choosing two edges at random for the origin and target of each hybridization and rejecting networks of level >1 . On 6 taxa all reticulations were hard to reconstruct with k = 4 , including a bad diamond I in the case h = 2 . On 10 and 15 taxa , both networks also had a diamond , of the bad type II for n = 10 . We varied the number of genes between 10 and 3000 . All analyses were run on 2 . 7–3 . 5 GHz processors . We first used the true simulated gene trees for inference . The rooted gene trees served as input for PhyloNet and the unrooted quartet CFs as observed in the g gene trees served as input for SNaQ . The semi-directed network returned by SNaQ was rooted by the outgroup species , when compatible with the estimated hybrid edges . Next , we used Seq-Gen [31] to simulate sequences of length 500 under HKY , κ = 2 , A , C , G and T frequencies of 0 . 300414 , 0 . 191363 , 0 . 196748 , 0 . 311475 and population mutation rate θ = 0 . 036 , as in [9] . Gene trees were estimated with MrBayes [28] using 106 generations sampled every 200 , 25% burnin and an HKY model . The consensus trees ( one per gene ) served as input for PhyloNet . The posterior tree samples were then used in BUCKy [26 , 27] for each 4-taxon set , to estimate quartet CFs and use them as input for SNaQ . For this pipeline , we used the tools implemened by [32] and available at https://github . com/nstenz/TICR . This procedure was replicated 30 times . The accuracy of each method was measured as the proportion of times that the estimated network matched the true network . To compare rooted networks we used the distance in [33] , which is a metric on reduced networks ( including level-1 networks ) and is implemented in PhyloNet . We used it to detect equality between rooted networks , but not to measure how “close” networks were , because this distance is very sensitive to small differences such as a change in the direction of a hybrid edge . Fig 6 summarizes the accuracy and speed of SNaQ and PhyloNet . On 10 or 15 taxa PhyloNet was too slow to run ( a single replicate with 10 taxa and 300 loci required over 400 hours ) , so we cannot provide a comparison of accuracy on these 2 larger networks . For networks with h = 2 or more , the accuracy of SNaQ decreased . So , for each semi-directed network estimated by SNaQ , we determined if its unrooted topology matched that of the true network . Fig 7 shows that in the vast majority of cases when the directed network was incorrectly estimated , its unrooted topology was still correctly inferred from true gene trees and for n = 6 with estimated gene trees . For n ≥ 10 , the inferred direction of hybrid edges degraded when gene trees were estimated . In most replicates on 10 taxa , this was because the bad diamond II near the root in the true network had a wrong estimated placement of the hybrid node . To detemine which features in the network were correctly estimated , we extracted the major tree from each network , that is , the tree obtained by keeping the major hybrid edge and suppressing the minor hybrid edge at each hybrid node . We then compared the true major tree ( from the true network ) to the estimated major tree using the Robinson-Foulds distance ( see Fig 8 ) . The major tree was correctly estimated from 300 or more genes in all scenarios , except when n = 6 , h = 2 and 300 genes ( 1 replicate out of 30 ) and 1000 genes ( 1 replicate out of 30 ) . In both cases , the true major tree was displayed in the estimated network but the major hybrid edge was estimated as a minor edge with γ < 0 . 5 . Therefore , the network’s “backbone” , i . e . the major vertical inheritance pattern , can still be estimated accurately even when the full network and hybrid edges are not ( Fig 7 ) . Among cases when the major tree was correctly estimated , we determined the detection accuracy of each true hybridization event . To do so , we compared each estimated hybridization with the true hybridization of interest . In each network ( true and estimated ) , we removed the other hybridizations by suppressing their minor hybrid edges and used the known outgroup to root both networks . We then calculated the hardwired cluster distance between the two resulting networks to determine if the estimated hybridization event matched the true hybridization of interest: connecting the same donor edge to the same recipient edge in the major tree ( Fig 9 ) . For n = 6 , the hybridizations forming a good diamond were recovered with high accuracy from 100 genes , but the hybridization forming a bad diamond I ( case h = 2 ) was very hard to recover , needing more than 1000 genes for an accurate inference of the hybrid edges’ direction . Still , the unrooted cycle was correctly estimated from 100 genes or more . For n = 10 and n = 15 taxa , the hybridization creating a cycle of k = 4 nodes was also very hard to detect with its correct direction , although its undirected cycle was accurately recovered from a few hundred genes . Hybridizations were recovered more accurately as their cycles spanned more nodes , with a high recovery rate for the hybridizations with k = 6 and k = 7 from 100 genes or more . We re-analyzed transcriptome data from [11] to reconstruct the evolutionary history of 24 swordtails and platyfishes ( Xiphophorus: Poeciliidae ) . Based on high CFs of splits in conflict with their species tree followed by a series of ABBA-BABA tests [35] , [11] concluded that hybridization or gene flow was widespread in the history of these tropical fishes . We re-analyzed their first set of 1183 transcripts . BUCKy was performed on each of the 10 , 626 4-taxon sets . The resulting quartet CFs were used in SNaQ , using hm = 0 to 5 and 10 runs each . The network with h = 0 and the major tree in the network with h = 1 were identical to the total evidence tree in [11] , with X . xiphidium placed within the grade of southern platyfishes ( SP ) , making the northern platyfishes ( NP ) paraphyletic ( see S1 Text ) . With h ≥ 2 the major tree was almost identical but with NP monophyletic ( Fig 10 ) because X . xiphidium was found sister to the rest of the NP species , but involved in a reticulation ( see below ) . With h ≥ 3 , a reticulation within the southern swordtails ( SS ) was found consistently ( γ = 0 . 43 ) , but with a direction in conflict with SS being an outgroup clade . Its cycle had only k = 5 nodes , 4 of them leading to a single taxon ( see S1 Text ) so we suspect an error in the inferred hybrid node and gene flow direction . The extra 2 reticulations found with h = 4 and 5 had low γ values ( in [0 . 006–0 . 16] ) . The network scores ( negative log-pseudolikelihood ) decreased sharply from h = 0 to h = 2 then slightly and somewhat linearly ( see S1 Text ) , suggesting that h = 2 best fits the fish data using a slope heuristic [45 , 46] . The network estimated with h = 2 ( Fig 10 ) found X . xiphidium involved in an ancient reticulation , contributing a proportion γ = 0 . 17 of genes to the lineage ancestral to northern swordtails ( NS ) . This reticulation might explain the placement of X . xiphidium closer to the root in [11] , from tree-based methods that do not account for potential gene flow . The second hybridization ( γ = 0 . 20 ) was found from the population ancestral to X . multilineatus and X . nigrensis into X . nezahuacoyotl , and relates to a high CF found by [11] for a clade uniting X . nezahuacoyotl and the nigrensis group . Bootstrap data sets were simulated by sampling each quartet CF from a uniform distribution on its 95% credibility interval ( conservatively ) then normalizing the sampled CFs across the 3 quartets on each 4-taxon set . For each bootstrap data set we estimated a network using 3 runs , and h = 3 ( instead of 2 ) because the third inferred reticulation had a high γ ( see S1 Text ) and to assess the ability of the bootstrap procedure to identify the best h value . If the bootstrap was consistent with the slope heuristic , we expected high bootstrap support for the placement of the first 2 reticulations and lower support for the third . As expected , this third reticulation and network topology within the SS clade was variable among bootstrap networks ( see S1 Text ) , suggesting uncertainty in the major tree within this clade ( Fig 10 ) . The rest of the tree was highly supported , as was the placement of the reticulation involving X . xiphidium . The reticulation involving X . nezahuacoyotl had split support for its donor lineage , with 75% support for a more ancestral lineage ( Fig 10 ) .
Network inference is theoretically and computationally challenging . Split networks can be estimated rapidly , yet lack an evolutionary model and biological interpretability . [39] proposed a very fast distance-based approach to reconstruct topological ancestral recombination graphs ( tARGs ) from a long alignment , but the biological interpretability of tARGs is still limited . The evolution model in [8] uses an explicit network and satisfyingly accounts for various processes: reticulation events , deep coalescences , and substitutions . Yet a full likelihood estimation of large network ( as in [9] ) seems beyond computational reach . Our pseudolikelihood method offers an alternative , allowing the estimation of bigger and more complex networks while maintaining biological interpretability and a flexible evolutionary model . We assumed a level-1 network throughout , where each hybrid node is part of a single cycle . This assumption is quite restrictive , but [40] showed that sequence data and gene trees on present-day species do not contain enough information to reconstruct complex networks , even from many loci . Therefore , some assumption has to be made to limit the network complexity . Extending our method to networks with intersecting cycles will need further work to restrict the search to candidate networks that are distinguishable from each other . Indeed , [40] show that different level-2 networks can have the exact same likelihood , and hence pseudolikelihood . So no method based on gene trees can ever decide which of these level-2 networks is true . Under a model without ILS , using full gene trees and branch length in substitutions per sites comparable across genes , [40] showed that level-1 networks are distinguishable but level-2 networks are not necessarily . Extending our approach to higher level networks , with or without ILS , will require extensive theory to work around this lack of identifiability . Our approach allows for multiple individuals per species . All alleles from the same species simply need to be treated as a known and fixed polytomy in the network . Future work could include this and other topology constraints on the network , to reduce the computational burden when there are known phylogenetic relationships . We allow hybrid edge lengths to be 0 , but we do not constrain them to be 0 ( unlike in [6 , 8] ) even though each gene flow event has to occur between contemporary populations . If one parental population went extinct or has no sampled descendants , the hybrid edge from this parent has a positive length in the observable network . A second reason is that a long branch can fit a population bottleneck , as might be expected in the formation of a new hybrid species . Not constraining hybrid branch lengths to 0 has a computational burden , however . Future implementations might enforce this constraint , when taxon sampling is thorough and extinction of parental populations can be ruled out . By considering quartet topologies only , we ignored branch lengths in gene trees . This choice frees us from various assumptions . Using gene tree branch lengths , which are in substitutions per site , would require some assumption on gene rates to make branch lengths comparable across trees , and a molecular clock on gene trees . Other assumptions would also be needed on population sizes , shared or not across lineages . The recent approach in [41] should scale well to many taxa , but makes these strong assumptions because it requires accurate distances obtained from branch lengths in gene trees . On the contrary , our approach should be robust to rate variation across genes and across lineages , and does not require any assumption on population sizes . Yu et al . [8] already noted a lack of identifiability from rooted gene trees for reticulations with k = 3 from only 4 taxa ( including the outgroup ) . We found a similar lack of identifiability from unrooted quartets if n < 5 . In practice , some reticulations are hard to detect even with 5 or more taxa , if some branches are long with no ILS ( close to violating A1 ) . However , in these cases the unrooted topology of the network can still be recovered , even if the direction of gene flow and the placement of the hybrid node is not . Therefore , heuristic strategies that keep the unrooted network unchanged , or that just slightly modify it , may improve the search for the best network . More tools are needed to study unrooted and semi-directed phylogenetic networks . For instance , no distance measure has been developed for such networks , that we know of . Distances between rooted networks would also be needed , that would be less sensitive to small changes in the unrooted or semi-directed topologies than the distance proposed in [33] . New notions of edge equivalence would also be needed on unrooted and semi-directed networks . It would help summarize a bootstrap sample of networks for instance , with no need for an outgroup . We propose here a tree-based but informative summary by extracting the major tree from each network , obtained by dropping any minor hybrid edge ( with inheritance γ < 0 . 5 ) . Because this tree summarizes the major vertical inheritance pattern at each node , it can be considered an estimate of the species tree . We found that recovering the underlying species tree can be much easier ( requiring fewer genes ) than recovering the horizontal signal . Even if the species tree is the main purpose of a study , [34] showed that species-tree methods can be inconsistent in recovering the vertical signal if there is gene flow , so using a network can be beneficial to avoid the possible inconsistency of tree-based coalescent methods . All data analyzed here had full taxon sampling from each gene , and we were able to use all 4-taxon sets . Future work could assess the impact of missing data ( gene sequences , or 4-taxon sets ) on the method’s accuracy . Missing 4-taxon sets will be necessary for large networks , because the number of 4-taxon sets grows very rapidly with the number of taxa ( ∼n4/24 ) . With many taxa , one may randomly select a collection of 4-taxon sets and/or choose them specifically . SNaQ calculates the number of quartets involving each taxon and provides information about under-represented taxa , if any . With many individuals per species , one may greatly reduce the collection of 4-taxon sets to be analyzed by randomly sampling from those containing at most one individual per species . If the assignment of individuals to species is correct , any 4-taxon set containing 2 individuals from the same species would be non-informative about the species-level relationships . This strategy is used in [42] to infer species trees under ILS . Model selection is necessary to estimate the number of hybridizations h , because the pseudolikelihood is bound to improve as h increases , like the likelihood or parsimony score in [43] . We used here the log pseudolikelihood profile with h . A sharp improvement is expected until h reaches the best value and a slower , linear improvement thereafter . Such data-driven slope heuristics can indeed be used with contrast functions ( like pseudolikelihoods ) for model selection in regression frameworks [45 , 46] . Information criteria have already been used to select h ( e . g . [44] ) , but these criteria are inappropriate if the full likelihood is replaced by a pseudolikelihood . Theory is missing to compare the pseudolikelihoods of different networks , because of the possible correlation between quartets from different 4-taxon sets . It can be shown , however , that quartets from two 4-taxon sets s1 and s2 are independent if s1 and s2 overlap by at most one taxon and if the true 4-taxon subnetworks share no internal edges . Future work could exploit this partial independence to construct hypothesis tests . Cross validation has been proposed by [9] , and was shown to have good performance . In our framework , the cross-valication error could be measured from the difference between the quartet CFs observed in the validation subset and the quartet CFs expected from the network estimated on the training set . Because K-fold cross-validation requires partitioning the loci into K subsets and re-estimating a network K times at each h value , this approach can be computationally heavy . Finally , [32] proposed a goodness-of-fit test , also based on quartet CFs , to determine if a tree with ILS fits the observed data or if a network is needed instead . This test could be extended to networks , to decide if a given h provides an adequate fit . One advantage to this approach is that testing the adequacy of a given h does not require to estimate a larger network with h + 1 hybridizations , whereas other approaches above would require estimation of both networks in order to decide that the simpler network is sufficient . After submission , we learned about similar work using subnetworks and a pseudolikelihood approach [47] , which scales to many taxa . In [47] , the pseudolikelihood is based on rooted triples whereas we use unrooted quartets . There are fewer triples , so the method in [47] is potentially faster . However , fewer triples means less information . For example , the networks Ψ1 and Ψ2 shown in Fig . 2 of [47] , which are not distinguishable from triplets , are in fact distinguishable from quartets ( see S1 Text ) . Our thorough study of the network identifiability allowed us to implement a search that avoids jumping between networks that are not distinguishable , which facilitates convergence . The downside of our approach is the assumption of a level-1 network . Instead , [47] do not assume any restriction on the network . Finally , our method does not require rooted gene trees as input , which we view as a major advantage because rooting errors are avoided . | Phylogenetic networks display the evolutionary history of groups of individuals ( species or populations ) including reticulation events such as hybridization , horizontal gene transfer or migration . Here , we present a likelihood method to learn networks from molecular sequences at multiple genes . Our model accounts for several biological processes: mutations , incomplete lineage sorting of alleles in ancestral populations , and reticulations in the network . The likelihood is decomposed into 4-taxon subsets to make the analyses scale to many species and many genes . Our work makes it possible to learn large phylogenetic networks from large data sets , with a statistical approach and a biologically relevant model . | [
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
] | [
"taxonomy",
"genetic",
"networks",
"population",
"genetics",
"vertebrates",
"animals",
"animal",
"phylogenetics",
"animal",
"models",
"osteichthyes",
"phylogenetics",
"data",
"management",
"model",
"organisms",
"phylogenetic",
"analysis",
"network",
"analysis",
"genome",
... | 2016 | Inferring Phylogenetic Networks with Maximum Pseudolikelihood under Incomplete Lineage Sorting |
Human mitochondrial DNA ( mtDNA ) replication is first initiated at the origin of H-strand replication . The initiation depends on RNA primers generated by transcription from an upstream promoter ( LSP ) . Here we reconstitute this process in vitro using purified transcription and replication factors . The majority of all transcription events from LSP are prematurely terminated after ~120 nucleotides , forming stable R-loops . These nascent R-loops cannot directly prime mtDNA synthesis , but must first be processed by RNase H1 to generate 3′-ends that can be used by DNA polymerase γ to initiate DNA synthesis . Our findings are consistent with recent studies of a knockout mouse model , which demonstrated that RNase H1 is required for R-loop processing and mtDNA maintenance in vivo . Both R-loop formation and DNA replication initiation are stimulated by the mitochondrial single-stranded DNA binding protein . In an RNase H1 deficient patient cell line , the precise initiation of mtDNA replication is lost and DNA synthesis is initiated from multiple sites throughout the mitochondrial control region . In combination with previously published in vivo data , the findings presented here suggest a model , in which R-loop processing by RNase H1 directs origin-specific initiation of DNA replication in human mitochondria .
Mitochondrial DNA ( mtDNA ) is a 16 . 6 kb circular , double-stranded DNA ( dsDNA ) molecule that contains genes for 13 components of the oxidative phosphorylation ( OXPHOS ) system as well as the 22 tRNAs and 2 rRNAs required for their translation . The two strands can be separated by CsCl gradient density centrifugation and are accordingly referred to as the heavy ( H ) and light ( L ) strands . Polycistronic transcription of the two strands is initiated from the L-strand and H-strand promoters ( LSP and HSP , respectively ) and carried out by a transcription machinery consisting of a single subunit RNA polymerase ( POLRMT ) , and transcription factors A ( TFAM ) and B2 ( TFB2M ) [1 , 2] . Transcription is further stimulated by the mitochondrial transcription elongation factor ( TEFM ) , which forms a stable ternary complex with the elongating POLRMT and template DNA [3–6] . Mitochondrial DNA is synthesized by the trimeric DNA polymerase γ ( POLγ ) that consists of one catalytic subunit ( POLγA ) and two identical accessory subunits ( POLγB ) , which act to increase processivity [7–10] . The replicative DNA helicase TWINKLE is a hexameric protein and its activity is stimulated by the mitochondrial single-stranded DNA binding protein ( mtSSB ) [11 , 12] . According to the strand displacement model for mtDNA replication , mtDNA synthesis is initiated from two separate origins of replication , OriH and OriL , one for each strand . DNA synthesis first commences at OriH and proceeds in one direction to produce a nascent H-strand [1 , 13] . When the replication machinery has synthesized approximately two thirds of the H-strand , it passes OriL , which is exposed in its single-stranded conformation and activated . OriL adopts a stem-loop structure and POLRMT initiates synthesis of short RNA primers from a poly-dT stretch in the loop region . These primers are used by POLγ to initiate L-strand synthesis with the parental H-strand as template [14 , 15] . H-strand and L-strand DNA replication subsequently proceed until two complete daughter molecules have been formed and separated in a Topoisomerase 3α-dependent process [16] . In the literature , there is some confusion with regard to the exact localization of OriH [1] . It is often defined as a single position , 191 in human mtDNA , since prominent free 5′-ends on DNA have been identified at this site [17] . It should however be noted that RNA to DNA transitions have never been identified at this position . How these 5′-ends are actually formed is not understood , but they are potentially the result of extensive primer processing far beyond the location of the actual RNA to DNA transition sites . The processing involves MGME1 and disease-causing mutations affecting this enzyme lead to accumulation of replication intermediates with incomplete processing of 5′-ends [18] . According to the current model for replication initiation , transcription initiated at LSP provides RNA primers for initiation at OriH [19–21] . In support of this notion , loss of POLRMT in a knockout mouse model abolishes primer synthesis in vivo [22] . The switch between primer formation and full-length transcription takes place in a region immediately downstream of LSP containing three evolutionary conserved sequence blocks , CSBI-III , and RNA to DNA transitions in newly synthesized H-strands have been mapped to multiple sites surrounding these elements [19–21] . Newly transcribed RNA remains associated with the CSB-region , forming R-loops that are resistant to RNase A and RNase T1 treatment [23–25] . The unique stability of these R-loops is explained by a G-quadruplex structure , which is formed co-transcriptionally between nascent RNA and the non-template DNA strand at CSBII . This G-quadruplex structure also stimulates premature transcription termination downstream of CSBII [26–28] . The 3′-ends of the terminated transcripts roughly overlap with RNA to DNA transitions mapped in the CSBII-region [26 , 29] , which has led to the hypothesis that sequence-dependent transcription termination may be responsible for primer formation . There is however no published experimental evidence demonstrating that transcripts prematurely terminated at CSBII can be directly used to prime DNA synthesis . In addition , a molecular understanding of the primer formation and replication initiation reaction is still missing . Recently , a knockout mouse model demonstrated that RNase H1 is required for R-loop degradation in vivo . In addition , depletion of RNase H1 caused a reduction in mtDNA levels , suggesting that the enzyme is required for mtDNA replication [30] . These in vivo findings raise the possibility that RNase H1 may play a role in R-loop processing and primer formation . RNase H1 is an RNase H enzyme capable of cleaving RNA-DNA hybrids . It can cleave hybrids that are down to approximately 6 nucleotides in length [31] . The enzyme can also cleave Okazaki fragment-like structures , leaving approximately two ribonucleotides next to the RNA-DNA junction [32] . In addition to the potential role in R-loop processing , RNase H1 has also been proposed to be involved in mitochondrial pre-rRNA processing by interacting with the mitochondrial protein P32 , which slightly enhances the RNase H1 enzymatic activity [33] . Initiation of mitochondrial DNA replication has been suggested to resemble replication of the E . coli plasmid ColE1 . In ColE1 replication , a transcript denoted RNAII associates with the template strand , forming an R-loop that serves as a primer for DNA synthesis [34 , 35] . The ColE1 origin of replication is situated downstream of a guanine-rich stretch that is essential for both replication initiation and R-loop formation [35 , 36] . For proper initiation of replication , RNAII has to be cleaved by E . coli RNase H , thereby creating a primer 3′-end that can be used by the replication machinery . The similarity between OriH of mtDNA and the ColE1 origin [27 , 28] , together with the recent in vivo findings of RNase H1 involvement in mtDNA synthesis , suggests that RNase H1 could be involved in primer processing in human mitochondria . To address this intriguing possibility , we here set out to reconstitute initiation of mtDNA replication in vitro .
In vivo analysis has suggested that RNase H1 is required to process R-loops in vivo [30] . We decided to investigate this process in vitro and therefore set out to reconstitute R-loop formation . For our work , we used purified human mitochondrial transcription proteins ( Fig 1A , lanes 1-4 ) . As templates , we employed supercoiled or relaxed plasmids containing the LSP promoter and the downstream CSBI-III-region ( pUC-LSP , S1 Table ) . On a relaxed template , a fraction of all transcription events initiated at LSP was prematurely terminated at CSBII ( Fig 1B , lane 1 ) [26 , 29] . This effect was stronger on a negatively supercoiled template , where nearly all transcription events were prematurely terminated at CSBII ( Fig 1B , compare lanes 1 and 5 ) . In agreement with previous reports , premature transcription termination was reduced by addition of the transcription elongation factor TEFM ( Fig 1B , lane 7 ) [4 , 5] . To detect if the pre-terminated transcripts remained as stable R-loops , we used an RNase A protection assay ( Fig 1C ) . RNase A digests free RNA , whereas RNA associated with DNA is resistant to this nuclease at high salt concentrations . No R-loops were formed with relaxed LSP template ( Fig 1B , lanes 2 and 4 ) . In the reaction with the supercoiled template however , we observed RNase A-resistant products at sizes just below the transcript terminated at CSBII ( Fig 1B , lane 6 ) . This observation suggested to us that long R-loops were formed , possibly encompassing nearly the entire region from LSP to CSBII . Interestingly , R-loops were not detected when TEFM was added to the transcription reaction ( Fig 1B , compare lanes 6 and 8 ) . In other systems , single-stranded DNA binding proteins can promote R-loop formation , probably by binding to the displaced DNA strand [37–39] . We therefore monitored the effects of mtSSB ( Fig 1A , lane 8 ) on R-loop formation in vitro . Addition of mtSSB did not affect the overall transcription patterns on a negatively supercoiled LSP template ( Fig 1D , compare lane 1 to lanes 3 and 5 ) , but the R-loops formed in the presence of mtSSB appeared less processed and more uniform in size ( Fig 1D , compare lane 2 to lanes 4 and 6 ) . When 120 nM mtSSB was added nearly 70% of all pre-terminated transcripts end up as R-loops ( Fig 1D , lanes 5-6 ) , as compared to 50% in the absence of mtSSB ( Fig 1D , lanes 1-2 ) . To ensure that our results were not due to mtSSB binding directly to single-stranded RNA and protecting it from RNase A degradation , we also monitored the effects of mtSSB on transcription from a linearized HSP plasmid ( pUC-HSP , S1 Table ) ; a template that does not contain a downstream R-loop-forming region . The presence of mtSSB did not protect transcripts formed by HSP transcription from RNase A degradation ( Fig 1D , lanes 7-10 ) . Finally , we investigated if RNase H1 ( Fig 1A , lane 9 ) could process the R-loops . We found that RNase H1 gradually degraded R-loops in a concentration-dependent manner , generating shorter RNA species . ( Fig 1E , compare lane 1 with lanes 2-7 ) . The shorter RNA species may be involved in hybrid G-quadruplex formation , and therefore partially resistant to RNase H1 degradation [27] . The observed degradation of R-loops in vitro was in agreement with in vivo data indicating that RNase H1 can process mitochondrial R-loops [30] . It has been suggested that transcripts prematurely terminated at CSBII may be used to prime initiation of mtDNA replication [26] . We therefore decided to analyze if POLγ ( Fig 1A lanes 5-7 ) can use the R-loops formed in vitro as primers for initiation of DNA synthesis . We first investigated if POLγ can use random RNA primers to initiate DNA synthesis on dsDNA . To this end , we utilized POLRMT’s ability to initiate transcription and produce short RNA molecules on negatively supercoiled dsDNA even in the absence of a promoter sequence [40] . In our experiments , we used an exonuclease deficient version of POLγ ( exo- ) , since this protein has strand displacement activity and can use dsDNA as a template even in the absence of a DNA helicase [41] . To monitor DNA synthesis we performed the reactions in the presence of [32P]dTTP ( Fig 2A ) . When both POLRMT and POLγ were added simultaneously to a negatively supercoiled dsDNA template ( without LSP ) , we observed formation of [32P]dTTP-labeled DNA products ranging in size from 50 to 700 nts ( Fig 2B , lane 3 ) . We also did the experiment with a template containing LSP but did not find any major effect of its presence ( Fig 2B , lane 6 ) . When we repeated the experiment in the presence of increasing amounts of purified mtSSB , replication products disappeared ( Fig 2C , lanes 2-4 and 6-8 ) . POLγ can thus initiate DNA synthesis from random primers generated by POLRMT and mtSSB represses this activity . The repressive effect of mtSSB is probably due to its ability to prevent transcription initiation by POLRMT on ssDNA [15 , 42] . We next monitored the effects of RNase H1 on replication . Using the LSP template , in the absence of mtSSB , we again observed abundant replication products ranging in size from 50 to 700 nts and we observed a mild stimulatory effect at high levels of RNase H1 ( Fig 2D , compare lanes 1 and 4 ) . Again , these non-specific products were reduced in the presence of low levels of mtSSB ( 20 nM , Fig 2D , lane 5 ) and abolished at high concentrations of mtSSB ( 120 nM , Fig 2D , lane 9 ) . Interestingly , addition of increasing levels of RNase H1 in the presence of mtSSB caused the formation of new group of DNA products , ranging in size between 25 and 100 nts ( Fig 2D , lanes 11-12 ) . We hypothesized that these products could be DNA replication events initiated from processed R-loops formed by transcription from LSP . To address this possibility , we repeated the experiments with a template lacking LSP . On this template , RNase H1 had no apparent effects on DNA synthesis and no short DNA products were observed when RNase H1 was added together with mtSSB ( Fig 2E , lanes 1-4 ) . To further demonstrate that priming was dependent on R-loops , we performed our experiments with the LSP-containing template in the presence of TEFM , which at high concentrations prevents R-loop formation . Increasing amounts of TEFM reduced the products in the size range between 25 and 100 nts , thus suggesting that these shorter replication products were indeed dependent on R-loop formation downstream of LSP ( Fig 2F , lanes 1-4 ) . In the presence of higher concentrations of TEFM , a high molecular weight species started to accumulate ( Fig 2F , lane 4 ) , which may be due to some longer RNA molecules being used as non-specific primers for replication . Interestingly , even if TEFM reduces replication initiation , it does not completely abolish the reaction . Shorter replication products were observed even when TEFM was present in excess ( Fig 2F , lane 4 ) . To further verify that the 25-100 nts replication products we had observed indeed originated from an LSP R-loop primer , we utilized the fact that the LSP R-loop region only contains a few guanines in the template strand . By adding ddCTP to our reactions we could therefore generate short specific length replication products . Before gel analysis , the replication products were treated with potassium hydroxide ( KOH ) to remove any RNA residues [43 , 44] ( Fig 3A ) . As expected , we could not observe any DNA synthesis in the absence of RNase H1 ( Fig 3B , lane 1 ) , again demonstrating that the unprocessed R-loops do not function as primers . Upon addition of RNase H1 , we observed replication products with sizes between 12 and 21 nts ( Fig 3B , lanes 2-8 ) . The highest levels of DNA synthesis were observed at 2 nM of RNase H1 ( Fig 3B , lane 5 ) . Furthermore , we found that mtSSB had a strong stimulatory effect on the reactions ( Fig 3C , compare lane 1 to lanes 2-7 ) . By introducing mutations that changed the length of C-less stretches downstream of LSP , we could verify that replication products observed after incubation with ddCTP was due to initiation near CSBII and CSBIII ( S1A–S1C Fig ) . Combined , our experiments therefore support the hypothesis that RNase H1 is required to process R-loops for primer formation and that mtSSB stimulates the process . For all subsequent experiments , we used an mtSSB concentration of 40 nM together with 2 nM of RNase H1 ( Fig 3C , lane 6 ) . We next examined the effects of differing template topology and CSB region mutations on replication initiation . Relaxation of the DNA template , which impairs R-loop formation , almost completely abolished replication initiation ( Fig 3D , compare lanes 1 and 2 ) . When CSBIII was mutated , the levels of most replication products shorter than 20 nts were slightly decreased , whereas longer products were left unaffected ( Fig 3D , compare lanes 1 and 3 ) . When CSBII was mutated all replication initiation events disappeared ( Fig 3D , compare lanes 1 and 4 ) . These results confirm that the LSP R-loops are essential for replication initiation and demonstrate that whereas the CSBII element is crucial for successful primer formation , CSBIII only has a minor effect on initiation ( Fig 3D , compare lanes 3 and 4 ) . Next , we examined the effects of TEFM on initiation of DNA synthesis ( Fig 3E ) . At lower concentrations , TEFM did not affect replication initiation . Interestingly , robust levels of initiation were observed even at equimolar concentrations of POLRMT and TEFM ( Fig 3E , lane 4 ) . Not even at very high TEFM concentrations ( molar ratio 4:1 relative POLRMT ) could we observe a complete inhibition of replication initiation ( Fig 3E , lane 6 ) . Our experiments so far had been performed with exonuclease deficient POLγ . We now repeated the experiments with WT POLγ ( Fig 1A , lane 5 ) . As previously demonstrated , initiation of DNA synthesis required the presence of POLγ , POLRMT , and RNase H1 ( Fig 3F ) . Furthermore , the reaction was stimulated by the addition of mtSSB ( Fig 3F , compare lanes 4 and 5 ) . The POLγ WT enzyme produced the same products as seen with POLγ exo- ( Fig 3F , compare lanes 5 and 6 ) , although the levels of replication products were lower , probably due to the weaker strand displacement activity of the WT enzyme . Based on our in vitro observations , we wanted to specifically analyze the effects of reduced RNase H1 activity on mtDNA replication initiation in mammalian cells . To this end , we decided to study the effects of disease causing mutations in RNASEH1 , which are associated with adult-onset mitochondrial encephalomyopathy [45] . First , we expressed two disease-causing mutant forms of RNase H1 , RNase H1:V142I and RNase H1:A185V ( Fig 4A ) as recombinant proteins and analyzed their effects on R-loop processing . Both RNase H1:V142I and RNase H1:A185V , alone or in combination ( Fig 4B , lanes 5-13 ) , displayed impaired R-loop processing activity compared to WT RNase H1 ( Fig 4B , lanes 2-4 ) . As a consequence , the mutant RNase H1 proteins could not support origin-specific initiation of DNA replication in vitro ( Fig 4C and S2 Fig ) . Second , we used fibroblasts isolated from a patient with two heterozygous RNASEH1 mutations , the V142I mutation ( see Fig 4A ) , and a second mutation , ( R157* ) , causing a truncated form of the enzyme lacking the active site . In our experiments , we used primer extension with DNA isolated from these RNase H1-deficient patient cells and a wild-type ( WT ) control to map initiation sites for mtDNA synthesis in vivo . We designed one primer to detect 5′-ends close to OriH and the CSB-region ( primer complementary to H-strand positions 8-29 , Fig 5A , Primer 1 ) and another to detect 5′-ends further downstream , in the D-loop region ( primer complementary to H-strand positions 16 , 231-16 , 251 , Fig 5A , Primer 2 ) . The isolated DNA was analyzed before and after treatment with E . coli RNase H2 , which will specifically degrade the RNA part of the covalently linked RNA-DNA molecule and thereby allow for exact mapping of RNA to DNA transitions . In control cells , 5′-ends were detected at mtDNA positions 191-194 , 171-176 , 148-153 and 110-113 ( Fig 5B , lane 1 ) as well as around mtDNA position 60 ( Fig 5C , lane 1 ) . The observed 5′-ends were not altered by RNase H2 treatment , demonstrating that they were fully processed i . e . had no RNA residues attached ( Fig 5B and 5C , compare lanes 1 and 2 ) . Using mtDNA isolated from RNase H1 deficient cells , we observed a strong increase in 5′-ends . Some 5′-ends were similar to , but sometimes more abundant , than the products observed in WT cells at positions 191-194 , 148-153 , 171-176 and around position 60 ( Fig 5B and 5C , lane 3 ) . We also noted a range of new 5′-ends not present in WT cells at positions ~305-315 , ~240 , 209-217 and 119 ( Fig 5B , lane 3 ) and throughout the D-loop region ( Fig 5C , lane 3 ) . Interestingly , at least three of these new 5′-ends were altered upon RNase H2 treatment suggesting that they represented new RNA to DNA transition sites not present in the WT control . These new sites were located in the CSB region at positions 305-315 and 209-217 ( Fig 5B , compare lanes 3 and 4 ) , and in the D-loop , at positions 16 , 371-16 , 383 ( Fig 5C , compare lanes 3 and 4 ) . Our data thus suggested that loss of RNase H1 activity caused initiation of mtDNA synthesis from multiple sites not used in WT cells . To further support our findings , we also used an alternative method to identify 5′-ends and RNA to DNA transitions . We performed 5′-End-seq and HydEn-seq on the mtDNA from control and patient cells deficient in RNase H1 . The 5′-End-seq method will detect all 5′-ends of DNA , including DNA molecules with ribonucleotides on the 5′-end . In contrast , HydEn-Seq will only detect DNA 5′-ends , since any RNA residues will be chemically removed [43] . By comparing results from 5′-End-seq and HydEn-seq , it is thus possible to identify RNA to DNA transition sites . The major DNA 5′-end in the control cells was mapped to position 111 ( Fig 5D ) . There were no major differences between the 5′-End-seq and HydEn-seq samples indicating that no RNA-DNA transitions are found in this region in control cells ( Fig 5D and 5E ) . We observed more reads in the control region of RNase H1-deficient patient cells ( Fig 5F ) compared to control cells , in agreement with the increased 5′-ends found by primer extension . Interestingly , 5′-ends of DNA were clearly shifted when compared to the control cells . New large peaks appeared around positions 235 and 56 and a region spanning 16 , 569/0 to 16 , 300 ( Fig 5F ) . When comparing the 5′-End-seq to the HydEn-seq data for the peaks in the 16 , 569/0 to 16 , 300 region , it was clear that all the peaks decreased in the HydEn-seq data , with the exception of a peak at 16 , 374 , which was slightly increased ( Fig 5G ) . One region with new peaks also appeared in the HydEn-seq data , around position 200 . In conclusion , the sequencing results agreed with the findings obtained by primer extension . The data from these two methods revealed an overall increase in 5′-ends in the control region of RNase H1-deficient patient cells as well as new RNA to DNA transition points . In conclusion , loss of RNase H1 activity leads to initiation of mtDNA synthesis at multiple sites not used in WT cells .
It has long been recognized that mammalian mtDNA replication is initiated at OriH [19–21] . Studies in mitochondrial extracts have also demonstrated that R-loops are formed in the region downstream of LSP and linked these structures to priming of mtDNA replication in the OriH-region [25] . The precise mechanisms of the proposed model have remained obscure , in part because of technical limitations . For instance , mitochondria cannot be transfected , a shortcoming that has prevented a detailed structural characterization of OriH . We here use an alternative approach to study OriH function , employing purified , recombinant proteins and defined DNA templates . Our work builds on recent in vivo observations , which demonstrated that RNase H1 is required for R-loop processing and mtDNA replication in a mouse knockout model system [30] . Early reports suggested that the LSP transcript involved in R-loop formation must be cleaved to form the 3′-OH termini required for initiation of replication at OriH . RNase MRP was identified as a candidate for this process [46–49] , but the idea was later abandoned , since experimental evidence argued against the existence of RNase MRP in mitochondria [50–52] . As demonstrated here , the R-loops formed by LSP-transcription are instead processed by RNase H1 and the 3′-ends formed are used to prime mtDNA synthesis by POLγ . The process is stimulated by mtSSB , which acts to stabilize R-loops , most likely by binding to the displaced DNA strand . In addition , mtSSB prevents non-promoter-specific initiation of transcription from single-stranded stretches of DNA [15] , thereby restricting replication initiation to the OriH-region . Our findings receive support from previous work , which has demonstrated that Rnaseh1 deletion in mice leads to mtDNA depletion [53] and that disease-causing mutations in RNASEH1 impair mtDNA replication [45 , 53 , 54] . In our work , we found that DNA replication in vitro was mainly initiated from the CSBII and CSBIII regions . The observed RNA to DNA transitions mapped in our recombinant system therefore correlate with the RNA to DNA transitions previously mapped in mtDNA by David Clayton and colleagues [19 , 20 , 47 , 48] . In later papers , primer extension and ligation mediated PCR was used to map RNA to DNA transition points and only one of the mapped regions , that downstream of CSBII , was identified as a replication initiation site in vivo [26 , 29] . What was not known at the time of these later studies was that the CSBII sequence forms strong G-quadruplex structures in both DNA and RNA [27 , 28] . Many DNA polymerases have problems bypassing such G-quadruples and primer extension can in fact be used to detect these non-B-form DNA structures [55] . The way that these experiments were devised may therefore have correctly mapped replication initiation sites located downstream of CSBII , but failed to reach 5′-ends on nascent mtDNA located upstream of the G-quadruplex-forming CSBII region , thereby missing initiation events near CSBIII . Interestingly , mutations that reduce RNase H1 activity in vivo lead to an increase of 7S DNA levels [45] . We believe that this effect is due to unregulated initiation of 7S DNA synthesis . In support of this notion , our primer extension and 5′-end sequencing results show an overall increase in free 5′-ends and identified multiple new RNA to DNA transition sites . Apparently , RNase H1 is required to restrict initiation of 7S DNA and mtDNA synthesis to OriH . The enzyme removes all RNA hybridized to the DNA template , with the exception of the RNA-loop molecules in the CSB region , which after RNase H1 processing can be used as primers . In the absence of RNase H1 , 7S DNA synthesis is instead initiated from multiple locations , possibly from any RNA molecule with a 3′-end hybridized to the DNA template , leading to unregulated initiation of 7S DNA synthesis and an increase in overall 7S DNA levels . In contrast , the mtDNA levels remain largely unchanged in RNASEH1 mutant cells . This observation is in agreement with the notion that regulation of mtDNA levels is not correlated to 7S DNA levels , but takes place at the end of the D-loop [56] . At this place , there is a regulated switch between abortive ( 7S DNA ) and genome length mtDNA replication , which explains why 7S DNA can be strongly up regulated , whereas mtDNA levels remain unaffected . The overall problem in the case of RNase H1 deficiency could instead be related to downstream events , such as DNA replication termination and mtDNA segregation , as the newly synthesized mtDNA will have shifted 5′-ends and thus also replication end points . TEFM prevents premature transcription termination at CSBII , which led to the suggestion that this protein can function as a regulator of primer formation , controlling the switch between transcription and mtDNA replication [4 , 5] . In support of this claim , addition of TEFM reduces R-loop formation in vitro . The effects of TEFM on replication initiation are less dramatic , as robust initiation is seen even when POLRMT and TEFM are present at equimolar levels . This observation does not necessarily argue against the idea that TEFM acts to regulate the switch between transcription and replication . It is still possible that variations in TEFM concentrations will affect the relative levels of primer formation and transcription elongation in vivo . However , TEFM does not cause an all or nothing effect , since the protein may be present at relatively high concentrations without abolishing replication initiation . Also arguing against an essential role for TEFM in regulating primer synthesis , depletion of this factor in cell lines does not lead to changed levels of mtDNA [3] . Clearly , more work is required to elucidate the precise role of TEFM in the regulation of mtDNA replication initiation . RNase H1 processing of R-loops may be influenced by a multitude of factors . The G-quadruplex structure formed between RNA and the non-template DNA strand at CSBII can reduce RNase H1 cleavage [27] . RNase H1 activity can also be influenced by other proteins . There could be e . g . direct protein interactions between RNase H1 and the transcription or replication proteins . In addition , RNase H1 has been shown to interact with the mitochondrial protein P32 and that this interaction significantly enhances the cleavage activity of RNase H1 on heteroduplex templates . If P32 also affects primer formation is not known [33] . We will address these intriguing possibilities in future studies . In conclusion , we here present a mechanism for primer formation and initiation of DNA replication in human mitochondria ( as shown in Fig 6 ) . As previously suggested [20 , 23 , 25] , there are indeed striking similarities between replication initiation in mammalian mitochondria and the process described for the ColE1 plasmid . Both systems depend on promoter driven transcription , R-loop formation and primer processing . Important aspects of our model are supported by in vivo evidence , but more work is clearly needed to validate our ideas and refine the proposed mechanisms . Of special importance will be to clarify how the switch between transcription and primer synthesis is regulated .
RNase H1 was expressed using the baculovirus system ( Sf9 cells ) . The protein coding sequence was PCR-amplified from human cDNA and cloned into the pBacPAK9 vector ( Clontech ) . The construct lacked the N-terminal MTS ( amino acids 1-26 ) , and carried a C-terminal 6×His-tag . Recombination and cell infection was performed as described in the BacPAK manual ( Clontech ) . The protein was expressed and purified as previously described for transcription proteins [5] . All other recombinant proteins were expressed and purified as described previously [5 , 41] . All templates used were either empty pUC19 plasmids or mitochondrial DNA sequences with or without modifications cloned in pUC18 . A list of all templates used in this study can be found in S1 Table . All supercoiled templates were carefully prepared with QIAprep Spin Miniprep Kit ( QIAGEN ) and kept at 4°C . To produce relaxed templates , supercoiled plasmids were treated with Topoisomerase I ( New England Biolabs ) . All transcription reaction volumes were 25 μL and contained 25 mM Tris-HCl pH 8 . 0 , 10 mM MgCl2 , 64 mM NaCl , 100 μg/mL BSA , 10 mM DTT , 400 μM ATP , 150 μM GTP , 150 μM CTP 10 μM UTP , 0 . 027 μM α-[32P]UTP ( 3000 Ci/mmol ) , 4 nM of indicated plasmid template , 20 nM POLRMT , 200 nM TFAM , 60 nM TFB2M , and 40 nM TEFM where indicated . The reactions were incubated at 32°C for 5 min and stopped by the addition of 200 μL stop buffer ( 10 mM Tris-HCl pH 8 . 0 , 0 . 2 M NaCl , 1 mM EDTA , 100 μg/mL glycogen ( Roche ) and 100 μg/mL proteinase K ( Ambion ) ) followed by incubation at 42°C for 45 min . The transcripts were recovered by ethanol precipitation and the pellets were dissolved in 20 μL gel loading buffer ( 98% formamide , 10 mM EDTA , 0 . 025% xylene cyanol FF , and 0 . 025% bromophenol blue ) and heated at 95°C for 3 min . The samples were analyzed on 4% denaturing polyacrylamide gels ( 1 × TBE and 7 M urea ) followed by exposure on photo film . Low Molecular Weight DNA Ladder ( NEB ) was used as a size marker . All experiments were performed multiple ( >3 ) times with similar results and each figure shows a representative gel image for that experiment . For R-loop detection , 1 . 5 μL of 5 . 0 M NaCl was added to each sample after the in vitro transcription reaction , followed by addition of 250 ng of RNase A ( ThermoFisher Scientific ) and incubation at 32°C for 5 min . The reactions were stopped and evaluated as the regular transcription reactions . Quantifications of R-loops were performed using ImageJ software ( https://imagej . nih . gov/ij/ ) . The intensity of R-loops larger than 100 nts was divided by the intensity of the CSBII transcript of the same reaction , to obtain a ratio indicating how efficient R-loop formation was . All experiments were performed multiple times with similar results and each figure shows a representative gel image for that experiment . All reaction volumes were 25 μL and contained 25 mM Tris-HCl pH 8 . 0 , 10 mM MgCl2 , 50 mM NaCl , 100 μg/mL BSA , 10 mM DTT , 400 μM ATP , 150 μM GTP , 150 μM CTP 150 μM UTP , 100 μM dATP , 100 μM dGTP , 100 μM dCTP or ddCTP as indicated , 10 μM dTTP , 0 . 027 μM α-[32P]dTTP ( 3000 Ci/mmol ) , and 8 nM of indicated template . In the case of RNA labeling the concentration of dTTP were 100 μM and UTP concentrations were 10 μM UTP and 0 . 027 μM α-[32P]UTP ( 3000 Ci/mmol ) . All reactions ( unless otherwise stated ) contained 200 nM of TFAM , 60 nM of TFB2M , 20 nM of POLRMT , 20 nM of D274A POLγA exo- ( or WT in Fig 3F lane 6 ) and 40 nM POLγB . RNase H1 was added at 2 nM and mtSSB was added at 40 nM unless otherwise stated . The reactions were incubated at 32°C for 30 min . Experiments with dCTP were stopped and evaluated as transcription reactions . Experiments with ddCTP were stopped by the addition of 5 μL stop buffer ( to a final concentration of 10 mM Tris-HCl pH 8 . 0 , 0 . 2 M NaCl , 1 mM EDTA , 660 μg/mL glycogen ( Roche ) and 100 μg/mL proteinase K ( Ambion ) ) followed by incubation at 42°C for 45 min . The reactions were treated with KOH ( 300 mM ) for 2 hrs at 55°C . The samples were recovered by ethanol precipitation in the presence of 0 . 5 volumes ammonium acetate ( 7 . 5 M ) , dissolved in 10 μL gel loading buffer ( 98% formamide , 10 mM EDTA , 0 . 025% xylene cyanol FF , and 0 . 025% bromophenol blue ) , heated at 95°C for 3 min . The products were analyzed on 6% denaturing polyacrylamide gels ( 1 × TBE and 7 M urea ) for samples with dCTP or 12% denaturing polyacrylamide sequencing gels ( 1 × TBE and 7 M urea ) for samples with ddCTP . The gels were exposed on photo film . All experiments were performed multiple times with similar results and each figure shows a representative gel image for that experiment . Patient and control fibroblast cells were grown in DMEM GlutaMAX medium , supplemented with 10% FBS , PEST and 10 μg/ml uridine in 5% humidified atmosphere at 37°C . Approximately 5 × 106 cells were collected and lysed for 30 min at 42°C in lysis buffer ( 10 mM Tris-HCl pH 8 . 0 , 5 mM EDTA , 10% SDS ) . An equal volume of phenol-chloroform was added , samples were mixed and centrifuged ( 15 , 000 g for 5 min , 4°C ) . The aqueous phase was saved and 100 mM NaCl and one volume isopropanol were added . After precipitation at -20°C for one hour , the samples were centrifuged ( 15 , 000 g , 20 min , 4°C ) and the pellets were washed with 70% EtOH . The DNA was resuspended in 100 μl TE buffer . DNA concentrations were measured using the Qubit fluorometric instrument ( ThermoFisher Scientific ) . Isolated DNA ( 1 . 8 μg ) was incubated for 1 hour at 37°C with or without RNase H2 ( NEB ) . Primer extension was performed with 2 U Taq DNA polymerase ( NEB ) in 1X ThermoPol buffer , 200 μM dNTPs and 1 . 5 pmol labeled primer . The primers were 5´-end labeled with PNK ( NEB ) and γ-[32P]ATP and were corresponding to L-strand positions 8-29 ( GGT CTA TCA CCC TAT TAA CCA C ) and 16 , 331-16 , 351 ( CAC ACA TCA ACT GCA ACT CCA ) . The primer extension reaction was performed with 5 minutes at 95°C , 30 seconds at 95°C , 30 seconds at 58°C , 45 seconds at 72°C and 5 minutes at 72°C with step 2-4 repeated in 20 cycles . The reactions were stopped and ethanol precipitated as for replication initiation reactions . The sequencing ladders were prepared with USB Sequenase Version 2 . 0 ( Affymetrix ) according to the manufacturers protocol . The primer extension experiment was performed multiple times with similar results and the figure shows a representative gel image for that experiment . Free 5′-ends of mtDNA from fibroblast cells were mapped by 5′-End-seq or HydEn-seq essentially as previously described [43 , 57] . In brief , 1 μg DNA was treated with 0 . 3 M KCl or 0 . 3 M KOH at 55°C for 2 hours . Samples were then phosphorylated with 3′-phosphatase-minus T4 polynucleotide kinase ( New England BioLabs ) for ligation to oligonucleotide ARC140 . After an adaptor was annealed ( ARC76–ARC77 ) , T7 DNA polymerase ( New England BioLabs ) was used to synthesize the second strand . Purified libraries were then sequenced using an Illumina NextSeq500 instrument . Reads were trimmed for quality and adapter sequence with cutadapt ( -m 15 –q 10match-reas-wildcards ) . Pairs with one or both reads shorter than 15 nts were discarded . Mate 1 of the remaining pairs was aligned to an index containing the sequence of all oligos used in the preparation of these libraries with bowtie sing bowtie 0 . 12 . 8 ( -m1 -v2 ) , and all pairs with successful alignments were discarded . Pairs passing this filter were subsequently aligned to the hg38 H . sapiens reference mitochondrial genome where the mitochondrial genome was cleaved at position 4 , 000 and OriH region was religated ( -m1 -v2 -X10000 --best ) . Single-end alignments were then performed using mate 1 of all unaligned pairs ( -m1 -v2 ) . The count of 5′-ends of all unique paired-end and single-end alignments were determined and these counts were converted to bedGraph format for visualization . Sequencing data have been deposited in the Gene Expression Omnibus under accession number GSE103612 . | Human mitochondria contain a double-stranded DNA genome that codes for key components of the oxidative phosphorylation system . The mitochondrial DNA ( mtDNA ) is replicated by a replication machinery distinct from that operating in the nucleus and mutations affecting individual replication factors have been associated with an array of rare , human diseases . In the present work , we demonstrate that RNase H1 directs origin-specific initiation of DNA replication in human mitochondria and that disease-causing mutations may impair this process . A unique feature of mtDNA replication is that primers required for initiation of leading-strand DNA replication are produced by the mitochondrial transcription machinery . A substantial fraction of all transcription events is prematurely terminated about 120 nucleotides downstream of the promoter and the RNA remains firmly associated with the genome , forming R-loops . Interestingly , the free 3′-end of these R-loops cannot directly prime initiation of DNA synthesis , but must first be processed by RNase H1 . The process is stimulated by the mitochondrial single-stranded DNA binding protein and faithfully reconstitutes replication events mapped in vivo . In combination with mapping of replication events in fibroblasts derived from patients with mutations in RNASEH1 , our findings point to a possible model for replication initiation in human mitochondria similar to that previously described in the E . coli plasmid , ColE1 . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"sequencing",
"techniques",
"nucleic",
"acid",
"synthesis",
"nucleases",
"mitochondrial",
"dna",
"enzymes",
"dna-binding",
"proteins",
"enzymology",
"dna",
"replication",
"forms",
"of",
"dna",
"mitochondria",
"dna",
"rna",
"synthesis",
"bioenergetics",
"molecular",
"bio... | 2019 | RNase H1 directs origin-specific initiation of DNA replication in human mitochondria |
DNA lesions are sensed by a network of proteins that trigger the DNA damage response ( DDR ) , a signaling cascade that acts to delay cell cycle progression and initiate DNA repair . The Mediator of DNA damage Checkpoint protein 1 ( MDC1 ) is essential for spreading of the DDR signaling on chromatin surrounding Double Strand Breaks ( DSBs ) by acting as a scaffold for PI3K kinases and for ubiquitin ligases . MDC1 also plays a role both in Non-Homologous End Joining ( NHEJ ) and Homologous Recombination ( HR ) repair pathways . Here we identify two novel binding partners of MDC1 , the poly ( ADP-ribose ) Polymerases ( PARPs ) TNKS1 and 2 . We find that TNKSs are recruited to DNA lesions by MDC1 and regulate DNA end resection and BRCA1A complex stabilization at lesions leading to efficient DSB repair by HR and proper checkpoint activation .
Maintenance of genome integrity is critical for both normal cellular functions and for suppressing mutagenic events that may lead to cancer [1] [2] . DNA damage can occur due to environmental agents such as UV light or irradiation , and endogenous sources such as oxidative by-products of cellular metabolism or stalled replication forks [2] . To prevent irreversible mutations that can occur throughout the life span of an organism , multiple repair systems have emerged during evolution [3] . Breaks that affect both DNA strands ( Double Strand breaks , DSBs ) are among the most lethal lesions as they can lead to the discontinuity of genetic information and chromosomal aberrations [2] . DSBs are repaired by two main pathways: Non Homologous End Joining ( NHEJ ) and Homologous Recombination ( HR ) [4] . NHEJ is used by cells to join broken ends by simple religation and although this pathway is active throughout the cell cycle , it mainly occurs during G1 [5] . The NHEJ pathway is often error prone and can drive chromosome translocations by joining distal DSBs from different parts of the genome [6] . HR functions predominantly when pairing of sister chromatids occurs during S/G2 and takes advantage of the information encoded by the homologous template to eliminate the DSB in an error-free manner [7] . During HR , DNA is processed to generate single stranded ends that are coated by RPA and subsequently by RAD51 . These nucleoprotein filaments are then prone to invade the homologous strand so that subsequent repair can take place [7 , 8] . Cells respond to DNA damage by initiating a signaling cascade , called the DNA damage response ( DDR ) , which leads to the activation of cell cycle checkpoints arresting the cell cycle and allowing the cell to repair the damage before division [9] . The DDR is initiated by the recruitment and extensive spreading of DDR proteins around the lesions that results in the formation of discrete foci [10] . A key stimulator of DDR spreading is the mediator of DNA damage checkpoint protein 1 ( MDC1 ) , which guides the perpetuation of the phosphatidylinositol 3-kinase ( PI3K ) –ataxia telangiectasia mutated ( ATM ) signaling pathway as well as the spreading of ubiquitination and subsequent recruitment of checkpoint mediators such as 53BP1 and BRCA1 [11 , 12] . BRCA1 is considered to be a master regulator of genomic integrity contributing to efficient repair of DSBs by HR , to DDR and to check-point activation [12] . Although it has been the focus of many studies , our knowledge of the 2089 amino acid ( aa ) long human MDC1 protein is not exhaustive . MDC1 was reported to directly interact with different DDR factors via its separate domains [13] . The Forkhead Associated Domain ( FHA ) of MDC1 was shown to be in contact with ATM , Chk2 and Rad51 . The MDC1 Ser-Asp-Thr ( SDT ) repeats interact with the MRE11-RAD50-NBS1 ( MRN ) complex , while the RNF8 Binding Domain ( RBM ) recruits the RNF8 ubiquitin ligase to MDC1-bound sites ( [13] and references therein ) . Furthermore , the BRCA1 C-terminal ( BRCTs ) repeats in the hMDC1 C-terminal domain were crystallized and shown to directly bind γ-H2AX [14] . Besides its major role as a platform for DDR signaling , MDC1 was also shown to play primordial roles in NHEJ and HR [15] [16] [17] . How one protein can fulfill these rather different roles is still an open question . Aiming bringing new knowledge concerning this point , we set forward to identify potential partners of MDC1 . We identified two novel MDC1 interacting partners , the poly-ADP-ribose polymerases ( PARPs ) Tankyrase 1 and 2 ( TNKS1/2 ) . We show that Tankyrases associate with DNA lesions in an MDC1-dependent fashion . Our data highlight the role of TNKSs in stabilizing the BRCA1-CtIP and BRCA1A complexes at DSBs and playing roles in HR and G2/M checkpoint activation .
To get insights into the mechanism of action of MDC1 , we searched for novel interacting partners . We conducted a yeast two-hybrid screen of a human placenta cDNA library using MDC1 as bait ( Fig 1A ) . Interestingly , two poly-ADP-ribose polymerases ( PARPs ) , TNKS1 ( PARP5a ) and TNKS2 ( PARP5b ) , were identified as proteins that interacted with the MDC1 “middle” domain ( Fig 1B ) . The clones identified in this screen harbored the ankyrin repeats of Tankyrases , a domain known for establishing protein-protein interactions ( Fig 1B ) [18 , 19] . For detailed results of the yeast two-hybrid screen see S1 Fig . To verify this interaction in human cells with the endogenous proteins , we performed co-immunoprecipitation experiments followed by Western Blot . As shown in Fig 1C , endogenous MDC1 interacts with TNKS1 and this association increases in the presence of the radiomimetic drug Neocarzinostatin ( NCS ) . The decrease in the levels of MDC1 upon DNA damage is possibly due to the proteolytic degradation of MDC1 , as shown in [20] ( see Fig 1C lane 3 ) . To further prove this interaction , we utilized our in vivo lacO-LacR targeting system ( Fig 1D ) that we have previously used to dissect the hierarchy of DNA repair factor recruitment at DDR foci [21] . Since the antibody recognizes mainly TNKS1 , we expressed tagged versions of TNKS1 or TNKS2 in the cells . Full-length mCherry-lacR-MDC1 efficiently recruited both flag-tagged TNKS1 and myc-tagged TNKS2 to the LacO array ( Fig 1E ) . To precisely map the MDC1 domain that recruits TNKSs , we generated deletion mutants of MDC1 fused to mCherry-lacR ( Fig 1A ) and expressed them in U2OS17 cells ( harboring the lacO repeats , Fig 1D ) . We found that the poorly characterized central region ( 742-1700aa , Fig 1A ) called hereafter the “middle” domain is sufficient to recruit both TNKSs ( Fig 2A , panels g-i and Fig 2B ) . TNKSs also interacted with equal efficiency with the MDC1 C-terminal region ( panels j-l in Fig 2A ) but not with the MDC1 N-terminal domain ( panels d-f in Fig 2A , see Fig 2B for quantifications ) . To map more precisely the TNKS Binding Domains ( TBDs ) of MDC1 , we continued our analysis by narrowing down the MDC1 regions that yield positive signals ( for the constructs see Fig 1A ) . Further deletion analysis of this “middle” domain showed that although the Proline-Serine-Threonine ( PST ) region ( 1130-1661aa ) does not interact with TNKSs ( panel p-r in Fig 2A ) , the domain between the known RNF8 binding domain ( RBM ) and PST , referred here as “interII” ( 800-1130aa ) ( Fig 1A ) , yielded in very efficient colocalization with TNKS1 or 2 ( panels m-o in Fig 2A , for quantifications see Fig 2B ) . Also , an interaction was detected between the BRCT domain of MDC1 ( 1881-2089aa ) and TNKS1/2 ( panels v-x in Fig 2A , for quantifications see Fig 2B ) . The finding that the colocalization frequency between tethered lacR-interII-MDC1 and TNKSs is higher than that of the full length MDC1 ( MDC1 FL ) or the “middle” region , points to an inhibitory role of the PST domain in this interaction . In line with this observation , deletion of the PST domain in the context of MDC1 FL increased the colocalization with TNKSs at the lacO array after tethering ( Fig 2C ) . A comparison of the amino acid sequences of the “interII” and “BRCTs” domains and that of known TBDs from the literature [18] allowed us to map the two potential TBDs within the MDC1 sequence at positions 948-955aa ( “interII” domain ) and 1993-2000aa ( “BRCTs” domain ) ( Fig 3A and 3B ) . Site directed mutagenesis of three amino acids of the eight present in the binding motifs to alanine ( Fig 3A and 3B ) showed that disrupting one TBD leads to decreased interaction , but mutations in both TBDs are required for complete loss of the MDC1 and TNKS colocalization signal ( Fig 3C and 3D ) . To confirm that the association of TNKSs with MDC1 depends on the above-described TBDs , we performed co-immunoprecipitation studies on extracts overexpressing mCherry-lacR-interII or mCherry-lacR-BRCT harboring or not the TBD mutations as well as the full-length mCherry-lacR-MDC1 ( wt or TBD mutant ) with FN-TNKS1 ( Fig 3E and 3F ) . We indeed found that mutations at the TBDs diminish the interaction with TNKS1 FL or deletion mutants ( Fig 3E and 3F ) . Overall , these observations establish the existence of two TBDs within the MDC1 sequence that ensure the interaction with Tankyrases in vivo . The above results concerning the interaction between MDC1 and TNKS1/2 suggest that TNKSs might localize to sites of DNA damage . To investigate this possibility , we first inflicted DNA damage in U2OS cells using an 800nm multiphoton laser . Efficient damage was visualized by the recruitment of GFP-MDC1 and the phosphorylation of H2AX γ-H2AX ) at the laser stripes ( Fig 4A ) . Both TNKS1 and TNKS2 as well as a PARP dead mutant version of TNKS1 were visualized at the sites of DNA damage that colocalized with the above DDR markers . ( panels c , h and l in Fig 4A ) . We next sought to follow the potential recruitment of Tankyrases at site-specific DSBs by taking advantage of the presence of a unique ISce-I homing endonuclease recognition site that lies adjacent to the lacO repeats in U2OS17 cells ( Fig 1D ) . Notably , we observed colocalization of the mCherry-lacR spot with TNKSs only in cells expressing the ISce-I endonuclease ( Fig 4B ) suggesting that TNKSs are recruited to “pure” DBSs . Subsequent quantification of TNKS colocalization with the ISce-I-induced breaks showed that TNKSs are recruited to DNA lesions in a fraction of cells ( around 30% , Fig 4B ) . Notably , the PARP dead mutant version of TNKS1 was recruited as efficiently as the wild type protein , suggesting that the PARP activity of TNKSs is not required for their localization to DNA damage sites ( Fig 4B , graph on the right , see also S2 Fig ) . Since the tethering of MDC1 to the lacO array is sufficient to recruit TNKSs ( see Figs 1E and 2B ) , we next assessed whether induction of ISce-I breaks next to MDC1-bound chromatin further increased the occupancy of the lacO array by TNKSs . Indeed , upon ISce-I break induction the MDC1-TNKS colocalization frequency was higher and the signal intensity on the array was stronger ( panels n-p and graph in Fig 4C ) . As MDC1 tethering was previously shown to trigger DDR [21] , we tested if TNKS loading requires this activity of the MDC1 . Our results show that the DDR depends on the C-terminal end of MDC1 ( Fig 4D ) , while the interII domain that recruits most efficiently TNKSs to the chromatin ( see Fig 2B ) does not activate such a response ( Fig 4D ) . To further test whether the interaction of TNKSs with the BRCT domain depends on its capacity to activate DDR , we inhibited DDR induced by BRCT after tethering using the ATM inhibitor . We find no effect on the recruitment efficiency of TNKS1 on tethered BRCT at the array ( Fig 4E ) . Thus , the abilities of MDC1 to trigger damage response and to load TNKSs to the chromatin are independent . The frequency of recruitment of TNKSs to pure DSBs ( ~30% , see Fig 4B ) suggests that a cell cycle-specific recruitment occurred . To test this idea , we quantified the percentage of TNKS and BRCA1 colocalization in ISce-I-inflicted breaks . We found that the vast majority of ISce-I breaks marked with TNKSs were also marked with BRCA1 ( ~80% , Fig 5A ) , which further points to a preferential recruitment of TNKSs to DNA damage in the S and G2 phases of the cell cycle . In line with this observation , the recruitment of TNKS1 at the lacO/ISce-I locus was reduced in cells arrested in G1/S by mimosine ( Fig 5B ) . The colocalization of TNKSs with naturally occuring γ-H2AX foci ( panel b and c in Fig 4B ) further reinforces the hypothesis about their presence at naturally occurring DNA lesions . Moreover , FN-TNKS1 exerted colocalization with γ-H2AX foci induced by NCS ( Fig 5C ) . This colocalization increased significantly in the case of FN-TNKS1 in cells arrested in G2 by RO-3306 ( Fig 5C ) . In line with this observation , the expression and the nuclear localization especially of TNKS1 seems to increase in S and G2 phases of the cell cycle ( S3A and S3B Fig ) . To establish if TNKS recruitment to breaks requires MDC1 , we quantified the frequency of TNKS1/2 localization to the lacO array in the presence or absence of ISce-I breaks in control and MDC1 depleted cells . Depletion of MDC1 significantly affected the recruitment of TNKSs to ISce-I-induced breaks ( Fig 5D ) , and hampered efficient TNKS1 foci formation after NCS treatment ( Fig 5E ) . On the other hand , downregulation of TNKSs had no effect on MDC1 recruitment to single DSBs in U2OS17 cells ( S4A Fig ) neither on MDC1 foci formation after DSB induction with the radiomimetic drug NCS ( S4B Fig ) . Efficient depletion of MDC1 and TNKSs was assessed by RT-PCR and Western blot analysis or immunofluorescence staining ( S5A and S5B Fig ) . Our findings thus provide evidence for MDC1-dependent recruitment of Tankyrases to DSBs in vivo as part of the early response to DNA damage . In a recent report , TNKS2 was shown to interact with RAD54 [18] . This result points to a potential role of TNKSs in DSB repair by HR via loading of HR factors at DNA lesions [22] [23] . Unfortunately , we were unable to visualize RAD54 foci after NCS treatment , and we therefore set forward to analyze the frequency of RAD51 foci in cells depleted for Tankyrases . We indeed observed a substantial drop in the number of cells forming RAD51 foci after NCS treatment ( Fig 6A and 6B ) . A similar result was obtained in U2OS17 cells depleted for Tankyrase1/2 and transfected with ISce-I . Cells in which Tankyrases were downregulated showed a lower frequency of RAD51 accumulation at lacO/ISce-I breaks ( Fig 6C ) . To verify if Tankyrases indeed favor the recruitment of RAD51 to pure DSBs , we tethered Tankyrase 1 fused to GFP-lacR to the chromatin and induced DSBs by ISce-I . Forced binding of TNKS1 to the lacO array resulted in increased RAD51 recruitment ( Fig 6D ) . To investigate whether the effect of TNKSs on RAD51 foci formation depends on their interaction with MDC1 , we performed rescue experiments . To this end , RAD51 foci were quantified in U2OS cells depleted for MDC1 and complemented with either wtMDC1 or the TBD mutant . As expected , depletion of MDC1 reduced RAD51 foci formation and this effect was rescued by expression of wtMDC1 but not the TBD mutant ( S6A Fig ) . Although this result point to the importance of MDC1-TNKS interaction on RAD51 foci formation , we noticed that the MDC1 TBD mutant exerted decreased ability to form foci compared to the wt ( S6B Fig ) . Therefore with this experiment we cannot separate the two functions . To overcome this limitation , we tethered wt and TBD mutant of MDC1 to the lacO array and assessed RAD51 recruitment after ISce-I cutting . Interestingly , tethering of the mutant leads to a significant drop in RAD51 recruitment after DSB induction with ISce-I compared to the wt MDC1 ( Fig 6E ) . In line with the effect of TNKSs’ depletion on RAD51 loading to lesions , Tankyrase knock down hampered the efficient repair by HR at a level comparable to our positive control , the BRCA1 knock down ( Fig 6F ) . It is worth noting that the cell cycle profile of TNKS KD cells proved similar to control cells ( S3C Fig ) and a more detailed analysis for the ratio of replicating cells showed no difference compared to the siSCR control either ( S3D Fig ) . The efficiency of BRCA1 depletion by siRNA was verified by WB ( S5C Fig ) . To investigate whether the role of TNKSs in HR depends on their PARP activity , we performed the same assay in cells depleted for TNKS1 and 2 and complemented with wild type TNKS1 and 2 or the combination of their PARP dead counterparts . Interestingly , wt as well as PARP dead TNKSs rescue the defect in HR , pointing to a structural role of TNKSs in DNA repair by HR that doesn't require the catalytic activity ( Fig 6F ) . In line with this observation , inhibition of TNKSs’ PARP activity using the inhibitor XAV-939 had no effect in HR ( S6C and S6D Fig ) . To get further insight into the mechanism by which TNKSs regulate HR , we investigated their role in DNA end resection . To this end , we assessed two markers of resection , CtIP and the phosphorylated form of RPA ( RPA-P ) at ISce-I breaks in control cells and cells depleted for TNKSs . Interestingly , downregulation of TNKSs reduced substantially the signal of these markers at the LacO/IsceI locus after break induction ( Fig 7A ) . Since CtIP was shown to act together with BRCA1 , we tested the efficiency of recruitment of BRCA1 at ISce-I and NCS induced breaks in cells depleted for TNKSs . We find that downregulation of TNKSs affect the recruitment of BRCA1 to DSBs in vivo ( Fig 7B and 7C ) . Moreover , TNKS1 tethering to the lacO array itself is enough to recruit BRCA1 ( Fig 7D ) . This effect is mediated by MDC1 since tethering of the Tankyrase binding mutant form of MDC1 does not lead to the recruitment of BRCA1 as compared with the wt ( Fig 7E ) . Altogether our data shed light on a role of Tankyrases in HR via mediating the recruitment of the CtIP-BRCA1 complex to DSBs . Besides its interaction with RAD54 , Tankyrase2 was also shown to bind to MERIT40 [18] . To verify if the reported interaction takes place on the chromatin , we investigated the ability to TNKS1 to recruit MERIT40 when tethered to the lacO array chromatin . Interestingly , the colocalization frequency of TNKS1 and MERIT40 was nearly 100% suggesting a strong interaction between the two proteins ( Fig 8A ) . Furthermore , the colocalization is independent of the PARP activity of TNKS1 , as the catalytic dead mutant TNKS proved just as efficient in recruiting MERIT40 as its wild type counterpart ( Fig 8A ) . These results are in agreement with previous results showing that MERIT40 is not a PARylation target of TNKSs and this interaction does not involve the PARP domain [18] . MERIT40 was shown to be required for the integrity of the BRCA1A complex containing RAP80 , BRCA1 , BRCC36 and CCDC98 [24] [25] . We thus tested if other subunits of this complex are also loaded to chromatin bound by TNKS1 . We indeed found that RAP80 was recruited to the lacO array with 100% efficiency in a PARP activity independent manner by TNKS1 ( Fig 8A ) . To verify that TNKS1 tethering does not trigger the DDR , we also observed γ-H2AX , MDC1 and 53BP1 accumulation at the TNKS1 bound array , but none of these factors or modifications could be detected ( S7A Fig ) . To test the potential role of TNKSs in recruiting BRCA1A subunits to chromatin after DNA damage , we depleted TNKS1 and 2 in U2OS17 cells and tested the efficiency of MERIT40 and RAP80 recruitment to ISce-I induced DSBs ( Fig 8B ) . Compared to the scrambled control , downregulation of TNKS expression decreased the association of all these factors with DSBs ( Fig 8B ) . On the other hand , downregulation of TNKSs had no effect on 53BP1 recruitment and H2AX phosphorylation ( S7B Fig ) . Similar results were obtained in U2OS cells treated with NCS; TNKS1/2 depletion resulted in a substantial decrease in the % of cells positive for RAP80 foci ( Fig 8C ) , while it had no effect on the ratio of foci-positive cells for γ-H2AX , MDC1 or 53BP1 ( S7C Fig ) . Furthermore , the proper loading of the subunits of the BRCA1A complex to pure DSBs is significantly affected when the Tankyrase binding mutant version of MDC1 is bound to the lacO array compared to the control ( Fig 8D ) . MERIT40 was previously reported to play a fundamental role in assuring the integrity of the BRCA1A complex . To determine whether BRCA1 loading to the array after TNKS1 tethering is mediated by MERIT40 or is due to a direct protein-protein interaction between BRCA1 and TNKSs , we depleted MERIT40 from U2OS17 cells using siRNA and tested the BRCA1 recruitment at the TNKS1 bound array . We found that siMERIT40 cells had a profound defect in BRCA1 recruitment in the absence of ISce-I break that represents the condition that BRCA1 protein is recruited to the lacO array only due to the interaction with TNKSs ( Fig 8E ) . Depletion of MERIT40 by siRNA was verified by IF ( S5D Fig ) . Altogether , our observations underline the role of TNKSs in mediating the association of two BRCA1-containing complexes with DNA lesions; the BRCA1A and the CtIP-BRCA1 complex . Our data point to a TNKS dependent recruitment of BRCA1A complex to DSBs mediated by MERIT40 . However , it was previously shown that RAP80 and MERIT40 recruitment to DNA lesions depends on RNF8-dependent ubiquitination [26 , 27] [28] . To investigate whether the recruitment of RAP80 , MERIT40 and BRCA1 by TNKSs depends on RNF8 , we depleted RNF8 by siRNA and we tested their recruitment efficiency on the lacO array after TNKS1 tethering in the absence or presence of ISce-I break . We find that RNF8 depletion does not affect the recruitment of RAP80 and MERIT40 to the lacO locus upon tethering of TNKS1 ( Fig 9A ) . At the same time siRNF8 transfected cells have only a minor reduction in the BRCA1 recruitment efficiency on tethered TNKS1 in vivo ( Fig 9A ) . To further test whether the ubiquitin binding property of RAP80 is responsible for the interaction with TNKSs , we assessed the recruitment of the UIM deletion mutant RAP80 not binding to DSBs to tethered TNKS1 [29] . Interestingly , we observed that the wild type and the UIM mutant RAP80 are equally recruited to the lacO array after TNKSs tethering ( Fig 9B ) . These findings suggest that TNKS interact with the BRCA1A complex in a ubiquitin independent manner . To investigate the relationship of the two modes of recruitment of BRCA1A complex to breaks ( the RNF8 dependent and the TNKS1 dependent ) , we sought to assess the kinetics of recruitment of RAP80 and MERIT40 to lesions induced by NCS upon depletion of RNF8 or TNKS1/2 . As it was reported before [26 , 27] , RNF8 depletion affects dramatically the initial recruitment of RAP80 to breaks . On the other hand , in cells that are depleted for TNKS1 and 2 , RAP80 foci formation is mainly affected at later time points , suggesting that RNF8 is essential for the recruitment of RAP80 and TNKSs for its stabilization at breaks ( Fig 9C ) . We were unfortunately unable to detect MERIT40 foci at NCS induced breaks . Since MERIT40 and BRCA1 were shown to be required for proper activation of the G2-M checkpoint after DNA damage [25 , 30] , we quantified the mitotic fraction of U2OS cells treated with phleomycin in control and TNKS depleted cells . In cells transfected with the non-targeting control siRNA there was a sharp drop in the number of cells undergoing mitosis after DNA damage induction , but in cells treated with siRNAs depleting TNKSs a significant percentage of cells could pass to the mitotic phase similarly to siMERIT40 and siBRCA1 conditions ( Fig 9D ) . Our data suggest that Tankyrases associate with DSBs to 1 . promote homology-mediated repair by regulating resection and RAD51 loading and to 2 . activate the check point , by stabilizing the BRCA1A complex ( Fig 9E ) .
MDC1 is a key regulator of DDR signaling at the chromatin surrounding DSBs [13] . It has been suggested to function as a platform for the recruitment of the ATM kinase and the perpetuation of ubiquitination by mediating the binding of the ubiquitin ligases RNF8 and RNF168 to DNA lesions [13 , 31] . Here we provide evidence for novel roles of MDC1 . The MDC1-dependent Tankyrase recruitment to DSBs reported here plays considerable roles in DSB repair by HR and in retention of the BRCA1A complex at lesions ( Fig 9E ) . Tankyrase 1 is a multifunctional poly ( ADP-ribose ) polymerase that was first shown to localize to telomeres through its interaction with the shelterin component TRF1 [32 , 33] . A recent study also showed that TNKS1 depletion by shRNA leads to DNA damage response activation exemplified by increased γ-H2AX and 53BP1 foci [33] . It has been proposed that DNA damage is caused by persistent cohesion of telomeric sequences after the completion of S-G2 and during mitosis [33] . In C . elegans , Tankyrase expression was shown to increases after DNA damage [34] . Our study provides evidence for a direct role of TNKSs in Double Strand Break Repair ( DSBR ) through their binding to MDC1 that can offer an additional explanation for the persistent DNA breaks observed in the absence of TNKS1 . We describe here different modes of action for TNKSs in DSBR . First we found that TNKSs affect Homologous Recombination by controlling DNA end resection and therefore the recruitment of RAD51 . Our data show that TNKS1 binding to chromatin is sufficient to promote binding of BRCA1 and CtIP and thereby promote resection as it was shown before [35] . Then we found that TNKSs load the BRCA1A complex to DSBs . MERIT40 is a regulator of BRCA1A complex integrity and its depletion destabilizes the complex and affects its recruitment to DSBs [24 , 25 , 30] . In agreement with this , we observe that TNKSs mediate the loading of MERIT40 , RAP80 and BRCA1 to chromatin . RNF8-dependent ubiquitination was shown to be important for MERIT40 and RAP80 recruitment to DNA lesions [26] [28] [24] . In this study , we demonstrate a novel mechanism for the recruitment of the BRCA1A complex to DSBs that involves interaction of MERIT40 with TNKSs . Our observations show that ubiquitination acts as the primary signal for the recruitment of the BRCA1A components and TNKSs are needed for the stabilization of the complex at the sites of damage . Moreover , our data clearly show that the stabilization and the interaction with TNKSs are independent of ubiquitin binding . The BRCA1A complex was reported to play an inhibitory role in resection and depletion of some of its subunits leads to increased Homologous recombination efficiency [36 , 37] . Our experiments show that TNKSs , although interact with the BRCA1A complex , regulate HR in a positive way . This can be explained by the fact that TNKSs promote the recruitment of two BRCA1 containing complexes to lesions , the BRCA1-CtIP and the BRCA1A complexes . Several recent reports highlighted the importance of poly ( ADP ) ribosylation in modulating the function of BRCA1 containing complexes . TNKSs were described to form a complex with ATM and BRCA1 during mitosis that assures proper mitotic progression through regulation of the poly ( ADP ) ribosylation of NuMA1 [38] . Also , poly ( ADP ) ribosylation was shown to be important for the rapid recruitment of the BRCA1-BARD1 heterodimer to DSBs [39] . Furthermore , PARylation of BRCA1 was recently reported to play a role in stabilizing the RAP80-BRCA1 interaction and modulating its DNA binding capacity [40] . As Tankyrase substrates were shown to get degraded [18] , an appealing hypothesis would be that MDC1 per se is PARylated and in a timely controlled manner gets erased from the chromatin by subsequent proteasomal degradation . In support of this idea , MDC1 is degraded upon DNA damage in an RNF4 dependent manner [20 , 41] . Also , in a recent study the Drosophila homologue of Tankyrase was reported to have a direct role in activating the 26S proteasome subunit’s activity through PARylating its partner , PI31 [42] . Moreover , RNF146 , a ubiquitin ligase that is reported to ubiquitinate proteins PARylated by TNKSs , was also reported to play a role in DNA repair [43] . Although this is a very exciting possibility that could provide evidence as to how DDR is switching off , our efforts to detect MDC1 PARylation by TNKS1 in vitro , or stabilization of MDC1 protein in cells in which TNKSs are downregulated , were unsuccessful . Nevertheless , we cannot exclude the possibility that our experimental conditions were not optimal to detect such a modification . An interesting feature of the above- described MDC1-TNKS interaction is that it is controlled by MDC1 itself . Indeed , we observed that in the absence of the MDC1 “PST” domain there is a substantial increase in the percentage of cells that display TNKS and tethered MDC1 colocalization , suggesting that the “PST” domain has an inhibitory effect on the interaction . Our finding that TNKS recruitment to DSBs occurs mainly during the S-G2 phases indicates that the “PST” domain may undergo post-translational modification ( s ) during cell cycle that induce conformational changes in MDC1 modulating its binding to TNKSs . Notably , the MDC1 “PST” domain was previously shown to get phosphorylated after DNA damage [44 , 45] and MDC1 is also phosphorylated during mitosis , further supporting our hypothesis that a “PST”-dependent structural change might occur [46] . Moreover , we showed that TNKS PARP activity is not required for the association with DSBs , which is in agreement with our yeast two-hybrid data showing that the TNKS ankyrin repeat domain interacts with MDC1 . The ankyrin repeat domain was previously shown to be important for the majority of interactions between TNKSs and their known binding partners [33] . The requirement for two TBDs is another interesting feature of the MDC1-TNKSs interaction . Both TBDs are highly conserved in MDC1 sequences from different species , highlighting their importance in MDC1 functions ( S8 Fig ) . It is noteworthy , that although in the context of the FL MDC1 the two TBDs have an additive role in the recruitment of TNKSs to chromatin , the “interII” domain including TBD1 is more proficient in loading TNKSs to the lacO array upon tethering . Of note , TBD2 lies in the hinge region between the two BRCTs and TNKS binding might thus be further regulated by the interaction of MDC1 with γ-H2AX . Future structural studies might highlight the importance of the two TBDs . Recently the idea of targeting Tankyrases in cancer treatment has emerged due to the discovery of their role in Wnt signalling [47] . Moreover , TNKS depletion was shown to be selectively lethal in a BRCA1 deficient background through a mechanism that was proposed to involve supernumerary chromosomes [48] . However , in the light of our new data about TNKS’s role in G2-M checkpoint activation , escape from the checkpoint might result in elevated DNA damage that can be also exemplified by supernumerary chromosomes [48] . Therefore , our results open a new perspective on targeting Tankyrases for cancer therapy .
U2OS , U2OS17 and U2OS19 cells were cultured at 37°C in DMEM supplemented with 4 . 5g/l glucose , 10% FCS and gentamicin . GCV6 cells and HRind cells were cultured as previously described [17 , 49] . JetPEI ( PolyPlus ) , Fugene6 ( Promega ) , Lipofectamine 2000 ( Invitrogen ) , INTERFERin ( PolyPlus ) , Neocarzinostatin ( NCS , Sigma ) Phleomycin ( Sigma ) , XAV-939 ( Tocris ) , Mimosine ( Sigma ) , RO-3306 ( Millipore ) , ATMi KU55933 ( Tocris ) . The following antibodies were used during our study: Anti-lacR ( gift from Dr A . Belmont ) , anti-flag ( F7425 , Sigma ) , anti-flag M2 ( F3165 , Sigma ) , anti-Myc ( clone 9E10 ) , anti- γ-H2AX ( ab22551 , abcam ) , anti-H2AX ( ab11175 , abcam ) anti-53BP1 ( NB100-304 , Novus ) , anti-RAD51 ( PC130 , Calbiochem ) , anti-BRCA1 ( OP92 and OP93 , Calbiochem ) , anti-MERIT40 ( gift from Dr J . Chen ) , anti-histone H3 S10P ( 51TA2H12 , Active Motif ) , anti-MDC1 ( MDC1-50 , Sigma ) , anti-TNKS ( ab13587 , abcam ) , anti-HA ( 11867423001 , Roche ) , anti-ATM ( Novus NB100-104 ) , anti-DNA-PKcs ( Millipore 04–1024 ) , anti-tubulin ( Sigma-Aldrich T5168 ) , anti-Merit40 ( A302-516A Bethyl ) , anti RAP80 ( A300-763A Bethyl ) , anti-CtIP ( Clone 14–1 Active motif ) , anti-phospho-RPA32 S4/8 ( A300-245A , Bethyl ) . Secondary antibodies for IF were purchased from Life Technologies , secondary antibodies for WB were from Jackson Immuno Research . The home-made rabbit polyclonal anti-MDC1 antibodies ( clones #2991 and #2992 ) were developed using the ETDAEEGTSLTASVVADVRK peptide corresponding to the 577–597 aa region of MDC1 as antigen . The peptide was coupled to ovalbumine and used to immunize two rabbits . All siRNAs used in the study were purchased from Dharmacon . Except if not precised , they were “On-Target plus SMARTpool” siRNAs containing four different sequences against the target mRNA . Custom designed siRNAs were the following: The FN-TNKS1 and MN-TNKS2 expressing plasmids in pCDNA 3 . 1 backbone having a flag or myc tag and NLS signal were gifts from Dr S . Smith . The PARP catalytic dead TNKS1 was subcloned together with its flag tag and the NLS from a retroviral vector ( gift from S . Smith ) to pCDNA 3 . 1 using the XhoI and HindIII sites . The flag-TNKS2 expressing plasmid and its mutant version were purchased from AddGene . The GFP-lacR-TNKS1 expressing plasmid and its mutant version were cloned in two steps: first the GFP cassette was added to the HindIII site of the FN-TNKS1 plasmid , and then the lacR coding sequence was subsequently cloned to the NotI site . mCherry-lacR-MDC1 deletion constructs were the results of amplifying by PCR the appropriate size fragment and subcloning it to the mCherry-lacR vector [21] . HA-ISce-I was a gift from Maria Jasin . The cDNA of RAP80 wt and UIM deletion mutant was amplified by PCR using the CFP-RAP80wt and CFP-RAP80ΔUIM vectors [50] and was subsequently cloned to the mCherrry-C2 vector . Directed mutagenesis to obtain the TBD mutant versions of MDC1 was done using the QuikChange site-directed mutagenesis kit . The siRNA resistant form of TNKS1 was generated by directed mutagenesis using the same kit . The mutations of the two Tankyrases rendering them catalytically inactive are: HEK293T cells were transfected with the indicated plasmids using Fugene 6 ( Promega ) following the supplier’s recommendations . Two days later cells were scraped in TENT buffer ( 25 mM Tris-HCl 7 . 5 , 2 mM EDTA 150 mM NaCl , 1% Triton X-100 protease inhibitor cocktail , ( Roche ) ) and incubated in ice for 15 min . The extracts were centrifuged and the supernatant was pre-cleared and subsequently incubated with the anti-Flag-M2 antibody for 2 hours . Protein G sepharose beads were added to the antibody-protein complex for 16 hrs . The beads were washed 3 times with TENT buffer and the bound fraction was eluted in SDS loading buffer . For the endogenous MDC1 IP , HEK293T cells were treated or not with 100 ng/ml NCS for 15 min and released for 1 hour prior to cell lysis and Immunoprecipitation procedure as described above . 1% input or the immunoprecipitated fractions were separated on SDS PAGE and transferred to Nitrocellulose membrane . The membrane was blocked in 5% milk PBS and incubated O/N at 4°C with the indicated antibodies . After three washes HRP-conjugated secondary antibodies were added for 1 hour and the membrane was developed using ECL detection kit ( GE Healthcare ) . U2OS cells -treated or not with NCS- were incubated in Hypotonic buffer ( 10 mM Hepes-KOH pH 7 . 9 , 1 . 5 mM MgCl2 , 10 mM KCl , 0 . 5 mM DDT ) on ice for 15 min followed by douncing using the loose pestle . After centrifugation of the samples the supernatant was considered the cytoplasmic fraction and the pellet ( nuclei fraction ) was further lyzed using RIPA buffer . The protein content of the cytoplasmic and nuclear fractions was determined by Bradford assay ( BioRad ) and equal protein amounts were analyzed by SDS-PAGE . FACS analysis of GFP ( for HR efficiency ) or propidium iodide staining ( for cell cycle analysis ) was conducted on a FACS Calibur ( Becton Dickinson ) and the results were quantified using FlowJo . RNA was isolated from the cells using TriReagent ( Molecular Research Center ) as suggested by the manufacturer . RT-qPCR reactions were done in triplicates using the QuantiTect SYBR Green PCR Kit ( Qiagen ) in a Roche Light Cycler . All the expression values were normalized to that of CyclophilinB . HR assay was described earlier [17 , 49] . Staining was done following standard IF protocol . Briefly , cells were washed with PBS , fixed in 4% PFA and permeabilized with PBS 0 . 1% tritonX-100 . Blocking and incubation in antibodies were performed in 10% heat inactivated FCS , washes were done with PBS 0 . 1% triton X-100 . Nuclei were counterstained with DAPI and cells were mounted using the ProLong Gold antifade reagent of Molecular Probes . Confocal microscopy pictures were taken at a Leica SP2 microscope , Z stack width was usually 0 . 5μm . For foci positive cell counting , at least 100 cells were analyzed for each condition . For RAD51 staining a cell was considered positive bearing >3 foci . For BRCA1 , RAP80 and MDC1 staining a cell was considered positive bearing >5 foci . Multiphoton laser striping was conducted on a Leica SP5 system , using a DMI6000 microscope with a 63x Plan Apo 1 . 4 objective and a Coherent Chameleon laser at 800nm wavelength . The microscope stage was equipped with a 37°C Life Imaging System box . For inflicting DNA breaks in several cells in a mosaic , the Taile-Scan option of the Leica program was applied . U2OS cells were seeded in 6well plates and transfected by the indicated siRNAs as suggested by the supplier , using Lipofectamine2000 ( Invitrogen ) . Twenty-four hours later cells were trypsinized and seeded at 5000cell/well concentration in glass-bottom , black 96 well plates . Another 24 hours later , cells were treated with 20μg/ml Phleomycin for 1 hour and released for 6 hours . Cells were fixed and processed for IF with an anti-histone H3 S10P antibody to mark mitotic cells . The percentage of positive cells was determined by the automated InCell1000 analyzer . Yeast two-hybrid screening was performed by Hybrigenics Services , S . A . S . , Paris , France ( http://www . hybrigenics-services . com ) . The coding sequence for human MDC1 aa 742–1698 ( GenBank accession number gi: 132626687 ) was PCR-amplified and cloned into pB66 as a C-terminal fusion to Gal4 DNA-binding domain ( N-Gal4-MDC1-C ) . The construct was checked by sequencing and used as a bait to screen a highly complex random-primed human placenta cDNA library constructed into pP6 . pB66 and pP6 derive from the original pAS2ΔΔ and pGADGH plasmids , respectively . 90 million clones ( 10-fold the complexity of the library ) were screened using a mating approach with Y187 ( matα ) and CG1945 ( mata ) yeast strains as previously described [52] . A total of 169 colonies were selected on a medium lacking tryptophan , leucine and histidine . The prey fragments of the positive clones were amplified by PCR and sequenced at their 5’ and 3’ junctions . The resulting sequences were used to identify the corresponding interacting proteins in the GenBank database ( NCBI ) using a fully automated procedure . A confidence score ( PBS , for Predicted Biological Score ) was attributed to each interaction as previously described [53] . Cells were blocked in G1 by a 20-hour treatment of 0 , 5mM mimosine , released from this block for 5 hours for S phase , or blocked for 20 hours by 10uM RO-3306 to enrich the population in G2 . Immunofluorescence staining of FN-TNKS1 and γ-H2AX was analyzed by microscopy . Cells were considered positive for colocalization after inspection of the merged images . Quantification was done in 50 cells for each condition . Intensity plots were generated by the Image J software . | MDC1 recruit Tankyrases to DNA lesions to regulate homologous recombination and to control check-point activation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunofluorescence",
"staining",
"gene",
"regulation",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"light",
"microscopy",
"dna",
"damage",
"immunoprecipitation",
"microscopy",
"confocal",
"microscopy",
"dna",
"epigenetics",
"chromatin",
"research",
... | 2016 | Tankyrases Promote Homologous Recombination and Check Point Activation in Response to DSBs |
Many loci in the human genome harbor complex genomic structures that can result in susceptibility to genomic rearrangements leading to various genomic disorders . Nephronophthisis 1 ( NPHP1 , MIM# 256100 ) is an autosomal recessive disorder that can be caused by defects of NPHP1; the gene maps within the human 2q13 region where low copy repeats ( LCRs ) are abundant . Loss of function of NPHP1 is responsible for approximately 85% of the NPHP1 cases—about 80% of such individuals carry a large recurrent homozygous NPHP1 deletion that occurs via nonallelic homologous recombination ( NAHR ) between two flanking directly oriented ~45 kb LCRs . Published data revealed a non-pathogenic inversion polymorphism involving the NPHP1 gene flanked by two inverted ~358 kb LCRs . Using optical mapping and array-comparative genomic hybridization , we identified three potential novel structural variant ( SV ) haplotypes at the NPHP1 locus that may protect a haploid genome from the NPHP1 deletion . Inter-species comparative genomic analyses among primate genomes revealed massive genomic changes during evolution . The aggregated data suggest that dynamic genomic rearrangements occurred historically within the NPHP1 locus and generated SV haplotypes observed in the human population today , which may confer differential susceptibility to genomic instability and the NPHP1 deletion within a personal genome . Our study documents diverse SV haplotypes at a complex LCR-laden human genomic region . Comparative analyses provide a model for how this complex region arose during primate evolution , and studies among humans suggest that intra-species polymorphism may potentially modulate an individual’s susceptibility to acquiring disease-associated alleles .
Genomic instability is a major contributor to de novo mutations that can occur in the process of human genome evolution [1–3] . Genomic rearrangements can be mediated by various mechanisms , including nonallelic homologous recombination ( NAHR ) , nonhomologous end joining , mobile element insertion ( e . g . long interspersed element ( LINE ) -mediated retrotransposition ) and replication based mechanisms [4] . Low copy repeat ( LCR ) mediated NAHR plays a significant role in genomic instability resulting in rearrangements associated with genomic disorders [5] . LCRs , also known as segmental duplications , are two or more repeated sequences that usually span 10–400 kilobases ( kb ) each and share >95% DNA sequence identity [6 , 7] . LCRs are highly homologous , and constitute ~5–10% of the human and great ape genomes [6 , 8 , 9] . LCRs provide substrates for NAHR-mediated crossing-over that results in structural variants ( SVs ) including copy number variants ( CNVs ) such as duplications and deletions of large genomic segments [5] or copy number neutral events such as inversions [10–12] . Numerous NAHR-mediated rearrangements are associated with genomic disorders by affecting dosage sensitive genes . For example , Potocki-Lupski syndrome ( PTLS , MIM #610883 ) or Smith-Magenis syndrome ( SMS , MIM #182290 ) are frequently caused by an ~3 . 7 megabases ( Mb ) NAHR-mediated common recurrent duplication or deletion , respectively . These recurrent rearrangements of 17p11 . 2 utilize directly oriented proximal and distal SMS-REPs as substrates for NAHR [13–18] . LCRs originated from genomic evolutionary processes and can facilitate responses to selective pressure by creating new genes that may contribute to lineage-specific phenotypes . LCRs can also configure local genomic structure in a manner that contributes significantly to disease susceptibility [19–24] . Because of their repetitive nature and structural complexity , LCRs can confound the accuracy of human and nonhuman mammalian genome assemblies . Discerning long stretches of paralogous , highly identical sequences can be difficult; this problem becomes particularly challenging when there are more than two copies in a haploid genome [6 , 25 , 26] , and consequently LCRs are likely under-represented in draft genome assemblies for many species . Mappability of the short sequencing reads from next generation sequencing techniques can be reduced within LCRs , and as a result multiple experimental molecular and computational approaches are often required to characterize SVs relative to the human haploid reference in a given personal genome . Several efforts have demonstrated the value of thoroughly scrutinizing complex genomic regions to better understand the human genome and discern variation that may be important to health , evolution , and susceptibility to diseases [27–33] . The human chromosomal region 2q13-2q14 . 1 represents the product of head-to-head fusion of two ancestral chromosomes forming human chromosome 2 [34] . This evolutionary fusion event is unique to the human genome , and is responsible for the chromosome number difference ( 46 versus 48 ) between human and the great apes including chimpanzee ( Pan troglodytes ) , gorilla ( Gorilla gorilla ) and orangutan ( Pongo abelii ) . The fusion of two subtelomeric regions from two ancestral chromosomes ( analogous chromosomes 2A and 2B in the great apes ) introduced substantial complexity to this region . A common recurrent 290 kb deletion encompassing Nephrocystin-1 ( NPHP1 , MIM *607100 ) , a gene that maps to the centromeric portion of the human 2q13 region , is associated with several diseases . Juvenile-onset nephronophthisis 1 ( NPHP1 , MIM #256100 ) is an autosomal recessive cystic kidney disorder causing chronic renal failure in children . Homozygous NPHP1 deletion is found in ~80% of patients born to consanguineous parents and in ~60% of sporadic cases [35] . In addition to nephronophthisis 1 , the same NPHP1 deletion has also been identified in patients with Senior-Loken syndrome-1 ( SLSN1 , MIM# 266900 ) and Joubert syndrome 4 ( JBTS4 , MIM# 609583 ) with distinct phenotypes [35–37] . Moreover , a recent study demonstrates that heterozygous NPHP1 deletion CNV in combination with NPHP1 point mutations ( SNVs ) can lead to Bardet-Biedl syndrome ( BBS , MIM# 209900 ) [38] . The NPHP1 deletion is recurrent , and results from NAHR-mediated unequal crossing-over involving the directly oriented flanking LCRs [39]; the frequency of heterozygous NPHP1 deletion is estimated to be approximately 1/400 in normal individuals from northern European descent [38] . Dittwald et al explored a clinical database containing chromosomal microarray ( CMA ) data from 25 , 144 patients , of which NPHP1 duplications ( N = 233 ) and deletions ( N = 118 ) were found to be the most commonly observed copy number aberrations ( combined ~1 . 4% ) compared to CNVs from other loci [5] . The complex genomic architecture of the human 2q13 region , especially the NPHP1 locus , provides the foundation for different SVs that may be observed in personal genomes among human populations . The polymorphic nature of this locus was previously demonstrated , and its evolution , prevalence and potential impact to disease susceptibility warrant further investigation . In this study , we utilized a combination of genomic technologies , including array-comparative genomic hybridization ( aCGH ) and optical mapping ( OM ) , to identify novel SV haplotypes at the NPHP1 locus and clarify the relative frequencies of specific haplotypes in the human population . We further utilized comparative sequence alignments of primate genome sequences and aCGH to construct a model for the evolution of the genomic architecture at the NPHP1 locus in nonhuman primates and human genomes . Unexpectedly , we found that this region displays evidence for incomplete lineage sorting , such that the structure of this region in humans is more similar to that of gorillas than to the orthologous region in chimpanzees or orangutans . The results also confirmed a dramatic genomic expansion of the NPHP1 locus during primate evolution and revealed a pattern of LCR evolution that may be explained by a model of multi-step , serial segmental duplication [32] .
The exceedingly complex and polymorphic genomic architecture of the NPHP1 locus presented difficulty during the assembly of the haploid human genome reference . This becomes readily apparent by the comparison between two recent updates of the human reference assembly , GRCh37/hg19 and NCBI36/hg18 , at the NPHP1 locus . The 800 kb sequences flanking each end of NPHP1 in hg18 and hg19 were compared by Miropeats [40] . Two ( Gaps II and III ) of the three major gaps ( Gaps I , II and III ) in hg18 are closed in hg19; Gap II corresponds to a region encompassing an ~45 kb LCR distal to NPHP1 that is only present in hg19 ( Fig 1A ) . We delineated the LCR structure of the NPHP1 locus ( hg19 , chr2: 110 , 080 , 914–111 , 762 , 639 ) using data from the UCSC Genome Browser “Segmental Dups” track ( http://genome . ucsc . edu/index . html ) . Two LCRs of approximately 358 kb in length , which we termed 358PROX ( centromeric ) and 358DIST ( telomeric ) , flanked NPHP1 in an inverted orientation–this can also be manifested by the Miropeats ( Fig 1A ) . Similarly , two LCRs of approximately 45 kb in length , termed 45PROX ( centromeric ) and 45DIST ( telomeric ) , were embedded within 358PROX and 358DIST , and thus had an inverted orientation . One additional LCR paralogous to 45PROX and 45DIST , termed 45MID , was revealed in build hg19 after the closure of Gap II from hg18 , resulting in a total of three copies of the 45 kb LCRs in the haploid reference . The 45MID and 45PROX , flanking NPHP1 in direct orientation , are presumably the LCR pairs responsible for the NAHR-mediated NPHP1 deletion ( Fig 1A ) . Self-alignments of the DNA sequences ( chr2: 110 , 080 , 914–111 , 762 , 639 ) using NCBI BLAST tool ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) confirmed the genomic structure of the NPHP1 locus in the reference by revealing the two major groups of LCRs and their relative orientations ( S1A Fig ) . We computationally characterized the LCRs at the NPHP1 locus using the human reference sequences . Pairwise alignments of both groups of 45 kb LCRs and 358 kb LCRs against their individual consensus sequences revealed a high percentage of sequence identities for both the 358 kb LCRs and 45 kb LCRs ( Table 1 ) . PR domain-containing protein 9 ( PRDM9 ) recognizes a degenerative 13-mer motif ( 5’-CCNCCNTNNCCNC-3’ ) that is critical to recruit recombination machinery required for crossovers in at least 40% of all human homologous recombination hot spots [4 , 41–44] . NAHR crossover studies suggest that the frequency of the PRDM9 hotspot motifs within LCR regions is one of the parameters correlated with the rate of NAHR mediated genomic rearrangement [5 , 17] . Characterization of the PRDM9 hot spot motif in the LCRs of the NPHP1 locus may elucidate potential crossover sites . At the NPHP1 locus , 12 motifs were found in each of the three 45 kb LCRs , while 161 and 154 hotspot motifs were found in 358PROX and 358DIST , respectively ( Table 1 ) . LCRs contribute to the complex genomic architecture of this region , and could incite genomic instability . The high degree of sequence identity between LCR pairs and the density of PRDM9 hotspot motifs ( Table 1 , S1B Fig ) may additionally contribute to the instability and increase the recurrent rearrangement frequency at the NPHP1 locus . The high similarity ( >99 . 6% ) between the corresponding paralogous LCRs in humans also indicates that gene conversion may occur frequently at the NPHP1 region [45] . NAHR events between directly oriented LCRs generate deletions or duplications; while NAHR events between inverted-oriented LCRs lead to inversions–such copy number neutral SVs may impose weaker selection forces than deletions and duplications do , and are thus more likely to be found as population polymorphisms [27–29] . In fact , the SV haplotype at the NPHP1 locus identified in the reference genome , arbitrarily designated as the H1 SV haplotype , is not the only SV haplotype in the human population . Experimental evidence suggested the presence of at least three alternative SV haplotypes ( H2 , H3 and H4 , Fig 1 ) before the first draft of the human genome assembly [39] . These alternative SV haplotypes share an inversion of the NPHP1 region , encompassing NPHP1 and 45MID , between 358PROX and 358DIST ( NPHP1 inversion ) . Besides the NPHP1 inversion , H2 and H4 appear to have several other SVs unique to themselves – 45PROX was lost in H2 , while both 45PROX and 45DIST were lost in H4 ( Fig 1B ) . We hypothesized that the majority of the structural polymorphism at the NPHP1 locus can derive from ( 1 ) the NPHP1 inversion , ( 2 ) the copy number loss of one or more of the 45 kb LCRs , or ( 3 ) a combination of these two events . To obtain further evidence to support these SV haplotypes and their prevalence across human populations , we examined the fosmid libraries from the Human Genome Structural Variation project ( HGSV , http://humanparalogy . gs . washington . edu/structuralvariation/ ) to search for individual discordant fosmids representing SVs [12] . We identified 78 discordant fosmids representing losses of either 45PROX or 45DIST and 31 discordant clones representing NPHP1 inversions between 358PROX and 358DIST in a total number of 17 individuals ( Fig 2A ) . Each of the 17 individuals had at least one discordant fosmid indicating loss of the 45 kb LCR , while 13/17 had at least one discordant fosmid indicating the NPHP1 inversion , suggesting that the current human reference genome actually presents a minor SV allele . We further examined the copy number distribution of the 45 kb LCRs utilizing the dataset published by Conrad et al [2] . CNVs in 450 individuals from different ethnicity groups were genotyped . These individuals include 180 CEU ( Utah residents with ancestry from northern and western Europe ) , 180 YRI ( Yoruba in Ibadan , Nigeria ) , 45 JPT ( Japanese in Tokyo , Japan ) and 45 CHB ( Han Chinese in Beijing , China ) . Various copy numbers of the 45 kb LCRs , ranging from two to six , were observed at different frequencies in each population ( Fig 2B ) . The distributions of copy numbers across different populations were not significantly different ( Kruskal-Wallis rank sum test , p-value = 0 . 6766 , Fig 2C ) . Aggregating all the populations , 1% , 13% , 56% , 25% and 5% of the entire examined population has two , three , four , five and six copies of the 45 kb LCRs , respectively ( Fig 2B ) . The frequencies of copy number derived from Conrad et al also correlated well with those derived from PFGE experimental data by Saunier et al [39] , in which 13% , 21% and 1 . 3% of 152 control individuals from an undefined ethnicity were found to harbor three , five and six copies of 45 kb LCRs , respectively ( Fig 2B ) . It is likely that the most common copy number of the 45 kb LCRs in a diploid genome is four , which deviates from the copy number of six that would exist in an individual with homozygous H1 , as in the haploid reference genome . Thus the six-copy state may be a minor genotype that was represented in only 1 . 3%-5% of the general population . The observation of polymorphic structural variants , including copy number polymorphisms of the 45 kb LCRs and NPHP1 inversion , prompted us to search for novel SV haplotypes . OM constructs ordered restriction maps ( Rmaps ) from single-molecules of DNA , which are assembled into genome-wide contigs that can be compared to an in silico restriction map from the human reference in order to discern SVs [46] . The OM can be used as an independent validation method for SVs revealed by other methods , such as fosmid sequencing data , which suggested SVs including the NPHP1 inversion and loss of the 45 kb LCRs in HapMap individual NA15510 ( Fig 3A ) . The SwaI OM contig assembly of NA15510 and its alignment to an in silico human reference ( hg19 ) further validated a homozygous H2 SV haplotype at the interrogated locus ( chr2: 109 , 943 , 987–111 , 547 , 676 ) , with four copies of the 45 kb LCRs and homozygous NPHP1 inversion ( Fig 3B ) . Additionally , NA10860 and NA18994 yielded OM results supporting genotypes identical to NA15510 ( Fig 3B ) . Fluorescence in situ hybridization ( FISH ) was performed in an attempt to delineate the organization of NPHP1 and the 45 kb LCRs at the NPHP1 locus . FISH experiments were designed with fosmid probes independently targeting NPHP1 ( fluoresces green , G ) and the 45 kb LCRs ( fluoresces red , R ) . A fosmid probe targeting ~1 . 3 Mb proximal to NPHP1 was used as an “anchor probe” ( fluoresces blue , B , S2A Fig ) . In the majority of interphase cells from the lymphoblastoid cell line of NA15510 ( 42/50 ) , we observed resolved signals of four red and two green ( R4G2 ) , which represented four copies of the 45 kb LCR and two copies of NPHP1 in a diploid genome ( S2B Fig ) . This result was consistent with the copy number data from OM . Additionally , a signal pattern of red-green-red , representing an SV haplotype also consistent with the OM data , was observed in the interphase cells ( S2B Fig ) ; however , such organization of signals was confounded by a signal pattern of yellow-red , the yellow of which was likely to represent an overlapping signal between red and green due to a two dimensional representation of a three dimensional reality and the close physical proximity , or overlapping in the z–plane , of a red and green signal . Moreover , using the blue “anchor probe” as a third color , we observed a signal pattern of blue-red-green-red in 25/50 interphase cells examined . This experimental result was also consistent with the OM data observed for this sample ( S2C Fig ) . However , such a pattern was not uniformly and consistently found ( S2C Fig ) . In aggregate , these results may be explained by the close proximity of the components being targeted , the three dimensional relative spatial positioning , and the less organized structure of chromosomes in interphase cells . We then performed custom-designed aCGH ( S3A Fig ) to validate the copy number estimations from OM . We used the DNA sample from NA10851 , a genome that has four copies of the 45 kb LCRs , as the universal reference for aCGH experiments . As a proof of principle , aCGH comparing six DNA samples ( NA18517 , NA15510 , NA18994 , NA10860 , NA18555 and NA12878 ) with NA10851 confirmed the copy number of the 45 kb LCRs estimated by the Conrad et al study ( Fig 3C , S3B Fig ) . These samples included the three samples ( NA15510 , NA18994 , NA10860 ) interrogated by OM . The consistency of copy numbers predicted by Conrad et al , aCGH , OM and the corroboration of the independent experimental assays including FISH further substantiated the accuracy of OM analysis for discerning CNV/SVs . A total number of eight DNA samples were analyzed by OM , and the contig assemblies at the NPHP1 locus revealed different SV haplotypes ( Table 2 ) . In addition to NA10860/NA18994/NA15510 , MM52 ( a multiple myeloma primary tumor sample ) [47] and HF087 ( an oligodendroglioma primary sample ) [48] were also found to be homozygous for the H2 SV haplotype by OM ( Table 2 ) . In the DNA from CHM ( complete hydatidiform mole , CHM1h-TERT ) [46] , the OM consensus map showed an allele with the loss of the 45PROX but without the NPHP1 inversion , presenting a potential novel SV haplotype ( termed H5 , Fig 3E and 3F , Table 2 ) . CHM is derived from fertilization of an enucleated egg with a single sperm [29] , and the haploid nature of the CHM genome facilitates accurate mapping and assembly by eliminating allelic variations from the diploid genomes . The OM analysis of H1-ES-P208 ( human embryonic stem cell line , passage 208 ) revealed alleles with three different SV haplotypes , including one H5 SV haplotype and two other novel SV haplotypes , one with the loss of the 45DIST and the NPHP1 inversion ( termed H6 ) and the other one with losses of both 45PROX and 45DIST without the NPHP1 inversion ( termed H7 ) ( Fig 3G and 3H , Table 2 ) . It is interesting that three different SV haplotypes in H1-ES-P208 were identified by OM , suggesting potential mosaicism . An NAHR-mediated inversion could potentially occur between the inverted 358 kb LCRs to convert H5 to H6 or vice versa . The highly identical paralogous LCRs may facilitate this rearrangement during the additional mitoses from the 208 cell culture passages of H1-ES-P208 [49 , 50] . aCGH analysis of H1-ES-P208 supported the copy number of three in its diploid genome as inferred by the OM analysis . Thus it led to the possibility that an admixture of cells with H5/H7 or H6/H7 combinations , both of which represent three copies of the 45 kb LCRs , could be present in the H1-ES-P208 cell line ( Table 2 ) . This finding should be further validated using orthogonal approaches that may delineate SV haplotypes . Unfortunately , the H1-ES-P208 cell line is no longer available . The DNA sample from HCC1937 ( a lymphoblastoid cell line from primary ductal carcinoma ) revealed two different SV haplotypes , H2 and H5 . These results emphasize the structural complexity of the NPHP1 locus and indicate that two ( or potentially sometimes more reflecting mosaic states ) SV haplotypes may be observed in the genome of one individual . However , although novel SV haplotypes are identified based on data from the aforementioned cell lines , it remains to be examined how representative they are of the different human populations worldwide . The 45 kb LCRs are responsible for the NAHR-mediated recurrent NPHP1 deletion . We have shown above that the copy number of this LCR is highly dynamic in the human population . To better understand the origin of this complexity and assess homologous genomic regions in closely related species , we investigated this region in great apes ( chimpanzees , gorillas and orangutans ) and Old World monkeys ( rhesus macaques [Macaca mulatta] and baboons [Papio anubis] ) . Analysis of the evolutionary history of this interval may represent a unique opportunity to characterize the emergence of a repeat sequence that causes susceptibility to a specific disease in humans . We were able to trace the human 45 kb LCR locus back to its ancestral origin by comparing several nonhuman primate genomes with the SV haplotypes observed in humans . Based on the OM data , the deletion of the 45 kb LCR resulted in an ~40 kb loss in the human subjects compared to the human reference , with an ~6 kb mismatching sequence remaining at the place of the 45 kb LCR loss ( Fig 3B ) . Alignments of human discordant fosmid clones with fully sequenced inserts , mapped to either the proximal or the distal side of human NPHP1 ( hg19 ) , revealed a shared “deletion/insertion” haplotype with the 45 kb LCR deletion and a 5936 bp insertion ( 5936Ins ) at the deletion breakpoint junction ( Figs 4A and 3B , S4 Fig ) . The 5936 bp stuffer sequence could not be uniquely mapped to any position of the human genome reference hg19 using BLAT ( http://genome . ucsc . edu/cgi-bin/hgBlat ) . We further investigated the origin of the 5936Ins using BLAST ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) with the Nucleotide collection ( nr/nt ) database . Interestingly , in addition to the five fully sequenced clones previously found in the human fosmid libraries , we found two chimpanzee BAC clones , CH251-328D3 and CH251-71L9 , which encompassed a sequence highly identical ( 98% ) to the human 5936Ins . The two clones mapped to a region in the chimpanzee chromosome 2A –a position syntenic to the human NPHP1 locus . This finding indicated that the 5936Ins that was present at some but not all human NPHP1 loci was also present in at least some chimpanzee genomes . We subsequently used BLAT to search for this sequence in the genomes of baboon , rhesus macaque , orangutan , gorilla and chimpanzee . This 5936Ins could be aligned to the reference genomes of all these species with increasing sequence identities , meaning it may exist as an ortholog in Old World monkeys and great apes ( Table 3 ) . From ~85 million years ago ( Mya ) to the present , primate genomes have undergone substantial sequence change during evolution . Genomic segmental duplications have been one significant aspect of that process [21 , 51 , 52] . The human 5936Ins that is conserved among baboons , rhesus macaques and great apes motivated us to use comparative genomic analyses to understand the dynamic structural changes that occurred at the NPHP1 locus during the evolution of the human genome . DNA sequences of 800 kb flanking each side of NPHP1 were downloaded from UCSC Genome Browser ( http://genome . ucsc . edu/index . html ) for baboon , rhesus macaque , orangutan , gorilla , chimpanzee and human . According to our previous analysis , a minor allele is presented at the NPHP1 locus in the hg19 . As a result , we modified the human reference sequence so that it reflected the more common H2 SV haplotype ( Table 2 ) . We manually constructed the H2 SV haplotype sequences by: ( 1 ) deleting the sequences of 45PROX , ( 2 ) inserting the 5936Ins and ( 3 ) inverting the sequences between 358PROX and 358DIST ( Fig 5A–5F ) . Miropeats [40] was subsequently used to perform local alignments between the human H2 and each individual primate reference genome . Alignment between human H1 and H2 revealed the gain of the 45PROX in H1 , the overall sequence similarity , and the inverted orientation of the paralogous 358 kb LCRs ( Fig 5A ) . Old World monkeys , including baboons and rhesus macaques , diverged from the ancestors of humans about 25–33 million years ago ( Mya ) and share 94 . 9% sequence identity with the human genome [53] . Due to lower sequence identity between baboon/macaque and human , the Miropeats threshold was lowered to display the alignments between more distant species; this change in the threshold correlated with more noise in the alignments . The diagram of alignments suggested that sequences homologous to the human 45 kb LCRs do not exist in the current baboon/macaque reference genomes , as the only traces between these references and human H2 were sparse , and likely indicated noisy alignments ( Fig 5E and 5F ) . A region on the left side of both baboon and rhesus macaque NPHP1 could be aligned to a portion of both human 358 kb LCRs , indicating that a smaller region orthologous to part of the human 358 kb LCRs exists in both of the baboon and macaque references , but it appears as only a single copy per haploid genome ( Fig 5E and 5F ) . In both baboon and rhesus macaque , the intra-species Miropeats alignments also revealed a lack of the pattern of paralogous LCRs , which was observed in the human reference genome ( S5E and S5F Fig ) . The great ape lineages including orangutans , gorillas and chimpanzees diverged from the human evolutionary lineage about 12–16 Mya , 6–8 Mya and 4 . 5–6 Mya , respectively , with increasing sequence identity to the human genome ( S1 Table ) [53] . Miropeats revealed similar patterns of alignment of orangutan and chimpanzee’s references versus human H2: two paralogous 45 kb LCRs in the human reference were either directly or invertedly aligned to a single genomic region in the references of orangutan and chimpanzee; while the two 358 kb LCRs in the human reference were partially aligned to a single genomic region on the left side of NPHP1 in the orangutan and chimpanzee references , with chimpanzee appearing to have a larger partial alignment ( Fig 5B and 5D ) . Although the 45 kb and 358 kb LCRs in human could be aligned to a region in orangutan and chimpanzee references with longer sequence homology , these genomic segments were lacking paralogous LCR partners . Consistently , intra-species alignments by Miropeats did not reveal any pattern of paralogous LCRs in their reference genomes ( S5B and S5D Fig ) . Gorilla is a more distant species from human than chimpanzee according to the standard evolutionary phylogeny; however , it had a Miropeats pattern more consistent with that of a human-human alignment . Each of the two human 45 kb LCRs in H2 could be apparently aligned to three genomic regions surrounding NPHP1 in the gorilla reference , suggesting an SV haplotype in gorilla that is orthologous but with lower sequence homology ( Fig 5C ) . Moreover , each of the two human 358 kb LCRs could also be aligned to the flanking sequences of gorilla NPHP1 at the paralogous positions , suggestive of a potential orthologous locus in gorilla ( Fig 5C ) . Intra-species alignment of the gorilla reference genome to itself revealed paralogous alignment of long sequences flanking NPHP1 , suggesting that a genomic architecture similar to the one observed in humans at the NPHP1 locus at least in the gorilla reference ( S5C Fig ) . Miropeats also revealed the conservation of the human 5936Ins in the nonhuman primate genomes . As shown by Miropeats , the 5936Ins observed in the human H2 was also identified in the reference genomes of baboon , rhesus macaque , orangutan and chimpanzee ( Fig 5B , 5D–5F ) . Interestingly , the gorilla reference contained two regions similar to the human 5936Ins that were adjacent to two gorilla 45 kb LCR orthologs ( Fig 5C ) . The relative transitions of the human 45 kb LCRs and 5936Ins in different nonhuman primates derived from Miropeats were confirmed by BLAT using the sequences of the human 45MID and the 5936Ins as templates ( Table 3 ) . Since Miropeats was applied to reference genomes to reveal similarities and differences between humans and nonhuman primates , the comparison results may not represent the general populations of queried species . Orthogonal experimental approaches , including aCGH and copy number analysis of genomic sequencing data generated from multiple primate individuals , were used to further investigate the copy number changes in the nonhuman primates comparing to humans . Interphase FISH experiments were performed on lymphoblastoid cell lines of one chimpanzee ( CRL-1868 ) and one gorilla ( CRL-1854 ) using the human fosmid probes described above to explore the genomic architecture at the NPHP1 locus in each species . The experiment illustrated the 45 kb LCR orthologs in both species tested . The majority of the scored gorilla interphase cells ( 46/50 ) and all of the scored chimpanzee interphase cells ( 50/50 ) showed R2G2 , suggesting one copy of the 45 kb LCR ortholog and one copy of NPHP1 on the haploid genome of each individual ( S2D and S2E Fig ) . Interestingly , resolved signals of R4G2 was observed in a minority population ( 4/50 ) of gorilla interphase cells , indicating a potential two-copy configuration of the 45 kb LCR ortholog in the gorilla haploid genome . Inter-species aCGH were performed to further validate the copy number alterations indicated by sequence alignments ( S6A Fig ) . Genomic DNAs from baboon ( N = 1 ) , rhesus macaque ( N = 2 ) , orangutan ( N = 1 ) , gorilla ( N = 3 ) and chimpanzee ( N = 7 ) were used to compare with human genomic DNA ( NA10851 ) on the previously described aCGH . The quality of the hybridization positively correlated with the sequence identities between different primates and human ( S1 Table ) . Comparing to the human genome , a large portion of genomic sequences flanking NPHP1 appeared to be nonexistent or have lower copy number in nonhuman primates , and the degree of similarity to human varied from baboon to chimpanzee ( S6 Fig ) . After careful examination of the aCGH data of the 358 kb LCR locus , we achieved an estimation of variation of the genomic content between nonhuman primates and human . Sequences totaling sizes of 146 kb ( 40 . 85% ) , 156 kb ( 43 . 6% ) , 88 kb ( 24 . 6% ) , 24 kb ( 6 . 7% ) and 72 kb ( 20 . 1% ) appeared to be nonexistent ( aCGH log2 ratio was lower than -1 ) in the genomes of baboon , rhesus macaque , orangutan , gorilla and chimpanzee , respectively , while approximately 123 kb ( 34 . 4% ) , 93 kb ( 26 . 0% ) , 115 kb ( 32 . 1% ) , 156 kb ( 43 . 6% ) and 197 kb ( 55 . 0% ) appeared to exist in the genomes of these primates , albeit at lower copy number ( aCGH log2 ratio was between 0 and -1 ) . The latter observation indicated that these genomic regions might constitute the ancestral nonduplicated segments orthologous to the human LCRs . The indication of copy number variants is derived from the log2 ratio of aCGH probes , which largely relies on the degree of hybridization based on sequence similarity . Thus , these data can reflect genome differences and phylogenic distance between the two species being compared . The sizes of genomic segments with copy number changes described above are estimates , and may be refined by testing a large cohort of nonhuman primates . Since the 45 kb LCRs in human are directly related to the recurrent NPHP1 deletion , and a haplotype consisting of a 45 kb LCR loss accompanied with a 5936Ins is frequently observed , we were interested in understanding the evolution of this haplotype . As previously shown , the copy number of the 45 kb LCRs ranged from two to six in humans without NPHP1 deletion . Moreover , aCGH of eight DNA samples from patients with homozygous recurrent NPHP1 deletion revealed two copies of the 45 kb LCRs in seven individuals and four copies in one individual ( Table 4 ) . NAHR-mediated NPHP1 deletion between a pair of directly-oriented 45 kb LCRs reduces one copy , creating a recombinant copy from two substrate copies , of the 45 kb LCRs after the deletion event . Thus the majority ( 14/16 ) of haplotypes examined prior to NPHP1 deletion would have two copies of the 45 kb LCRs–this is consistent with the observation that copy number of two is the most frequently observed copy number in a haploid human genome . Copy numbers of the 45 kb LCR orthologs in nonhuman primates were also estimated by aCGH using DNA from NA10851 as reference . In the tested baboon ( N = 1 ) and rhesus macaques ( N = 2 ) , the species average log2 ratios targeting the human 45 kb LCRs were consistent with a complete absence of the 45 kb LCR ortholog ( Fig 6B ) . In the tested orangutan ( N = 1 ) and chimpanzees ( N = 7 ) , the species average log2 ratios indicated a reduced copy number of the 45 kb LCR orthologs in comparison to the reference human DNA used in aCGH ( Fig 6C ) . In the tested gorillas ( N = 3 ) , the species average log2 ratio was close to zero , indicating that the copy number of the human 45 kb LCRs equaled the copy number of the gorilla orthologs ( Fig 6C ) . Mean log2 ratios of the 45 kb LCRs or orthologous region for each individuals tested , including the hybridizations between human DNA samples without NPHP1 deletion ( N = 10 ) , are shown in Table 4 . Similar copy number of LCRs at the human and gorilla NPHP1 locus could be independently validated using existing and larger datasets . Dumas et al performed cDNA aCGH to survey genome-wide gene CNVs across a number of primate lineages [21] . We investigated CNVs at the NPHP1 locus using their published dataset , and correlated them with the CNV findings from our comparative analysis . Examination of two data points of human cDNAs ( AA937147 and AI820499 ) located at the human 45 kb LCR locus revealed the relative copy numbers examined in each individual comparing to the human reference . Consistently , the gorilla genomes showed roughly comparable signal intensity compared to the human genomes , while the chimpanzee and orangutan genomes presented lower signal intensity for these two data points ( S7 Fig ) . Although both AA937147 and AI820499 in rhesus macaque and baboon presented lower signal intensity than human , large variations were observed between these two data points ( S7 Fig ) . These latter findings in rhesus macaque and baboon suggest that AA937147 and AI820499 , although located at the same LCR locus , might have different copy numbers inside each species . Alternatively , the large variation could also be due to the essential absence of the 45 kb LCR ortholog in the macaque and baboon genomes . Moreover , Sudmant et al analyzed read-depth profiles from whole genome sequencing ( WGS with a median coverage of ~25× ) data of 10 humans , 32 gorillas , 23 chimpanzees and 17 orangutans . Copy number analysis based on sequencing read-depth revealed a similar copy number of the 45 kb LCRs and their orthologous regions in the tested humans and gorillas ( p-value = 0 . 1705 , Welch Two Sample t-test ) . Moreover , the copy number of the 45 kb LCRs in human is higher when comparing to the orthologous region in the tested chimpanzees ( p-value < 0 . 0003 , Welch Two Sample t-test ) and orangutans ( p-value < 0 . 0003 , Welch Two Sample t-test ) ( Fig 6D and S7 Fig ) . These aggregated data provided evidence to partially support the observation of the similar genomic architectures between humans and gorillas .
Combined OM and aCGH approaches appear to be a versatile route for delineating SV haplotypes in a structurally complex locus like NPHP1 . Array CGH can detect CNVs as small as a few hundred base pairs in size in test samples when compared to a reference . However , balanced SVs , e . g . inversions , cannot be detected by aCGH . Unlike aCGH , OM creates large datasets of ordered Rmaps from individual genomic DNA molecules , which through analysis reveal genome structures . Alignments between an optical map from test samples and the in silico generated reference maps reveal both CNVs and copy number neutral inversions and translocations . Errors associated with enzymatic cleavage or fragment sizing are inevitable , but can be modeled and dealt with by algorithms and software designed to work with large Rmap datasets for the construction of accurate maps [55–57] . Importantly , the CNVs were called consistently by both OM and aCGH in the samples tested in this study ( Table 2 ) . These data suggest that OM and aCGH , two orthogonal genomic approaches for SV characterization , can complement each other to provide a comprehensive and accurate SV haplotype . Furthermore , sequencing technologies may greatly facilitate the delineation of SV haplotypes in a region with complex genomic structures . The highly complex and repeating nature of genomic regions enriched with LCRs can challenge short read sequencing approaches and introduce mapping artifacts , some of these limitations may be potentially overcome by single-molecule long-read sequencing technologies , such as single-molecule real-time ( SMRT ) sequencing [58] . The evolutionary history of structural changes in a complex region such as 2q13 can be reconstructed by appropriate comparisons among related populations and species . The formation of human chromosome 2 through a telomeric fusion of chromosomes 2A and 2B was originally documented using high-resolution G-banding technique [34] . Later , using cosmid sequencing , two inverted arrays of telomeric repeats ( 5’ ( TTAGGG ) n3’ ) in a head-to-head orientation ( 5’ ( TTAGGG ) n- ( CCCTAA ) m3’ ) were found at the 2q fusion breakpoint ( 2qFus ) [59] . In our study , we analyzed the origin of the SV haplotypes of the NPHP1 locus , which is about 3 Mb proximal to 2qFus . The comparative genomic analyses among nonhuman primates and humans suggest a trend of genomic expansion at the NPHP1 locus during primate evolution . Aggregating the sequence alignments and copy number analyses using aCGH and WGS data of baboon , rhesus macaque , orangutan , gorilla , chimpanzee and human , we propose the following model . First , the present-day human 45 kb and 358 kb LCRs may be formed by gradual expansion and propagation of primate orthologous sequences into paralogous regions early in the evolution of the apes , after they diverged from the Old World monkeys . The nonhuman primate orthologs of the human 45 kb LCRs , not found in the current baboon and rhesus macaque reference sequences , emerge prior to the divergence of orangutan and human as they are found in both lineages ( 45MID , Fig 7 ) . This 45kb sequence expanded in the lineage leading to chimpanzees and exhibit increasing sequence identity to the human 45 kb LCRs . These sequences then propagated into paralogous regions and eventually formed the 45 kb LCRs now observed in the human genome . A similar unknown mechanism may be suitable for explaining the expansion and paralogous propagation of the nonhuman primate orthologs of the human 358 kb LCRs ( Fig 7 ) . In the human lineage , the order of appearance of the 45 kb LCRs may be inferred by molecular clock analysis based on reference DNA sequence comparisons excluding insertion/deletion ( indel ) events . Comparative analysis suggests that 45MID in the human haploid reference is the ancestral copy among the three copies . The sequence identity between 45MID and 45PROX or 45DIST is slightly lower than that between 45PROX and 45DIST using the Smith-Waterman local alignment algorithm ( Table 1 ) . This suggests that 45DIST and 45PROX are paralogous propagation products of 45MID . The observation of the 45MID ortholog location adjacent to NPHP1 in both chimpanzee and orangutan supports the contention that the 45MID in human is the ancestral copy . However , the sequence-based molecular clock analysis is based on a priori hypothesis that the paralogous LCRs evolved independently . Therefore , it may be confounded by the potential interactions between paralogous sequences , such as homologous recombination leading to gene conversion , considering the fact that the 45PROX and 45MID are imbedded within the two highly homologous and inverted 358 kb LCRs . Secondly , the human 45 kb LCRs imbedded in the 358 kb LCRs may be formed by paralogous insertion of the duplicated segments at the site where the 5936Ins were lost . This model is supported by the identification of the human 5936Ins orthologous sequences in all the primate reference genomes with increasing identities . The 5936Ins orthologs are present in all lineages studied , whereas the 45 kb LCR orthologs are only present in the orangutan , gorilla and chimpanzee genomes ( Fig 7 ) . Thus , the haplotype in human with the 5936Ins embedded in the 358 kb LCRs is the likely ancestral haplotype , whereas the haplotype possessing the imbedded 45 kb LCRs in the 358 kb LCRs may be more recent and could be formed by the deletion of the 5936Ins followed by a 45 kb LCR insertion . Therefore , the haplotype without 45PROX and 45DIST may be the most ancestral haplotype in the human lineage . The structural complexity at the NPHP1 locus may also exist at a population level within the human species . We delineated various SV haplotypes in seven diploid human genomes and one haploid human genome using OM . These include five diploid genomes ( NA10860 , NA18994 , NA15510 , MM52 and HF087 ) that are homozygous for H2 ( H2/H2 ) , one ( H1-ES-P208 ) with a possible admixture of cells with H5/H7 and H6/H7 combinations , one ( HCC1937 ) with H2/H5 , and one haploid genome ( CHM ) with only H5 . Within15 chromosomes/SV haplotypes , NPHP1 inversions are found in at least 11 , which account for 73 . 3% of the total alleles . Moreover , cross-correlation between aCGH ( testing copy number of 45 kb LCRs ) and OM ( testing NPHP1 inversion ) shows that none of the genomes investigated above harbor the haploid reference allele ( Table 2 ) . On a population basis , 11% of YRI , 3% of CHB/JPT and 1% of CEU harbor 6 copies of the 45 kb LCRs , thus possess at most two reference alleles ( homozygous ) ; 31% of YRI , 31% of CHB/JPT and 16% of CEU harbor 5 copies of the 45 kb LCRs , thus possess at most one reference allele ( heterozygous ) in each examined individual . The aggregate experimental evidence above strongly suggests that the SV haplotype at the NPHP1 locus is highly variable , as has been observed for the complex LCRs at the iso17q susceptibility locus [33] . Furthermore , the allele represented in the human reference genome may actually be a minor allele , suggesting a potential necessity for improvement of the reference genome and again reiterating the limitations of the haploid human reference and the lack of representation of CNV/SV variant alleles . In the current study , we used OM to identify various SV haplotypes in seven diploid human genomes and one haploid human genome . However , haplotype analysis of the NPHP1 locus was not performed in a large general population or across ethnicities in either humans or nonhuman primates studied . Thus the frequency of each SV haplotype in the general human population cannot be estimated . Perhaps large-scale assays , with methods such as single-molecule and long-read sequencing , will benefit the further investigation of the complex NPHP1 locus . Highly identical LCRs at the NPHP1 locus lead to local genomic instability , which subsequently results in variable SV haplotypes . A new SV haplotype may be generated from an existing one by a simple inversion between two invertedly oriented LCRs , e . g . H3 from H1 ( Fig 1B ) . In some cases , depending on the breakpoint , an inversion may generate a novel SV haplotype that may change the CNV susceptibility of the disease-associated genes , and the resultant SV haplotype may have potential clinical significance [28 , 29] . At the NPHP1 locus , the two directly oriented 45 kb LCRs flanking NPHP1 are the substrates for the NAHR event resulting in the common recurrent deletion of the gene . However , intra-chromosomal NAHR-mediated deletions would be inhibited if a chromosome lacks the flanking directly oriented LCRs . Interestingly , in our study , we identified SV haplotypes that appear to be resistant to NAHR-mediated NPHP1 deletion as a result of the loss of either one or both of the directly-oriented 45 kb LCRs flanking NPHP1 . For example , OM analysis of the haploid genome of CHM reveals an H5 SV haplotype losing the 45 kb LCR on the centromeric side of NPHP1 which is utilized as a flanking substrate for unequal crossing-over ( Fig 3 ) . Therefore , these findings suggest that there may exist protective alleles that potentially inhibit intra-chromosomal NAHR , while the inter-chromosomal event may also be prevented if the protective allele exists in a homozygous state in an individual . These SV haplotypes could potentially reduce the frequency of the NPHP1 common recurrent deletion in the human population . This study works in concert with the previous studies regarding the correlation between local genomic structure and individual’s susceptibility to acquiring disease-associated alleles [28 , 29] . Both protective and susceptible SV haplotypes likely exist at other disease-associated loci with similar structural complexity [22] . The SV haplotypes ( H5 , H6 and H7 ) elucidated by OM were identified in human cell lines . Thus , they may potentially reflect tissue culture events generated in the cell lines tested . However , these results are parsimoniously explained by the underlying genomic architecture and mechanistic first principles , and thus the observed results in cell lines likely represent the organismal genome structure , reflect the genomic instability at this locus and indicate the potential existence of H5 , H6 , and H7 in the personal genomes of individuals in human populations . Nevertheless , it remains to be examined how representative these SV haplotypes are for human populations at large . The NPHP1 locus in the gorilla reference genome is an interesting example of evolutionary complexity at two levels: complexity of sequence structure and complexity of population-level evolutionary genetics . Although whole genome comparisons indicate that gorillas diverged from human ancestors before chimpanzees did [60] , the gorilla reference genome has a configuration of SV haplotype more similar to human than the chimpanzee does at the NPHP1 locus ( Fig 7 ) . Distinctly , instead of completely deleting the gorilla 5936Ins ortholog as observed in the human SV haplotypes , the gorilla 45 kb LCR orthologs appears to be imbedded inside of the gorilla 5936Ins ortholog . The coexistence of these orthologs indicates that the SV haplotype found in the gorilla reference may be an intermediate state different from the SV haplotypes in human . Alternatively , the observation of the human pattern of SV haplotype in gorilla could potentially reflect an assembly error due to the local complexity in the gorilla reference genome; this complexity involves both genomic gaps and the presence of the partial 45 kb LCR and 5936Ins orthologs within the 358 kb LCR orthologs . Analyses using genome-wide cDNA arrays [21] , the WGS read-depth analysis , and the genomic aCGH ( performed in this study ) confirm a comparable copy number of the 45 kb LCR and its ortholog between the human H2 SV haplotype and the gorilla genomes tested ( Fig 6 , S7 Fig , S8 Fig , Table 4 ) . Thus , the aCGH and sequence alignment data are suggestive of an analogous structural haplotype at the NPHP1 locus between gorilla and human that differs from that of chimpanzee ( S6 Fig ) . Copy number analyses of DNA samples from a large number of gorillas will provide additional evidence supporting this hypothesis . It is intriguing that the gorilla genomic structure for this region appears to be more similar to human than chimpanzee . The estimated divergence time of human-gorilla lineages is approximately 6–8 Mya , which is earlier than estimated human-chimpanzee divergence ( approximately 4 . 5–6 Mya ) . Moreover , the sequence identity of human versus chimpanzee ( 99% ) is slightly higher than that of human versus gorilla ( 98 . 4% ) [53 , 60] . The genome-wide evidence reflects the most commonly accepted evolutionary phylogeny that has humans more closely related to chimpanzees than to gorillas , i . e . ( ( H-C ) -G ) . However , a recent study of the gorilla genome shed light on the complex evolutionary phylogeny by providing compelling evidence that incomplete lineage sorting affects 15% of the human-chimpanzee-gorilla genomes [60] . That is , whole genome analyses demonstrate that 70% of the gorilla genome follows the ( ( H-C ) -G ) standard phylogeny , while 15% of the gorilla genome segments exhibit ( ( H-G ) -C ) whereas another 15% exhibit ( ( C-G ) -H ) . The structural variation results using aCGH and WGS read-depth analysis suggest that the region including NPHP1 and its adjacent LCRs fall in one of the segments with the alternative ( ( H-G ) -C ) pattern . It is plausible that the last common ancestor of humans , chimpanzees and gorillas was polymorphic for NPHP1 haplotypes , and segregating for haplotypes that resemble the human/gorilla structure and the chimpanzee structure . If this were true , the pattern of variation across the three species can be explained by the retention of one ancestral haplotype in humans and gorillas , and the loss of that haplotype with retention of the more ancestral form in chimpanzees [52 , 61] . Furthermore , these results lead to the prediction that gorillas will be more susceptible than other nonhuman primates to mutations that delete NPHP1 and thus cause a disease similar to NPHP1 . Our findings also suggest that the complex structure of the human NPHP1 region was established prior to the fusion of the two ancestral chromosomes that formed the present human chromosome 2 , as neither gorillas nor chimpanzees exhibit this fusion . In summary , we computationally and experimentally characterized the genomic architecture and identified novel SV haplotypes of the NPHP1 locus in the human 2q13 region . The more commonly observed alternative SV haplotypes suggest the current human genome reference represents a minor allele . For such complex loci enriched with LCRs , the accuracy of assembly may be compromised . Thus detailed exploration using various comparative genomic analytical methods is needed to document the human genome structure and the stages of its evolution in a more comprehensive way . NAHR-mediated NPHP1 deletion occurs between the two flanking directly oriented LCRs . Here , we found that potential “protective alleles” lacking directly oriented LCR flanking NPHP1 also exist in the examined human genomes , and such structure may protect those alleles from NPHP1 deletion mediated by intra-chromosomal NAHR , or inter-chromosomal events in the homozygous state . Such potential “protective alleles” may also exist in other “NAHR-susceptible” loci , with the drawback of such alleles being a decrease in genome plasticity that facilitates evolution . A large number of genomic disorders are associated with loci with complex genomic architectures that introduce risk of genomic rearrangements ( e . g . 17p11 . 2 and SMS/PTLS ) . The assessment of the disease risk of these disorders can be facilitated by accurate determination of the alternative SV haplotypes via single molecule analysis , including OM or long-molecule/long-read sequencing . Moreover , we assessed the origins of the complexity of the NPHP1 locus using inter-species comparative genomic analysis , and we found evidence supporting the genomic expansion and propagation of LCRs during primate evolution . The generation of LCRs may occur in a multi-step manner , and the higher order of genomic complexity constituted by LCRs may render the genome susceptible to instability and DNA rearrangements .
Sequences of LCRs were downloaded from UCSC Genome Browser ( http://genome . ucsc . edu/cgi-bin/hgTables ) . The coordinates used for sequence downloading are: chr2:110688766–110733137 ( 45PROX ) , chr2:110983705–111031088 ( 45MID ) , chr2:111153517–111197896 ( 45DIST ) , chr2:110494432–110852754 ( 358PROX ) and chr2:111033788–111392192 ( 358DIST ) . Pairwise alignments of the LCRs in the same group were performed using “pairwiseAlignment” R package [62] . It was performed as a type of local alignment that considers the penalty from end gaps . Sequence identity was calculated after the alignment . We aligned the LCR sequences using the Clustal W2 algorithm [63] and determined the percentage of identical positions over a 100 base pair window along the length of the LCR . We created a position weight matrix ( PWM ) based on a previously reported recombination hotspot motif [41] and subsequently assessed each LCR for matches to the motif’s PWM and its reverse complement using the Biostrings package implemented in the R Statistical Programing Language ( http://www . r-project . org/ , http://www . bioconductor . org/packages/release/bioc/html/Biostrings . html ) . We indicate the positions of strong ( >85% of the maximum possible score ) matches along the edge of each LCR with a triangle . Fosmid libraries of 17 individuals ( ABC7 , ABC8 , ABC9 , ABC10 , ABC11 , ABC12 , ABC13 , ABC14 , ABC16 , ABC18 , ABC21 , ABC22 , ABC23 , ABC24 , ABC27 , WIBR2 , JVI ) were downloaded from Human Genome Structural Variation Project ( HGSV , http://humanparalogy . gs . washington . edu/structuralvariation/ ) . All end sequence pairs ( ESPs ) mapped to hg19 build were manually filtered according to the mapping quality and chromosomal location . Discordant fosmids were selected based on the annotations judging the distance and orientation between the ESPs . A UCSC Genome Browser custom track was created for the discordance fosmids identified in NPHP1 locus based on genomic coordinates and relative orientations of the ESPs ( http://genome . ucsc . edu/cgi-bin/hgCustom ) . Miropeats program was used to descriptively illustrate the genomic architecture by plotting the inter-/intra-species alignments of the reference genome . ICAass ( v 2 . 5 ) algorithm was used to perform DNA sequence comparisons , and Miropeats ( v 2 . 01 ) was then applied for converting the comparisons into graphical display based on the position and matching quality ( a threshold set up by users ) [40] . According to the time of divergences and overall sequence similarities upon evolution , different thresholds gauging the length of DNA sequence homology ( “seed” ) were chosen in order to show the feature of alignments between different primates . In our study , a threshold of 500 were set for baboon/human , macaque/human and orangutan/human pairs , while 1000 were used for gorilla/human , chimpanzee/human and all the intra-species alignments . Miropeats were performed between sequences from genomic intervals +/-800 kb of NPHP1 of each genome build used . Optical mapping [46 , 48 , 55–57 , 65–71] is a single-molecule , whole-genome analysis system for the comprehensive discovery and characterization of SVs . Large genomic DNA molecules ( from 300 kb to multi-Mbs ) were extracted , stretched and immobilized on positively charged glass surfaces via capillary flow within microfluidic devices fabricated using soft lithography [65] . Hydrodynamic forces generated by capillary flow combine with DNA/surface electrostatic interactions to stretch and immobilize very long molecules . DNA molecules , thus presented , were restriction digested ( SwaI or BamHI , New England Biolabs ) , stained with YOYO-1 ( an intercalating fluorochrome; Invitrogen ) and imaged using a custom-designed , fully-automated , epifluorescence microscopy imaging system [65] . Restriction endonuclease sites undergo double-stranded breakage followed by DNA relaxation at the cut ends , which present as micron-sized gaps along stretched DNA molecules . Acquired images were then automatically analyzed using custom machine vision software [65 , 66] , which yielded large datasets of single molecule ordered Rmaps . Using an iterative assembly process that leverages Bayesian inference approaches and cluster computing [46 , 55–57 , 67 , 68] , the Rmaps datasets were then assembled into multimegabase map contigs that were later joined to span entire chromosomes . The iterative assembler clusters single-molecule maps using pairwise alignments to a reference genome , and then assembles these map clusters using a maximum-likelihood Bayesian assembler to generate contigs and consensus restriction maps . The assembled genomes were then viewed within a custom genome visualization environment ( Genspect ) that allows detailed inspection of the primary data underlying called SVs . Lastly , the assembly/analysis pipeline automatically tabulates a list of SVs that were inspected and manually curated in another custom visualization software ( GnomSpace ) to characterize all SVs in the analyzed genomes . Great ape CNV data derived from whole genome sequencing ( WGS ) were downloaded as bigBed files from http://eichlerlab . gs . washington . edu/greatape-cnv/tracks/ , and bigBedToBed was used to extract annotations within the region of interest . The human genome hg18 coordinates of the CNV annotations were converted to hg19 coordinates using liftOver with default parameters . A UCSC track hub was generated in order to visualize the CNVs lifted over to hg19 . The average number of copies for each sample ( great apes or human ) was calculated for each genomic window of 500 bp in the region of interest . Fosmids clones ( G248P81805G1 , G248P88963C6 , and G248P88660A3 ) were obtained from BACPAC Resources Center ( BPRC ) as “stab-cultures” . Clones were cultured in LB medium containing 12 . 5μg/ml chloramphenicol ) . Fosmids were extracted from a suspension culture with QIAGEN Plasmid Midi Kit . G248P81805G1 is from NPHP1; G248P88963C6 is from the 45 kb LCRs; and G248P88660A3 is from a conserved region ~1 . 3 Mb proximal to NPHP1 . Cultured lymphoblastoid cell lines from the individuals NA15510 ( human ) , CRL-1854 ( gorilla ) , and CRL-1868 ( chimpanzee ) were harvested using 10ug/ml Colcemid ( Roche ) for 30 minutes followed by 0 . 075M ( hypotonic ) treatment for 10 minutes at 37°C . The cells were then fixed using Carnoy’s fixative ( 3 methanol: 1glacial acetic acid ) . The cell pellet obtained from the harvesting was used to prepare the “dropped slides” for FISH . Directly labeled custom “home-brewed” probes were produced from fosmid clones mentioned above . The home brewing process was performed using Nick Translation Kit ( Abbott molecular ) . Green dUTP , Orange dUTP ( herein referred to as red ) and Aqua dUTP ( Abbott Molecular ) were used to label G248P81805G1 , G248P88963C6 and G248P88660A3 , respectively . Signal validation was also verified by observing the metaphases on an inverse DAPI function . The slides were aged using 2X SSC at room temperature for 2 minutes followed by sequential dehydration using 70% , 80% and 100% ethanol for 2 minutes each . Metaphases on slides were marked and 5ul of the probe mixture was added . Target DNA and the probes were co-denatured at 75°C for 5 minutes followed by hybridization at 37°C for 16 hours . The slides underwent post wash with 2X SSC at 37C for 2 minutes . The slides were left to air dry after post wash , and 10ul of DAPI-II counterstain ( Dako , Agilent Technologies ) was added to the slides . Metaphases were viewed using Olympus Florescence microscope ( Olympus America ) and analyzed using Cytovision software v3 . 6 . | Genomic instability due to the intrinsic sequence architecture of the genome , such as low copy repeats ( LCRs ) , is a major contributor to de novo mutations that can occur in the process of human genome evolution . LCRs can mediate genomic rearrangements associated with genomic disorders by acting as substrates for nonallelic homologous recombination . Juvenile-onset nephronophthisis 1 is the most frequent genetic cause of renal failure in children . An LCR-mediated , homozygous common recurrent deletion encompassing NPHP1 is found in the majority of affected subjects , while heterozygous deletion representing the nephronophthisis 1 recessive carrier state is frequently observed amongst world populations . Interestingly , the human NPHP1 locus is located proximal to the head-to-head fusion site of two ancestral chromosomes that occurred in the great apes , which resulted in a reduction of chromosome number from 48 in nonhuman primates to the current 46 in humans . In this study , we characterized and provided evidence for the diverse genomic architecture at the NPHP1 locus and potential structural variant haplotypes in the human population . Furthermore , our analyses of primate genomes shed light on the massive changes of genomic architecture at the human NPHP1 locus and delineated a model for the emergence of the LCRs during primate evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Comparative Genomic Analyses of the Human NPHP1 Locus Reveal Complex Genomic Architecture and Its Regional Evolution in Primates |
The morphogenesis of sex combs ( SCs ) , a male trait in many species of fruit flies , is an excellent system in which to study the cell biology , genetics and evolution of a trait . In Drosophila melanogaster , where the incipient SC rotates from horizontal to a vertical position , three signal comb properties have been documented: length , final angle and shape ( linearity ) . During SC rotation , in which many cellular processes are occurring both spatially and temporally , it is difficult to distinguish which processes are crucial for which attributes of the comb . We have used a novel approach combining simulations and experiments to uncover the spatio-temporal dynamics underlying SC rotation . Our results indicate that 1 ) the final SC shape is primarily controlled by the inhomogeneity of initial cell size in cells close to the immature comb , 2 ) the final angle is primarily controlled by later cell expansion and 3 ) a temporal sequence of cell expansion mitigates the malformations generally associated with longer rotated SCs . Overall , our work has linked together the morphological diversity of SCs and the cellular dynamics behind such diversity , thus providing important insights on how evolution may affect SC development via the behaviours of surrounding epithelial cells .
Morphogenesis concerns the development of forms in organisms . It is a major theme in biology not only because of its importance in its own right , but also because of its essential relationship and interactions with biological evolution . Darwin pointed out that the adult form of an organism depends on its developmental trajectory . Hence , if heredity is important in determining the adult form , it must exert its influence during development [1] . As with many natural phenomena , physical processes involving morphogenesis inherently span many hierarchical levels of biological organization and time scales , in that events from one level ( e . g . genetic level ) can alter events at another level ( e . g . cellular level ) . A central research topic in evo-devo is to understand how those hierarchical levels integrate leading to evolution [2–4] . One example of morphogenesis that lends itself to exploring physical processes at genetic , cellular and tissue levels through time is the formation of a morphological feature called the sex comb ( SC ) in species of the Drosophilidae family ( Fig 1A ) [5–7] . Existence of SCs is a male-specific trait of many species of flies . The phylogenetic relationships of these species have been studied showing that SCs have evolved independently several times [8–10] . In D . melanogaster , the SC is a linear , almost vertical arrangement of modified bristles on the first-tarsal segment of their forelegs . During development , SCs of D . melanogaster were imputed to rotate from a horizontal to an almost vertical position ( Fig 1B ) [11] and this has been corroborated using modern genetic and microscopical tools [6 , 12–14] . In related experiments , Malagón [15] had initial evidence that the major force driving SC rotation was provided by cell expansion distal to ( below ) the SC , and that the cells proximal to ( above ) the SC passively responded by diminishing in area and disappearing from the epithelium . However , not all species of flies with vertical , linear SCs develop them through rotation . Some species utilize different bristles which are already in a vertical position and merely have to come together to form a vertical column . Indeed some species with vertical SCs on more than one segment use a different method on each segment [8] . Drosophila SCs display spectacular developmental and morphological variations during evolution . Some examples include comb shape ( Fig 1E ) , comb length ( Fig 1F ) , number of combs per tarsal segment , tooth size and pigmentation . Possibly , the most interesting comb feature involves its orientation [9] , which constantly changes between three positions relative to joint: transverse , diagonal , and vertical ( Fig 1D ) . Malagon and Larsen [16] suggest that genetic perturbations in D . melanogaster can easily phenocopy changes in comb variation . Thus , the SC system provides a rich developmental and evolutionary phenomenology with which to explore the strategies and tactics involved in morphogenesis and its evolution . Understanding the dynamics of cell behaviours and the mechanical constraints underlying SC morphogenesis represents an important step towards linking the genetics of cellular behaviours which occur during development to their evolution over time . Combined use of different approaches is essential for further progress in evolutionary-developmental biology . We previously used a combination of developmental and experimental approaches and showed the role of developmental constraints and interaction between development and selection in the rotation and evolution of SCs in D . melanogaster [6] . Here , we use a combination of computational modelling ( cellular Potts model , or CPM , [17] ) with experimental evidence to investigate and quantify the spatio-temporal dynamics and interplay of various mechanical characteristics of cells critical for the proper rotation of SCs in D . melanogaster . Although computational modelling or a hybrid computational-experimental approach has been used successfully to describe various cellular processes , including examples in morphogenesis [18–26] , to our best knowledge this work represents the first hybrid approach used to explore the cellular dynamics driving SC morphogenesis . Furthermore , we emphasize that in addition to replicating experimental results in our simulations , we have generated quantitative and falsifiable hypotheses which can guide future experiments , thus showing the synergy between experiments and computational modelling in generating novel insights in broad areas of biology .
In this section we show via simulation how the initial spatial arrangement of distal cells is critical for proper SC rotation . Fig 2A and 2B show snapshots of two SC rotation simulations . These two examples share identical initial spatial cell configurations . Importantly , under this specific initial cell configuration , every distal cell has an initial cell size roughly equal to each other ( when t = 0 mcs , top panels of Fig 2A and 2B ) . Moreover , a t a r g e t t e r m ( Eq 5 ) is set to be equal for every distal cell in each simulation of Fig 2A and 2B . The only difference in parameter setup between Fig 2A and 2B is that a t a r g e t t e r m of distal cells of Fig 2A is smaller than that of Fig 2B . ( a t a r g e t t e r m ( c = E P 1 ) = a t a r g e t t e r m ( c = E P 2 ) = 13 pixels in Fig 2A , while a t a r g e t t e r m ( c = E P 1 ) = a t a r g e t t e r m ( c = E P 2 ) = 20 pixels in Fig 2B . ) Taken together , expansion rates of distal cells are different across simulations ( and with Fig 2B having a higher expansion rate than Fig 2A ) , even though the expansion rates are roughly uniform across distal cells within a simulation . It is clear that proper rotation is unlikely to occur with such uniform expansion rates of distal cells , as evidenced by the severe curvature of ( Fig 2A ) or broken ( Fig 2B ) SC at the end of rotation . These rotation defects are not rescued by a blanket change in expansion rates of distal epithelial cells ( SC rotations in both Fig 2A and 2B are defective ) . The curvatures of the rotated SCs suggest that uniformity of expansion rate across epithelial cells causes unequal rotation along the length of the SC , with the base end receiving more rotation than the tip end . To restore proper rotation , distal cells have to be arranged in a manner that provides more rotation to the SC tip . One way to achieve this is to make the initial sizes inhomogeneous across distal cells , with the distal epithelial cells having smaller initial sizes closer to the tip end of the SC ( top panel of Fig 2C ) . When coupled with similar a t a r g e t t e r m values ( Table 2 ) across distal cells , this inhomogeneous spatial arrangement of epithelial cells creates a differential push which largely maintains the shape of the SC during the entire rotation , therefore increasing the likelihood of proper SC rotation ( Fig 2C ) . Even though we have shown only three examples , the above phenomenon is robust in all simulations . In fact , we are unable to produce a normally straight and vertical SC in any simulation with initial spatial cell configurations and a t a r g e t t e r m values that represent uniform expansion rates across distal epithelial cells . Proper SC rotations can only occur in simulations with inhomogeneous distal epithelial cell expansion rates in the manner as in Fig 2C . We thus deduce that such inhomogeneous expansion rates amongst distal epithelial cells , brought about by the inhomogeneous initial spatial cell configuration , must be a critical biological mechanism underlying proper SC rotation . Having established that the initial spatial configuration of epithelial cells is crucial for proper SC rotation , we now investigate the dependence of SC orientation on the expansion parameters of epithelial cells . Fig 3A , 3B and 3C show respectively three representative SC simulations in which we obtain increasing final rotation angles α ( definition in Fig 3E ) while controlling expansion rates of distal cells . In each of the three simulations , we start with the common initial condition ( top panel of Fig 3 ) where the starting areas of distal cells are different , with the larger cells concentrated towards the base ( pivot point ) of the SC and the smaller cells concentrated towards the SC “tip” . This common initial condition used here is identical to the one used in Fig 2C . To obtain different final rotated SC angles α , we control the expansion characteristics by using different values of a t a r g e t t e r m for distal cells in each of the three representative simulations . We set smaller values of a t a r g e t t e r m to obtain a smaller α , and vice versa . Specifically , we set a t a r g e t t e r m ( c = E P 1 ) = 6 . 5 and a t a r g e t t e r m ( c = E P 2 ) = 5 in Fig 3A; a t a r g e t t e r m ( c = E P 1 ) = 9 and a t a r g e t t e r m ( c = E P 2 ) = 7 in Fig 3B . In Fig 3C , the default values of a t a r g e t t e r m are used ( Table 2 ) . Spatially inhomogeneous distal cell expansion is observed in each of these three simulations from the summary statistics Fig 3D . Specifically , in each of these three simulations , the magenta distal cells EP2 ( closer to the SC “tip” ) expand the most and the blue distal cells EP1 ( closer to the base pivot point ) expand the least . As discussed in Fig 2C , such spatially inhomogeneous and differential expansion in distal epithelial cells is critical for providing the appropriate amount of push across the length of SC , so that it is still well-formed and intact towards the end of the rotation . We illustrate here that the degree of the inhomogeneity determines the final SC rotation angle α . An SC rotation with a higher angle α requires a higher spatial inhomogeneity of expansion between distal cells at each of the extreme ends of SC , as one readily observes from the summary statistics that the spatial inhomogeneity in distal cell expansion is highest in Fig 3C but lowest in Fig 3A . Do we find spatial heterogeneity during distal cell expansion as predicted from our simulations ? To test this , we performed in vivo experiments with ♀wt , ♂babPR72 and ♂wt ( Fig 4 ) . SC rotation is absent from ♀wt , while rotation is partial in ectopic heterozygous babPR72 mutants with sex comb on the second tarsal segment t2 . Although five pupae were measured for each of ♀wt , ♂babPR72 and ♂wt for statistical analyses ( Table A in S1 Text ) , we only display images taken at the start and at the end of the rotation of the same representative pupa for each of ♀wt , ♂babPR72 and ♂wt respectively . Following our conventions , we label the distal cells closer to the SC base as “EP1” ( blue ) , the distal cells closer to the SC tip as “EP2” ( magenta ) . The line separating EP1 and EP2 cells is located at the half length mark of the SC on every image on which area analysis was performed . Three observations are clear from Fig 4: 1 ) In rotating sex combs studied ( ♂wt and ♂babPR72 ) , average initial apical area ( obtained by dividing the coloured area with the number of cells in the area ) of EP2 ( magenta ) is smaller than that of EP1 ( blue ) , showing the spatial inhomogeneity of initial apical areas amongst distal cells ( median P r ( E P 2 I n i t i a l E P 1 I n i t i a l < 1 ) = 1 . 00 with P r ( E P 2 I n i t i a l E P 1 I n i t i a l < 1 ) > 0 . 5 at least 99 . 98% of times for ♂wt; median P r ( E P 2 I n i t i a l E P 1 I n i t i a l < 1 ) = 0 . 99 with P r ( E P 2 I n i t i a l E P 1 I n i t i a l < 1 ) > 0 . 5 at least 99 . 98% of the times for ♂babPR72 ) , while spatial inhomogeneity of distal cell areas is not apparent in ♀wt . ( Median P r ( E P 2 I n i t i a l E P 1 I n i t i a l < 1 ) ∼ 0 . 54 with P r ( E P 2 I n i t i a l E P 1 I n i t i a l < 1 ) > 0 . 5 approximately 65% of times–S1 Text , Table B in S1 Text and Figure A in S1 Text ) . 2 ) In rotating sex combs studied ( ♂wt and ♂babPR72 ) , EP2 on average expands at a faster rate ( defined as final area initial area × 1 time taken ) than EP1 , providing a differential push to the SC during rotation ( median Pr ( Cinhomogeneity > 1 ) = 1 . 00 with Pr ( Cinhomogeneity > 1 ) > 0 . 5 at least 99 . 98% of the times for ♂wt , median Pr ( Cinhomogeneity > 1 ) = 1 . 00 with Pr ( Cinhomogeneity > 1 ) > 0 . 5 at least 99 . 98% of times for ♂babPR72–Table B in S1 Text and Figure A in S1 Text ) , and 3 ) how much EP2 expands faster on average than EP1 is related to α , with a higher α correlated with a greater disparity in expansion rates between EP2 and EP1 ( median Pr ( ΔCinhomogeneity ( ♂wt , ♀wt ) > 0 ) = 1 . 00 with Pr ( ΔCinhomogeneity ( ♂wt , ♀wt ) > 0 ) > 0 . 5 at least 99 . 98% of the times; median Pr ( ΔCinhomogeneity ( ♂wt , ♂babPR72 ) > 0 ) = 0 . 70 with Pr ( ΔCinhomogeneity ( ♂wt , ♂babPR72 ) > 0 ) > 0 . 5 90% of the times; median Pr ( ΔCinhomogeneity ( ♂babPR72 , ♀wt ) > 0 ) = 0 . 99 with Pr ( ΔCinhomogeneity ( ♂babPR72 , ♀wt ) > 0 ) > 0 . 5 at least 99 . 98% of the times–Figure A in S1 Text and Table B in S1 Text ) , further buttressing the differential push claim in simulations . ( Pr means probability . Please see , for example [30–32] , for Monte Carlo simulations and bootstrapping method for statistical analyses . ) Next , we investigate the dependence of SC rotation on the length of the SC and the adhesion between successive SC teeth . To quantify SC rotation , we have already introduced α in the previous section and used it in Fig 3 to describe how much SCs rotate . We now introduce ABASCT , angle between adjacent sex comb teeth , which is useful to quantify the departure of rotated SCs from linearity . In Fig 5A , we illustrate the standard deviation ( SD ) of ABASCT is related to SC shape variation . β1 , β2 , β3 and β4 are respectively the angles between SC teeth 1 and 2 , 2 and 3 , 3 and 4 , and 4 and 5 of example SC1 . The SD of these angles is a measure of curvature of SCs . An almost straight SC , SC2 , has a lower ABASCT SD value , but an SC that is highly non-linear , such as SC3 , has a relatively large ABASCT SD value . We emphasize that ABASCT SD and α measure separate properties of the SC . For example , SC2 and SC3 have identical α but distinct ABASCT SDs . Hitherto we have mostly discussed how spatial properties of distal epithelial cells may help proper SC rotation . However , even with inhomogeneous spatial patterning of distal cells , the rotated simulated SCs suffer from a decreasing intact ratio as the length of SC is increased ( Fig 6A ) . In particular , the intact ratio of 11-tooth SCs drops considerably across virtually all examined SC adhesion parameters when compared with the shorter 9-tooth SCs . In this section we explore whether introducing temporal dynamics to the expansion of distal epithelial cells , in addition to spatial patterning , could improve the breaking statistics of rotated SCs . We perform additional simulations in which the blue distal cells ( EP1 ) expand later in the simulations , while keeping everything else identical . There are 3 × 36 = 108 additional sets of simulations , which cover 3 magnitudes of delayed expansion of EP1 cells ( 40 , 120 and 200 mcs ) . Visual examination of the aggregate statistics suggests that these extra simulations share similar α and ABASCT SD distributions with the regular , non-delayed simulations ( Figure G in S1 Text and Figure H in S1 Text ) . However , comparison of intact ratios between the delayed and non-delayed simulations ( Fig 8A ) shows that while there is no apparent improvement on intact ratios for shorter 5 , 7 , 9-tooth SCs , there is a consistent improvement for 11-tooth SCs across almost all SC adhesion parameters ( Table D in S1 Text ) . In fact , with an appropriate delay parameter ( ∼ 120 mcs ) , intact ratios of 11-tooth SC approach that of a 9-tooth one . Too high a delay parameter ( > 150 mcs ) does not make intact ratios higher . On the contrary , it causes compression of cells at the centre of the distal region , a phenomenon not readily observed in experiments . Fig 8B shows two examples of 11-tooth SC simulations in which a delayed expansion of EP1 affects the final outcome of the SC , with the one having a delayed expansion of EP1 enjoying an intact SC while the non-delayed simulation has a broken rotated SC . To determine whether such a temporal delay exists in experiments , we collect data on the time course of growth of distal cells in each of our short , wildtype and long SC examples ( Fig 9 ) . The percentage growth in area of these cells ( as compared to the start of measurement at either 22 or 23 hours AP ) is then fitted with logistic functions to determine the time lag between the expansion phase of distal epithelial cells in different areas . The results of curve fitting suggest that , in our long SC example ( Fig 9C ) , there is a delay in the expansion phase of distal epithelial cells closer to the base of the SC relative to the distal epithelial cells closer to the SC tip , with the start of expansion phase of magenta cells ( closer to the SC tip ) at 23 . 27 hours ( 22 . 25-24 . 28 hours 95% CI ) , while that of blue cells ( closer to the SC base ) is 27 . 42 hours ( 26 . 34-28 . 51 hours 95% CI ) . In the wildtype example ( Fig 9B ) , however , there is no statistically significant difference between the start of the expansion phase of the selected magenta and blue cells . Our experimental results provide evidence that longer SCs may take advantage of the time delay to achieve proper rotations . We emphasize that even though delayed expansion of blue coloured cells is also detected in our short SC example ( Fig 9A ) , it does not represent an inconsistency between modelling and experiments because delayed expansion of EP1 ( blue ) cells does not statistically affect the intact ratio of shorter SCs ( Fig 8A ) .
SC rotation in D . melanogaster involves a complex pattern of cell behaviours ( such as cell proliferation , shape change and movement ) [15 , 27 , 34 , 35] . However , how these multiple cellular processes contribute to SC morphogenesis has remained elusive . We wished to understand how the temporal and spatial properties of each cell behaviour contribute to normal comb rotation . Although this would be experimentally impractical , it is possible to explore these properties computationally by simulating the effects of varying cell behaviours . From extensive simulations , we deduce that proper rotation of the SC depends on precise spatio-temporal dynamics of distal cells . The initial spatial cell size distribution in the distal epithelium was found to be important to minimize SC bending and malformation . In particular , distal cells closer to the SC base ( pivot point ) have larger initial apical areas than cells farther away from the base . Such inhomogeneity in initial apical areas of distal cells translates into a differential push across the length of SC as the cells expand , ensuring proper SC rotation in simulations . In addition to spatial distribution , a temporal component was also discovered in our simulations , when delay in expansion of distal cells closer to SC base relative to cells closer to the SC tip reduces incidence of long SC breaking during rotation . These simulations are not only useful in showing that cell behaviours distal to the rotating comb are likely sufficient to provide the motive force for comb rotation , they also suggest relationships of which importance was not previously realized . In fact , having found in simulations of the spatio-temporal dynamics of the distal cells predicted to be crucial for proper SC rotations , we were able to go back to the original data in the 4D movies and confirm that this was indeed occurring biologically . Not only did the simulations produce testable hypotheses concerning developmental mechanisms in the model species D . melanogaster they also suggested hypotheses concerning how changes in developmental processes might produce different SC patterns during fly evolution . For instance , changes in rotation angle might be due to changes in expansion of distal cells ( Fig 4 ) . Our work also suggests potential avenues for these changes and which paths might be forbidden . As an example , rotation of long SCs in our simulations are more difficult because increasing comb length generates multiple problems for movement including comb breaking , atypical alignment , among others ( Fig 6 ) . These results are consistent with the differing strategies to achieve SC morphogenesis taken up by multiple Drosophila species with long SCs . Most of the SCs from these species lack rotation , appearing already in a vertical position early in the developmental process [6 , 9 , 13] . Furthermore , other species with long rotating SCs such as Drosophila guanche display frequent broken and misaligned bristles [6] , consistent with our simulations . From our biological Drosophila experiments , we observed a relatively consistent linear shape ( quantified by ABASCT SDs , [8 , 13 , 27] ) of rotating SCs amongst Drosophila species , yet the orientation of SCs ( i . e . α ) has a larger variation [13] . On the other hand , we have discovered from our simulations that SC shape is primarily ( though not exclusively ) governed by the initial inhomogeneity of distal cell sizes ( Fig 2 ) , while SC orientation can be controlled by the subsequent rate of expansion of these cells ( Fig 3 ) . Therefore , if these two phenotypes ( shape and orientation of SCs ) follow distinct evolutionary trajectories , our simulation results suggest that there may be signatures left by evolution on the two corresponding governing cellular characteristics ( initial inhomogeneity in distal cell sizes and subsequent expansion of these cells , respectively ) . In other words , a canalized SC shape implies a conserved spatial inhomogeneity in initial distal cell areas , but variation in SC orientation implies diversity in subsequent expansion rates of these cells . Preliminary evidence from Fig 4 suggests this is indeed the case . Summary statistics Fig 4D and subsequent statistical analyses in particular from Figure A in S1 Text show an approximate , albeit imperfect conserved existence of inhomogeneity of initial distal cell areas across the male fly genotype examples ( both wildtype and mutant ) , which exhibit different SC orientations but similar SC shapes . It is primarily the variation in the subsequent expansion of epithelial cells between the fly genotypes that determines the SC orientation . Our results suggest the possibility of separate molecular mechanisms , which are born out of evolution and are continuously modified by it , underlying these two cell characteristics . In addition , our results raise the question of whether Drosophila species using developmental mechanisms other than rotation also canalize their SC shape while exhibiting higher variation in comb orientation . There are several shortcomings in this first work to study SC rotation in silico which warrant future improvements . First , although we attempt to replicate as close to experimental results as possible , αs in our simulations generally have slightly lower values than expected for normal rotation . Our results thus suggest that other cellular processes may also contribute to SC rotation . For example , joint formation is not included in our simulations yet previous studies have pointed out that it affects the basal part of SC [13] and constitutes the male-specific movement of sensillum campaniforme [15] . Effects of joint formation on SC rotation can be studied using photo-activated lines that can disrupt development precisely [36 , 37] . We also did not systematically investigate the possibility of the “pulling” effect from proximal cells during SC rotation . A primary reason we focused on the distal , not proximal cells is that even without significant contraction in apical areas in the proximal region , proper SC rotation could still occur [15] . There is also an absence of experimental observation of any initial spatial arrangement of proximal cells similar to , but opposite in direction of , distal cells . The apparent absence of such spatial arrangement means that proximal cells are unlikely to exert a coordinated pulling force similar to the pushing force of distal cells during SC rotation . Moreover , we did not observe from experiments a clear spatial or temporal sequence in which proximal cells extrude from the epithelium , meaning that it is more likely that the proximal cells were passively “squeezed” due to external pressure rather than actively contracting generating a pulling force . Nevertheless , for future considerations it would be desirable to quantify the effects ( if any ) of proximal cells in next iterations of simulation models . Currently , although cell extrusion is included in our simulations , its effects are not examined thoroughly . While cell extrusion is shown to be the consequence of cell crowding rather than a major driving force of development [38 , 39] , future simulations could better quantify its effects above the SC and test whether it buffers comb shape . From the simulation perspective , the parameters used here are chosen to reproduce SC dynamics as observed in experiments . A consequence of such an approach is that these parameters may not correspond to the mechanical characteristics of actual SC or epithelial cells . In theory , mechanical characteristics of cells are related to the spatial derivatives of the Hamiltonian Heff used in CPM . However , the exact correspondence is usually complicated and may depend on the specific topology of the cellular systems studied . Although there have been attempts to perform CPM simulations with parameters based on matching the terms of Heff and the mechanical characteristics of cells [40] , the majority of simulation examples in the literature use our approach of choosing parameters based on the correct reproduction of observed cellular behaviours [41] . We expect the major conclusions established in this work should be largely independent of the specific parameters used in the simulation model , so long as those parameters also demonstrate observed cellular behaviours . For example , it would be extremely difficult to imagine something as fundamental as the spatio-temporal properties of distal cells not to hold if another set of more “realistic” parameters were used for simulations , in particular when these previously unnoticed spatio-temporal properties are themselves demonstrated in experiments ( Figs 4 and 9 ) . Future simulations could address this issue by further refining the parameters so as to reflect the actual mechanical characteristics of cells . In addition , some of the cells in the simulations have a very small initial lattice size ( ≤ 5 pixels ) . This could potentially create issues when the stochastic effects of the simulations become large . For example , these small cells could disappear due to stochastic effects early in the simulations with no chance to expand . To check whether our simulation results are scalable in size , we performed some sample simulations with identical cell arrangements but with a resolution 4 times as much ( i . e . 228 × 168 pixels in the sample simulations vs . 114 × 84 pixels–Figure I in S1 Text ) . We discovered that SCs were able to properly rotate just as in the original simulations . Thus , although some anomalous effects could be present in the simulated cells in the original simulations due to their smaller grid sizes , these are likely not severe enough to have impacted the conclusions of this work . In the area analysis of Fig 4 and Table A in S1 Text , we classified distal cells by a demarcation line drawn at the halfway length mark of the SC on every confocal image where area analysis was performed . The demarcation line is horizontal for 30° ≤ α ≤ 90° but perpendicular to the SC for α < 30° . Any distal cell above the line ( closer to SC tip ) was defined as EP2 and any distal cell below it ( closer to the SC base ) EP1 . Such an approximate method of classifying distal cells in experiment could be problematic because movement of cells during SC rotation means that some of the distal cells may have different classifications at the start and at the end of SC rotation . To ensure that our classification method is robust , we have performed individual cell tracking on a sample fly leg ( Figure J in S1 Text ) . The advantage of individual cell tracking is that the cell labels ( whether the cells are defined as EP1 or EP2 ) are consistent throughout the entire area analysis . We obtained identical conclusions with individual cell tracking as with our original method , thus showing that our approximate method of cell classification does not significantly affect the conclusions established in this work . Separately , a certain degree of cell compression was observed along the EP1/EP2 boundary in the simulations of Fig 3A and 3B , while no such simulation artefact was observed in Fig 3C . Cell compression occurs in some simulations because of the inhomogeneous expansion rates amongst the distal cells due to their different initial areas , terminal target areas and timing of expansion . Therefore , some EP1 or EP2 cells are receiving more pushing forces than others , causing compression . In particular , the imbalance in pressure along the EP1/EP2 boundary is more acute in Fig 3A and 3B simply because we reduced the expansion of magenta EP2 cells in Fig 3A and 3B ( as compared to Fig 3C ) , but we did not reduce as much expansion for the blue EP1 cells . We believe that such cell compression can be further mitigated by fine-tuning initial distal cell area configurations , terminal target areas and λ parameters . As well as expansion , distal cells were observed experimentally to elongate along the y or the ordinate axis during SC rotation . While we have included cell polarization ( ϵ in Table 2 ) in our model , it remains an interesting question the extent of distal cell polarization contributes to SC rotation . Although we have not systematically examined the issue from a simulation standpoint , in the early model development we ran limited sets of simulations in which distal cell expansion was isotropic ( i . e . cell polarization was absent ) , we noticed that simulated SCs could still properly rotate but the rotation angle ( α ) was generally smaller by 10-15° as compared to simulations with cell polarization . More extensive simulations are required to fully elucidate the relationship between cell polarization and SC rotation . Finally , during the measurement of temporal changes in apical areas of distal cells ( Fig 9 ) , we noticed that in addition to the primary growth pattern following logistic function ( fitted curves of Fig 9A , 9B and 9C ) , there also exists oscillatory behaviours for several of the cell samples measured , in particular the ones in Fig 9A and 9C . Although actin-myosin binding dynamics are thought to be responsible for similar oscillatory behaviours with a much shorter time scale [42] , we remain agnostic about the possible mechanisms for cell area oscillations in our case because of their long periods ( up to a few hours ) . Future experiments are required to answer the open question about the mechanism underlying these oscillations and whether such oscillatory behaviours contribute to SC rotation .
Unless otherwise specified , simulations were performed in sets . A set consists of 48 independent simulations . These 48 simulations share identical parameters and initial cell configurations , but differ in the seed for pseudo random number generation ( Knuth’s subtractive method , [31 , 51] ) which affects the selection of pixels for attempted flips . Since these simulations can be regarded as multiple numerical experiments under the same parameters , we can extract attributes from these outputs and perform analyses to determine how different parameters affect outputs in a statistical manner . In other sections we have developed various metrics which quantify the well-formedness of the rotated SCs in simulations . These metrics were extracted from the final image output file of each simulation via Python with the NumPy package and APIs for vtk files . R ( version 3 . 0 . 2 , https://www . r-project . org , [52] ) with the multcomp [53] and ggplot2 [54] packages was used for statistical analyses and visualization of results . Visualization of simulated SC rotations and data plotting were performed with paraview ( version 4 . 0 . 1 , www . paraview . org , [55] ) and gnuplot ( version 4 . 6 , www . gnuplot . info ) respectively . The grofit package of R [56] was used to fit cell expansion data of Fig 9 , while gimp ( version 2 . 8 . 10 , www . gimp . org ) was used for colour highlighting of cells there . The default Mersenne-Twister [57] in R was used as the pseudo random number generator for statistical analyses of Fig 4 ( Table B in S1 Text ) . | The sex comb ( SC ) is a series of modified bristles on the male forelegs of many species of fruit flies . The size , position and shape of these sex combs vary drastically across different fly species . Therefore , SCs are a model system which illustrates the interaction between evolution and organism development influencing phenotypic features . In this work , we use a combined simulation-experimental approach to study the cellular processes involved in the rotation of developing SCs in common fruit flies ( D . melanogaster ) . Our results indicate that despite the appearance of a complicated set of motions of surrounding cells associated with SC rotation , the final SC attributes only depend on a few selected parameters . We showed that changes in the timing and extent of cell size increase in distal cells altered the extent of SC rotation and breakage . Furthermore , these changes were sufficient to account for the observed variations in SC rotation between different fly species . Thus , our computational model has given us important insights on how evolution may use various cellular processes as a means to manifest the diversity of SCs across different fly species . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"statistics",
"animals",
"epithelial",
"cells",
"simulation",
"and",
"modeling",
"animal",
"models",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"mathematics",
"experimental",
"o... | 2018 | Rotation of sex combs in Drosophila melanogaster requires precise and coordinated spatio-temporal dynamics from forces generated by epithelial cells |
The strength and breadth of an individual’s antibody repertoire is an important predictor of their response to influenza infection or vaccination . Although progress has been made in understanding qualitatively how repeated exposures shape the antibody mediated immune response , quantitative understanding remains limited . We developed a set of mathematical models describing short-term antibody kinetics following influenza infection or vaccination and fit them to haemagglutination inhibition ( HI ) titres from 5 groups of ferrets which were exposed to different combinations of trivalent inactivated influenza vaccine ( TIV with or without adjuvant ) , A/H3N2 priming inoculation and post-vaccination A/H1N1 inoculation . We fit models with various immunological mechanisms that have been empirically observed but have not previously been included in mathematical models of antibody landscapes , including: titre ceiling effects , antigenic seniority and exposure-type specific cross reactivity . Based on the parameter estimates of the best supported models , we describe a number of key immunological features . We found quantifiable differences in the degree of homologous and cross-reactive antibody boosting elicited by different exposure types . Infection and adjuvanted vaccination generally resulted in strong , broadly reactive responses whereas unadjuvanted vaccination resulted in a weak , narrow response . We found that the order of exposure mattered: priming with A/H3N2 improved subsequent vaccine response , and the second dose of adjuvanted vaccination resulted in substantially greater antibody boosting than the first . Either antigenic seniority or a titre ceiling effect were included in the two best fitting models , suggesting a role for a mechanism describing diminishing antibody boosting with repeated exposures . Although there was considerable uncertainty in our estimates of antibody waning parameters , our results suggest that both short and long term waning were present and would be identifiable with a larger set of experiments . These results highlight the potential use of repeat exposure animal models in revealing short-term , strain-specific immune dynamics of influenza .
Natural infection with influenza stimulates a complex and multifaceted immune response to neutralise and clear the infection [1] . The adaptive immune response is of particular interest for seasonal epidemic and pandemic preparedness , as responses from previous exposures provide some long-term protection against reinfection and disease via antibody and T-cell mediated immunity [2 , 3] . Focusing on the adaptive immune response is also advantageous because it can be ( i ) induced in advance of an epidemic through vaccination and ( ii ) measured and compared against correlates of protection to improve public health forecasting [4–7] . Influenza is an antigenically variable virus and undergoes continual antigenic drift , whereby mutations in immunodominant epitopes are selected by immunological pressure , allowing influenza lineages to escape population herd immunity [8–10] . This results in the continual loss of long-term immunity as antibodies effective against past strains fail to neutralise novel variants [11] . The current strategy for combating antigenic drift is to regularly update the seasonal influenza vaccine to better represent circulating strains , resulting in a competition between virus and vaccine formulation . Vaccines are more effective in some years than others due to factors such as: the antigenic match between the vaccine and circulating strain , and prior exposure histories of the population [12 , 13] . Consequently , there has been a recent push towards a universal influenza vaccination strategy , either through new vaccines or improved strength and breadth of immunity using existing technologies [14 , 15] . Whilst there is some cross-reactivity and cross-protection within influenza A virus subtypes and within influenza B virus lineages , humans experience numerous infections over their lives [16–18] . Each successive influenza exposure , which may be vaccination or infection , can strengthen the available repertoire of T and B cells which target epitopes on circulating and previously encountered strains [19 , 20] . In the humoural response , this occurs by boosting antibodies produced by pre-existing long-lived plasma cells and activated memory B cells ( MBCs ) , and through generating a novel B cell response targeting unrecognised epitopes [21–23] . Given that individuals experience repeated infections and vaccinations from antigenically varied influenza viruses , interpreting the composition of an observed antibody response is confounded by the complex interaction of an individual’s immunity from prior infections with the infecting virus [24–27] . A large body of experimental and observational work exists describing the contribution of infection histories to observed influenza susceptibility profiles , antibody landscapes and vaccination responses , often under the terms “original antigenic sin” or “antigenic seniority” [1 , 28–34] . Furthermore , next generation assays to characterise antibody diversity and B cell identity have provided a detailed understanding of immune dynamics , including short term immune kinetics , duration of the humoural response , and immunodominance of different antigenic sites [35–40] . However , few studies have integrated these mechanisms into quantitative frameworks which can be used to explain and predict serological data from human populations , which often rely on simpler and less finely resolved assays [18 , 22 , 41 , 42] . Animal models , in particular ferrets , have been used to generate much of our understanding of influenza immunology due to opportunity for intensive observation and control [36 , 43–48] . Here , we exploit the experimental flexibility and transparency of a ferret model to find evidence for and quantify multiple immunological mechanisms that may be important in characterising antibody landscapes generated from complex exposure histories , yet are observable using only routine antibody assays . Quantifying short term mechanisms in a ferret system might reveal patterns that could be used to improve the predictability and interpretation of human antibody landscapes following exposure [49] . We developed a mathematical model of antibody boosting and biphasic waning to describe antibody kinetics using previously published antibody titre data from a group of ferrets with varied but known exposure histories [43] . Previous models of antibody kinetics have often focused on the response to a single immunogen following one exposure [50–52] . Here , we take into account previously described immunological phenomena to describe cross-reactive antibody titres arising from varied exposure histories . These phenomena included exposure type-specific homologous and cross-reactive antibody kinetics , the role of priming on subsequent vaccination , titre-dependent antibody boosting ( or titre ceiling effects ) and reduced antibody boosting with each subsequent exposure ( antigenic seniority ) [32 , 53–59] . By fitting models with various combinations of these mechanisms to haemagglutination inhibition ( HI ) titre data from ferrets , we sought to identify immunological mechanisms that are important in describing observed antibody profiles arising from multiple exposures . Parameter estimates from these model fits allowed us to quantify the impact of prior infection and adjuvant inclusion on antibody levels following vaccination and to compare homologous and cross reactive boosting profiles of different exposure types .
Antibody titre data were obtained from a previously published ferret study [43] . The experimental protocol was originally designed to reflect different possible human infection and vaccination histories at the time of the 2009 pandemic . Here we present a secondary analysis of these data , with the intention of characterizing underlying immunological processes . Briefly , five experimental groups each of three ferrets underwent different combinations of infection with seasonal influenza A and/or vaccination with Northern and Southern Hemisphere trivalent inactivated influenza vaccine ( TIV ) , with or without Freund’s incomplete adjuvant ( IFA ) , over the course of 70 days ( Table 1 ) . Serum samples were collected at days 0 , 21 , 37 , 49 and 70 from all ferrets ( Fig 1 ) . HI titres were used to determine antibody titres to each infection and TIV strain . Dilution plates with 12 wells were used , such that the highest possible recorded dilution was 1:40960 , and the lowest detectable titre was 1:20 . Undetectable titres were recorded as <1:20 . All analyses here were carried out using log titres , defined as k = l o g 2 ( D 10 ) , where D was the recorded dilution . Observed log titres were therefore assigned values between 0 and 12 , where <1: 20 = 0 , 1: 20 = 1 and ≥1: 40960 = 12 . Full adult doses of human TIV were used in groups A , B , C and D . The first vaccination ( Southern Hemisphere 2008 TIV ) contained A/Solomon Islands/3/2006 ( H1N1 ) , A/Brisbane/10/2007 ( H3N2 ) and B/Brisbane/3/2007 , administered at day 28 ( TIV 1 ) . The second vaccination ( Northern Hemisphere 2007/2008 TIV ) contained A/Solomon Islands/3/2006 ( H1N1 ) , A/Wisconsin/67/2005 ( H3N2 ) and B/Malaysia/2506/2004 , administered at day 42 ( TIV 2 ) . Vaccines used in groups B and D were emulsified in an equal volume of IFA immediately before administration ( TIV 1/2 + adjuvant ) . All vaccines contained 15μg of HA of each strain , and were delivered to sedated animals intramuscularly in the quadriceps muscles of both hind legs . Infections were carried out by dropwise intranasal challenges with 103 . 5 50% tissue culture infectious doses ( TCID50 ) in 0 . 5 mL with A/Panama/2007/1999 ( H3N2 ) in groups C , D and E , and with A/Fukushima/141/2006 ( H1N1 ) in all groups . The mathematical model describes the kinetics of homologous and heterologous antibody titres following exposure . Fig 2 depicts the example of an individual becoming infected and later vaccinated , though the model may characterise any sequence of exposures . Conceptually similar mathematical models of boosting followed by biphasic waning have been used previously to describe antibody secreting cell ( ASC ) and antibody kinetics [50 , 52 , 60 , 61] . After an infection ( start at time ξ1 ) , homologous antibody titres undergo boosting rising linearly ( on the log scale ) by μ1 log units to a peak after time tp1 , ignoring any delay between exposure and the start of antibody production . Titres then quickly drop by a fixed proportion , d1 , over ts1 days ( in the timescale of a few days to weeks ) , representing the initial short-term waning phase as free antibodies and early short-lived ASCs begin to decay following clearance of the initial antigen dose [33] . Antibody waning then switches to a constant rate m1 ( log titre units lost per day ) for the remainder of time ( representing the population of persistent ASCs , lasting months to years ) until subsequent vaccination ( syringe at time ξ2 ) , when antibody dynamics become dominated by a new set of boosting and waning parameters . We did not include a third , steady state phase due to the short time frame of these experiments [62] . Antibodies effective against heterologous strains experience boosting and biphasic waning in proportion to the exposure strain , with the proportion dependent on the antigenic distance between the measured and exposure strains . We assumed that the lower bound of detection of the HI assay ( a log titre of 0 ) was synonymous with a true absence of antibodies , such that model predicted titres could not wane below this level . We then built on this base model to incorporate additional immunological mechanisms that are important in describing antibody boosting and waning . These included: biphasic or monophasic antibody waning; exposure-type specific or type non-specific cross-reactivity; antigenic seniority; the impact of priming infection on subsequent vaccine response; and titre-dependent boosting . We considered models with different numbers of exposure types to match the experimental design: either 3 ( infection , TIV , TIV + adjuvant ) or 6 ( priming infection , secondary infection , initial TIV , secondary TIV , initial TIV + adjuvant , secondary TIV + adjuvant ) . The base boosting and waning model remains the same across model variants , but these mechanisms add complexity to the boosting parameter , μ , and link different exposures with common parameters . A full description of each of these mechanisms and their implementation is described in S1 Supporting Protocol . We fit each of the 64 potential model variants in a Bayesian framework using parallel-tempering Markov chain Monte Carlo ( PT-MCMC ) to estimate the posterior medians and 95% credible intervals ( CI ) of all free model parameters . For each model , we ran 3 chains each for 5000000 iterations . Where the effective sample size ( ESS ) was <200 or the Gelman-Rubin diagnostic ( R ^ ) was <1 . 1 for any estimated parameter ( calculated using the coda R package [63] ) , we ran 5 chains each for 10000000 iterations and obtained upper 95% confidence intervals for R ^ of <1 . 1 for all estimated parameters presented here . ESS and R ^ estimates for all parameters are provided in S5 Table . We then performed a model comparison analysis using Pareto-smoothed importance sampling leave-one-out cross-validation ( PSIS-LOO ) with the loo R package [64 , 65] . Briefly , the purpose of this analysis was to compare the expected log point-wise predictive density ( ELPD ) of different model fits to compare their out-of-sample prediction accuracy . Comparing ELPD estimates serves a similar purpose to comparing other information criteria , where a lower ELPD suggests greater predictive power penalised by model complexity . Results shown in the main text are from the most complex model ( most free parameters ) variant with δELPD<1 compared to the lowest ELPD . Parameter estimates from all model variants with a δELPD<20 are shown in S7–S14 Figs . Posterior parameter estimates are shown as medians and 95% CIs . Further details of the model fitting and comparison are described in S1 Supporting Protocol . All code and data are available as an R package at https://github . com/jameshay218/antibodyKinetics .
To validate our boosting and biphasic waning model for a single exposure , we fit the base model to HI titres against A/Panama/2007/1999 ( H3N2 ) from group E alone ( Fig 3 ) . Ignoring the later exposure to A/Fukushima/141/06 ( H1N1 ) at day 56 , from which we do not expect any cross-subtypic antibody reactivity , these data in isolation reflect a typical antibody trajectory following exposure to a single immunogen and measurement of antibodies against it [34 , 50] . The models with biphasic waning ( both with estimated long term waning rate , m , and fixed m = 0 ) were better supported than the models with monophasic waning or no waning ( ELPD -20 . 6 ( standard error ( SE ) , 3 . 37 ) and -20 . 5 ( SE , 3 . 77 ) compared to -23 . 4 ( SE , 3 . 71 ) and -28 . 1 ( SE , 3 . 48 ) respectively ) , although we note that these differences are small with respect to the standard error of the ELPD estimates . The biphasic waning models with estimated long-term waning m and fixed long-term waning m = 0 had a difference in ELPD of <1 , suggesting that both models had similar predictive performance . Overall , these results suggest that the model with monophasic waning is justified over the version with no waning , and that the biphasic waning model is better still than the monophasic waning model . Posterior estimates for model parameters were: μ = 9 . 91 ( median , 95% CI 7 . 08–12 . 7 ) ; d = 0 . 551 ( median , 95% CI 0 . 183–0 . 695 ) ; ts = 19 . 5 days ( median , 95% CI 6 . 39–27 . 4 days ) ; m = 0 . 0414 ( median , 95% CI 0 . 00405–0 . 103 ) . Overall , ferrets that received more frequent and immunogenic exposures achieved the highest , most broadly reactive and long-lived antibody titres . The full data show substantial variation in observed antibody titres across the groups driven by different exposure types and combinations . Following two doses of unadjuvanted TIV , ferrets achieved only modest increases in titres against the vaccine strains ( Fig 4A ) , with 2 out of 3 ferrets failing to generate H3N2 titres that persisted past day 37 . The addition of an adjuvant resulted in increased and persistent titres against the vaccine strains in all ferrets by day 49 . Titres against A/Fukushima/141/2006 ( H1N1 ) , which is antigenically similar to A/Solomon Islands/3/2006 ( H1N1 ) , were also increased at this time point ( Fig 4B ) . Similarly , priming infection resulted in higher and long-lived titres to the vaccine strains and A/Fukushima/141/2006 ( H1N1 ) relative to ferrets in the unprimed , unadjuvanted TIV protocol ( Fig 4C ) . Observed titres at day 21 against A/Panama/2007/1999 ( H3N2 ) were consistently high following priming infection in groups C-E , with one ferret in each of groups C and E also experiencing some boosting of antibodies against the other H3N2 strains . All ferrets were infected with A/Fukushima/141/2006 ( H1N1 ) at day 56 , leading to elevated titres to both H1N1 strains by day 70 in all ferrets . The top two model variants had ELPD estimates of -412 . 1 ( SE , 21 . 4 ) and -412 . 6 ( SE , 20 . 7 ) respectively . Both of these models included: a role for priming infection in increasing subsequent vaccine response; different boosting profiles between vaccination and infection; different boosting profiles with adjuvant versus without adjuvant; and biphasic antibody waning . The model with the lowest ELPD ( model ID 21 , S6 Table ) had 30 free parameters and also included titre-dependent boosting , no antigenic seniority , and no exposure type-specific cross reactivity . The other model ( model ID 62 , S6 Table ) had 33 free parameters and did not include titre-dependent boosting , but did include antigenic seniority and exposure type-specific cross reactivity . Fig 4 shows the latter ( more complex ) model variant fitted to the data . Parameter estimates for these two models are shown in S4 Table . The remainder of the results refer to the latter model with more free parameters . Other model variants may provide latent titre predictions more in line with biological expectations , though we note that they are not as well justified based on the model comparison analysis . For example , the predicted latent titres at the time of secondary vaccination ( day 42 ) were unexpectedly lower in group D than group C under the chosen model variant ( Fig 4C & 4D ) . Oil-in-water adjuvants are hypothesised to increase recruitment of neutrophils , antibody presenting cells and antigen bearing B cells at draining lymph nodes , and we would therefore expect antibody titres following adjuvanted TIV to be higher than unadjuvanted TIV throughout this time frame [66 , 67] . These unexpected results are likely due to limitations of the flexible model structure , which finds the set of parameter estimates best supported by all of the data , potentially at the cost of some biological realism . For example , a model variant identical to the one used in Fig 4 with the addition of titre-dependent boosting provided waning parameter estimates in line with the expectation of adjuvanted vaccination leading to higher antibody titres at all times relative to unadjuvanted vaccination ( S4 Fig , model ID 54 , S5 Table ) . Overall , ELPD estimates ranged from -412 . 1 ( SE , 21 . 4 ) in the highest ranked model to -543 . 6 ( SE , 21 . 8 ) in the lowest ranked model ( S6 Table ) . The simplest model with 8 free parameters was the third lowest ranked model ( ELPD -539 . 0 ( SE , 21 . 9 ) ) , whereas the most complex model with 35 free parameters was the the 7th highest ranked model ( ELPD -417 . 0 ( SE , 21 . 2 ) ) . We note that some of the simpler model variants may have similar predictive performance to the best fitting model and may therefore be more suitable in a general predictive application . For example , further constraining the antibody trajectories in Fig 4E from day 56 is possible by assuming shared kinetics parameters for the A/Panama/2007/1999 ( H3N2 ) and A/Fukushima/141/2006 ( H1N1 ) infections , as we would expect these trajectories to be similar given that they are both primary exposures to that subtype . One of the fitted model variants ( model ID 64 , S6 Table ) that was identical to the one in Fig 4 but assumed 3 rather than 6 distinct exposure types produced tighter 95% CIs post A/Fukushima/141/2006 ( H1N1 ) infection ( S3 Fig ) . However , this model variant is less well supported based on the model comparison analysis ( ELPD -441 . 7 ( SE , 20 . 3 ) , δELPD = 29 . 6 ) and provided estimates of post-infection waning that were almost identical to those for A/Panama/2007/1999 ( H3N2 ) infection in the 6 exposure type model; overall , suggesting that the A/Fukushima/141/2006 ( H1N1 ) data did not contribute to the posterior estimates ( model ID 64 , S5 Table ) . Our aim was not to predict unseen data but rather to quantify immunological mechanisms . As a crude measure of mechanism importance , we performed Pseudo-Bayesian model averaging ( Pseudo-BMA+ ) to estimate the relative weights of each model variant and thereby weights of models with a particular mechanism relative to models without [68] . Although comparison of variable importance using information criteria must be interpreted with caution ( for example , changing the sample size or experimental protocol may change the results ) , Pseudo-BMA+ serves as a rough estimate of which mechanisms are most important in explaining these data [69] . Variable weights were: 1 . 00 for the presence of priming; 0 . 999 for the presence of 6 exposure types; 0 . 836 for the presence of biphasic waning; 0 . 579 for the presence of titre dependent boosting; 0 . 572 for the presence of type specific cross reactivity; and 0 . 406 for the presence of antigenic seniority . The top two models had Pseudo-BMA+ weights of 0 . 331 and 0 . 303 , with a drop off to the third model with a weight of 0 . 0977 . The top two models included only titre-dependent boosting and antigenic seniority respectively , suggesting that inclusion of at least one of these mechanisms improved predictive performance . The consistency of parameter estimates across the best fitting model variants is demonstrated in S7–S14 Figs . The level of homologous boosting resulting from priming infection ( Infection 1 ) and secondary infection ( Infection 2 ) was similar , shown by similar estimates for μ from both infections ( Fig 5A ) . We inferred that antibody titres fell only marginally following the initial waning phase ( μ ( 1 − d ) , Fig 5B ) . The antibody waning rate was not identifiable for secondary infection due to the lack of observations following this exposure . We found evidence for only low levels of homologous antibody boosting following both initial and secondary doses of unadjuvanted TIV ( TIV 1 and TIV 2 ) that quickly waned to near undetectable levels during the initial waning phase . The addition of an adjuvant appeared to have no significant impact on the homologous antibody response to the first vaccine dose , but did improve the response to a second dose of vaccine ( TIV 1 compared to TIV 1 + adjuvant and TIV 2 compared TIV 2 + adjuvant , Fig 5B ) . Titres against A/Brisbane/10/2007 ( H3N2 ) and A/Solomon Islands/3/2006 ( H1N1 ) were similar following the first unadjuvanted vaccine dose and the first adjuvanted vaccine dose ( TIV 1 compared to TIV 1 + adjuvant , Fig 4A & 4B ) . However , the second adjuvanted TIV dose appeared to elicit a significant persistent boost to the vaccine strains , which resulted in peak titres near the limit of detection of this assay ( TIV 2 compared to TIV 2 + adjuvant , Fig 4A & 4B ) . In models with type-specific cross-reactivity , we found differences in the width of cross reactivity elicited by the 6 exposure types shown in Fig 5 . Secondary infection appeared to elicit a level of cross reactivity in line with that of the priming infection , whereas cross reactivity for both unadjuvanted and adjuvanted vaccination appeared to be narrower and only boosted antibodies that were effective against antigenically similar viruses ( Fig 6 ) . σ describes the degree by which antibody titre decreases as a function of antigenic distance , where higher values of σ suggest lower cross reactive breadth . When a single cross reactivity gradient was assumed for all exposure types ( as in the highest ranked model ) , we estimated the cross reactivity gradient to be 2 . 33 ( median; 95% CI: 1 . 74–3 . 01 ) , suggesting narrower cross reactivity than would be expected given the definition for cross reactivity based on ferret antisera ( an antigenic distance of 1 unit should see a reduction in antibody boosting of 1 log titre unit ) [70] . Fig 6 demonstrates that homologous boosting ( the y-intercept ) was too small to elicit any measurable cross reactive boosting at these antigenic distances . The cross reactivity gradient parameter , σ , could therefore not be identified for the second dose of unadjuvanted TIV and first dose of adjuvanted TIV , and we were only able to recover the prior distribution for these parameters . These values were therefore excluded from Fig 5F . Our model provided support for the presence of an initial short-term , rapid waning phase followed by a secondary long-term , sustained waning phase . For all vaccine doses , we estimated that the majority of the antibody boost waned within two weeks of reaching the peak ( upper 95% CI 17 . 3 , 5 . 15 , 12 . 5 and 2 . 16 days for TIV 1 , TIV 2 , TIV 1 + adjuvant and TIV 2 + adjuvant respectively , Fig 5A & 5B ) . Conversely for priming infection , we estimated that the antibody titre was maintained at near peak levels with an estimated initial waning phase duration of 18 . 9 days ( median; 95% CI 11 . 0–29 . 3 ) and a 21 . 6% ( median; 95% CI: 2 . 02–48 . 8% ) drop in log titre relative to the peak . We estimated similar long-term waning rates for second unadjuvanted TIV , second adjuvanted TIV and priming infection ( Fig 5E ) . We could not produce constrained estimates for the waning phases that take place following infection with A/Fukushima/141/2006 ( H1N1 ) at day 56 , given that only one subsequent observation was made at day 70 . Although the 95% CI does not exclude biphasic waning rates consistent with the other exposures , any single trajectory in this range that passes through the single observation is similarly likely given these data . Prior to receiving non-adjuvanted TIV , experimental group C was infected with H3N2 Panama/2007/1999 at day 0 , which represented a host being primed by natural infection prior to vaccination . Our model allowed us to identify additional homologous and cross-reactive antibody boosting that resulted from priming improving the subsequent vaccine response , as comparable experimental groups were given the same vaccination schedule with or without priming infection . The model suggested that priming infection added a substantial boost ( 7 . 28 log units ( median , 95% CI 5 . 85–8 . 92 ) ) to antibodies against the A/H1N1 and A/H3N2 vaccine strains at the time of vaccination in addition to that provided by the vaccine itself ( Fig 5A ) . Given our assumption that a log titre of 0 represents the true absence of antibodies , we cannot be certain that the higher titres observed at day 37 are due to a single large boost from primed vaccination rather than an antibody boost below the limit of detection from the priming infection followed by a small subsequent TIV boost . However , previous antibody kinetics results showing higher vaccine-induced antibody boosting following priming from the same detectable starting titre suggest that the former explanation is likely [71] . We estimated the cross reactivity of this additional boost to be broad with a gradient of 0 . 882 ( median; 95% CI: 0 . 531–1 . 49 ) , suggesting that priming increases the cross-reactive breadth of the vaccine response . It should be noted that whilst additional priming-induced vaccine boosting is well supported by the model fit , the model overestimates the antibody titre to A/Fukushima/141/2006 ( H1N1 ) at day 37 elicited by initial dose of TIV following priming by H3N2 infection ( Fig 4C ) . This may be a result of subtype-specific interactions that are not captured by our model . Despite the relatively short duration of these experiments , we found some evidence for a trend of decreasing antibody response with increasing number of prior exposures and/or higher pre-exposure titres . In the best fitting model with antigenic seniority , we estimated τ to be 0 . 213 ( median; 95% CI: 0 . 134–0 . 300 ) , suggesting that antibody boosting decreased substantially with increasing number exposures after taking into account exposure type and priming . τ measures the proportion of the full boost that is lost with each successive exposure experiences relative to the first ( ie . boosting decreases linearly as a function of increasing prior exposures ) . A higher value of τ therefore indicates more boosting suppression with an increasing number of prior exposures . Based on these estimates , the amount of antibody boosting would be reduced by over 50% following 4 exposures . In the best fitting model with titre-dependent boosting , we estimated the titre-dependence gradient γ to be 0 . 0898 ( median; 95% CI: 0 . 0788–0 . 102 ) applying to all titres below 10 . 9 ( median; 95% CI: 8 . 62–11 . 95 ) . γ gives the proportion of full boost that is lost per unit increase in log titre at the time of exposure ( with no suppression from a starting log titre of 0 ) . The full posterior estimate for the titre-dependent boosting mechanism is shown in S5 Fig . We note that the inferred titre-dependent boost relationship may be different if the limit of detection of the HI assay was lower . The top two model variants incorporated one of antigenic seniority or titre dependent boosting , suggesting that either one significantly improves model fit relative to the model variants with neither . The two mechanisms are correlated in these experiments , and antigenic seniority was not well identified with estimates for τ that did not exclude 0 for models with both titre-dependent boosting and antigenic seniority . However , all of the top models with antigenic seniority but no titre-dependent boosting give constrained estimates for τ away from 0 ( S14 Fig ) .
In this study , we used a mathematical model of antibody kinetics to describe boosting and waning following influenza vaccination or infection in a group of well characterised ferrets . We fit various subsets of the model with different immunological mechanisms and found that the two best supported models both included: type-specific antibody boosting; type-specific biphasic waning; 6 distinct exposure types; and a role for priming in increasing subsequent vaccine response . Antigenic seniority , antigenic distance-mediated cross reactivity specific to each exposure type and titre-dependent boosting were also included amongst these top models , suggesting that these mechanisms may be important in accurately describing observed antibody titres following multiple exposures . We found quantitative differences in the level of homologous and cross-reactive antibody boosting between vaccination , infection and adjuvanted vaccination in this ferret model . A single TIV dose with or without adjuvant elicited negligible levels of homologous and cross reactive boosting . A second dose of TIV with adjuvant resulted in significant , broadly reactive antibody boosting , whereas a second dose of TIV without adjuvant did not elicit significant antibody boosting . The profile of boosting for primary infection was consistent across experimental groups , and similar in magnitude to secondary infection . Furthermore , we found that priming infection induced a significantly broader and stronger boosting profile following subsequent vaccination . Our work has a number of limitations . The model predicted latent antibody titres were broadly in line with expected immune dynamics , though there were some exceptions . The model variant presented in the main text was selected based on the optimal balance between fewest parameters and accuracy of predicted antibody trajectories with respect to the full set of observed titres . Other model variants or designs may provide results more in line with biological expectations , but would require estimation of more parameters or collection of additional data . Given the relative sparsity of samples across time , some aspects of the model were poorly identified or included spurious features for some subsets of the data . In particular , sampling around the biphasic waning period of the vaccinations and following the final exposure event was limited , resulting in poor identifiability for some of the waning and timing parameters . We therefore restricted our reported results to estimates that were consistent across the best supported model variants ( S7–S14 Figs ) . Experiments of a similar design with fewer exposures and more frequent sampling would power the model to elucidate these waning phases further and look for differences in response longevity by exposure type . Our experimental timeline was much shorter than the typical human exposure timescale of months to years , with a minimum gap of 14 days between the two vaccine doses and 28 days between infection and vaccination [18] . No further increase in antibody titre was detected from 14 days post TIV 1 in group A or post A/Panama/2007/1999 ( H3N2 ) infection in group E , suggesting that serum antibody titres consistently peaked within 14 days and therefore before each subsequent exposure . Furthermore , germinal centre ( GC ) structures and GC-derived ASCs had likely developed within this time , as it has been shown in other small mammals ( mice ) that GC B cells are present within 14 days post infection [39 , 72 , 73] . However , given that GC responses peak after 4 weeks and persist for months in mice , there may not have been sufficient time for the MBC population , which is a significant contributor to post-vaccination ASCs in humans [23] , to fully develop . Recruitment of naïve B cells may also be impacted by both the presence of pre-existing GC structures from primary infection and the inclusion of IFA [66 , 67 , 74] . The inferred antibody kinetics of the two TIV doses might therefore be different if they were further apart . It is possible that the time to peak response parameter would also differ depending on the relative contributions of memory and de novo ASCs; however , models that allowed the time to peak parameter to vary were less well supported based on ELPD , resulted in reduced identifiability of some parameters , and did not change estimates for identifiable parameters ( S1 Supporting Protocol ) . We captured the impact of immune memory on subsequent antibody responses in two ways: antigenic seniority and titre-dependent boosting [26] . Titre-dependent boosting was a function of only homologous antibody titres , whereas antigenic seniority was assumed to be a function of all previous exposures regardless of exposure strain , subtype and administration route . If antigenic seniority is primarily a result of HA reactive antibodies , then it would only act within a subtype in contrast to our assumed mechanism [32] . However , it is not clear that immune memory effects do not extend between influenza subtypes [75 , 76] . For example , immune imprinting to a particular subtype has been proposed as an explanation for age-specific mortality in the 2009 H1N1 pandemic and to explain the age distribution of avian H5N1 and H7N9 cases , though these observations do not necessarily suggest any cross-subtype or cross-group impact on subsequent antibody responses [30 , 77] . It has been proposed that cross-subtype effects might act through cross-reactive memory T-cell responses that act to deplete heterosubtypic antigen load , which may in turn lead to lower antibody boosting [24] . HI assays measure only the aggregated activity of polyclonal antibodies targeting multiple epitopes on the haemagglutinin head , and we were therefore unable to investigate epitope- or B cell-specific contributions to serum antibody titres [78] . Understanding the immunological mechanisms of imprinting effects , short-term kinetics due to GC overlap , and the relative contributions of MBC derived and de novo antibody boosting would require data either on epitope-specific antibodies or single-cell assays [23 , 39 , 79] . Epitope masking , wherein pre-existing antibodies targeting recognised , but poorly conserved epitopes ( ie . the epitopes that generate cross-reactivity in the HI assay between eg . A/Panama/2007/1999 ( H3N2 ) and A/Wisconsin/67/2005 ( H3N2 ) antibodies ) sterically mask access to previously unseen and conserved , immunosubdominant epitopes ( eg . at the receptor binding site ) , has been proposed as a mechanism for the recall bias of subsequent responses [19 , 31 , 80–82] . A model that captures the contribution of MBCs and naïve B cells to antibodies targeting a variety of epitopes may explain how immune imprinting contributes to observed antibody titres , and may include additional insights such as decreased cross-reactive breadth and magnitude with each repeated exposure . Our data included only trivalent vaccination with and without IFA , and estimates of any boosting parameters are therefore conditional on the presence of three antigens in a single vaccination . It would be interesting to compare the inferred homologous and cross-reactive boost of different vaccination strategies ( eg . a three antigen TIV compared to a monovalent vaccine , or comparison by inoculum dose ) , and for adjuvants more relevant to human vaccination such as MF59 [83 , 84] . There may also be underlying heterogeneities in antibody response between and within influenza subtypes as well as between vaccine types [85 , 86] . For example , Live Attenuated Influenza Vaccines ( LAIV ) , as well as newer DNA vaccines may provide different antibody kinetic profiles and may elicit broader antibody responses , or provide different priming effects [58 , 87] . We found evidence for biphasic waning following both primary infection and secondary vaccination . There was some evidence that the magnitude and duration of waning differed between exposure types: TIV 1 and 2 and adjuvanted TIV 1 waned very quickly , whereas Infection 2 and TIV 2 + adjuvant were more persistent . Heterogeneity in antibody waning rates between individuals and vaccine types have been shown for other pathogens [60 , 88] . Although studies of influenza antibody response duration have been carried out in humans , quantifying waning rates independent of subsequent exposures that cause repeated boosting is difficult [2 , 61 , 89–91] . Our model fit to the single exposure ferrets provides an estimation of the waning rate of homologous antibodies in the absence of further exposure , but the cut off of 70 days limits the applicability of this waning rate to a timescale more relevant to humans . Extrapolating our estimated waning rate following primary infection would suggest that antibody titres would wane to non-detectable within a few months , whereas antibody responses against many viruses are known to persist for decades [88 , 92] . Longer term studies investigating the longevity of the antibody response in the absence of repeated exposure would be useful to quantify a long-term , steady state antibody waning rate [52] . Further mechanisms such as differential waning rates between cross-reactive and homologous antibodies are likely to be important , but were not identifiable here [21 , 59] . Although animal models are potentially useful , identification of these mechanisms in human populations is likely possible given long-term , frequent sampling of human sera combined with robust statistical methods [18 , 22] . Our results have implications for comparing different vaccination strategies . Achieving high HI titres against currently circulating strains is a key endpoint in influenza vaccine trials due to its correlation with clinical protection [4 , 93–95] . However , there are a number of obstacles to achieving these high titres in some populations including antigenic interactions , age specific responses and antibody waning [11 , 96–99] . One approach to improving vaccine effectiveness may therefore to elicit a broader antibody response to compensate for potential strain mismatch [100] . Adding adjuvants such as MF59 and AS03 has been shown to induce higher antibody titres that have greater cross-reactive properties [55 , 56 , 101 , 102] . Quantitative comparisons of cross reactivity profiles , as we have provided here , could be a useful tool in comparing the effectiveness of different adjuvants , which would provide a measurable benefit to trade-off against safety and immunogenicity concerns [103 , 104] . In addition to modelling boosting suppression due to prior immunity , we considered potential enhancement via priming infection . “Prime-boosting” has been described previously as a strategy to induce broadly reactive immune responses that may be rapidly boosted in advance of exposure to an antigenically novel virus [53 , 54 , 71] . Models that included a priming mechanism were ranked systematically higher in our model comparison analysis than those that did not , suggesting that this phenomena is important in explaining titres arising from repeated exposures . We found that vaccine responses to A/H1N1 strains were higher and more broadly reactive in A/H3N2 primed ferrets compared to unprimed ferrets , though our model did not account for subtype specific interactions and subsequently overestimated post vaccination A/H1N1 titres in primed ferrets . Although the phylogenetic relationship between the priming and subsequent boosting strain is likely to be important , heterosubtypic protection has been shown previously in animal models , potentially via cytotoxic T lymphocyte responses [45 , 47] . Our results suggest that mathematical models of antibody kinetics that explicitly consider immunological mechanisms and exposure-type specific parameters would be valuable for the prediction of antibody landscapes in human populations . Human cohort studies tracking infants from birth as they experience their first few influenza exposures are also now underway [105] . Combining these studies with single-cell immune profiling and mathematical models of multiple exposure kinetics will help to elucidate the role of these immunological mechanisms in building human antibody profiles . Direct inference from long-term observational data in humans may be difficult , but experimental models , such as the ferret system described here , provide an excellent alternative data source for the inference of short-term immunological mechanisms that may map onto models recovered using human sera [18 , 21 , 41 , 42] . | Despite most individuals having some preexisting immunity from past influenza infections and vaccinations , a significant proportion of the human population is infected with influenza each year . Predicting how an individual’s antibody profile will change following exposure is therefore useful for evaluating which populations are at greatest risk and how effective vaccination strategies might be . However , interpretation of antibody data from humans is complicated by immunological interactions between all previous , unobserved exposures in an individual’s life . We developed a mathematical model to describe short-term antibody kinetics that are important in building an individual’s immune profile but are difficult to observe in human populations . We validated this model using antibody data from ferrets with known , varied infection and vaccination histories . We were able to quantify the independent contributions of various exposures and immunological mechanisms in generating observed antibody titres . These results suggest that data from experimental systems may be included in models of human antibody dynamics , which may improve predictions of vaccination strategy effectiveness and how population susceptibility changes over time . | [
"Abstract",
"Introduction",
"Materials",
"&",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"influenza",
"immunology",
"vertebrates",
"animals",
"mammals",
"vaccines",
"preventive",
"medicine",
"immunologic",
"adjuvants",
"infectious",
"disease",
"control",
"antibodies",
"vaccination",
"and",
"immu... | 2019 | Characterising antibody kinetics from multiple influenza infection and vaccination events in ferrets |
The carbon storage regulator protein CsrA regulates cellular processes post-transcriptionally by binding to target-RNAs altering translation efficiency and/or their stability . Here we identified and analyzed the direct targets of CsrA in the human pathogen Legionella pneumophila . Genome wide transcriptome , proteome and RNA co-immunoprecipitation followed by deep sequencing of a wild type and a csrA mutant strain identified 479 RNAs with potential CsrA interaction sites located in the untranslated and/or coding regions of mRNAs or of known non-coding sRNAs . Further analyses revealed that CsrA exhibits a dual regulatory role in virulence as it affects the expression of the regulators FleQ , LqsR , LetE and RpoS but it also directly regulates the timely expression of over 40 Dot/Icm substrates . CsrA controls its own expression and the stringent response through a regulatory feedback loop as evidenced by its binding to RelA-mRNA and links it to quorum sensing and motility . CsrA is a central player in the carbon , amino acid , fatty acid metabolism and energy transfer and directly affects the biosynthesis of cofactors , vitamins and secondary metabolites . We describe the first L . pneumophila riboswitch , a thiamine pyrophosphate riboswitch whose regulatory impact is fine-tuned by CsrA , and identified a unique regulatory mode of CsrA , the active stabilization of RNA anti-terminator conformations inside a coding sequence preventing Rho-dependent termination of the gap operon through transcriptional polarity effects . This allows L . pneumophila to regulate the pentose phosphate pathway and the glycolysis combined or individually although they share genes in a single operon . Thus the L . pneumophila genome has evolved to acclimate at least five different modes of regulation by CsrA giving it a truly unique position in its life cycle .
The Gram negative , environmental bacterium Legionella pneumophila is proliferating in aquatic environments where it parasitizes in fresh water protozoa [1–3] . When contaminated water is aerosolized , mainly within man-made devices and installations , L . pneumophila can gain access to the human lung and cause a severe pneumonia called Legionnaires’ disease [4] . The capacity of this environmental bacterium to cause disease in humans evolved from the interaction with aquatic amoebae , as the same strategies used for persistence in protozoa also allow this pathogen also to replicate within alveolar macrophages [5 , 6] . In amoeba as well as in human macrophages the L . pneumophila life cycle consists of two distinct stages , a replicative form that proliferates when nutrients are available and a transmissive or virulent form that is able to escape from the spent host when nutrients are exhausted and to infect a new host cell [7 , 8] . In the transmissive form traits like virulence , motility and resistance against several stress factors are induced , whereas these are typically repressed during replication [8 , 9] . A key regulator of the switch between replicative and transmissive L . pneumophila is the RNA-binding protein CsrA [10 , 11] . CsrA is a global , posttranscriptional regulator of gene expression in many bacteria where it plays important roles in regulating motility , virulence and metabolism [12] . To fulfill its regulatory role , CsrA binds to the 5’ untranslated region ( 5’ UTR ) or in the start region of the coding sequence of the mRNA of its target genes . CsrA modulates translation , and alters mRNA turnover and/or transcript elongation [12 , 13] . The current model of the L . pneumophila life cycle regulation is that starvation of amino acids and altered fatty acid biosynthesis lead to the production of ( p ) ppGpp and subsequently the activation of the two-component system ( TCS ) LetA/LetS and the alternative sigma factor RpoS [14 , 15] . Both promote the transcription of the small non-coding RNAs RsmX , RsmY and RsmZ , which in turn bind and sequester CsrA leading to the expression of transmissive and repression of replicative traits [16–18] . The letA/S-mutants instead are non-motile , less pigmented , sodium stress resistant , but oxidative stress sensitive and also show a reduced infectivity for A . castellanii [16–18] . Analyses of a strain overexpressing CsrA or a conditional csrA-mutant revealed that CsrA represses typical post exponential ( PE ) , transmissive phenotypes of L . pneumophila such as cell shape shortening , pigmentation , motility , sodium sensitivity and cytotoxicity [10 , 11 , 19] . Additionally , the quorum sensing regulatory system LqsTS/LqsR and the TCS PmrB/PmrA regulate CsrA activity [16 , 20–23] . In Legionella , PmrBA was shown to regulate the expression of several Dot/Icm effector proteins and positively regulates the transcription of csrA [16 , 24] . Different studies have reported indirect evidence linking L . pneumophila CsrA and virulence by identifying putative CsrA binding motifs in the mRNAs of secreted Dot/Icm effectors , or analyzing CsrA overexpressing strains [10 , 11 , 16 , 25] . However , the direct targets of CsrA and whether these are regulated by the classical regulatory mechanism described for CsrA or not , are not known . By using transcriptomics , proteomics , RNA-Immunoprecipitation followed by deep sequencing ( RIPseq ) , together with biochemical , phenotypical and molecular analyses we identified the L . pneumophila CsrA targets genome wide and discovered a new mode of action of CsrA that allows to regulate genes comprised in the same operon , independently .
To study the regulatory consequences of CsrA in detail , we constructed a mutant csrA- by inserting an apramycin-resistance cassette after the amino acid Tyr48 of the lpp0845 gene encoding the major CsrA in L . pneumophila Paris ( S1A Fig ) . CsrA is essential for L . pneumophila , but such a truncated CsrA variant has a strongly reduced expression of CsrA ( S1B Fig ) , similarly to what was reported for Escherichia coli [26 , 27] , possibly by immediate degradation of a miss folded protein due to the distorted C-terminal helical structure . Furthermore , it has been reported that mutations leading to a dislocation of the alpha-helix are totally devoid of biological activity [28] , thus allowing to study the RNA targets of CsrA . Indeed , when analyzing flagelllin expression , which is a hallmark of the PE-phase in L . pneumophila it was expressed already in E-phase in the csrA- strain indicating that CsrA mediated repression in E phase was released ( S1C Fig ) . Phenotypic analyses of the mutant showed that in contrast to the wt strain the csrA- strain was motile already in E phase , as judged by microscopic observation of actively moving bacteria . Furthermore it showed a significantly higher pigment production and sodium sensitivity , a phenotype that reflects activity of the Dot/Icm T4SS system ( S1D and S1E Fig ) . Furthermore , the csrA- strain was more resistant to oxidative stress ( S1F Fig ) and showed increased tolerance against moderate acidification ( pH 4 . 8 ) as compared to the wt ( S1G Fig ) . Thus , a csrA mutant shows clear evidence of a transmissive/virulent phenotype already during exponential growth ( replication ) . To evaluate the impact of CsrA on intracellular replication we compared its growth within Acanthamoeba castellanii to that of the wt strain , a ΔletA and a ΔrsmY/Z double mutant [17] . As shown in Fig 1 , mutation of the csrA locus has a drastic effect on the overall replication capacity ( 100x less than wt ) comparable to what is observed for the RsmYZ and LetA mutants . However , the ΔletA and ΔrsmYZ strains exhibit both a strong defect in the entry and the initial stages of infection ( 1h ) but once established in the host cell , they show similar replication efficiency as the wt strain during the first infection cycle ( between 1h and 24h ) , but are not replicating anymore during the second infection cycle ( >24h ) . In contrast , the entry of the csrA- and the wt strain is identical , but then replication is significantly diminished in absence of CsrA , in particular during the first infection cycle up to 15h post infection and at the end of the second infection cycle ( Fig 1 ) . Similar results were observed for a conditional csrA-strain when infecting bone-marrow-derived macrophages [11] . Thus the ΔletA and ΔrsmYZ strains seem to be halted in the non-virulent , replicative phase and are not able to efficiently infect amoeba [17 , 29] whereas the csrA- strain seems to be forced into a premature transmissive phenotype with a strong deficit in the reconversion from the virulent into the replicative stage . To assess the influence of CsrA on gene expression , we compared the transcriptome and the proteome profile of the wt and the csrA- strain during exponential growth phase using whole genome microarrays [9] and mass spectrometry-based shotgun proteomics ( LC-MS/MS ) . Transcriptome analyses revealed that 431 genes showed different RNA levels due to the loss of CsrA , of which 236 were significantly upregulated and 195 significantly downregulated in the csrA- strain as compared to the wt strain , when a 1 . 5 fold change in gene expression was taken as cut-off ( S1A Table ) . In contrast , at post-exponential growth only 6 genes showed a significant different transcript level between the wt and csrA- strain ( S1B Table ) . Interestingly , among them are csrA ( upregulated ) and the three small ncRNAs rsmX , rsmY and rsmZ , ( downregulated ) , giving a first indication that CsrA may be involved in controlling its expression . The proteome analysis by LC MS/MS of the mutant and the wt in E phase identified 1662 proteins in total of which 1448 could be quantified . Cluster analysis showed that expression of 1353 out of the 1448 proteins was affected by the mutation of CsrA with about half of the identified proteins up- and half downregulated ( Fig 2 ) . About 15% of the proteome ( 216 proteins ) was strongly affected with 131 proteins significantly up- and 85 significantly downregulated in the csrA mutant ( S2 Table ) . One of the most striking differences between the wt and the csrA- strain is the timing of flagella expression as demonstrated by flagelline expression already in E phase ( S1C Fig ) . Furthermore the loss of CsrA leads to motility of the bacteria already in E phase as judged by microscopy . Indeed , 32 genes of the flagellar biosynthesis gene cluster were significantly up-regulated in the csrA- strain already in E phase . Among those are FlaA ( lpp1294 ) , the sigma factor FliA ( lpp1746 ) and the response regulator FleR ( lpp1726 ) ( S1 Table ) . FlaA is the most abundant structural protein of the L . pneumophila flagellum , whereas FliA and FleR , together with RpoN ( lpp0542 ) and FleQ ( lpp0915 ) , are the master regulators for flagellar assembly [30] . In line with these results , the major regulator of flagella biosynthesis FleQ together with other flagella proteins was also significantly up-regulated in our proteomic data ( S2 Table ) . Thus our results confirm that CsrA impacts flagella biosynthesis negatively during replication and shows that CsrA-mediated repression must be relieved to activate motility . Similar to E . coli and other bacteria where CsrA was studied in more detail , our microarray and proteomic data showed that CsrA regulates the central carbon flux in L . pneumophila . CsrA positively impacts the pyruvate/2-oxoglutarate dehydrogenase complex ( lpp1515-lpp1517 ) and proteins that participate in the Entner-Dudoroff ( lpp0483-lpp0487 ) pathway , an alternative pathway to catabolize glucose to pyruvate . Additionally , the transcript of the glycolysis protein triosephosphate isomerase ( lpp2838 ) is down-regulated in csrA- strain . Furthermore , the ribose-5-phosphate isomerase A ( lpp0108 ) of the pentose phosphate pathway , the pyruvate dehydrogenase complex ( lpp1461 ) and the 2-oxoglutarate dehydrogenase ( lpp0597 ) of the TCA cycle were down-regulated at the protein level in absence of CsrA ( S1 and S2 Tables ) . In late stages of the infectious cycle L . pneumophila contains poly-β-hydroxybutyrate ( PHB ) granules [31 , 32] . PHB is an important carbon and energy storage polyester produced under nutrient-limited conditions in numerous microorganisms [33] . The regulation of this pathway in L . pneumophila is not known . We found that two genes that are part of the poly-3-hydroxybutyrate biosynthesis pathway , the acetoacetyl-CoA reductase ( lpp0621 ) and the 3-hydroxybutyrate dehydrogenase ( lpp2264 , bdhA ) are induced on transcript and protein level in the absence of CsrA . Expression of bdhA was shown to be dependent on RpoS [34] . As RpoS is slightly up regulated in the csrA- additional , indirect influence of RpoS cannot be excluded . Thus , CsrA seems to play a vital role in regulating the carbon metabolism in particular for the decision whether pyruvate is metabolized for energy and metabolite production or if it is transferred into its storage compound . Finally , a prominent feature of a csrA- mutant is its reduced intracellular growth ( Fig 1 ) . The reason for this phenotype is clearly reflected in the transcriptome and proteome data , as at least 40 substrates of the Dot/Icm secretion system are differentially expressed between the wt and the csrA- strain ( S1 and S2 Tables ) . Among them the Sid family effectors SidE , SidJ , SidK , SdbC , SdeD , of which e . g . the SidE family proteins have recently be shown to ubiquitinate multiple Rab small GTPases associated with the endoplasmic reticulum [35 , 36] . Furthermore , the expression of the effectors RavA , RavH , MavH , MavT , MavQ , LepA , LepB , VipE , PieF and YlfB is influenced by CsrA . For some of these ( e . g . RavH , MavT , MavQ , YlfB ) it has been shown previously that their expression is under the direct control of the LetA/Rsm/CsrA regulatory cascade and that the corresponding transcripts might contain CsrA-regulatory elements mainly in the 5'UTR/RBS regions suggesting a regulation by CsrA [16 , 25] . Eighteen of the Dot/Icm substrates differentially expressed in the transcriptome were also differentially expressed in the proteome ( S1 and S2 Tables ) . An example is the eukaryotic-like sphingosine-1-phosphate lyase ( LpSPL , Lpp2128 ) that targets the sphingolipid metabolism of the host cell to restrain autophagy [37] . Taken together , genome wide transcriptome and proteome analyses showed that CsrA activity has a major impact on motility and the central carbon metabolism . Furthermore an important role of CsrA in virulence and stress response was clearly seen by the differential expression of many Dot/Icm secreted proteins and the intracellular replication defect of a the csrA- strain . Our proteome and transcriptome data show that CsrA influences the expression of many major regulatory proteins like FleQ and FleR , the alternative sigma factor RpoS ( lpp1247 ) , the nucleoid-associated proteins Fis2 ( lpp1324 ) , Fis3 ( lpp1707 ) and HU-beta ( lpp1826 ) or the transmission trait enhancer protein LetE . As it is well known , transcriptome and proteome data overlap only partially and they do not allow distinguishing between direct or indirect regulations . Thus we analysed the direct interaction between CsrA and its target-RNAs by co-immunoprecipitation experiments followed by massive sequencing ( RIPseq ) . Five independent RIPseq libraries obtained with epitope-tagged CsrA protein were created and deep sequencing of CsrA-bound transcripts using an Illumina platform was performed . We identified in total 479 RNAs localized in the untranslated ( UTR ) and/or coding regions of mRNAs or of known non-coding sRNAs of L . pneumophila ( S3 Table ) . To identify CsrA targets , we used a script developed by Dugar and colleagues [38] , that calculates in sliding windows the coverage enrichment of the co-IP versus the control . For the comparison , the coverage files were normalized according to the number of mapped base pairs of each sample ( control and co-IP ) . A peak was defined as an at least five times sequence enrichment in the co-IP as compared to the control IP . The values for the enrichment of each CsrA target are recorded in S3 Table . Calculation of the enrichment of A ( N ) GGA motifs in the CsrA target peaks defined in the co-IP as compared to those found in the control IP revealed a 1 . 24 to 2 . 94 enrichment of GGA motifs in the peaks of the co-IP ( S4 Table ) . Epitope tagged CsrA was expressed about 2 times more than the native copy ( S2 Fig ) thus few RNAs that were bound at low affinity , but are not true targets might have been included . When comparing results obtained from the protein and transcriptome data the correlation factor was R2 = 0 . 314 , a factor comparable with data from other bacteria where a correlation of 0 . 20–0 . 47 was reported [39] . The RIPseq , transcriptome and proteome data combined showed a concordance of 32% . S5 Table reports the 51 targets that were identified in all three approaches . Table 1 summarizes the distribution of the proteins encoded by these CsrA targeted RNAs according to their functional category and Table 2 shows all targets that are discussed here in detail . As suggested from the RIPseq , proteome and transcriptome data combined , CsrA affects all major metabolic pathways like the carbon , amino acid and fatty acid metabolism as well as energy transfer , transport and uptake of nutrients or response . Furthermore , the mRNA of many proteins related to the biosynthesis of cofactors , vitamins and secondary metabolites including thiamine , pyridoxal , inositol phosphate , S-adenosylmethionine or riboflavin are directly interacting with CsrA ( Table 1 and S3 Table ) . Secondly , CsrA seems to have a major influence on translational and transcriptional processes as well as DNA replication and repair as judged from its interaction with numerous RNAs of proteins of these functional groups . Thirdly , CsrA directly binds to RNAs of proteins implicated in virulence , stress response and adaptation to environmental changes ( Table 1 and S3 Table ) . RIPseq analyses identified two RNAs directly bound to CsrA , one located in the 5’UTR of the mRNA coding for the transcriptional regulator FleQ and the other one in that coding for the two-component response regulator FleR . Both regulators are indispensible for flagella biosynthesis in L . pneumophila as the deletion of one or the other led to the down regulation of all flagellar genes and consequently to a complete loss of motility [30] . Indeed , EMSA assays confirmed the interaction of the fleQ mRNA with CsrA in vitro , and the addition of an excess of unlabeled RsmZ as control , abolished this interaction ( Fig 3A ) . To identify the binding sites of CsrA we predicted the RNA structure with the Mfold program [40] . This revealed two A ( N ) GGA-binding motives in the fleQ mRNA that are present in loops of ( Fig 3B ) . To analyze whether these were indeed the CsrA binding sites , we mutated the FleQ1 AAGGA-loop motif to AAAAA and the FleQ2 AGGA-loop motif to AAAA ( Fig 3B ) . The mutation of the FleQ2 motif had only little consequences whereas mutation of FleQ1 led to a partial loss of the interaction . However , mutation of both sites completely abolished binding of CsrA ( Fig 3C ) . The existence of two binding sites might reflect an independent binding of one or two CsrA proteins to the FleQ-mRNA with different affinities or could also indicate a serial interaction of the homodimeric CsrA with both loops initiating at FleQ1 . The motif is overlapping with the ribosomal binding site ( RBS ) of the fleQ gene , indicating that CsrA has a negative effect on FleQ translation by preventing ribosome binding , which is the most predominant operational mode of CsrA [12 , 13] . This result was further substantiated by the result of BlaM reporter assays where the potential CsrA-binding region identified by RIPseq was fused upstream of the BlaM gene ( S3A Fig ) . Indeed , in absence of CsrA or mutations of the CsrA-binding motifs resulted in a higher ß-lactamase activity due to higher expression of the BlaM protein ( S3A Fig ) . Furthermore , this results are in agreement with the phenotypic observation that flagella biosynthesis is dependent on CsrA and our proteomic data , in which FleQ as well as FleR are significantly up-regulated in the csrA- strain ( S2 Table ) . Thus , the regulatory function of CsrA seems to be exerted upstream the regulatory cascade for flagellar biosynthesis by directly preventing the efficient translation of FleQ and FleR during replication . L . pneumophila produces a quorum sensing autoinducer molecule , 3-hydroxypentadecane-4-one ( LAI-1 ) , that is synthesized by LqsA and sensed by the two sensor kinases LqsS and LqsT , which subsequently phosphorylate the response regulator LqsR [20 , 41 , 42] . Expression of LqsR was shown to be dependent on RpoS and LetA [22] . LqsR-knock out mutants are defective in the formation of the Legionella containing vacuole ( LCV ) and in the replication in amoeba and macrophages [23] . Here we show that LqsR is linked to the LetA/Rsm-regulatory cascade through the binding of CsrA to the RBS region of LqsR ( Fig 4A ) . A clear band shift in presence of increasing concentrations of purified CsrA was observed whereas adding a surplus of unlabeled RsmZ outcompeted the interaction between CsrA and the biotin-labeled lqsR-mRNA ( Fig 4A ) . Mfold analyses predicted two binding loops containing an A ( N ) GGA regulatory motif . Mutating one of these motives had nearly no influence on CsrA-binding whereas the mutation of both led to a complete loss of the interaction of CsrA with the 5’UTR region of the LqsR mRNA ( Fig 4B ) . This could be the consequence of independent binding of CsrA to one or the other motif , or analogous to FleQ , might indicate a cooperative binding to both loops that would act equivalently in this case . Yet , both binding sites together might not be necessary to ensure efficient CsrA interaction , but may increase the affinity of CsrA to the LqsR mRNA . We also undertook BlaM reporter assays that showed that the mutation of the CsrA-binding loci led to elevated BlaM expression ( S3B Fig ) hence further supporting our model in which CsrA negatively regulates the LqsR expression most likely on the translational level . Thus the two major regulators CsrA and LqsR act complementary , with CsrA governing the transition to transmissive phase whereas LqsR facilitates the switch from transmissive phase back to replicative phase . This direct connection between CsrA and LqsR uncovers the missing link between the stringent response pathway and the response to local population density via quorum sensing . Strikingly , in our RIPseq analyses we identified RelA ( Lpp1413 ) , RpoS ( Lpp1247 ) , PmrA ( Lpp1255 ) and LqsR ( Lpp2788 ) as targets of CsrA . These four regulators have been shown to be major players in the switch from replicative to transmissive/virulent L . pneumophila [15 , 16 , 23 , 24] . Indeed , in vitro interaction assays with CsrA and the mRNAs of LqsR , RpoS and RelA , respectively confirmed the interaction of these mRNAs with CsrA ( Fig 5A and 5B ) . The putative CsrA-binding sites for these regulators were all located in the translation initiation region suggesting a negative regulation of the translation , a model that is in line with the results of the BlaM reporter assays for LqsR and RpoS ( S3B and S3C Fig ) . Surprisingly we identified in our proteomic analyses RelA down-regulated in the csrA- mutant ( S2 Table ) , a finding that was also supported by the results of the BlaM reporter assay ( S3D Fig ) . Thus although , the CsrA-binding site is overlapping the RBS/start codon of RelA , we cannot exclude that the presence of CsrA might have also an auxiliary effect on RelA translation . However , due to the complexity of the network and the fact that CsrA directly regulates the expression of a high number of major regulators in Legionella , like LetE ( S4 Fig ) , the three Fis proteins , HU-beta or RpoH ( Table 2 and S3 Table ) , also secondary regulatory effects are captured by our proteomic data . Additionally , we discovered that CsrA might bind also to its own mRNA . To test whether CsrA might also bind to the mRNA of Hfq , another major RNA binding protein of L . pneumophila that was not identified in our RIPseq data but whose mRNA contains also GGA motifs , we tested in vitro binding by EMSA assays . In accordance with our RIPseq results , no interaction was found , indicating that CsrA is not influencing the Hfq regulation regulatory pathways ( S5 Fig ) . Taken together , our data suggest that several feedback loops coordinate the different signals derived from the stringent response and from quorum sensing and an auto-regulation to fine-tune CsrA activity itself exists in L . pneumophila ( Fig 5C ) . Moreover , CsrA regulates the switch to virulent L . pneumophila not only by targeting the major regulators , but also by directly targeting secreted effector proteins . In total , 41 of the known Dot/Icm effectors are directly targeted by CsrA according to our RIPseq analyses of which 32 are also found to be differentially expressed in the transcriptome and/or proteome approach ( S1 , S2 and S3 Tables ) . To substantiate this finding , we tested the interaction of CsrA with the mRNAs of three of them by EMSA analyses ( S6 Fig ) . This confirmed that the eukaryotic ectonucleoside triphosphate diphosphohydrolase ( ecto-NTPDase ) ( Lpp1033 ) required for optimal intracellular replication [43] , LidA , ( Lpp1002 ) implicated in Rab1 sequestration and development of the LCV [44] , and YlfA ( Lpp2264 ) that inhibits endosomal trafficking [45 , 46] , are all directly targeted as purified CsrA showed clear binding in vitro to the corresponding YlfA- , LidA- and NTPase-RNAs ( S6 Fig ) . Interestingly , as judged from the combined approach of total protein mass spectrometry analyses and RIPseq experiments , the regulatory effect of CsrA is not only repressive . For example the Dot/Icm secreted eukaryotic-like sphingosine-1-phosphate lyase LpSpl [37] and others are downregulated in the csrA- strain in our proteome and transcriptome dataset ( S1 and S2 Tables ) , indicating that the effect of CsrA allows a highly coordinated life cycle regulation in a spatial and timely manner . Taken together , CsrA regulates virulence formation via two routes , first by regulating several major regulatory proteins and secondly by directly interacting with the transcript of over 40 Dot/Icm-secreted effector proteins . Another CsrA-binding site was located within the operon lpp0151-lpp0154 , encoding the transketolase ( Lpp0154 , Tkt ) , the glyceraldeyd 3-phosphate dehydrogenase ( Lpp0153 , Gap ) , the phosphoglycerate kinase ( Lpp0152 ) and the pyruvate kinase ( Lpp0151 ) one of the rate determining enzymes of the glycolysis ( Fig 6A ) . The transcriptional start site ( TSS ) of this operon lies upstream of lpp0154 as determined by transcriptional start site mapping using dRNAseq [18] . Even though CsrA-binding was seen at several distinct regions inside the operon , the most significant site was located at the 5’ region of the gap-mRNA ( Fig 6A and S3 Table ) . We confirmed binding of CsrA to this site by EMSA assays with in vitro transcribed mRNA comprising this region and purified CsrA ( Fig 6B ) . Interestingly , in contrast to the targets discussed above , the CsrA-binding site A ( N ) GGA was not associated with the ribosome-binding region suggesting a regulatory mechanism different from translation hindrance . To get insight how CsrA affects this operon , we performed qRT-PCR and compared the levels of transcription upstream ( at the end of the tkt gene ) and downstream ( inside the gap gene ) of the CsrA-binding region . As shown in Fig 6C , the transcript level of tkt and gap were similar in the wild type strain during the growth in broth , but showed significant differences in the csrA- mutant , a phenotype that was complemented when reintroducing csrA . Interestingly , the relative amount of the Tkt-transcript was identical in the wt and the mutant whereas the transcription of gap was significantly lower in absence of CsrA . Mfold analysis of the secondary structure of the RNA region containing the CsrA binding site predicted two energetically favored conformations: a ) the A ( N ) GGA motif was buried in a double-stranded sequence followed by a strong hairpin and b ) the CsrA regulatory motif was present in a loop and the subsequent hairpin was destabilized ( Fig 7A ) . Furthermore , adjacent to the hairpin we identified a short sequence closely related to an auxiliary element assisting in Rho-dependent termination of transcription [47] . To test our hypothesis that this region might be involved in preliminary transcription termination of the gap operon , we performed in vitro transcriptional assays in presence of NusG , an additional factor known to facilitate recognition of the termination signal [47 , 48] . The full-length transcript appeared in absence of the Rho factor whereas premature termination was observed when purified Rho was added to the reaction ( Fig 7B ) . Strikingly , this effect could be gradually reverted by adding increasing concentrations of CsrA ( Fig 7B ) . Northern blot analyses of the wt , csrA- and the complemented csrA- mutant showed indeed that in the csrA- strain more tkt transcript is present and less transcript of the complete operon as compared to the wt and the complemented csrA- mutant strain ( S7 Fig ) . Mutation of the CsrA-binding motif abolished that effect , but differently as expected . The introduction of AAAA replacing the AGGA-motif led to complete transcription run-off even in presence of Rho and independent of CsrA . This is probably due to the fact that the mutation energetically disfavors the development of the hairpin by preventing the auxiliary double-stranded region upstream of it . Thus , it stabilizes the conformation in which the termination structure is disrupted 'mimicking' permanent CsrA-binding . However , a double mutated template comprising the AAAA motif together with its complementary part of the dsRNA region ( S8A Fig ) to UUUU resulted in a truncated transcription fragment independent of CsrA ( S8B Fig ) . Surprisingly , due to this double mutation we observe a premature termination already in absence of Rho and CsrA that was not detected under wt condition ( Fig 7B ) . Possibly , the extended nucleotide mutations introduced provoke the stabilization of conformational changes in the RNA structure that leads to a transcriptional interruption even independent of the Rho protein . However , in presence of Rho the termination process is strongly enhanced compared to the absence of Rho . Furthermore , EMSA assays with this RNA showed that the double mutation led to a complete loss of interaction with the CsrA protein ( S8C Fig ) . It was shown that CsrA is able to remodel the RNA secondary structure in the leader sequence of the pgaA gene of E . coli to promote Rho-dependent transcription termination [48] , but here we demonstrate for the first time that CsrA may participate in the negative regulation of transcriptional termination events in bacteria . Our data suggest that CsrA is actively stabilizing RNA anti-terminator conformations inside a coding sequence rather than preventing a premature stop of transcription . This new model of regulation as proposed in Fig 7C would lead to an efficient expression of the glycolysis part of the tkt/gap-operon only in presence of CsrA . In this model , CsrA-dependent polar transcriptional effects enable L . pneumophila to regulate the pentose phosphate pathway ( transketolase ) and the glycolysis combined ( when CsrA is present ) or individually ( when CsrA is absent ) according to the needs of the cell even though both pathways share genes in a single operon . Among the targets identified by RIPseq , additional enzymes of the glycolysis ( fba , tpi , eno ) , the gluconeogenesis ( pps , ppc ) the pentose phosphate ( rpiA , prsA ) and Entner-Doudoroff pathway ( zwf-operon ) or enzymes for the supply of ATP and reducing equivalents through the pyruvate dehydrogenase complex ( pdh ) and the TCA cycle ( e . g . acnA , icd , sdh/suc , sfcA ) were present ( Table 2 and S3 Table ) . Thus , results from our combined datasets indicate that CsrA is a critical regulator of the carbon flux from the nutrient source—in L . pneumophila mainly amino acids like serine and threonine , but also glucose [49 , 50]—to obtain energy via oxidative phosphorylation . CsrA seems to play a pivotal role in the production of metabolic intermediates and cell components including the interconversion of amino acids or the biosynthesis of nucleotides and fatty acids/lipids essential for efficient replication and cell proliferation ( Table 1 and S3 Table ) . Metabolite flux analyses are under way to further substantiate our results obtained from the RIPseq , transcriptome and proteome data . In contrast , CsrA is hindering the synthesis of storage molecules and short-chain fatty acids according to our microarray and proteomic data . In absence of a functional glycogen biosynthesis pathway , these molecules , in particular the poly-3-hydroxybutanoate ( PHB ) , are used for carbon and energy storage during nutrient starvation [51 , 52] . Several enzymes of these pathways were identified in our RIPseq analyses , including the 3-hydroxybutyrate dehydrogenase ( Lpp2264 ) , the acetoactetyl-CoA reductase ( Lpp0620 ) and the polyhydroxyalkanoate synthase ( Lpp2038 PhbC ) ( S3 Table ) . For PhbC , the CsrA-binding region was located in the 5’UTR overlapping the RBS and the start codon assuming negative regulation of translation initiation ( S9A Fig ) . The EMSA assay undertaken with the in vitro transcribed phbC region , confirmed interaction of CsrA with the predicted region ( S9B Fig ) . To support our findings , we analyzed the PHB concentration of the wild type and csrA- strain during different growth stages . Bacteria were treated with BODIPY493/503 , a molecule known to stain nonpolar lipids , a method for quantification of cellular PHB in yeast and bacteria [53] . The fluorescence of the different L . pneumophila strains was measured by flow cytometry at exponential , post-exponential and stationary phase and the percentage of PHB positive cells was determined . We observed lower amounts of cells containing the storage polymer in replicating cells ( E phase ) and increasing numbers during the transition to transmissive phase . During stationary ( S- ) phase , the PHB amount dropped again in the wild type most likely due to its utilization for maintenance of vital physiological function ( S9C and S9D Fig ) . In contrast to the wt , no changes during the life cycle were observed in absence of CsrA . In particular during E and S phases , the quantity of PHB positive cells was significantly higher in the mutant compared to wt . Therefore , we assume that CsrA affects the blocking of the production of the storage component PHB during replication and its utilization during S phase as suggested by the missing decline in cellular fluorescence . However , four different polyhydroxyalkanoate synthases , but no PHB hydrolyzing poly ( 3-hydroxybutyrate ) depolymerase were identified in the L . pneumophila genome . Thus , catabolic and anabolic pathways are possibly executed by the same enzymes in a bidirectional manner . Directly related to the energetic status of the cell is the biosynthesis of co-factors and vitamins that are indispensible to ensure an unobstructed flow of metabolites . Several of their biosynthesis pathways are directly affected by CsrA ( Table 1 and S3 Table ) . One example is the regulation of thiamine synthesis in L . pneumophila . The thiamine derivate , thiamine pyrophosphate TPP , acts as a cofactor in the catabolism of sugars and amino acids and is closely linked to the CH-metabolism through the TPP-depending enzymes , like pyruvate dehydrogenase , 2-oxoglutarate dehydrogenase , transketolase and branched-chain 2-oxo acid dehydrogenase [54] . We identified a CsrA-binding region upstream the thi-operon of L . pneumophila ( lpp1522-1527 ) and a structure with similarity to a THI-box sequence of the TPP riboswitch in the 5'UTR region ( Fig 8A ) that was also identified by RibEx analysis [55] . Additionally , a putative Rho-independent terminator is located directly downstream of the TPP riboswitch ( Fig 8A ) . EMSA assays with the in vitro transcribed 5'UTR RNA region and purified CsrA showed interaction as a function of CsrA concentration ( Fig 8B ) . To study the functionality of the regulatory region , we established a reporter assay by fusing the TPP-riboswitch sequence to the β-Lactamase ( BlaM ) encoding gene . L . pneumophila wt csrA-mutant containing this plasmid were grown in a minimal medium [56] with and without defined concentrations of thiamine pyrophosphate . In fact , we observed a dependency of BlaM activity on TPP ( Fig 8C ) and that in absence of additional TPP , BlaM activity was significantly reduced in the mutant compared to the wt . This effect was even more pronounced in presence of 1 mM and 2 mM of TPP indicating that indeed CsrA is beneficial for the expression of the thi-operon . A similar situation has been described for the E . coli riboswitch responding to the molybdenum cofactor , as CsrA is able to activate the expression of the corresponding moaA genes by binding to the mRNA-leader sequence [57] . To strengthen our observations , we mutated the conserved region of the predicted TPP riboswitch that is known to be indispensible for binding the thiamine moiety of the TPP ( S10A Fig ) . Indeed , this mutation led to an uncoupling of the BlaM activity from the extracellular TPP concentration ( S10B Fig ) . Whereas mutation of the CsrA-binding ( S10A Fig ) site resulted in a lower BlaM activity already at low TPP concentrations compared to the non-mutated reporter assay ( S10B Fig ) similar to what was observed in the csrA- strain . Our results suggest a model where a concerted response between TPP and CsrA regulates the production of thiamine in L . pneumophila . As shown in Fig 8D two predominant conformational riboswitch-structures are predicted by Mfold . One is the known tandem-loop conformation of the riboswitch , the OFF state , with high affinity to TPP and stabilized by its binding . In this state the operon expression is inhibited as indicated from qPCR analysis ( S11 Fig ) , probably by transcriptional termination due to the presence of a terminator sequence in the 5' UTR leader sequence of the thi-operon . In the alternative conformation , the predicted ON state , the typical secondary structure of the TPP-riboswitch is unfolded and an alternative , third loop is formed containing the CsrA-regulatory motif A ( N ) GGA . We postulate that this structure has less affinity to TPP and diminishes the premature termination of the thi-leader sequence . Additionally , interaction with CsrA to the newly-formed A ( N ) GGA-loop is able to stabilize this open structure and shifts the reaction balance to the right even in presence of moderate concentrations of TPP . This thiamine dependent riboswitch is the first riboswitch identified in L . pneumophila and is another example for the diversified functionality of CsrA in the bacterial cell beyond translational interference . We assume that a CsrA-dependent fine-tuning mechanism of the TPP riboswitch ensures sufficient production of thiamine in the cell during high metabolic activity by altering the threshold of the TPP inhibitory feedback regulation . To deprive pathogens from the availability of iron , the infected host cells go into an “iron-withhold defense mode” leading to a cross-regulatory interaction between iron homeostasis and the immune response [58] . Therefore pathogens have developed many mechanisms to optimize iron acquisition from the host cell . Iron-uptake in L . pneumophila comprises iron-chelating siderophore production and ferrous iron uptake systems [59] . However excess of iron can be toxic for the cells because of the high reducing potential and the generation of reactive oxygen species . Therefore , a tight control of the intracellular iron concentration is inevitable . The uptake is coordinated in most bacteria by the ferric uptake regulator Fur and an iron-responsive regulatory sRNA [60] . Here we found that CsrA interacts with the transcript of the fur gene suggesting a growth phase-dependent control of iron acquisition . This is in agreement with the previous finding that iron starvation stimulates virulence formation , motility and stress resistance in L . pneumophila [61] . Interestingly , the observed CsrA-binding sites are located inside the CDS region ( S12A Fig ) , but EMSA assays confirmed binding of CsrA to pFur2 under in vitro conditions , but not to pFur1 , where another potential CsrA interaction region could be located according to the RIPseq analyses ( S12B Fig ) . Thus our results suggest that a sole CsrA-binding site inside the Fur-coding sequence is present . Growth phase-dependent analysis of the Fur transcript by RT-qPCR revealed increased transcription during E phase , which was slightly , but significantly reduced in the csrA- background ( S12C Fig ) . Therefore , CsrA seems to positively influence the fur transcript levels in metabolically active cells . To better understand how CsrA may function in this context we analyzed the stability of the Fur-mRNA in presence of rifampicin over time . This revealed a significant reduction of the half-life of the fur-mRNA in absence of CsrA in vivo and a higher RNA stability when over-expressing CsrA ( S13A Fig ) . According to this observation , we predict a stabilizing effect of CsrA on the Fur transcript , similar to what was described in E . coli for the glgC or flhDC transcripts [62 , 63] . The impact of CsrA on iron acquisition is also seen when growing the wt and the csrA- strain under different iron conditions . Indeed , loss of functional CsrA led to a growth defect at very low ( minimal medium without additional iron ) and very high iron concentrations , similar but in a lesser extent as for a fur-knock out strain ( S13B Fig ) . We further quantified the siderophore secretion capacity of the wt compared to the csrA- strain using the CAS assay as previously described [61] after transfer from iron-rich to iron-starvation medium . This showed a clear correlation between CsrA and CAS activity suggesting a vital role of CsrA in siderophore production and/or secretion ( S13C Fig ) . Furthermore , the RIPseq analysis identified numerous CsrA-RNA interactions with proteins that are directly linked to iron acquisition ( pyoverdine biosynthesis protein PucA ( Lpp0236 ) , heme oxygenase ( Lpp0288 ) , the zinc/iron transporter ( Lpp2018 ) or the heme-binding protein Hbp ( Lpp2164 ) and iron using proteins , like the catalase/peroxidase KatG ( Lpp0252 ) , the Fe-S cluster assembly complex ( Lpp0651-Lpp0658 ) , the ferredoxin-like 4Fe-4S binding protein ( Lpp1898 ) or the L-serine dehydratase ( Lpp0854 ) ( S3 Table ) . Thus , in L . pneumophila CsrA has a crucial impact on iron homeostasis highlighting yet another global function of CsrA .
The RNA-binding protein CsrA is the key regulator governing adaption of L . pneumophila to its hosts and therewith the transition between replicative bacteria to transmission competent bacteria . CsrA was first identified in Escherichia coli but has now been recognized in several bacterial species among those in many pathogens [12 , 13 , 27] . CsrA is known to post-transcriptionally control metabolism , motility and virulence by binding to mRNAs of its targets [12 , 13] . Here we report that 479 direct CsrA targets exist in the L . pneumophila genome ( S3 Table ) , which is in a similar range as the 467 targets reported for Salmonella Typhimurium [64] or the 154 targets reported for Campylobacter jejuni [38] when compared to the different genome sizes ( 3 . 4Mb for L . pneumophila , 4 . Mb for S . Thyphimurium and 1 . 6Mb for C . jejuni ) . Furthermore , we describe a new way by which CsrA may regulate gene expression differently within one operon and demonstrate that CsrA governs the transition from replicating to virulent bacteria in multilayered and complex circuitries with several unique features of L . pneumophila . We found that CsrA governs the expression of the virulent phenotype in a dual way . First by directly binding to the mRNAs of major virulence regulators like LqsR , PrmA , LetE and RelA and secondly by interfering with the expression of at least 41 Dot/Icm secreted proteins to assure their timely activity ( S3 Table ) . The direct regulation of 26 Dot/Icm effectors by CsrA was previously predicted by searching for the CsrA binding motif A ( N ) GGA and by showing that their expression was under the control of the LetA/Rsm/CsrA regulatory cascade [25] . Indeed , for 14 of these 26 substrates binding to CsrA was confirmed by our RIPseq analyses . Essential for virulence and successful infection is also the access to iron ( for review see [60] ) . In contrast , elevated iron concentrations can be toxic for the pathogen , thus iron acquisition , usage and storage have to be well coordinated and fine-tuned . Here we show that the maintenance of the iron homeostasis is tightly connected to CsrA as fur gene expression , encoding the main regulator of iron homeostasis , is under the direct influence of CsrA . It binds inside the coding sequence of fur potentially mediating therewith the stabilization of the mRNA ( S12 Fig ) . The increased transcript stability might be related to a reduced endonucleolytic accessibility in which the cleavage sites are occluded by CsrA . Indeed , the presence of a potential RNase E binding site ( A/G ) AUU ( A/U ) directly adjacent to the CsrA-binding motif may suggest RNase E degradation dependent on CsrA . In line with this model , a reduced growth rate at low iron levels and a lower siderophore secretion was detected in a csrA- strain ( S13B and S13C Fig ) . Moreover , the bacterial iron storage protein bacterioferritin ( Lpp2460 ) is transcribed in an RpoS- and LqsR-dependent manner [23 , 65] and both regulators are targets of CsrA . Thus CsrA participates in the control of iron homeostasis at the level of iron acquisition , usage and storage . It was previously shown for E . coli , that autoregulatory loops regulate CsrA expression and activity [66] . Here we uncovered that CsrA of L . pneumophila also binds its own mRNA and in addition it directly interacts with the 5' leader/RBS region of its own transcriptional activator , the response regulator PmrA . These findings indicate the presence of autoregulatory circuits to control expression and activity of the Csr system in L . pneumophila . However the regulatory circuitry differs partly from E . coli , as we observed inhibition of translation of CsrA by self-binding similar to what is reported for E . coli , and besides a negative autoregulatory effect of CsrA on its own transcript level what is different to E . coli . Indeed the L . pneumophila csrA transcript is downregulated during transmissive/virulent phase [9] . Thus in L . pneumophila extensive autoregulatory circuitries regulate CsrA expression and translation , but in contrast to E . coli the transcript levels seem also to be regulated by the TCS PmrBA , as described previously [24] and not only by CsrA itself ( Fig 5 ) . Most interestingly , we also observed interaction of CsrA with RelA mRNA indicating that a regulatory mechanism exists whereby CsrA directly affects the synthesis of the alarmone ( p ) ppGpp and hence the stringent response in L . pneumophila ( Fig 5 ) . Surprisingly , our proteomic data suggest that RelA expression is positively regulated by CsrA as a csrA- strain exhibits lower amounts of RelA protein than the wt strain during exponential growth ( S2 Table ) . Thus , a negative feedback regulation of the ( p ) ppGpp production in L . pneumophila seems to be present where the inactivation of CsrA by its antagonistic sRNAs weakens the stringent response . This regulatory circuitry is also different to the E . coli system where CsrA deficiency led to an increase in relA expression and elevated levels of ( p ) ppGpp [67] . The regulatory complexity is further increased as the transcription of relA is highly dependent on RpoS and DksA ( [15] ) . Furthermore , CsrA directly interacts with the 5' leader/RBS region of its transcriptional activator , the response regulator PmrA . Thus , the stringent response in L . pneumophila is controlled by several positive and negative feedback loops including the simultaneous inhibition of RelA on transcript level via diminution of RpoS but post-transcriptional activation by CsrA ( Fig 5 ) . This multilayered regulation and regulatory redundancy might be necessary to increase its robustness and for fine-tuning of the Csr system . The prominent role of CsrA is further underlined as we provide evidence that CsrA is the link between the stringent response and quorum sensing ( QS ) in L . pneumophila . It was shown for Vibrio that CsrA could affect the QS circuit by modulating the transcription of the response regulator LuxR or modulating the activity of LuxO indirectly via the sRNAs CsrB , C and D [68 , 69] . In L . pneumophila , we show direct interaction of CsrA to the 5'UTR/RBS region of LqsR suggesting a negative regulation of translation ( Fig 4 ) . The link of the stringent response with quorum sensing through CsrA allows L . pneumophila a balanced regulation of the biphasic life cycle . Only the integration of both pathways together ensure the expression of virulence factors , stress adaptations or cell motility with optimal precision to the metabolic state of the cell . Thus , crucial environmental signals like amino acid starvation likely lead to the coordinated expression of stationary phase traits when in parallel the cell density has reached a critical level . This might also be a signal for the necessity for the bacterium to become motile to be able to escape the host cell and find a new host . Indeed , the dependence of motility on the LetA/Rsm/CsrA cascade was shown earlier and an impact of CsrA on the flagellar biosynthesis sigma factor FliA was predicted [10] . Furthermore the QS response via the Lqs system is known to mediate motility in L . pneumophila [21] . We show here , that flagella biosynthesis is regulated through direct binding of CsrA to FleQ and FleR that are controlling the transcription of class II and class III flagella genes , respectively ( Fig 3 ) . Proteomic data suggest that both are negatively regulated via CsrA-binding to the RBS/start condon region of the transcript . Thus , sequestering CsrA by RsmXYZ from the target-RNAs FleQ , FleR and LqsR results in concerted activation of the flagella biosynthesis and links QS and motility via CsrA . Taken together , CsrA is the central regulator that integrates and coordinates the varying extra- and intracellular stimuli and merges them into a global cell response . The interconnection of flagella biosynthesis , stringent response and QS allows simultaneous preparation of the entire bacterial community to complete the replicative life cycle and to enter into the virulent stage in a cooperative manner . Similarly to what was described in other bacteria , CsrA of L . pneumophila is a major regulator of metabolism as numerous enzymes of various metabolic pathways are under the direct control of CsrA ( Table 1 and S1–S3 Tables ) . Indeed , CsrA may also act prominently at the interface of the bipartite metabolism observed in L . pneumophila [70] balancing the two distinct parts containing the Entner-Douderoff ( ED ) pathway/ Gluconeogenesis/Pentose Phosphate pathway ( PPP ) on one hand and the TCA cycle/Amino acid metabolism/Fatty acid metabolism on the other hand . By studying the regulation of the carbon metabolism by CsrA we discovered a new way by which CsrA may modulate transcription and regulates carbon flux . Indeed the particular location of a CsrA binding site within the coding sequence of the glyceraldehyde-3-phosphate dehydrogenase ( gap ) gene of the tkt/gap-operon was intriguing ( Fig 6 ) . Our qPCR analyses showed strong CsrA dependency for the transcription of gap , but not for the tkt transcript although they are organized in an operon . This effect was not related to changing RNA stability between the transcripts , but in vitro transcription assays in presence of Rho , NusG and CsrA disclosed that the premature transcriptional stop caused by Rho was completely absent when CsrA was present . These results led us to propose a model in which CsrA-binding stabilizes the alternative secondary conformation that cover the transcription termination sites . Hence , under optimal nutrient availability the genes of the operon are transcribed together as a unit . In contrast , under certain metabolic conditions , like stress or starvation , it can be advantageous to uncouple the expression of the PPP ( transketolase ) and the glycolysis to conserve energy and to prevent the synthesis of unused transcripts . Thus , transcription of the inhibitory sRNA and subsequent CsrA-sequestration unmasks the premature terminator promoting the transcriptional block of the downstream glycolytic genes without affecting the transcription of the tkt gene . Excitingly , in L . pneumophila CsrA is able to mediate transcriptional polarity effects by preventing rho-dependent termination , a regulatory mode that might be present also in other bacteria encoding CsrA . In addition , CsrA affects the production of secondary metabolites and vitamins that are necessary for an undisturbed metabolic flow . Among those are thiamine pyrophosphate , which is essential for the functioning of central enzymes of the carbohydrate metabolism like the pyruvate dehydrogenase , the 2-oxoglutarate dehydrogenase , but also the transketolase . Interestingly , we detected a THI-box riboswitch structure in the 5'UTR leader sequence of the thi-operon , a widely used control element found in all kingdoms of life [55] that was overlapping with a CsrA-binding site . Our experiments showed that CsrA is necessary for the fine-tuning of the thi-operon expression most likely by modulating the secondary structure of the TPP riboswitch ( Fig 8 ) . During periods of high metabolic activity the amount of TPP used in the cell needs to be elevated compared to carbon starvation phases . Consequently , the TPP threshold value that leads to the transcriptional or translational blockage via the riboswitch must be adapted to the actual conditions . We postulate that the interaction of CsrA with the thi-leader sequence reduces its affinity to TPP . This consequently would lead to an increased expression of the genes necessary for the thiamine biosynthesis even in presence of considerable amounts of TPP . Collectively , our data show that CsrA is linking the fine-tuned regulation of the stringent response , quorum sensing , metabolism and virulence and revealed that the decision of the cell between energy production via the TCA cycle and the synthesis of the carbon and energy storage molecule poly-3-hydroxybutyrate ( PHB ) , is coordinated through the activity of CsrA ( Fig 9A ) . Furthermore we discovered the first riboswitch in Legionella and found that L . pneumophila CsrA has evolved a mechanism by which it is able to regulate genes organized in the same operon differently , according to the needs of the cell ( Fig 9B ) adding thereby another example for the astonishing diversity of CsrA functions in bacterial cells .
L . pneumophila strain Paris was cultured in N- ( 2-acetamido ) -2-aminoethanesulfonic acid ( ACES ) -buffered yeast extract broth or on ACES-buffered charcoal-yeast ( BCYE ) extract agar at 37°C . A . castellanii ATCC50739 was cultured in PYG 712 medium ( 2% proteose peptone , 0 . 1% yeast extract , 0 . 1 M glucose , 4 mM MgSO4 , 0 . 4 M CaCl2 , 0 . 1% sodium citrate dihydrate , 0 . 05 mM Fe ( NH4 ) 2 ( SO4 ) 2 x 6H2O , 2 . 5 mM NaH2PO3 , 2 . 5 mM K2HPO3 ) at 20°C . A . castellanii infection assays were conducted as previously described [9][46] . Intracellular multiplication was monitored using a 300μl sample , which was centrifuged ( 14500 rpm ) and vortexed to break up amoeba . The number of colony forming units ( CFU ) was determined by plating on BCYE agar . Each infection was carried out in duplicates or triplicates . Mutant strains of L . pneumophila were constructed as described previously [71] . In brief , the gene of interest was inactivated by introduction of an apramycine resistance ( apraR ) cassette into the chromosomal gene by 3-steps PCR using the following primers: CsrA_F TTGCAATATAAGCTCAAGATAC and CsrA_Inv_R gctgatggagctgcacat gaaTAAATTTCTTCACGATGAACAG , CsrA_Inv_F gagcggatcggggattgtcttAAAGAAT CTGATGATTCGGAAC and CsrA_R ATTGTTGATAACAAAAGTATCC . To amplify the apramycine cassette the primers Apra_F TTCATGTGCAGCTCCATCAGC and Apra_R AAGACAATCCCCGATCCGCTC were used . The final product was cloned into the pGEM-T easy vector ( Promega ) . For the beta-lactamase reporter assay , the predicted TPP riboswitch region was amplified ( TPP_F GAATTCGGCGCGGGGTGTCGGGAAATC , TPP_R GAATTCAAAAGGGAACCATGCCTTAAAAAGG ) and cloned into the pXDC61 upstream of the blaM gene using the EcorI restriction side . BODIPY 493/503 ( Molecular Probes ) was solubilized in DMSO at a concentration of 100μg/ml . Bacterial cultures of wt and csrA- mutant were grown in BYE and 500μl were centrifuged for 3 min at 5000g at different ODs . Pellets were resuspended in 35% ethanol , adjusted to OD 0 . 1 and incubated for 20 min at room temperature ( RT ) . After centrifugation , the pellet was resuspended in 990μl PBS and 10μl BODIPY stock solution and incubated for 5min , RT . The cells were pelleted and washed once with 1ml PBS before resuspended in 100μl PBS . Fluorescence was analyzed with a MACSQuant flow cytometer ( Miltenyi Biotec ) . For the iron deficiency assays , L . pneumophila wt and csrA mutant strains were grown in Minimal Medium at an initial OD600 of 0 . 1 containing 0 ( +/- DFX ) , 25 , 50 , 100 , 250 , 500 and 1000μM of additional iron-pyrophosphate . After 24h of growth at 37°C , 170rpm , absorption of the cultures were measured at OD600 . To quantify the siderophore secretion wt and csrA- mutants were grown in BCYE medium until E phase . The cells were pelleted and washed twice with Minimal medium without additional iron and resuspended at an OD of 0 . 2 . At time points 0 , 2h , 4h , 6h , 8h and 24h , cells were centrifuged and 150μl of the supernatant were mixed with 30μl of CAS solution ( 60 . 5mg Chromeazurol S in 50m H2O , 2 . 7mg FeCl3*6H20 in 10mM HCl , 73mg HDTMA in 40ml H2O were mixed and autoclaved ) and the OD655 was measured after 30 min of incubation at RT . Strains containing the TPP-pXDC61 plasmid were grown in Minimal Medium containing 0 , 1 , 2mM thiamine pyrophosphate at 37°C , 200rpm without IPTG until reaching early exponential phase . Cells were harvested , resuspended in PBS and sonicated . Protein concentration was quantified by Bradford . 10μg of total protein in 100μl PBS were mixed with 50μl Nitrocefin ( 0 . 5mg/ml in PBS + 5% DMSO ) and the enzyme kinetic was followed with a spectrophotometer at 486nm . BlaM activity was calculated from the initial slope . For CsrA expression in L . pneumophila Paris ( containing a double Flag-Tag at the C-terminal end or without Flag-Tag ) full-length cDNAs encoding CsrA ( lpp0845 ) were amplified by PCR using primer CsrA-F tctagaATGTTGATTTTGACTCGGCGTATAG and CsrA-R ctgcagTTATACTGCTTGTTCCGAATCATC or CsrA-Flag-R ctgcagTTACTTATCGTCA TCGTCCTTGTAGTCCTTATCGTCATCGTCCTTGTAGTCTACTGCTTGTTCCGAATCATC , respectively , and cloned into pGEM-T easy vector ( Promega ) . The fragment was verified by sequencing , cut with XbaI/PstI and ligated into the pBC KS vector under the control of the Mip ( lpp0855 ) promoter region of L . pneumophila . Competent L . pneumophila bacteria were transformed by electroporation and positive colonies were selected on 10μg/ml chloramphenicol and sequenced . L . pneumophila expressing CsrA+2xFlagTag and L . pneumophila expressing CsrA without Flag-Tag as negative control were grown in broth until exponential growth ( OD 2 ) . Cells were cross linked with formaldehyde ( final concentration 1 , 1% ) over night at 4°C on a rotating platform , then formaldehyde was quenched by adding 125 mM glycine and pellets were rinsed twice with PBS . Pellets were resuspended in lysis buffer ( 50mM HEPES-KOH pH7 . 5 , 150mM NaCl , 1mM EDTA , 1% Triton X-100 , 0 . 1% Na-deoxycholate , protease inhibitor ) , sonicated and total protein concentrations were adjusted to 1mg/ml . The total protein of both samples , CsrA+2xFlagTag and negative control , was cleared separately by BSA-blocked Dynabeads protein G ( Invitrogen ) and subsequently incubated with Dynabeads protein G coupled to Anti-Flag antibodies ( Sigma ) over night at 4°C on a rotating platform . Samples and negative control were washed twice with Lysis buffer containing 350mM NaCl and 5 times with wash buffer ( 10mM Tris-HCl pH 7 . 5 , 250 mM NaCl , 0 . 5% NP-40 , 0 . 5% Na-deoxycholate , 1mM EDTA ) . Subsequently beads were washed with TE buffer , resuspended in elution buffer ( 50mM Tris-HCl pH 8 . 0 , 1mM EDTA , 1% SDS ) and incubated at 65°C for 30 min . Cross-linking was reversed and DNA and protein were digested . The RNA was metal-catalyzed heat fragmented to a size around 100-200nt using the RNA fragmentation kit . The RIPseq library IP1 and IP2 were constructed as described previously [72 , 73] . For IP7 , 8 , and 9 the RNA of the independent samples was purified and further processed according to the TruSeq stranded mRNA sample preparation guide of Illumina . The two parallel processed samples were ligated with adaptor 6 ( positive CsrA+2xFlagTag library ) and adaptor 12 ( negative control , minus FlagTag ) , respectively , the quantity was determined using a Qubit 2 . 0 ( Invitrogen ) and the quality was checked using a Bioanalyzer . High quality libraries were sequenced using an Illumina HiSeq platform . This experiment was done in 5 replicates . The reads in FASTQ format generated by Illumina sequencing were filtered using FastXclipper from the FastX toolkit ( http://hannonlab . cshl . edu/fastx_toolkit/ ) and Tagdust [67] for adapter removal . Trimming of reads was performed using Sickle ( https://github . com/najoshi/sickle ) . To assure high sequence quality , we used a cutoff phred score of 20 . After trimming , all the sequence reads shorter than 18nt were eliminated . Reads were aligned to the L . pneumophila strain Paris chromosome sequence ( NCBI Acc . -No: NC_006368 . 1 ) using Bowtie 2 software [68] . Only the uniquely mapped reads are kept , so , a read cannot contribute for the coverage value at different positions . Duplicate reads were removed from mapping results ( with samtools rmdup ) and BAM files were built using the Samtools software [69] . For sample and control , coverage files ( wiggle files ) were generated for the forward and the reverse strand and normalized according to the number of mapped base pairs . For each couple sample/control , a scale factor was applied to scale the small sample up to the bigger sample . To define CsrA-bound RNA regions , enrichment regions ( peaks ) were detected using a python script “sliding_window_peak_calling_script . py” previously described [38] and available at Zenodo ( https://zenodo . org/record/49292 ) . Peak detection was performed separately for the forward and reverse strand of each replicon , according to the author description and with a minimum required fold change ( FC ) of 5 for the enrichment . To identify CsrA targets , overlaps between gene annotations and enrichment regions were performed with bedtools suite [74] . Depending on the library preparation , we used peaks with sequence reads in the sense in which the genes are described for IP-1 and IP-2 and peaks with reads in the reverse sense as the genes are transcribed for IP-7 , IP-8 and IP-9 . Overlaps with less than 10 nts were discarded . BAM files were imported into Artemis [75] to manually validate the identified CsrA targets . The sequence reads after adapter removal as well as coverage files of the RIP-seq libraries have been deposited in the NCBI Gene Expression Omnibus ( GEO ) [76] under the accession number GSE94068 . Full-length cDNAs encoding CsrA ( lpp0845 ) were amplified by PCR using primer CsrA-F ccatggTGATTTTGACTCGGCGTATAG and CsrA-R aagcttTACTGCTTGTTCCGAATCA TCAG . Fragments were cloned in frame into the expression vector pET-28a ( Novagen ) , which adds a hexa-histidine tag to the C-terminus of the protein; positive clones in E . coli DH5α were selected on 50μg/ml kanamycine and sequences were verified . E . coli BL21 ( Invitrogen ) cells were used for protein overexpression . The cells were grown in LB medium , containing , 50 μg ml-1 kanamycine , at 20°C . Protein production was induced by adding 0 . 5 mM IPTG at A600 ~0 . 5 , and cells were harvested in late exponential phase by centrifugation at 4°C . Cells from 1 liter culture were resuspended in 1 ml buffer A– 100 mM Tris/HCl ( pH 7 . 5 ) , 150 mM NaCl , 100mM KCl , 5 mM MgCl2 , 2 mM DTT , 5% ( v/v ) glycerol and a cocktail of protease inhibitors ( Sigma ) at the concentration recommended by the manufacturer . Cells were disrupted by sonication , centrifuged ( 18 , 000 × g , 30 min , 4°C ) and the supernatant was applied to a Ni-NTA affinity column ( GE Healthcare ) equilibrated with buffer A including 10 mM imidazole . The column was washed with the same buffer containing 100 mM imidazole then the enzyme was eluted with 500 mM imidazole in buffer A . Fractions were dialyzed against buffer A and concentrated by centrifugation ( Microcon , 3 kDa cut-off , Millipore ) to a final concentration of 0 . 5 mg protein/ml . Proteins were quantified according to Bradford using BSA as standard . Aliquots of 20μl were frozen in liquid N2 and kept at -80°C until further use . The quality of the purification was determined after SDS-PAGE analysis ( 4% stacking gel and 12% running gel ) and staining with Brilliant Blue G—Colloidal Concentrate ( Sigma ) . The region corresponding to target RNAs selected according to the RIPseq experiments was amplified from bacterial DNA adding a T7 promoter at the 5' end . PCR products were used in MEGAshortscript T7 Kit ( Ambion ) to produce in vitro RNA ( S6 Table ) . 2μM of Biotin-11-CTP was added into the reaction mix for later detection . This reaction mix was incubated at 37°C for 2h and the RNA was purified by Phenol/Chloroform extraction . The RNA concentration was estimated by UV absorption at 260nm . For 10μl interaction assays , 200nM of RNA was combined with varying concentrations of purified CsrA-His ( 0–5μM ) and incubated in buffer A in presence of 250ng tRNA yeast ( Invitrogen ) for 30min at RT . Subsequently , samples were fractionated under non-denaturing conditions on Blue-Native PAGE and blotted to BrightStar-Plus transfer membranes ( Ambion ) . Membranes were blocked in PBS buffer containing 0 . 1% Tween-20 and 1% ECL blocking agent ( GE Healthcare ) for 1h at RT and , afterwards , incubated for 1h in the same buffer including mouse anti-biotin antibodies ( Invitrogen ) . After washing and binding of secondary antibodies ( anti-mouse Ig-HRP , Dako ) , the RNA-bands were visualized with ECL Plus Western blotting solutions ( GE Healthcare ) and detected with a G-box ( Syngene ) . CsrA was over-expressed as described above . Full-length cDNA of the ( transcription termination factor Rho ( lpp3002 ) and NusG ( lpp0382 ) were amplified by PCR using Rho-F ggatccATGAATCTTAGTGAACTTAAGCAATTAC/Rho-R ctcgagTTATTCCTGACGCTT CATTGCATC and NusG-F ccATGGTCGAGGAAAACAAAGCAAAACAG/NusG-R ctcgagTGTTTTTTCTACTTGACTGAACTC , respectively . Fragments were cloned in frame into the expression vector pET-28a ( Novagen ) , which adds a hexa-histidine tag to the N-terminus of the protein . The Rho and NusG proteins were over-expressed and purified as described for CsrA , except buffer R was used instead– 50 mM Tris/HCl ( pH 7 . 5 ) , 50mM KCl , 2 mM DTT , 10% ( v/v ) glycerol and a cocktail of protease inhibitors ( Sigma ) at the concentration recommended by the manufacturer . Template DNA for the T7 RNA polymerase was amplified from wt L . pneumophila Paris using primer pairs GapTer-F TGTAATACGACTCACTATAGGATCTGGCATCGATGTGACCG and GapTer- R TATGACCCATTGCCGCATCTC . Transcription termination assays were performed as follows: 20μl reaction mixtures containing 70nM template DNA , 20U T7 RNA polymerase + transcription buffer ( Thermo Scientific ) , 1μM NusG , 0–2μM Rho , 0–5μM CsrA , 0 . 5μl RNaseOut ( Life Technologies ) , 100μM ATP , GTP , UTP; 25μM CTP + 25μM CTP-11-Biotin ( Roche ) were incubated for 2h at 37°C . The reaction was stopped with 115μl 5mM EDTA before extraction in phenol:chloroforme:isoamyl alcohol and precipitation with ethanol and Na-acetate . Pellets were resolved in 1x RNA loading dye ( Thermo Scientific ) , incubated for 10 min at 65°C and analyzed by 10% urea-polyacrylamide electrophoresis . Samples were blotted to BrightStar-Plus transfer membranes ( Ambion ) and visualized as described for EMSA assays . Ten micrograms of total RNA isolated from wt , csrA- and the complemented csrA- strain were size-separated on a 1 . 5% denaturing agarose gel and transferred onto a positively charged nylon membrane by capillarity . The gel was photographed under ultraviolet light to capture ethidium bromide staining of ribosomal RNA bands for loading controls . RNA was cross- linked to membranes by exposure to UV light for two minutes and membranes were prehybridized in Ultrahyb buffer ( Ambion , AM8670 ) for 1h . RNA probes radioactively labelled with α-P33-UTP ( Perkin-Elmer , BLU007X500UC ) were generated using the T7 Maxiscript kit ( Ambion , AM1314 ) and PCR templates were amplified from genomic DNA using primers T7-gapNB_F , gapNB_R , T7-tktNB_F and tktNB_R listed in S7 Table . The membrane was hybridized at 68°C by adding the radiolabeled probes overnight . Blots were washed twice at the hybridization temperature in 2X SSC , 0 . 1% SDS followed by two washes in in 0 . 1X SSC , 0 . 1% SDS . Membranes were wrapped in saran wrap and subsequently used to expose to films . Total RNA was extracted as previously described [77] . Paris wt and mutant strains were grown in AYE medium , and harvested for RNA isolation at the E ( OD 2 . 5 ) and PE growth phase ( OD 4 . 3 ) . RNA was prepared in triplicates ( three independent cultures ) and each RNA sample was hybridized twice to the microarrays ( dye swap ) . RNA was reverse-transcribed with Superscript indirect cDNA kit ( Invitrogen ) and labeled with Cy5 or Cy3 ( Amersham Biosciences ) according to the supplier’s instructions . The design of microarrays containing gene-specific 70mer oligonucleotides based on all predicted genes of the genome of L . pneumophila strain Paris ( CR628336 ) and its plasmid ( CR628338 ) was previously described [9] . Hybridization was performed following the manufacturers’ recommendations ( Corning ) using 250 pmol of Cy3 and Cy5 labeled cDNA . Slides were scanned on a GenePix 4000A scanner ( Axon Instruments ) . Laser power and/or PMT were adjusted to balance the two channels and the resulting files were analyzed using Genepix Pro 4 . 0 software . Spots were excluded from analysis in case of high local background fluorescence , slide abnormalities , or weak intensity . Data normalization and differential analysis were conducted using the R software ( http://www . R-project . org ) . No background subtraction was performed , but a careful graphical examination of all the slides was conducted to ensure a homogeneous , low-level background in both channels . A loess normalization [78] was performed on a slide-by-slide basis ( BioConductor package marray; http://www . bioconductor . org/packages/bioc/stable/src/contrib/html/marray . html ) . Differential analysis was carried out separately for each comparison between two time points , using the VM method ( VarMixt package [79] ) , together with the Benjamini and Yekutieli [80] p-value adjustment method . If not stated otherwise , only differently expressed genes with 1 . 5-fold-changes were taken into consideration . Empty and flagged spots were excluded from the data set , and only genes with no missing values for the comparison of interest were analyzed . Legionella was grown in triplicates to E phase ( OD2 . 5 ) and cells were lysed in 20 mM HEPES pH 8 . 0 , 8 M urea , 1 mM sodium orthovanadate , 2 . 5 mM sodium pyrophosphate and 1 mM glycerophosphate by sonication . The protein concentration in the supernatants of each replicate was measured using a Bradford assay ( Biorad ) and equal protein amounts , each containing 1 mg total protein , were used for further analysis . Proteins in each sample were reduced with 5 mM DTT and incubation for 30 minutes at 55°C and then alkylated by addition of 100 mM iodoacetamide and incubation for 15 minutes at room temperature in the dark . Both samples were further diluted with 20 mM HEPES pH 8 . 0 to a final urea concentration of 2 M and proteins were digested with 10 μg trypsin ( Promega ) ( 1/100 , w/w ) overnight at 37°C . Peptides were then purified on a Sep-Pak C18 cartridge ( Waters ) and 50 μg peptides of each sample was re-dissolved in 50 μl solvent A ( 0 . 1% formic acid in water/acetonitrile ( 98:2 , v/v ) ) of which 2 μl was injected for LC-MS/MS analysis on an EASY-nLC 1000 system ( Proxeon , Thermo Fisher Scientific ) in line connected to a Q Exactive HF mass spectrometer with a Nanospray Flex Ion source ( Thermo Fisher Scientific ) . Peptides were loaded in solvent A ( 0 . 1% formic acid in water ) on a reverse-phase column ( made in-house , 75 μm I . D . x 250 mm , 1 . 9 μm beads C18 Reprosil-Pur , Dr . Maisch ) and eluted by an increase in solvent B ( 0 . 1% formic acid in acetonitrile ) in linear gradients from 5% to 23% in 100 minutes , then from 23% to 40% in 40 minutes and finally from 40% to 55% in 20 minutes , all at a constant flow rate of 250 nl/min . The mass spectrometer was operated in data-dependent mode , automatically switching between MS and MS/MS acquisition for the 15 most abundant ion peaks per MS spectrum . Full-scan MS spectra ( 300–1700 m/z ) were acquired at a resolution of 60 , 000 after accumulation to a target value of 3 , 000 , 000 with a maximum fill time of 20 ms . The 15 most intense ions above a threshold value of 400 , 000 were isolated ( window of 1 . 4 Th ) for fragmentation by HCD at a normalized collision energy of 28% after filling the trap at a target value of 100 , 000 for maximum 25 ms with an underfill ratio of 10% . The S-lens RF level was set at 60 and we excluded precursor ions with single , unassigned and charge states above six from fragmentation selection . Data analysis was performed with MaxQuant ( version 1 . 5 . 3 . 8 ) [81] using the Andromeda search engine [82] with default search settings including a false discovery rate set at 1% on both the peptide and protein level . Spectra were searched against two databases with L . pneumophila Paris proteins encoded by the chromosome ( database containing 3142 protein sequences ) and the plasmid ( database containing 142 protein sequences ) ( http://genolist . pasteur . fr/LegioList/ ) with a mass tolerance for precursor and fragment ions of 4 . 5 and 20 ppm , respectively , during the main search . Enzyme specificity was set as C-terminal to arginine and lysine , also allowing cleavage at proline bonds and a maximum of two missed cleavages . Oxidation of methionine residues was set as variable modification and carbamidomethyl formation of cysteine residues was set as a fixed modification . Only proteins with at least one unique or razor peptide were retained leading to the identification of 1662 L . pneumophila proteins ( S8 Table ) . Proteins were quantified by the MaxLFQ algorithm integrated in the MaxQuant software [83] . A minimum ratio count of two unique or razor peptides was required for quantification . Further data analysis was performed with the Perseus software ( version 1 . 5 . 3 . 0 ) after loading the protein group file from MaxQuant . Proteins only identified by site , reverse database hits and contaminants were removed and replicate samples were grouped . Proteins with less than three valid values in at least one group were removed and missing values were imputed from a normal distribution around the detection limit . After log2 transformation of the LFQ intensity values , a t-test was performed ( FDR = 0 . 05 and S0 = 1 ) to compare both strains and reveal significantly up- and downregulated proteins ( S2 Table ) . After Z-scoring , the intensity values of each protein were also visualized on a heat map after non-supervised hierarchical clustering ( Fig 2 ) . The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository [84] with the dataset identifier PXD004730 . | The RNA binding protein CsrA is the master regulator of the bi-phasic life cycle of Legionella pneumophila governing virulence expression in this intracellular pathogen . Here , we have used deep sequencing of RNA enriched by co-immunoprecipitation with epitope-tagged CsrA to identify CsrA-associated transcripts at the genome level . We found 479 mRNAs or non-coding RNAs to be targets of CsrA . Among those major regulators including FleQ , the regulator of flagella expression , LqsR , the regulator of quorum sensing and RpoS implicated in stress response were identified . The expression of over 40 type IV secreted effector proteins important for intracellular survival and virulence are under the control of CsrA . Combined with transcriptomics , whole shotgun proteomics of a wild type and a CsrA mutant strain and functional analyses of several CsrA-targeted RNAs we identified the first riboswitch in L . pneumophila , a thiamine pyrophosphate riboswitch , and discovered a new mode of regulation by CsrA that allows L . pneumophila to regulate the pentose phosphate pathway and the glycolysis combined or individually although they share genes in a single operon . Our results further underline the indispensable role of CsrA in the life cycle of L . pneumophila and provide new insights into its regulatory roles and mechanisms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"electrophoretic",
"mobility",
"shift",
"assay",
"gene",
"regulation",
"pathogens",
"messenger",
"rna",
"microbiology",
"operons",
"dna",
"transcription",
"regulator",
"genes",
"gene"... | 2017 | The Legionella pneumophila genome evolved to accommodate multiple regulatory mechanisms controlled by the CsrA-system |
The use of conventional chemical insecticides and bacterial toxins to control lepidopteran pests of global agriculture has imposed significant selection pressure leading to the rapid evolution of insecticide resistance . Transgenic crops ( e . g . , cotton ) expressing the Bt Cry toxins are now used world wide to control these pests , including the highly polyphagous and invasive cotton bollworm Helicoverpa armigera . Since 2004 , the Cry2Ab toxin has become widely used for controlling H . armigera , often used in combination with Cry1Ac to delay resistance evolution . Isolation of H . armigera and H . punctigera individuals heterozygous for Cry2Ab resistance in 2002 and 2004 , respectively , allowed aspects of Cry2Ab resistance ( level , fitness costs , genetic dominance , complementation tests ) to be characterised in both species . However , the gene identity and genetic changes conferring this resistance were unknown , as was the detailed Cry2Ab mode of action . No cross-resistance to Cry1Ac was observed in mutant lines . Biphasic linkage analysis of a Cry2Ab-resistant H . armigera family followed by exon-primed intron-crossing ( EPIC ) marker mapping and candidate gene sequencing identified three independent resistance-associated INDEL mutations in an ATP-Binding Cassette ( ABC ) transporter gene we named HaABCA2 . A deletion mutation was also identified in the H . punctigera homolog from the resistant line . All mutations truncate the ABCA2 protein . Isolation of further Cry2Ab resistance alleles in the same gene from field H . armigera populations indicates unequal resistance allele frequencies and the potential for Bt resistance evolution . Identification of the gene involved in resistance as an ABC transporter of the A subfamily adds to the body of evidence on the crucial role this gene family plays in the mode of action of the Bt Cry toxins . The structural differences between the ABCA2 , and that of the C subfamily required for Cry1Ac toxicity , indicate differences in the detailed mode-of-action of the two Bt Cry toxins .
In recent decades , agriculture has increasingly come to rely on toxins encoded by the Gram-positive bacterium Bacillus thuringiensis ( Bt ) for the production of insect–resistant transgenic crops . Narrow spectrum insecticides such as the protoxin crystals produced by Bt during sporulation are highly specific for certain insect groups including the Lepidoptera , Diptera and Coleoptera ( e . g . , [1] ) . Bt sprays have been used for many years , gaining widespread acceptance in pest management due to their relative target-specificity and their safety for humans , most other organisms , and the environment . However , the increasing cultivation of Bt transgenic crops poses a significant risk with various field populations of major lepidopteran pests reported to have developed resistance [2–4] , threatening the sustainability of this strategy for crop protection . Indeed a major reason for the uptake of Bt was the evolution of resistance to chemical insecticides such as organochlorides , synthetic pyrethroids , and organophosphates in pests such as the cotton bollworm Helicoverpa armigera . This species is one of the most damaging and economically important lepidopteran pests known worldwide in a variety of crops , and one of four major pests in the genus Helicoverpa , the others being H . punctigera , H . zea and H . assulta [5–8] . For control of lepidopteran pests , genes encoding members of the Cry1A family were the first to be used in transgenic crops . However resistance alleles to the Cry1Ac toxin have been reported in field-collected H . punctigera from Australia [9 , 10] , and H . armigera from China [11–13] . Resistance to Cry1Ac has also been reported in Indian populations of Pectinophora gossypiella [14] and H . zea from the New World [15–17] . Genetic studies on the field-derived strains have provided critical insight into the mode of action of this toxin , by identifying key receptors present on the surface of midgut epithelial cells ( e . g . , [18–23]; see also [24–26] ) . This binding of the activated toxins to specific receptors is crucial for formation of pores in the affected cells , leading eventually to the death of the larvae . Resistance to the Cry1Ac toxin in the Lepidoptera was first shown to be associated with mutation of a gene encoding a 12-cadherin domain protein . Deletions of different lengths were observed in various regions of the gene in , e . g . , Heliothis virescens [27] , P . gossypiella [28] and H . armigera [11–13 , 24] , as well as insertions by transposable elements [12 , 29 , 30] . Down regulation of cis-mediated transcription of the trypsin gene HaTryR allele due to mutations at the promoter region , mis-splicing of the ABCC2 gene , and a deletion mutation of the Aminopeptidase N ( APN ) gene have also been demonstrated to lead to resistance to Cry1Ac in H . armigera [31–33] . Recently , mutations in the ABCC2 gene belonging to Family C of the ATP-Binding Cassette ( ABC ) transporter family have also been shown to confer resistance to Cry1Ab and Cry1Ac in H . virescens [34] , Plutella xylostella [35] , Bombyx mori [36] , Spodoptera exigua [37] , and H . armigera [32] . Heterologous expression of ABCC2 from resistant and susceptible B . mori has shown that it aids in pore formation [38] , and modification or deletion of ABCC2 is hypothesized to block the final step in the toxin's mode of action [39] . To prolong the efficacy of individual Bt toxins as transgenic control agents , multiple Bt genes , encoding different toxins with different modes of action , have been incorporated into plants . The second Bt gene adopted in many countries in transgenic plants has been a member of the Cry2A family , Cry2Ab . Cry1A resistance due to mutations in the cadherin or ABCC2 genes is not known to confer cross-resistance to Cry2Ab [22 , 34] . However , Cry2Ab resistance has now been reported in various lepidopteran pests ( e . g . , P . gossypiella [40]; H . zea [41] ) . In Australia , field-derived Cry2Ab resistance alleles in H . armigera and H . punctigera were first isolated in the summer of 2002/2003 and 2004/2005 respectively , and were used to establish the homozygous resistant lines SP15 [42] and Hp4-13 respectively [43] . All Cry2Ab resistant H . armigera and H . punctigera alleles isolated from Australia to-date have been shown to be recessive [9 , 21 , 42 , 43] . Isolates were captured using the "F2 screen" method [44] with a discriminating dose of Cry2Ab toxin , and confirmed as allelic by complementation tests [83] . These isolates are being used in F1 tests which involved crossing them to a field-collected insect ( of unknown genotype ) , and screening the F1 offspring for resistance [45–47] . These F1 screens have estimated the frequency of Cry2Ab-resistance-conferring alleles in H punctigera and H . armigera field populations to range between 0 . 010 and 0 . 047 , and 0 . 015 and 0 . 044 , respectively , with no significant linear trend over time ( from 2007 to 2014 ) [48–50] . The F1 and F2 screen techniques do not directly reveal the molecular identity of such resistance alleles , and the molecular basis of the Cry2Ab resistance is likely due to specific target-site alterations located within the midgut [21] . Whether the H . armigera and H . punctigera Cry2Ab resistance genes are homologous is not known , although parallel evolution in orthologous ABCC2 genes leading to Cry1Ac resistance in different species has been reported ( e . g . , [34 , 35] ) . A third generation of transgenic cotton ( Bollgard III ( BGIII ) ) expressing three Bt toxin genes ( Cry1Ac , Cry2Ab and Vip3A ) will soon become available in Australia . In light of these developments and whilst the Australian industry adopts a pre-emptive strategy to manage resistance to Bt , several key assumptions of this strategy are theoretically sound but empirically untested . Marker-assisted detection of the resistance alleles in insect populations will therefore not only enable a more efficient monitoring effort but will also enable assumptions about the ecology of resistance to be rigorously examined . In this paper we report on the identification of the Cry2Ab resistance gene in H . armigera using linkage mapping and a chromosome walk with the assistance of exon-primed intron-crossing ( EPIC ) -PCR markers . This gene , which is expressed in the midgut , encodes an ABC transporter in the A subfamily—ABCA2—and is the likely site of mutations conferring resistance to Cry2Ab . By screening additional lab-isolated resistant lines derived from field-collected materials , we show that resistance to the Cry2Ab toxin in H . armigera occurred through independent evolutionary events involving different mutations , all of which were located in different exons of the same ABCA2 gene in both species . This work therefore provides the first insight into the detailed mode of action of a Cry2A toxin , which is conserved across different lepidopteran species , and is of considerable significance for the management of Bt resistance globally .
The absence of crossing-over in female Lepidoptera makes it possible to map a recessive trait such as the Cry2Ab resistance in SP15 to a linkage group using biphasic linkage analysis with AFLPs as genetic markers [18] . Progeny from a female-informative backcross family were bioassayed with a discriminating dosage of Cry2Ab; 161 AFLPs segregating in this family were grouped into 31 independently assorting linkage groups . Linkage to Cry2Ab resistance was tested by comparing bioassayed survivors with untreated control progeny . Only AFLP linkage group ( LG ) 8 showed a significant association with resistance ( χ2 = 19 . 44 , P < 0 . 001; Fig 1 ) ; all 29 treated survivors were homozygous for the SP15 homolog of this linkage group . Southern blot analysis of RFLPs in an unrelated Cry2Ab-susceptible H . armigera family showed that one AFLP from LG8 was linked to ribosomal protein gene RpL22 . RpL22 in B . mori is located on chromosome 17 ( BmChr17 , KAIKObase [51] ) . Using specific probes for additional ribosomal protein genes , the Cry2Ab-resistance-associated linkage group in H . armigera was also shown to carry genes for RpL38 and RpS24 , confirming homology with BmChr17 [52] . This assignment excludes a number of previously-identified genes as candidates for Cry2Ab resistance . Chromosomes harbouring homologs of previously-identified Cry1Ac resistance mutations in H . virescens include BmChr06 with the 12-cadherin-domain protein [27] , BmChr15 with the ABCC2 protein [34] , and BmChr21 with the BtR-5 gene [53] . Moreover , genes for previously identified Cry1Ac binding proteins map to chromosomes other than BmChr17: several aminopeptidase genes are located on BmChr09 [54] , a membrane-bound alkaline phosphatase gene maps to BmChr03 [23] , and the P252 glycoprotein gene is on BmChr25 [55 , 56] . Although different levels of cross-resistance between Cry1A and Cry2A toxins have been reported in H . armigera from China [57 , 58] , in H . virescens [59] , H . zea and P . gossypiella ( [40] , see also [60] ) ; independent segregation of BmChr17 relative to all of these other chromosomes is nevertheless consistent with the absence of cross-resistance between Cry1Ac and Cry2Ab in both the SP15 H . armigera [42] and Hp4-13 H . punctigera [9 , 10] lines , thereby supporting the notion that the two toxins have different modes of action . However , we found that the ortholog of the bre-5 glycosyltransferase gene in a mutant of the nematode C . elegans resistant to the Cry4B toxin [19] is located on BmChr17 ( Fig 2 ) . This gene was therefore further investigated as a candidate gene for Cry2Ab resistance . Additional linkage mapping in a male-informative backcross and two F2 families was performed to further localize the resistance locus . A preliminary map based on 72 progeny from these families gave the gene order and recombination values as follows: Bre−5– ( 0 . 16 ) –Cry2Ab resistance locus– ( 0 . 10 ) –RpL22– ( 0 . 06 ) –RpS24 The order and spacing of the three marker loci was similar to that in B . mori on BmChr17 . However , the large fraction of recombinants between the resistance locus and bre-5 ruled out the latter as a candidate ( Fig 2 ) . The Cry2Ab resistance gene was further localised within BmChr17 using recombinational mapping in backcrosses with F1 males . For this work , markers were developed from H . armigera orthologs for genes mapped along BmChr17 . Sequences allowing design of EPIC-PCR primers for the H . armigera orthologs for these genes were obtained from transcriptome sequencing of midgut RNA extracted from larvae of the GR susceptible colony . Recombinational analysis of selected markers in H . armigera showed the linkage order of these markers to be the same as in B . mori ( Fig 2 ) , greatly assisting the subsequent analysis which employed the B . mori genome as a reference framework . Analysis of recombination rates between 3 of these markers ( RpL38 , Zip2 , VGCal-A ) and the Cry2Ab resistance allele placed it between Bre-5 ( BGIBMGA005534 ) and RpL22 ( BGIBMGA006986 , at nt 14152986 ) . The markers for the voltage-gated channel protein gene ( orthologous to BGIBMGA007009 , starting at 12341859 on BmChr17 –see S1 Table ) further restricted the area containing the resistance gene . The target area could however be more narrowly defined , since the gene is ~10cM from BGIBMGA005534 and ~16cM from BGIBMGA006986; these genes are located at ~3Mbp and ~10Mbp respectively on the BmChr17 sequence ( see S1 Table ) . In fine scale analysis of this region , markers for the genes BEACH , ANK_SAM , NaPT , DUF410 , and AN_Peroxidase all showed recombination with the resistance trait . Of these , the marker for DUF410 ( BGIBMGA007299 , located at 7124451 on BmChr17 ) most closely restricted the target region on the proximal side , corresponding to less than 3Mbp of BmChr17 , and containing fewer than 30 genes ( S1 Table ) . Two ABC transporter A subfamily genes are located adjacently between nts 8466000–8564000 on BmChr17 . The first of these , termed BmABCA1 , is well-predicted as BGIBMGA007221 , while the other , BmABCA2 , includes the partial predictions BGIBMGA007218 and BGIBMGA007217 ( see [61 , 62] for analysis of the original uncorrected gene models ) . The sequence of HaABCA1 , the H . armigera ortholog to the BGIBMGA007221 BmABCA1 gene , was obtained from RNAseq libraries , from total larvae of the susceptible GR colony . The EPIC-PCR marker for HaABCA1 ( ABCA1; S2 Table ) gave a genotype profile consistent with tight linkage to the Bt Cry2Ab resistance allele; the F2 bioassayed offspring ( homozygous allele size of 272bp / 272bp ) was identical to the SP15 grandmother ( 272bp / 272bp ) in 100% of all samples tested ( final n = 72 ) . The GR grandfather was heterozygous with alleles 264bp / 282bp , leading to the F1 male being heterozygous with allele sizes 272bp / 282bp . The F2 control ( n = 20 ) gave the expected 50:50 ratio with n = 11 being 272bp / 272bp homozygous and n = 9 heterozygous ( 272bp / 282bp ) . This tight linkage between the HaABCA1 gene and resistance made it a candidate for being the target of the resistance mutation . To assess whether HaABCA1 was the real location of resistance mutations , we checked whether this gene is expressed in the midgut . No evidence for significant midgut expression of ABCA1 was found in either B . mori ( Bm-MDB , B . mori Microarray Database [63]; see S1 Table ) or H . armigera , making it unlikely that this gene is actually involved in resistance . Initially detected in total larval and pupal transcripts , its expression was more specifically evident in larval foregut , hindgut , trachea and haemocytes . However the adjacent ABCA2 gene was significantly expressed in the midgut of both Bombyx and H . armigera; of the genes in this region of BmChr17 , it is among the most highly expressed in the midgut [63] ( see S1 Table ) . The full-length transcript of the H . armigera ortholog HaABCA2 ( Fig 3 ) encodes a protein of 1 , 742 amino acids with 67 . 28% identity to the BmABCA2 gene in B . mori ( S1 Fig ) . We therefore further explored whether any changes were evident in the transcripts of the ABCA2 gene in midgut RNA from larvae of the resistant line . The ABCA2 candidate gene cDNA was fully sequenced from a SP15 resistant individual that identified a 73bp deletion at exon 16 ( ‘ G CTA GGA GTT CTG CGT TAC GTC ATG TCT TTA TCA CCA ACC ATT AGA ACT AGG TGG TTG TCG TTG GAA GAA GGG’ from nucleotides 2 , 889–2 , 961 ) / 8bp ( ‘C GGT TAA G’ ) insertion mutation ( = allele 1 , Ha2Ab-R01 ) ( Fig 4a and 4b ) of the coding sequence which resulted in the replacement of leucine ( L ) by glycine ( G ) at position 964 , followed by a stop codon downstream from the mutation site at position 965 . The 8bp ‘C GGT TAA G’ nucleotides matched completely to part of intron 16 ( intron 16 nucleotide positions 56 to 62 ) of the SP15 H . armigera line . From the sequence of this mutated cDNA , we designed primers to screen additional independently isolated field resistant lines from 2005 ( line 5–405 ) ; 2006 ( lines 6–364 and 6–798 ) , 2009 ( line 9–4802 ) , 2010 ( line 10–485 ) , and 2012 ( line 12–2169 ) , four of which ( i . e . , 5–405 , 9–4802 , 10–485 , 12–2169 ) showed the same 73bp deletion/8bp insertion mutation as identified in the SP15 individual . A second resistance allele ( = Ha2Ab-R02 ) with a 5bp ( ‘ACA AG’ ) deletion mutation at nucleotides 3 , 127–3 , 131 of the coding sequence was identified in a homozygous individual from the Cry2Ab resistant lines 6–364 . A heterozygous resistant individual from line 6–798 was identified to possess one Ha2Ab-R02 allele , and a third resistance allele ( = Ha2Ab-R03 ) at nucleotides 4 , 104–4 , 108 that represented a 5bp ( GAATA ) nucleotide duplication ( Fig 4b ) , similar to the target site duplication ( TSD ) signature that is widespread in the H . armigera genome due to transposable element transposition activities ( e . g . , see [64] ) . cDNA sequencing of both the 6–364 and 6–798 lines identified the presence of the Ha2Ab-R02 allele as homozygous in the 6–364 line , and also both Ha2Ab-R02 and Ha2AB-R03 alleles at exons 18 and 24 respectively in line 6–798 , thereby confirming that this resistant line was heterozygous for the ABCA2 gene ( see S2 Fig ) . All three mutations identified to date in the Cry2Ab resistant lines are located at the 3’ region of the 5 . 1Kb coding sequence , and result in truncation of the protein . A summary of the three Cry2Ab resistance alleles identified in H . armigera is presented in Fig 4a and 4b . Sequence analyses of the SP15 , 6–364 and 6–798 resistant individuals confirmed that no other INDELs or nonsense mutations were present in coding regions of the ABCA2 gene . Similarly , additional sequence analyses from multiple susceptible H . armigera individuals showed the predicted fully functional ( non-truncated ) ABCA2 gene , while nucleotide variation ( <1% ) between Cry2Ab susceptible H . armigera ‘GR’ lines resulted in nine amino acid changes , six of which involved nonsynonymous substitutions between amino acids with hydrophobic side-chains ( e . g . , valine ( V ) , isoleucine ( I ) , tyrosine ( Y ) , phenylalanine ( F ) , methionine ( M ) , and leucine ( L ) ) , and one amino acid substitutions of each between lysine ( K ) and threonine ( T ) , alanine ( A ) and proline ( P ) , and glycine ( G ) and glutamic acid ( E ) ( S2 Fig ) . To confirm the significance of the HaABCA2 mutations in resistant lines , we asked whether susceptible individuals collected from the field carried the same INDEL mutations or other inactivating mutations in the H . armigera ABCA2 gene . Starting with single pairs of field-collected insects , F1 pools from each mating pair generated F2 progenies ( n = 90 ) that were screened against a discriminating dose of Cry2Ab and identified those pairs whose F2 progeny all died as carrying only susceptible alleles . This test is enough to exclude any resistance-conferring alleles ( occurring as heterozygotes ) amongst the grandparents . Ten individual 3rd instar larvae representing 10 different field-collected susceptible populations were RNA extracted and RT-PCR used to generate cDNA . To screen for any evidence of mutations in the ABCA2 gene , PCR of the cDNA and sequencing using appropriate primer pairs ( see S1 Table ) were performed . No evidence for any inactivating mutations was found in any of these 10 individuals , i . e . all contained untruncated transcripts at exons 16 , 18 and 24 where Ha2Ab-R01 , R02 and R03 alleles were detected , respectively . A total of two transmembrane domains ( TMDs; i . e . , TMD 1 , TMD 2 ) with each consisting of six transmembrane helices ( TM I-VI in TMD 1; TM VII-XII in TMD 2 , Fig 5 ) were predicted from the HaABC2 sequence that corresponded to those characteristic of the ABC transporter subfamily A . As for other members of this subfamily , N-glycosylation sites were also predicted for both of the extracellular domain ( ECD ) loops between TM I and TM II , and between TM VII and TM VIII ( Figs 3 and 5 ) . The intracellular loop between TMD 1 and TMD 2 ( i . e . , between TM VI and TM VII ) and after TM XII contained the highly conserved regions for ATP Nucleotide Binding Fold 1 ( NBF1 ) ( including the Transporter signature Motif 1; TpM1 ) , and NBF2 ( and TpM2 ) , respectively . Each of the mutant alleles Ha2Ab-R01 , Ha2Ab-R02 and Ha2Ab-R03 introduced stop codons into the reading frame , causing significant truncations of the HaABCA2 protein ( Fig 4 ) . The first two mutations introduced stop codons in the extracellular loop between TM VII and TM VIII of TMD 2 , whereas the 5bp insertion that resulted in the Ha2Ab-R03 allele occurred just one amino acid after TM XII of TMD 2 . All three mutations therefore truncated the protein before the second nucleotide-binding domain NBF2 ( Fig 5 ) , which would render the ABC transporter completely inactive , even if the protein were expressed and integrated into the cell membrane . In the Cry2Ab-resistant H . punctigera Hp4-13 strain , the allele ( Hp2Ab-R04 ) encoding the homolog of HaABCA2 was found to contain a 14bp deletion ( Fig 4 ) . This deletion disrupts the coding region of the transcript by introducing frame shifts that lead to a missense mutation and to the loss of the TpM2 transporter motif at the NBF2 ( Figs 4 and 5 ) . Although linkage mapping was not performed to conclusively associate this mutation with the resistant phenotype in H . punctigera , the fact that the same gene is mutated in a resistant strain in this species strongly supports the role of this gene as the target of mutations conferring resistance to Cry2Ab . We next asked whether sequences for these ABCA proteins existed in genomes of other Lepidoptera , including some ( H . virescens [65] , Plutella . xylostella [66]; B . mori [67] ) known to be susceptible to Cry2Ab . We examined the published genome sequences of Danaus plexippus [68] , Heliconius melpomene [69] , and P . xylostella [70 , 71] . Existing predictions of the genes were often inaccurate , e . g . , as with the two partial predictions for BmABCA2 , and ABCA1 was predicted as two separate partial proteins for the D . plexippus genome [68] . We used Scipio [72] and FGENESH [73] as well as additional transcriptomic data to generate complete predictions ( GenBank Accession numbers KP219762-KP219770 ) . In each species , the ABCA1 and ABCA2 genes were present , and situated adjacently in a tail-to-tail orientation just as in B . mori . Orthology was further confirmed by conserved flanking genes; in each species the ortholog of B . mori mitochondrial ribosomal protein S7 ( GenBank Accession XP_004927481 ) occurred upstream of ABCA2 , and the ortholog of B . mori ubiquitin carboxyl-terminal hydrolase ( GenBank Accession XP_004927480 ) occurred upstream of ABCA1 . An alignment of the predicted ABCA1 and ABCA2 proteins along with the respective Drosophila protein showing greatest similarity ( S1 Fig ) was used to construct a phylogenetic ( ML ) tree . This tree indicated that the gene duplication creating ABCA1 and ABCA2 likely occurred in the common ancestor of the Lepidoptera shown ( Fig 6 ) .
The importance of Bt toxins for insect pest and disease control has stimulated enormous interest in the study of their mode of action . For Cry1A toxins , there is specific and saturable binding to membrane targets , and a sequential mode of action has been proposed [25 , 39] . The toxin first binds to the 12-cadherin domain protein , resulting in processing and accelerated oligomerization before binding to membrane-bound glycosylated proteins such as aminopeptidases , alkaline phosphatase and other glycoproteins [25 , 26]; the integral membrane ABCC2 protein then facilitates pore insertion . Cry1A and Cry2A proteins have comparable three-domain structures [74 , 75] , making them likely to act in similar ways as pore-forming toxins . Specific and saturable binding to membranes was also recently shown for Cry2Ab [22 , 75] , and resistance is associated with a loss of binding [21] . Despite these similarities , toxicity of Cry2Ab in general is unaffected by mutations conferring Cry1Ac resistance . Specifically , mutations in APN or cadherin or ABCC2 do not render insects resistant to Cry2Ab , so that this toxin must be binding to one or more different targets . The identification of ABCA2 suggests a mode of action of Cry2Ab differing slightly from that proposed for Cry1A toxins . ABCA2 carries two extracellular domains that are present as long loops between helices TM I and TM II , and between helices TM VII and TM VIII ( Figs 3 and 5 ) . Both of these loops are glycosylated in mammals [76] , and six glycosylation sites are predicted for HaABCA2 ( Fig 3 ) . In contrast , for the lepidopteran ABCC2 , the corresponding loops are very short and contain no glycosylation signals [34] . We hypothesize that Cry2Ab also has a sequential mode of action in which the ABCA2 protein itself is able to provide both binding and pore insertion functions . Specifically , Cry2A toxins would , upon activation [75] bind to the glycosylated ECD loops in TMD 1 and/or TMD 2 . This binding could form the basis of oligomerization and bring the pre-pore structure close to the TMDs for pore insertion , as proposed for ABCC2 [34] . It is possible that other proteins may also be involved in Cry2Ab binding and pore formation , particularly since mammalian ABCAs have been suggested to occur in multi-protein complexes in the membrane [77] . Interestingly , the ABCA2 mutations confer resistance to very high concentrations of Cry2A [42] , as would be expected if both receptor and pore insertion functions are simultaneously blocked . Similarly , in H . virescens the ABCC2 mutation results in higher levels of resistance to Cry1Ac than does the mutation in cadherin , but when both are homozygous in the same strain and both receptor and pore insertion functions are blocked , extremely high resistance levels result [34] . To what extent are these findings likely to apply to other Bt toxins of the Cry2A family ? Phylogenetic analyses based on the shared common three-domain structure [78] showed that Cry2Aa and Cry2Ai are sister toxin groups and occupy a basal position to both Cry2Ab/Cry2Ag and Cry2Ae/Cry2Ah clades . Cross-resistance between Cry2Ab and Cry2Aa has been demonstrated in the SP15 strain of H . armigera [42] , and between Cry2Ab and Cry2Ae in both H . armigera and H . punctigera [21] . Resistance to Cry2Aa has also been identified in H . virescens [53 , 79] , in H . zea [15] , in P . gossypiella [40] , and in Ostrinia nubilalis [80]; ABCA2 remains to be investigated in these species . Cry2A toxins are also toxic to some Diptera [81 , 82] ( but see [83] ) , with Cry2Ab recently shown to be effective against the malaria mosquito vector Anopheles gambiae [82] . Cry2Ab was ineffective against Aedes aegypti [82] but the similar toxin Cry2Ag was highly effective [84] . Seven ABC transporters of the A subfamily are present in the B . mori genome ( three on BmChr17 , and one each on chromosomes 5 , 14 , 16 , and 19; see [61 , 62] for analysis of the initial gene models ) , and similar numbers have been found in other Lepidoptera with sequenced genomes . This subfamily has been well characterised in vertebrates; there are 12 members known in humans [85] and a similar number in the mouse; the nomenclature for these differs from that of insects . The HaABCA2 gene and the other insect genes shown in Fig 6 belong to an insect-specific clade within this subfamily , with no direct orthologs among the vertebrate genes [86] . The human ABCA genes have been extensively analysed and are expressed in a variety of tissues , with most being involved in lipid transport and trafficking . Mutations in human ABCA2 ( not orthologous to lepidopteran ABCA2 ) are associated with early-onset of Alzheimer’s disease [87 , 88] . The mouse ABCA2 ( an ortholog of the human ABCA2 ) has a possible role in regulating cholesterol homeostasis and low-density lipoprotein receptor metabolism in N2 neuroblastoma cells [89] , with knock-out causing a ‘shaky’ ( tremor ) phenotype [90] . Mutations in the mouse ABCA3 gene , which is expressed in lung tissues , are associated with a foetal surfactant deficiency that is fatal . However , in H . armigera ( but not H . punctigera ) , the Cry2Ab resistant line homozygous for ABCA2-inactivating mutation has no demonstrated substantial fitness costs compared to Cry2Ab susceptible insects [91] . Whether this is due to functional redundancy with another of the midgut-expressed ABCAs remains to be determined . The frequency distribution of resistant ABCA2 alleles identified to-date is non-uniform . Seven resistant H . armigera lines , isolated independently from the field between 2002 and 2012 , produced three resistant alleles . The Ha2Ab-R01 allele was present in five lines: SP15 , 5–405 , 9–4802 , 10–485 and 12–2169; the Ha2Ab-R02 allele was homozygous in 6–364 and present in heterozygotes in 6–798; the Ha2Ab-R03 allele was the alternative allele in the heterozygous 6–798 line . A single mutant allele , Hp2Ab-R04 , was found in the one H . punctigera resistant line HP4-13 . Thus some alleles in H . armigera were common enough to be recovered several times from the field . Whether this is due to some selection by an unknown agent in the Australian environment as proposed [92] remains to be tested . The still-rare resistance-conferring alleles identified in field populations occur at a limited number of locations in the gene . If confirmed by studies of further alleles , this raises the possibility that DNA-based screens will allow monitoring of the spread of Bt resistance in H . armigera and H . punctigera . Although PCR-based screens [29 , 93] for mutations in the 12-cadherin domain protein of H . armigera that confers resistance to Cry1Ab and Cry1Ac identified the same allele ( r1 ) in field material from northern China as that originally identified as conferring resistance [12] , this has not always been the case . For H . virescens [94] , for example , different mutations in the same gene were identified by the F1 screen ( i . e . , mating field-caught individuals with an existing homozygous resistant strain and testing the F1 offspring ) . For P . gossypiella , screening in India [95] found different cadherin gene mutations to those originally identified in Arizona [96] . Every Cry2Ab-resistant line from an F2 screen that has been molecularly characterized has shown mutations in the ABCA2 gene . This confirms the value of using the less expensive F1 screen with ABCA2-mutant lines to extend the estimation of Cry2Ab resistance allele frequencies in Australia . Incorporating PCR-based screening will further improve detection efficacies of ABCA2-based resistance in the field , enabling more accurate and faster estimates of resistance allele frequencies , and is especially relevant for the analysis of historical field material generated through F1/F2 screening methods [43 , 92] , and for tracking spatial and temporal movement patterns of resistance alleles across the landscape . Further characterisation of resistance-conferring ABCA2 alleles will also help to resolve the current discrepancy between the F2 screen and the F1 screen in estimating allele frequencies [91] . It will be important to determine whether Cry2Ab-resistance-conferring ABCA2 mutations occur in H . armigera elsewhere in its geographic range , including its recent incursions into the Americas [97 , 98] . Finally , examination of ABCA2 may provide insight in several species where the Cry2A resistance mechanism is still unknown , including H . virescens [53] , P . gossypiella [40] , H . zea [41] , and Trichoplusia ni [99] .
As a result of an F2 screen in 2002 , the first H . armigera Cry2Ab resistant strain ( Sp15 ) was established from a single pair of moths collected as eggs on corn near Griffith , New South Wales ( NSW ) , Australia [42] . Detailed descriptions of the techniques employed have been provided [42 , 92] . F1 progeny from that pair were intercrossed and the resultant F2 larvae exposed to a screening concentration of Cry2Ab in ground leaf material of the cotton variety Sicala V-2 transformed with the B . thuringiensis variety kurstaki cry2Ab gene construct . Survivors among the F2 formed the basis of the resistant colony Sp15 . Since 2003 , F2 screens with H . armigera and H . punctigera performed as part of a resistance monitoring program have isolated additional lines [10] . The isolated H . punctigera Cry2Ab resistant line ( Hp4-13 ) was established from eggs collected at St George , Queensland , Australia in 2004 [43] . Complementation tests for allelism established that one or more alleles at the same locus conferred resistance to Cry2Ab in five lines of H . armigera derived from the field in Australia from 2002 to 2006 , including SP15 and 5–405 ( previously named NA405 ) [100] . Assignment of the Cry2Ab resistance locus to an AFLP linkage group was carried out using the resistant line SP15 and the susceptible GR line . The initial cross was a SP15 Cry2Ab r/r ♂ x GR Cry2Ab s/s-♀ , yielding family G . An F1 female from family G was crossed to an SP15 male to produce the female informative backcross family F2031 . Backcross progeny were bioassayed using the discriminatory Bt Cry2Ab concentration [92] to select for homozygous resistant ( Cry2Ab r/r ) individuals . Additional backcross progeny were not exposed to Cry2Ab , to serve as controls . AFLPs [101] from genomic DNA of grandparents , parents and 59 progeny of family F2301 were analysed for linkage using the method of Heckel et al . [18] . Twenty-nine progeny were survivors of exposure to Cry2Ab and 30 were untreated controls . AFLPs were grouped using the program DBM3Lnk . p as in Heckel et al . [18] . As expected from achiasmatic oogenesis in female Lepidoptera , no recombinants were found within AFLP linkage groups . Linkage to resistance was tested for each linkage group using signed interaction chi-squared tests with one degree of freedom [102] , with a Bonferroni correction for 31 linkage groups . One AFLP band from the only linkage group with a significant association with resistance was cut out of the gel , reamplified , cloned and sequenced ( GenBank Accession No . KJ419919 ) . The insert was hybridized to a Southern blot made from an unrelated Bt-susceptible H . armigera family in which several ribosomal protein genes had previously been mapped , enabling comparison to the homologous ribosomal protein genes of Bombyx mori . This showed AFLP group 8 in family F2301 to correspond to B . mori chromosome 17 ( BmChr17 ) . Additional linkage mapping in a male-informative backcross ( G2016 ) and two F2 families ( G2020 , G2029 ) was performed to further localize the resistance locus . Offspring from these families that had survived the discriminating concentration and were presumably homozygous for the SP15-derived resistance allele were examined for recombinants at marker loci . H . armigera homologs of ribosomal protein genes RpL22 and RpS24 on B . mori BmChr17 were sequenced to identify polymorphisms to be used in mapping . The gene bre-5 on BmChr17 was considered a candidate for the resistance gene because of its role in Cry4B resistance in the nematode Caenorhabditis elegans [19] , and was also mapped using sequence variation in the coding region and a PCR-RFLP using a polymorphic PstI restriction site . To establish an appropriate mapping family , a GR Cry2Ab susceptible homozygous male ( Cry2Ab s/s ♂ ) was mated with a SP15 Cry2Ab resistant homozygous female ( Cry2Ab r/r ♀ ) . The F1 susceptible heterozygous male ( Cry2Ab r/s ♂ ) was back-crossed to a SP15 female to obtain F2 offspring of either homozygous resistant ( Cry2Ab r/r ) or heterozygous susceptible ( Cry2Ab r/s ) genotypes in equal proportions . Approximately 300 F2 offspring were bioassayed using the discriminatory Bt Cry2Ab concentration [92] to select for homozygous resistant ( Cry2Ab r/r ) individuals . Control ( n = 100 ) F2 offspring were not bioassayed and were included in subsequent genotyping experiments using EPIC-PCR markers as described below . EPIC PCR markers used in this study were designed using the primer designing criteria previously reported by [103] for H . armigera . Briefly EPIC-PCR primers were designed using the primer analysis software Oligo Version: 7 . 17 ( Molecular Biology Insights , Inc . , Cascade , CO 80809 , USA ) and avoiding false primer annealing sites for both forward and reverse primer , with no or minimal hairpin structures and primer dimmer formation . We also designed the EPIC-PCR primers with intron amplicon of typically less than 500bp such that polymorphisms in F2 cross can be easily scored . Intron sizes were estimated based on B . mori gene annotation . EPIC-PCR primers were optimised prior to having a fluorescent tag ( FAM , HEX or TET ) attached to the 5’ end of the forward primer . Amplicons of the mapping family from individual EPIC-PCR primer pairs were visualised on 1–1 . 5% agarose gels prior to being purified by acetic acid/ethanol precipitation and sent to Genetic Analysis Facility ( GAF ) at James Cook University ( JCU ) for genotyping . PCR conditions , and genotyping procedures were previously described [103 , 104] . A list of all EPIC-PCR primers used in this study can be found in S2 Table . Genomic DNA was extracted using the Qiagen Blood and Tissue extraction kit ( Qiagen Cat . #69506 ) . For the founding grandparents ( i . e . , F0 ) and parents ( i . e . , F1 ) one leg each was used in gDNA extraction , with gDNA eluted in 200°L of the AE buffer . Bioassayed and control F2 samples were collected as 3rd instar larvae and gDNA was extracted as for the parents and grandparents . All genotyping with EPIC-PCR markers involved screening of grandparents , F1 parents , 72 bioassayed ( Cry2Ab r/r ) offspring and 20 control F2 offspring ( i . e . , either Cry2Ab r/r or Cry2Ab r/s ) . Under the linkage mapping pattern , genome/chromosome walking towards the resistance gene should generate reduced recombination rates in the resistant F2 as one approaches the genomic region of interest . Messenger RNA sequencing was done in order to generate full-length transcripts in H . armigera for candidate genes , identify resistant alleles where cDNA amplification failed and to identify the homologous candidate genes in H . punctigera . Total RNA was extracted from the midgut of third-instar larvae or whole larvae using the TRIzo Plus RNA purification kit ( Life Technologies , Cat # 12183555 ) and dried down for shipping with RNAstable Tube Kit ( Biometrica Cat . # 93221–001 ) . RNAseq library preparation , sequencing and bioinformatic analysis was done according to standard Illumina protocols by the Beijing Genomics Institute ( BGI ) in Shenzen , China . Except for the resistant line 7–183 which used gDNA as a template for sequencing , candidate genes from the remaining resistant lines were completely sequenced using a cDNA template from 3rd instar larvae prepared using an RNA extraction kit ( Qiagen RNeasy mini kit , Cat . # 74106 ) , and trace genomic DNA contaminants removed using the Qiagen RNase-Free DNase set ( Cat . # 79254 ) . First strand cDNA was synthesised using the Invitrogen SuperScript III RT First Strand Synthesis System for RT PCR ( Cat . # 18080–051 ) , in the presence of RNase H . All sequencing was performed at the John Curtin School of Medical Research , Australian National University ( ANU ) , and used the ABI BigDye v3 chemistry . Contig assembly used the Staden pregap4 and Gap4 software [105] and was visualised using Artemis ( Release 12 . 0 ) [106] . Sequences generated and used in this study have been deposited in GenBank ( Accession numbers KP259910 , KP259911 , KP259912 ) . The amino acid sequence predicted from a complete mRNA sequence from a Cry2Ab susceptible individual belonging to the GR-line was used to predict the domain structure of the H . armigera ABCA2 protein . The protein prediction software Split V3 . 5 <http://split4 . pmfst . hr/split/4/> [107] was used to search for transmembrane protein secondary structure ( i . e . , transmembrane helices ) . In the mouse RmP ABC transporter , several N-glycosylation sites were predicted on the protein’s extracellular domains [76 , 108] . We used the NetNGlyc 1 . 0 Server <http://www . cbs . dtu . dk/services/NetNGlyc/> developed to predict N-Glycosylation sites in human proteins for the purpose of predicting N-Glycosylation sites in the protein sequences . The software uses artificial neural networks to examine for Asn-Xaa-Ser/Thr sequence context . Sequences in the transcriptome databases corresponding to candidate genes were identified by standalone BLAST . Homologs in GenBank were identified using BLAST and homologous gene clusters identified in NCBI and in Ensembl . Orthologous genes from other lepidopteran genomes were retrieved using their online databases from public domains as cited in the appropriate sections below . Protein sequences were aligned using Multiple Alignment using Fast Fourier Transform ( MAFFT ) [109] <http://www . ebi . ac . uk/Tools/msa/mafft/> and phylogenetic tree ( maximum likelihood ( ML ) with rapid bootstrapping ) inference using RAxML-HPC2 on XSEDE ( 8 . 0 . 24 ) ( available at the CIPRES Science Gateway V3 . 3 ) <http://www . phylo . org/sub_sections/portal/> [110–112] , and redrawn using Dendroscope version 2 . 4 [113] . | Transgenic crops expressing the insecticidal protein Cry2Ab from Bacillus thuringiensis ( Bt ) are used worldwide to suppress damage by lepidopteran pests , often used in combination with Cry1Ac toxin to delay resistance evolution . Until now , the Cry2Ab mode of action and the mechanism of resistance were unknown , with field-isolated Cry2Ab resistant Helicoverpa armigera showing no cross-resistance to Cry1Ac . In this study , biphasic linkage analysis of a Cry2Ab-resistant H . armigera family followed by EPIC marker mapping and candidate gene sequencing identified three independent INDEL mutations in an ATP-Binding Cassette transporter subfamily A gene ( ABCA2 ) . A deletion mutation was identified in the same gene of resistant H . punctigera . All four mutations are predicted to truncate the ABCA2 protein . This is the first molecular genetic characterization of insect resistance to the Cry2Ab toxin , and detection of diverse Cry2Ab resistance alleles will contribute to understanding the micro-evolutionary processes that underpinned lepidopteran Bt-resistance . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Insect Resistance to Bacillus thuringiensis Toxin Cry2Ab Is Conferred by Mutations in an ABC Transporter Subfamily A Protein |
Coastal marine ecosystems can be managed by actions undertaken both on the land and in the ocean . Quantifying and comparing the costs and benefits of actions in both realms is therefore necessary for efficient management . Here , we quantify the link between terrestrial sediment runoff and a downstream coastal marine ecosystem and contrast the cost-effectiveness of marine- and land-based conservation actions . We use a dynamic land- and sea-scape model to determine whether limited funds should be directed to 1 of 4 alternative conservation actions—protection on land , protection in the ocean , restoration on land , or restoration in the ocean—to maximise the extent of light-dependent marine benthic habitats across decadal timescales . We apply the model to a case study for a seagrass meadow in Australia . We find that marine restoration is the most cost-effective action over decadal timescales in this system , based on a conservative estimate of the rate at which seagrass can expand into a new habitat . The optimal decision will vary in different social–ecological contexts , but some basic information can guide optimal investments to counteract land- and ocean-based stressors: ( 1 ) marine restoration should be prioritised if the rates of marine ecosystem decline and expansion are similar and low; ( 2 ) marine protection should take precedence if the rate of marine ecosystem decline is high or if the adjacent catchment is relatively intact and has a low rate of vegetation decline; ( 3 ) land-based actions are optimal when the ratio of marine ecosystem expansion to decline is greater than 1:1 . 4 , with terrestrial restoration typically the most cost-effective action; and ( 4 ) land protection should be prioritised if the catchment is relatively intact but the rate of vegetation decline is high . These rules of thumb illustrate how cost-effective conservation outcomes for connected land–ocean systems can proceed without complex modelling .
Widespread degradation and loss of coastal marine ecosystems has occurred over the previous centuries and has accelerated in recent decades [1–5] . These changes compromise the delivery of important ecosystem services to human society [6] . Coastal marine ecosystems pose a particular challenge to environmental managers because they are exposed to threats occurring both in the ocean ( e . g . , overfishing , direct damage ) and on land . The conversion of native terrestrial vegetation for agriculture , urbanization , and industry increases runoff [7] , causing degradation and die-offs of coastal ecosystems such as coral reefs [8] and seagrass meadows [2] . These declines threaten the functional integrity of coastal and marine ecosystems and the services they provide , such as food supplies , coastal protection , and climate regulation [9–11] . Consequently , the conservation of coastal species and ecosystems requires a mixture of both marine and terrestrial conservation actions [12–16] . Conservation prioritisation of marine ecosystems and adjacent landscapes traditionally focuses on protecting intact habitats , in either marine and or terrestrial realms , from future degradation [e . g . 17 , 18–20] [but see 21] . Ecological restoration is commonly considered a less preferred management strategy than protection [22] , particularly in marine environments , where restoration costs are high and success rates are low [23] . However , restoration can deliver better ecological outcomes than protection , depending on existing land uses , conservation intervention costs , and ecosystem expansion rates [24] . Compared to other actions , restoration is rarely considered [25] , and trade-offs between restoration and protection actions have never been evaluated across complex land–sea systems . Comparing the costs of conservation actions , both on land and in the ocean , with the benefits accrued in the marine ecosystem ( ‘cost-effectiveness’ ) is at the forefront of conservation planning for land–sea ecosystems [e . g . 17] . Incorporating exchanges across the land–sea interface is challenging , requiring the integration of data and models across the terrestrial , freshwater , and marine realms [12–14 , 20] . Recent advances have allowed the benefits of terrestrial actions on marine ecosystems to be estimated [17–20 , 26–33] , but in practice , land–sea conservation planning has rarely explicitly quantified how the management of terrestrial threats impacts marine ecosystems . For instance , recent implementations of the ‘Reef 2050 Long-term Sustainability Plan’ for the Great Barrier Reef and the ‘Chesapeake Bay Total Maximum Daily Load’ programs aim to minimise sediment , nutrient , and pollutant delivery to the ocean and assume that marine ecosystems will respond positively [34 , 35] but do not predict the effect size of the marine ecosystem response . As a result , it is not clear that their terrestrial focus will outperform actions in the marine environment , which have the advantage of directly affecting the management goal . To compare and prioritise actions across the land–sea interface , we need to identify the links between ( 1 ) the amount of land-based actions required to reduce a threat on receiving marine environments and ( 2 ) the amount of change in the marine ecosystem triggered by such a reduction . We propose that integrated land–sea planning must compare the cost-effectiveness of 4 broad conservation actions: protect habitat on the land , protect habitat in the ocean , restore habitat on the land , and restore habitat in the ocean . Here , we develop a repeatable and transferable approach to determine which of those 4 actions maximises the extent of intact marine habitat for a given budget and project timeframe ( Fig 1 ) . The model extends an existing terrestrial model [24] across the land–sea interface . It is general in structure and could potentially apply to any marine system that is affected by sediment runoff and marine-based threats . In the original terrestrial model [24] , the landscape is divided into 4 states describing the condition of the native vegetation—intact and unprotected , intact and protected , cleared , or restoring . The act of restoring or protecting habitat moves it between these different states . In our expanded model , we consider the state of habitat in both a landscape and adjacent seascape , which are connected by sediment runoff from the land into the ocean . Cleared terrestrial habitat increases sediment loads , which reduces water clarity in the adjacent ocean . The resulting decrease in light reaching the seafloor reduces the amount of habitat suitable for light-dependent species [13] . We focus on suspended sediments because they are a key driver of marine ecosystem condition in many inshore areas [2 , 8 , 36] but acknowledge the importance of the other components of runoff more broadly , including toxicant and nutrient loads . Importantly , our model assumes that the marine ecosystem is sensitive to both sediment runoff [37] and marine-based threats and that there is habitat in appropriate condition for marine restoration; if these conditions are not met , then approaches targeted towards either land- or ocean-based threats would be required . The model is spatially implicit , i . e . , it is parameterised by spatial data ( see Materials and methods , S1 Text , S1 Table ) . We apply this model to a case study of seagrass meadows in Moreton Bay and riparian areas in adjacent catchments in Queensland , Australia ( Fig 2 ) . Seagrass meadows are an excellent test system because they provide a suite of ecosystem services and are strongly influenced by both land-based processes and direct local impacts in the ocean [2 , 38–40] . The catchments draining into Moreton Bay are heavily modified , with only 20% intact remnant vegetation . Historical and ongoing land-clearing has significantly increased soil erosion , primarily through the process of gully erosion [41–42] . Suspended sediment delivery has negatively impacted marine ecosystems in the region [43–45] . The site is therefore representative of the wider global challenges posed to marine ecosystems by increased sediment runoff [14 , 15 , 39] . We ultimately aim to identify key factors that determine which broad conservation action is most effective under different circumstances . Therefore , we use the model output and sensitivity analyses to answer 2 questions . One , which of the 4 conservation actions maximises the extent of intact seagrass after 30 years ? And two , under which conditions would our decision-making vary ? Using the results from this and other studies [24 , 37] , we propose simple ‘rules of thumb’ that can help decision makers identify whether restoration or protection , either on the land or in the ocean , will be the most cost-effective approach to improving the state of marine ecosystems . These rules of thumb are likely specific to our study system but may be used as guidelines ( or , alternatively , viewed as hypotheses ) to inform decision-making in other regions until models are parametrised for those sites .
Using our dynamic model of seagrass meadows and riparian areas in adjacent catchments in Southeast Queensland , Australia , we investigated the effect of investment of $50 million ( all costs are in 2015 USD unless otherwise stated ) per year over 30 years into each of 4 separate conservation actions . The model was used to quantify the area of intact seagrass habitat resulting from marine restoration , marine protection , terrestrial restoration or terrestrial protection ( see below , Materials and methods and S1 Text , for detailed descriptions ) . We found that if the objective is to increase the amount of habitat suitable for ( but not necessarily occupied by ) light-sensitive species ( in this case , seagrass ) , then restoration of riparian areas on land is the most cost-effective strategy ( Fig 3A ) . However , this will not necessarily maximise the area of occupied ( ‘intact’ ) marine habitat immediately , as that depends on how fast the marine ecosystem can recover and expand into habitat which was previously unsuitable due to low light availability; there is substantial uncertainty in this parameter ( S2 Table ) . Controversially , we find that the most cost-effective way to maximise the extent of intact marine habitat over decadal timescales is to directly restore the marine ecosystem , despite the higher cost [23] ( Fig 3B ) . Obviously , this conclusion depends on the availability of suitable , unoccupied habitat . If all marine habitat is unsuitable for marine restoration due to low water clarity , then revegetation of riparian vegetation to minimise sedimentation stress is required [46] . Below , we discuss the costs and benefits of each specific conservation action in turn for our study system and then examine how these decisions may vary for other systems . Marine restoration was defined as planting seagrass transplants into habitat that is suitable for , but not presently occupied by , seagrass and was the most cost-effective action for achieving the highest coverage of seagrass habitat after 30 years ( Fig 3B ) . Our modelling assumes that: ( 1 ) it takes 3 years for seagrass transplants to grow , fill in meadow gaps , and become a ‘healthy’ , self-sustaining meadow [47]; ( 2 ) it costs $418 , 000 per ha [23] to source , transplant , and monitor the seagrass; ( 3 ) restoration has a high failure rate ( 62% ) [23]; and ( 4 ) a maximum of 0 . 1% of the existing meadow can be in a restoring state in any year . Surprisingly , despite an expected cost of over $1 million per ha for successful restoration , restoring seagrass is a better strategy to maximise seagrass coverage than marine protection or land-based actions . Larger areas of intact seagrass are achieved if the area of seagrass that is in a restoring state in any year is less conservative ( e . g . , 1% [S1 Fig] ) . Our results imply that , given sufficient funding , effort , and suitable habitat , large-scale marine restoration projects could achieve significant gains in ecosystem extent , as recently assessed [46] . Marine protection was defined as the installation of environmentally friendly moorings , which avoid seafloor damage caused by traditional moorings and minimise the effects of dragging anchor chains [48] . In other regions , where trawling or dredging are the main threats to seagrass , the implementation of Marine Protected Areas ( MPAs ) , which minimise seafloor damage by excluding destructive activities , may be a more appropriate conservation action [49] . Our study area is a Marine Park where seafloor destructive fishing techniques are forbidden and environmentally friendly moorings are the approach currently used to increase seagrass protection [50] . Our model predicts that marine protection yields the fastest initial increase and greatest total area of protected seagrass habitat ( Fig 3C ) , because it is relatively cheap compared to the other actions . Marine protection increased the overall area of seagrass through time by a small amount ( Fig 3B ) , because seagrass habitat decline rates are proportional to the amount of unprotected habitat . At $131 , 000 per ha , the 8 . 8% of seagrass habitat that is suitable for protection in the study region ( S1 Text ) can be protected in the first year so that , over decadal scales , the impact of marine protection on seagrass habitat area is limited . Land restoration was defined as using revegetation and other actions in the riparian zone to reduce erosion in riverine locations where native vegetation had been previously cleared , at a cost of $17 , 310 per ha and with a probability of success of 50% ( personal communication , J . O’Mara , SEQ Catchments ) . The resulting reduction in runoff ( Fig 3D ) increases the area of suitable marine habitat ( Fig 3A ) , as the increased water clarity improves light availability on the seafloor . Our model advances and operationalises our understanding of the impacts of sediment input on light-dependent benthic marine species by factoring in an ‘action–response curve’ [51] describing the relationship between sediment loads and illuminated seafloor area that is suitable for light-dependent species ( S1 Text ) . This relationship was generated using modelled daily sediment loads , monthly observed water clarity , and a species distribution model of seagrass habitat ( S1 Text ) [52] and is applicable in geomorphic and ecological contexts where sediment runoff impacts marine ecosystems [37] . Our results show that land restoration only offers small increases in seagrass coverage ( Fig 3B ) because there is a substantial 10-year time lag between restoration actions and the mitigation of sediment erosion and because we estimate that seagrass colonises newly available areas slowly ( 1 . 13% per year , [53] ) ( see S1 Text ) . Varying this parameter changes the results substantially , which we explore further below . Land protection was defined as purchasing privately held land containing intact native vegetation and designating it as a nature reserve , at a relatively low cost of $3 , 530 per ha ( S1 Text ) . Land protection only provides second-order benefits to marine habitat ( Fig 3D ) : It reduces terrestrial habitat decline rates , leading to relatively less erosion and less sediment within the rivers . It therefore had relatively little impact on any metric of seagrass habitat ( Fig 3A , 3B and 3C ) . Land protection therefore offers little benefit to catchments that are already highly degraded and where riparian habitat decline rates are low , such as in our case study . While the model presented here can in theory be applied to any sensitive marine ecosystem affected by both land- and ocean-based threats , it is not straightforward to source the data needed for accurate parameterisation . We therefore varied key model parameters to identify contexts where the optimal conservation strategy may differ from our results , including rates of marine ecosystem decline and expansion , rates of terrestrial ecosystem decline , and the magnitude of previous land clearing . For instance , we can find the optimal investment strategy for landscapes with extensive historic and ongoing land clearing , such as parts of Malaysia and Indonesia [54] , or high magnitudes of degradation but lower rates of ongoing land clearing , such as our study system [41] and Mediterranean countries including Albania , Algeria , and Bosnia [55] . Similarly , we can identify optimal approaches to marine ecosystems with different rates of habitat decline and expansion . For instance , kelp beds can undergo rapid declines yet can also recover rapidly when conditions are suitable [56] . In contrast , Posidonia oceanica seagrass meadows in some Mediterranean regions , such as Corsica , are declining slowly [57] but also have slow expansion rates [58] . We discovered that the relative rates of decline and expansion in the marine ecosystem , as well as the rate and magnitude of degradation on land , are key factors in our decision-making ( See Fig 4 , S2 , S3 and S4 Figs ) . When marine habitat can recover more quickly than the rate of marine habitat decline , then land-based conservation can yield optimal results . Specifically , if the ratio of marine habitat expansion to decline rates is greater than approximately 1:1 . 4 ( ratio of x- and y-axis values indicated by red dashed line in Fig 4A ) , then actions on land typically deliver the greatest cost-effectiveness ( Fig 4 , A-3 , B-3 , D-3 , but see C-3 below ) . In that case , land restoration is the best option ( Fig 4A and 4B , A-3 , B-3 ) , unless the catchment is relatively intact and the rate of land decline is high , in which case land protection is most cost-effective ( Fig 4D , D-3 ) . If the catchment is relatively intact with a low rate of loss , then marine protection is optimal ( Fig 4C , C-3 ) . Conversely , if marine habitat decline rates are greater than expansion rates , we should act in the ocean , essentially regardless of what occurs on land . Specifically , if the ratio between expansion and decline rates for the marine ecosystem is less than approximately 1:1 . 4 ( Fig 4 , A-1 , A-2 , B-1 , B-2 , C-1 , C-2 , D-1 , D-2 ) , then we should act in the ocean . If the rates of marine ecosystem decline and expansion are similar and relatively low ( less than 1% per year ) , then restoration in the ocean is the most cost-effective strategy ( Fig 4 , A-1 , B-1 , C-1 , D-1 ) . Marine protection is the most cost-effective action for our system when rates of seagrass decline outside MPAs are high ( Fig 4 , A-2 , B-2 , C-2 , D-2 ) . The findings from our analyses are factored into a generic decision-making protocol for conservation investment in marine ecosystems influenced by both land- and ocean-based threats ( Fig 5 ) . A number of conditions must be met for land-based stressors to critically impact marine ecosystems [37] . In addition to the assumptions outlined previously , the nearshore marine region must be within the impact radius of 1 or more rivers and be in an enclosed or shallow region , and land uses within the catchments must have increased the erosion of sediments or nutrients on a large scale [37] ( Fig 5 ) . If these criteria are met , then the results from our model can be used as a first step to guide decision-making , without the need for complex modelling .
The conservation dynamics of coastal ecosystems are driven by a combination of terrestrial and marine drivers . Efficient conservation investment will require decision-support tools that are repeatable , transparent , and that quantitatively describe the connections between the land- and sea-scape . The optimisation model we describe here provides a robust and extendable method to support these decisions . Our study drew an unexpected conclusion: Despite high costs and low success rates , direct restoration of marine ecosystems may be the most cost-effective method to maximise marine habitat extent over decadal timescales . If marine restoration is not likely to succeed due to poor water quality , lack of suitable substrate , or other factors , then this conclusion will clearly not hold true . Nonetheless , we propose that a paradigm shift is occurring , whereby restoration is being recognised in particular contexts as an important option for the recovery of biodiversity [24 , 59] . This is supported by recent findings that marine restoration is more likely to succeed if it is conducted on larger spatial scales [46] and if care is taken to select appropriate sites and techniques [23] . Our study highlights several factors that are essential elements in determining the most cost-effective management action but which are not often present in decision-support tools . These include the effects of time lags and the complex , nonlinear relationship between activities on the land and benefits in the sea . Land-based impacts are key drivers of seagrass extent and condition , suggesting that terrestrial protection should be a high-priority conservation action [40] . However , from a management perspective , it is essential to compare the outcomes of actions against a specific objective and timeframe and to factor in economic constraints . Our findings align broadly with those of Klein et al . [60] , who report that the cost-effectiveness of marine conservation was almost always higher than that of terrestrial conservation within any ecoregion in the Coral Triangle . Our findings differ from those of Gilby et al . [21] , whose findings support the more commonly held view that the most effective actions to benefit inshore coral reefs in Moreton Bay , Australia , would be expansion of the marine reserve network and reductions in sediment inputs from land , without considering variation in the costs of management actions or time lags . Other studies have focused on quantifying the effects of land-based impacts or protection on marine ecosystems but have not compared the results to those obtained from protection from ocean-based threats or have not quantified the effects of restoration in either marine or terrestrial realms [18 , 19 , 31] . In reality , a combination of approaches—on land and in the ocean—will be required to achieve ecological improvements in many marine regions . For instance , land-based actions would be required first if there was no suitable marine habitat available for restoration ( e . g . , Fig 5 ) . For our study system , all seagrass that is suitable for protection using environmentally friendly moorings could be achieved using the budget in the first year . Similarly , the budget for marine restoration was not completely allocated in a given year when we assume that only small areas of marine restoration can be achieved at one time . This means that marine protection or restoration could be implemented first and that the budget could be used for other strategies concurrently or in later years . In practice , budget and regulatory agencies often do not span the land–sea interface , which means land- and ocean-based actions will likely proceed independently of one another . Our model predicts the area of seagrass habitat change resulting from management actions on either side of the land–sea interface . The modelling framework provides a substantial advance in our ability to quantify the costs and benefits associated with conservation actions on land and in the ocean . In the present study , our objective was to maximise the extent of seagrass , but , there are multiple benefits from catchment restoration that are unrelated to marine ecosystems , such as enhanced freshwater biodiversity , reduced drinking water treatment costs , and increased public amenity . If we instead aim to maximise the delivery of ecosystem services provided by both seagrass and riparian habitats , which are worth $26 , 226 per ha per year and $27 , 021 per ha per year in 2007 USD , respectively , based on [61] , then the optimal decision would always be to restore the catchment ( S1 Text , S5 Fig ) . Although we used habitat area as a proxy for the delivery of ecosystem functions and services by a habitat , metrics of habitat condition ( e . g . , cover , biomass , species composition ) could be explored in the future . The ultimate aim of this approach is to identify the most cost-effective ways of maximising ecosystem functions and services ( e . g . , food or habitat for other species , water filtration , and wave attenuation ) . There are a number of uncertainties and assumptions that affect model outputs and interpretation . First , we quantify the seagrass decline rate from satellite imagery in a clear water region [62] , which is likely less impacted than nearshore areas and therefore may underrepresent the decline rate . Second , the rates at which marine organisms can colonise new areas vary among species and regions ( S2 Table ) , are scale-dependent , and the factors influencing expansion rates are not well understood . Low seagrass expansion rates may indicate bistability of seagrass and base substratum systems , where feedback mechanisms hinder the re-establishment of vegetation [63] . High seagrass expansion rates following improvement of environmental conditions can occur in some contexts [64] , typically when seagrass seed banks are present [65] . The influence of rates of seagrass decline and expansion are explored in the sensitivity analyses . Predicting the area that is suitable for restoration in marine habitats requires a habitat distribution model , the results of which contain several uncertainties , including whether all relevant environmental variables have been included in the model , whether the species is in equilibrium with environmental variables , and which method is used to select the threshold value delineating species presence and absence [52] . The effectiveness of catchment restoration is also uncertain; erosion may in fact temporarily increase following riparian restoration before eventually diminishing [66] , although our uncertainty in the parameter representing time lags in restoration is not likely to impact whether we should be acting on land or in the ocean ( S2 Fig ) . There is large variability in the costs and success rates of restoration [23] , yet we do not find these to be the most important factors affecting whether actions should occur on land or in the ocean ( S3 Fig ) . Furthermore , the costs and success rates of conservation actions will vary across the land- and sea-scape; when costs vary spatially , managers can target lower-cost areas preferentially , which reduces the average cost of management activities . Lastly , our analysis does not factor in the impact of nutrient runoff , pesticides and herbicides , climatic variability , extreme events , or climate change , the impacts of which are extremely challenging to predict [67] and therefore beyond the scope of our work but which are important areas of future research . Further uncertainty lies in the impacts of sediment runoff from the catchment on marine habitat dynamics . Previous studies linking land and the ocean have used simple distance-based relationships between sediment load and marine habitat metrics , such as ‘relative condition’ of coral reefs [e . g . , 17 , 31] . Advancing this approach , Tulloch et al . [19] used a sediment plume model that accounted for depth , bathymetry , currents , and particle size of modelled sediment runoff [28] to quantify reduction in relative coral condition due to sediment . Here , we apply a model that uses spatial empirical time series data of water clarity to estimate habitat suitability for light-dependent species; it would be interesting to compare how results vary based on the different approaches . Our approach is developed for seagrass but is applicable to other benthic marine habitats that are influenced by light availability , such as algae and coral reefs , although the link between sediment loads and the marine ecosystem would need to be modified to represent the dynamics of other ecosystems . We also tested how sensitive our model was to the functional form of the relationship between sediment loads and suitable marine habitat area by running the model with a separate linear relationship between sediment loads and suitable marine habitat area ( S1 Text ) . Surprisingly , while the amount of seagrass habitat that can be achieved by each conservation action varies depending on which relationship is used , the optimal management action does not ( S4 and S6 Figs ) . This is a relatively well-known finding in environmental decision theory , where uncertainties in the input parameters alter predictions but do not change the relative priority of management options [68 , 69] . Finally , reductions in sediment supplies , such as those resulting from the construction of dams , negatively influence marine ecosystems such as mangroves [70]; discrepancies in ecological impacts of increases versus decreases in sediment supplies to coasts is a challenge for managers . Despite structural and parametric uncertainty in the model , a quantitative optimisation framework that explicitly links conservation across the land–ocean interface provides a major conceptual advance . Specifically , it provides a quantifiable and repeatable structure for understanding the costs and benefits of taking different conservation actions and a transparent justification for acting either on the land or in the ocean . Thus , not only are we able to argue that marine conservation actions deliver the best outcomes for marine ecosystems , but this framework also offers a mechanistic explanation for why land-based management may be inferior: Multiple time lags separate terrestrial restoration projects from marine conservation outcomes . Although there are documented instances of land-based actions delivering measurable improvements in coastal water clarity and marine habitat extent [53] , these have only materialised following delays in the effect of land-based management on runoff , further delays in the improvement of coastal water clarity , and final delays in the expansion of those marine habitats . Following decades or centuries of land- and ocean-based impacts on marine ecosystems globally [1] , the challenge now is to reverse the resultant declines . Using transparent , transferable , and cost-effective approaches is critical to this process .
The study was parameterised for seagrass meadows in Moreton Bay , Queensland , Australia , and adjacent riparian areas below dams , which are considered the primary sources of sediments to the ocean in the region [42] ( Lat: -27 . 0–28 . 3; Lon: 151 . 9–153 . 4 ) ( Fig 2 ) . Moreton Bay is a shallow coastal embayment adjacent to Brisbane , the capital city of Queensland . It is home to 18 , 000 ha of seagrass comprised of 7 species , which provide grazing areas for iconic , vulnerable , and threatened species such as green sea turtles , dugongs , and migratory shorebirds . Riparian areas in the catchment have been heavily cleared since European colonisation in the mid-1800s , mainly for agriculture and urbanisation , causing ongoing increased sedimentation in riverways and marine environments [41–42 , 45] . Local direct threats to seagrass are mainly from physical damage from anchoring and mooring [71] . We extended the dynamic landscape modelling methodology of [24] to apply to both a seascape and adjacent landscape , which are connected together by sediment runoff from degraded landscapes into the ocean ( Fig 1 ) . Cleared terrestrial habitat increases sediment loads , which reduces water clarity in the adjacent ocean . The resulting decrease in light reaching the seafloor reduces the area suitable for light-dependent species . Model parameters were obtained from a variety of sources , including raster or shapefile spatial datasets ( see below and S1 Table ) , but the dynamic landscape model is not spatially explicit . See below for additional details . Each area of land at time t is classified as being in 1 of 4 states: intact and unprotected , A ( t ) ; intact and protected , P ( t ) ; degraded or cleared , C ( t ) ; or undergoing restoration , R ( t ) , with the total amount in each state described as a proportion of the landscape , and with A ( t ) + P ( t ) + C ( t ) + R ( t ) = 1 at all times ( Fig 1 ) . The landscape area is constrained to riparian habitats , because those are the major determinant of sediment input to the ocean in Queensland and the primary target of current restoration projects [42 , 72 , 73] . The seascape is split into suitable habitat ( sufficient light , soft sediments , and suitable wave energy , based on [52] ) and unsuitable ( US ) habitat . While the total area is constant , the amount of each habitat changes in each time step based on sediment loads . The area that is suitable for seagrass is divided into the same categories as on land , but denoted by the subscript S , with AS ( t ) + PS ( t ) + CS ( t ) + RS ( t ) = 1 at all times . These categories can be considered available and suitable , protected and suitable , etc . When an increase in sediment causes a decrease in suitable habitat , that decrease is taken in the appropriate proportions from each category of suitable habitat , which then becomes unsuitable . When sediment decreases allow for an increase in suitable habitat , this newly suitable habitat is added to the cleared and suitable state ( CS ( t ) ) . Transitions between habitat categories are determined by 4 rates ( degradation and revegetation on land and in the ocean ) and 6 processes ( restoration and protection on land and in the ocean , expansion in the ocean , and change in suitable habitat area in the ocean ) , which are described below . A major challenge to integrated land–sea planning is quantifying the relationship between actions undertaken on land and their effects on the marine environment . For seagrass meadows in Moreton Bay , this relationship is primarily defined by the effects of terrestrial sediment runoff on the amount of illuminated seafloor available in the ocean . We used an ‘action–response’ curve [sensu 51] describing the relationship between sediment loads and seagrass-suitable habitat area calculated in ( S1 Text ) , which in turn uses the habitat distribution model published in [52] , monthly water quality data , and monthly sediment load data ( S1 Text , S10 Data ) . This relationship predicts the area of habitat that is suitable for seagrass in each year based on the sediment loads delivered from the catchment in the previous year . This approach provides a simplification of a complex system , whereby sediment distribution and resuspension are affected by sediment composition , rainfall , and oceanographic processes , among other factors . Factoring in spatially explicit hydrodynamic modelling of sediment distribution would be an important next step to this research . Transitions in the area of suitable and unsuitable marine habitat are determined by the amount of intact land ( A + P ) in the previous time step , which affects the quantity of sediment delivered to the ocean and the area of habitat available for light-dependent marine species , like seagrass . If the area of suitable habitat is greater in t than in t−1 , then the newly suitable habitat area is added to the Cleared ( Cs ) fraction , since this habitat would not contain seagrass at the outset . If the area of suitable habitat decreases in t compared to t−1 , then habitat is removed proportionally from AS , PS , CS , and RS . Information on the definitions of the 4 conservation actions ( marine restoration , marine protection , land restoration , and land protection ) , as well as on their costs and probabilities of success , are given in the Results , S1 Text and S1 Table . Here , we provide a summary of the model parameters describing the rates , processes , and initial conditions . Further information is given in S1 Text and S1 Table . We run the model for 2 scenarios . For both scenarios , we run 4 allocation simulations , where the budget is allocated to each of the actions in isolation . Each simulation expends a budget of $50 million per year over 30 years in 0 . 1-year time intervals , for a total of $1 . 5 billion ( not accounting for inflation and discount rates ) . By comparison , it will cost $5–$10 billion over 10 years to mitigate sedimentation issues in the Great Barrier Reef catchments [84] . The time horizon of 30 years aligns with other management policies aimed at mitigating sediment issues , e . g . , the ‘Reef 2050 Long-term Sustainability Plan’ for the Great Barrier Reef [34] . All results are standardised to the outcomes achieved with no investment . For the first scenario , we invest the budget according to the parameters that we believe best describe the system . For the second scenario , we run a sensitivity analysis to examine uncertainty or variability in some of the key parameters identified in the model development process . Specifically , the model is run for rates of marine ecosystem decline and expansion varying by 0%–7% and 0%–10% per year , respectively . This approach is replicated for 4 different ecological and resource use contexts , encompassing historic land clearing extent = 20 or 80% and rate of land clearing = 0 . 75 or 7% per year . Further sensitivity analyses on the effects of time lags in restoration success , costs of actions , and the maximum area suitable for marine restoration or protection are provided in the Supporting Information and are outlined in S3 Table . At the beginning of each time step , the area of suitable marine habitat is calculated according to the area of intact ( A + P ) habitat on land . Next , changes in land and ocean habitat in each time step Δt are modelled as described by the following equations , where B is the annual budget , S is the area of marine habitat , and L is the area of land habitat . Subscripts to S and L describe the proportional areas of different habitat categories transitioning between fractions . A representative model output describing the land–sea dynamics through time is provided in S7 Fig , and additional detail is given in S1 Text . | Many coastal marine ecosystems are threatened by anthropogenic activities , but often , the best way to restore and protect these important ecosystems is unclear . Conventional wisdom suggests that the 2 most effective conservation actions to benefit coastal marine ecosystems are implementation of marine protected areas or , alternatively , reduction of land-based threats . Active marine restoration is typically considered a low-priority option , in part due to high costs and low success rates . But does this conventional wisdom hold up to closer scrutiny ? We developed a model to ask: should we restore or protect , on either the land or in the ocean , to maximise the extent of coastal marine ecosystems ? We based the model on seagrass meadows and adjacent catchments in Queensland , Australia . Surprisingly , we found that direct , active marine restoration can be the most cost-effective approach to maximising extent of marine ecosystems over longer ( decades-long ) timescales . There is , however , substantial uncertainty in our understanding of the dynamics of complex linked land–sea ecosystems . Further , geomorphological and ecological conditions vary geographically . Therefore , we also used the model to investigate how uncertainty in key parameters affects decision-making outcomes . Our results can be used to guide investment into coastal marine conservation in the absence of complex , region-specific modelling . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"ecology",
"and",
"environmental",
"sciences",
"marine",
"ecosystems",
"oceans",
"marine",
"biology",
"aquatic",
"environments",
"geology",
"habitats",
"bodies",
"of",
"water",
"marine",
"environments",
"petrology",
"ecosystems",
"marine",
"and",
"aquatic",
"sciences",
... | 2017 | Simple rules can guide whether land- or ocean-based conservation will best benefit marine ecosystems |
In order to record the stream of autobiographical information that defines our unique personal history , our brains must form durable memories from single brief exposures to the patterned stimuli that impinge on them continuously throughout life . However , little is known about the computational strategies or neural mechanisms that underlie the brain's ability to perform this type of "online" learning . Based on increasing evidence that dendrites act as both signaling and learning units in the brain , we developed an analytical model that relates online recognition memory capacity to roughly a dozen dendritic , network , pattern , and task-related parameters . We used the model to determine what dendrite size maximizes storage capacity under varying assumptions about pattern density and noise level . We show that over a several-fold range of both of these parameters , and over multiple orders-of-magnitude of memory size , capacity is maximized when dendrites contain a few hundred synapses—roughly the natural number found in memory-related areas of the brain . Thus , in comparison to entire neurons , dendrites increase storage capacity by providing a larger number of better-sized learning units . Our model provides the first normative theory that explains how dendrites increase the brain’s capacity for online learning; predicts which combinations of parameter settings we should expect to find in the brain under normal operating conditions; leads to novel interpretations of an array of existing experimental results; and provides a tool for understanding which changes associated with neurological disorders , aging , or stress are most likely to produce memory deficits—knowledge that could eventually help in the design of improved clinical treatments for memory loss .
To function well in a complex world , our brains must somehow stream our everyday experiences into memory as they occur in real time . An “online” memory of this kind , once termed a “Palimpsest” [1] , must be capable of forming durable memory traces from a single brief exposure to each incoming pattern , while preserving previously stored memories as long and faithfully as possible ( Fig 1 ) . This combined need for rapid imprinting and large capacity requires that the memory system carefully manage both its learning and forgetting processes , but we currently know little about how these processes are implemented and coordinated in the brain . A number of quantitative models have been proposed for palimpsest-style online memories , and have addressed a variety of different issues , including: how memory capacity scales with network size , how metaplastic learning rules can increase memory capacity , and the tradeoff between initial trace strength and memory lifetimes [1–8] . A few studies with a more empirical focus have addressed the biological mechanisms underlying recency vs . familiarity memory [9]; the coordination of online learning with long-term memory processes; and the details of memory-related neuronal response properties during online learning tasks [10–12] . Nearly all previous models of online learning have assumed that the neurons involved in memory storage are classical "point neurons” , that is , simple integrative units lacking any representation of a cell’s dendritic tree . This simplification is notable , given the now substantial evidence from both modeling and experimental studies that dendritic trees are powerful , functionally compartmentalized information processors that can augment the computing capabilities of individual neurons in numerous ways [7 , 13–59] . Beyond their contributions to the computing functions of neurons , it is also increasingly apparent that dendrites help to organize and spatially compartmentalize synaptic plasticity processes [7 , 40 , 60–86] . Thus , given that dendrites can act as both signaling and learning units within a neuron , it is important to understand how having dendrites could affect the brain’s online learning and memory processes . In this paper , we focus on the role that dendrites may play in familiarity-based recognition , a function most closely associated with the perirhinal cortex [87 , 88] . Here , we introduce a mathematical model that allows us to calculate online storage capacity from the underlying parameter values of a previously proposed dendrite-based memory circuit [7] . The model includes biophysical parameters ( dendritic learning and firing thresholds , network recognition threshold ) , wiring-related parameters ( number of axons , number of dendrites , number of synapses per dendrite ) , and input pattern statistics ( pattern density , noise level ) ( see Table 1 ) . As an example of the model’s use , we study the interactions between memory capacity , dendrite size , and pattern statistics , and cross-check the results using full network simulations . We found that dendrites containing a few hundred synapses ( as opposed to a few tens or a few thousand ) maximize storage capacity , providing the first normative theory that accounts for the actual sizes of dendrites found in online memory areas of the brain .
The network structure and plasticity rules have been previously described in [7] , but are repeated here for clarity . A population of neurons with a total of M separately thresholded dendrites receives inputs from NA input axons ( Fig 2b ) . Each dendrite receives K synaptic contacts randomly sampled from the NA axons , for a total number of synapses NS=M∙K . The connectivity matrix is assumed to be fixed . Input patterns are binary-valued vectors x = {x1 , … , xNA} for which component xi is 1 if the ith axon is “firing” and 0 otherwise . We quantify density/sparsity of the patterns by the fraction of axons fA firing in each pattern; the value of fA ranged from 0 . 008 to 0 . 18 in this study , as we found empirically in previous work that sparse patterns maximize capacity in this type of memory [7] . To model a biologically realistic form of input variability , we assumed that each active axon ( xi=1 ) produces a burst of spikes , where the number of spikes in the burst is drawn from a binomial distribution with mean μburst=Nburst·Pburst=4 spikes/burst . Pburst ranged from 1 ( no noise ) to 0 . 4 ( high noise ) , with Nburst varying inversely . Inactive axons ( xi=0 ) were assumed to produce no spikes . We denote the noisy spike count version of an input component x~i~xi∙Binom ( Nburst , Pburst ) . Synapses are characterized by both a weight wij , where the subscript indicates a connection between axon i and dendrite j , and an additional scalar parameter αij , representing the synapse’s “age” . The weight of each synapse is binary-valued , and can change between weak ( w = 0 ) and strong ( w = 1 ) states when the dendrite containing the synapse undergoes a learning event; the conditions that trigger a learning event are discussed below . The age variable at each synapse tracks the number of learning events that have occurred in the parent dendrite since the synapse last participated in learning . Two different measures of a dendrite’s activation level determine how the dendrite responds to an input , and whether it undergoes a learning event . The “presynaptic” activation measure is based on the activity levels of the set of axons Dj that make contact with the jth dendrite apre ( j ) =∑iϵDjx˜i . In words , apre ( j ) is the total number of presynaptic spikes arriving at all the synapses impinging on the jth dendrite , regardless of their postsynaptic weights , and is thus a measure of the maximum response the dendrite could muster to that input pattern assuming all of the activated synapses were strong ( w=1 ) . The more conventional “postsynaptic” activation level takes account of the synaptic weights in the usual way: apost ( j ) =∑iϵDjwij·x˜i . When the postsynaptic activation level exceeds the “firing” threshold θF , the dendrite is said to fire , that is , generates a response rj = 1 . The responses of all dendrites within a neuron sum linearly to produce the neuron’s response ( Fig 2b ) , and the responses of all neurons in the network sum linearly to produce the overall network response r . The overall response of the network can therefore be written directly as a sum over all the M dendritic responses: r=∑jϵ[1 , M]rj so that the network can be viewed as a single “super neuron” with M dendrites . Finally , an input pattern is classified as “familiar” if r≥θR , and “novel” if r<θR , where θR is the recognition threshold ( Fig 2b ) . The goal of learning is to ensure that learned patterns going back as far as possible in time produce suprathreshold network responses ( r≥θR ) , while randomly drawn patterns do not . Learning of any given pattern occurs in only the small fraction of dendrites that cross both the presynaptic and postsynaptic learning thresholds ( apre ( j ) >θLpre and apost ( j ) >θLpost ) . When this occurs , a “learning event” is triggered in the dendrite , and all active synapses belonging to that dendrite “learn” , as follows . If an active synapse is currently in the weak state , it is “potentiated” ( i . e . both strengthened and “juvenated”: wij→1 , αij→0 ) , or if it is already in the strong state , then it remains strong but is juvenated ( wij=1 , αij→0 ) . All strong synapses in the dendrite that are not active during the learning event remain strong but grow older ( wij=1 , αij→αij+1 ) . Thus αijcounts the number of learning events that have occurred in the dendrite since the synapse last learned , and thus represents the age of the most recent information that that synapse is involved in storing . Note that a synapse’s age variable counts learning events within its parent dendrite only , and any given dendrite learns only rarely , so the counter need have only a small number of distinct values , on the order of ~12 under the simulation conditions explored in this paper . To maintain a constant fraction of strong synapses ( we used fs=0 . 5 ) , and thereby to prevent saturation of the memory , in each dendrite undergoing learning , a number of strong synapses are depressed ( wij→0 ) equal to the number of weak synapses potentiated during that learning event . A key feature of the learning rule is that the synapses targeted for depression are those that learned least recently ( i . e . having the largest values of αij ) , so that the information erased during depression is the “oldest” stored information . This “age-ordered depression” strategy substantially increases online storage capacity [5] , especially in a 2-layer dendrite-based memory where the very sparse use of synapses during pattern storage gives each strong synapse , and the information it represents , the opportunity to grow old [7] . One of the key quantities involved in calculating storage capacity is L , the length of the age queue within a dendrite ( see Fig 3 ) . An approximate expression for L is given here; the derivation can be found in the Methods . L is a measure of the time a pattern feature persists in a dendrite , and given that age queues progress at roughly equal rates in all the dendrites involved in storing a pattern , it also effectively measures a pattern’s lifetime in memory–counted in units of dendritic learning events . L can be understood intuitively through an oversimplified example: If 10 synapses are strengthened on a dendrite during a learning event , and there are 120 strong synapses on the dendrite , then L would be ~12 . That is , after ~12 learning events have elapsed since a pattern was first stored , the 10 synapses involved in storing the pattern are now the oldest on the dendrite and must be depressed , and the memory is lost . The actual expression for L is more complex as it takes into account the fact that strong synapses do not inexorably progress to the ends of their age queues–they can be rejuvenated one or more times during the course of their lifetimes , in which case the same strong synapse participates in the representation of more than one pattern . To convert from L to a number of training patterns , we must multiply L by the approximate number of patterns per dendritic learning event , or “learning interval” 1PL , where PL is the probability that an arbitrary dendrite learns a particular pattern . This gives an expression for capacity: C≈LPL=1PL[log ( 1−fS ) log ( 1−θLpreK⋅μburst ) −1] ( 2 ) Although PL is conceptually simple , its expression is complicated since it depends on pattern density , noise level , two learning thresholds , dendrite size , and fS , and so it is omitted here for clarity ( see in the Methods section for the full expression and some discussion ) . The expression for C measures how long patterns persist in memory , but a different calculation is needed in order to predict the memory’s recognition performance , that is , the false positive and false negative error rates ϵ+ and ϵ- that we can expect to obtain during a pattern’s lifetime . These error rates depend on the separation of the distributions of responses to trained vs . untrained patterns ( Fig 1 ) . These two distributions can be computed from the network parameters to determine whether the allowable error rate tolerances θ+ and θ- will be met during the lifetime calculated in Eq 2 ( see Methods ) . How can the expression for online storage capacity ( Eq 2 ) be exploited ? Given that one of the unique features of our model is that dendrites are the learning units , we used the model to determine how capacity varies with dendrite size , which in turn allows us to determine the optimal dendrite size . In particular , we asked: for a fixed total number of synapses in the memory network ( NS=M∙K ) , if the goal is to maximize online storage capacity , is it better to have many short dendrites ( i . e . large M , small K ) , a few long dendrites ( small M , large K ) , or something in between ? Furthermore , how does the optimal dendrite size vary with properties of the input patterns , such as pattern density and input noise level ? To address these questions , we fixed network parameters Nsand fs and then for varying combinations of the pattern-related parameters ( fA , Nburst , Pburst ) , we computed C as a function of dendrite size K , using values of the learning , firing , and recognition thresholds ( θLpost , θLpre , θF , θR ) optimized for each value of K through a semi-automated grid search . The “optimal” dendrite size under a particular set of input conditions was the value of K that maximized capacity , subject to the constraint that immediately after training , responses to trained patterns were strong enough , and responses to random patterns were weak enough , that both the false positive ( ϵ+ ) and false negative ( ϵ- ) error rates fell below specified tolerances ( we used 1% for both ) . Note that though K appears explicitly only once in Eq 2 , as a result of the capacity optimization process , all of the thresholds , and consequently θLpre and PL in Eq 2 depend implicitly on K . The net effect of these dependencies is analyzed in detail in the sections below on penalties for long and short dendrites . Capacity is plotted in Fig 4a as a function of K for pattern density values ranging from 0 . 8% to 18% . In the case with fA=1 . 5% , capacity peaked at ~30 , 000 patterns when dendrites each contained 256 synapses , and declined substantially for both short ( K<100 ) and long ( K>1000 ) dendrites . As the pattern density increased ( to 18% ) or decreased ( to 0 . 8% ) , peak capacity varied nearly 5-fold , favoring sparser patterns , but over the more than 20-fold range of pattern densities tested , peak capacity always occurred for dendrites ranging from 100–500 synapses ( grey shaded area ) . Focusing on the high-capacity ( sparse ) end of the range with fA<3% , peak capacity was confined to the narrower range of 200–500 ( i . e . “a few hundred” ) synapses . We also observed that sparser patterns led to a preference for longer dendrites , an effect we unpack below using full network simulations . It is important to clarify that the higher recognition capacity seen for sparser patterns does not result from the fact that sparser patterns contain less information , thereby reducing storage costs per pattern ( see S1 Text ) . We also note that in the more realistic conditions modeled in the full network simulations ( see below and Fig 5 ) , peak capacity saturates at slightly higher pattern activation densities ( around 1 . 5% ) than is predicted by the analytical model , and the optimal pattern density may be higher still under conditions of increased background noise ( S1 Fig shows strong susceptibility to background noise even at 3% pattern density ) . To test the effect of pattern noise on capacity , we varied the input noise level by choosing combinations of Nburst and Pburst whose product was always μburst=4 spikes , but that yielded narrow or broad spike count distributions for each active pattern component ( Fig 4b , see histogram insets ) . In this way , we varied the degree to which a trained pattern resembled itself upon repeated presentations . The variation in event counts arising from the above scheme could be viewed as representing either variation in the number of action potentials arriving at the presynaptic terminal from trial to trial , or variation in the number of synaptic release events caused by a given number of action potentials , or a combination of both effects . As expected , higher noise levels reduced peak capacity ( Fig 4b ) , except in the long dendrite range ( K>1000 ) where central limit effects rendered dendrites insensitive to this type of noise . In keeping with this effect , optimal dendrite size increased slightly as the noise level increased , but again , peak capacity was consistently seen for dendrites in the “few hundred” synapse range . Even higher levels of noise were not considered because a simple , biologically available saturation strategy that maps multiple release events into a relatively constant post-synaptic response can largely mitigate the effects of this type of noise . ( We did not include a multi-input saturation mechanism in our model to avoid the added complexity ) . To verify that the preference for dendrites in the few hundred synapse range was not an artifact of “small” network size , we generated capacity curves from Eq 2 for networks scaled up 256-fold from a base size of N = 5 . 12 million synapses to ~1 . 3 billion synapses . The results are shown on a log plot in Fig 4c . As shown in Fig 4d , the scaling power for dendrite sizes K = 64 , 256 , and 1024 were , respectively , 0 . 98 , 0 . 97 , and 0 . 97 , confirming earlier observations that storage capacity in an optimized dendrite-based memory grows essentially linearly with network size [7] . All the while , the preference for dendrites containing a few hundred synapses remained essentially invariant . To cross-check the results of the analytical model , we simulated a full memory network , and measured capacity empirically as a function of K . Unlike the analytical case , in which capacity was assumed to be proportional to the calculated length of dendritic age queues , in the network simulations we performed explicit old-new recognition memory tests , and optimized system parameters to achieve false positive and false negative error rates of 1% . In the interests of greater biological realism , we replaced the hard dendritic firing threshold and binary input-output function with a continuous sigmoidal input-output function given by 11+e-sx-θF , and optimized over the slope parameter s along with the 4 threshold parameters . In addition , we relaxed the strict assumption of the analytical model that every input to the network was statistically independent of every other , and instead arranged for each input axon to form ρ synaptic contacts within the memory area , rather than just one . This “redundancy” factor , ρ , set by default to 200 , introduced some degree of correlation in the input patterns , and lowered peak capacity somewhat , but had no effect on our main conclusions . Fig 5a depicts one such simulation with 5 . 12 million synapses . In the top panel , blue dots show responses to trained patterns , red dots show responses to randomly drawn ( untrained ) patterns that establish the baseline trace strength ( green dashed line ) above which stored pattern traces must rise to be recognized . Consistent with the analytical model , responses to trained patterns remain essentially constant during an extended post-training period , in this example spanning ~10 , 000 patterns . After the flat post-training phase , in contrast to the relatively abrupt fall in trace strength envisioned by the analytical model , a more gradual decline is seen , reflecting the variable times at which the synapses encoding each pattern reach the end of their age queues in different dendrites . Note that the false negative error rate begins to climb during this trace decay period , as the lower fringe of the trained response distribution ( blue ) progressively merges with the untrained background distribution ( red ) . In this simulation , capacity was reached at ~21 , 000 patterns , which by our specification is the point where both false positive and false negative error rates equaled 1% . Mirroring the approach taken with the analytical model , multiple simulations were run with varying firing , learning , and recognition thresholds to find the combination of parameters that maximized capacity for each value of K , subject to the same error rate constraints as before . As an additional check of the analytical model , we histogrammed synapse ages within a dendrite ( for many dendrites ) ( Fig 5b ) , and found that they conformed to a geometric distribution as predicted ( red line shows a fitted exponential decay ) , up to the “cliff” at the end of the age queue ( blue dashed line ) . Capacity was measured for dendrite sizes between 32 and 4 , 096 synapses , and the results are shown in Fig 5c and 5d , which are the analogues of Fig 4a and 4b , respectively . When compared to the curves produced by the analytical model , the capacity curves produced by full network simulations had similarly placed capacity peaks and similar qualitative dependence on pattern density and noise levels . In one minor difference , we noted that under the more realistic conditions modeled in the full network simulations , peak capacity saturated at slightly higher pattern activation densities ( around 1 . 5% ) than was predicted by the analytical model ( Fig 4a ) . To determine whether the predictions regarding optimal dendrite size would survive under even more challenging “real world” operating conditions , we added increasing amounts of background noise ( spurious spikes added to nominally inactive pattern components ) , on top of the pre-existing burst noise and pattern correlations . As in the case of burst noise , the background noise level varied between 2 extremes: zero noise , which maximized capacity , and a “high noise” level that reduced storage capacity by roughly a factor of 2 compared to the no-noise case . As in the case of burst noise , we did not consider very high noise levels on the grounds that the deleterious effects of background noise can be compensated by a relatively simple mechanism , for which there is evidence: pre-synaptic terminals with low release probability for “singleton” spikes , along with paired pulse facilitation [89] , would allow the effects of sporadic background spikes to be suppressed while maintaining strong responses to signal-carrying bursts . Even at background noise levels capable of causing a significant reduction in peak capacity , the effect of background noise on optimal dendrite size was negligible ( S1 Fig ) . Only at very high levels of background noise , where capacity was reduced more than twofold , did optimal dendrite size change significantly , moving outside of the of the “few hundred” synapses per dendrite range ( S1 Fig ) . Next we examined the effect of increasing correlations in the input patterns . Given that a single axon can in fact form many thousands of synaptic contacts , corresponding to a much higher redundancy factor than we used in our base simulation , we ran simulations using redundancy factors ρ=5 , 000 and ρ=10 , 000 ( Fig 5f ) , which meant that groups of 5 , 000 or 10 , 000 synapses scattered across the memory were activated identically . Given previous reports that input correlations can be very deleterious to capacity [10] , we speculated that these drastic reductions in the effective dimensionality of the input patterns would severely challenge a memory architecture that was designed to perform optimally with random inputs , or at least significantly alter its behavior . As shown in Fig 5f , however , even in the high-redundancy case ( with a 10 , 000-fold reduction in input space dimensionality ) , peak capacity dropped by only a factor of ~2 compared to the case with ρ=200 , with little to no change in optimal dendrite size . We next took advantage of the full network simulations to probe the mechanisms that lead to the capacity costs associated with both short and long dendrites . Fig 5e shows two important quantities: the average number of dendrites ( μLD ) and synapses ( μLS ) used to store a single pattern in the simulations from Fig 5c . The significance of these quantities is discussed below as we work through the distinct capacity penalties for long and short dendrites . As shown in Fig 5e , as dendrites grow longer , dendrite usage per stored pattern drops from a value around 10 ( at peak capacity ) to a “floor” of roughly ~7 dendrites at the long-dendrite end of the range , whereas synapse usage climbs steeply from a baseline of around 150 synapses . To understand the source of the lower bound of ~7 on the average number of dendrites used to store each pattern , it is useful to consider the situation that holds when , in the interests of resource efficiency , we attempt to store each pattern with the minimum possible trace strength: one dendrite . One dendrite firing in response to a familiar pattern is in principle sufficient for recognition , if it is reliable ( i . e . occurs > 99% of the time ) , and if the network’s response to untrained patterns is reliably zero ( i . e . > 99% of the time ) . In a large network , given that each dendrite participates in learning with equal ( small ) probability , the distribution of the number of dendrites that undergoes a learning event is approximately Poisson with mean μLD=PL·M . Given that a Poisson distribution is characterized fully by its mean , setting μLD=1 by adjusting the learning thresholds , which control PL , means that one dendrite will undergo a learning event for each presented pattern–on average–which is the goal . However , with a mean of 1 , the probability that zero dendrites learn is surprisingly high: ~37% ( Fig 6a , top plot ) . Thus , in aiming to use a single dendrite to encode a pattern on average , more than a third of all patterns presented to the network would produce no memory trace at all , leading to a false negative error rate far above the 1% acceptable threshold . To avoid this pitfall , it is critical to reduce the probability to below 1% that zero dendrites learn , which according to the Poisson distribution requires a mean μLD=5 dendrites . This requires a remarkable 5-fold increase in PL relative to the theoretical minimum , with a corresponding 5x increase in synapse resource consumption ( Fig 6a , middle plot ) . Worse , given increased variability in the number of learning dendrites as well as increased readout failures due to input noise and correlations , storage capacity turns out to be maximized when an even higher value of PLis used , achieved by further loosening the learning thresholds , which for our combination of system parameters leads to the empirically obtained optimal value of μLD=~7 dendrites at the long-dendrite end of the range . Given this floor of ~7 dendrites , it becomes clear why synapse usage increases as dendrites grow longer: the number of synapses used in a dendrite that undergoes a learning event is roughly proportional to the dendrite length K , since the number of synapses that learn is roughly proportional to the number of synapses activated in the dendrite , which is proportional to dendrite size . Tied to this increase in synapse usage per pattern , as the total number of dendrites M in the system decreases ( because each one contains a larger fraction of the synapses ) , the frequency with which each dendrite must participate in learning increases , which speeds the per-pattern rate at which synapses move along their age queues . Thus , from a capacity standpoint , it is ideal to choose system parameters such that the minimum encoding bound of 7 dendrites is actually used ( or whatever minimum number of dendrites is needed , given the settings of the error rate thresholds and noise level ) , but having met this lower bound , dendrites should be kept as short as possible . The reasons capacity declines as dendrites grow shorter are complex , and are discussed only briefly here ( see the S1 Text and S3 and S4 Figs for more details ) . We first consider why dendrite usage increases for short dendrites , rather than remaining at the minimal encoding bound . Short dendrites are intrinsically more susceptible to variability in crossing their learning and firing thresholds , since fewer active synapses are involved . As dendrites become very short , this requires the network to increase dendrite usage far above the nominal lower bound of μLD=5 . For example , under sparse activation ( fA=1% ) , medium noise conditions ( Pburst=47 , Nburst=7 ) with dendrites containing ~200 synapses , when the system is optimized for capacity , μLD≈15 ( blue solid curve in Fig 5e ) , substantially more than the number of dendrites used under maximum capacity conditions . While this increase in dendrite usage is more than offset by the reduced dendrite size , which tends to reduce synapse usage , the total number of synapses altered during learning in fact remains approximately constant , implying that a larger fraction of synapses is modified within each short dendrite that engages in learning . This higher synapse burn rate in short dendrites leads to shorter age queues , and in the end lowers capacity .
Why are dendrites of “medium” size optimal for storage capacity in the context of an online familiarity-based recognition memory ? The simplest explanation is that short dendrites suffer from one set of disadvantages , and long dendrites suffer from another , leaving the optimal dendrite size somewhere in the middle . Short dendrites have relatively noisier post-synaptic response distributions because fewer synapses contribute to the response . As a result , a larger fraction of the synapses on a short dendrite must be modified during learning to ensure that the dendrite's response to previously trained patterns remains comfortably at the upper tail of the untrained pattern response distribution . Increasing the fraction of synapses used within a dendrite during each learning event shortens the dendrite's age queue , which comes at a capacity cost . This effect leads to a preference for longer dendrites . But long dendrites also have their disadvantages . An online recognition memory should aim to store the weakest possible trace of each learned pattern , which in our framework corresponds to learning in a small number of dendrites near the "minimum encoding bound" ( corresponding to ~7 dendrites under the conditions used in our study; see Fig 5e ) . This means that the longer the dendrites become , the more synaptic resources are consumed by each dendrite that learns , since the number of synapses used per dendrite during a learning event is roughly proportional to dendrite size . Clearly from this perspective , it's best to keep dendrites as short as possible . The compromise between the need to keep dendrites long enough to avoid noise and age queue problems , and short enough to avoid excessive synapse use per learning dendrite , puts the optimal size around a few hundred synapses for biologically reasonable values of pattern activation density and noise . Of course , our assumptions regarding "biologically reasonable" pattern activation densities and noise levels are informed guesses rather than certain knowledge , and are not likely to be universal constants across brain areas , species and operating conditions . It is therefore possible that the natural dendrite sizes found in medial temporal lobe memory areas are determined in part by factors other than capacity optimization according to Eq 2 . For example , developmental constraints , energy constraints , space constraints , and combinations thereof , may have been responsible for pushing the actual dendrite size in one direction or another , away from the optimal length as determined by capacity considerations alone . Nonetheless , it is useful to capture basic relationships between biophysical parameters , wiring parameters , input pattern statistics , and capacity , as a starting point for a more complete online memory model . That mid-sized dendrites optimize capacity can be understood from another perspective . Eq 2 shows capacity is given by the ratio of L , the length of a dendrite's age queue , to PL , the probability that a dendrite learns . PL , in the denominator , grows larger as dendrites grow in size because the same average number of dendrites is always used to learn , but when dendrites are long , there are fewer of them to choose from . L , in the numerator , grows smaller as dendrites shrink in size because of the higher value of fpot needed to compensate for noise effects . Balancing these two effects , capacity is maximized for dendrites of intermediate size , for which L is not too small , and PL is not too large . Thus , among the many roles that dendrites may play in the brain , in the context of an online familiarity-based recognition memory , separately thresholded dendrites play the critical role that they downsize the learning units from neuron-sized units ( ~20 , 000 synapses ) to units containing a few hundred synapses , which are much more numerous , while still containing enough synapses to avoid the capacity costs associated with noise effects and shortened age queues . Simply put , having separately thresholded dendrites provides the memory system with more learning units of a better size . If dendrite-sized learning units were not available , so that it was necessary to construct an online recognition memory from neuron-sized units , storage capacity would be cut by an order of magnitude or more ( see Fig 5c ) . A general theme that emerges from our study is the importance of variability control for a recognition memory . The goal of a neural-style online recognition memory is to store a trace of each learned pattern that consumes as few synaptic resources as possible , but that nonetheless allows the network to produce a reliable recognition response on future encounters with a stored pattern . Variability in the magnitude of network responses to either learned or unlearned patterns , such as that produced by burst noise , or low pattern density , complicates this goal in at least two ways . First , increased variability in the responses to unlearned patterns raises the level of background noise , and thus the required minimum encoded signal strength that learned patterns must obtain . This in turn increases the number of synapses that must be devoted to storing each new memory . Second , increased variability in signal strength for learned patterns increases the rate of readout failures ( for fixed firing and recognition thresholds ) . This increase in false negative errors must again be compensated for by increasing memory trace strength for all patterns , which wastefully strengthens patterns whose traces were already well above the recognition threshold . These effects imply that a brain system devoted to recognition memory is under intense pressure to include response normalization mechanisms , presumably involving local inhibitory circuits [91–99] . It is intriguing to note that if network behavior could be perfectly normalized , so that every pattern is stored by learning in the exact same number of dendrites , e . g . 1 dendrite , then this would represent a 7-fold resource savings , presumably leading to a corresponding boost in capacity compared to the peak capacity conditions shown in Fig 5 ( where an optimized high capacity network chooses to learn using 7 dendrites ) . Several of the mechanisms and processes in our dendrite-based learning scheme are consistent with known biological mechanisms , including that: The main speculative/predictive features of our model pertain to the specific conditions for LTP and LTD . First , following [7] we assumed here that the triggering of a learning event in a dendrite , which induces both LTP and LTD , depends on a compound threshold: in order to learn , a dendrite must both ( 1 ) receive an unusually strong presynaptic input , that is , unusually many axons impinging on the dendrite must be firing and releasing glutamate; and ( 2 ) experience an unusually strong post-synaptic response , that is , unusually many of the firing axons must be driving synapses that are already in a strong state . Note that a traditional Hebbian learning rule would tie learning to the post-synaptic response alone ( ∑wixi ) , placing no explicit condition on the number of axons participating ( ∑xi ) . The pre-synaptic condition was incorporated into our model opportunistically , when we observed that doing so doubled the memory's storage capacity [7] . We call the existence of a compound learning threshold a "prediction" of our model on the grounds that the brain would have been under evolutionary pressure to discover any small functional modifications that significantly boost storage capacity , and so the brain might have “discovered” this optimization–as we did . The prediction is weak , however , given that the memory can function in basically the same fashion with a single , conventional post-synaptic threshold , albeit with reduced capacity . Unlike our weak prediction of a compound dendritic learning threshold , which could be falsified without dire consequences for the model , the prediction that synapses involved in an online familiarity memory should have a prescribed lifetime in the potentiated state , after which they are actively depotentiated , is a more deeply rooted feature of our model . This prediction is also a nearly inevitable consequence of the statement of the learning problem itself: any online recognition memory whose memory retention is much shorter than the animal's lifetime will be "full" at all times , except for a transient period at the beginning of the animal's life when the memory is first filling up . Once it reaches its chronically full state , each time a new pattern is written into the memory by strengthening synapses , as a matter of homeostatic necessity the equivalent of one stored pattern must be erased by weakening synapses , and in the interests of optimal performance , that one erased pattern should be the oldest stored pattern . The alternative–partially degrading many patterns of varying ages–is a poor strategy for a recognition memory , since any pattern whose signal strength is prematurely degraded to the point where it falls below the recognition threshold is functionally lost , yet its unerased detritus continues to uselessly consume space in the memory . Furthermore , since it is most efficient from a resource allocation point of view to store memory traces that are just strong enough to cross the recognition threshold , and no stronger , the system cannot abide gradual attrition of pattern traces . Thus the problem statement itself , and simple logic , dictate that a memory network in the brain devoted to online familiarity/recognition memory should attempt to target the oldest information for erasure as each new pattern is stored . It is difficult to imagine how selective erasure of old information could occur unless synapses keep track of their ages , and unless a dendrite is able to target its oldest synapses for depression as it undergoes each new learning event . Age-based depression of synapses was previously explored as a strategy for increasing online learning capacity in the context of a 1-layer Willshaw network [5] . It is only in the context of a 2-layer memory , however , in which synaptic learning probabilities can be driven down to extremely low values without compromising signal strength , that synapses are given the opportunity to actually grow old [7] . In the 2-layer dendrite-based memory scheme we have studied , storage capacity is increased ( ~linearly ) by increasing the number of dendrites , without altering the synapse model or the plasticity rule . As an alternative , Stefano Fusi and colleagues have developed two elegant models of online learning that boost capacity instead by increasing the complexity of individual synapses [4 , 8] . Both models share the following basic framework: the memory consists of N synapses abstracted away from any particular network architecture; by default , every synapse is modified during the storage of every pattern; to store a pattern , synapses are strengthened and weakened in equal numbers; and all instructed weight changes during pattern storage overwrite previously stored information . The goal of these models is to carefully manage the plasticity-stability tradeoff that exists when each synapse is asked to encode information about many patterns that have been stored over time: synapses that are very plastic are good at rapidly storing new information but poor at preserving old information , whereas synapses that are very stable are good at preserving old information but poor at rapidly storing new information ( synopsis adapted from [8] ) . In the "Cascade" model [4] , synaptic weights are binary valued ( strong and weak ) , but can exist in states of varying lability/stability . The state diagram within each synapse operates according to two main principles . First , repeated potentiation instructions push a strong synapse into an increasingly stable strong state , that is , a state that shows an increasing resistance to depression . Similarly , repeated depression instructions by the learning rule have the effect of pushing a weak synapse into an ever more stable weak state , one that increasingly resists potentiation . Second , at "deeper" levels of the cascade , corresponding to more stable strong and weak states , the transitions to even deeper levels corresponding to even more stable states , and the transitions in synaptic weight value from strong to weak or weak to strong , all become increasingly improbable , so that synapses in deeper cascade states remain stable over longer and longer time scales . The variation in these transition probabilities across cascade levels can be considerable: according to [4] the optimal cascade model with 10^6 synapses has 15 cascade levels . With this many levels , the most labile synapses at the top of the cascade change weight with probability 1 ( i . e . deterministically ) in response to a weight change instruction , whereas the most stable synapses deep in the cascade only change weight with probability 1/16 , 384 in response to a weight-change instruction . Thus , a weak synapse in its most stable state would need to receive ~10 , 000 potentiation instructions in a row in order to reach a 50% chance of actually undergoing potentiation . These two operating principles of the Cascade model are clearly distinguishable from those governing synaptic plasticity in our model . First , in the Cascade model , all synaptic state transitions are probabilistic , whereas in the dendrite-based model , all synaptic state changes are deterministic: during learning , weak synapses receiving the instruction to potentiate do so fully and immediately , and during forgetting , strong synapses that reach the end of their lifetimes are fully and immediately depressed . The logic of synapse durability is also different in the Cascade vs . dendrite-based models . In the Cascade model , when a synapse is first potentiated , it is in its most labile strong state , and therefore most vulnerable to depression . In the dendrite-based model , a synapse that has just been potentiated is in its most durable state , in the sense that it will withstand the largest number of consecutive learning events in which it does not participate before it "ages out" and finally succumbs to synaptic depression . In the Benna and Fusi model [8] , the machinery contained within each synapse consists ( metaphorically ) of a chain of connected fluid-filled beakers . The first beaker represents the synapse’s ( graded ) strength value by the level of virtual liquid relative to equilibrium , and the last beaker is tied to the equilibrium liquid level . Synaptic potentiation occurs deterministically , and consists of adding a fixed amount of liquid "weight" to the first beaker; synaptic depression consists of removing that amount of liquid from the first beaker . The equilibration of liquid levels in the beaker chain following an instructed weight change , and particularly the equilibration of the first beaker , captures the time course of the memory decay at each synapse . In the example shown in [8] , a synapse consisted of a chain of 12 virtual beakers that doubled in capacity with each step down the chain ( so that the last beaker had a capacity 2 , 048 times that of the first beaker ) , and whose fluid levels were governed by differential equations with pre-determined rate constants linking each pair of buckets . As a practical matter , the authors found the number of discrete levels per beaker could be reduced linearly from 35 in the first ( smallest ) beaker , corresponding to 35 levels of visible synaptic weight , down to 2 levels in the last ( largest ) beaker . This parameterization yielded a total of ~10^14 possible memory states within each synapse . Interestingly , unlike the cascade model whose synapses only change state in response to plasticity instructions ( which can occur asynchronously ) , the chain-of-beakers model , if taken literally , continues to equilibrate—i . e . forget—even during periods when the rate of new learning slows or stops , such as during quiet wakefulness or sleep . Thus , an additional layer of mechanism is presumably needed to modulate the inter-beaker flow rates in a coordinated fashion depending on the external learning rate . In summary , both of these models [4 , 8] achieve longer memory lifetimes by increasing the complexity of the synapse model as the size of the memory increases . In terms of cost , the machinery inside these more complex synapses requires more parameters ( >10 ) , and those parameters must span large dynamic ranges ( >1000 ) to reach realistic memory sizes . How does a dendrite-based model grow storage capacity without increasing the complexity of the individual synapses ? Within virtually any recognition memory model , the conceptually simplest way to increase storage capacity is to reduce the fraction of synapses that are modified during the storage of each pattern ( the signal ) , while correspondingly reducing the response of the memory to random input patterns ( the noise ) . Practically , this can be achieved by sparsifying the input patterns inversely with pattern size as the memory grows larger . Thus , if the memory increases in size from N to c·N synapses , in order to increase capacity c-fold , the pattern density 'a' must be reduced c-fold so that the same number of synapses is activated by each pattern as before . Assuming the learning rule instructs each activated synapse to become strong if it was weak , a·N/2 weak synapses would be potentiated on average ( under the assumption that half of the synapses are strong ) , and an equal number of strong synapses would be depressed to maintain homeostasis ( drawn from the N/2 strong synapses ) . To a rough approximation , this leads to a capacity of ~1/a patterns . Thus , if a = 1% of synapses are changed during the storage of each pattern , then after ~100 patterns are stored , the memory will have turned over completely . This simple scaling approach runs into the biological plausibility problem that very large capacities require very low pattern densities , and very low depression probabilities . To achieve a capacity of 100 , 000 patterns , for example , only 1 in 100 , 000 input neurons could be active , and synaptic depression would occur in only 1 in 100 , 000 strong synapses . Reliably controlling such small activity and plasticity probabilities could be difficult to achieve in neural tissue . As an alternative both to this very simple sparsification approach , and to the "complex synapse" approach developed by Fusi and colleagues , adding a layer of dendritic learning units allows the memory to push further into the sparse plasticity regime without the need for very low pattern densities or plasticity probabilities . Relative to a flat ( 1-layer ) memory model , dendritic learning thresholds can restrict learning to just a few dendrites from a very large pool . For example , in a simulation of a 5 million-synapse network discussed previously , with a moderate pattern sparseness level of a = 3% , the dendrite learning probability after optimization was PL = 0 . 0005 , ( corresponding to 1–2% of neurons in the network having one dendrite that crosses the learning threshold ) . Beyond the sparsification of learning attributable to dendritic learning thresholds , learning is sparsified even further by the fact that within each learning dendrite , only the active 3% of synapses receives ( and obeys ) the instruction to potentiate or refresh , and that same small fraction of synapses is depressed . Thus , in the above scenario , relative to a 1-layer network with the same coding density of 3% , the existence of a dendritic learning threshold sparsifies learning by a factor of 1/PL = 2000 , significantly boosting capacity without requiring extreme , biologically unrealistic coding sparseness . In our model the formation of new memories is achieved through long-term potentiation ( or rejuvenation ) of a few activated synapses on a few strongly activated dendrites that undergo learning events . The "forgetting" of old memories involves heterosynaptic depression of the least-recently-potentiated/rejuvenated synapses in the same dendrites that are undergoing learning . Given the pressure to keep memory traces at their bare minimum strength , when our model is optimized for capacity , synaptic changes are exceedingly sparse , involving only a small fraction of the synapses on a minute fraction of dendrites . ( The finding that memory capacity is optimized by sparse patterns has also been reported for 1-layer models: [2 , 111–114] ) . For example , in a memory network containing ~5 million synapses , under conditions that optimize storage capacity ( i . e . with dendrites containing ~256 synapses , and patterns of 3% density ) , we found that each time a pattern is learned , only 150 of the 5 million synapses learn ( 0 . 003% ) , less than half of which are overtly strengthened ( i . e . some are only rejuvenated ) , and those few altered synapses are confined to just 10 of the 20 , 000 dendrites contained within the network . If we consider extremely sparse synaptic plasticity to be a prediction of our model , could such sparse changes be detected experimentally ? The likelihood of detecting changes in this few dendrites seems higher when it is considered that 20 , 000 dendrites corresponds to 500–1 , 000 neurons . We would thus expect that 10 ( i . e . 1–2% ) of the neurons in the network would contain a dendrite that participates in learning . In vivo imaging techniques with a field of view containing hundreds of neurons should make this level of detection possible . What role might structural plasticity play in online learning ? We previously explored the role that active dendrites might play in familiarity-based recognition in the very different scenario where patterns can be trained repeatedly [46 , 115] . The opportunity for repeated , interleaved training of patterns gives the system time to exploit wiring plasticity mechanisms [116] , wherein existing connections between axons and dendrites can be eliminated and new ones formed in such a way that correlated inputs end up forming contacts onto the same dendrites . This type of wiring plasticity is not an option in an online learning scenario , since each pattern is experienced only once , such that all learning-related synaptic changes must be immediate–or at least immediately induced . We showed that correlation-based sorting of inputs onto different dendrites using a Hebb-type learning rule increased the storage capacity of a neuron by more than an order of magnitude compared to a neuron with the same total number of synaptic inputs that lacked dendrites . Furthermore , as here , we found that dendrites of intermediate size optimized capacity–though for different reasons . It is interesting to note that in our current model , structural turnover of weak synapses has no effect on what is stored in the memory , as long as new weak synapses are added to the system at the same rate that existing weak synapses are removed . If weak synapses form a substantial fraction of the total synapse population–we have assumed 50% here ( but the percentage may actually be closer to 90% in CA1 –see [117] ) –then high rates of spine elimination and new spine formation could be tolerated within the memory area without any loss of stored information–again , as long as the turnover is restricted to weak synapses . What would be the advantage of eliminating existing weak connections and forming new ones ? Under the assumption that input axons are uncorrelated , as we have assumed in this work for simplicity , we can see no advantage to this type of structural turnover . However , if meaningful correlations between input axons do exist , then structural turnover could be a sign that wiring plasticity mechanisms are attempting to co-locate correlated synapses on the same dendrites [118 , 119] , which could lead to a significant capacity advantage [46 , 115 , 116] . Familiarity-based recognition is a very basic form of memory , and is most closely associated with the perirhinal cortex [10 , 87 , 88] . However , currently available data regarding the responses of familiarity ( vs . novelty ) neurons in the PRC is complex , and not easily related to our findings here ( see S1 Text for an in depth discussion ) . Further work will be required to determine whether the dendrite-based architecture of Fig 2b will be helpful in explaining familiarity-based recognition processes in the brain . What can the dendrite-based architecture we have studied here tell us about other types of memory systems ? A trivial extension of our architecture in which N copies of the memory network are concatenated would allow the construction of a full N-bit binary online associative memory . This type of memory would behave exactly as ours , but would allow an arbitrary N-bit output pattern to be one-shot associated with each input pattern , as in a Willshaw network . In this scenario , only the subset of the N networks whose outputs are instructed to be 1 would learn each input pattern , while any networks instructed to produce 0 responses would simply ignore the input pattern . If the output patterns are sparse ( which they needn't be ) , only a small fraction of the networks would need to participate in the learning of each association . It might also be desirable to assign extended lifetimes to particularly important patterns; this could be accomplished in either of two ways: 1 ) Extended-lifetime synapses could be established during the learning of important patterns , so that the synapses representing those patterns would remain invulnerable to depotentiation for longer times , or even permanently . Doing so would of course reduce the lifetimes of other patterns in the memory . 2 ) The memory could be composed of multiple subnetworks having a range of pattern lifetimes , and important patterns could be stored in longer-lifetime ( i . e . larger capacity or more rarely used ) networks . The decision as to which or how many networks participate in the storage of each pattern could be gated by an "importance" signal provided by another brain area . In other cases it might be valuable to store different trace strengths for different patterns , rather than uniform , bare-bones recognition traces for all patterns . Note this goal is inconsistent with the goal to maximize storage lifetimes for all patterns , but could also be useful in certain ecological situations . Our simple architecture allows for this directly: nothing is to prevent a larger or small number of dendrites from being used in the learning of any particular pattern , such that it's memory trace would be stronger or weaker than the norm . Regardless of trace strength , a pattern’s lifetime would remain roughly the same , since lifetimes are determined mainly by the lengths of the dendritic age queues , which do not depend on the number of dendrites used for storage . The trace strength assigned to each pattern could again be determined by a signal generated by another brain area , whose effect is to raise or lower dendritic learning thresholds . In yet another scenario it might be useful to store gradually decaying memory traces so that trace strength can represent recency of learning ( which is again a different goal than maximizing recognition capacity ) . A graded recency signal can be efficiently produced by storing each pattern simultaneously in multiple networks with a range of capacities/sizes/memory lifetimes . Early in its storage lifetime , the pattern would evoke a memory trace from all networks , so that it's total trace strength would be high , but as time progresses , and its trace progressively expires from the lower-capacity networks , its overall trace strength would gradually decay . This use of such a tiered system to achieve a graded decay time course is more resource-efficient than certain other forms of trace decay that have been considered in the online memory literature , in that the stored information in a tiered network with synapse age management expires in a controlled fashion [109] . Finally , it will require future work to determine which of our results can carry over to Hopfield-style recurrent networks [120–123] constructed from neurons with thresholded dendrites , where the goal in that case would be to maximize recall capacity . In one obvious difference , the ability to recall entire patterns from partial cues requires that the entire patterns be stored ( in stark contrast to the need to generate only a reliable familiarity signal ) , so synapse resource consumption per pattern will be much higher than in the basic familiarity network . Furthermore , the need to modify recurrent synapses during the initial learning of a pattern implies that the participating neurons must fire action potentials during initial learning in order to activate those recurrent connections , which implies that their dendrites must cross both the learning and firing thresholds during learning . Interestingly , this requirement would seem to render such a memory useless for familiarity-based recognition , since the neurons that participate in the learning of a pattern must already fire on a pattern’s first presentation to the memory . This incompatibility could be one reason why the functions of familiarity and recall memory have been assigned to distinct areas within the medial temporal lobe [87 , 88] .
As discussed in the main text , after a certain number of learning events has occurred following the storage of a pattern feature in a dendrite , the strong synapses encoding the stored feature begin to “fall off” the end of the dendrite’s age queue , and the memory trace in the dendrite is effectively lost . We refer to the number of learning events that can be endured before this loss occurs as the length of the age queue L . If we assume that the frequency of learning events is constant across dendrites in the network , given that the queue length L is also constant across dendrites , most of the strong synapses encoding a particular pattern’s features will be depressed roughly simultaneously ( in different dendrites ) , leading to a relatively rapid decay of the network’s overall response r to that pattern . The value of L is therefore a measure of the length of time that a pattern’s trace persists in the memory , and is therefore effectively a measure of capacity in units of dendritic learning events . L can , in principle , be determined by framing learning as a Markov process with the state diagram shown in Fig 3 . Consider a single synapse on a given dendrite . If p→ is the ( L+1 ) ×1 vector containing the probability that , at a given time , this synapse is in each of the L+1 states shown in Fig 3 , and T is the L+1×L+1 matrix containing the state transition probabilities , then with each learning event , p→ will change as p→→Tp→ . After many learning events , p→ will approach the equilibrium distribution , characterized by the condition that learning leaves it unchanged: p→∞=Tp→∞ . Using the fact that for the equilibrium distribution p→∞ , fs of the synapses must be strong , one can solve for L ( since the ( L+1 ) ×1 vector p→∞ implicitly depends on L ) . Using the eigenvectors and eigenvalues of T , one can also compute the distribution p→t after any number of learning events . However , while the Markov approach is very general , the simple dynamics of the age queue allow for a more direct and transparent derivation of L . To find L , we might naively divide the total number of strong synapses per dendrite ( fS∙K ) by the average number of synapses potentiated in each dendrite that experiences a learning event μpot . where μpot≈θLpreμburst . In words , μpot is approximately equal to the total number of spikes impinging on all activated synapses on the dendrite , given by the threshold value θLpre ( since in most cases learning dendrites will have just crossed this threshold ) , divided by the average number of spikes per participating synapse μburst . This gives L≈fS∙K∙μburstθLpre . However , this would underestimate L because synapses that are only juvenated ( i . e . that were already strong ) do not contribute to the aging of synapses further along the age queue , so that the average rate of progression along the age queue slows as strong synapses grow older . To estimate L more accurately , consider the equilibrium distribution of synapse ages in the queue of a single dendrite ( blue histogram in Fig 3 ) . The age of the right-most column of the age histogram is an indicator of the expected age ( measured in learning events ) at which the synapses encoding a pattern are depressed and moved to the unordered collection of weak synapses . During each learning event , a large fraction ( fage ) of synapses in each column that were not activated move rightward to the next older column , while a small fraction ( 1-fage ) are juvenated ( promoted to the first column ) . This process leads to a bias towards younger synapses in the queue , and can be well-approximated by a finite geometric sequence with length L , decay ratio fage , sum fS∙K ( note the sum of the columns is the total number of strong synapses ) , and first column height μpot ( the average number of synapses that learn per dendrite per learning event ) , so that: fS⋅K=μpot⋅1−fageL+11−fage . Assuming that the synapses in a dendrite are all equally likely to be potentiated ( ignoring the effects of the postsynaptic threshold–see below ) , with μpot≈θLpreμburst , then we have that fage≈1-θLpreK∙μburst and can solve the above equation for L . Note that L counts the number of dendritic learning events before a memory is eroded , whereas memory capacity C should count the number of training patterns . Thus , to approximate C , we must multiply L by the approximate number of patterns per dendritic learning event , or “learning interval” 1PL , where PL is the probability that an arbitrary dendrite learns a particular pattern . Although PL is conceptually simple , its expression is complicated since it depends on pattern density , noise level , two learning thresholds , dendrite size , and fS ( see expression below ) . Collecting these results , we can approximate memory capacity by C≈LPL=1PL⋅[log ( 1−fS ) log ( 1−θLpreK⋅μburst ) −1] . For simplicity , the expression for L in the capacity equation does not include the effect of the postsynaptic threshold θLpost , which makes strong synapses more likely to learn , lowers fage and increases absolute capacity . The synapse age distribution remains roughly geometric , however ( see Fig 5b ) , and we observed that the qualitative behavior of the system depends only weakly on θLpost , justifying its omission from the analysis . Synaptic activation on a dendrite is governed by 4 binomial random variables: as , the number of active strong synapses; ss , the number of spikes received by strong synapses; aw , the number of active weak synapses; and sw , the number of spikes received by weak synapses . These random variables have the distributions shown below . Learning occurs when presynaptic activation crosses the presynaptic learning threshold , or ss+sw>θLpre , and postsynaptic activation crosses the postsynaptic learning threshold , or ss>θLpost . Using the distributions for as , aw , ss , and sw , and the fact that PL=pss+sw>θLpre , ss>θLpost , we can write an explicit expression for PL: as~Bi ( fs·K , fA ) ss~Bi ( Nburst·as , pburst ) aw~Bi ( ( 1−fs ) K , fA ) sw~Bi ( Nburst·aw , pburst ) PL=∑i∈[0 , fs·K]j∈[0 , ( 1−fs·K ) ]k∈[θLpost+1 , Nburst·i]l∈[θLpre−k+1 , Nburst·j]Bi ( Nburst·j , pburst ) [l]⋅Bi ( Nburst·i , pburst ) [k]⋅Bi ( fs·K , fA ) [i]⋅Bi ( ( 1−fs ) K , fA ) [j] where BiN , p[k] is the binomial pdf with parameters ( N , p ) evaluated at k . A simpler alternative to evaluating this expression directly is to estimate it by generating a large number of samples of as , aw , ss , and sw according to the above distributions , and directly observing the fraction of cases that cross both learning thresholds . Once the capacity formula is used to calculate how long a given memory trace will last , we must verify that during its lifetime , the trace is sufficiently strong . We do this by checking whether the error tolerances ϵ+ and ϵ- are met immediately after training . First , we compute ϵ+ , the probability that an untrained pattern will be recognized . To be recognized , a pattern must activate at least θR dendrites in the network . For a randomly selected untrained pattern , the distribution of the number of activated dendrites will be approximately Poisson with mean PF⋅M , where M is the number of dendrites in the network and PF is the probability that a given dendrite fires in response to a randomly selected pattern . For a pattern to fire a dendrite , it must cause a postsynaptic activation >θF , or ss>θF , using the notation of above . Since the distribution of ss is known , it is relatively easy to write an expression for PF and ϵ+ explicitly: PF=p ( ss>θF ) =∑i∈[0 , fs·K]j∈[θF , Nburst·i]Bi ( Nburst·i , pburst ) [k]⋅Bi ( fs·K , fA ) [i] ϵ+=∑r≥θRPoiss ( PF⋅M ) [r] As for ϵ- , the probability that a previously trained pattern is forgotten , we approximate this quantity with ϵ-0 , or the immediately post-training false negative rate ( justified by the fact that during the “lifetime” of the memory , C , the trace strength is roughly constant ) . To calculate ϵ-0 , we use the following observation: when training a new pattern , it will learn in a certain set of dendrites . Immediately after training , if the pattern is re-presentated to the network , all of these dendrites should respond , since learning has significantly boosted the pattern’s features in these dendrites . In other words , dendrite readout failures immediately after learning should be very rare . Therefore , for a pattern to be too weak for recognition immediately after training , it must have learned in too few dendrites . The number of learning dendrites for a given pattern will have a Poisson distribution with mean PF⋅M . Therefore , ϵ- can be written ϵ−≈ϵ−0=∑l<θRPoiss ( PL⋅M ) [l] If for the given settings of the learning and firing thresholds θLpre , θLpost , θF , θR , the error tolerances are met–that is , ϵ+ , ϵ-<1%–then the memory lifetime is compared to the best memory lifetime found so far . Otherwise , we continue the search through threshold space . All data contained in figures as well as simulation code is available in S1 Data file titled "Plos data/code" . | Humans can effortlessly recognize a pattern as familiar even after a single presentation and a long delay , and our capacity to do so even with complex stimuli such as images has been called "almost limitless" . How is the information needed to support familiarity judgements stored so rapidly and held so reliably for such a long time ? Most theoretical work aimed at understanding the brain's one-shot learning mechanisms has been based on drastically simplified neuron models which omit any representation of the most visually prominent features of neurons—their extensive dendritic arbors . Given recent evidence that individual dendritic branches generate local spikes , and function as separately thresholded learning/responding units inside neurons , we set out to capture mathematically how the numerous parameters needed to describe a dendrite-based neural learning system interact to determine the memory's storage capacity . Using the model , we show that having dendrite-sized learning units provides a large capacity boost compared to a memory based on simplified ( dendriteless ) neurons , attesting to the importance of dendrites for optimal memory function . Our mathematical model may also prove useful in future efforts to understand how disruptions to dendritic structure and function lead to reduced memory capacity in aging and disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"learning",
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"neural",
"networks",
"engineering",
"and",
"technology",
"nervous",
"system",
"signal",
"processing",
"membrane",
"potential",
"social",
"sciences",
"electrophysiology",
"neuroscience",
"learnin... | 2019 | How Dendrites Affect Online Recognition Memory |
Four SIV-infected monkeys with high plasma virus and CNS injury were treated with an anti-α4 blocking antibody ( natalizumab ) once a week for three weeks beginning on 28 days post-infection ( late ) . Infection in the brain and gut were quantified , and neuronal injury in the CNS was assessed by MR spectroscopy , and compared to controls with AIDS and SIV encephalitis . Treatment resulted in stabilization of ongoing neuronal injury ( NAA/Cr by 1H MRS ) , and decreased numbers of monocytes/macrophages and productive infection ( SIV p28+ , RNA+ ) in brain and gut . Antibody treatment of six SIV infected monkeys at the time of infection ( early ) for 3 weeks blocked monocyte/macrophage traffic and infection in the CNS , and significantly decreased leukocyte traffic and infection in the gut . SIV – RNA and p28 was absent in the CNS and the gut . SIV DNA was undetectable in brains of five of six early treated macaques , but proviral DNA in guts of treated and control animals was equivalent . Early treated animals had low-to-no plasma LPS and sCD163 . These results support the notion that monocyte/macrophage traffic late in infection drives neuronal injury and maintains CNS viral reservoirs and lesions . Leukocyte traffic early in infection seeds the CNS with virus and contributes to productive infection in the gut . Leukocyte traffic early contributes to gut pathology , bacterial translocation , and activation of innate immunity .
The importance of monocyte/macrophages as a critical cell type bringing human immunodeficiency virus ( HIV ) to the central nervous system ( CNS ) is often assumed [1] , [2] , but has not been directly investigated . Similarly , the function of leukocytes seeding the gut early during infection has not been directly assessed . HIV infection of the CNS is associated with compromised motor , behavioral , and cognitive functioning , collectively referred to as HIV-associated neurocognitive disorders ( HAND ) [3] . Neuropathologic correlates of these clinical conditions include accumulation of perivascular macrophages , microglial activation , decreased synaptic/dendritic densities , neuronal damage and loss [4] . Combination antiretroviral therapies ( cART ) restore peripheral immune function and control viral replication , however effective cART does not prevent the formation of a CNS viral reservoir early in infection [5] . Consequently , neuroinflammation remains and neurologic impairment affects the majority of HIV-infected individuals [6] , [7] . Gut-associated lymphoid tissues ( GALT ) are another important reservoir of HIV RNA and DNA that is established during acute infection and persists despite long-term effective therapy [8] , [9] . SIV infection in rhesus macaques results in a disease course similar to HIV-infected humans in the pre-ART era [10] . Experiments in SIV-infected rhesus macaques have provided important insights into the role of innate and adaptive immune cell types in viral persistence and maintenance of tissue reservoirs [11] . SIVmac251 infection with CD8 lymphocyte depletion results in uncontrolled plasma viremia during the first two weeks of infection and rapid progression to AIDS . This rapid and predictable progression to AIDS also allows for therapeutic treatment studies in monkeys because we achieve >85% incidence of AIDS and SIV encephalitis ( SIVE ) within months of infection compared to approximately 25% of non-depleted animals developing SIVE [11] . Similar to HIV infection in humans , virus is detected very early in the CNS , within perivascular macrophage cuffs . But in the rapid monkey model CNS pathology occurs more quickly , and histopathology is more severe with several fold more monocyte/macrophages accumulating early ( 21 days post infection ) , productive infection is easily detectable , and multi-nucleated giant cells ( MNGC ) are present . Within the CNS of HIV-infected humans and SIV-infected monkeys early , and terminally with AIDS , CD4+ T lymphocytes are rare , and not usually detected . Early after exposure to HIV and SIV , virions and infected cells enter the gut and infect resident CD4+ T lymphocytes . These cells harbor virus and propagate infection , resulting in CD4+ T cell loss within days [12] , [13] . With CD4+ T cell depletion , there is expansion of activated immune cells and virus in blood that can infect draining lymph nodes , brain , and other tissues [14] . CD4+ T cell apoptosis during acute HIV and SIV infection is thought to contribute to aberrant immune activation and translocation of microbial products , which can cause increased trafficking of monocytes into the CNS . It is postulated that this is closely linked to the development of HAND and SIVE [15] , [16] . Similar to the gut , SIV and HIV are found in the CNS as early as 3 [17] , [18] and consistently by 14 days post infection ( dpi ) [5] , [19] , and occur concurrently with accumulation of perivascular macrophages , some of which are infected [20] , [21] . Although neurons are not infected , neuronal damage is evident even during the acute phase of infection [11] , [22] , [23] . 1H MRS is a sensitive method of non-invasively measuring neuronal injury by decreased levels of neuronal metabolites N-acetylaspartate+N-acetylaspartylglutamate ( collectively NAA ) . Neuronal injury ( decreased NAA/Cr ) correlates with the expansion of activated monocytes in the periphery , indicating that neuroinvasion , likely through entry of activated or infected monocytes into the brain , is required for CNS pathogenesis [24] . We have previously shown with non-CD8 depleted SIV infected animals , decreased NAA/Cr ratios with neuronal injury that then reverse when inflammation subsides [25] . In contrast with CD8 depletion and SIV infection there is a steep and drastic decrease in NAA/Cr [26] . Using cART [27] and more recently minocycline [28] with ongoing infection and neuronal injury , we have reported a reversal of decreased NAA/Cr consistent with recovery of neuronal injury or lack of further injury . A decrease in peripheral activation of monocytes correlated with a reversal of decreased NAA/Cr [24] . Using BrdU , we have shown that the magnitude of blood monocyte expansion as early as 8 dpi is highly predictive of the rate of disease progression and severity of CNS neuropathology [29] . It is widely considered that monocyte/macrophage traffic and accumulation in CNS drives neuronal injury , though no study has tested whether directly blocking such traffic affects neuronal injury , or blocks CNS infection . In this study , we used the anti-α4 antibody natalizumab ( Biogen Idec ) , which selectively binds the α4 subunit of α4β1 and α4β7 integrins , blocking the interaction between α4 and its' ligands [30] . Natalizumab prevents accumulation of leukocytes ( B cells , T lymphocytes , and monocyte/macrophages ) in the CNS of patients with relapsing-remitting Multiple Sclerosis [31] and small intestine of patients with Crohn's disease [32] , but does not affect normal leukocyte traffic through lymph nodes in humans [33] , [34] or monkeys [35] . Increased expression of α4 on leukocytes and endothelial expression of VCAM-1 are critical for the migration of monocytes and T cells during both HIV and SIV infection [36]–[38] , indicating that this interaction is likely involved in leukocyte migration across the brain and intestinal barriers with immune activation occurring early in disease . In this study , we used natalizumab primarily to assess the requirement of leukocyte trafficking on SIV neuropathogenesis , and secondarily to assess the impact of SIV pathogenesis in the gut . To examine the requirement of leukocytes for neuronal injury and maintenance of viral reservoirs , macaques were treated later in infection ( late; 28 , 34 , and 41 dpi; n = 4 ) and compared to SIV infected non-treated controls ( n = 4 ) , all sacrificed when they developed AIDS ( 49 to 62 dpi ) . To determine if leukocyte traffic is responsible for seeding and/or maintaining viral infection of the brain and gut , animals received natalizumab at the time of infection ( early; 0 , 7 , and 14 dpi; n = 6 ) and were compared to untreated controls ( n = 3 ) , all sacrificed on 22 dpi . In animals treated late , we found decreased accumulation of SIV-infected monocyte/macrophages in the CNS and stabilization of neuronal injury . Early natalizumab treatment prevented macrophage traffic and infection in the CNS , and decreased the number of productively infected cells in the gut . Overall , these data underscore the requirement of monocyte/macrophage traffic for neuronal injury and maintenance of the CNS lesions , and indicate that early leukocyte traffic is critical for seeding the CNS and contributes to seeding of gut with virus .
The eight SIV-infected macaques in the later cohort ( n = 4 natalizumab treated , n = 4 non-treated ) developed AIDS . One of the four natalizumab treated animals and two of four untreated macaques developed SIVE , defined by productive viral replication , the presence of MNGC , and macrophage accumulation in the CNS . Plasma viral loads in all animals remained elevated regardless of treatment and there were no differences in viral loads between control and experimental animals ( Figure S1 ) . Interestingly , CSF viral load was elevated in early in animals that were treated at the time of infection . We assessed the requirement of continuous monocyte/macrophage traffic for neuronal injury and maintenance of CNS reservoirs with three weekly natalizumab treatments ( 30 mg/kg ) beginning on 28 dpi , when significant neuronal damage had already occurred [22] , [23] . Neuronal injury ( decreased NAA/Cr ) was measured in frontal cortex ( FC ) , parietal cortex ( PC ) , basal ganglia ( BG ) , and white matter semiovale ( WM ) of the four natalizumab treated and four untreated macaques by MR spectroscopy biweekly ( Fig . 2 ) . The mean NAA/Cr ratio declined from pre-infection to 4 weeks post infection ( wpi ) in FC ( −13% , p = 0 . 0028 ) , PC ( −8 . 3% , p = 0 . 0016 ) , BG ( −9 . 7% , p = 0 . 008 ) , and WM ( −8% , p = 0 . 036 ) of all animals ( Fig . 2A–D ) , consistent with neuronal damage as previously reported [23] , [26] , [27] . Following natalizumab treatment , NAA/Cr decreases stabilized in the FC ( +0 . 5% , p = 0 . 892 ) , PC ( −3 . 3% , p = 0 . 596 ) , BG ( −2 . 1% , p = 0 . 757 ) and WM ( −5 . 7% , p = 0 . 046 ) . In contrast , SIV infected , non-treated animals had continued reductions of NAA/Cr in the FC ( −13 . 2% , p = 0 . 016 ) , PC ( −12 . 5% , p = 0 . 0008 ) , and WM ( −11 . 9% , p = 0 . 0001 ) , and a trend towards decline in the BG ( −6 . 8% , p = 0 . 13 ) as previously demonstrated ( Fig . 2E–H ) [26]–[28] . When comparing the NAA/Cr slopes following natalizumab treatment , there was a significant difference between groups in the PC ( untreated r2 = 0 . 76 , treated r2 = 0 . 21; p = 0 . 038 ) and WM ( untreated r2 = 0 . 83 , treated r2 = 0 . 76; p = 0 . 031 ) , but not the FC ( untreated r2 = 0 . 77 , treated r2 = 0 . 0005; p = 0 . 057 ) and BG ( untreated r2 = 0 . 49 , treated r2 = 0 . 28; p = 0 . 362 ) . Along with NAA/Cr , changes in Myo-inositol ( MI ) and Choline ( Cho ) were monitored throughout infection , with no significant differences in MI/Cr or Cho/Cr between treatment groups at any point during the study . Relative to control animals , there were increased numbers of leukocytes and leukocyte precursors in the blood with natalizumab , which is likely due to their inability to traffic to parenchymal , non lymphoid tissues ( Table 1 ) . In late treated macaques , we found elevated numbers of CD4+ lymphocytes ( control: 450 cells/µl , treated: 1444 cells/µl ) , CD14+ monocytes ( control: 260 cells/µl , treated: 879 cells/µl ) CD14l°CD16+ monocytes ( control: 15 cells/µl , treated: 48 cells/µl ) , CD20+ B cells ( control: 464 cells/µl , treated: 1140 cells/µl ) , and CD34+ hematopoetic stem cells ( HSCs ) ( control: 1 . 9 cells/µl , treated: 21 cells/µl ) . With early natalizumab treatment , there were elevated numbers of circulating CD4+ lymphocytes ( control: 536 cells/µl , treated: 903 cells/µl ) , CD14+ monocytes ( control: 355 cells/µl , treated: 725 cells/µl ) , CD14+CD16− monocytes ( control: 260 cells/µl , treated: 467 cells/µl ) , CD14+CD16+ monocytes ( control: 55 cells/µl , treated: 123 cells/µl ) , CD14l°CD16+ monocytes ( control: 17 cells/µl , treated: 96 cells/µl ) , CD20+ B cells ( control: 91 cells/µl , treated: 415 cells/µl ) , and CD34+ HSCs ( control: 1 cell/µl , treated 30 cells/µl ) . These data indicate that α4 blockade with natalizumab prevented cell trafficking , as previously described in humans [31] , [39] , [40] and monkeys [35] . Although there was accumulation of all immune cell subsets in the blood , this increase only affected the percentage of CD14l°CD16+ monocytes , B cells , and HSCs in late natalizumab treated animals , and the proportions of CD14+CD16+ and CD14l°CD16+ monocytes , CD20+ B cells , and CD34+ HSCs with early natalizumab treatment ( Table 1 ) . All control and natalizumab treated animals remained CD8 lymphocyte depleted throughout the study ( Table 1 ) . In tissues we found no CD3+CD8+ lymphocytes in the CNS of early and late natalizumab treated or control animals , nor were there CD3+CD8+ lymphocytes in lymph nodes . This is consistent with these animals being persistently CD8 lymphocyte depleted as we have previously described [11] , [24] , [27] , [29] , [41] . Similarly , we did not find CD3+CD8− lymphocytes ( consistent with CD3+CD4+ cells ) in the CNS . In gut regions sampled , we did not find CD3+CD8+ lymphocytes , but did find CD3+CD8− lymphocytes ( see below ) that represent CD4+ lymphocytes and/or NK cells . Together , these data are consistent with blocking leukocyte traffic to the CNS and gut with natalizumab . In all brain regions examined ( frontal cortex , parietal cortex , occipital cortex , brainstem ) , numbers of SIV p28+ and RNA+ cells were markedly lower in late natalizumab treated versus untreated animals ( p28+ p = 0 . 0202 , RNA+ p = 0 . 0005; Fig . 3A ) . There were significantly fewer activated CD68+ resident macrophages ( p = 0 . 0017; Fig . 3B ) and recently infiltrating MAC387+ monocytes ( p = 0 . 0003; Fig . 3C ) in late treated macaques . We have previously observed significant numbers of BrdU+ macrophages in the CNS of animals receiving BrdU even 24-hours prior to sacrifice [29] , yet no BrdU+ macrophages were found in brains of animals that received BrdU after natalizumab treatment began ( 33 dpi , 24-hours prior to necropsy ) ( Fig . 3D ) . In animals that received BrdU throughout infection ( −9 dpi , 26 dpi , and 24-hours prior to necropsy ) , we found lower numbers of BrdU+ cells in natalizumab treated animals versus controls ( Fig . 3D ) . These data demonstrate that natalizumab treatment with ongoing infection blocks monocyte/macrophage traffic , reduces the CNS reservoir of productively infected monocyte/macrophages , and stabilizes neuronal injury . In the gut ( duodenum , jejunum , colon ) , there were fewer SIV p28+ cells in treated animals ( p = 0 . 0187 ) , but no difference in the number of RNA+ cells observed between late natalizumab treated animals and untreated controls ( Fig . 4A ) . There were lower numbers of CD68+ macrophages ( p = 0 . 0460; Fig . 4B ) , MAC387+ monocytes ( p = 0 . 0182; Fig . 4C ) , and CD3+ T lymphocytes ( p = 0 . 0001; Fig . 4D ) in the guts of natalizumab treated macaques , indicating that late treatment was not sufficient to stop viral infection that has already occurred , but did reduce subsequent traffic of lymphocytes and monocyte/macrophages . Next we sought to determine whether weekly natalizumab treatment at the time of infection ( early ) blocks viral seeding of the CNS and gut . Relative to untreated controls ( n = 3 ) that were also sacrificed at 22 dpi , there were fewer SIV p28+ ( p = 0 . 0004 ) and RNA+ ( p = 0 . 0024 ) cells ( Fig . 3A ) , and CD68+ macrophages ( p = 0 . 0016; Fig . 3B ) in the CNS of early natalizumab treated animals ( n = 6 ) . When present , SIV p28+ and RNA+ cells were primarily found in vessels outside the parenchyma . Numbers of MAC387+ cells were lower in brains of early treated macaques ( p = 0 . 0179; Fig . 3C ) , and recently trafficking BrdU+ cells were absent ( Fig . 3D ) . In the guts of animals receiving early natalizumab , there was a significant reduction in SIV p28+ ( p = 0 . 0012 ) and RNA+ cells ( p = 0 . 0013 ) ( Fig . 4A ) . There were similar numbers of CD68+ macrophages in early treated and control groups ( Fig . 4B ) , but lower numbers of MAC387+ monocytes ( p<0 . 0001; Fig . 4C ) and CD3+ T lymphocytes ( p = 0 . 0001; Fig . 4D ) with natalizumab treatment . Interestingly , early treated macaques had significantly lower plasma LPS at 8 ( p<0 . 0001 ) and 12 dpi ( p = 0 . 0019 ) than untreated controls ( Fig . 5A ) . In contrast , equivalent LPS levels were seen in late treated and non-treated animals . Early natalizumab also resulted in reduced soluble CD163 in plasma , with treated macaques exhibiting significantly lower concentrations than untreated macaques at 12 ( p = 0 . 0488 ) and 21 dpi ( p<0 . 0001 ) ( Fig . 5B ) . To determine whether natalizumab treatment on the day of SIV infection blocked latent viral infection in brain and gut , we analyzed tissues for SIV gag DNA using qPCR . Proviral DNA was undetectable in brains of five of six early natalizumab treated macaques ( Fig . 6A ) . One animal had a low level of SIV DNA that was detected only in brainstem , which may be explained by brainstem trauma resulting from a CSF tap . Although natalizumab significantly reduced the number of productively infected cells in the gut , similar numbers of viral DNA copies were found in the duodenum and jejunum of natalizumab and untreated animals . SIV gag DNA levels were lower in colon with treatment , however this difference did not reach significance ( Fig . 6B ) . There was elevated SIV provirus in axillary lymph nodes and similar levels of SIV DNA in mesenteric lymph nodes of early natalizumab treated relative to untreated controls ( Fig . 6C ) , probably reflecting differing degrees of α4β1 and α4β7 utilization in these different compartments . This was not surprising , as comparable numbers of SIV p28+ and RNA+ infected cells were detected in lymph nodes from treated and untreated animals in both late and early cohorts ( Fig . 7A ) . Natalizumab treated animals had fewer CD3+ T lymphocytes in lymph nodes than matched controls ( Late p = 0 . 0011 , Early p = 0 . 0006; Fig . 7D ) , yet similar numbers of CD68+ ( Fig . 7B ) and MAC387+ monocytes ( Fig . 7C ) were observed in all animals , suggesting that natalizumab did not affect immune recirculation in lymph nodes .
While it has been suggested that monocyte/macrophage traffic drives CNS infection and neuron damage , this has not been demonstrated experimentally . Here , we examined whether continuous neuronal injury with HIV and SIV infection depends on monocyte/macrophage traffic , and if cell trafficking to CNS and gut is required for viral seeding . NAA/Cr was monitored throughout infection in four SIV-infected rhesus macaques treated with natalizumab beginning after 28 days of infection , when significant neuronal injury had already occurred . We have previously shown these decreases to correlate with increased monocyte/macrophage activation , accumulation , viral infection , and neuronal injury by immunohistochemical and neuropathologic examination [11] , [22] , [24] , [42] . Despite significant reductions in NAA/Cr , blocking cell traffic with natalizumab stabilized NAA/Cr declines , consistent with limiting further neuronal injury . Because natalizumab also blocks lymphocyte traffic , it is possible that lymphocytes might also play a role in neuronal injury , however it has been repeatedly demonstrated that there are low-to-no CD4+ T cells in the CNS with HIV and SIV infection [11] , [43] , [44] , and our animals were CD8 lymphocyte depleted . Although not absolutely demonstrated , this underscores the importance of monocyte/macrophage more so than lymphocyte traffic in SIV neuropathogenesis . To determine whether leukocyte traffic is required for initial seeding of brain and gut , animals were treated with natalizumab beginning on the day of infection . At sacrifice 22 days later , no SIV p28+ or RNA+ cells were found in the CNS , indicating that traffic of leukocytes from the periphery is necessary for initial viral dissemination in the brain . This is further supported by the absence of SIV gag DNA in brain tissues of five of six natalizumab treated animals . Provirus in the brainstem of the sixth macaque may be a result of a CSF tap trauma , and a lower concentration of SIV gag DNA was found in this animal than in brains of non-treated controls . There were no BrdU+ cells in early or late natalizumab treated animals , indicating that α4 blockade was sufficient to prevent BrdU+ monocyte/macrophages from entering the brain . We have previously reported that the majority of BrdU+ cells in the CNS of SIV-infected animals are MAC387+ [29] , [41] , underscoring the role of recently recruited MAC387+ monocytes in active CNS inflammation [11] . The few scattered MAC387+ monocytes and CD68+ macrophages seen in brains of treated animals suggests that despite SIV infection and CD8 lymphocyte depletion , very little inflammation occurred in the CNS following natalizumab treatment . Blocking leukocyte traffic later in animals with ongoing inflammation and lesions reduced inflammation to almost undetectable levels . These observations with low numbers of SIV p28+ and RNA+ cells and rapid stabilization of NAA/Cr in the brains of late natalizumab treated animals suggest that ongoing traffic maintains not only neuronal injury , but also productive infection of the CNS . The small intestine is a primary site for SIV infection , with interaction between the α4β7 integrin and MAdCAM-1 facilitating traffic of leukocytes [45] , [46] . Natalizumab reduced numbers of CD3+ T lymphocytes , MAC387+ monocytes , and SIV p28+ cells relative to controls , suggesting that treatment suppressed traffic of cells responsible for early viral replication . It has previously been shown that loss of α4β7HIGHCD4+ T cells in blood is an indication of decreased numbers of CD4+ T cells in gut [47] . Whether we directly blocked trafficking of α4β7HIGHCD4+ T cells was not assessed . SIV DNA was detected in gut tissues of early natalizumab treated animals , however it is plausible this is non-integrated DNA , as very low numbers of SIV p28+ and RNA+ cells were observed in these tissues . Others have demonstrated that with early infection of GALT with ART given four hours after infection , there is protection against rapid depletion of CD4+ T cells , yet SIV RNA and DNA were detected [48] . Despite high levels of SIV provirus in the gastrointestinal tract with natalizumab treatment , productive viral infection appeared to be controlled . This is in contrast to what was seen in the CNS , which might be accounted for in part by the BBB . It is established that in the CNS , the BBB controls traffic of cells , which can be blocked by natalizumab . Therefore , blocking α4β1 and α4β7 likely has a more limited impact in gut than in the CNS . It is important to note that the majority of T cells trafficking to GALT utilize α4β7 , but this is a small population in the blood [46] , [49] . The viral envelope protein gp120 can bind to the α4β7 receptor expressed by leukocytes homing to the small intestine , which may not affect cell infection , but can result in activation and apoptosis of T lymphocytes by HIV and SIV [50] . This could be why we observed low numbers of CD3+ T cells despite similar levels of SIV DNA in the guts of natalizumab treated animals . We did not find differences in plasma and CSF viral loads with and without natalizumab treatment ( early and late ) . The lack of differences in plasma underscores that plasma viral load in this study is not driving tissue pathogenesis , and is more likely the result of cell-associated virus and inflammatory cells . The significance of these observations is not clear . It is unclear whether treatment late , once virus has entered the CNS , would be expected to have an effect on CSF virus since CNS parenchymal infection has already been established . The lack of differences in early treatment could be the result of virus entering the CNS as free virus , but to date this has not been convincingly demonstrated and there was no productive infection in the parenchyma . A more likely scenario is that cell-associated virus enters the CNS via the area postrema that has no BBB , and/or the choroid plexus , which has tight junctions ependymal cells , but not endothelial cells . Thus , there may be minimal receptors that are blocked with natalizumab and therefore cell traffic may not be blocked . Early inflammation and traffic of leukocytes to the choroid plexus with HIV and SIV infection and evidence of productive viral infection does occur in the choroid plexus . These observations support that the choroid plexus is a likely source of elevated virus we find in the CSF , despite natalizumab treatment . With regard to decreased traffic of cells to the gut , we found low levels of plasma LPS with early natalizumab treatment , but no difference with late treatment , suggesting that inflammation in the intestine early during infection contributes to mucosal damage and endotoxin translocation . In addition to inhibiting release of microbial products such as LPS from gut , low levels of sCD163 in plasma were also observed . Because significant reductions of chemokine/cytokine production in blood and CSF have previously been shown in natalizumab treated patients [51] , this may also have contributed to an overall reduction in peripheral immune activation , as suggested by the reduced sCD163 and chemokine expression on monocyte and T lymphocyte populations in blood . We found decreases in ex vivo transmigration and adhesion of PBMCs from natalizumab treated animals , further supporting the diminished ability of cells to traffic to the brain and gut . Similar numbers of MAC387+ and CD68+ monocyte/macrophages in lymph tissues of untreated and treated macaques suggest that natalizumab did not significantly affect traffic of these cells to lymph nodes , a finding made by others using natalizumab in monkeys [35] , [52] . These same studies also demonstrated normal regulatory immune function in natalizumab treated macaques , but increased numbers of lymphocyte precursors , monocyte/macrophages , and T cells in blood . We observed a similar expansion of CD14+ monocytes , CD4+ T lymphocytes , CD34+ hematopoietic progenitors , and CD20+ B lymphocytes ( Table 1 ) , as well as a decline of CD49d ( α4 integrin ) expression in the periphery of all treated animals . It was surprising to observe fewer CD3+ T lymphocytes in lymph nodes of both groups of natalizumab treated macaques , however this may be explained by higher numbers of SIV p28+ and RNA+ cells and elevated SIV DNA copies in lymph nodes , and therefore high numbers of infected leukocytes that are susceptible to apoptosis . Several papers have shown normal lymphoid follicle function and no major differences in immune function with natalizumab treatment , as both monocytes and T cells use the interaction between leukocyte function antigen ( LFA ) -1 with ICAM-1 or ICAM-2 in order to traffic into high endothelial venules [35] , [53] , [54] . Early initiation of effective cART reduces CNS disease [55] , suppresses virus to non-detectable levels , and reduces HIV transmission , however current therapies are not sufficient to eradicate viral reservoirs [56] . Furthermore , many cART therapies have low CNS penetration and do not target monocyte/macrophages that drive cardiac and CNS pathology . While we do not suggest using natalizumab long-term in HIV-infected patients , one might consider whether natalizumab treatment early , in combination with antiretroviral therapy , could stop productive infection of the brain and gut , preventing the establishment of these tissue reservoirs . While PML is a concern in patients receiving natalizumab for extended periods , all reported incidents have occurred after more than a year of antibody treatment . Additionally , patients with JC viral antibodies have received effective natalizumab treatment for 24 months without the development of PML [57] . Regardless , the experiments described here underscore the critical role of monocyte/macrophage traffic in ongoing neuronal injury , and establishment and maintenance of viral reservoirs in the CNS and intestinal tissues .
A total of seventeen rhesus macaques ( Macaca mulatta ) were intravenously inoculated with SIVmac251 ( 20 ng SIV p27; a generous gift from Dr . Ronald Desrosiers , NERPC ) . CD8 lymphocyte depletion was achieved using cM-T807 , an α-CD8 antibody that was administered subcutaneously ( 10 mg/kg ) on day 6 post infection ( pi ) and intravenously ( 5 mg/kg ) on days 8 and 12 pi [58]–[60] . Eight macaques ( n = 4 late natalizumab treated , n = 4 untreated ) were sacrificed at similar time points with progression to AIDS ( 49 to 62 dpi ) . Nine animals ( n = 6 early natalizumab treated , n = 3 untreated ) were sacrificed at 22 dpi . Plasma and CSF SIV RNA were quantified in all animals at various time points throughout infection using real-time PCR as previously described [60] . The recombinant humanized IgG4 monoclonal anti-α4 integrin mAb ( natalizumab ) was kindly provided by Biogen Idec ( Cambridge , MA ) in a sterile concentrated solution . This antibody has specificity for the α4 subunit of α4β1 ( very late activation antigen 4 , VLA-4 ) and α4β7 integrins expressed on the surface of all leukocytes except neutrophils [61] . The rhesus macaque α4 sequence exhibits 96% homology with the human sequence ( NCBI ) , and the anti-α4 antibody binds to the α4 subunit with affinity comparable to that in humans ( Kd = 0 . 04–0 . 07 µg/ml ) [35] . The pharmacokinetic half-life of natalizumab in humans is 11±4 days , however more than 70% of α4 integrin sites remain saturated 4 weeks after infusion and cell counts in the CSF are significantly reduced for up to 6 months [33] . The antibody was administered once weekly for three weeks beginning on the day of infection ( 0 dpi , n = 6 ) or 28 days after infection ( 28 dpi , n = 4 ) . On the day of infusion 30 mg/kg of α-VLA-4 was injected into a 250 mL bag of 0 . 9% NaCl and administered intravenously ( iv ) over 30–60 minutes . We chose a high dose of natalizumab and only treated three times with one-week intervals between each treatment to avoid hypersensitivity responses by the monkeys to the humanized antibody . This regimen has previously been shown to maintain high serum levels of natalizumab throughout treatment in rhesus macaques [35] . Chemistry panels including alanine aminotransferase ( ALT ) and aspartate aminotransferase ( AST ) were examined at various time points throughout infection and remained below 100 IU/L , indicating that natalizumab treatment did not induce hepatotoxicity . 5-bromo-2′-deoxyduridine ( BrdU ) ( Sigma ) was prepared as a 30 mg/mL stock solution in 1× PBS ( Ca2+/Mg2+ free; Mediatech Inc . ) and given intravenously at 60 mg/kg as described previously [29] . To monitor levels of monocyte/macrophage trafficking out of the bone marrow , in blood , and into the CNS and gut , BrdU was administered prior to infection ( −9 dpi ) , at peak infection ( 26 dpi ) , and 24 hours prior to necropsy in two macaques given natalizumab beginning on 28 dpi and two untreated control animals . In the other thirteen animals , BrdU was administered once natalizumab treatment was initiated , on days 33 and 47 post infection ( n = 2 late treated , n = 2 untreated ) or days 6 and 20 post infection ( n = 6 early treated , n = 3 untreated ) . Flow cytometric analyses were performed as previously published [24] , [62] using 100 µl samples of blood stained with the following fluorochrome-conjugated primary antibodies: anti-CD3-Alexa Fluor 700 ( SP34-2 ) , anti-CD4-PerCp-Cy5 . 5 ( L200 ) , anti CD8-APC ( RPA-T8 ) , anti-CD11b-Alexa Fluor 700 ( 1CRF44 ) , anti-CD14-Pacific Blue ( M5E2 ) , anti-CD16-PE-Cy7 ( 3G8 ) , anti-CD20-APC ( 2H7 ) , anti-CD20-APC-Cy7 ( L27 ) , anti-CD25-PE ( M-A251 ) , anti-CD34-PE ( 563 ) , anti-CD49d-PE-Cy5 ( 9F10 ) , anti-CD95-FITC ( DX2 ) , anti-CD195-APC ( 3A9 ) , and isotype control anti-IgG1 , κ-FITC ( DX2 ) from BD Biosciences , HLA-DR-Texas Red-PE ( Immu-357; Beckman Coulter ) , anti-CD163-PerCp-Cy5 . 5 ( GHI/61; Biolegend ) , anti-CD8-PE ( DK25; Dako ) , anti-CD28-PE-Cy7 ( CD28 . 2; eBioscience ) , anti-CD8-Qdot-655 ( 3B5; Invitrogen ) ; anti-CD44v6-Biotin ( VFF-7; Invitrogen ) , anti-CD4-Qdot-605 ( 19Thy-5D7; NIH Nonhuman Primate Reagent Resource ) , anti-CCR2-PE ( 48607; R&D systems ) , and anti-CD64-FITC ( 22; Trillium Diagnostics ) . Samples were fixed in PBS containing 2% formaldehyde , acquired on a FACSAria cell sorter ( Becton-Dickinson ) and analyzed with Tree Star Flow Jo version 8 . 7 . Monocytes and lymphocytes were first selected based on size and granularity using forward scatter ( FSC ) area vs . side scatter ( SSC ) area . From this gate , doublets were excluded ( FSC area vs . FSC height ) . Populations were further identified using negative selection and positive expression of various cell markers using 12-color flow cytometry panels . Complete blood counts were obtained using a CBC Hematology Analyzer ( Hema-True , HESKA ) and the absolute number of peripheral blood cell subsets was calculated by multiplying the total white blood cell count by the total percentage of each population as determined by flow cytometric analysis . To determine if blocking monocyte/macrophage traffic impacted neuronal injury , n = 4 rhesus macaques were treated with natalizumab beginning on 28 dpi . These animals and non-treated controls ( n = 4 ) were scanned prior to infection ( 2× ) and biweekly thereafter until sacrifice . For imaging , each animal was tranquilized , intubated , and monitored continuously throughout the scanning procedure as previously described [26] , [28] . Briefly , MR imaging and spectroscopy were performed on a 3 Tesla whole-body imager ( Magnetom TIM Trio , Siemens ) with a circularly polarized transmit-receive extremity coil . First a three-plane localizer scan , used for positioning and to ensure 1H voxel reproducibility , was acquired . The 1H MRS volumes of interest ( VOI ) were then chosen as previously described [26] , [28] . Single-voxel proton spectra were acquired from the parietal cortex ( PC ) , frontal cortex ( FC ) , basal ganglia ( BG ) and white matter semiovale ( WM ) using the point resolved spectroscopy sequence ( PRESS ) with WET [63] water suppression . Spectroscopic data were processed using LCModel software and concentrations of NAA ( N-acetylaspartate+N-acetylaspartylglutamate ) and creatine-containing compounds ( Cr ) were quantified using the unsuppressed water signal as an internal intensity reference . Levels of sCD163 in plasma were determined using an ELISA kit , according to the manufacturer's protocol ( Trillium Diagnostics ) as previously described [29] . Endotoxin lipopolysaccharide ( LPS ) levels in heat-inactivated plasma were measured using the Limulus Amebocyte Lysate ( LAL ) test ( Associates of Cape Cod Inc . ) as previously described [29] . Samples were diluted fivefold with endotoxin-free water and heated ( 30 min at 65°C ) to inactivate plasma components . Following incubation with LAL ( 30 min at 37°C ) and chromogen , duplicate samples were read at 570 nm in a photometric plate reader . LPS concentrations were expressed in endotoxin units ( EU ) , with an assay sensitivity range of 0 . 005 EU/mL–50 EU/mL . On the day of sacrifice , animals were anesthetized with ketamine-HCl and euthanized by intravenous pentobarbital overdose . Axillary lymph node , intestinal ( duodenum , jejunum , and colon ) , and cerebral ( brainstem , frontal cortex , parietal cortex , and occipital cortex ) tissues were collected in 10% neutral buffered formalin , embedded in paraffin , and sectioned at 5 µm . For immunohistochemistry , tissue sections were deparafinized , rehydrated and incubated with blocking reagents . Mature resident monocyte/macrophage and activated microglia were assessed using anti-CD68 ( KP1; Dako ) and newly infiltrating monocytes were identified by the expression of myeloid/histiocyte antigen MAC387 ( MAC387; Dako ) as previously described [41] . T-lymphocytes were double stained with anti-CD3 ( A 0452; Dako ) followed by anti-CD8 ( 1A5; Vector Laboratories ) , and BrdU+ cells were examined using anti-BrdU ( Bu20A; Dako ) as previously described [29] . Because the CD4 antigen is not optimally detected in routine paraffin embedded sections , we used double CD3 and CD8 to detect CD3+ cells and then determined if they were CD8+ or CD8− . CD3+CD3− cells in the CNS would include CD3+CD4+ lymphocytes . Productive SIV infection was determined with anti-SIV-p28 ( 3F7; Fitzgerald Industries International ) and by in situ hybridization for SIV RNA using anti-digoxigenin labeled SIVmac239 antisense riboprobes that span the entire SIVmac genome ( Lofstrand Labs ) as previously described [64] . Hybridization specificity was confirmed in each experiment using the SIVmac239 sense probe and matched tissue from uninfected rhesus macaques . For quantification , at least 3 non-serial blind-coded sections from all tissues were stained for each marker . Tissue sections were examined with a Zeiss Axio Imager M1 microscope ( Carl Zeiss MicroImaging , Inc . ) using a Plan-Apochromat x20/0 . 8 Korr objective and analyzed by one unblinded and one blinded observer using Adobe Photoshop v11 . 0 . 2 software . The minimum number of arbitrary visual fields analyzed in each tissue was 24 . From this number a median number of cells per tissue region was calculated . Data are represented as the number of positive cells per unit area ( cells/mm2 ) , and each point represents the mean number of positive cells in the three tissue regions examined from a single animal . For each tissue examined , ten 15 µm frozen sections were homogenized and washed in 1× PBS ( Ca2+/Mg2+ free; Mediatech Inc . ) prior to genomic DNA isolation using the AllPrep DNA/RNA Mini Kit ( Qiagen ) according to manufacturers instructions . For each sample , 100 ng of gDNA was loaded in triplicate wells . The concentration of the gDNA was calculated using the Qubit 2 . 0 Fluorometer ( Invitrogen ) . A standard curve was added to each PCR plate , consisting of a plasmid containing 1 copy of the SIV gag gene that was serially diluted from 1e9 copies down to 1 copy per microliter . Each quantitative PCR reaction contained 5 µl of a standard serial dilution or sample ( diluted to 20 ng/µl ) and 20 µl of reaction master mix containing 12 . 5 µl Invitrogen 2× TaqMan Universal Mastermix 2 , 2 . 25 µl each of 10 uM forward and reverse primers , 0 . 625 µl of 10 µM TaqMan probe , and 2 . 375 µl of water . The forward and reverse primers ShehuF 5′-AATTAGATAGATTTGGATTAGCAGAAAGC and ShehuR 5′-CACCAGATGACGCAGACAGTATTAT and the MGB TaqMan probe ShehuP 6FAM-CAACAGGCTCAGAAAA-MGBNFQ were used as described previously [65] . The PCR was performed using Applied Biosystems 7500 Fast Real-Time PCR System under the following conditions: 95°C 10 min followed by 45 cycles of 94°C 15 s and 60°C 60 s . The lowest limit of detection of the assay was 50 copies per reaction . The number of viral gag gene DNA copies per 100 ng of total tissue gDNA was calculated using Applied Biosystems 750 Software v2 . 0 . 5 . Statistical analyses were conducted using Prism version 6 . 0 ( GraphPad Software , Inc . ) . To detect significant changes in NAA/Cr metabolite ratios during disease progression , analysis of variance with repeated measures ( RM-ANOVA ) was used . If significant by RM-ANOVA ( P<0 . 05 ) , Holm-Šídák post-tests were used to isolate significant differences between time points and treatment groups . To determine whether the NAA/Cr slopes between treatment groups were significantly different , linear regression analyses were used to calculate r2 values for each treatment group and P values for differences in NAA/Cr slopes between groups in each brain region . All other P values were calculated using Student's two-tailed , unpaired t tests . Statistical significance was defined as P<0 . 05 . Data are presented as the mean ± the standard error of the mean ( SEM ) . The treatment of animals was in accordance with the Guide for the Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources ( 8th edition ) . The studies were performed with the approval of the Massachusetts General Hospital Subcommittee on Research and Animal Care , and the Institutional Animal Care and Use Committee of Harvard University . Animals were housed according to the standards of the American Association for Accreditation of Laboratory Animal Care . After infection with SIV , animals were individually housed , but received all other components of the NEPRC Environmental Enrichment Program . The Enrichment Program was supervised by NEPRC veterinarians in collaboration with Animal Behavioral staff , not by the PI . Enrichment was provided through manipulatable devices , food items , structural and environmental enhancements , and positive human interaction . Animals did not undergo food or water deprivation at any time during the study and were monitored daily for evidence of disease and changes in appetite and behavior . Clinical support was administered under the direction of an attending veterinarian and included antibiotics , analgesics , and intravenous fluids . Animals were anesthetized with ketamine-HCL and euthanized by intravenous pentobarbital overdose . The New England Primate Research Center ( NEPRC ) Protocol Number for this study is 04420 and the Animal Welfare Assurance Number is A3431-01 . | To determine whether ongoing cell traffic is required for SIV-associated tissue damage , we blocked monocyte and T lymphocyte traffic to the brain and gut during a ) ongoing infection or , b ) at the time of infection . When animals were treated at four weeks post infection ( late ) , once significant neuronal injury and accumulation of infected macrophages had already occurred , neuronal injury was stabilized , and CNS infection and the number of CNS lesions decreased . In the gut , there were significantly fewer productively infected cells and decreased inflammatory macrophages post treatment . Treatment at the time of infection ( early ) blocked infection of the CNS ( SIV –DNA , RNA , or protein ) and macrophage accumulation . In the gut , treatment at the time of infection blocked productive infection ( SIV –RNA and protein ) but not SIV –DNA . Interestingly , with treatment at the time of infection , there was no evidence of microbial translocation or elevated sCD163 in plasma , demonstrating that leukocyte traffic early plays a role in damage to gut tissues . Overall , these data point to the role of monocyte traffic and possibly lymphocytes to the CNS and leukocyte traffic to the gut to establish and maintain viral reservoirs . They underscore the role of monocyte/macrophage traffic and accumulation in the CNS for neuronal injury and maintenance of CNS lesions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"immune",
"cells",
"cytokines",
"immunology",
"neuroscience",
"histology",
"developmental",
"biology",
"molecular",
"development",
"immunologic",
"techniques",
"immunohistochemical",
"analysis",
"neuroimaging",
"research",
"and",
"analysis",
"methods",
"wh... | 2014 | Anti-α4 Antibody Treatment Blocks Virus Traffic to the Brain and Gut Early, and Stabilizes CNS Injury Late in Infection |
The analysis of microbiome compositions in the human gut has gained increasing interest due to the broader availability of data and functional databases and substantial progress in data analysis methods , but also due to the high relevance of the microbiome in human health and disease . While most analyses infer interactions among highly abundant species , the large number of low-abundance species has received less attention . Here we present a novel analysis method based on Boolean operations applied to microbial co-occurrence patterns . We calibrate our approach with simulated data based on a dynamical Boolean network model from which we interpret the statistics of attractor states as a theoretical proxy for microbiome composition . We show that for given fractions of synergistic and competitive interactions in the model our Boolean abundance analysis can reliably detect these interactions . Analyzing a novel data set of 822 microbiome compositions of the human gut , we find a large number of highly significant synergistic interactions among these low-abundance species , forming a connected network , and a few isolated competitive interactions .
An important current trend in the analysis of microbiome compositions is to relate co-abundance patterns with functional capabilities of the microbial species [1–3] . Examples of such analyses include the use of phylogenetic relationships as a proxy for functional similarity [4] , the statistical analysis of an overlap in enzyme content [5] , up to the study of metabolic networks of interacting species via the definition of environmental boundaries of metabolic networks [6] to the concept of metabolic interactions between microbial species [7] . Network approaches are an important ingredient in this endeavor of relating ecological and functional aspects of microbiome composition [8 , 9] . Mathematical models of the microbiome system that compare to data still are rare or address specific situations . For instance , [10] modeled the primary succession of murine intestinal microbiota . This is a case where relatively well-defined initial conditions are available such that theory-experiment comparison becomes feasible . Also concepts from ecological community theory seem promising in explaining microbiome composition . [11] review approaches where environmental selection , habitat types , and invasion processes after disturbance lead to different scenarios of community assembly . But in general , a “theory of the microbiome” , especially with the ambition to provide clinical relevance , is far from being established . In addition to the obvious challenge of finding a suitable representation of functional capabilities , it is also not clear , how co-abundance patterns reliably reveal the set of ( synergistic and competitive ) microbial interactions . Network effects ( i . e . the multitude of positive and negative influences acting upon each microbial population and affecting the abundance pattern ) will impede the link-by-link inference of such microbial interactions . Using simulated abundance patterns and analyzing a large number of stool microbiome samples from a community-based sample , we address the statistical question , how reliably co-abundance patterns reveal ( the set of synergistic and competitive ) microbial interactions . Recently there has been evidence for few discrete stable compositional states , or enterotypes [12] . While this viewpoint has been challenged in the last years [13 , 14] , the hypothesis that microbiome compositions may follow a few distinct patterns remains actively discussed in research . It is especially not clear in general whether microbiome abundances are comprised of clustered states ( enterotypes ) or merely assume different values in a “gradient”-like landscape [13] . Over the last years the potential relevance of the human microbiome for aspects of health and diseases have become ever more apparent [2] . Along with this development goes an increased interest from theoretical biology and systems biology to understand the microbiome , its stability , its contributions to disease onset and progression , as well as its response patterns to perturbations . These debates , together with the availability of ever more data sets on microbiome compositions , emphasize the importance of a theoretical understanding of the dynamical properties of the microbiome . In this context , the inference of microbial association networks from species abundance data has a very important role [8] . Commonly applied methods are regression analysis , local similarity analysis and statistical validations via suitable null models ( see [8] for a detailed review on network inference methods in the context of microbiome compositions ) . In [9] a data analysis pipeline with several correlation and similarity measures as well as Generalized Boosted Linear Models have been used to reconstruct microbial interaction networks across different body sites based on the Human Microbiome Consortium data . In general the inference of the underlying interaction network is a nontrivial task . To address side effects of normalization and statistically under-powered data , [15] introduced a new transformation and graphical inference framework and demonstrated improved detectability of the interaction network for a suitably selected sparsity parameter . A striking feature of microbiome compositions is the wide spread of abundances: Although often dominated by a few highly abundant species , a typical microbiome in the human gut also consists of a wide range of low-abundance species . It seems plausible that the main ‘housekeeping functions’ of such a system of interacting microorganisms are installed by the major , high-abundance components , while subtle adjustments to environmental changes and differences in the host phenotype are achieved rather via this broad range of low-abundance species ( see e . g . , [16 , 17] ) . However their detection is more challenging , as network reconstruction methods based on abundance tend to priorize interactions involving the high-abundance species . In this work , we present and analyze a comparatively large dataset ( consisting of 822 samples ) and introduce a new network inference method based on Boolean networks . The purpose of our investigation is two-fold: First , we introduce a new method for inferring microbial interaction networks from abundance data and test this method using simulated ‘data’ . Second , we apply this method to a new data set of human gut microbiome compositions and show that the co-abundance patterns among low-abundance species contains a multitude of highly systematic , statistically significant interactions . The microbial interaction networks obtained from most analyses are dominated by highly abundant species . Here the negative interaction between Firmicutes and Bacteriodetes is a prominent example , which is also a main basis of the concept of enterotypes . As discussed above , here we assess , whether the low-abundance segment of the microbiome contains evidence for systematic interactions . Therefore , throughout this study we focus on binary data ( i . e . the presence or absence of a microbial species in a particular microbiome sample ) . A simple example illustrates the statistical signal we focus on in these binarized vectors: On a link-by-link basis ( i . e . , for a single pair of microbial species ) , a preference for ( 1 , 1 ) in positive interactions or ( 1 , 0 ) and ( 0 , 1 ) in negative interactions can be expected . It is not clear , however , whether this tendency is also visible in a whole network , where nodes tend to have more than one such positive or negative interaction . Lastly , it is an open question , whether ‘snapshots’ of steady states of the system ( instead of , e . g . , time courses ) allow a reliable reconstruction of such interactions . The main difference of our approach to previous studies is that we binarize the data and make use of the full methodological scope available on this binary state space . This is seen ( i ) in the simulation model we employ to calibrate and test our method ( section Method for testing the analysis method using simulated data ) , ( ii ) in the computational technique we use to distinguish between cooperative and competitive interactions ( sections Boolean abundance analysis and Analysis of simulated data ) as well as in ( iii ) the possibility to treat some of the properties of the binarized abundance vectors analytically ( section Background ) . For the simulation model , we construct random graphs of positive and negative interactions , simulate time courses starting from random initial abundance patterns and thus obtain a set of attractors ( i . e . binary steady state vectors ) , which represent steady-state microbiome compositions on a binarized ( species is present or absent ) level , arising in a network of cooperative ( synergistic ) and competitive ( antagonistic ) interactions . In general , Boolean network models [18] have been very successful in the context of gene regulation: Ignoring the gradual changes of gene activity and rather focusing on the logical organization of the system , this logical circuitry determines the patterns of ‘on’ and ‘off’ states of genes [19] . The binary nature of our data allows us to use the change of co-occurence vectors under Boolean operations as a predictor of the sign of the interaction . A decisive test is , how our new measure performs under noise and with increasing connectivity .
Here we introduce a new method for the inference of microbial interaction networks from microbiome composition data . The method , called Entropy Shifts of Abundance Vectors under Boolean Operations ( ESABO ) , evaluates the information content of pairs of binary abundance vectors , when combined via Boolean operations . In contrast to purely descriptive abundance diversity measures used in population ecology , here we introduce an approach which is directly targeted for the detection of ( synergistic and competitive ) relationships among microbial species . The ESABO method starts from a set {k} ( 1 ≤ k ≤ NA ) of samples with abundances b k ( i ) for each species i . Let OP be a Boolean operation ( OP ∈ {AND , OR , NAND , NOR , …} ) . Then the ESABO score for operation OP of species i with respect to species j is defined as the z-score ( compared to a null model of shuffled abundance vectors ) of the entropy of abundances after pointwise application of a Boolean operation x i j i OP j k ⟵ b k ( i ) OP b k ( j ) ( 1 ) where the z-score is obtained from comparison to an ensemble of x ˜ i j i OP j k ⟵ b k ( i ) OP b ˜ k ( j ) ( 2 ) where b ˜ k ( j ) for each fixed j is an abundance vector randomly reshuffled in k . The entropy H = −∑i pi ln pi ( in physics with prefactor kB or in Shannon theory of information with base 2 ) of any normalized set of probabilities pi is a measure of uncertainty . Here , the sum is over the two states 0 and 1 occurring in the vector ( xijiOPj ) k . An entropy shift therefore can be associated with a gain of information . In it is shown that the occurring patterns can be extracted from one operation and we choose the AND operation which appears more straightforward to interpret . The ESABO method only uses species occurrences as a binary information . In order to calibrate and test our analysis method , we therefore opted for a minimal model , which creates such occurrence patterns on a binary level . We generate a random undirected graph with N nodes , representing N microbial species , connected by M+ positive and M− negative interactions . Starting from random ( binary ) compositions , we update the state of the system according to the following update rule: si ( t+1 ) ={ 1 , ∑j=1NGijsj ( t ) >0si ( t ) , ∑j=1NGijsj ( t ) =00 , ∑j=1NGijsj ( t ) <0 ( 3 ) where G = ( Gij ) is the generalized adjacency matrix of the interaction graph G: Gij = 1 , −1 , 0 for positive , negative or no interactions between species i and j , respectively . For each species i at time t , si ( t ) specifies whether the species is abundant ( si ( t ) = 1 ) or absent ( non-abundant , si ( t ) = 0 ) . From the networks we go via simulated time series across 1000 random initial conditions to asymptotic compositions . In most cases , the observed attractor is a steady state ( see Supporting Information in S1 Text for details ) . In cases where a cyclic attractor is observed , the recorded asymptotic composition will be one of the time points from the cycle . For each interaction network G , we thus obtain a list of attractor vectors a → ( j ) with j = 1 , … , NA , where NA denotes the number of ( numerically observed ) attractors . Each such vector can be seen as an experimental sample of microbial abundancies ( preserving only the information , whether a species is present or absent ) . Such a row vector a → ( j ) of the data matrix A i j = a i ( j ) is in the following called the occurrence vector of the jth sample . The column vector b → ( i ) = { a i ( 1 ) , a i ( 2 ) , … , a i ( N A ) } are named the abundance vector of the ith species across all NA samples . Fig 1 illustrates this setup . In the following , we analyze co-abundances , i . e . the relative frequencies ( approximating pair probabilities ) p k l ( i , j ) of entries ( k , l ) in pairs of abundance vectors ( b → ( i ) , b → ( j ) ) with k , l ∈ {0 , 1} . Furthermore , let p k ( i ) denote the relative frequency of the entry k in the abundance vector b → ( i ) . Then the Jaccard index with respect to ( 1 , 1 ) is defined as J 11 ( i , j ) = p 11 ( i , j ) / min ( p 1 ( i ) , p 1 ( j ) ) . We analyze pairs of abundance vectors via their transformation under Boolean operations . Let x i j ( AND ) be the binary vector obtained from applying a logical AND to the two vectors b → ( i ) and b → ( j ) , i . e . , ( x i j ( AND ) ) k = b k ( i ) AND b k ( j ) , leading to relative frequencies p 0 ( x i j ( AND ) ) and p 1 ( x i j ( AND ) ) of zeros and ones in the resulting vector ( which is of length k ) . The entropy H ( AND ) ( b → ( i ) , b → ( j ) ) = - ∑ k p k ( x i j ( AND ) ) log p k ( x i j ( AND ) ) is then an indicator , whether the vector has become simpler or less simple under the Boolean operation . This ‘entropy shift’ is the main observable in the ESABO method introduced in section Boolean abundance analysis . These entropies can now be compared with entropies obtained from shuffled versions of the original abundance vectors b → ( i ) and b → ( j ) , leading to a z-score ( of entropies compared to the entropies from the shuffled versions ) . This comprises the ESABO score for species pair ( i , j ) , which can be expected to be markedly different for positive and negative interactions between species i and j . In Table 1 the ESABO scores for the network from Fig 1A are shown . The z-scores were always calculated with respect to 1000 randomized networks . Except for one outlier in the case of synergistic links , our method successfully classifies the respective signs of interaction links with high significance |z—score| ≫ 1 . Simulating ‘abundance data’ already in a binarized form allows us to study interaction patterns not masked by the extreme hierarchy of species abundances . However , the Boolean model here only serves the purpose of testing and calibrating the method . It is by no means intended to produce ‘data’ which are in all aspects similar to the true microbial abundance data . In particular , in this minimal model the number of attractors decreases rapidly with the number of links ( see Fig D in the Supporting Information S1 Text ) . 822 individuals from a community-based sample from Schleswig-Holstein ( Germany ) were used as discovery sample set . The stool samples , as well as corresponding phenotypic data and information on diet and nutrition were collected by the PopGen Biobank ( Schleswig-Holstein , Germany ) [20] . Study participants collected fecal samples at home in standard fecal tubes . Samples were shipped immediately at room temperature to the PopGen laboratory . Upon arrival into study center ( within 24 hours ) samples were stored at −80°C until processing . Studies exploring the impact of storage conditions on the samples quality and stability of the microbial communities indicated that storage in RT for 24 hour is recommended for optimal preservation [21 , 22] . Written , informed consent was obtained from all study participants and all protocols were approved by the institutional ethical review committee in adherence with the Declaration of Helsinki Principles . 16S rRNA sequencing , genotype , nutritional , and phenotype data used for the herein described study has been made available to other scientists through PopGen’s biobank general data transfer agreement . Dense single nucleotide polymorphisms ( SNP ) genotype data set ( n = 1 , 074 , 163 SNPs ) derived by combining and quality controlling—using standard methods of data filtering—from Affymetrix 6 . 0 , Affymetrix Axiom arrays and the custom Illumina Immunochip and Illumina Metabochip was used for verification of gender and ancestry of study individuals . Individuals who showed statistically relevant genetic dissimilarity to the other subjects ( population outliers identified by PCA-based mapping against the HapMap III CEU , CHB , JPT and YRI population ) or who showed evidence for cryptic relatedness to other study participants ( unexpected duplicates , first- or second-degree relatives identified by identity by descent estimated using the R-package SNPRelate ( vs . 0 . 9 . 19 ) ) were removed . All gender assignments could be verified by reference to the proportion of heterozygous SNPs on the X chromosome . The final data set consisted of 784 samples . The bacterial genomic DNA for the discovery sample set was extracted manually using MoBio PowerSoil DNA Isolation Kit . The discovery sample set was sequenced using primers amplifying V1-V2 regions of 16S rRNA gene combined with Multiplex IDentifiers ( MIDs ) and adapters established for the a 454 Life Sciences GS-FLX using Titanium sequencing technique as described in [23] . Quality filtering of the 454 GS-FLX data was performed according to [24] in summary only reads that are at least 250 bp long and average quality >25 were kept . The microbiome of discovery sample set was subsetted to 1000 reads per sample and taxonomical census matrix from phylum to genus level were constructed accordingly . Phylogeny based alpha-diversities ( Faith PD ) and beta-diversities ( weighted and unweighted Unifrac ) were calculated with FastTree produced maximum-likelihood tree and Mothur .
First , we test the ESABO method using simulated data , as discussed in Section Boolean abundance analysis . In order to better understand the prediction quality of the ESABO method within this framework of the simulated species interaction networks we evaluated the z-scores of entropy shifts under a Boolean AND for an ensemble of 20 networks ( N = 15 , M+ = M− = 15 ) for positive interactions , negative interactions and a random selection of absent interactions ( see Fig 2 ) . The histogram for negative interactions is clearly centered at negative z-scores , while the positive interactions are predominantly in the positive range , even though some values are in the negative z-score range as well . These outliers will be discussed in more detail below . The sample of absent links yields a narrow distribution of z-scores around zero , confirming that we can expect only a small contribution from false positives in the ESABO method . In the subsequent analysis , we will condense the information contained in the ESABO score even further and define the prediction quality in the following way ( see also Supporting Information ) : The prediction quality of positive interactions is the number of times a z-score larger than 1 is observed minus the number of times a z-score smaller than −1 is found , divided by the number of positive interactions . For negative interactions , negative z-scores are expected . Correspondingly , the prediction quality is the number of times a z-score smaller than −1 is observed minus the number of times a z-score larger than 1 is found , divided by the number of negative interactions . In the case of the Jaccard index J , the prediction quality is the number of times with J > 0 . 6 minus the number of cases J is greater than 0 . 4 minus the number of cases J is smaller than −0 . 4 . The range of connectivity values is limited by two requirements: ( 1 ) We only consider connected networks . ( 2 ) We require more than 100 distinct steady states . Furthermore , we analyze networks with the same number of positive and negative interactions ( M+ = M− ) . Even on the level of the pair probabilities p k l ( i , j ) , the difference between positive and negative interactions is clearly seen . Fig 3 shows some examples of histograms of the corresponding relative pair abundances , p k l ( i , j ) / min ( p k ( i ) , p l ( j ) ) for the small example from Fig 1A . This systematic difference of positive and negative interactions derived from a large set of steady state composition is a key result of our investigation . The standard Jaccard index , for example , would pick up a systematic enhancement ( suppression ) of the co-occurences of 1’s ( i . e . the pair ( 1 , 1 ) ) for positive ( negative ) interactions . It is our hypothesis that the amount of change ( amount of simplification ) two vectors display under a Boolean operation ( e . g . , logical AND or logical OR ) is very different for synergistic and competitive interactions . In addition , this systematic change is quite robust against ‘cross talk’ generated by additional links and against ‘noise’ generated by measurement errors in the data . In the following , we will use the simulated data to investigate the prediction quality of such entropy shifts under increasing connectivity and noise , and benchmark it against the Jaccard index , which is a more standard analysis method of species co-abundances . We find that the entropy shift performs less well than the Jaccard index in identifying positive interactions , but substantially better in indentifying negative interactions ( Fig 4 ) . Both measures are similarly robust with respect to connectivity and random entries in the data ( noise ) . The interesting observation of a maximal prediction quality of the Jaccard index at intermediate noise levels ( Fig 4D ) might call for additional investigations . The Jaccard index is here used on the binary level as follows . For positive links , the frequency of ( 1 , 1 ) in two binarized vectors is normalized by the minimum number of 1s in each vector . For negative links , the frequency of ( 0 , 0 ) in two binarized vectors is normalized by the minimum number of 0s in each vector . The comparison with the Jaccard index only serves the purpose of showing that our assessment based on the entropy shift achieves a similar quality . The prediction quality here is defined as ( normalized ) number of correctly classified links minus the number of incorrectly classified links . A prediction quality of 0 . 5 thus means that 50 percent more links are correctly classified than incorrectly classified . The ESABO method is about the statistics of pairs of binary values . The main variant is the one , where entropy shifts under Boolean operations are evaluated . In Fig 4 this standard version is compared with a variant of the Jaccard index applied to the binary vectors ( see the Supporting Information S1 Text for the detailed definition ) . In spite of the high prediction quality obtained with the Jaccard index , the disadvantage of the ESABO version using the Jaccard index is that the thresholds for determining a positive or negative interactions are somewhat arbitrary , while in the original ESABO score ( i . e . , the z-score of entropy shifts under a Boolean AND ) the threshold has a clear interpretation as the number of standard deviations away from random data . In subsequent versions of ESABO we will study particularly , how combinations of Boolean operations and such simple indices can be employed to enhance prediction quality further . It is important to sample the system’s dynamical ‘possibility space’ ( i . e . , the set of steady states ) homogeneously . We found that a sampling according to the system’s attractor basin sizes systematically reduces the detectability of edges ( see Fig . C in S1 Text ) . In order to verify that the entropy shifts evaluated within the ESABO method are robust against a certain amount of randomness ( detection errors in microbial species ) , we introduce binary noise in the simulated data . A noise level p means that p percent of entries in a binarized abundance vector are substituted by a random choice of 0 and 1 . We observe that the prediction quality remains rather high up to noise levels of 20 percent ( p = 0 . 2; see Fig 4 ) . With the inclusion of simple binary noise we can verify that the reconstructed links are robust against detection errors in the data , an issue that can be expected to be of much higher relevance in the case of low-abundance species than in the high-abundance regime . As seen in Table 1 and Fig 2 there are occasional ‘outliers’ in the z-score distributions ( positive interactions with a large negative z-score ) . We have performed several analyses to understand , whether these outliers can be predicted from the topology of the species interaction network . So far , we have not found a topological explanation for this effect . Based on 40 random species interaction networks ( N = 15 , M+ = M− = 15 ) and 500 runs on each of the networks we estimate the number of such outliers ( z-score ≤ − 1 to be around 9 . 6 percent of all positive interactions ( see Supporting Information S1 Text for details ) . We have observed that the outliers are associated with strong compositional differences between the two binarized vectors entering the ESABO score . This point will be investigated in more detail in the future . In the previous section we have shown that the abundances and co-abundances stemming from positive and negative interactions can be detected from ESABO scores of the dynamically generated attractor states . To apply this to biological abundance data , we analyze the co-occurences on phyla level for the dataset described in subsection Study subjects and sample collection and the subsequent three Methods subsections . A binarization threshold of 1 has been used ( i . e . values of zero are mapped to zero , while all other values are mapped to one ) , as the distinction between the presence and absence of a species seems quite reliable ( see Supporting Information S1 Text ) . We observe that the pairs with highest ( and lowest ) ESABO scores are strongly symmetric , i . e . , we observe positive mutualisms or antagonims , respectively . We here compute the ESABO scores with respect to logical AND operations . A large number of z-scores is observed in the range of absolute values between −1 and 1 . 5 . We note that the overall resource competition is expected to lead to a more ubiquitious-type connectivity ( i . e . , highly clustered or even close to all-to-all coupled within the subgraph ) such that only highest z-scores are considered here . Correspondingly , the threshold for positive co-occurence has to be adjusted independently . In total , we obtain 4 competitive ( ESABO score ) resp . 8 lowly co-abundant ( by z-score ) pairs of nodes as listed in Table F and C , and a fairly more extensive list of mutualistic ( and highly co-abundant ) pairs of phyla shown in Tables D+E and F . in the Supporting Information S1 Text . From the co-abundance data and their respective ESABO scores we can extract a network of significantly mutualistic links between species ( Fig 5 ) and a corresponding network of mutually inhibiting links ( Fig 6 ) . Interestingly , the competitive and cooperative links form different networks . For competition and thus low co-occurence , the nodes are fragmented into 4 subgraphs ( see Fig 6 ) . For mutualism and thus high co-occurence , the nodes form a connected graph which contains Tenericutes , Actinobacteria , and Spirochaetes as the three nodes with highest node degree ( Fig 5 ) . The histogram of abundances for the three main hubs in the network of positive interactions , Fig 7 , illustrates that our method is sensitive to interactions among low-abundance species . As a contrast , the abundance histograms for the two dominant phyla , Bacteroidetes and Firmicutes are also shown . | Over the last years the composition of microbial communities in the human gut , the gut microbiome , has gained prominence in clinical research . Providing an estimate of the microbial interaction network from compositional data is an important prerequisite for clinical interpretation and for a better theoretical understanding of such microbial communities . Many studies have focused on the dominant interactions of species that are highly abundant such as , on the phyla level , Bacteriodetes and Firmicutes . Using binarized abundance vectors ( recording only the presence and absence of microbial species ) we show that the low-abundance segment of the microbiome also contains a large number of systematic interactions . For low-abundant species , our inference method evaluates the transformation of pairs of such vectors ‘binary co-abundance’ under Boolean operations . First we calibrate our new method using simulated data . Then we apply it to novel microbiome data from a human population study . The method reveals a large number of significant positive interactions and several significant negative interactions among low-abundance microbial species . It can be argued that important inter-individual differences and adaptations to changes in environmental conditions rather occur on the level of the low-abundance species than in the few main highly abundant species . This hypothesis could explain the broad distribution of abundances in microbiome compositions . | [
"Abstract",
"Introduction",
"Methods",
"Results"
] | [
"ecology",
"and",
"environmental",
"sciences",
"microbiome",
"metabolic",
"networks",
"microbiology",
"species",
"delimitation",
"simulation",
"and",
"modeling",
"network",
"analysis",
"speciation",
"thermodynamics",
"microbial",
"genomics",
"research",
"and",
"analysis",
... | 2017 | Boolean analysis reveals systematic interactions among low-abundance species in the human gut microbiome |
Trypanosoma brucei is the etiological agent of Human African Trypanosomiasis , an endemic parasitic disease of sub-Saharan Africa . TbCatB and rhodesain are the sole Clan CA papain-like cysteine proteases produced by the parasite during infection of the mammalian host and are implicated in the progression of disease . Of considerable interest is the exploration of these two enzymes as targets for cysteine protease inhibitors that are effective against T . brucei . We have determined , by X-ray crystallography , the first reported structure of TbCatB in complex with the cathepsin B selective inhibitor CA074 . In addition we report the structure of rhodesain in complex with the vinyl-sulfone K11002 . The mature domain of our TbCat•CA074 structure contains unique features for a cathepsin B-like enzyme including an elongated N-terminus extending 16 residues past the predicted maturation cleavage site . N-terminal Edman sequencing reveals an even longer extension than is observed amongst the ordered portions of the crystal structure . The TbCat•CA074 structure confirms that the occluding loop , which is an essential part of the substrate-binding site , creates a larger prime side pocket in the active site cleft than is found in mammalian cathepsin B-small molecule structures . Our data further highlight enhanced flexibility in the occluding loop main chain and structural deviations from mammalian cathepsin B enzymes that may affect activity and inhibitor design . Comparisons with the rhodesain•K11002 structure highlight key differences that may impact the design of cysteine protease inhibitors as anti-trypanosomal drugs .
The protozoan parasite Trypanosoma brucei is the cause of Human African Trypanosomiasis ( HAT , sleeping sickness ) in humans and nagana in domestic livestock [1] , [2] , [3] . With over 60 million people at risk and 50 , 000–70 , 000 infected , new drugs are required to control the spread of disease and associated mortality . Only four drugs are approved for treatment [4] , however , most are limited by parasite resistance [4] and marked host toxicity [5] , [6] , [7] , [8] . A new combination therapy for treating HAT has been approved recently by the World Health Organization ( WHO ) , however no newly developed drugs are on the horizon [9] , [10] Clan CA cysteine proteases play central roles during the lifecycle of many parasitic organisms [11] and have been established as effective drug targets in treating many parasitic diseases [12] , [13] , [14] , [15] . Bloodstream T . brucei parasites express two papain family cysteine proteases , rhodesain ( brucipain , trypanopain ) , a cathepsin L-like enzyme and TbCatB , a cathepsin B-like enzyme . Rhodesain is the more abundant of the two cathepsins and is required to cross the blood-brain barrier [16] . RNA interference of TbCatB is able to rescue mice from a lethal T . brucei infection [16] . RNAi knockdown of rhodesain , however , only prolongs mouse survival [17] . TbCatB may therefore represent the more promising target for novel cysteine protease inhibitors targeting T . brucei infection . A cysteine protease inhibitor , Z-Phe-Ala-CHN2 , has been shown to be lethal to T . brucei both in vitro and in vivo [18] , [19] and our efforts are focused on elucidating key features of inhibitors that will optimize both specificity and potency . Both cathepsin B-like and cathepsin L-like proteases share the common features of Clan CA cysteine proteases , including a conserved catalytic triad ( Cys/His/Asn ) and a substrate-binding site comprised of many structurally conserved residues [20] . One major difference between cathepsin L and cathepsin B enzymes is the presence of an ‘occluding loop’ of approximately 20 amino acids located on the surface of cathepsin Bs that confers an additional exopeptidase activity to these cysteine proteases [21] . To elaborate on the structural and biochemical differences between cathepsin L and cathepsin B-like cysteine proteases in T . brucei , and to aid in the design of better inhibitors , we have determined the high-resolution crystal structures of rhodesain•K11002 and the first crystal structure of T . brucei cathepsin B , TbCatB•CA074 .
Recombinant TbCatB was modified from a previously described protocol [22] . The gene encoding the full-length zymogen ( minus the N-terminal signal sequence ) was sub-cloned from a pPICZαB-TbCatB construct into the pPICZαA expression vector ( Invitrogen ) . Site-directed mutagenesis using the QuickChange system ( Stratagene ) was used to add a C-terminal His tag and to incorporate an N216D mutation ( full-length numbering ) at a predicted glycosylation site . The expression of a glycoslylation site mutant in P . pastoris is a strategy that has recently been successful in our structural studies of the homologous parasite cysteine protease cruzain [23] , [24] . Pichia pastoris strain ×33 was transformed with 20µg of BstXI linearized TbCatB-N216D according to the manufacturer's instructions ( Invitrogen ) . A single transformed colony was used to inoculate 2–3 ml of YPD-zeocin media and the culture was grown overnight at 30°C . The following day , 2 liters of YPD-zeocin media were inoculated with the starter culture and incubated at 30°C , with constant shaking at 250rpm , until cell density reached an OD600 of 3–4 ( typically 2–3 days ) . Cells were harvested at 1500g for 15min and the pellet was rinsed with 100ml BMM media and centrifuged at 1500g for 15min to remove residual YPD media . Cells were then harvested and resuspended in BMM media to an OD600 of approximately 1 . 0 . Induction of protein overexpression was carried out in a BioFlo110 Fermentor/Bioreactor ( New Brunswick Scientific ) with the addition of 1% methanol , twice a day . Supernatant was collected after 3 days incubation and concentrated to 50ml using an Ultrasette™ lab tangential flow device with a 30kDa cut-off ( Pall Corporation ) . The concentrated sample was adjusted to a final concentration of 300mM NaCl and 10mM Imidazole and incubated with 2ml of Ni-NTA beads ( Qiagen ) overnight at 4°C . The beads with the bound sample were transferred to an empty PD-10 column ( GE Healthcare ) , rinsed with 100mM phosphate pH 6 . 0 , 300mM NaCl , 10mM Imidazole , and eluted with 100mM phosphate pH 6 . 0 , 300mM NaCl and 200mM Imidazole . Eluted proteins were dialyzed against 1L buffer containing 20mM Tris-HCl pH 8 . 0 , 1mM EDTA and 5mM β-mercaptoethanol . Dialysis buffer was changed after 2 hours and continued overnight at 4°C . To produce the mature form of the protease , purified enzyme was auto-activated in a buffer containing 100mM sodium acetate pH 4 . 5 , 10mM DTT , 1mM EDTA , 100mM NaCl and 100ug/ml dextran sulfate ( MW 5000 ) . Protease activity was monitored every hour using Z-Phe-Arg-AMC as the substrate [22] , [25] , [26] . After reaching its maximum , activity was completely abolished with the addition of 10-fold molar excess CA-074 ( Sigma-Aldrich ) . The mixture was incubated overnight , with gentle stirring , to ensure complete inhibition of the activated enzyme . Inhibited enzyme was then dialyzed against 20mM Tris-HCl pH8 . 0 , 1mM EDTA and 5mM β-mercaptoethanol and applied to a Mono-Q column ( GE Healthcare ) with TbCatB•CA074 eluting at approximately 150mM NaCl ( gradient 0–1M NaCl ) at a flow rate of 1ml/min . Fractions corresponding to the protease were pooled and further purified on a Superdex 200 gel-filtration column ( GE Healthcare ) . The purified sample was concentrated to 3mg/ml and the buffer exchanged to 20mM Tris-HCl pH 8 . 0 . Three separate aliquots of TbCatB from the same batch were activated , purified and inhibited with CA074 as described above . 10–20µg of activated , inhibited protein was run on an SDS-PAGE gel and transferred on to a PVDF membrane . The blot was run for 90mins hrs at a current of 125mA . The same experiment was performed with a single purified , unactivated , sample of the enzyme . Bands containing mature and full-length TbCatB were excised from the membrane and sent for N-terminal Edman sequencing at the Protein and Nucleic Acid Facility ( PAN ) , Stanford University Medical Center ( http://cmgm . stanford . edu/pan/ ) . T . brucei were cultured in HMI-9 medium to a density of approximately 1 . 5×106 tryps per/ml . To obtain crude extracts , tryps were pelleted in 50 ml conical centrifuge tubes by centrifugation at 2500 rpm . The medium was aspirated and the pellet re-suspended in 250 of lysis buffer ( 50mM Sodium Acetate pH 5 . 5 , 1mM EDTA , 1% Tx-100 ) . The lysate was clarified by centrifugation and the protein concentration of the supernatant was measured by Bradford assay ( Bio-Rad ) . Recombinant , activated TbCatB was prepared and purified to the stage of anion exchange chromatography ( Mono Q ) , as above . The crude lysate containing native TbCatB and the activated recombinant sample were both prepared for analysis by adding 5× SDS loading buffer and boiling for 2-minutes . 40 µg of the crude lysate or 0 . 3µg of purified activate TbCatB was loaded into the well of a Novex 12-well Bis-Tris mini gel ( Invitrogen ) and resolved by SDS-PAGE at 180 volt with constant current for 1 . 5 hours . The gel was transferred onto a PVDF membrane ( Bio-Rad ) and blocked for 2 hours in buffer containing 3% milk and 0 . 5% BSA . After blocking , the blots were incubated with rabbit anti-TbCatB antiserum and diluted 1∶1 , 000 overnight at 4°C . The blots were washed 3 times for 5 minutes with TBST and then incubated at room temperature for 1 hour with goat anti-rabbit serum ( GE Healthcare ) diluted at 1∶1 , 000 in TBS . Afterwards , the blots were washed 3 times for 5 minutes with TBST and once with TBS . The immunoblots were then analyzed by ECL reagent ( GE Healthcare ) ( Figure S1 ) . Crystallization conditions were screened with a Mosquito drop-setting system ( TTP Labtech ) against a number of commercially available kits . Optimization of crystal conditions was performed manually on the basis of initial screening hits . Hanging drops of 1–2µl were set up with 3mg/ml TbCatB•CA074 , and 1M LiCl , 10% PEG 3350 , 0 . 2M Tris-HCl pH 7 . 6 . Rod-shaped crystals formed after 3 days and reached a maximum size after 5–10 days . Crystals were flash-cooled in well solution supplemented with 30% glycerol and mounted for the Stanford Auto Mounter ( SAM ) system [27] . Rhodesain was expressed in P . pastoris and purified and activated as described previously [25] , [26] , [28] with a Ser172Ala mutation incorporated to remove an N-glycosylation site from the mature domain of rhodesain . Active rhodesain was incubated with a 10-fold molar excess of the inhibitor K11002 , dissolved in DMSO . Complete inhibition of enzymatic activity was confirmed by fluorometric assay against the substrate Z-Phe-Arg-Nmec ( Bachem ) . Purified rhodesain was concentrated to approximately 8 mg/ml in preparation for crystallization . Crystals of maximum size were obtained after approximately 10 days via the sitting drop method , from a precipitating solution of 1 . 6M ammonium sulfate , 0 . 1M Bicine pH 9 . 0 at 18°C . Crystals were flash-cooled in liquid nitrogen in well solution supplemented with 20% ethylene glycol . All diffraction data were collected at the Stanford Synchrotron Radiation Lightsource ( SSRL ) . Rhodesain•K11002 data were collected to 1 . 16Å on BL9-1 after selecting an optimal crystal from screening performed with the robotic SAM system [27] . TbCatB•CA074 data were collected following a similar protocol on SSRL BL7-1 with the best crystals diffracting to 1 . 6A resolution . For both datasets , reflections were indexed and integrated in MOSFLM [29] and scaled and merged in SCALA [30] . The TbCatB•CA074 structure was solved by molecular replacement using MOLREP [31] with a homology model built by MODELLER [32] from the ensemble coordinates of human cathepsin B ( 1GMY ) , rat cathepsin B ( 1CTE ) and cruzain ( 1F2A ) . Two clear rotation function solutions were obtained in space group P21 with peak heights/sigma of 13 . 27 and 12 . 91 respectively , corresponding to two molecules in the asymmetric unit . The translation function yielded a clear solution for the dimer with a score of 0 . 53 and initial Rfactor of 54 . 6% . The structure of rhodesain•K11002 was solved by molecular replacement using PHASER [33] with a model derived from a prior structure of rhodesain bound to a different vinyl sulfone containing inhibitor ( PDB ID 2P7U ) . A single , strong solution was obtained in space group P212121 with a rotation function Z-score of 19 . 3 , a translation Z-score of 28 . 7 and an initial LLG ( log likelihood gain ) of +896 , which improved to +2022 . 7 with 6 cycles of rigid body refinement . Following rigid body and maximum likelihood restrained refinement in REFMAC5 [34] , the inhibitor molecules were placed in clear mFo-DFc difference electron density in using COOT [35] . During these initial stages of refinement the occluding loop residues in TbCatB were removed from the model and rebuilt as the difference density became clear . Both models were completed through iterative rounds of manual model building and refinement with COOT [35] and REFMAC5 [34] . TLS parameterization was used to refine the TbCatB•CA074 structure , while anisotropic temperature factors were refined in the case of rhodesain•K11002 . Water molecules were placed in each structure with COOT and manually assessed . The final rhodesain•K11002 model contains 1 molecule of rhodesain , 1 inhibitor molecule , 402 water molecules and 11 ethylene glycol molecules . The final TbCatB•CA074 model contains 2 molecules of TbCatB , 2 inhibitor molecules , 572 water molecules , 6 glycerol molecules , 1 Tris molecule , a lithium ion and a magnesium ion . Statistics for data collection and refinement are given in Table 1 . The coordinates and observed structure factors amplitudes for rhodesain•K11002 and TbCatB•CA074 have been deposited in the Protein Data Bank under accession codes 2P86 and 3HHI respectively .
The structure of rhodesain•K11002 was determined to 1 . 16Å resolution and refined to an Rfree of 13 . 0% and an Rfactor of 11 . 0% ( Figure 1 ) . The complex crystallized in spacegroup P212121 with one complete copy of the rhodesain mature catalytic domain ( residues 1–215 ) in the asymmetric unit . The amino acid residues at the beginning of mature rhodesain structure are APAA , consistent with the predicted cleavage site at the N-termini of mature rhodesain , cruzain and other cathepsin L-like proteases . We recently reported the first crystal structure of rhodesain in complex with the vinyl sulfone inhibitor K11777 ( PDB ID 2P7U ) [36] . Superimposition of this vinyl sulfone complex with the K11002 complex reported here matches 214 α-carbons with root mean square distances of 0 . 27Å . An interesting feature of the K11002 complex is the observation of a dual conformation of the phenylsulfone moiety at the P1′ position of the inhibitor ( Figure 2 ) . The optimal model to data agreement was obtained by refinement of the two conformations as a 70%/30% combination of relative occupancies . We have previously observed that this group can flip out of the S1′ pocket [36] . The TbCatB•CA074 complex crystallized in spacegroup P21 with two molecules in the asymmetric unit and was refined to 1 . 60Å resolution to a final Rfree of 17 . 8% and Rfactor of 14 . 7% ( Figure 1 ) . Chains A and B comprise residues 78–335 and 78–337 of full-length TbCatB , respectively . With the predicted maturation cleavage site of the enzyme between Pro93 and Leu94 [18] , we were surprised to observe that our crystal structure of the activated , mature form of TbCatB contains an additional 16 residues at the N-terminus preceding the predicted site of activation . In light of the non-standard start of the catalytic domain , this structure is numbered from the N-terminal signal-sequence , with methionine as residue 1 . For clarification in the text , TbCatB residues have the ‘standard’ mature domain numbering , according to human cathepsin B , in superscript . In the occluding loop of TbCatB ( Pro189105-Phe213126 ) the electron density is somewhat diffuse in parts and these more flexible regions were therefore modeled at reduced occupancy . Interestingly , in chain B ( where the electron density for the occluding loop is clear ) we observed a dual conformation for the backbone carbonyl of His194110 and the backbone amide of His195111 . A nearby electron density peak was modeled and refined as a water molecule . Although there was no significant difference density around this water when refined at full occupancy , we decided to model this position at 50% occupancy ( HOH565 ) to reflect its transient interaction with the peptide chain in this region . Refinement of this peak as a common cellular/buffer ion at full occupancy ( Li+ , Na+ , Mg2+ , Ca2+ ) yielded significant positive or negative difference density . In light of the unexpected residues observed N-terminal to the predicted maturation site , three separate aliquots of TbCatB were analyzed by Edman sequencing to determine , under the activation conditions described above , the full-length sequence of the mature domain prior to crystallization . All three reactions yielded LREAKRLNNV as the N-terminal peptide sequence . This corresponds to a peptide 61 residues into the full-length TbCatB sequence . The Asn at position 69 is Gly according to the published sequence ( Genbank accession code AAR88085 ) and we presume this to simply be an experimental error due to the low amount of sample used for the blot . Additional N-terminal sequencing studies using E64 as the inhibitor and a larger quantity of sample confirm this residue to be Gly69 ( data not shown ) . To rule out an activation event during cell culture and expression of the recombinant enzyme in P . pastoris , we also sequenced a purified , unactivated sample . This yielded the peptide EAEFALVAED; the first four residues are a cloning artifact from the 5′ end of the expression vector . The remaining six residues correspond to the expected beginning of the cloned sequence ( minus the signal sequence ) . The implications of these findings are discussed below . To determine whether an unexpected activation event occurs in-vivo , we compared the size , by Western Blot analysis , of our recombinant , purified mature TbCatB with a sample of the native protein present in crude T . brucei lysates ( Figure S1 ) . With the predicted size of mature TbCatB calculated to be 26–27kDa , the blot clearly shows that both recombinant and activated TbCatB are larger than predicted , with the recombinant form being the slightly larger of the two . While there are no modifications to prevent glycosylation of the native form , we have previously observed the size of the glycosylated and ( Endo H-treated ) deglycosylated protein to be the same ( data not shown ) .
We report the crystal structures of rhodesain•K11002 and TbCatB•CA074 , two papain family cysteine proteases implicated in the pathogenesis of Trypanosoma brucei infection ( Figure 1 ) . The structure of rhodesain•K11002 is similar to that of rhodesain•K11777 ( PDB ID 2P7U ) [36] with the bound inhibitor varying only at the P3 position ( N-methyl piperazine in K11777 , morpholino urea in K11002 ) . While a number of hydrogen bonds are formed between residues lining the substrate-binding site and the inhibitor backbone , a number of hydrophobic residues also provide binding energy , principally in the S2 subsite ( Figure 2 ) , the subsite that confers selectivity for this class of enzyme . This is in contrast with the TbCatB•CA074 complex where hydrogen bonding between the enzyme and inhibitor dominate over hydrophobic interactions . The phenylsulfone moiety at P1′ is a common motif represented in many parasite cysteine protease•vinylsulfone complexes . The dual conformation of this moiety in the rhodesain•K11002 structure is unique for a parasite cysteine protease•vinylsulfone complex and we have not observed this in other high resolution structures of rhodesain or the closely related cruzain from Trypanosoma cruzi . The TbCatB crystal structure , the first reported for this enzyme , is similar in overall structure to homologous cathepsins B-like enzymes studied ( Table S1 ) , with the majority of the variation found in the occluding loop region ( discussed below ) . Our crystal structure also reveals several interesting features that are atypical of a cathepsin B-like cysteine protease . Cathepsin B family members were originally defined in vertebrate systems as possessing an acidic residue at the bottom of the S2 subsite that allows for the accommodation of basic residues in the pocket [37] , [38] . TbCatB has a Gly at this position , which opens up the pocket allowing larger P2 substituents to be targeted to this part of the active site cleft ( Figure 3 ) . Homology modeling previously indicated an acidic functionality around the S2 subsite of TbCatB [39] , that may be able to take advantage of a positive charge at the P2 position of small molecule inhibitors . These acidic residues line the sides ( Asp16675 , Asp16877 , Asp258175 ) and bottom ( Asp327244 ) of the pocket . In our structure Asp258175 and Asp327244 are available for binding and in each copy of TbCatB interact with a glycerol molecule from the cryoprotectant solution ( Figure 4 ) . Superimposition of TbCatB•CA074 and rhodesain•K11002 highlights structural differences that cause rhodesain to be more sterically restricted at the S2 subsite ( Figure 5 ) . Firstly , Asp16675 in TbCatB is substituted for a Leu in rhodesain ( Leu67 ) ; Asp16675 in TbCatB is able to pack itself against helix α3 where it establishes a number of hydrogen bonding interactions ( Figure 5a ) . Leu67 packs more favorably against the hydrophobic environment of the rhodesain S2 subsite with the large hydrophobic phenylalanyl at the P2 position of K11002 . This residue therefore points in toward the substrate-binding site in rhodesain . Secondly , rhodesain has the larger Ala208 ( Gly328245 in TbCatB ) at the bottom of the S2 subsite , making the pocket shallower ( Figure 5b ) . Finally , the loop between strands β2 and β3 in rhodesain is anchored to an adjacent loop ( between strands β5 and β6 ) by a disulfide bridge between Cys155 and Cys203 . A number of direct and water-mediated hydrogen bonding interactions stabilize this conformation and Gln159 and Leu160 are pulled into the S2 subsite to further narrow the pocket ( Figure 5c ) . In TbCatB , the β2–β3 loop lacks the cysteine required to form the anchoring disulfide bridge , and is glycine-rich ( Gly269186 , Gly276193 , Gly280197 and Gly281198 ) when compared with rhodesain . The additional flexibility allows the C-terminal portion of the TbCatB loop to adopt a conformation similar to that found in human cathepsin B , removed from the S2 subsite and oriented towards the prime sites . Of note , the mobility of this loop was recently alluded to in homology modeling studies by Mallari et al . in comparison with human cathepsin L [40] . TbCatB has an ‘occluding loop’ , a unique feature of cathepsin B-like enzymes , which spans the prime side of the substrate binding site and distinguishes them from the cathepsin L-like enzymes [11] , [21] . In TbCatB , the loop is three residues longer than in mammalian homologs and we note a dual peptide conformation between His 194110 and His195111 . The occluding loop in TbCatB further deviates from homologous structures between residues 206120–210123 . Human , rat and bovine cathepsin B have an invariant “GEGD” motif in this region . The glycine residues flanking Glu122 confer additional flexibility in this region such that the negatively charged residue is able to flip in and out of the active site [41] ( Figure 3 ) . The corresponding motif in TbCatB , “FNFD” , lacks this flexibility and both Phe208121 and Phe210123 stack with the N-terminal residue ( Phe189105 ) of the occluding loop , creating a more stable opening around S1′ . This feature of the TbCatB occluding loop presents the possibility to engineer additional specificity into inhibitors targeting this enzyme . Indeed Mallari et al . have shown that out of a series of 56 compounds , only those with a specific N9 substituent ( hydroxypropyl ) were reasonable human CatB inhibitors . The authors propose this may be due to the ability of this substituent to stabilize the flexible loop in a favorable conformation . This stabilizing interaction was not expected to be important in TbcatB; indeed TbcatB was tolerant of a wide range of substitutions at this position on the inhibitor scaffold . An interesting aspect of mammalian cathepsin B-like enzyme structure is the presence of two salt bridges ( His110-Asp22 and Arg116-Asp224 ) that stabilize the “closed” conformation of the loop in the mature form ( Figure 6 ) . Mutations that disrupt either ion pair are correlated with a major increase in endopeptidase activity [42] , presumably due to a corresponding increase in loop flexibility . While the His-Asp pair is conserved in TbCatB , Arg116 is substituted for Tyr202 and Asp224 is substituted for Glu307 . In TbCatB , the acidic residue does not interact directly with Tyr202 , but instead stabilizes the occluding loop at an insertion ( relative to mammalian enzymes ) through an interaction with Asn200 ( Figure 6 ) . It is tempting to speculate on the role that these substitutions might play , if any , on altering the characteristic pH dependance of cathepsin B activity/inhibition . However , at present , we have no biochemical evidence to support this assumption and clearly this is a point that requires further investigation through mutational analysis . Cysteine proteases are expressed as inactive “zymogens” containing a “pro”-domain that aids in the proper folding of the full-length protein and suppresses the activity of the catalytic ( mature ) domain . Autoproteolysis results in cleavage between the pro and catalytic domains yielding the fully active , mature enzyme . Comparison of this TbCatB mature domain with crystal structures of the mature domains of mammalian cathepsin B enzymes , as well as the mature domains of rhodesain and papain , shows that the TbCatB structure has an unusually long N-terminus . However , further comparison with parasite cysteine proteases reveal that an elongated N-terminus is shared with the malarial proteases falcipain-2 ( FP-2 ) and falcipain-3 ( FP-3 ) [43] , [44] ( Figure S2A ) . Superimposition with TbCatB reveals the N-terminal extension of these proteases to be of similar length to that found in our structure ( 16 residues in falcipain-2 and 18 residues in falcipain-3 ) . Furthermore , the extension in TbCatB establishes several polar and hydrophobic interactions with the L and R domains of the main α/β fold ( Figure S2B ) , as is observed in structures of FP-2 and FP-3 ( although this results in the malarial extensions adopting more extended secondary structure ) . While comparisons can be drawn between TbCatB and the malarial proteases , the atypical N-terminus of FP-2 and FP-3 was already identified before the structures were known , including the lack of a typical papain-family mature cleavage site [43] . Conversely , TbCatB does contain such a cleavage site and , contrary to our findings , residues upstream were expected to form part of the pro-domain . Our Edman sequencing data suggest the possibility of an even longer end ( 33 residues N-terminal to the predicted cleavage site ‘LPSS’ ) . Analysis of the crystal packing in our TbCatB model suggests these residues may occupy a nearby solvent channel in the crystal and are therefore disordered . Alternatively , they may be lost during crystallization . Comparisons with the human and rat unactivated zymogens ( PDB IDs 1MIR and 3PBH ) show that , in the full-length “pro” form , the equivalent residues form a long loop and short helix that occlude the active site . The possibility of an additional 33 residues at the N-terminus of the mature TbCatB therefore remains an intriguing puzzle . While our sequencing data exclude the possibility of the recombinant enzyme being activated during yeast cell culture , we cannot exclude cellular activation of the endogenous enzyme as expressed by the native parasite . The Western Blot data show the latter to be larger upon activation than predicted by sequence analyses but slightly smaller than the recombinant form . We can only speculate that perhaps the native enzyme undergoes further processing during expression in T . brucei . Future experiments will be guided towards shedding further light on the unusual processing of this parasite cysteine protease . | Proteases are ubiquitous in all forms of life and catalyze the enzymatic degradation of proteins . These enzymes regulate and coordinate a vast number of cellular processes and are therefore essential to many organisms . While serine proteases dominate in mammals , parasitic organisms commonly rely on cysteine proteases of the Clan CA family throughout their lifecycle . Clan CA cysteine proteases are therefore regarded as promising targets for the selective design of drugs to treat parasitic diseases , such as Human African Trypanosomiasis caused by Trypanosoma brucei . The genomes of kinetoplastids such as Trypanosoma spp . and Leishmania spp . encode two Clan CA C1 family cysteine proteases and in T . brucei these are represented by rhodesain and TbCatB . We have determined three-dimensional structures of these two enzymes as part of our ongoing efforts to synthesize more effective anti-trypanosomal drugs . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biophysics/biomacromolecule-ligand",
"interactions",
"biochemistry/protein",
"chemistry",
"computational",
"biology/macromolecular",
"structure",
"analysis"
] | 2010 | Crystal Structures of TbCatB and Rhodesain, Potential Chemotherapeutic Targets and Major Cysteine Proteases of Trypanosoma brucei |
Aberrant O-glycosylation of serum immunoglobulin A1 ( IgA1 ) represents a heritable pathogenic defect in IgA nephropathy , the most common form of glomerulonephritis worldwide , but specific genetic factors involved in its determination are not known . We performed a quantitative GWAS for serum levels of galactose-deficient IgA1 ( Gd-IgA1 ) in 2 , 633 subjects of European and East Asian ancestry and discovered two genome-wide significant loci , in C1GALT1 ( rs13226913 , P = 3 . 2 x 10−11 ) and C1GALT1C1 ( rs5910940 , P = 2 . 7 x 10−8 ) . These genes encode molecular partners essential for enzymatic O-glycosylation of IgA1 . We demonstrated that these two loci explain approximately 7% of variability in circulating Gd-IgA1 in Europeans , but only 2% in East Asians . Notably , the Gd-IgA1-increasing allele of rs13226913 is common in Europeans , but rare in East Asians . Moreover , rs13226913 represents a strong cis-eQTL for C1GALT1 that encodes the key enzyme responsible for the transfer of galactose to O-linked glycans on IgA1 . By in vitro siRNA knock-down studies , we confirmed that mRNA levels of both C1GALT1 and C1GALT1C1 determine the rate of secretion of Gd-IgA1 in IgA1-producing cells . Our findings provide novel insights into the genetic regulation of O-glycosylation and are relevant not only to IgA nephropathy , but also to other complex traits associated with O-glycosylation defects , including inflammatory bowel disease , hematologic disease , and cancer .
N- and O-glycosylation are fundamental post-translational modifications of proteins in mammalian cells . Abnormalities in glycosylation have been linked to a broad range of human diseases , including neurologic disorders , immune-mediated and inflammatory diseases as well as cancer . Protein glycosylation is mediated by a large family of enzymes that have cell- and tissue-specific activity , and can generate highly diverse glycan structures that are important for signaling , cell-cell and cell-matrix interactions . The combinatorial possibilities of glycan structures imparted by the large number of glycosylation enzymes complicate a systematic analysis of protein glycosylation patterns and identification of critical steps involved in the activity , concentration , and regulation in any given cell or tissue . In such a setting , genetic studies of congenital defects of glycosylation in humans have provided significant insight into non-redundant regulatory nodes in this pathway[1] . The majority of these Mendelian disorders arise from loss of function mutations that severely perturb protein glycosylation across a range of tissues and produce a wide range of organ dysfunction in early life . However , less pronounced abnormalities in protein glycosylation have also been detected in complex disorders such as autoimmunity and cancer , suggesting that more subtle defects in this pathway can have important consequences for human health . IgA nephropathy ( IgAN ) , the most common cause of glomerulonephritis and a common cause of kidney failure worldwide , is a prototypical example of an immune-mediated disorder characterized by abnormal glycosylation[2] . In humans , the hinge-region segments of the heavy chains of immunoglobulin A1 ( IgA1 ) have 3 to 6 O-glycans , resulting in a variety of IgA1 glycoforms in circulation . In healthy individuals , the prevailing O-glycans include the N-acetylgalactosamine ( GalNAc ) -galactose disaccharide and its sialylated forms . In IgAN , galactose-deficient IgA1 ( Gd-IgA1 ) glycoforms are significantly more abundant compared to those of healthy controls[3] . These under-galactosylated glycoforms are secreted by IgA1-producing cells while galactosylation of other circulating O-glycosylated proteins is preserved , suggesting a specific defect within IgA1-producing cells[4] . The pathogenetic mechanism of IgAN involves an autoimmune response resulting in production of IgA or IgG autoantibodies against circulating Gd-IgA1 , and formation of immune complexes ( Gd-IgA1 complexed with autoantibodies ) that deposit in the kidney and cause tissue injury[2 , 5] . Consistent with this mechanism , Gd-IgA1 is the predominant glycoform in circulating immune complexes and in the glomerular immune deposits in patients with IgAN[6–9] and elevated serum levels of Gd-IgA1 ( autoantigen ) and anti-glycan antibodies ( autoantibody ) are associated with more aggressive disease and accelerated progression to end-stage kidney failure[10 , 11] . The design of a simple lectin-based ELISA assay , using a GalNAc-specific lectin from Helix aspersa ( HAA ) , enables screening of sera to quantify the levels of circulating Gd-IgA1[3] . Using this assay , we have demonstrated that the serum levels of Gd-IgA1 represent a normally distributed quantitative trait in healthy populations , but up to two thirds of IgAN patients have levels above the 95th percentile for healthy controls . Examining family members of probands with familial and sporadic forms of IgAN , we also showed that elevated serum Gd-IgA1 levels segregate independently of serum total IgA levels and have high heritability ( estimated at 50–70% ) [12 , 13] . Moreover , many healthy family members exhibited very high Gd-IgA1 levels , identifying elevated Gd-IgA1 as a heritable risk factor that precedes the development of IgAN . To date , multiethnic genome-wide association studies involving over 20 , 000 individuals have identified 15 risk loci predisposing to IgAN , highlighting the importance of innate and adaptive immunity in this disorder . Power analyses indicated that discovery of additional risk loci using the case-control design will require significant expansion in sample size . However , a systematic analysis of quantitative endophenotypes that are linked to disease pathogenesis , such as Gd-IgA1 , has not been conducted to date and may provide the opportunity to discover additional pathogenic pathways using a smaller sample size . In this study , we performed the first GWAS for serum Gd-IgA1 levels , and successfully mapped new loci with surprisingly large contributions to the heritability of the circulating level of Gd-IgA1 independently of IgA levels .
In order to test if serum levels of Gd-IgA1 remain stable over time , we first performed measurements of total serum immunoglobulin levels along with Gd-IgA1 levels at baseline and at four years of follow-up in 32 individuals of European ancestry followed longitudinally ( Fig 1 ) . While serum total IgG and IgA levels varied with time , Gd-IgA1 levels ( normalized for total IgA ) remained remarkably stable over a 4-year period ( r2 = 0 . 92 , P = 1 . 8 x 10−13 ) , demonstrating that O-glycosylation of IgA1 is minimally affected by random environmental factors . We next used HAA lectin-based ELISA to analyze single time-point sera of 1 , 195 individuals in our discovery cohorts composed of 950 individuals of East-Asian ancestry ( 483 biopsy-documented IgAN cases and 467 controls ) and 245 individuals of European ancestry ( 141 biopsy-documented IgAN cases and 104 controls , Table 1 ) . As previously reported , serum Gd-IgA1 levels were positively correlated with age ( East Asians r = 0 . 13 , P = 8 . 9x10-5; Europeans r = 0 . 15 , P = 1 . 7x10-2 ) and total IgA levels ( East Asians r = 0 . 75 , P < 2 . 2x10-16; Europeans r = 0 . 56 , P < 2 . 2x10-16 ) , but were independent of gender ( P > 0 . 05 ) . In both cohorts , Gd-IgA1 levels were also significantly higher in IgAN cases compared to controls independently of age and total IgA levels ( adjusted P < 2 . 2x10-16 in each individual cohort ) , providing a large-scale replication of prior findings . We next performed a GWAS for serum levels of Gd-IgA1 in these cohorts with and without adjustment for total IgA levels . For genome-wide analysis , we used a linear model with individual SNPs coded as additive genetic predictors , and the outcome defined as standardized residuals of serum Gd-IgA1 after normalization and additional adjustment for case/control status , age , ancestry and cohort membership ( see Methods ) . Each ethnicity-defined discovery cohort was analyzed separately and the results were meta-analyzed to prioritize top signals for follow-up . With this approach , we observed minimal genomic inflation in the combined genome-wide analyses ( λ = 1 . 01 ) , indicating negligible effect of population stratification . We first examined potential associations with known IgAN susceptibility loci , but found no statistically significant or suggestive signals between Gd-IgA1 levels and known IgAN risk alleles ( S1 Table ) . In addition , we found no association between the global polygenetic risk score for IgAN , which captures the combined effect of all IgAN risk loci , and Gd-IgA1 levels . We also did not detect any associations of Gd-IgA1 levels with loci previously linked to variation in total IgA levels[14–16] , IgA deficiency[17] , or N-glycosylation of IgG[18] . At the same time , we replicated previously reported association of total IgA with ELL2 ( rs56219066 , P = 8 . 5x10-3 ) [15] , confirming that genetic regulation of IgA levels is distinct from that for Gd-IgA1 levels . These data thus indicated the presence of yet undiscovered loci controlling variation in Gd-IgA1 levels . We next examined genome-wide distribution of P-values from the discovery stage to identify novel loci associated with Gd-IgA1 levels . Although no signal reached genome-wide significance in the discovery stage , we observed several suggestive ( P < 5x10-4 ) loci that we followed up in 1 , 438 additional individuals of East-Asian ( N = 653 ) and European ( N = 785 ) ancestry ( S1 Fig ) . Subsequently , we analyzed all cohorts ( N = 2 , 633 ) jointly to identify genome-wide significant loci ( Table 2 , S2 Table ) . Our power calculations demonstrate that our design provides adequate power to detect variants explaining ≥1 . 5% of overall trait variance at a genome-wide significant alpha 5x10-8 ( S3 Table ) . In the combined analysis , two distinct genomic loci , on chromosomes 7p21 . 3 and Xq24 , reached genome-wide significance ( Fig 2A ) . The strongest association was located within a 200-kb interval on chromosome 7p21 . 3 ( Fig 2B ) , explaining 4% of trait variance in Europeans and ~1% in East Asians ( S4 Table ) . The only gene within this locus is C1GALT1 , encoding core 1 synthase , glycoprotein-N-acetylgalactosamine 3-beta-galactosyltransferase 1 . The top signal was represented by rs13226913 ( P = 3 . 2x10-11 ) , an intronic SNP within C1GALT1 . This locus is further supported by rs1008897 ( P = 9 . 1x10-10 ) in partial LD with rs13226913 ( r2 = 0 . 33 , D’ = 0 . 91 in Europeans and r2 = 0 . 52 , D’ = 0 . 73 in East Asians ) . After mutual conditioning , both SNPs continue to be associated with the phenotype , suggesting a complex pattern of association at this locus ( S5 Table ) . The protein encoded by C1GALT1 generates the common core 1 O-glycan structure by transferring galactose ( Gal ) from UDP-Gal to GalNAc-alpha-1-Ser/Thr . Core 1 O-glycans are the main glycans in the hinge region of circulating IgA1 , as well as precursors of many extended mucin-type O-glycans on cell-surface and secreted glycoproteins . In humans , C1GALT1 is abundantly expressed in IgA1-secreting cells[19] , as well as in EBV-transformed lymphocytes , gastrointestinal tract , lungs , and kidneys[20] . The top SNP , rs13226913 , is not in LD with any coding variant , but it perfectly tags several SNPs intersecting the ENCODE and Roadmap enhancers and promoters in immune cells , including EBV-immortalized B cells and primary CD19+ cells ( S6 Table ) . Interrogation of eQTL databases revealed that rs13226913 has a highly significant cis-eQTL effect on C1GALT1 in peripheral blood cells ( P = 3 . 9 x 10−23 ) with the T allele associated with lower mRNA levels ( S7 Table ) . Consistent with this finding , rs13226913 imparts an additive effect with each T ( derived ) allele increasing Gd-IgA1 levels by 0 . 22 standard deviation units ( 95% CI: 0 . 10–0 . 30 ) . The second genome-wide significant locus comprises a 100-kb interval on chromosome Xq24 ( Fig 2C ) and explains an additional 2 . 7% of the overall trait variance in Europeans and 1 . 2% in East Asians ( S4 Table ) . The top signal at this locus is represented by rs5910940 ( P = 2 . 7x10-8 ) , a SNP 3’ downstream from C1GALT1C1 . The T ( derived ) allele increases serum Gd-IgA1 levels by 0 . 14 standard deviation units per allele ( 95%CI: 0 . 11–0 . 17 ) . Our post-hoc examination of genotypic effects suggests a dominant effect of the rs5910940-T allele in females ( dominant model P = 7 . 9x10-9 , S8 Table ) , although skewed inactivation of chromosome X in IgA1-producing cells could also potentially explain this effect . C1GALT1C1 encodes a transmembrane protein that is similar to the core 1 beta1 , 3-galactosyltransferase 1 encoded by C1GALT1 . However , its gene product ( known as COSMC ) lacks the galactosyltransferase activity , and instead acts as a molecular chaperone required for the folding , stability , and full activity of C1GALT1[21] . C1GALT1C1 is also ubiquitously expressed in multiple tissues , including IgA1-secreting cells[19] , other blood cells , gastrointestinal tract , kidneys , and lungs[20] . Because sex chromosomes are not included in most eQTL analyses , we were not able to confirm if rs5910940 has an effect on the expression of C1GALT1C1 based on available datasets . However , rs5910940 tags a 2-bp insertion in the active promoter of C1GALT1C1 in B-lymphocytes and leukemia cell lines ( S9 Table ) . Considering the known functional dependency of C1GALT1 and C1GALT1C1 , we also tested for potential epistasis between these two loci , but did not detect any significant genetic interaction . Taken together , these data predict an additive regulatory effect of rs13226913 and rs5910940 , resulting in lower C1GALT1 and C1GALT1C1 expression , and leading to increased production of Gd-IgA1 . We next performed siRNA knock-down studies in human cultured IgA1-secreting cell lines to confirm the effect of lower C1GALT1 and C1GALT1C1 transcript abundance on the production of Gd-IgA1 ( Fig 3 ) . Consistent with the observed genetic effect , in vitro knock-down of C1GALT1 resulted in 30–50% increased production of Gd-IgA1 by the cells derived from IgAN patients ( P = 0 . 025 ) as well as from healthy controls ( P = 0 . 011 ) . Similar to C1GALT1 , in vitro siRNA knock-down of C1GALT1C1 in IgA1-producing cell lines significantly increased the production of Gd-IgA1 in healthy individuals ( P = 0 . 032 ) and a similar trend was observed in IgAN patients ( P = 0 . 066 , Fig 3 ) . Consistent with the genetic data , there was no multiplicative effect on Gd-IgA1 production with combined siRNA knock-down in IgA1-secreting lines . Jointly , the newly discovered C1GALT1 and C1GALTC1 loci explain up to 7% of variance in serum Gd-IgA1 levels in Europeans and 2% in East Asians ( S4 Table ) . Further examination of effect estimates by ethnicity confirms that the European cohorts predominantly drive these associations ( S10 Table ) . Notably , the derived ( T ) allele of rs13226913 at C1GALT1 locus is considerably more frequent in Europeans ( freq . 47% ) compared to East Asians ( freq . 10% ) , additionally contributing to the difference in variance explained between ethnicities . Subsequent examination of allelic frequencies in the Human Genome Diversity Panel ( Fig 4 ) confirms that the derived allele of rs13226913 is rare or absent in some Asian populations , while being the predominant ( major ) allele in Europeans ( freq . ≥50% ) . In contrast , the T ( derived ) allele of rs5910940 at C1GALT1C1 locus is equally frequent in Asian and European populations ( freq . ~50% ) , but nearly fixed in selected African populations . These findings suggest potential involvement of geographically confined selective pressures acting on the loci controlling the O-glycosylation process . Lastly , we detected additional suggestive signals , including a locus on chromosome 7p13 that warrants further follow-up in larger cohorts ( S2A and S2B Fig ) . This locus is represented by rs978056 ( P = 3 . 3x10-5 ) , an intronic SNP in HECW1 ( encoding E3 ubiquitin ligase ) previously studied in the context of colon and breast cancer ( S3 Fig ) . Based on the analysis of known protein-protein interactions , HECW1 is a second-degree neighbor of C1GALT1 and COSMC , with ubiquitin C as a common interacting protein ( S2C Fig ) .
Genetic studies of immune endophenotypes have provided novel insights into the genetic architecture of complex traits and enhanced sub-classification of several autoimmune and inflammatory disorders . The power of immune endophenotypes is best exemplified by recent genetic studies of ANCA titers in vasculitis[22] , IgE levels in asthma[23 , 24] , and studies of IgG N-glycosylation and autoimmunity[18] . Taking a similar approach , we performed the first GWAS for aberrant O-glycosylation of IgA1 . Abnormalities in the O-glycan synthesis have been linked to several human diseases , including IgAN , inflammatory bowel disease ( IBD ) , hematologic diseases , and cancer . Dense O-glycosylation of various mucins produced by epithelial cells is critical for the formation of a protective viscous barrier with anti-microbial properties at the mucosal surfaces of the gastrointestinal , urogenital and respiratory systems . Recent studies indicate that proper O-glycosylation of mucins is required for intestinal integrity in mice[25 , 26] and may play a role in human susceptibility to IBD[27 , 28] . In addition , O-glycosylation can affect the structure and immunogenicity of the modified proteins . For example , defective O-glycosylation represents the key pathogenic feature of Tn syndrome[29] , where acquired enzymatic defect in the addition of galactose to O-glycans leads to exposed terminal GalNAc residue ( Tn antigen ) on the surface of red blood cells , triggering polyagglutination by naturally occurring anti-Tn antibodies[29] . Moreover , Tn and sialyl-Tn represent oncofetal antigens that are over-expressed in human cancers and may directly influence cancer growth , metastasis and survival , but the exact molecular perturbations that lead to O-glycosylation defects in tumor cells are presently not known[30] . Similar to Tn syndrome , the pathogenesis of IgAN involves autoimmune response to Tn antigens . In this case , the Tn antigen is exposed at the hinge region of IgA1 molecules as a result of aberrant O-glycosylation of IgA1 in the Golgi apparatus of IgA1-producing cells[9] . In patients with IgAN , the galactose-deficient IgA1 ( Gd-IgA1 ) is recognized by circulating anti-Tn autoantibodies [5] , leading to the formation of nephritogenic immune complexes[6–9] . Several independent studies , including in healthy twins and in families with IgAN , have demonstrated that serum levels of Gd-IgA1 have high heritability , providing high level of support for a genetic determination of this trait and a strong rationale for this study[12 , 13 , 31] . In this study , we quantified the levels of Gd-IgA1 in sera of 2 , 633 subjects of European and East-Asian ancestry using a simple lectin-based ELISA assay . Using GWAS approach , we discovered two genome-wide significant loci , on chromosomes 7p21 . 3 and Xq24 , both with large effects on circulating levels of Gd-IgA1 . The 7p21 . 3 locus contains C1GALT1 gene , that encodes human core 1 β1–3-galactosyltransferase ( C1GALT1 ) , the key enzyme responsible for the addition of galactose to the Tn antigen . Mice deficient in C1GALT1 protein develop thrombocytopenia and kidney disease attributed to defective O-glycosylation of cell-surface proteins[32] . Moreover , C1GALT1 deficiency in mice results in a defective mucus layer , leading to spontaneous colonic inflammation that is dependent on the exposure to intestinal microbiota[25 , 26] . C1GALT1 requires a molecular chaperone , COSMC , that ensures the enzyme is properly folded within the endoplasmic reticulum; loss of COSMC activity results in C1GALT1 being degraded in the proteosome[29] . Interestingly , COSMC is encoded by C1GALT1C1 residing within our second genome-wide significant locus on chromosome Xq24 . We also localized a suggestive locus on chromosome 7p13 that encodes an E3 ubiquitin ligase , but it is presently not known if this protein participates in the proteosomal degradation of C1GALT1 . This signal will require further follow up . Importantly , our study demonstrates that there are several common genetic variants with relatively large effects on IgA1 O-glycosylation . These effects are conveyed by different genes , but converge on a single enzymatic step in the O-glycosylation pathway . While we successfully identified two novel loci for serum Gd-IgA1 levels , several important limitations of our study design need to be acknowledged . First , our GWAS has a two-stage design and involves bi-ethnic cohorts . Although to date this is the largest study of individuals worldwide with measured serum Gd-IgA1 levels , this sample size is still not adequate to detect ethnicity-specific loci . Thus , our design is presently limited to the discovery of alleles that have similar effects in both Europeans and East Asians . At the same time , the bi-ethnic composition of our cohorts clearly enabled identification of the C1GALT1 locus . The lead allele at this locus has a direction-consistent effect in both ethnicities , but because of the Gd-IgA1-increasing allele is relatively rare or even fixed in some Asian populations , this signal would have been missed if the discovery were performed entirely in East Asians . The second limitation relates to genomic resolution of our discovery study . Although genome-wide imputation was not performed at the time the study was conducted , post-hoc imputation using the latest 1000 Genomes reference revealed no additional loci outside of the regions that were originally selected for follow-up . Moreover , our conditional analyses revealed no additional signals among the imputed SNPs after controlling for the lead alleles at each locus , suggesting that our top SNPs captured most of the signal at the newly reported loci ( S4 Fig ) . Given the observed distributional differences in serum Gd-IgA1 levels between cases and controls ( S5 Fig ) , we estimate that we would require a sample size of at least 24 , 000 cases and 24 , 000 controls to detect the effect of C1GALT1 and C1GALT1C1 loci in a bi-ethnic GWAS for IgAN ( S11 Table ) [33] . This sample size requirement is more than 3-fold greater than the largest bi-ethnic GWAS for IgAN published to date[34] . Moreover , considering weaker effect of these loci in East Asians , an even larger sample size ( over 40 , 000 cases and 40 , 000 controls ) would be required for a GWAS involving only East-Asian participants ( 5-fold greater than the largest published study for Chinese[35] ) . Yet , our endophenotype-based approach uncovered these loci in a minute fraction of the sample size required by a conventional case-control designs . Our power calculations also clearly indicate that much larger follow-up studies will be needed to conclusively demonstrate that Gd-IgA1-increasing alleles have a direct effect on the disease risk . In summary , our results contribute new insights into the genetic regulation of O-glycan synthesis , and demonstrate that a simple lectin-based assay can be used effectively to map genetic regulators of O-glycosylation of serum proteins . Given the high heritability of this trait , it is likely that additional loci contribute to variation in Gd-IgA1 levels . In particular , the inheritance pattern in IgAN kindreds suggested segregation of a major dominant gene , indicating a potential role of additional rare alleles with large effects[12] . A search in larger population-based studies that includes both common and rare variants is likely to uncover additional genetic determinants of O-glycosylation defects and elucidate mechanisms leading to IgAN and related disorders .
This study has a two-stage design ( S1 Fig ) . In Stage 1 ( the discovery phase ) we performed a genome-wide meta-analysis of two discovery cohorts: the Chinese cohort of 950 individuals ( 483 cases and 467 controls , all Han Chinese ancestry , genotyped with Illumina 660-quad chip ) , and the US cohort of 245 individuals ( 141 cases and 104 controls , all European ancestry , genotyped with the Illumina 550v3 chip ) . Genome-wide scan was performed in both cohorts and fixed-effects meta-analysis was applied to prioritize signals for follow-up studies . In Stage 2 we performed targeted genotyping of the top signals from Stage 1 in five cohorts of European and East-Asian ancestry ( 1 , 438 individuals in total , Table 1 ) . We estimated the power of our study design for a range of effect sizes under the following assumptions: standard normal trait distribution , additive risk model , no heterogeneity in association , marker allelic frequency of 0 . 25 ( average MAF for the microarrays used ) , a follow-up significance threshold of P<5×10−4 , and a combined significance level of P<5×10−8 . These calculations demonstrate that we have adequate power to detect variants explaining ≥1 . 5% of overall trait variance ( S3 Table ) . Our study was conducted according to the principles expressed in the Declaration of Helsinki; all subjects provided informed consent to participate in genetic studies , and the Institutional Review Board of Columbia University as well as local ethics review committees for each of the individual cohorts approved our study protocol . The serum level of total IgA was determined using standard ELISA[36] . The serum level of Gd-IgA1 was determined using a custom HAA-based ELISA assay[12 , 13 , 36] . This method relies on the detection of HAA binding to desialylated galactose-deficient glycans ( Tn antigens ) of serum IgA1 immunocaptured on ELISA plates . Because in humans , IgA1 , but not IgA2 , has O-glycans , this assay effectively quantifies the serum level of Gd-IgA1 in units/ml . We have optimized this assay for high-throughput . Briefly , 96-well plates were coated with F ( ab’ ) 2 fragment of goat IgG anti-human IgA at 3 μg/ml . Plates were blocked with 1% BSA in PBS containing 0 . 05% Tween 20 , and serial two-fold dilutions of samples and standards in blocking solution were incubated overnight at room temperature . To remove terminal sialic acid , the samples were treated with 100 μL ( 1 mU ) per well of neuraminidase ( Roche ) in 10 mM sodium acetate buffer ( pH = 5 ) for 3 h at 37°C . Next , the samples were incubated with GalNAc-specific biotinylated HAA lectin ( Sigma-Aldrich ) for 3 h at 37°C . The bound lectin was detected with avidin-horseradish peroxidase conjugate , followed by the peroxidase substrate , o-phenylenediamine-H2O2 ( Sigma ) ; the reaction was stopped with 1 M sulfuric acid . The concentration of Gd-IgA1 was calculated by interpolating the optical densities at 490 nm on calibration curves constructed using a myeloma Gd-IgA1 standard . The intra-assay coefficients of variation ( CVs ) for calibration curves , plotted by a 4-parameter model , ranged from 2–10% for the extremes of the curves and 1–5% in the middle region . If higher values were noted , the samples were re-analyzed . The inter-assay CV was also consistently under 5% and our prior studies demonstrated excellent reproducibility of this assay[36] . In the final analysis , we applied a correction for potential plate effects using the same replicate samples across all plates . After corrections , serum Gd-IgA1 levels for each cohort were tested for normality by the Shapiro-Wilk test , assisted by visual inspection of histograms and QQ-plots . Non-normal trait distributions were transformed using logarithmic transformation . The log-transformed traits were regressed against age and case-control status to derive standardized residuals . Summary statistics ( mean , SD , skewness , and kurtosis ) were derived for the distribution of standardized residuals , that were then used as a quantitative trait in GWAS analysis . Summary statistics , normality testing , transformations , plots , and regression analyses were performed with R 3 . 0 software package ( CRAN ) . The discovery cohorts have been published , including details of the genotyping , genotype quality control , and ancestry analyses[34 , 37] . Briefly , we implemented strict quality control analyses for each of the discovery cohorts , removing individual samples with low call rates , duplicates and samples with cryptic relatedness ( pi-hat > 0 . 10 ) , ancestry outliers , and samples with a detected sex mismatch . After all quality-control steps , the Chinese Discovery Cohort was composed of 950 individuals typed with 508 , 112 SNPs , while the US Discovery Cohort was composed of 245 individuals typed with 531 , 778 SNPs . In total , 468 , 781 SNPs overlapped between the cohorts , and this set of common markers was used for the discovery meta-analysis . To reduce any potential bias from population structure , we used modified PCA-based ancestry matching algorithms ( Spectral-GEM software ) [38 , 39] , as described in our prior studies of these cohorts[34 , 37] . Primary association testing for the Gd-IgA1 phenotype ( expressed as standardized residuals ) was performed for each individual cohort under an additive linear model in PLINK[40] . We included significant principal components of ancestry as covariates in linear models used for association testing . Additionally , we performed regression analyses with and without adjustment for serum total IgA levels . We derived adjusted effect estimates with standard errors for each SNP , and we combined these results using an inverse variance-weighted method ( METAL software ) [41] . We visually examined genome-wide distributions of P-values using QQ-plots for each individual cohort , as well as for the joint analysis of both cohorts . We estimated the genomic inflation factors[42] , that were negligible for each individual discovery cohort ( lambda = 1 . 011 and 1 . 013 for the Chinese and US cohorts , respectively ) . The overall genomic inflation factor was estimated at 1 . 010 and the final meta-analysis QQ-plots showed no global deviation from the expected distribution of P-values ( S1 Fig ) . We next prioritized the top 50 SNPs for replication among the top suggestive SNPs with P<5x10-4 from the GWAS analyses . First , we clustered the top signals into distinct loci based on their genomic coordinates and metrics of LD . Conditional regression analysis was carried out to detect independent association between signals within the same genomic regions . For genotyping in replication cohorts , we prioritized the independent SNPs that had the lowest P-value at each genomic locus . In addition , we required that each SNP be successfully genotyped in both discovery cohorts . We excluded 'singleton signals' defined as loci supported by only a single SNP in the absence of supporting signals with P<0 . 01 within the same LD block . If the genotyping assay failed for the top SNP , a back-up SNP was selected on the basis of its strength of phenotypic association , LD with the top SNP , genotyping quality , and ability to successfully design working primers . Moreover , we added SNPs for which the signals became more significant ( P<5x10-4 ) after adjustment for serum total IgA levels . In all , we successfully genotyped 50 carefully selected SNPs in 1 , 438 independent replication samples across five cohorts . Similar to our prior studies , the genotyping of replication cohorts was performed using KASP ( Kompetitive Alelle Specific PCR , LGC Genomics ) . In our prior studies , this technology had >99 . 8% accuracy rates[43] . Table 1 summarizes the ethnic composition of our replication cohorts along with the genotyping method and average genotype call rates . We first carried out association analyses individually within each of the cohorts using the same methods as in the discovery study . Next , we combined the results using a fixed-effects model ( S2 Table ) . For each of the genotyped SNPs , we derived pooled effect estimates and their 95% confidence intervals . To declare genome-wide significance , we used the generally accepted threshold of P<5x10-8 , initially proposed for Europeans genotyped with high-density platforms based on extrapolation to infinite marker density[44] . We performed two types of association tests for X-linked markers . Our primary association test involved sex-stratified meta-analysis of chromosome X markers: each male and female sub-cohort was analyzed separately and the association statistics were combined across all sub-cohorts using fixed effects meta-analysis . This approach is not affected by the type of allele coding in males and allows for different effect size estimates between males and females[27] . In secondary analyses , we assumed complete X-inactivation in females and a similar effect size between males and females . In this test , females are considered to have 0 , 1 , or 2 copies of an allele as in an autosomal analysis while males are considered to have 0 or 2 copies of the same allele ( i . e . , male hemizygotes are equivalent to female homozygotes ) . The main limitation of this approach relates the assumption of complete X inactivation . Because approximately 15–25% of X-linked genes escape inactivation in female-derived fibroblasts[45] and chromosome X inactivation has not been studied in IgA1-secreting cells , this analysis was performed only on an exploratory basis , but the results were consistent with sex-stratified analyses . For the genome-wide significant loci , we explored two alternative genetic models ( dominant and recessive ) and compared these models using Bayesian Information Criterion ( S8 Table ) . We also tested for all pairwise genetic interactions between the suggestive and significant loci using two different tests . First , we used a 1-degree of freedom likelihood ratio test to compare two nested linear regression models: the model with main effects only versus the model with main effects plus additive interaction terms . Second , a more general 4-degree of freedom genotypic interaction test was performed . In this test , we compared a model with allelic effects , dominant effects , and their interaction terms with a reduced model without any of the interaction terms . All models were stratified by sex and cohort . The analyses were performed in R 3 . 0 software package ( CRAN ) . To interrogate any potential SNPs that were not directly typed in our dataset , we downloaded the latest release of the 1000 Genomes ( Phase 3 ) and imputed our discovery cohorts using ethnicity-specific reference panels . The haplotypes were phased using Markov Chain Haplotyping software ( MACH ) and the imputations were carried out with Minimac3 . For downstream analyses , we applied strict quality control filters post-imputation , including only markers that were either genotyped or imputed with high confidence ( R2 ≥ 0 . 8 ) . Association testing of imputed SNPs was performed assuming an underlying additive linear model and including cohort-specific significant principal components as covariates . Primary analysis was performed using a dosage association method in PLINK , that accounts for uncertainty in prediction of the imputed data by weighting genotypes by their posterior probabilities . We used a similar approach to perform conditional analyses across the top loci , with conditioning SNPs added as additional covariates in linear models . Using the imputed results for the C1GALT1 , C1GALT1C1 , and HECW1 regions , we examined all of the top most associated variants as well as all SNPs in LD with the lead SNP ( r2>0 . 5 ) at each locus . We annotated these variants using ANNOVAR[46] , SeattleSeq[47] , SNPNexus[48] , FunciSNP[49] , HaploReg4[50] , and ChroMos[51] . The transcripts whose expressions were correlated with the lead SNPs in cis- or trans- were also identified using available eQTL datasets , including: ( 1 ) peripheral blood eQTLs based on meta-analysis of 5 , 311 Europeans[52] , ( 2 ) primary B-cell and monocyte eQTLs from 288 Europeans[53] , and ( 3 ) the latest release of GTEx data across multiple tissue types[20 , 54] . We utilized , automated MEDLINE text mining tools to assess network connectivity between genes residing in implicated GWAS loci , including GRAIL[55] , e-LiSe[56] , and FACTA+[57] . We also interrogated all known protein-protein interaction networks for connectivity between candidate genes using the Disease Association Protein-Protein Link Evaluator ( DAPPLE ) [58] and Protein Interaction Network Analysis platform ( PINA2 ) [59] . We used Cytoscape v . 2 . 8 to visualize network graphs . IgA1-secreting cell lines from five patients with IgAN and five healthy controls were transfected using ON-TARGETplus SMARTpool siRNAs ( Thermo Fisher Scientific ) specific for human C1GALT1 , COSMC , or both . The ON-TARGETplus Non-targeting Pool siRNAs was used as a control . We followed our previously published protocol for Amaxa nucleofector II ( Lonza ) [60] . Twenty-four hours after transfection , the knock-down efficiency was determined by qRT-PCR with previously described primers[9 , 60] . The knockdown was expressed as cDNA level of the individual gene normalized to GAPDH after respective siRNA treatment , divided by the respective value obtained after treatment by non-targeting siRNA . The effect of siRNA knock-down on the phenotype ( the degree of galactose-deficiency of IgA1 ) was based on the reactivity of secreted IgA1 with a lectin from Helix aspersa specific for terminal GalNAc , as described[9 , 60] . | O-glycosylation is a common type of post-translational modification of proteins; specific abnormalities in the mechanism of O-glycosylation have been implicated in cancer , inflammatory and blood diseases . However , the molecular basis of abnormal O-glycosylation in these complex disorders is not known . We studied the genetic basis of defective O-glycosylation of serum immunoglobulin A1 ( IgA1 ) , that represents the key pathogenic defect in IgA nephropathy , the most common form of primary glomerulonephritis worldwide . We report our results of the first genome-wide association study for this trait using serum assays in 2 , 633 individuals of European and East-Asian ancestry . In our genome scan , we observed two significant signals with large effects , on chromosomes 7p21 . 3 and Xq24 , jointly explaining about 7% of trait variability . These signals implicate two genes that encode molecular partners essential for enzymatic O-glycosylation of IgA1 and mucins , and represent potential new targets for therapy . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"genome-wide",
"association",
"studies",
"gene",
"regulation",
"mathematics",
"statistics",
"(mathematics)",
"genome",
"analysis",
"molecular",
"biology",
"techniques",
"genotyping",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"genomic",
"sig... | 2017 | GWAS for serum galactose-deficient IgA1 implicates critical genes of the O-glycosylation pathway |
Sri Lanka’s Anti Filariasis Campaign distributed 5 rounds of mass drug administration ( MDA with DEC plus albendazole ) to all endemic regions in the country from 2002–2006 . Post-MDA surveillance results have generally been encouraging . However , recent studies have documented low level persistence of Wuchereria bancrofti in Galle district based on comprehensive surveys that include molecular xenomonitoring ( MX , detection of filarial DNA in mosquitoes ) results . The purposes of this study were to demonstrate the use of MX in large evaluation units ( EUs ) and to field test different mosquito sampling schemes . Galle district ( population 1 . 1 million ) was divided into two EUs . These included a coastal EU with known persistent LF and an inland EU with little persistent LF . Mosquitoes were systematically sampled from ~300 trap locations in 30 randomly selected clusters ( health administrative units ) per EU . Approximately 28 , 000 Culex quinquefasciatus were collected with gravid traps and tested for filarial DNA by qPCR . 92/625 pools ( 14 . 7% ) from the coastal EU and 8/583 pools ( 1 . 4% ) from the inland EU were positive for filarial DNA . Maximum likelihood estimates ( MLE ) for filarial DNA rates were essentially the same when the same number of mosquito pools were collected and tested from 75 , 150 , or 300 trap sites ( range 0 . 61–0 . 78% for the coastal EU and 0 . 04–0 . 07% for the inland EU ) . The ability to use a smaller number of trap sites reduces the cost and time required for mosquito sampling . These results suggest there is widespread persistence of W . bancrofti infection in the coastal Galle EU 8 years after the last round of MDA in 2006 , and this is consistent with other data from the district . This study has shown that MX can be used by national programs to assess and map the persistence of W . bancrofti at the level of large EUs in areas with Culex transmission .
Lymphatic filariasis ( LF ) is a disfiguring and disabling disease that affects approximately 120 million people in 73 countries . The World Health Organization ( WHO ) initiated a Global Programme to Eliminate Lymphatic Filariasis ( GPELF ) in 2000 that is largely based on repeated annual cycles of mass drug administration ( MDA ) with albendazole together with either ivermectin ( in Africa ) or diethylcarbamazine ( outside of Africa ) to reduce infection rates and interrupt transmission . GPELF aims to eliminate LF as a public health problem in all countries by the year 2020 . With a target population of 1 . 4 billion , this is easily the largest disease intervention program initiated to date based on MDA; more than 6 billion medication doses were distributed to some 600 million people between 2000 and 2014 [1] . MDA has dramatically reduced LF infection rates and prevented new cases in many countries [2 , 3] . However , more work is needed to develop and test methods for determining whether W . bancrofti has been eliminated from countries and regions following MDA . The WHO now recommends use of transmission assessment surveys ( TAS ) to demonstrate interruption of transmission . TAS surveys use systematic sampling protocol to test 1st and 2nd grade primary school children for filarial antigenemia , and they are powered to provide 95% certainty that infection rates in children are less than 2% . While this is a useful surveillance tool , a recent study showed that antibody testing of sentinel populations and molecular xenomonitoring ( MX , detection of filarial DNA in vector mosquitoes by PCR ) were more sensitive than TAS for detecting persistent W . bancrofti in Sri Lanka [4] . MX relies on the ability of mosquitoes to efficiently take up filarial parasites from human blood , and parasite DNA is detected in mosquito pools by PCR . Although MX has been used extensively in recent years to demonstrate the impact of MDA on W . bancrofti in communities [4–8] , no national LF elimination program has adopted MX as a routine method for post-MDA surveillance . Sri Lanka was one of the first countries to implement a LF elimination program based on WHO guidelines . The Sri Lankan Ministry of Health’s Anti Filariasis Campaign ( AFC ) distributed 5 annual rounds of MDA with albendazole plus diethylcarbamazine to all endemic regions of the country between 2002 and 2006 [9] . Various types of surveillance have been conducted since the MDA program ended in 2006 . Post-MDA surveillance results ( based on detection of parasites in human blood by microscopy ) have consistently shown microfilaremia rates lower than the target value of 1% in all sentinel and spot check sites . However , recent studies have provided evidence of persistent LF infection and transmission in several districts , and this is especially worrisome in the Southern district of Galle . Galle District , with a population of approximately 1 . 06 million , is located about 125 km south of the country’s capital in Colombo . The district area is about 1 , 652 km2 ( 636 square miles ) with 73 km of coastline . Ecological conditions ( annual rainfall 200–250 cm , mean temperature 25–28 C , and relative humidity levels often >80% ) are favorable for mosquitoes . Survey data before and after the MDA program have consistently shown higher LF rates in coastal areas that have low-lying low plains , low hills , extensive surface water , and higher human population densities than in inland areas of the district that have steep hills and better drainage . It is well documented that Culex quinquefasciatus is the major W . bancrofti transmission vector in Sri Lanka [10–13] , and the traditional filariasis belt extends along the coast from Puttalam in the West around to Matara in the South . Some of these areas have poorly drained polluted water , latrine catch pits , coconut husk pits , and rice fields that are primary breeding areas for Culex and other types of mosquitoes [14] . A devastating Tsunami in 2004 in this coastal belt affected the topography in some areas in ways that may have increased mosquito densities and disease transmission . However , there are no mapping data available on distribution of breeding sites for C . quinquefasciatus in Sri Lanka or in Galle district . The central Ministry of Health and Nutrition ( MOH ) divides districts into progressively smaller health administrative units . Galle is divided into large Ministry of Health areas ( MOOH ) with populations in the range of 22 , 000–95 , 000 . Each MOOH area is comprised of Public Health Inspector ( PHI ) areas , and these sub-district health administration units are comprised of smaller Public Health Midwife ( PHM ) areas . Galle had high rates of LF prior to the national LF elimination program , and government surveys have shown that sentinel sites in this district have consistently had higher microfilaria ( Mf ) rates than most sites in other districts in the years since the MDA program ended . Although the AFC used the district as an implementation unit for MDA , Galle was divided into two evaluation units ( a high-risk coastal EU and low-risk inland EU ) for monitoring and evaluation . The AFC conducted transmission assessment surveys ( TAS ) in these EUs in 2012–2013 according to WHO guidelines , and both EUs easily passed the WHO threshold . However , recent surveys in two public health inspector ( PHI ) areas in Galle district showed high rates of filarial DNA in Culex quinquefasciatus and high rates of anti-filarial antibodies in primary school children [4] . These results suggested that there are hotspots with persistent W . bancrofti transmission in Galle . The current study was conducted to test a scheme for systematic sampling of mosquitoes so that MX could complement TAS for assessing residual filariasis activity by national filariasis elimination programs . The study also aimed to map areas with persistent W . bancrofti in Galle district and to compare MX results with results from an extensive population survey for microfilaremia that was conducted by the AFC in 2013 . While MX has been previously used to assess the impact of MDA on persistent W . bancrofti in populations , this study reports the first time that it has been used by a national LF elimination program to detect persistent W . bancrofti at the scale of large EUs .
We obtained consent from household members to place CDC gravid traps on their property . The Microfilaria surveys were conducted as a public health activity by the Anti Filariasis Campaign , Sri Lanka Ministry of Health . Written consent was obtained from all adults . Participation of children required written consent from at least one parent or guardian plus assent by the child . Unique identifiers of human participants were not used in this study . Galle district is divided into 19 Medical Officers of Health ( MOOH ) divisions , and each of these are comprised of smaller health administrative units called Public Health Midwife ( PHM ) areas that were used as evaluation areas ( EAs ) for the cluster surveys described in this paper . There are 340 PHM areas in Galle district with a mean population of approximately 3 , 000 ( range 669–8025 ) . The district was divided into two evaluation units ( EU ) for post-MDA surveillance . These included a high risk coastal EU with 210 PHMs in 11 MOOH areas and a low risk inland EU with 126 PHMs in 8 MOOH areas . Three of the 11 high risk MOOH areas are not located near the coast , but in this study they were considered to be part of the high risk coastal EU based on historical LF data , detection of Mf positive cases in routine surveys , and because of they are bordering areas with high Mf rates . In the 2013 census , approximately 24 , 600 households were listed in the coastal EU and 17 , 400 households were in the inland EU . The mosquito surveys used cluster sampling based on systematic selection of households ( HH ) to assess prevalence of Wuchereria bancrofti DNA in C . quinquefasciatus species [15] . The study was designed to collect and test 300 pools of 25 mosquitoes per pool collected from 300 , 150 , and 75 HH locations . PHM maps , census data and voter registries were used to identify the HH locations in each of these areas . Thirty PHMs were randomly selected as evaluation areas ( EAs ) in each of the two EUs studied using Survey Sample Builder ( SSB ) software ( http://www . ntdsupport . org/resources/transmission-assessment-survey-sample-builder ) . The sampling interval was calculated by dividing the estimated number of HH in each EU by the number of HH that were needed for mosquito trap placement in the EU ( 300 ) . The starting HH for each PHM was selected at random from the census list , and other HH for trap placement in that PHM were selected using the sampling interval . Subsets of 150 and 75 trapping sites were randomly chosen from the pool of 300 per EU for collection of 2 and 4 mosquito pools per trap . To allow for anticipated 10% HH refusal rate for permitting mosquito trap placement , trap locations were increased to 320 , 156 , 76 or 81 , respectively . The mosquito surveys were performed between December 2013 and September 2014 . The long duration was due to administrative issues and not because of time requirements for the work . Mosquitoes were collected with CDC gravid traps ( Model 1712 , John W . Hock Company , Gainesville , FL ) with liquid bait that attracts Cx . quinquefasciatus [4 , 16] . The traps were placed outdoors in shaded areas adjacent to houses after obtaining consent from the residents . Traps were set to collect mosquitoes from dusk to dawn for 1 to 3 nights . A second trap was placed next to selected houses in some locations in order to collect the required number of mosquitoes . This was commonly done when four pools were required from a trapping location . Traps that did not yield enough mosquitoes to form pools were moved to a new location . Mosquitoes were collected from an average of ten ( range 6–19 ) locations in each of the 60 PHM areas . Culex quinquefasciatus were manually sorted to select only gravid , semigravid or blood fed females and pooled in groups of up to 25 . Pools were placed in 2 ml seal-rite tubes and dried as previously described [4] . Tubes with mosquitoes were labeled with barcode stickers and these numbers were scanned into cellphones and associated with GPS coordinates . The dried mosquitoes were transferred to the AFC central laboratory in Colombo for DNA isolation and qPCR testing . Genomic DNA extraction of pools of mosquitoes ( 1–25 ) was performed as previously described [17] . The extracted DNA samples in 200 μl of Diethylpyrocarbonate ( DEPC ) treated water was stored at -20 C in 1 . 5 ml sterile polypropylene tubes with barcode stickers . W . bancrofti DNA was detected in mosquito DNA samples that were tested in duplicate by quantitative PCR ( qPCR ) in microtiter plates with an Applied Biosystems 7300 Real Time PCR System ( Life technologies , California , USA ) as previously described [18 , 19] . Positive and negative control samples were included in each qPCR assay . Mosquito DNA samples with borderline Ct values ( ≥38 ) and those with inconsistent results in duplicate wells were retested . Samples with Ct values >38 were considered to be negative . All DNA extractions and qPCR assays were performed in AFC laboratories in Colombo . The AFC conducted a large scale night blood smear survey between March and August , 2013 . Households were selected for the survey using the WHO guidelines [20] based on the probability proportional to estimated size ( PPES ) of the population with the goal of testing approximately 3% of the population ( 38 , 000 people drawn from all PHMs in the district ) . A maximum of 4 people per household were sampled between the ages of 2 and 70 years . Two to four teams collected blood in night blood surveys that started no earlier than 20:30 hr . Each team included a public health inspector , a blood collector , a helper , and a supervisor . It took six months to complete the night blood surveys , but additional time was required for processing of blood smears and microscopy . Microfilaria ( Mf ) testing was performed with a measured volume of 60 μl of night blood . Finger prick blood samples were collected with One Touch Ultra Soft lancet holders with sterile , single use lancets ( LifeScan , Inc . , Milpitas , CA ) . Two-spot blood smears of 30 μl each were stained with Giemsa , and Mf were detected by microscopy . PHM numbers , household locations , trap numbers , pool numbers , and pool PCR test results were entered into Motorola ( Motorola Solutions , Inc . , Schaumburg , IL ) Blur phones using preloaded survey forms with a LINKS data collection platform ( http://www . linkssystem . org/ ) . Specimens and laboratory qPCR test results were linked to trap locations using barcode stickers ( Partnered Print Solutions , Dacula , GA ) . Data were downloaded as Microsoft Excel files ( Microsoft Corporation , Redmond , WA ) for analysis at AFC and at Washington University . Demographic information and Mf smear results were entered onto paper forms for the Mf surveys , and this information was later transferred into Excel spreadsheets ( Microsoft Corporation , Redmond , WA ) . GPS coordinates for mosquito sampling sites were obtained with cellphones . Coordinates were plotted using ArcGIS 10 . 2 . 1 ( ESRI , Redlands , CA ) . qPCR results were expressed as the percentages of positive pools and trap sites by PHM and EU . Filarial DNA rates in mosquitoes ( maximum likelihood estimates with 95% confidence intervals ) were calculated with PoolScreen 2 . 0 . 3 software as previously described [21 , 22] . Separate PoolScreen estimates ( maximum likelihood estimates or MLE with 95% confidence intervals ) were calculated for approximately 300 pools from 300 trap sites , 300 pools from 150 trap sites , and 300 pools from 75 trap sites from 30 PHMs ( evaluation areas ) in each of the two EUs . In order to assess filarial DNA rates in mosquitoes collected in Galle district as a single EU , the merged results from 2 EUs were analyzed by calculating mosquito DNA rates in 30 PHMs randomly selected from the 60 PHMs in two EUs that were sampled in the study . This process was repeated 30 times to assess variability in estimates obtained with different PHM samples . Results were analyzed by ANOVA and the significance of differences was assessed by the Tukey method . Data analysis was performed with Statistical Analysis Software ( SAS , version 9 . 2 , SAS Institute Inc . Cary , North Carolina ) . Some figures were produced with Graphpad Prism 6 ( GraphPad Software , Inc . , La Jolla , CA ) . We obtained consent from household members to place CDC gravid traps on their property . The Mf surveys were conducted as a public health activity by the Anti Filariasis Campaign , Sri Lanka Ministry of Health . Written consent was obtained from all adults . Participation of children required written consent from at least one parent or guardian plus assent by the child . Unique identifiers of human participants were not used in this study . However , AFC followed up to treat Mf carriers identified during night blood surveys with anti-filarial medications according to WHO guidelines .
Approximately 28 , 700 blood fed , gravid or semigravid C . quinquefasciatus mosquitoes were collected from more than 600 trap locations in 60 PHMs ( 30 in each EU ) between December 2013 and September 2014 ( Table 1 ) . Mosquitoes were sorted into 1208 pools with 1–25 mosquitoes per pool ( see Tables 1 and 2 ) . Each pool contained mosquitoes from a single trap location . While 89 . 5% of pools contained 25 mosquitoes ( 1081/1208 ) , some pools were smaller , because few mosquitoes were trapped in some collection sites despite placement of traps for up to 3 nights . Indeed , zero mosquitoes were collected in 45 trap locations . Factors such as higher altitude , very effective water drainage , and paradoxically , excessive rain affected collections at these locations . qPCR results by EU are summarized in Table 1 . Combined results from both EUs showed that 100 of 1208 pools ( 8% ) were positive for filarial DNA . 16/30 PHM areas in the coastal EU and 3/30 ( 10% ) in the inland EU had at least one positive pool ( range 1–16 ) . Filarial DNA rates in mosquitoes exceeded provisional targets for filarial DNA [4 , 19] of 0 . 25% for maximum likelihood and 1% for the upper confidence limit of the estimate in 12 of 30 PHMs in the coastal EU and in 4 of 30 PHMs in the inland EU ( Fig 1 ) . Thus W . bancrofti DNA rates were high in most areas within the coastal EU and much lower in the inland EU . Some of the inland PHMs with positive qPCR results were adjacent to areas in the coastal EU with high filarial DNA rates in mosquitoes . Figs 2 and 3 provide information on the general location of mosquitoes with filarial DNA in the PHM areas sampled in this study . Filarial DNA was detected in mosquitoes in 22% ( 70/317 ) of trap locations in the coastal EU and in 3% ( 8/317 ) of trap locations in the inland EU . Although the number of mosquitoes collected from each trap location varied because of the sampling scheme , Fig 3 shows that mosquitoes with filarial DNA were widely distributed in coastal areas . Table 2 shows that similar results were obtained when similar numbers of mosquitoes were tested in 300 pools that were collected from approximately 75 , 150 , or 300 sampling locations with 4 , 2 or 1 pool per site , respectively . Mosquito DNA rates were also similar when different numbers of collection sites were used in the inland EU , although it was not possible to obtain enough mosquitoes for 4 pools from some collection sites . Table 3 shows results that would have been obtained if Galle district had been considered as a single EU rather than as two EUs . The results shown are the mean and SD of 30 separate randomly selected samples of 30 EAs drawn from the entire district . The results show that the estimates were essentially the same whether mosquitoes were sampled from 316 , 153 , or 70 trap locations . More importantly , the results show that the parasite DNA rate in the combined EU was lower and intermediate between the rates observed in the coastal and inland EUs . Although the filarial DNA rate in mosquitoes from the district wide EU exceeded the provisional target of 0 . 25% , the diluting effect of results from the inland areas is clear ( the MLE for the coastal EU and for the entire district were 0 . 67% vs . 0 . 35% , respectively ) . This emphasizes the importance of creating EU’s so that transmission risks within them are as uniform as practicable [23] . Night blood testing was performed with 38 , 065 samples collected from all PHMs and all MOOH areas in Galle district . Only 52 slides were positive for an overall rate of 0 . 14% ( Table 4 ) . 12 out of 19 MOOH areas recorded at least 1 Mf-positive subject ( range 1–22 ) , and the mean ( SD ) Mf count in positive slides was ( 17 ± 30; range 1–177 ) . As expected , Mf rates were much higher in MOOH areas in the coastal EU than in the inland EU , and 47 of 52 ( 90 . 4% ) Mf positives identified in the district-wide survey lived in coastal areas . The Balapitiya MOOH area in the coastal EU had the highest number of Mf-positive subjects and the highest Mf rate ( 0 . 9% ) ; 22 of 52 ( 42% ) Mf positive subjects were from this MOOH area . However , Mf rates were much higher ( range ≥1%-4 . 64% ) in 12 of the 340 PHMs surveyed . These were Talpe ( 1 . 39% ) , Panagamuwa ( 1 . 72% ) , Paragahathota ( 4 . 64% ) , Galmangoda ( 3 . 45% ) , Balapitiya ( 4 . 42% ) , Randombe ( 1 . 73% ) , Wellabada ( 1 . 12% ) , Dalawella ( 1 . 01% ) , Maliduwua ( 1 . 41% ) , Kumbhalwella-2 ( 2 . 75% ) , Aluthwatta ( 1 . 26% ) , and Mawita ( 1 . 13% ) , respectively . Ten of the PHMs with Mf rates >1% were in the coastal EU and two were in the inland EU . Table 4 shows Mf survey results and MX results by MOOH area . Note that all PHM areas were sampled in the Mf survey while MX was performed in only a few PHM areas per MOOH area . Despite this difference , the MX and Mf data both indicated widespread persistence of LF in the coastal EU with much lower rates in the inland EU . The two MOOH areas with the highest Mf rates ( Balapitiya and Galle ) also had the highest MX signals ( % mosquito pools positive for filarial DNA ) . However , not all areas had consistent results , and three MOOH areas ( Ambalangoda , Hikkaduwa , and Habaraduwa ) with very high MX positivity had low Mf rates . This may have been due to small samples in the Mf surveys or to differences in the areas sampled , as mentioned above . Two PHM areas ( in Imaduwa and Yakkalamulla MOOHs in the inland EU ) also had mosquitoes that were positive for filarial DNA ( 7% and 2% pools positive , Fig 1 ) and the same MOOH areas had Mf carriers detected by night blood surveys ( Mf rate 0 . 75% ) ( Table 4 ) . On the other hand , no positive mosquito pools were detected in other PHMs that were located in MOOH areas in the inland EU where Mf carriers were identified ( Table 4 ) . Again , this may be because not all PHMs were sampled for MX , and because persistent filarial infections following MDA can be highly focal . Table 5 summarizes Mf and MX results for 15 PHM areas in 6 coastal MOOH areas that were tested by both survey methods . These results illustrate the dramatic heterogeneity of persistent W . bancrofti in Galle district . They also show that MX surveys based on relatively small numbers of mosquito pools are more sensitive for detecting persistent W . bancrofti infections than Mf testing with sample sizes in the range of 91–267 per PHM area . Only 5 of 15 PHMs had positive signals by Mf testing while 10 of 15 PHMs had positive mosquito pools by qPCR . On the other hand , one PHM with no positive mosquito pool ( out of 17 tested ) had a positive Mf signal with 1 positive smear out of 101 tested . Fig 3 shows the location of gravid traps that yielded mosquitoes that were tested for filarial DNA by qPCR . Mosquitoes with filarial DNA were widely distributed in coastal areas and also present in a few inland areas . S1 Fig shows a more detailed map for MX results from three PHM areas that were tested in the Balapitiya MOOH area . Filarial DNA was detected in mosquitoes collected in 58% of trapping sites in Balapitiya . These results show that MX is useful for identifying areas with persistent filariasis , and this information could be useful for targeting additional measures to interrupt transmission .
The Sri Lanka Anti Filariasis Campaign has succeeded in most formerly endemic areas , and the program has fulfilled many of the criteria required for validation according to WHO guidelines . A prior study compared several different types of tests for detecting persistent W . bancrofti at the level of Public Health Inspector areas with populations of ~25 , 000 . This study has used MX and Mf testing to document persistence of W . bancrofti in EUs with populations of 500 , 000 or more . It is well known that the last mile is often the most difficult for elimination programs . The last 5% is as important as the first 95% when programs are in eradication mode [35] . We have shown that MX can be used as a sensitive method for detecting and mapping areas with persistent W . bancrofti . This type of information could be very helpful for planning interventions to carry programs across the last mile to the finish line . | Lymphatic filariasis ( LF ) is a disabling neglected tropical disease that affects millions of people in 73 countries . The Global Program to Eliminate Lymphatic Filariasis ( GPELF ) is using mass drug administration ( MDA ) to reduce infections in humans below levels required for sustained transmission by mosquitoes . Prior studies have shown that molecular xenomonitoring ( MX , detection of filarial DNA in insect vectors ) is useful for detecting low-level persistence of W . bancrofti infection following MDA , but technical challenges have prevented national programs from adopting this approach on a large scale . This paper reports the first use of MX for post-MDA surveillance in large evaluation units . CDC gravid traps were effective for collecting large numbers of Culex quinquefasciatus mosquitoes , and filarial DNA was detected by qPCR . We found that MX is more sensitive than other surveillance methods for detecting persistent W . bancrofti in communities and that there is widespread , low-level persistence of infection in coastal Galle district ( Sri Lanka ) 8 years after the last round of MDA . The systematic mosquito sampling protocol used in this study should be feasible for use by national filariasis elimination programs to assess and map the persistence of W . bancrofti in regions where the infection is transmitted by Culex mosquitoes . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"geographical",
"locations",
"parasitic",
"diseases",
"animals",
"filariasis",
"molecular",
"biology",
"techniques",
"insect",
"vectors",
"research",
"and",
"analysis",
"methods",
"public",
"and... | 2016 | Programmatic Use of Molecular Xenomonitoring at the Level of Evaluation Units to Assess Persistence of Lymphatic Filariasis in Sri Lanka |
Almost immediately after a human being is born , so too is a new microbial ecosystem , one that resides in that person's gastrointestinal tract . Although it is a universal and integral part of human biology , the temporal progression of this process , the sources of the microbes that make up the ecosystem , how and why it varies from one infant to another , and how the composition of this ecosystem influences human physiology , development , and disease are still poorly understood . As a step toward systematically investigating these questions , we designed a microarray to detect and quantitate the small subunit ribosomal RNA ( SSU rRNA ) gene sequences of most currently recognized species and taxonomic groups of bacteria . We used this microarray , along with sequencing of cloned libraries of PCR-amplified SSU rDNA , to profile the microbial communities in an average of 26 stool samples each from 14 healthy , full-term human infants , including a pair of dizygotic twins , beginning with the first stool after birth and continuing at defined intervals throughout the first year of life . To investigate possible origins of the infant microbiota , we also profiled vaginal and milk samples from most of the mothers , and stool samples from all of the mothers , most of the fathers , and two siblings . The composition and temporal patterns of the microbial communities varied widely from baby to baby . Despite considerable temporal variation , the distinct features of each baby's microbial community were recognizable for intervals of weeks to months . The strikingly parallel temporal patterns of the twins suggested that incidental environmental exposures play a major role in determining the distinctive characteristics of the microbial community in each baby . By the end of the first year of life , the idiosyncratic microbial ecosystems in each baby , although still distinct , had converged toward a profile characteristic of the adult gastrointestinal tract .
The adult human body typically comprises ten times more microbial cells than human cells , due largely to the extremely high density of microbes found in the human intestinal tract ( typically 1011–1012 microbes/ml of luminal content ) . This microbial ecosystem serves numerous important functions for its human host , including protection against pathogens , nutrient processing , stimulation of angiogenesis , and regulation of host fat storage [1–7] . It is clear that this list is not yet complete; as this field of study expands , we are continually discovering new roles and relationships . Studies of gnotobiotic mice have been particularly enlightening , illustrating the essential role of the gastrointestinal ( GI ) microbiota in normal gut development [2 , 5] . In addition , numerous diseases in both adults and infants have known or suspected links to the GI microbiota , including stomach cancer [8] , mucosa-associated lymphoid tissue lymphoma [9] , inflammatory bowel disease [10 , 11] , and necrotizing enterocolitis [12 , 13] . The composition of the adult GI microbiota has been intensely studied , using both cultivation and , more recently , culture-independent , small subunit ( SSU ) ribosomal DNA ( rDNA ) sequence-based methods [14] . The human colon ecosystem alone has been estimated to contain more than 400 bacterial species , belonging to a limited number of broad taxonomic divisions [15] . Members of the anaerobic genera Bacteroides , Eubacterium , Clostridium , Ruminococcus , and Faecalibacterium have typically been found to comprise a large majority of the human adult gut microbial community . Still , each adult's gut appears to have a unique microbial community , with a structure that remains stable on the time scale of months [3 , 15 , 16] . In contrast , the infant GI microbiota is more variable in its composition and less stable over time . In the first year of life , the infant intestinal tract progresses from sterility to extremely dense colonization , ending with a mixture of microbes that is broadly very similar to that found in the adult intestine [17] . Although the beginning and end points of this time course are well defined , the path between these points is poorly understood . There are conflicting reports in the literature regarding the composition of the neonatal GI microbiota and the factors that shape it . Several studies have reported that Bifidobacteria almost always dominate the GI microbiota of breast-fed infants by several weeks of age [17–20] , while others find that they occur in only a small fraction of infants , or are not numerically dominant [21 , 22] . The effect of diet on the composition of the infant GI microbiota is also controversial—numerous studies have found a lower abundance of Bifidobacteria and a higher abundance of aerobic bacteria in the GI microbiota of formula-fed infants relative to breast-fed infants [20 , 21 , 23–25] , yet other reports have found no such difference [26 , 27] . Mode of delivery has frequently been cited as one of the key factors that shape the infant microbiota [18 , 28 , 29] . The GI microbiota of infants delivered by caesarean section has been reported to differ from that of infants delivered vaginally , both in the timing of colonization and in composition [18 , 30–32] , and in some cases , there are clearly traces of the maternal vaginal microbiota in the neonatal GI microbiota [33] , yet the relative importance of mode of delivery on GI microbiota is unclear . Because of the increased incidence of GI problems in premature infants , the effect of gestational age has also been extensively studied . These studies have consistently shown that the microbiota of hospitalized , preterm infants differs from that of healthy , full-term babies [32 , 34–36] . Attempts to associate specific microbes with the occurrence of necrotizing enterocolitis , a condition with suspected bacterial etiology that is an important cause of morbidity and mortality in premature babies , have yielded mixed results [32 , 36] . Clearly , there is still much to be learned about the origins and development of the infant GI microbiota and its influence on health and disease . We focused our study on describing the range of profiles that constitute a healthy infant GI microbiota in the hopes of discovering themes that govern its development , and in order to provide a detailed reference and a solid foundation for later studies examining the factors that influence the GI microbiota . Our study participants included 14 healthy , full-term babies , born to 13 healthy mothers ( thus including one set of fraternal twins ) ( Table 1 ) . Stool samples were collected according to a prescribed schedule , beginning with the first stool produced after birth: samples were collected daily at first and then with decreasing frequency over the course of 1 y , with additional sampling around key events ( e . g . , introduction of solid food and administration of antibiotics ) , yeilding an average of 26 stool samples per baby ( Table 2 ) . In addition , stool samples were collected from parents and siblings , and vaginal swabs and breast milk were collected from the mothers . We analyzed the microbiota of each of these specimens using a newly developed SSU rDNA microarray designed to give nearly comprehensive coverage of known SSU rDNA species . A subset of these samples was also analyzed by SSU rDNA clone library sequencing , for the purposes of calibrating and validating our microarray results .
To survey the composition of our sample set and to provide a basis for quantitative calibration of the microarray results , we created a reference pool by combining equal amounts of amplified SSU rDNA from each PCR-amplifiable sample ( except for samples collected when the infants were ≥1 y old ) . We obtained 3 , 458 high-quality clone sequences from a library constructed from this pool , and taxonomically assigned each sequence using Ribosomal Database Project's Classifier [37] . The taxonomic distribution of these sequences is summarized in Table 3 . To assess the performance of our new microarray design relative to SSU rDNA sequencing , we sequenced SSU rDNAs amplified from each of 12 individual biological samples obtained in this study , selected for their diverse profiles by 16S rDNA microarray analysis . This study set included DNA extracted from eight baby stools , two maternal stools , one vaginal swab , and one breast milk sample . For each of these samples , we amplified SSU rDNA sequences using the same PCR primers that were used in the microarray analysis , then cloned and sequenced several hundred ( mean = 342 ) of the amplified products for a total of 4 , 100 sequences . We focused our comparison at levels 2 , 3 , and 4 of the prokaryotic multiple sequence alignment ( prokMSA ) hierarchy , which very roughly correspond to the phylum , class , and order levels in the classical taxonomic hierarchy . At these broader levels , most sequences are expected to have homology to at least one probe in our current microarray design , and rDNA sequences can generally be unambiguously classified . Microarray-based relative abundance estimates were obtained for 2 , 149 species and taxonomic groups by integrating data from all probes that represented any subset of the class in question , as fully described in Materials and Methods . Sequence-based estimates were obtained by taxonomically classifying each sequence by assigning the prokMSA operational taxonomic unit ( OTU ) code of the best BLAST match in the 2004 prokMSA database of 86 , 453 SSU ribosomal RNA ( rRNA ) gene sequences [38] ( Datasets S1 and S2 ) . Although the relative abundance of a bacterial species cannot be precisely determined from its proportional representation in a pool of amplified rDNA sequences , we expect that such estimates should be accurate within an order of magnitude and usually within a few-fold [39–41] , based on previous studies that compared abundance levels estimated from sequencing SSU rDNA amplicons with counts based on in situ hybridization . Overall , the microarray results were very similar to those obtained by sequencing , both qualitatively and quantitatively . Figure 1A shows the comparison of the community profiles of each of the 12 samples derived from our microarray analysis and by sequencing , for each taxonomic group at level 2 of the prokMSA taxonomic tree . Note that the levels ( e . g . , level 2 ) in the prokMSA taxonomy do not have a consistent correspondence with the levels ( e . g . , phylum ) in the classical taxonomic hierarchy , and thus some of the conventional names associated with prokMSA level 2 groups can appear somewhat incongruous . Both the sequence analysis and the microarray analysis showed that the samples were dominated by a limited number of taxonomic groups—99% of the 4 , 100 sequences were encompassed by just three of the 22 level 2 prokMSA divisions: 2 . 15 ( Flexibacter-Cytophaga-Bacteroides ) , 2 . 28 ( Proteobacteria ) , and 2 . 30 ( Gram-positive bacteria [including Firmicutes and Actinobacteria] ) , and the remaining 1% belonged to groups 2 . 10 ( Prosthecobacter ) , 2 . 29 ( Fusobacteria ) , or 2 . 21 ( Cyanobacteria and Chloroplasts ) . As shown in Figure 1B and 1C , the population profiles obtained by microarray and sequencing analysis were also quantitatively similar—the Pearson correlation of the microarray- and sequencing-based estimates of relative abundance for the 12 samples was 0 . 97 at prokMSA taxonomic level 2 ( Figure 1B ) , 0 . 89 at level 3 ( Figure 1C ) , and 0 . 80 at level 4 ( unpublished data ) . We estimated the overall density of bacteria in each sample by a real-time quantitative PCR ( qPCR ) assay , using a broad-range bacterial primer and probe set ( see Materials and Methods ) . We used the total number of rRNA gene copies ( typically about five per genome [42] ) per gram of stool , as estimated by this assay , to approximate the total density of bacteria . As shown in Figure 2 , the total number of rRNA gene copies was relatively unstable throughout the first week of life , then persisted in most babies in the range of 109 to 1010/g of stool ( wet weight ) . Although there was no clear effect of method of delivery on the timing of the colonization , it is noteworthy that babies 13 and 14 ( the dizygotic twins ) , who were the only babies delivered by a planned caesarean section , and thus without rupture of the amniotic membrane and exposure to maternal birth canal microbiota during labor or delivery , had low bacterial counts ( <108 rRNA gene copies/g ) until the seventh day of life . We analyzed the bacterial composition of 430 samples—363 infant stool samples , 43 adult stool samples , two sibling stool samples , 12 breast milk samples , and ten maternal vaginal swabs—by hybridization to the DNA microarray developed in this study . By combining information across multiple probes ( see Materials and Methods ) , we obtained relative abundance estimates for 2 , 149 nested taxonomic groups and species in each of these samples ( All probes are listed in Dataset S3; All taxa are listed in Dataset S4 ) . As shown in Figure 3 , the phylum-level diversity in the stool samples analyzed in this study was extremely limited . The vast majority of samples were dominated by just three of the 22 level 2 bacterial groups represented by our microarray: 2 . 15 ( Flexibacter-Cytophaga-Bacteroides ) , 2 . 28 ( Proteobacteria ) , and 2 . 30 ( Gram-Positive Bacteria [Firmicutes and Actinobacteria] ) . A second major finding was the remarkable degree of interindividual variation in the colonization process . Although the taxa that populate the infant GI tract were limited at the broadest levels , each baby was distinct in the combination of microbial species that it acquired and maintained , and in the precise temporal pattern in which those species appeared and disappeared . Bacteroides , for example , dominated the early microbiota of some babies but were virtually absent at this stage in other babies . A third striking feature of this dataset was the relative stability of the microbial populations over time—even early in the course of the colonization of the infant GI tract , most taxonomic groups persisted over intervals of weeks to months . The main dimensions of variation among the colonization profiles of different taxonomic groups were timing of colonization and temporal stability . Consistent with previous studies [28 , 35 , 43 , 44] , the earliest colonizers were often organisms predicted to be aerobes ( e . g . , Staphylococcus , Streptococcus , and Enterobacteria ) , whereas the later colonizers tended to be strict anaerobes ( Eubacteria , and Clostridia ) . The Bacteroides varied greatly from baby to baby in the timing of their first appearance , but were consistently present to some degree in nearly all babies by 1 y . Several other taxa , including Prevotella , Acinetobacter , Desulfovibrio , Veillonella , and Clostridium perfringens , tended to appear only transiently , sometimes appearing and disappearing repeatedly within a baby's first year of life . We explored the similarities and differences in the composition of all of our samples by hierarchically clustering the 430 samples based on their similarity with respect to their abundance profiles for the set of 53 prokMSA level 4 taxonomic groups that had at least two samples with a relative abundance estimate greater than 1% . The clustering pattern , as reflected in the dendrogram at the top of Figure 4 , highlights several critical features of the colonization program , and shows that the stool microbiota of babies 1 y of age and older is distinctly different from that at earlier ages and much more similar to that of adults . Prior to 6 mo of age , stool samples tended to cluster by baby , indicating that the differences from baby to baby are much greater than the changes over periods of weeks or months in the composition of any individual baby's microbiota . There were two notable exceptions to this baby-specific clustering . First , samples from the first few days of life often clustered away from the rest of a given baby's samples , sometimes clustering with other very early samples and sometimes with samples from other sites ( e . g . , baby 8 day 1 with vaginal samples ) . Second , samples from babies 13 and 14 , who are fraternal twins , tended to intermingle . Figure 4B shows examples of several of the clustering patterns described above . Most of the breast milk and maternal vaginal samples clustered perfectly by anatomic site of origin . As expected , all but one of the vaginal samples were overwhelmingly dominated by lactobacilli , with Staphylococci , Bacteroides , Clostridia , and Veillonella among the groups variably present as minority constituents . The vaginal sample from one of the mothers ( mother of baby 11 ) had a distinctly different population profile , dominated instead by members of the Gamma Proteobacteria group . The microbial populations found in the milk samples were diverse , often including mixtures of enterics and species of Bacteroides , Pseudomonas , Haemophilus , Veillonella , and Streptococcus . In order to compare the infants more systematically , we determined the nearest-neighbor sample for each sample as measured by the Pearson correlation of level 4 relative abundance estimates . Using this metric , the nearest-neighbor sample of any given baby sample was usually another sample from the same baby—the average percentage of samples from a given baby for which the most similar sample was from the same baby was 82% . Figure 5 summarizes this analysis and illustrates the interesting finding that by this measure , the most similar pair of babies by far was babies 13 and 14—fraternal twins raised in the same environment—8 of 23 ( 35% ) of baby 13′s nearest-neighbor samples were from baby 14 ( the next most similar pair was babies 11 and 14 , at 17% ) . The similarity of the microbial community profiles of stool samples from babies 1 y and older to each other and to those of the adult stool samples suggested that the infant GI communities converged over time toward a generalized “adult-like” microbiota . We explored this phenomenon by calculating , for each age interval , the average pairwise Pearson correlation of the population profiles of all infant samples collected at that age . As shown in Figure 6A , this analysis revealed that as time progressed , the babies' microbiota consistently converged toward a common profile . We also calculated , for each time point , the average correlation of infant samples at that time point to a generalized adult profile ( centroid of 18 adult samples—nine fathers and nine mothers from this study ) . This analysis , shown in Figure 6B , confirmed that the profile toward which the infants' microbiota converges is similar to that of adults , and highlighted an apparent tendency for a population rearrangement to occur around 5 d after birth . Notably , the infants' GI microbiota was not significantly more similar to that of their parents than to that of other adults , as measured by the Pearson correlations of their level 4 taxonomic profiles ( mean baby–parent correlation of 0 . 55 for within family , versus 0 . 62 between families for nine “triads” of contemporaneously obtained samples from baby , mother , and father obtained at 1–1 . 5 y of age ) . To visualize the temporal patterns in the particular phylogenetic groups that populate the infant gut , we charted the relative abundance of the nine level 4 taxonomic groups that had a mean relative abundance of 1% or greater over time in each infant ( Figure 7 ) . This analysis enabled us to identify common themes and interesting differences among the colonization profiles of these babies . First , we observed that “uneven” populations ( populations heavily dominated by a single taxonomic group ) were common in the first several weeks but rare later in the time courses . Another notable feature in the temporal program of many of the babies was the occurrence of one or more dramatic shifts in the population structure—such shifts were frequently stabilized within one sampling interval . We were unable to identify any specific age or signal event consistently associated with such transitions , although the transition to an “adult-like” profile often followed the introduction of solid foods . Several of the babies were treated with antibiotics either in the neonatal period ( day 0–28 ) or in the later months ( see Table 1 and Figure 2 for more details ) . In some cases , the treatment was associated with a striking alteration in the density or composition of the GI microbiota . For example , baby 8 received two courses of amoxicillin , one at 4 mo and one at 6 mo . In both cases , both the total density of bacteria ( Figure 2 ) and the community composition were dramatically altered ( Figures 3 and 6 ) . Indeed , in this baby , the bacterial density in fecal samples decreased so much during the antibiotic courses that we were unable to amplify sufficient SSU rDNA for microarray analysis , so we could only compare the populations before and after the antibiotic course . However , we did not identify any consistent consequences of antibiotic treatment . The results of both the sequence analysis of the reference pool and the microarray data analyses indicated that Bifidobacteria were only minor components of the population—a result at odds with the conventional wisdom [20 , 21 , 26] . The primers we used for broad-range PCR amplification of the reference pool ( the source of the sequences ) and samples for microarray analysis were potentially suboptimal for amplification of Bifidobacteria [21 , 26] due to three mismatches in the rDNA sequence of Bifidobacterium longum to the forward primer 8F used in this study . A survey of the 5′ sequences of full-length SSU rDNA genes showed that Bifidobacteria are outliers in their divergence from the generally conserved 8F primer sequence . We therefore carried out two independent analyses to determine whether and how the quantitative estimates of Bifidobacteria from the microarray hybridization results would need to be adjusted . First , we quantitatively evaluated the relative efficiency with which the 8F/1391R primer pair amplified SSU rDNA from two Bifidobacteria species—Bifidobacterium longum and Bifidobacterium infantis—compared to a set of three diverse common fecal bacteria—Escherichia coli , Clostridium perfringens , Bacteroides fragilis—all of which have SSU rDNA sequences with one or more mismatches to the 8F/1391R PCR primer sequences . Using a range of stoichiometric mixtures of chromosomal DNA extracted from these species , we found that after 20 cycles ( the number of cycles used for our microarray analyses and for amplification of the reference pool prior to sequencing ) , efficiency of amplification of the two Bifidobacterial species' DNA was consistently 8-fold lower than that of the three other species , all of which amplified with nearly identical efficiencies ( unpublished data ) . This result suggests that both the reference pool sequencing results and the microarray-based quantitation underestimated the abundance of the Bifidobacteria group by a factor of eight . Second , we used a real-time qPCR assay with a primer pair and probe optimized for detection of Bifidobacteria to obtain an independent estimate of the abundance of Bifidobacteria in each sample . The results confirmed the finding from the microarray analysis that Bifidobacteria were almost always only minor constituents of the fecal microbiota of both the infants and adults in our study population ( Dataset S5 and Figure S1 ) . The majority of bacterial species identified in our sample set were previously reported constituents of the human GI microbiota . There were , however , a number of cases in which the microarray results indicated the presence of a bacterial species or group that was both unexpected and not represented in our sequenced reference pool . We investigated several of these cases using independent assays . For 12 of the suspect species/taxa , we used the cognate group-specific primers in a PCR assay applied to most or all of the samples in which the suspect species/taxa appeared to be present based on the microarray results , as well as a small set of samples in which the suspect species was not detected by the microarray . In one case , that of Sutterella wadsworthia , sequencing of the species-specific PCR product confirmed its presence . In seven of 12 cases , none of the array positive ( or negative ) samples yielded an amplified product in the PCR analysis . For four remaining cases , the ostensibly species-specific PCR assay yielded an amplified product of the expected size , but the clones sequenced from this product did not correspond to the expected species . We further investigated these four cases by sequencing a clone library obtained by amplification with the same broad-range primers that were used in preparation for microarray analysis . Although the sequencing did not confirm the presence of any of the four questionable species/taxa , it provided strong evidence for a major source of false-positive hybridization signals . Specifically , in three of the four cases , we identified a relatively abundant species whose rDNA sequence was sufficiently similar to the probe sequence that it was likely to account for the observed signal . In one case ( Legionella pneumophila ) , which was predicted to be present at approximately 1% , we were unable to identify any candidate species that could account for the hybridization signal ( i . e . , none with best BLAST matches scores ≥30 ) , among our set of 192 sequences . Since our power to detect a species present at a partial abundance of 1% was only 85% , it remains possible that this species , or another species with a similar SSU rDNA sequence , could have been present at a low abundance in the suspect samples . Both our DNA extraction and rDNA amplification methods were optimized for bacteria and suboptimal for eukaryotes and archaea , thus we separately tested for the presence and abundance of fungi or archaea by means of qPCR assays with broad specificity for the respective taxonomic groups . Based on our qPCR analysis , fungi were intermittently detectable in stool samples at relatively low abundance ( 104–106 rRNA genes/g fecal wet weight ) , persisting for varying durations in individual babies , through the first year of life . One of the babies in this study ( baby 10 ) was noted to have a diaper rash , as well as oral thrush , both of which are commonly caused by a fungus ( Candida ) , and which were treated with an antifungal agent ( nystatin ) . The qPCR analysis detected especially high levels of fungal rDNA in stool samples from this baby , particularly during the period in which these findings were described . This baby's mother also had notably high levels of fungal SSU rDNA sequences in her prenatal vaginal swab sample , but not in her “day 0” stool sample . The prevalence of archaea was considerably lower and more variable than that of fungi or bacteria; qPCR analysis detected archaeal rRNA genes ( in the range of 103–106 rRNA genes/g ) in only seven babies during their first year of life , and in four of these babies , they were detected in only a single sample . In these babies , archaea appeared only transiently , and almost exclusively in the first few weeks of life; they were detected in only one infant after the fifth week of life . Limited analysis of archaeal sequences amplified from the three maternal stool samples that tested positive for archaea ( mothers 4 , 9 , and 12 ) revealed a predominance of Methanobrevibacter smithii ( 7/8 archaeal clones identified , including at least one clone from each mother ) , with one additional ( uncultured ) archaeal phylotype . Results of qPCR analysis of fungi and archaea are included in Dataset S5 and shown graphically with bacterial qPCR results in Figure S2 .
The microbial colonization of the infant GI tract is a remarkable episode in the human lifecycle . Every time a human baby is born , a rich and dynamic ecosystem develops from a sterile environment . Within days , the microbial immigrants establish a thriving community whose population soon outnumbers that of the baby's own cells . The evolutionarily ancient symbiosis between the human GI tract and its resident microbiota undoubtedly involves diverse reciprocal interactions between the microbiota and the host , with important consequences for human health and physiology . These interactions can have beneficial nutritional , immunological , and developmental effects , or pathogenic effects for the host [2 , 5 , 7 , 18 , 45] . This study began with the development of a DNA microarray with nearly comprehensive coverage of the bacterial taxa represented in the available database of SSU rRNA gene sequences . Our microarray design and experimental methods were based on lessons learned in the validation of a less comprehensive SSU rDNA microarray [46] . These previous experiments enabled us to optimize our methods for computational prediction of SSU rDNA hybridization behaviors , and to develop an experimental protocol that maximized hybridization specificity . The excellent concordance in the measurements of individual taxa determined using the new microarray design in comparison with sequencing results from corresponding SSU rDNA clone libraries ( Figure 1 ) suggests that these design principles hold true for this platform across a diversity of taxa and give us confidence in both the comprehensiveness and accuracy of the results obtained with our new microarray probe set . It is important to note , however , that our methods of array design and analysis are imperfect and still evolving . Several of the unexpected species predicted by the microarray to be present in one or more samples could not be corroborated by sequencing . In most of these cases , sequence analysis of the sample ( s ) in question revealed that low-level cross hybridization of a highly abundant species was responsible for the false-positive prediction , a result that will be taken into consideration in future rounds of array design and analysis . We used this microarray in a detailed , systematic , and quantitative study of bacterial colonization of the newborn human GI tract . We used freshly collected stool samples as surrogates for samples taken from the lumen and mucosa of the colon . Although there are undoubtedly differences in the population profiles of stool samples and corresponding mucosa , we found in a previous study that the profiles are nonetheless remarkably consistent—sufficiently so that individual stool samples can readily be matched to colonic biopsy samples from the same individual , based on the similarity in their bacterial profiles [15 , 46] . Thus , we believe that the results of our temporal analysis of the bacterial populations in infant stool samples provide a useful window on the resident colonic microbiota . In view of the importance of the symbiosis between human host and gut commensals for both human host and microbial colonist , it would be easy to imagine that the program of microbial colonization of the neonatal GI tract would have evolved under strong selective pressure , acting on both the intestinal niche and its microbial colonists , to be highly deterministic and stereotyped . We might have expected that a highly restricted group of co-evolved commensals would be exceptionally well adapted to this environment and consistently dominate the colonization process in a stereotyped fashion . Indeed , the bacteria that we found in infant and adult feces , presumably reflecting the colonic microbiota , were largely restricted to only a small subset of the bacterial world—Proteobacteria , Bacteroides , Firmicutes , Actinobacteria , and Verrucomicrobia . Yet , surprisingly , we found that in the first days to months of life , the microbiota of the infant gut , and the temporal pattern in which it evolves , is remarkably variable from individual to individual . The seemingly chaotic progression of the early events in colonization , and the similarity in bacterial composition of some early infant samples to breast milk or vaginal swabs , suggests that the bacterial population that develops in the initial stages is to a significant extent determined by the specific bacteria to which a baby happens to be exposed . Notably , these maternal “signatures” did not persist indefinitely , as evidenced by our failure to find a significantly higher correlation of the overall taxonomic profiles of baby/parent pairs from the same household versus different households . An important exception to the tale of individuality and uniqueness in the early profiles was the remarkable similarity of the temporal profiles of the fraternal twins ( babies 13 and 14 ) ( Figures 4 and 5 ) . These twins shared both a common environment and approximately 50% genetic identity , making it impossible to determine from this study to what degree each of these commonalities is responsible for their similar colonization patterns . However , evidence from this and other studies suggests that the shared environment is a major factor . One argument in favor of this view is the lack of comparable similarity in the microbial communities of other pairs that also share 50% genetic identity , including mother:baby , father:baby , and sibling:baby ( unpublished data ) , although this dissimilarity may be due in part to their differing stages in development . Another argument in favor of a strong environmental influence is the coincidental transient appearance of specific organisms in both twins—it is hard to imagine that the appearance of a particular microbe on a particular day could be genetically programmed . Our final argument rests on evidence from a previous study that the microbiota of genetically equivalent families from a cross of inbred mice was more similar among members of the same “household” ( mother and offspring who share a cage ) than between households [1] . The definition of a “healthy” intestinal microbiota encompasses a remarkable diversity of community profiles in the first 6 mo of life . Although diverse and idiosyncratic in the early months , these microbial communities became progressively more similar to one another ( Figure 6A ) , converging toward a generic adult-like profile ( Figure 6B ) characterized by a preponderance of Bacteroides and Firmicutes , common occurrence of Verrucomicrobia , and very low abundance of Proteobacteria and aerobic Gram-negative bacteria in general . We hypothesize that the earliest colonization events are determined to a large extent by opportunistic colonization by bacteria to which a baby is exposed in its environment . Common environmental exposures are likely to include the maternal vaginal , fecal , or skin microbiota , as is suggested by the observed similarity of some infants' early stool microbiota to these samples , which is consistent with previous evidence of vertical transmission of microbes [33 , 47 , 48] . The diversity and variation would thus reflect the corresponding individuality of these accidental exposures . Over time , however , the fitness advantage of the taxa that typically dominate the adult colonic microbiota apparently overcomes the initial advantage of early-colonizing opportunists that are less well adapted to the intestinal environment . In addition , progressive changes in the gut environment , due to intrinsic developmental changes in the gut mucosa , transition to an “adult” diet , and the effects of the microbiota itself [44 , 49–51] , may impose increasingly stringent selection for the most highly adapted bacteria . Thus , despite the unexpectedly chaotic early months , the establishment of the gut ecosystem in human infants turns out after all to follow a conserved , conventional program . The transformation of the intestinal microbiota to an adult-like pattern implicitly involved replacement of species found in infants , but rarely in adults , with species characteristically found in the adult colon . One potential driving force for such a demographic change might be that the adult-like community members eventually dominate by virtue of their greater ability to establish themselves stably and irreversibly once they colonize a host . We looked for evidence of this differential “stickiness” by comparing the autocorrelations over time of the abundance of each “species” ( see Materials and Methods ) . We found no clear evidence that the species characteristic of adult microbiota were able to establish more intrinsically stable colonization than the species characteristic of infant microbiota . We and others have found that the individual-specific characteristics of the bacterial microbiota of adults are stable , in the sense that they remain consistently more similar within an individual over time than between individuals , for periods of a year or more , and one of the striking results of this study was the identification of relatively stable , individual-specific patterns of bacterial colonization even in the first weeks and months of life . These observations raised the interesting possibility that opportunistic colonization events in early infancy might play a significant role in defining the distinct characteristics of the same individual's microbiota into adulthood . We looked for evidence of this by comparing the intraindividual and interindividual correlations of bacterial profiles at 1 or 2 mo and 1 y , and found no significant difference ( unpublished data ) . Thus , although these results certainly do not exclude the possibility that early colonization events play an important role in determining the adult microbiota , there does not appear to be a strong , direct correlation between the two . Our results and conclusions differ considerably from many previous reports in several respects . One notable discrepancy between our studies and many others was the relatively low frequency and abundance of Bifidobacteria in the fecal microbiota at any age from birth to adulthood . Bifidobacteria have received disproportionate attention , in part because of their reputed beneficial effects , and many studies have reported ( and reviews have repeated ) that the microbiota of breast-fed infants is dominated by Bifidobacteria [17–19] . We were thus surprised by , and initially skeptical of , the apparent paucity of Bifidobacteria in nearly all of our samples , and took steps to verify that our results were accurate . Bifidobacteria-specific qPCR corroborated the conclusion from our microarray results that Bifidobacteria were rarely major constituents of the GI microbiota , at least in this study population , and that in most babies , they did not appear until several months after birth , and thereafter persisted as a minority population . Although it is conceivable that there are geographical or demographic differences in the prevalence of Bifidobacteria , we suspect that the emphasis on Bifidobacteria in studies and reviews of the infant GI microbiota may be out of proportion to its prevalence , abundance , and relevance to health . The results presented here suggest numerous future avenues of research . An intriguing feature of the bacterial population dynamics was the occurrence of abrupt shifts punctuating intervals of relative stability . Except in one case ( the antibiotic treatment of baby 8 ) , we could not readily identify a strong candidate for the cause of the shifts we observed . Some possibilities include bacteriophage outbreaks that can selectively decimate a dominant taxonomic group [52]; stochastic , opportunistic invasion of a metabolic or anatomic niche by a fitter species; and subtle developmental or diet-induced changes in the gut environment tipping the fitness balance in the population . Other important avenues for future research will be comparing the composition and evolution of microbial communities encountered in these healthy babies to those of preterm or otherwise unhealthy babies and to investigate the effects of antibiotics , diet , and mode of delivery on the development and evolution of these communities . Even though the healthy babies in this study assumed a large range of microbial community profiles , they were similar in several respects , most notably in the major contributing phyla , in the acquisition of certain key phyla over time , and in the relative stability of their profiles over time . It may be that in other states of health or disease , we will find either species or groups that are novel to this environment , or unusual combinations of this newly defined set of “usual” species . Importantly , although we have shown that the gut microbiota becomes increasingly stereotyped over the first year , it is clearly established that stable interindividual differences persist even in adults [15 , 16] . When and how do these stable “intrinsic” characteristics of the microbiota of each individual develop ? How long do they persist ? How do the differing stabilities of colonization by different bacteria relate to their microanatomic ( e . g . , crypt vs . villus vs . mucous layer ) or metabolic niches ? Identifying the environmental and genetic factors that determine the distinctive characteristics of each individual's microbiota , and determining whether and how these individual specific features affect the host's physiology and health , will be an important goal for future investigations , in which the microarray described in this study will be a useful tool .
The microarray contained 10 , 500 DNA probes ( 10 , 265 unique sequences ) . The probes comprised 1 , 379 control probes ( 1 , 144 unique sequences ) and 9 , 121 unique taxonomically specific probes ( 5 , 938 group-level and 3183 species-level probes ) , consisting of 40-nucleotide ( nt ) sequences derived from the SSU rRNA genes , and selected for their specificity to the corresponding species or taxonomic group . The basic principles of the design are detailed in a previous report [46] . The design was based on the 2004 prokMSA SSU rDNA sequence database and phylogenetic tree [38] , containing 86 , 453 prokaryotic SSU rDNA sequences ( 5 , 672 archaeal and 80 , 781 bacterial ) organized into 672 archaeal and 15 , 765 bacterial OTUs . OTUs were defined in prokMSA as groups of sequences with identity scores ( as defined in [38] ) greater than either 95% or 98% ( in certain medically relevant genera ) . We distilled this database by selecting a single high-quality sequence representative for each OTU , and trimmed the sequences to within the regions amplified by universal bacterial ( Bact-8F: 5′-AGAGTTTGATCCTGGCTCAG-3′ [53] + 1391R: 5′-GACGGGCGGTGTGTRCA-3′ [54] ) or archaeal ( Arch344 [55]/1391R ) primers . The OTUs in our pruned prokMSA database were organized into 945 ( 53 archaeal + 892 bacterial ) taxonomic groups ( nodes ) containing multiple OTUs , and 92 single OTU “nodes . ” The nomenclature of the database was such that each OTU was designated by a numerical code that indicated its prokMSA taxonomy , e . g . , species “1 . 2 . 3 . 4 . 5 . 6 . 007” belongs to superkingdom 1 , phylum 1 . 2 , class 1 . 2 . 3 , order 1 . 2 . 3 . 4 , family 1 . 2 . 3 . 4 . 5 , genus 1 . 2 . 3 . 4 . 5 . 6 . In this manuscript , taxonomic levels are referred to according to their depth in the OTU code , e . g . , species “1 . 2 . 3 . 4 . 5 . 6” belongs to the more-inclusive group “1” at level 1 and less-inclusive group “1 . 2 . 3 . 4: at level 4 . We generated a large set of candidate probes by using BLAST [56] to predict the potential for hybridization of overlapping 40-nt sequences from each OTU in our distilled sequence database with each of the other OTUs in the database ( by tiling across each sequence with window of two ) . A sequence was deemed a candidate probe for a specific taxonomic group if it was predicted to hybridize to at least 10% of that group's members , and not to any non-group members , using an empirically determined threshold of 28 out of 40 BLAST match–mismatches ( total nucleotide matches minus mismatches to the best BLAST hit for a given rDNA sequence ) as our cutoff for potential hybridization [46] . From the resulting set of candidate probes , we selected , for each node/taxonomic group , the two probes predicted to hybridize to the largest fraction of that group . We also selected probes from our candidate set such that each OTU in our distilled database was represented by probes at as many taxonomic levels as possible . Due to space limitations on the array , we were unable to print species-specific probes for every OTU in our database . Instead , we supplemented the set of group-level probes designed as described above with three additional sets of probes . First , we compiled a list of bacterial species that were either medically relevant or known human commensals . We identified the prokMSA codes for each of these species and tested the selected sequence from that OTU for species-specific 40-nt sequences , as defined by a BLAST match–mismatch score no greater than 27 to any other sequence . We also evaluated the species-specific probes from our previous array design [46] in the context of the prokMSA database , and included 467 such species-specific probes , representing 286 species . The final category of taxonomic probes were the “novel OTU” probes—316 probes designed to represent recently discovered SSU rDNA “species” that were identified in studies of the adult human colon [15] or stomach [57] ( novel OTUs were defined by a 99% identity cutoff as described in the original studies ) . Finally , our microarray design included 1 , 153 control probes—positive and negative—designed for normalization and systematic examination of hybridization behaviors . The negative controls included both non-rDNA sequences and reverse complement rDNA sequences , whereas the positive controls included primer sequences and several sets of overlapping probes covering complete bacterial , archaeal , and eukaryotic SSU rRNA genes . Surface-attached oligonucleotide probes were synthesized in situ as previously described [58] , with a 10-nt poly-T linker used to tether the specific 40-nt DNA probe ( Agilent Technologies , http://www . chem . agilent . com ) . All arrays also included 307 standard Agilent control probes . All probe sequences and annotations are available in Dataset S6 . Our microarray probe set included one or more group-specific probes for 649 of the 950 taxonomic groups in prokMSA and species-specific probes for 1 , 590 bacterial and 39 archaeal species . Taken together , these probes ensured that 15 , 406 ( 94% ) of the 16 , 437 species represented in the prokMSA database had at least one representative probe at some level in the tree from phylum to species , and most prokMSA species ( 74% ) had representative probes at multiple taxonomic levels ( mean of 2 . 4 levels per species ) . Thirteen healthy pregnant women were recruited at the Stanford University Medical Center . All study participants , including 14 babies ( one set of fraternal twins ) , nine fathers , 13 mothers , and two siblings ( 1–2 y old ) provided informed consent or were consented for by their parents . The study design was approved by the Stanford University Administrative Panel on Human Subjects in Medical Research . At 36–40 wk of gestation , vaginal swabs ( Copan Diagnostics , http://www . copanusa . com ) were obtained from ten of the 13 mothers and immediately frozen at −20 °C . After birth , infant stool samples were obtained by the parents using stool collection vials ( Sarstedt , http://www . sarstedt . com/php/main . php ) , which contained a spoon for standardized collection of approximately 300 mg of material . Infant stool samples were collected according to the prescribed schedule ( Table 1 ) and immediately stored in home freezers . A maternal “day 0” stool sample was obtained within 0–5 days following delivery . Stool samples were transported on ice to the laboratory for processing 2 wk , 3 mo , and 6 mo after birth of the baby . Twelve of the mothers provided breast milk samples ( ~20 ml ) 3–9 mo after delivery , and one of them also provided breast milk 6 d after delivery; these samples were collected in 50-ml tubes and frozen immediately . Nine of the study families also provided contemporaneous stool samples from mother , father , and baby 12–17 mo after the baby's birth . Upon arrival in the laboratory , all samples were immediately transferred to a −80 °C freezer , and stored there until processing . A total of 548 samples were collected , including 471 stool samples from babies , 39 stool samples from mothers , nine stool samples from fathers , two stool samples from siblings , 16 breast milk samples , and 11 vaginal swabs . Parents were instructed to keep a journal recording key events in the categories of illness , medication , dietary change , and travel . Table 2 contains selected information for each baby ( e . g . , gender and method of delivery ) . DNA was extracted from stool samples using the QIAamp Stool DNA mini kit ( Qiagen , http://www1 . qiagen . com ) . Vaginal swabs were processed using the QIAamp DNA mini kit ( Qiagen ) . Milk samples were first concentrated by spinning 2 ml in a microcentrifuge for 10 min at 5 , 000 g , removing 1 , 800 μl of the supernatant . The pellet was resuspended in 200 μl of the remaining supernatant , and DNA was extracted using the QIAamp DNA mini kit . Samples were processed in batches of approximately 16 , and multiple extraction controls were included in each run to monitor contamination . The ratio of samples to extraction controls was 6 . 8 for stool , 2 . 8 for vaginal swabs , and 5 . 3 for milk . SSU rDNA was amplified from extracted DNA using broad-range bacterial-specific primers Bact-8F ( 5′-AGAGTTTGATCCTGGCTCAG-3′ ) [53] and T7-1391R ( 5′-AATTCTAATACGACTCACTATAGGGAGACGGGCGGTGTGTRCA-3′ ) [46 , 54] . These primers amplify approximately 90% or more of the full-length bacterial SSU rRNA coding sequence , and provide a promoter for T7 RNA polymerase . PCR mixtures were composed of 1× PCR buffer II ( Applied Biosystems , http://www . appliedbiosystems . com ) , 1 . 5 mM MgCl2 , 0 . 05% Triton X-100 , 20 mM tetramethylammonium chloride , 2% dimethyl sulfoxide , 0 . 1 mM concentrations of each deoxyribonucleoside triphosphate , 0 . 4 μM concentrations of each primer , 2 . 5 U of AmpliTaq DNA polymerase ( Applied Biosystems ) , and 5 μl of extracted DNA in a final volume of 50 μl . The PCR conditions used were 5 min at 95 °C , 20 cycles of 30 s at 94 °C , 30 s at 55 °C , and 90 s at 72 °C , followed by 8 min at 72 °C . Amplification was carried out by using a GeneAmp PCR system 9700 ( Applied Biosystems ) . In cases of extremely low yield , multiple 20-cycle reactions were pooled . PCR reactions were cleaned up in 96-well format using Invitrogen's Charge Switch PCR Purification bead-based system ( Invitrogen , http://www . invitrogen . com ) , and stored at −20 °C . Our common reference was an equimolar mix of SSU rDNA amplicons from each sample ( infant or maternal stool , vaginal , or breast milk ) collected before the subject infant was 1 y old . To create the equimolar mix , purified 20-cycle PCR products were quantitated in 96-well format using the Quant-It PicoGreen dsDNA kit ( Molecular Probes , http://probes . invitrogen . com ) and pooled in equal amounts . The approximate fractions of stool- , vaginal- , and milk-derived SSU rDNA in the resulting pool were 90% , 5% , and 5% , respectively . This DNA pool was used both as a template for in vitro transcription ( for microarray hybridizations ) and for construction of a SSU rDNA clone library . The reference pool ( an equimolar mix of SSU rDNA amplicons from most samples , described above ) was cloned and sequenced as previously described [15] . We obtained 3 , 458 high-quality bacterial rDNA sequences of length greater than 800 nt , including both 3 , 163 double reads and 295 single reads . These sequences were taxonomically classified using the Ribosomal Database Project ( RDP ) classifier ( summarized in Table 3 ) . We also cloned and sequenced several hundred SSU rDNA amplicons ( mean = 342; range = 125–557 adequate sequences ) from each of 12 diverse individual samples in the same way , yielding a total of 4 , 100 high-quality sequences of length greater than 800 nt . The 12 samples sequenced consisted of ten stool samples ( day 11 from baby 2; day 1 from mother of baby 3; week 12 from baby 3; month 6+ from baby 8; day 1 and day 2B from baby 10; day 12 from baby 12; month 7 from baby 13; month 7 from mother of twins: babies 13 and 14; and month 7 from baby 14 ) , one milk sample from the mother of twins 13 and 14 , and one vaginal sample from the mother of twins 13 and 14 . Each sequence from these 12 individual samples was taxonomically classified according to the 2004 prokMSA taxonomy [38] using BLAST . Specifically , we used BLAST to find the sequence with the most matches in the entire prokMSA database ( also trimmed to within 8F and 1391R ) . The top two hits were reported and compared ( hits with fewer than 600 matched nucleotides were not considered ) , and if these two hits mapped to the same OTU , then the sequence was classified to that OTU . If the top two hits differed in their taxonomic code , then the sequence was classified only at the most-specific level shared by the top two hits . In cases in which the second-best hit was considerably worse ( matches 2nd/matches 1st < 0 . 9 ) , only the best hit was considered . The prokMSA OTU codes explicitly define the taxonomic classification of a sequence at all phylogenetic levels for all of the 4 , 100 high-quality “individual sample” sequences . Purified SSU rDNA amplicons were used as a template for in vitro transcription-based synthesis of amino-allyl–labeled single-stranded RNA using the MEGAScript T7 In Vitro Transcription Kit ( Ambion , http://www . ambion . com ) . Transcription reactions were cleaned up in 96-well format using Ambion's MagMax RNA Purification system and stored at −20 °C . Large batches ( 5–10 μg ) of reference RNA ( equimolar pool of all samples; see below ) were labeled using Cy3 NHS ester and stored for several weeks for repeated use . On the day of hybridization , 1 μg of each sample RNA was labeled using Cy5 NHS ester as described previously [46] . We then combined 100 ng of Cy5-labeled sample and 100 ng of Cy3-labeled reference pool in a volume of 48 μl ( in nuclease-free water ) . We then added 2 μl of Agilent's 25× fragmentation buffer , and fragmented the RNA by heating at 70 °C for 30 min before stopping the reaction by putting it on ice and adding 50 μl of Agilent's 2× hybridization buffer . Immediately before hybridization , we heated the reaction to 95 °C for 5 min , then cooled it on ice before adding 120 μl of 1× hybridization buffer to 100 μl of fragmented , labeled RNA , and loaded 200 μl of this mixture into a hybridization chamber ( Agilent ) . Arrays were hybridized at 60 °C in a rotating oven for 14–18 h . Slides were washed in 6× SSC , 0 . 005% TritonX-102 for 5 min at room temperature , then in 0 . 1× SSC , 0 . 005% TritonX-102 for 5 min , and scanned immediately using an Agilent DNA Microarray Scanner . Washing and scanning were performed in a low-ozone environment [59] . Data were extracted from the scanned microarray image using the most current version of the Agilent Feature Extraction software ( Versions 7 . 1 . 1–8 . 1 . 1 ) . The raw background-subtracted Cy5 ( or Cy3 ) fluorescence intensity values for each probe were normalized by dividing the Cy5 ( or Cy3 ) values by the median Cy5 ( or Cy3 ) value of the universal probe “UNIV2” ( extended version of 3′ PCR primer 1391R: 5′-GTGGGGAGCGAACAGGATTAGATACCCTGGTAGTCCACGC-3′ ) from the corresponding array , and multiplying by 100 . At this stage , values ranged from 0 . 01 to approximately 100; values greater than 100 were rare , occurring only when the fluorescent signal for a specific probe was brighter than that of the universal probe . The normalized Cy5 values were “decompressed” by the following correction: decompressed Cy5 equals 10 to the power of log5 of the normalized Cy5 intensity . This decompression corrects for the nonlinear relationship of the hybridization signals to the relative abundance of the target species , which we observed in a series of serial dilution experiments described in a previous report [46] . In those experiments , we found that a 10-fold change in abundance translated into approximately a 5-fold change in the corresponding fluorescence intensity . Following this transformation , the expanded range of values was 0 . 001 to approximately 700 . We used BLAST to predict the hybridization of each of the 3 , 458 common reference rDNA sequences greater than 800 base pairs , using a weighting scheme described previously [46] , such that a probe with a perfect match to every sequence would have an expectation of 100% , and probes with fewer or imperfect matches to the sequences in the reference pool would have correspondingly lower expected hybridization values . We then calculated the log ( observed/expected ) ratio for each probe in the context of our pooled sample reference mix and applied these probe-specific correction factors to the decompressed Cy5 values . Microarray analysis of synthetic pools of defined composition allowed us to identify poorly performing probes , which were then excluded from further analyses . We evaluated probe performance using five synthetic pools ( four pools of six unique rDNA PCR products , one pool of 230 unique rDNA PCR products [from [46]] ) and one biological pool ( the common reference described above ) , whose composition was characterized by deep sequencing . Probes with fluorescence intensity values greater than 0 . 5% of that of the universal probes despite the lack of predicted sequence homology ( defined as a BLAST match–mismatch value less than 25 out of a possible 40 ) to any of the species in the sample were excluded from subsequent analyses . We also discarded data from probes that had an observed ( decompressed ) signal intensity 100-fold higher or lower than expected , in the biological reference pool analyses , ( as described above in the Microarray data normalization section ) . The remaining set of 6 , 381 well-behaved probes was used in all subsequent analyses . For each sample , we derived an estimate of the relative abundance of each taxonomic group in our phylogenetic tree using an algorithm that ensures that no species contributes more than once to the estimate of a taxonomic group abundance , and that the downstream probes ( probes that represent distinct subsets of species belonging to that phylogenetic group ) are incorporated into the cumulative group abundance estimate . Specifically , for each phylogenetic group in each sample , we sorted all of the downstream probes according to their microarray-based relative abundance estimates and calculated the total sum of the relative abundance estimates of all nonoverlapping probes . As a result , the specific probes added together to represent a given taxonomic group varied across samples , depending on which specific probes had the greatest hybridization signal in that sample . Estimates of the relative abundance of each prokMSA group , at each level of the taxonomic hierarchy , in the reference pool and in the 12 individual samples analyzed by sequencing , were derived by calculating the proportion of the sequences in the corresponding rDNA library that were assigned by BLAST to that group . These sequence-based abundance estimates were directly compared to those derived from the microarray data as described in the previous section . We used autocorrelations as a way to measure the tendency of a taxonomic group to persist once established ( “stickiness” ) . For a given time series from time a to time b , we calculated the Pearson correlation for each baby of the vector ( a + 1 , … , b ) and vector ( a , … , b − 1 ) representing the log ( relative abundance ) estimates . The autocorrelation of each taxonomic group was then taken to be the median autocorrelation across all of the babies for which the taxonomic group in question was present at least once ( defined as abundance >0 . 1% ) in the time interval in question . In our “stickiness” analysis , we performed this analysis separately for two different sets of samples , collected at different sampling intervals , in order to avoid the confounding effects of variation in sampling intervals: ( 1 ) once weekly samples from weeks 1–12; and ( 2 ) monthly samples from months 1–6 . To screen for archaeal and fungal rDNA sequences . we first screened pools comprising all of the extracted DNA samples from each baby ( i . e . , one pool per baby ) , all parental stool samples , all vaginal samples , and all milk samples , respectively . For each pool that gave a positive result , all the component samples were then analyzed individually . To screen for archaeal rDNA sequences , we used two sets of broad-range archaeal-specific primers: A751F and U1406R [60]; and Arch333F ( 5′-TCCAGGCCCTACGGG-3′ [61] ) and Arch958R ( 5′-YCCGGCGTTGAMTCCAATT-3′ [62 , 63] ) . To screen for fungi , we used the broad-range fungal-specific primers 817F [64] and 1536R-rev ( 5′-AATRCAATGCTCTATCCCCA-3′ , adapted from [64] ) . PCR mixtures were similar to those used for the broad-range bacterial PCR described above , but with the following changes: for the fungal-specific PCR , the MgCl2 concentration was increased to 2 mM , and BSA was added in a final concentration of 1 mg/ml . To both fungal-specific and archaeal-specific PCR reactions , no tetramethylammonium chloride was added , and 2 μl of pooled DNA was added , with a final volume of 50 μl . The cycling program consisted of 5 min at 95 °C , 35 cycles of ( 30 sec at 94 °C , 30 sec at 55 °C , and 30 sec at 72 °C ) , followed by 8 min at 72 °C . Amplification reactions were analyzed on agarose gels . To investigate whether mismatches to the broad-range bacterial primers could affect amplification , DNA was extracted from five bacterial reference strains: Escherichia coli ( TOP10 cells , Invitrogen ) , Clostridium perfringens ( ATCC 13124 ) , Bacteroides fragilis ( ATCC 25285 ) , Bifidobacterium longum ( ATCC 15707 ) , and Bifidobacterium infantis ( ATCC 15697 ) , using the QIAamp DNA mini kit . DNA concentrations were measured using a UV spectrophotometer , and adjusted to correct for DNA yield , genome size , and number of SSU rRNA gene copies per genome to obtain “normalized” DNAs , each containing the same number of SSU rRNA gene copies per μl . Bacterial DNA from lysates was amplified individually or in pairs , using primers 8F and T7-1391R as described above , using either 20 or 35 cycles . For the paired reactions , normalized Bifidobacterium longum DNA was mixed with undiluted or serial dilutions of normalized DNA from Escherichia coli , Clostridium perfringens , or Bacteroides fragilis . After amplification , PCR mixtures were purified using the QIAquick PCR purification kit ( Qiagen ) and digested using a set of restriction enzymes selected to distinguish between the two PCR products obtained with the paired species . Digestions were analyzed on agarose gels , to quantitate the relative abundance of the PCR products representing Bifidobacterium longum and the species in the comparison group , respectively . A separate real-time qPCR assay was used to amplify and quantify rDNA from each of four microbial groups: total bacteria , Bifidobacteria , total fungi , and total archaea . Total bacterial qPCR was performed using a 10:1 mixture of the universal forward primer 8FM ( 5′-AGAGTTTGATCMTGGCTCAG-3′ , adapted from [53] ) and corresponding Bifidobacterium longum forward primer 8FB ( 5′-AGGGTTCGATTCTGGCTCAG-3′ , this study ) , with reverse primer Bact515R ( 5′-TTACCGCGGCKGCTGGCAC-3′ , adapted from [54] ) and TaqMan probe Bact338K ( 5′-FAM/CCAKACTCCTACGGGAGGCAGCAG/TAMRA-3′ , adapted from [65] ) . We supplemented the universal bacterial forward primer with the Bifidobacterium longum forward primer because an analysis of SSU rDNA sequences showed that this group was an outlier in that it had three mismatches to our forward primer 8FM . Pilot studies showed that this primer mixture allowed comparable amplification of SSU rRNA genes from representative Bifidobacteria , Bacteroides , Enterobacteria , and Clostridia . Bifidobacterium genus qPCR was performed using primers Bif42F [26] and Bif164R ( 5′-CATCCGGCATTACCACCCGTT-3′ , adapted from [66] ) , and probe Bif126_Taqman [26] . Total fungal qPCR was performed using primers ITS1F-F ( 5′-CTTGGTCATTTAGAGGAAGTAA-3′ [67] ) and ITS4-R ( 5′-TCCTCCGCTTATTGATATGC-3′ [68] , and TaqMan probe 5 . 8S ( 5′-FAM/CATTTCGCTGCGTTCTTCATCGATG/TAMRA-3′ , adapted from [68] . Archaeal qPCR was performed using primers Arch333F ( 5′-TCCAGGCCCTACGGG-3′ ) [61] and Arch958R ( 5′-YCCGGCGTTGAMTCCAATT-3′ ) [63] , and TaqMan probe 515F ( 5′-FAM/GTGCCAGCMGCCGCGGTAA/TAMRA-3′ , adapted from [54] ) . For all qPCR assays , each 20-μl reaction contained 1× TaqMan Universal PCR master mix ( Applied Biosystems ) , 0 . 9 μM of each primer ( 0 . 09 μM of primer 8FB in the total bacterial assay ) , 0 . 2 μM of the probe , 1 U of AmpliTaq Gold DNA polymerase ( Applied Biosystems ) , and 2 μl of extracted DNA . The thermal cycling program consisted of 95 °C for 10 min , followed by either 40 cycles ( bacterial and bifidobacterial assays ) or 45 cycles ( fungal and archaeal assays ) of 95 °C for 30 s , 55 °C for 30 s , 60 °C for 45 s , 65 °C for 15 s , and 72 °C for 15 s [69] . Reactions were carried out in a Prism 7900HT Sequence Detection System ( Applied Biosystems ) . Ten-fold serial dilutions of known quantities of rDNA from the appropriate microbial group ( i . e . , bacteria , Bifidobacteria , fungi , or archaea ) were used to generate standard curves . Absolute rDNA abundance was calculated based on the standard curves using SDS software version 2 . 1 ( Applied Biosystems ) with the baseline set at cycles 3–15 ( bacterial and bifidobacterial assays ) or cycles 3–13 ( fungal and archaeal assays ) , and the cycle threshold set within the geometric phase of the amplification curve . Sensitivity of each assay was approximately 100 rDNA molecules per PCR reaction well . Every qPCR reaction plate included two types of negative controls ( reagent control and aliquot control ) , each in triplicate . Specificity of the bifidobacterial real-time PCR assay was tested using genomic DNA extracted from 17 bacterial reference strains: Bacillus subtilis ( ATCC 6633 ) , Bacteroides fragilis ( ATCC 25285 ) , Bacteroides thetaiotaomicron ( ATCC 29148 ) , Bifidobacterium longum ( ATCC 15707 ) , Bifidobacterium infantis ( ATCC 15697 ) , Clostridium perfringens ( ATCC 13124 ) , Clostridium putrefaciens ( ATCC 25786 ) , Enterococcus faecalis ( ATCC 19433 ) , Escherichia coli ( TOP10 cells; Invitrogen ) , Haemophilus haemolyticus ( ATCC 33390 ) , Lactobacillus acidophilus ( ATCC 4356 ) , Lactobacillus delbrueckii ( ATCC 4797 ) , Megasphaera elsdenii ( ATCC 17752 ) , Proteus vulgaris ( ATCC 13315 ) , Pseudomonas aeruginosa ( ATCC 10145 ) , Staphylococcus aureus ( ATCC 25923 ) , and Streptococcus salivarius ( ATCC 13419 ) . We investigated the origin of array hybridization signals representing 12 species/taxa whose presence in the samples was unexpected and uncorroborated by sequences in the reference pool . For each analysis , we used one or both of two independent assays . First , we attempted to amplify the sequences apparently detected by the array analysis , using species/taxa specific primers; when a product was obtained , it was further analyzed by sequencing . To amplify the sequences , we used a primer identical to the 40-mer probe that yielded a hybridization signal in our microarray as the 5′ primer and a 40-mer universal sequence ( the reverse complement of sequence UNIV2 given above ) as the 3′ primer . In some cases , we also tried to amplify the suspect sequence using a truncated ( 23-mer ) version of the corresponding oligonucleotide probe from the microarray paired with a known group-specific PCR primer from ProbeBase [55] . All PCRs were performed under conditions identical to those used in the original amplifications of samples for microarray analysis . Positive bands of expected size were cloned and sequenced . As a second approach , four samples predicted from the microarray results to contain sequences from unexpected species were further analyzed by sequencing of SSU rDNA clone libraries ( 96–288 clones ) , generated by amplification using broad-range bacterial primers 8F and 1391R ( as above ) , and cloning and sequencing as previously described [15] . The relative abundances predicted by microarray analysis and the numbers of clones sequenced were as follows: Vibrio ( 13% ) : 96 , Deinococcus ( 0 . 1% ) : 192 , Spirochaetes ( 1% ) : 288 , and Legionella pneumophila ( 1% ) : 192 . | It has been recognized for nearly a century that human beings are inhabited by a remarkably dense and diverse microbial ecosystem , yet we are only just beginning to understand and appreciate the many roles that these microbes play in human health and development . Knowing the composition of this ecosystem is a crucial step toward understanding its roles . In this study , we designed and applied a ribosomal DNA microarray-based approach to trace the development of the intestinal flora in 14 healthy , full-term infants over the first year of life . We found that the composition and temporal patterns of the microbial communities varied widely from baby to baby , supporting a broader definition of healthy colonization than previously recognized . By one year of age , the babies retained their uniqueness but had converged toward a profile characteristic of the adult gastrointestinal tract . The composition and temporal patterns of development of the intestinal microbiota in a pair of fraternal twins were strikingly similar , suggesting that genetic and environmental factors shape our gut microbiota in a reproducible way . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"ecology",
"obstetrics",
"pediatrics",
"and",
"child",
"health",
"immunology",
"microbiology",
"homo",
"(human)",
"eubacteria"
] | 2007 | Development of the Human Infant Intestinal Microbiota |
To determine whether variation in the preservative , pore size of the sieve , solvent , centrifugal force and centrifugation time used in the Ridley-Allen Concentration method for examining faecal specimens for parasite stages had any effect on their recovery in faecal specimens . A questionnaire was sent to all participants in the UK NEQAS Faecal Parasitology Scheme . The recovery of parasite stages was compared using formalin diluted in water or formalin diluted in saline as the fixative , 3 different pore sizes of sieve , ether or ethyl acetate as a solvent , 7 different centrifugal forces and 6 different centrifugation times according to the methods described by participants completing the questionnaire . The number of parasite stages recovered was higher when formalin diluted in water was used as fixative , a smaller pore size of sieve was used , ethyl acetate along with Triton X 100 was used as a solvent and a centrifugal force of 3 , 000 rpm for 3 minutes were employed . This study showed that differences in methodology at various stages of the concentration process affect the recovery of parasites from a faecal specimen and parasites present in small numbers could be missed if the recommended methodology is not followed .
The use of a concentration method to examine faeces for parasites increases the likelihood of finding ova , cysts and larvae , particularly in those specimens where they are present in insufficient numbers to be seen on direct microscopy . Concentration methods have been employed in clinical laboratories since 1948 when Ritchie [1] demonstrated their effectiveness . The method was improved by Ridley and Hawgood in 1956 [2] and simplified by Ridley and Allen in 1970 [3] using 10% formalin in water as a fixative , ether as a solvent to extract fat and debris , followed by filtration through a sieve with a pore size of 425 microns and centrifugation at 3000rpm for 1 minute to leave the ova , cysts and larvae in the sediment at the bottom of the centrifugation tube . The method uses several pieces of apparatus which are washed after the concentration of each sample and also has health and safety implications since formalin is an irritant and ether is flammable . The hazardous aspects and labour-intensive nature of the method encouraged commercial companies to promote enclosed , disposable faecal concentration systems which claim to have comparable sensitivity to the Ridley Allen Method in recovering parasites . These kits reduce the hazards of formalin by being enclosed and by using ethyl acetate rather than ether , since it is less flammable and more stable . Young et al [4] showed ethyl acetate to be an acceptable alternative to ether in the recovery of parasites . In 2011 , UKNEQAS Parasitology sent a questionnaire to all participants of the Faecal Parasitology Scheme to ask for their routine method for concentrating faecal samples for the detection of parasites . This was in response to concerns expressed by UKNEQAS Parasitology who noted poor performance among participants examining faecal specimens that contained low numbers of parasites . Those concerns were also raised by participants themselves who reported recovering lower numbers of parasites than those seen by UKNEQAS Parasitology in the pre-distribution examination of the EQA specimens . Furthermore , where parasites were present in low numbers they were failing to see them at all , thus adversely affecting their score in the NEQAS Faecal Parasitology Scheme . Following an analysis of the questionnaire it was apparent that although 96% of respondents used a concentration method based on the Modified Ridley Allen [3] technique , there were differences regarding whether the formalin used was diluted in water or saline; pore size of the sieve ( 0 . 35mm– 1 . 5mm or no sieve used ) ; the centrifugal speed ( 500–3 , 500 rpm ) ; centrifugal time ( 1–10 minutes ) ; and the solvent used ( ether , ethyl acetate or no solvent ) depending on the commercial kit deployed . These findings prompted UKNEQAS to conduct a study to investigate how differences in the various stages of the concentration method affected the recovery of parasites from faecal specimens .
A questionnaire to ascertain the concentration methods routinely used in clinical laboratories to recover parasites from faecal specimens was sent to all 580 participants in the UKNEQAS Faecal Parasitology Scheme . Two hundred participants returned a completed questionnaire . The questionnaire sought information on the kit , fixative , pore size of the sieve , solvent , plus the centrifugation speed and time used to deposit the parasites . After analysis of the participants’ responses , UKNEQAS Parasitology conducted a study to examine those aspects of the concentration method that deviated from the modified Ridley Allen method . The stages examined were: Twenty specimens preserved in 10% formalin in water were concentrated using 10% formalin in water or 10% formalin in saline as a fixative . Eight specimens preserved in 10% formalin and water were examined using 3 different pore sizes of sieve; 425μm , 800μm and 1 , 500μm . Fifty four specimens preserved for between 6 months and 2 years in 10% formalin and water and 24 fresh unfixed specimens containing ova , cysts and larvae were concentrated and examined to compare the recovery of parasites using ether or ethyl acetate as an extractor of fat and debris . Since the polarity and miscibility indices [5] of ether result in its being a better fat extractor than ethyl acetate , the surfactant Triton X 100 is added to the faecal/formalin solution to compensate if ethyl acetate is used . Triton X 100 helps ethyl acetate to break up faecal matter and results in a less dense deposit , thus facilitating the identification of parasite stages . A concentration of 0 . 1% Triton X 100 i . e . 1 mL added to 1 litre of formalin/water was noted to be the most effective concentration since excess Triton X 100 results in a soapy deposit which makes it difficult to examine . The effect of not using a solvent was not examined in this study as it has already been shown that the recovery of parasite stages is significantly lower and more debris is present if a solvent is not used [3] . Eight samples preserved in 10% formalin and water were examined to assess the effect of parasite recovery using different centrifugation times , 1 minute , 2 minutes , 3 minutes , 4 minutes , 5 minutes and 10 minutes and different centrifugation speeds ( forces ) , 500rpm ( 34G ) , 1000rpm ( 134G ) , 1 , 500rpm ( 300G ) , 2000rpm ( 537G ) , 2 , 500rpm ( 840G ) , 3 , 000rpm ( 1 , 200G ) and 3 , 500rpm ( 1643G ) . The centrifugal times and speeds were selected in accordance with those used by participants as reported on the returned questionnaires . All centrifugations were carried out on the same centrifuge . The Midi Parasep faecal parasite concentrator , an enclosed system that employs the principle of the Ridley-Allen formol-ether sedimentation technique was used to concentrate samples for the comparison of ether and ethyl acetate [6] . A pea-sized amount ( equivalent to approximately 1 gram ) of faeces was mixed with 6 millilitres ( mL ) of 10% formalin in water in the mixing chamber . Two mL of ether or ethyl acetate was added ( formalin/Triton-X 100 mixture was added to the mixture if ethyl acetate was used as it helps to emulsify the faecal matter ) . Parasep was assembled and sealed by screwing the filter thimble and sedimentation cone onto the mixing chamber . The mixture was vortexed for 15 seconds and the system inverted to allow the mixture to be filtered through the filter thimble ( pore size of 425μm ) and centrifuged at 1200g or 3000 rpm for 1–3 minutes according to the manufacturer’s instructions at the time . The mixing chamber and the filter thimble were unscrewed together and discarded . Like the conventional Ridley-Allen sedimentation method , there is an upper ethyl acetate layer , fatty plug , formalin supernatant and deposit . The fatty plug was loosened and the supernatant was safely discarded according to the Control of Substances Hazardous to Health regulations 2002 [7] . The concentration procedure apart from the centrifugation stage was undertaken in a Class 1 safety cabinet . The Parasep faecal parasite concentrator with ethyl acetate as a solvent and Triton X 100 as a surfactant was used to compare the different centrifugal speeds and times . The recommended speed of 3000rpm and time of 1–3 minutes were altered according to the variations in speed and time quoted by participants . The Parasep faecal parasite concentrator with ethyl acetate as a solvent and Triton X 100 as a surfactant according to the manufacturer’s instructions was used to compare 10% formalin diluted in water and 10% formalin diluted in saline as a preservative . The conventional Ridley-Allen method with ethyl acetate and Triton X as a solvent was used to compare the sieves with different pore sizes . Prior to the microscopic examination , all faecal deposits were re-suspended in 75μL of saline ( three drops ) and thoroughly mixed . In each case , 50μL of the diluted deposit was dispensed onto a microscope slide and a 22mm by 22mm coverslip applied . The whole of the coverslip was examined and the number of ova , cysts and larvae recorded . All specimens were processed in duplicate and the mean calculated . The Wilcoxon Matched-Pairs Signed-Ranks Test which uses the sizes of the differences was the statistical test used to compare the recovery of parasites in formalin in water versus formalin in saline , ether versus ethyl acetate and the different pore sizes of the sieves .
The commercial kits used by participants , the manufacturer’s recommended procedure , the procedure used by participants , and the number of participants who followed the manufacturer’s procedure are shown in Table 1 . The centrifugal speeds and times used by participants are shown in Figs 1 and 2 respectively . The number of ova and cysts for each parasite species in individual specimens when 10% formalin in water or 10% formalin in 0 . 9% saline was used as a fixative are shown in Table 2 . The Wilcoxon Matched-Pairs Signed-Ranks Test showed that formalin in water was significantly more effective than formalin in saline in the recovery of parasites ( p < = 0 . 005 ) . The number of ova recovered using 3 different pore sizes of sieve during the filtration stage are shown in Table 3 . There was a significant difference in the recovery of parasites between the 3 different sieves . The pore size of 425 μm , as recommended in the Ridley-Allen Method resulted in the best recovery of ova ( Table 3 ) . The amount of debris seen in the deposit increased with pore sizes above the recommended , making ova more difficult to see . The number of ova and cysts for each parasite species in each specimen ( some parasites were present in multiple specimens ) are shown in Table 4 for the formalin-preserved specimens and Table 5 for the fresh , unpreserved samples . Use of ethyl acetate and Triton X recovered more parasites in both the preserved ( p < = 0 . 001036 ) and unpreserved specimens ( p < = 0 . 003732 ) than using ether . Although ether recovered more parasites than ethyl acetate and Triton X 100 from 11/54 preserved samples and 8/24 fresh unpreserved samples , ether failed to recover parasites from 5/54 preserved samples and 1/24 fresh unpreserved samples , whereas use of ethyl acetate and Triton X 100 recovered parasites from all of them . When using ether , ova and cysts became trapped in the fatty layer and were consequently disposed of with the supernatant , though the deposit contained less debris than when ethyl acetate was used . In order to assess the effect of centrifugal speed and time on the recovery of parasites , 8 organisms were selected; Ascaris lumbricoides , Trichuris trichiura , Hookworm species , Taenia species , Entamoeba coli , Entamoeba histolytica/dispar , Entamoeba hartmanni and Endolimax nana . They were concentrated using the different speed and time combinations reported in the questionnaire . The results are shown in Figs 3–10 . The number of parasite stages detected increased with an increase in centrifugal speed and time . However , although the recovery of parasites was greatest if the samples were centrifuged for 3 , 500rpm for 10 minutes , the deposit was more difficult to examine due to excess debris masking the parasites present .
A concentration method is essential to increase the chance of finding parasite stages in a faecal specimen . This study has shown that the preservative used , solvent used , differences in centrifugation time and speed , and pore size of the sieve all affect the recovery . It also showed that the modified Ridley Allen method was optimised when using ethyl acetate with triton X instead of ether and the centrifugation time was increased from one to three minutes . Manufacturers and users of concentration kits alike should adhere to the recommendations given below to increase the recovery of parasite stages from faecal samples , not least because failing to do so will have an adverse effect on their detection in clinical specimens . Our recommendation for the concentration method is that approximately 1 gram or a pea-sized amount of faeces should be concentrated and that 10% formalin in water should be used instead of formalin in saline . A surfactant ( Triton X 100 ) should be added when using ethyl acetate as a solvent . The most effective concentration of Triton X 100 is 0 . 1% . The sample must be sieved after adding it to formalin water using a pore size no greater than 0 . 5mm . The sample must be vortexed for at least 15 seconds after the addition of ethyl acetate and Triton X . Adequate centrifugal force must be used , i . e . 3000 rpm which equates to 1 , 200g . This confirms the centrifugal speed recommended in the modified Ridley Allen method [3] . A centrifugal time of 3 minutes is recommended . Three drops of saline should be added to the deposit prior to examination . Failing to do so results in the deposit being too dense to examine . | UKNEQAS Parasitology and participants themselves noted poor performance in examining faecal specimens that contained low numbers of parasites . Our survey of methods used showed that 96% of respondents used a concentration method based on the Modified Ridley-Allen technique , but there were differences regarding whether the formalin used was diluted in water or saline; the pore size of the sieve; the centrifugal speed; centrifugal time and the solvent used , depending on the commercial kit deployed . UKNEQAS conducted a study to investigate how differences in the various stages of the concentration method affected the recovery of parasites from faecal specimens . This study has shown that the preservative used , solvent used , differences in centrifugation time and speed , and pore size of the sieve all affect recovery . The modified Ridley Allen method was optimised when using ethyl acetate with triton X instead of ether and the centrifugation time was increased from one to three minutes . A concentration method is required for adequate detection of faecal parasites . Poor concentration technique leads to poor recovery of parasites . Variation in the procedure can reduce the recovery of parasite stages , particularly if present in small numbers . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"chemical",
"compounds",
"laboratory",
"equipment",
"engineering",
"and",
"technology",
"organic",
"compounds",
"animals",
"parasitology",
"preservatives",
"developmental",
"biology",
"ascaris",
"ascaris",
"lumbricoides",
"centrifugation",
"research",
"and",
... | 2016 | Faecal Parasitology: Concentration Methodology Needs to be Better Standardised |
Electroencephalography ( EEG ) provides a non-invasive measure of brain electrical activity . Neural population models , where large numbers of interacting neurons are considered collectively as a macroscopic system , have long been used to understand features in EEG signals . By tuning dozens of input parameters describing the excitatory and inhibitory neuron populations , these models can reproduce prominent features of the EEG such as the alpha-rhythm . However , the inverse problem , of directly estimating the parameters from fits to EEG data , remains unsolved . Solving this multi-parameter non-linear fitting problem will potentially provide a real-time method for characterizing average neuronal properties in human subjects . Here we perform unbiased fits of a 22-parameter neural population model to EEG data from 82 individuals , using both particle swarm optimization and Markov chain Monte Carlo sampling . We estimate how much is learned about individual parameters by computing Kullback-Leibler divergences between posterior and prior distributions for each parameter . Results indicate that only a single parameter , that determining the dynamics of inhibitory synaptic activity , is directly identifiable , while other parameters have large , though correlated , uncertainties . We show that the eigenvalues of the Fisher information matrix are roughly uniformly spaced over a log scale , indicating that the model is sloppy , like many of the regulatory network models in systems biology . These eigenvalues indicate that the system can be modeled with a low effective dimensionality , with inhibitory synaptic activity being prominent in driving system behavior .
Neural population models are high-order , multi-parameter , dynamical systems . It has long been known that , even in simple dynamical systems , there can exist very different parameter combinations which generate similar predictions [16–27] . This many-to-one mapping between parameter inputs and model observables is referred to as structural unidentifiability , if the predictions are exactly identical , or practical identifiability , if the predictions are nearly identical . Any fitting of an unidentifiable model to data results in large , correlated parameter uncertainties . Many developments in the study of identifiability in differential equation models have been motivated by problems in systems biology involving biomolecular regulatory networks [28–35] . Model unidentifiability is closely related to , though distinct from , model sloppiness . A model is referred to as sloppy if the sensitivity of its predictions for different parameters covers a broad range [36–41] . These sensitivities , quantified by the eigenvalues of the Fisher information matrix , are roughly uniformly spaced over a log scale . This characteristic has been discovered in a variety of nonlinear models and arises from the geometry of nonlinear models projected into data space [37 , 38] . Parameters that sensitively affect model predictions are termed ‘stiff’ while those that can be changed with little effect on predictions are termed ‘sloppy’ . While sloppy parameters are often unidentifiable as well , the terms are not synonymous [42 , 43] . Like unidentifiability , sloppiness has been found to be prevalent in models of biomolecular networks [44–47] . Unidentifiability and sloppiness are pervasive in nonlinear fitting problems , the simplest examples of which are fits to polynomials or to multiple exponentials [36 , 38] . Since parameter estimation in differential equation models always involves nonlinear fits ( to exponential impulse responses in the time domain or rational transfer functions in the spectral domain , for example ) , unidentifiability and sloppiness should always be a concern in dynamical systems . This is true even for linear , time-invariant systems [19 , 22 , 26] . Of course , explicitly nonlinear functions of parameters certainly exacerbate the problem—a nonlinear function at saturation will give the same response for a range of different parameter inputs , for example . Unidentifiabilities also arise when a model supports phenomena at significantly different timescales . For instance , if only dynamics on a slower timescale can be observed , parameters which determine behavior at the faster timescale would not be constrained by data [48] . Unlike in systems biology modeling , in neurophysical modeling there has been little recognition of the problem of unidentifiability , beyond select examples in neural code models [49] , a thalamo-cortical neural population model [50] , and dynamic causal models [51] . This has been cited in [52] as an example of how approaches used in systems biology can help address problems in computational neuroscience [53] . However , despite this lack of formal discussion , implicit recognition of unidentifiability in computational neuroscience has been widespread , with several studies , including those for models of single neurons [54–58] , occulomotor integration [59] , and neural populations [60–62] , detecting the large , correlated parameter uncertainties that are the hallmark of unidentifiability and sloppiness . In this paper we examine identifiability and sloppiness in a well-known neurophysical model [11 , 61 , 63 , 64] with 22 unknown parameters . We concentrate on fitting EEG data exhibiting alpha-oscillations in resting state human subjects in an attempt to understand the mechanistic origin of this prominent , yet still poorly understood , phenomenon . We fit the model to the EEG spectrum from each of 82 subjects using both a particle swarm optimization and Markov chain Monte Carlo method . When viewed across all subjects , only 1 of the original parameters , the decay rate of inhibitory synaptic activity , emerges as being identifiable . This indicates that inhibitory synaptic activity is essential for explaining alpha-rhythmogenesis . Examination of the Fisher information matrix shows that there are ∼5 parameter combinations that are identifiable , a considerable reduction from the original 22 . This indicates that , although most parameters are unidentifiable , their values cannot in general be selected arbitrarily to fit the data . The neural population model used in this paper is well established and has been described previously [11 , 65] . Semi-analytical and numerical solutions of these equations have revealed a rich repertoire of physiologically plausible activity including noise driven , limit cycle and chaotic oscillations at the frequency of the mammalian alpha rhythm [11 , 61 , 66 , 67] . Here we use the spatially homogeneous version given by the following coupled set of first and second order ordinary differential equations: τ e d h e ( t ) d t= h e r e s t - h e ( t ) + h e e q - h e | h e e q - h e r e s t | I e e ( t ) + h i e q - h e | h i e q - h e r e s t | I i e ( t ) , ( 1 ) τ i d h i ( t ) d t= h i r e s t - h i ( t ) + h e e q - h i | h e e q - h i r e s t | I e i ( t ) + h i e q - h i | h i e q - h i r e s t | I i i ( t ) , ( 2 ) d 2 I e e ( t ) d t 2 + 2 γ e e d I e e ( t ) d t + γ e e 2 I e e ( t ) = Γ e γ e e e ( N e e β S e ( h e ) + p e e ( t ) ) , ( 3 ) d 2 I i e ( t ) d t 2 + 2 γ i e d I i e ( t ) d t + γ i e 2 I i e ( t ) = Γ i γ i e e ( N i e β S i ( h i ) + p i e ( t ) ) , ( 4 ) d 2 I e i ( t ) d t 2 + 2 γ e i d I e i ( t ) d t + γ e i 2 I e i ( t ) = Γ e γ e i e ( N e i β S e ( h e ) + p e i ( t ) ) , ( 5 ) d 2 I i i ( t ) d t 2 + 2 γ i i d I i i ( t ) d t + γ i i 2 I i i ( t ) = Γ i γ i i e ( N i i β S i ( h i ) + p i i ( t ) ) , ( 6 ) where Sj ( hj ) =Sjmax ( 1+exp ( −2 ( hj−μj¯ ) σj ) ) ;j=e , i . ( 7 ) These equations describe the interactions between inhibitory and excitatory neuronal populations in a macrocolumn . Table 1 lists the parameters along with their physiological ranges as assumed by Bojak and Liley [61] . The temporal dynamics of mean soma membrane potentials for the excitatory ( he ( t ) ) and inhibitory ( hi ( t ) ) populations are described in Eqs ( 1 ) and ( 2 ) . The temporal dynamics of the synaptic activity , Iee ( t ) , Iie ( t ) , Iei ( t ) , and Iii ( t ) , are given by Eqs ( 3 ) – ( 6 ) . The relationship between the mean population firing rate , Sj , and the mean soma membrane potentials of the respective population is given in Eq ( 7 ) . It has been shown that the local field potential measured in the EEG is linearly proportional to the mean soma membrane potentials of the excitatory populations , he ( t ) [68 , 69] . In this study we fit the above model to EEG recordings from 82 different individuals . This data is a subset of a larger dataset which , in its full version , consists of 14 experimental tasks performed by each of the 109 subjects with recordings on 64 electrodes according to the International 10-10 System . The full set , collected and contributed by Schalk et al [70] using the BCI2000 instrumentation system , is available for public access in PhysioNet [71] ( https://www . physionet . org/pn4/eegmmidb/ ) . For the purpose of studying alpha-rhythm , we restricted our analysis to signals from the Oz electrode , selecting data from individuals whose EEG spectrum exhibited clear alpha peaks during the associated eyes-closed task . Welch’s method of averaging the spectra derived from multiple overlapping time segments [72] was used to estimate the single spectrum for a particular individual . This approach improves the precision of the power spectral density estimate by sacrificing some spectral resolution . A one-minute EEG signal associated with a particular individual , sampled at 160 Hz , was divided into segments using a 4-second Hamming window with an overlap of 50% . Since the computational demands of fitting our model directly to EEG time series data are prohibitive , we fit the EEG spectrum instead . This approach is in accordance with earlier fits of neural population models [50 , 61 , 62] , which involved fewer unknown parameters than we have here , and generally only fit a single EEG spectrum . We are thus assuming stationarity of the system over the one-minute EEG signal , where stationarity here means that it is the parameters that are constant; the states are allowed to vary about a stable fixed point . We furthermore assume that deviations of the state away from the fixed point are small enough to allow linearization of the model . Though parameters are assumed to be constant within a given EEG recording they can of course vary between different recordings ( and thus individuals ) . Because of well-known nonlinearities and nonstationarities in EEG recordings , our linearized model was used in inference procedures only for frequencies between 2 Hz and 20 Hz . It is well known that well over 95% of the spectral power in the resting M/EEG falls below 30 Hz . Indeed , typical estimates of resting M/EEG spectral edge frequency ( SEF95 ) ( i . e . the frequency below which 95% of the spectral power is contained ) are in the range of 24-26 Hz ( see e . g . [73–75] ) . To demonstrate the accuracy of our inference methods , we also show an example of parameter estimation from a simulated spectrum where the underlying parameter set is known ( and referred to as the ground truth ) . In order to choose a plausible parameter set for this test , we use the maximum likelihood estimate found for Subject 77 ( the estimate found from any other subject would also have been suitable ) . The simulated spectrum was then calculated by sampling each frequency channel from the gamma-distributed model prediction . To examine the identifiability and sloppiness of the neural population model , we fit to an EEG spectrum and estimate the posterior distribution over the 22 unknown parameters . We then characterize the properties of this distribution to diagnose the signatures of unidentifiability and sloppiness . To ensure that our results are not specific to a particular fitting algorithm , we use two independent methods: particle swarm optimization ( PSO ) and Markov chain Monte Carlo ( MCMC ) . To ensure that our results are not specific to a given individual , we estimate the 22 posterior distributions , using both methods , for each of the 82 different EEG spectra . A full description of the methods for fitting the data and analyzing the results is given in Section “Methods” where we first describe the procedure for calculating the predicted model spectrum , along with the likelihood function for the spectral estimate . We then outline the two fitting schemes , focusing on how we use them to sample from the 22-dimensional posterior distribution . Finally , we describe use of the Kullback-Leibler divergence ( KLD ) to summarize how much we learned about individual parameters , and the Fisher information matrix ( FIM ) , to assess the sloppiness and identifiability of the model . Our implementation of the methods and all datasets are publicly available at https://github . com/cds-swinburne/Hartoyo-et-al-2019-DATA-n-CODE .
Figs 1 and 2 illustrate best fits using two different methods: PSO , which finds the best fits in least squares ( LS ) manner , and MCMC , which samples solutions based on maximum likelihood ( ML ) estimations . Although the fits are generally similar , subtle differences between the two methods can be observed . For example , in subject 72 the the ML fit performs better on regions with lower power but less well in regions with higher power . This difference is expected: while LS are computed over unweighted power spectra , ML favors frequencies with lower variances which typically are those with lower power spectra . The posterior marginal distributions for each parameter , from a few subjects , are shown in blue in Fig 3 ( from PSO sampling ) and Fig 4 ( from MCMC sampling ) . These are compared to the uniform prior distributions ( green ) . The top row , which corresponds to analysis of the simulated spectrum , also shows the ground truth value ( red ) . Each parameter is plotted in normalized coordinates , where -1 corresponds to the lower limit of the plausible parameter interval and +1 corresponds to the upper limit ( see Table 1 ) . Posterior distributions found using PSO sampling are generally broader than those found using MCMC sampling . This behavior is expected from the differences between the sampling methods: while MCMC sampling can retain correlations between samples even with significant subsampling , the different PSO samples are independent from one another . This demonstrates the superiority of the PSO approach , at least under the sampling conditions employed here . Nevertheless , both methods show that it is the postsynaptic potential rate constant of the inhibitory population , γi , which is consistently constrained by the data across different subjects . To better quantify how much we have learned about each parameter , the KLDs for each parameter , from all 82 subjects , are shown in Fig 5 ( for PSO ) and Fig 6 ( for MCMC ) . These confirm that it is γi that is best-constrained by the data . Most other posterior distributions are only slightly narrower than their prior distribution . Furthermore , by analysis of a simulated spectrum ( see Table 2 ) , we find that the γi estimate is accurate as well as precise . Eigenvalues of the Fisher information matrix ( FIM ) are shown on log scale for the selected subjects are shown in Fig 7 ( for PSO ) and Fig 8 ( for MCMC ) , i . e . those computed around LS best fits and ML best fits , respectively . In all cases presented in the figures , the eigenvalues are spread over many decades with approximately uniform spacing over a log scale . This indicates that this neural population model is sloppy [36–41] . Comparison of these eigenspectra across different subjects suggests that there are usually ∼5 identifiable parameter combinations for each subject . The larger FIM eigenvalues define eigenvector directions corresponding to identifiable parameter combinations . To understand the parameters that contribute the most to each ( identifiable ) eigenvector we compute the angular distance between each parameter direction and a given eigenvector . The closer the angular distance to 0° or 180° , the greater the parameter contributes to the parameter combination and thus the more identifiable that parameter . Fig 9 ( LS fits ) and Fig 10 ( ML fits ) show the distributions , across all 82 subjects , of angular distances between each parameter and the three stiffest parameter combinations ( red ) . For a null comparison , the angular distances to vectors randomly pointed in the 22-dimensional space are also shown ( blue ) . For both LS and ML fits , γi again stands out . It has the greatest contribution to the stiffest parameter direction , once again showing that it is identifiable . Interestingly , the postsynaptic potential rate constant of the excitatory population , γe , dominates the third stiffest parameter combination . This indicates that it may also play an identifiable role in driving system dynamics , though to a lesser extent than γi .
Fitting a neural population model to EEG data is an ill-posed inverse problem , where a wide range of parameter combinations are consistent with the observed spectrum . Our approach to fitting such an unidentifiable model is to generate many samples of parameter estimates , all of which give a good fit to the data , and then characterize the structure of these samples . The steps we used can be summarized as follows: Characterization of unidentifiability and sloppiness helps quantify the degree to which a model is over-parameterized . This in turn helps to illuminate how it can be simplified . The existence of correlations between parameter estimates suggests that model complexity can be reduced by grouping together , eliminating , or averaging subsets of parameters . This produces an effective model , with fewer degrees of freedom , without compromising predictive ability . A number of model reduction techniques have been proposed for dynamical systems , such as balanced truncation [76–78] , singular perturbation [79] , and the manifold boundary approximation [80 , 81] . In physical theory , model reduction techniques such as mean field and renormalization group methods [82] have long been used to quantify the effective parameters in complex physical systems . The concept of entropy , which enumerates the number of ( unidentifiable ) microstates that are consistent with a single ( observable ) macrostate , can be thought of as a measure of unidentifiability . Our finding that there are only ∼5 identifiable eigenvalues in the FIM spectrum indicates that the number of effective parameters in our model is only about 5 . The challenge is to understand what these effective parameters mean physiologically . Some insight into their role can be obtained by examining the derivatives of the modelled spectra in the directions of the leading eigenvectors of the Fisher information matrix [83] ( See S3 Appendix for details ) . In broad terms , for most subjects , it appears that the leading eigenvector ( effectively the inhibitory decay rate γi ) is related to variation in the location of the alpha peak; the second eigenvector is related to the height of the alpha peak compared to the overall background level and ( somewhat more loosely ) the third eigenvector is associated with the width of the alpha peak . The remaining combinations do not appear to be related to easily identifiable features of the spectrum . Importantly , although each eigenvector could have contributions from all 22 of the original parameters , the leading 3 eigenvectors appear to be influenced by just a few of these ( See S5 Fig for details ) . This indicates that , even though γi is the only parameter that is clearly identifiable , a subset of the other parameters is being constrained as well . This suggests that the essential character of the alpha peak —its position , height and width—is determined by only a few of the original parameters . Neural population models are coarse-grained approximations to networks of single neurons . When trying to interpret a measurement such as the EEG , our results show that even neural population models are not coarse-grained enough , since most parameters are unidentifiable . The fact that only one of the original parameters , out of 22 , is consistently identifiable , a result confirmed by comparisons over 82 subjects and two different fitting routines , would seem to be a bleak result: despite the considerable effort required to fit the model , we appear to still be ignorant of 21 of the 22 parameters . However , when fitting a nonlinear model with many parameters , there is no guarantee that any of them will be identifiable . The fact that one has been found hints that it has a special role . This has parallels in physical systems where the effective model parameters are the ones that remain identifiable under coarse-graining . For example , it has been shown [39] that in diffusion processes and magnetic phase transitions , most of the microscopic parameters become unidentifiable at macroscopic scales , with only parameters such as the diffusion coefficient and average magnetization emerging unscathed . It has been suggested [84] that there may exist organizing principles that create ‘protectorates’ at mesoscopic scales , corresponding to particular parameters or parameter combinations that are robust to coarse-graining . The suggestion here then is that γi is an effective parameter in neural population models , one that plays a central role in generating and modulating the alpha-rhythm in cortex . This discovery has potential clinical significance since the majority of agents used to induce a state of surgical anaesthesia are thought to function by altering the time course of postsynaptic inhibition . Being able to determine γi , the postsynaptic inhibitory rate constant , by fitting to EEG data could provide real-time physiological insights into the functional effect of anaesthesia , improving upon standard signal processing methods [85] which are not built on an underlying theory of brain dynamics . We note that it is likely that the alpha band activity obtained from a single electrode ( in our case Oz ) represents the superposition of multiple independent , spatially distributed , alpha rhythm generators [86–88] . In future , it may be possible to differentiate between these different sources by jointly fitting the model to signals from multiple electrodes . We conclude by remarking that there are deep parallels between model identifiability , dynamic compensation [89–91] and evolvability [92] in a dynamical system . If the function of the system is robust , or insensitive , to changes in some of its underlying parameters , it can be impossible to infer those parameters by studying functional observables alone . Thus the study of identifiability and sloppiness is not simply a study of fitting problems but is also an examination of which parameter values are functionally essential and which are not .
The model spectrum is calculated from the spatially homogeneous version of the full model equations [11] . We make the additional assumptions that: Under these assumptions , it can be shown that the linear system transfer function , T ( s ) , is ( to within an overall sign ) that of a simple feedback system as shown in Fig 11 involving two third order filters: T ( s ) = H 1 ( s ) 1 + H 1 ( s ) H 2 ( s ) , ( 8 ) H 1 ( s ) = k 13 k 13 k 31 - k 11 ( s ) k 33 ( s ) , ( 9 ) H 2 ( s ) = k 15 k 24 k 41 k 52 k 13 { k 22 ( s ) k 55 ( s ) - k 26 k 62 } . ( 10 ) The polynomials k11 ( s ) and k22 ( s ) are linear in s and k33 ( s ) and k55 ( s ) are quadratic in s . The derivation of this result and detailed expressions for the factors appearing in these equations are given in S1 Appendix . Given that the spectra are assumed to arise from a white noise spectrum filtered by this transfer function , the expected value of the spectral estimate at frequency ω , given a vector of model parameters ( θ ) , has the form: ⟨ S ( ω ) ⟩ = α S ^ ( ω | θ ) = α | H 1 ( i ω ) 1 + H 1 ( i ω ) H 2 ( i ω ) | 2 , ( 11 ) where the constant α accounts for the unknown driving amplitude and for attenuation due to volume conduction and other ( frequency-independent ) effects . The value for α is found using a least-squares fit to the measured spectral estimates . The analytic result is: α = ∑ n S n S n ^ ( θ ) ∑ n S ^ n 2 ( θ ) , where S n ≡ S ( ω n ) ; n = 1 , … , N . ( 12 ) For a spectrum of the form described above , with sufficiently high sampling rates and negligible measurement noise , the spectral estimate from the Welch periodogram at each sampled frequency {ωn = 2π fn; n = 1 , 2 , 3 , … , N} is approximately an independent random variable with a known distribution [96] . The exact form is computationally involved and for our immediate purposes we will ignore the effects of window overlap and non-uniform window shape on the resulting distributions . With this simplification the probability distribution function ( pdf ) for the spectral estimate , Sn , is a gamma distribution: f ( S n ; K , Θ ) = S n K - 1 e - S n θ n Θ n K Γ ( K ) ; S n > 0 . ( 13 ) Here , for non-zero frequencies , the shape parameter K is found from the number of epochs averaged in the periodogram . For zero frequency , replace K with K/2 throughout . The scale parameter is given by Θ n ( θ , α ) = α S ^ n ( θ ) K , where S ^ n ( θ ) ≡ S ^ ( ω n | θ ) ; n = 1 , ⋯ , N . ( 14 ) The likelihood function for the vector of spectral estimates , S = [S1 S2 S3 ⋯ SN]T , given the parameter values θ , is then the product of the distributions of the individual spectral estimates: p ( S | θ , α ) = ( 1 Γ ( K ) ) N ( ∏ n ( S n S ^ n ( θ ) ) K - 1 S ^ n ( θ ) ) ( K α ) N K exp ( - K α ∑ n S n S ^ n ( θ ) ) . ( 15 ) The constant α is adjusted to give the maximum likelihood fit of the model spectrum to the target experimental spectrum . The analytic result is that α = 1 N ∑ n = 1 N S n S ^ n ( θ ) . ( 16 ) The likelihood based on model parameters alone is then p ( S | θ ) = ( K K e - K Γ ( K ) ) N ( 1 1 N ∑ n = 1 N S n S ^ n ( θ ) ) N K ( ∏ n ( S n S ^ n ( θ ) ) K - 1 S ^ n ( θ ) ) . ( 17 ) A convenient measure of the information gained about individual parameters as a result of the measurement of the spectrum is the Kullback-Leibler divergence ( KLD ) [100 , 101] . Here we use the KLD to measure the change in the marginal posterior distribution of each parameter relative to its marginal prior: D K L ( i ) ( S ) = ∫ p i ( θ i | S ) ln p i ( θ i | S ) p 0 ( θ i ) d θ i . ( 22 ) When KLDs are used to measure how posteriors differ from priors based on MCMC samples , the integral is numerically evaluated using marginal distributions approximated by kernel density estimates using 1000 parameter values resampled from the full MCMC sampled parameter set for the given spectrum . The prior distributions are uniform over their support . For consistency , the posterior kernel estimates are truncated to have the same support . The kernel density estimate for a given parameter is sampled at 100 points over its support and the integral estimated numerically . For the PSO samples , due to the limited number of independent samples , the integral is estimated using a 10 bin histogram approximation . To assess the sloppiness of the model fit , we examine the eigenvalues of the Fisher information matrix ( FIM ) , the definition of which for the pdf P ( S|θ ) is given by: I μ ν ( θ ) = ∫ P ( S | θ ) ∂ ln P ( S | θ ) ∂ θ μ ∂ ln P ( S | θ ) ∂ θ ν d N S ( 23 ) In general the integration here could present considerable difficulty , however , for the distribution given by Eq ( 13 ) , it can be shown that a simplification is possible , resulting in an expression involving only the derivatives of the model spectral estimates , evaluated at the desired parameter values: I μ ν ( θ ) = K ∑ n ∂ ln S ^ n ( θ ) ∂ θ μ ∂ ln S ^ n ( θ ) ∂ θ ν ( 24 ) ( For a derivation of this result , see S2 Appendix . ) The derivatives , with respect to normalised parameters at the LS or ML estimated values , are evaluated numerically using a 5-point finite difference approximation , and the resulting products summed over the sampled frequencies . The Matlab® eig command is used to find the eigenvalues and eigenvectors of the resulting matrices . Numerical experiments with surrogate matrices suggest that the eigenvalues calculated using eig are reliable over some 10 orders of magnitude . For our modelled spectra we expect the FIM to be positive semidefinite and of less than full rank , so negative eigenvalues and eigenvalues smaller than 10−10 times the largest eigenvalue are taken as zero . | Electroencephalography ( EEG ) , where electrodes are used to measure electric potential on the outside of the scalp , provides a simple , non-invasive way to study brain activity . Physiological interpretation of features in EEG signals has often involved use of collective models of neural populations . These neural population models have dozens of input parameters to describe the properties of inhibitory and excitatory neurons . Being able to estimate these parameters by direct fits to EEG data holds the promise of providing a real-time non-invasive method of inferring neuronal properties in different individuals . However , it has long been impossible to fit these nonlinear , multi-parameter models effectively . Here we describe fits of a 22-parameter neural population model to EEG spectra from 82 different subjects , all exhibiting alpha-oscillations . We show how only one parameter , that describing inhibitory dynamics , is constrained by the data , although all parameters are correlated . These results indicate that inhibitory synaptic activity plays a central role in the generation and modulation of the alpha-rhythm in humans . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"markov",
"models",
"membrane",
"potential",
"brain",
"electrophysiology",
"electrophysiology",
"neuroscience",
"clinical",
"medicine",
"systems",
"science",
"mathematics",
"physiological",
"parameters",
"algebra",
"brain",
"mapping"... | 2019 | Parameter estimation and identifiability in a neural population model for electro-cortical activity |
DNA double-strand breaks ( DSBs ) are formed during meiosis by the action of the topoisomerase-like Spo11/Rec12 protein , which remains covalently bound to the 5′ ends of the broken DNA . Spo11/Rec12 removal is required for resection and initiation of strand invasion for DSB repair . It was previously shown that budding yeast Spo11 , the homolog of fission yeast Rec12 , is removed from DNA by endonucleolytic cleavage . The release of two Spo11 bound oligonucleotide classes , heterogeneous in length , led to the conjecture of asymmetric cleavage . In fission yeast , we found only one class of oligonucleotides bound to Rec12 ranging in length from 17 to 27 nucleotides . Ctp1 , Rad50 , and the nuclease activity of Rad32 , the fission yeast homolog of Mre11 , are required for endonucleolytic Rec12 removal . Further , we detected no Rec12 removal in a rad50S mutant . However , strains with additional loss of components localizing to the linear elements , Hop1 or Mek1 , showed some Rec12 removal , a restoration depending on Ctp1 and Rad32 nuclease activity . But , deletion of hop1 or mek1 did not suppress the phenotypes of ctp1Δ and the nuclease dead mutant ( rad32-D65N ) . We discuss what consequences for subsequent repair a single class of Rec12-oligonucleotides may have during meiotic recombination in fission yeast in comparison to two classes of Spo11-oligonucleotides in budding yeast . Furthermore , we hypothesize on the participation of Hop1 and Mek1 in Rec12 removal .
Meiosis is a special type of cell division generating haploid gametes . One round of DNA replication is followed by two successive divisions separating homologous chromosomes ( Meiosis I ) and sister chromatids ( Meiosis II ) . Recombination during meiotic prophase ensures the formation of physical connections , called chiasmata , between the homologous chromosomes , indispensable for proper homolog separation in most species . In prophase , before the first meiotic division , homologous chromosomes become aligned and recombination is initiated by the topoisomerase-like protein Spo11 in Saccharomyces cerevisiae , which catalyzes double strand break ( DSB ) formation by a transesterification reaction [1] . Afterwards Spo11 remains covalently attached to each 5′ end of the meiotic DSB [2] , [3] . An amino acid change of tyrosine-135 to phenylalanine abolishes covalent linkage of Spo11 with DNA [1] . This covalent linkage is evolutionarily conserved in the Schizosaccharomyces pombe Spo11 homolog Rec12 ( tyrosine-98 , [1] , [4] ) . True topoisomerase activity ( religation of DNA ) has not been demonstrated for Spo11 . Instead , Spo11 was shown to be excised from the DNA by endonucleolytic cleavage , resulting in free 5′ ends accessible for further strand resection [5] . The released Spo11-oligonucleotide consisted of two classes of oligonucleotides heterogeneous in length . As the most straightforward explanation for these two classes of oligonucleotides , Neale et al . hypothesized asymmetric cleavage . The remaining distinct DNA ends would present different loading platforms for recombination proteins , e . g . Rad51 and Dmc1 [5] . DSB formation does not exclusively depend on Spo11/Rec12 action , but requires multiple auxiliary proteins in both yeasts ( reviewed in [6] ) . The evolutionarily conserved Mre11/Rad50/Nbs1 ( MRN ) complex , in S . cerevisiae termed MRX ( Mre11/Rad50/Xrs1 ) , has a central role in mitotic and meiotic DNA repair ( reviewed in [7] ) . Any null mutant of the MRX complex in S . cerevisiae abolishes meiotic DSB formation [8]–[10] . Point mutations in MRE11 or RAD50 , e . g . mre11S or rad50S , accumulate unrepaired DSBs with Spo11 still covalently bound to the DNA [2] . Therefore the MRX complex is required not only for DSB formation , but also for removing Spo11 from meiotic DSB ends . In addition Sae2/Com1 was shown to be required for DSB repair , but not for DSB formation [11] , [12] . The sae2Δ mutant shows a phenotype similar to rad50S . In-vitro experiments suggested a cooperative action of Sae2 and the MRX complex [13] . In S . pombe the MRN complex is not required for DSB formation [14] . A point mutation , also called rad50S in S . pombe , accumulates DSBs , resulting in low spore viability , which can be partially rescued by deletion of rec12 [15] , [16] . Similar to sae2Δ in S . cerevisiae , deletion of the S . pombe homolog ctp1 does not affect DSB formation [17] , but abolishes Rec12 removal from DNA , like the nuclease dead mutant of Rad32 ( Rad32-D65N ) [16] . Besides proteins directly involved in DSB formation , other meiosis-specific proteins affect DSB formation in S . cerevisiae , among them Red1 , Mek1 , and Hop1 , which associate with lateral elements of the synaptonemal complex ( SC ) [18] , [19] . The SC is formed in S . cerevisiae and many other eukaryotes to mediate homologous chromosome pairing . In S . pombe proteinaceous structures , called linear elements ( LEs ) , are formed instead [20] , [21] . S . cerevisiae MEK1 and HOP1 genetically interact with RED1 [22] . Meiotic intergenic ( 7 to 60 fold ) and intragenic ( 4 to 25 fold ) recombination are reduced in mek1 and hop1 mutants , depending on the interval [23] , [24] . Loss of both proteins lead to DSB reduction of 5 to 15% compared to wild type [25]–[27] . Unequal recombination between sister chromatids is increased in mek1 and red1 mutants , suggesting a participation of these three proteins in a “barrier to sister chromatid repair” during meiotic recombination [28] . S . pombe Hop1 and Mek1 localize to Rec10 , a main component of the LEs and the distant homolog of S . cerevisiae Red1 [21] , [29] . As in S . cerevisiae , Mek1 was identified as a meiosis-specific kinase participating in the regulation of meiotic cell cycle progression in fission yeast [29] . Intergenic recombination was reduced in a mek1 mutant compared to wild type [29] . Here we present evidence for the removal of S . pombe Rec12 from DNA by endonucleolytic cleavage . Unlike in S . cerevisiae , S . pombe Rec12-oligonucleotides were found to be homogeneous in length , which may indicate symmetric cleavage . Rec12 removal depends on Ctp1 , Rad50 , and the nuclease activity of Rad32 . Furthermore , we present evidence that Rec12 removal in a rad50S mutant can be partially restored by deletion of hop1 or mek1 with subsequent DSB repair .
pat1-114 meiosis has the advantage that cells enter and proceed through meiosis in a highly synchronous way [30] . Although pat1-114 meiosis leads to chromosome segregation defects in the first meiotic division [31] , it has been used systematically for studying the timing of early meiotic events , including DSB formation and repair [4] , [15] . Strains carrying the pat1-114 mutation were arrested in G1 by nitrogen starvation . A temperature shift to 34°C induces meiosis . Meiotic DSB formation starts at 3 hours after meiotic induction , reaches a maximum at 4 hours , and ends at 5 hours [32] . Rec12 expression was monitored in the pat1-114 meiosis by Western blot analysis . Pre-meiotic DNA replication and completion of meiotic division was checked by FACS analysis and DAPI staining ( Figure S1 , data not shown ) . As expected , myc epitope-tagged Rec12 taken four hours after meiotic induction migrated at 68 kDa ( Figure 1A , right panel ) . A more slowly migrating protein at 75 kDa was also detected . Presumably , the 68 kDa protein corresponded to free Rec12 . We examined whether the 75 kDa protein could be a removed Rec12-oligonucleotide compound , like it was found in budding yeast [5] . Terminal transferase ( TdT ) catalyzes the addition of deoxynucleotides to 3′ hydroxyl termini of DNA molecules . Treating immmunoprecipitated Rec12 with TdT together with a chain termination desoxynucleotide analog ( α-32P-cordycepin ) revealed that the 75 kDa protein is a Rec12-oligonucleotide ( Figure 1A , left panel ) . In a Rec12Y98Fmyc active site mutant no DSB formation occurs [4] and no Rec12Y98F-oligonucleotide appeared . Rec12-oligonucleotide appearance was specific for Rec12myc and anti-myc antibody ( Figure 1B ) . Neither untagged Rec12 nor Rec12myc immunoprecipitation with an anti-HA antibody led to the detection of Rec12-oligonucleotide . In a rad50S mutant , where DSBs are not repaired [15] and Rec12 stays covalently attached to the DNA , no Rec12-oligonucleotide was detectable ( Figure 1B ) . Furthermore , Rec12-oligonucleotide was not detectable in rad50Δ , ctp1Δ , and rad32-D65N nuclease dead mutant ( Figure 1C ) . This reveals that Ctp1 and the MRN complex are required for removing Rec12 from the DNA . Investigation of meiotic DSB formation by pulsed-field gel electrophoresis ( PFGE ) and Rec12-oligonucleotide appearance in the same pat1-114 meiotic culture revealed similar timing of these events ( Figure 2 ) . DSBs became visible at around 3 hours after meiotic induction and reached a maximum at 4 hours . During further meiotic progression , DSBs became repaired and intact chromosomes reappeared ( Figure 2A ) . Shortly after the start of DSB formation Rec12-oligonucleotide appeared , reached a maximum at 4 to 5 hours , and disappeared at later time points ( Figure 2B ) . Rec12myc expression was followed by immunodetection in dot-blot analysis ( Figure 2C ) . The mass difference of about 7 kDa between free Rec12 and Rec12-oligonucleotide seen in the Western blot analysis ( Figure 1A , right panel ) , suggested an oligonucleotide length of approximately 23 nucleotides ( mean nucleotide mass of 308 Da ) . To elucidate further the length of this oligonucleotide , immunoprecipitated Rec12 extract was deproteinized with pronase after TdT labeling and separated on a 20% denaturing polyacrylamid gel ( Figure 3 ) . Oligonucleotides of known length were also labeled with TdT , and used as size markers . We detected a signal between 17 to 27 nucleotides ( Figure 3 ) , which is in accordance with the results from Western blot analysis and autoradiography ( Figure 1A ) . The signal was specific for the myc-tag and Rec12 activity ( Figure 3 ) . Thus , Rec12 is removed from the DNA by endonucleotic cleavage , which releases Rec12 bound oligonucleotides with a mean length of about 22 nucleotides . Investigations of DSB formation and repair by PFGE in a rad50S hop1Δ mutant indicated partial DSB repair ( our unpublished data ) . Direct analysis of Rec12 oligonucleotide appearance in time-course experiments of rad50+ and rad50S strains in combination with hop1Δ , mek1Δ , or hop1Δ mek1Δ , would clarify whether this partial repair was actually occurring by Rec12-oligonucleotide removal as in hop1+ ( Figure 4 ) . Radiolabeled Rec12-oligonucleotide was analyzed after separation with a PhosphorImager and quantified by the AIDA software . The amount of Rec12-oligonucleotide at each time point was normalized to a standard meiotic sample N ( see Materials and Methods ) . The summation of Rec12-oligonuclotide values observed throughout the time-course in the pat1-114 strain was defined as 100% . Values of Rec12-oligonucleotide in mutant strains were estimated accordingly . We used equal amounts of cells for the IP experiments ( see Materials and Methods ) . Furthermore , the wild type and mutant strains used for the experiments in Figure 4 , showed overall Rec12myc abundance with similar kinetics and amounts throughout the time-courses ( dot blots , see Figure S2 ) . Rec12-oligonucleotide release occurred in the mutant strains with comparable timing to the pat1-114 strain and started at 3 hours , reached a maximum at 4 to 5 hours , and diminished afterwards . Less Rec12-oligonucleotide was detected in mek1Δ ( 70% ) , hop1Δ ( 53% ) , and hop1Δ mek1Δ ( 22% ) mutants ( Figure 4A and 4B , left panels ) . As shown above , in a rad50S mutant no Rec12-oligonucleotide was detectable . Remarkably , Rec12-oligonucleotides were detectable to 7% and 3% in strains carrying mek1Δ or hop1Δ , respectively ( Figure 4A and 4B , right panel ) . No Rec12-oligonucleotide was detectable in a rad50S hop1Δ mek1Δ mutant . These findings indicate that in the rad50S mutation absence of the LE associated proteins Hop1 or Mek1 leads to at least some removal of Rec12 from the chromatide fragments . In the rad50S mutant , where unrepaired DSBs accumulate [15] , low spore viability was detected [16] . We confirmed the low spore viability at the restrictive temperature of the rad50S mutant ( 6% , Table 1 ) . Surprisingly , the spore viability was increased in homozygous crosses of rad50S mek1Δ ( 17% ) and rad50S hop1Δ ( 20% ) in comparison to rad50S . The rad50S mek1Δ hop1Δ mutant showed rad50S single mutant level ( 8% ) . Furthermore , we studied intragenic recombination ( gene conversion ) in the non-hotspot interval ade7-50×ade7-152 at the restrictive temperature of 34°C for rad50S using the same material from the crosses for spore viability determination ( Table 1 ) . Notably , conversion frequency in the homozygous rad50S crosses was reduced about 70-fold compared to wild type . rad50S mek1Δ and rad50S hop1Δ mutants were only 40-fold reduced compared to wild type , whereas the conversion frequency in the rad50S mek1Δ hop1Δ crosses was comparable to the rad50S value . To elucidate whether Ctp1 and the MRN-complex contribute to the observed Rec12 removal in rad50S hop1Δ and rad50S mek1Δ , further mutants were analysed . There was no Rec12-oligonucleotide detectable in rad50S hop1Δ ctp1Δ , rad50S mek1Δ rad32-D65N , hop1Δ ctp1Δ , or mek1Δ rad32-D65N ( Figure 5 ) . Therefore , it seems that Ctp1 and the nuclease activity of Rad32 are required for removal of Rec12 from DNA in rad50S hop1Δ , rad50S mek1Δ , as well as in rad50+ ( Figure 1C ) . Moreover , neither deletion of hop1 nor mek1 is able to restore Rec12 removal in ctp1Δ or rad32-D65N; the suppression is specific for rad50S .
Is was hypothesized that S . cerevisiae Spo11 is removed by asymmetric cleavage , resulting in two classes of oligonucleotides , heterogeneous in length ( 10–15 and 24–40 nucleotides , respectively ) . Here we show that fission yeast Rec12 is removed from DNA by endonucleolytic cleavage , releasing one class of Rec12-oligonucleotide only ( Figure 1 and Figure 3 ) . Appearance of this repair byproduct correlates with DSB presence ( Figure 2 ) . The length of covalently bound DNA to Rec12 is in the range of 17 to 27 nucleotides ( Figure 3 ) , which is larger than the short class ( 10 to 15 nucleotides ) , but not as large as the longer class ( 24 to 40 nucleotides ) of Spo11 bound oligonucleotides in S . cerevisiae . The signal of TdT-labeled oligonucleotides suggested one class with an average length of 22 nucleotides in S . pombe ( Figure 3 ) , which is in accordance with the more slowly migrating Rec12-oligonucleotide in Western blot analysis ( Figure 1A ) . This interpretation is strengthened by the single peak profile of the oligonucleotide signal ( Figure 3 ) . However , we cannot ascertain that the released oligonucleotides fall into two discrete length classes with mean lengths differing by only few nucleotides . During submission of this paper an analogous study investigating the removal of Rec12 after DSB formation in S . pombe has been published [33] . Although the experimental setting slightly differs , one class of oligonucleotides with a mean length of 22 nucleotides was found as well . A single class of oligonucleotides would lead to the prediction that Rec12 is , contrary to Spo11 in S . cerevisiae and mouse [5] , removed by symmetrical cuts from the chromatid fragments . What causes and consequences may symmetric Rec12 removal have for subsequent DSB processing ? For initiation of meiotic DSB formation , several meiosis-specific proteins are required . In S . cerevisiae , several protein sub-complexes assemble and load in a regulated fashion at emerging DSB sites [34] , [35] . In addition , the MRX complex ( Mre11/Rad50/Xrs1 ) loads independently of Spo11 to meiotic chromatin [36] and is indispensable for DSB formation [8]–[10] . In S . pombe , only few of the S . cerevisiae proteins required for DSB formation are conserved . Instead , several S . pombe specific proteins have been identified . Sterical constraints based on asymmetric cleavage present in S . cerevisiae may thus not exist in S . pombe . Neale et al . suggested the following model for the presumed asymmetric Spo11 removal in S . cerevisiae: the shorter Spo11-oligonucleotide would be released for immediate resection , whereas the longer one would remain basepaired leading to a delay of resection . This creates timely ordered accessibility of the 3′OH ends , facilitating an orchestrated preferential strand invasion into a chromatid of the homologous chromosome . Rad51 and Dmc1 could be loaded differentially on the distinct DNA ends , in accordance with the finding of side-by-side formation of Rad51 and Dmc1 foci [37] . Physical analysis of joint molecules ( JMs ) in budding yeast , revealed that the majority of JMs are formed between homologous chromosomes [38] , [39] . Furthermore , Dmc1 specifically promotes interhomolog JMs [38] . In addition , the dmc1 mutant shows a meiotic arrest with extensively resected DSB ends [40] . In fission yeast , however , sister chromatid JMs outnumber homolog JMs by a factor of four [41] . The S . pombe dmc1Δ mutant forms and repairs DSBs , and meiotic prophase is not arrested [14] . Weak reductions of recombination in the single mutants , but strong reduction in rad51Δ dmc1Δ , indicates redundant meiotic functions of Dmc1 and Rad51 [42] . A differential loading of Rad51 and Dmc1 may not occur in fission yeast , and thus an asymmetric Rec12 removal would not be necessary . It has been suggested , that Sae2/Com1 , as well as the Rad50 and Mre11 components of the MRX complex , are required for Spo11 removal in S . cerevisiae [5] . However , since these proteins are also required for DSB formation , investigation of their actual requirement for Spo11 removal from the DNA is difficult in this organism . On the contrary , the S . pombe MRN complex ( Rad32/Rad50/Nbs1 ) is not required for DSB formation [14] , but obviously Rad50 , as well as Ctp1 and Rad32 endonuclease activity are essential for Rec12 removal ( Figure 1C ) . These findings are in accordance to experiments showing Rec12 retention on chromosomal DNA in ctp1Δ or rad32D65N nuclease dead mutant [16] . Thus , we conclude that Ctp1 and Rad32 are key players in the removal of Rec12 from the DNA . The Rec12-oligonucleotide amounts in hop1Δ , mek1Δ , and hop1Δ mek1Δ strains were reduced ( see Figure 3A ) . Probably , this reduction of repair byproduct is due to reduced DSB formation . Since DSB formation and repair are simultaneous processes , full quantification of DSB formation is not possible in these mutants . Rec7 is required for meiosis specific DSB formation and aggregates in foci on LEs in immuno-stained nuclear spreads [43] . The amount of Rec7 foci in hop1Δ and mek1Δ mutants was reduced , suggesting less Rec7 cooperation with Rec12 for DSB formation [43] . Furthermore , Rad51 foci , marking DSB repair sites , were also reduced in hop1Δ and mek1Δ [43] . Taken together , these findings strengthen the argument for reduction of DSB formation in these mutants . However , alternative Rec12 removal that does not involve endonucleolytic cleavage , for instance involvement of a phosphodiesterase like Tdp1 , cannot be excluded . The S . cerevisiae homolog of Tdp1 has been shown to participate in Top2 removal , which like Spo11/Rec12 is bound to the 5′ ends of DNA [44] . We did not detect Rec12-oligonucleotide in a rad50S mutant ( Figure 4 ) . This finding is in agreement with the described accumulation of unrepaired DSBs and low spore viability [15] , [16] . Surprisingly , Rec12-oligonucleotide was detectable in low amounts in rad50S mek1Δ and rad50S hop1Δ . In addition , spore viability and intragenic recombination in both mutants were increased compared to rad50S ( Table 1 ) . Together these findings indicate partial restoration of Rec12 removal in rad50S by hop1Δ or mek1Δ deletion . Furthermore , we were able to show that the observed suppression depends on Ctp1 and Rad32 nuclease activity , like Rec12 removal in rad50+ . Obviously , once Rec12 is removed , DSB repair is possible resulting in elevated intragenic recombination and the increased spore viability . No Rec12-oligonucleotide was detectable in rad50S hop1Δ mek1Δ . We interpret this finding differently than the one in rad50S . The amount of Rec12-oligonucleotide detected in the rad50+ mek1Δ hop1Δ mutant was reduced about 2-fold compared to rad50+ hop1Δ ( see above ) . Assuming a 2-fold reduction in rad50S mek1Δ hop1Δ compared to rad50S hop1Δ , the Rec12-oligonucleotide amount would probably be below the detection limit . In S . cerevisiae , genetic interactions between components associated with the SC's lateral elements ( Mek1 , Hop1 , Red1 ) , Sae2 , and the MRX complex have been reported: Firstly , a hop1Δ mutant suppresses the rad50S spore lethality [26] , and secondly , a hop1 and red1 deletion suppresses a sae2Δ sporulation defect [45] . In-vitro experiments validate the cooperative action between the MRX complex and Sae2 [13] , [46] . Furthermore , co-localization of Sae2 and Mre11 in nuclear spreads is impaired in the rad50S mutant , suggesting a role of Rad50 in mediating the interaction between the MRX complex and Sae2 in meiosis [47] . Recently , S . pombe Ctp1 and the nuclease activity of Rad32 ( homologs of S . cerevisiae Sae2 and Mre11 ) were shown to be required for Rec12 removal from the DNA , suggesting conserved action [16] . The S . pombe rad50S mutation is , in contrast to the analogous mutation in budding yeast , a temperature sensitive allele . At the restrictive temperature ( 34°C ) the activation of the endonuclease needed for the removal of Rec12 is probably impaired . This might be either due to a conformational change within the M ( Rad50S ) N complex and/or lost interactions with associated proteins , e . g . Ctp1 . How might a deletion of Mek1 or Hop1 provoke endonuclease activity in the rad50S background ? The association of Hop1 and Mek1 with the LEs might indicate their participation in meiosis-specific chromatin organization to promote chromosome pairing . The endonuclease activity required for removing Rec12 from the broken DNA is , as described above , provided by the cooperative action of the MRN complex and Ctp1 ( Figure 1C ) . We suggest that the presence of Hop1 and Mek1 in a rad50S mutant impairs the interaction between the MRN complex and Ctp1 , and thus Rec12 remains covalently bound to the DNA at the site of DSB . Absence of Hop1 or Mek1 would then loosen the chromatin organization and advance the encounter of MRN and Ctp1 , causing some Rec12 removal . Obviously , the same proteins are responsible for Rec12 removal in wild type , rad50S hop1Δ and rad50S mek1Δ meiosis since no Rec12 removal was detectable in rad50S hop1Δ ctp1Δ or rad50S mek1Δ rad32D65N triple mutants ( Figure 5 ) . In summary , the single class of Rec12-oligonucleotide described adds to the list of differences between the two model organisms S . pombe and S . cerevisiae . Perhaps , this also occurs in other eukaryotes , particularly in those that do not possess Dmc1 homologs , e . g . Drosophila melanogaster and Caenorhabditis elegans [48] . We have demonstrated genetic interaction between hop1 and rad50S in S . pombe . In addition , we found a genetic interaction between mek1 and rad50S . In both cases , the partial restoration of Rec12-oligonucleotide formation as a byproduct of DSB processing does occur . Moreover , we were able to shed light on the participation of Ctp1 and Rad32 in Rec12 removal from DNA ends after DSB formation . Rec12 removal seems to be achieved by joint action of proteins , including Ctp1 , Rad50 , and Rad32 . Both potential endonucleases , Ctp1 and Rad32 , are required for cleavage . Until now , PFGE was used to monitor global DSB formation and repair . By measuring the appearance of the Rec12-oligonucleotide byproduct as presented here , a new tool for monitoring global DSB processing is now available in S . pombe .
Media and general methods are described in [49] , [50] . Synthetic medium , EMM with 2% glucose and EMM without nitrogen source ( EMM-N ) with 1% glucose were described in [51] . 75 and 15 mg/liter adenine were added to EMM and EMM-N . Meiotic time-course experiments were carried out as described elsewhere [30] . Samples were checked by FACS analysis and DAPI staining for synchronicity and completion of meiosis ( see Figure S1 ) . S . pombe strains are listed in Table S1 . The rec12Δ::hygR deletion was constructed according to the Baehler method [52] . Rec12myc functionality has been tested previously [53] . The rec12Y98Fmyc allele was constructed as follows: an integrative plasmid ( pYC36A , [31] ) , containing 2 kb of 5′ non-translated promoter region of rec12 , the rec12 ORF , and 13 copies of the myc epitope tags ( from the pFA6a series [52] ) , was targeted for site-directed mutagenesis ( QuikChange Lightning Site-directed mutagenesis Kit from Stratagene ) with the primers #KL241 ( 5′-CGAAAATTCAGATTTAAGTTCTAATTACATTTGCAGAGATATCTATTTCAGAGATGTAGATTTATTCAAG-3′ ) and #KL242 ( 5′-CTTGAATAAATCTACATCTCTGAAATAGATATCTCTGCAAATGTAATTAGAACTTAAATCTGAATTTTCG-3′ ) , introducing an Eco RV restriction-recognition site together with the amino acid change tyrosine to phenylalanine at position 98 . The plasmid was stably integrated at the lys1 locus by homologous recombination of the truncated lys1-N gene on the plasmid and the lys1-131 point mutation in a rec12Δ::hygR deletion strain , and correct integration confirmed by sequencing . Parental strains for crosses were grown in YEL plus supplements , cell material was mixed and plated on MEA plus supplements , followed by incubation for three days at 34°C ( involving the temperature-sensitive allele rad50S ) . The cross material was then suspended and treated with glusulase [1∶1000 ( v/v ) Helix pomatia juice , Biosepra] solution over night at 30°C . 0 . 1 ml undiluted or ten-fold diluted spore suspension was plated onto YEA plus supplements . The plates were incubated at 30°C for at least 20 hours . Afterwards the plates were inspected under the microscope of a Singer tetrad dissection apparatus ( allows systematic inspection of non-overlapping fields of view ) [54] . Two classes of spore fates were distinguished and quantified: one spore/cell and 2–4 cells ( no division , no further growth ) , in comparison to >4 cells ( microcolonies ) . Only the latter contributed to spore viability . 50 ml cells of OD595 0 . 8 were harvested from meiotic cultures at the indicated time points ( hours after temperature shift to 34°C ) , washed with water and disrupted with glass beads in 400 µl lysis buffer ( 50 mM HEPES , 140 mM NaCl , 1 mM EDTA , 1% Triton-X100 , 0 . 1% ( w/v ) Na deoxycholate , 1 mM Pefabloc SC ( Roche ) , 0 . 5% ( v/v ) Pefabloc SC protector solution , Complete Protease Inhibitor tablet ( Roche ) ) using the FastPrepMachine ( Bio101 ) at level 6 , three times 30 seconds with at least 2 minutes cooling on ice in between . The crude extract was centrifuged for 10 minutes at 13 , 000 rpm in a microcentrifuge at 4°C and the supernatant was centrifuged again for 5 minutes . [53] For each immunoprecipitation ( IP ) sample , 30 µl Dynabeads Protein G ( Invitrogen ) were pre-coated in 50 µl 0 . 1 M citrate-phosphate buffer with 6 µg monoclonal anti-myc antibody ( 9E10 , SantaCruz ) or anti-HA ( 12CA5 , Roche ) , agitated at 22°C . IPs were carried out on a rotating wheel ( 4°C , 2 hours ) by adding the cleared extract directly to the pre-coated beads . Immune complexes were collected with a magnet , washed twice with 1× NEB4 buffer ( 50 mM K-acetate , 20 mM Tris-acetate , 10 mM Mg-acetate , 1 mM DTT , NewEnglandBiolabs ) and incubated in 50 µl 1× terminal transferase labeling buffer ( 50 mM K-acetate , 20 mM Tris-acetate , 10 mM Mg-acetate , NewEnglandBiolabs ) with 2 . 5 mM CoCl2 ( NewEnglandBiolabs ) , 40 U terminal transferase ( TdT , NewEnglandBiolabs ) and 8 µCi [alpha 32P] cordycepin triphosphate ( 5000 Ci/mmol , Perkin Elmer ) for 2 hours at 37°C with agitation . IP complexes were washed twice with 1× NEB4 buffer . Elution was carried out by boiling in 50 µl 2× Laemmli buffer ( 100 mM Tris-HCl pH 6 . 8 , 20% glycerol , 1 mM EDTA , 4% SDS , 0 . 05% bromphenol blue ) . Rec12-oligonucleotide was separated by SDS-PAGE ( 10% ) , fixed , dried , and exposed . Quantification of Rec12-oligonucleotide was done with AIDA 1D- quantification software ( raytest ) . A 4-hour sample from a pat1-114 meiotic wild-type time-course experiment served as a normalization standard ( N ) . For each time point the amount of Rec12-oligonucleotide was determined by normalizing the amount of radioactivity to the meiotic sample N that was processed in parallel and loaded next to each experiment . The summation of Rec12-oligonuclotide values at each time point in wild type ( pat1-114 ) corresponded to 100% . Values of Rec12-oligonucleotide in mutant strains were determined accordingly . For Western blot analysis , protein was transferred onto a PVDF membrane in transfer buffer ( 25 mM Tris , 190 mM glycine , 20% methanol ) , probed with monoclonal anti c-myc-peroxidase antibody ( 1∶10 , 000 , clone 9E10 , Roche ) , and detected with the ECL Plus Western Blot Detection Kit ( Amersham ) . For detection of Rec12myc abundance during whole time course experiments 10% of immunoprecipitated Rec12 was retained and eluted in 20 µl elution buffer ( 4% SDS , 100 mM Tris-HCl pH 6 . 8 , 1 mM EDTA ) by heating ( 7 minutes , 97°C ) . 5 µl were dropped on PVDF membrane and probed with monoclonal anti c-myc-peroxidase antibody ( 1∶10 , 000 , clone 9E10 , Roche ) , and detected with the ECL Plus Western Blot Detection Kit ( Amersham ) . After TdT treatment , immunocomplexes were eluted in 30 µl Pronase elution buffer ( 0 . 5% SDS , 100 mM Tris ( pH 7 . 5 ) , 10 mM CaCl , 10 mM EDTA ) by boiling for 5 minutes . Pronase ( Roche ) was added for deproteinization ( final concentration: 0 . 2 mg/ml , 2 hours at 50°C ) . After four phenol/chloroform/iso-amylalcohol ( 25∶24∶1 ) and one chloroform/iso-amylalcohol extractions , the supernatant containing radiolabeled DNA was then directly mixed with 30 µl formamide loading buffer ( 80% formamide , 0 . 1% xylene xyanol , 0 . 1% bromphenol blue ) and boiled for 2 minutes . Extracts from wild type , tagged Rec12 ( Rec12myc ) , and Rec12 active site mutant ( Rec12Y98Fmyc ) , were processed in parallel and the whole sample was separated on a denaturing 20% poly-acrylamid gel . Gels were fixed , dried on Whatman paper prior to exposure . Genomic DNA embedded in agarose plugs was prepared as described in [53] . The chromosomal DNA in plugs was analyzed by pulsed-field gel electrophoresis ( PFGE ) as previously described [4] . Crosses of h− ade7-50 and h+ ade7-152 were carried out . Appropriate dilutions of spores were plated on selective ( MMA ) and non-selective media ( YEA ) and incubated for 5 days at 30°C . Prototroph frequencies were calculated as the number of prototrophs per 106 viable spores ( ppm ) . | A diploid zygote arises by fusion of two haploid gametes . The specific cell division leading to haploid gametes with haploid chromosome number , accompanied by recombination of genetic material , is called meiosis . It is essential for sexually reproducing eukaryotes . During meiotic prophase , shortly after DNA replication , programmed DNA double-strand breaks mark the initiation of recombination . In budding yeast , the protein responsible for DNA double-strand break formation , Spo11 , creates a covalent DNA-Spo11 intermediate , which needs to be removed for subsequent recombination . Presumably , asymmetric endonucleolytic cleavage of DNA next to bound Spo11 leads to distinct DNA ends in budding yeast and mouse . Here we show that the fission yeast Spo11 homolog Rec12 is removed by endonucleolytic cleavage as well . However , only a single oligonucleotides class can be detected suggesting symmetric cleavage . We show that Rec12 removal depends on Ctp1 and the MRN-complex . Furthermore , we applied this new method to monitor DNA double-strand break repair in mutants . Until now , pulsed-field gel electrophoresis was used to monitor global DNA double-strand break formation and repair . The assay presented here provides a new tool to monitor global DNA double-strand break processing only . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"biology/dna",
"repair",
"molecular",
"biology/recombination",
"cell",
"biology/developmental",
"molecular",
"mechanisms",
"genetics",
"and",
"genomics/chromosome",
"biology"
] | 2009 | Ctp1 and the MRN-Complex Are Required for Endonucleolytic Rec12 Removal with Release of a Single Class of Oligonucleotides in Fission Yeast |
Leptospirosis has become an urban health problem as slum settlements have expanded worldwide . Efforts to identify interventions for urban leptospirosis have been hampered by the lack of population-based information on Leptospira transmission determinants . The aim of the study was to estimate the prevalence of Leptospira infection and identify risk factors for infection in the urban slum setting . We performed a community-based survey of 3 , 171 slum residents from Salvador , Brazil . Leptospira agglutinating antibodies were measured as a marker for prior infection . Poisson regression models evaluated the association between the presence of Leptospira antibodies and environmental attributes obtained from Geographical Information System surveys and indicators of socioeconomic status and exposures for individuals . Overall prevalence of Leptospira antibodies was 15 . 4% ( 95% confidence interval [CI] , 14 . 0–16 . 8 ) . Households of subjects with Leptospira antibodies clustered in squatter areas at the bottom of valleys . The risk of acquiring Leptospira antibodies was associated with household environmental factors such as residence in flood-risk regions with open sewers ( prevalence ratio [PR] 1 . 42 , 95% CI 1 . 14–1 . 75 ) and proximity to accumulated refuse ( 1 . 43 , 1 . 04–1 . 88 ) , sighting rats ( 1 . 32 , 1 . 10–1 . 58 ) , and the presence of chickens ( 1 . 26 , 1 . 05–1 . 51 ) . Furthermore , low income and black race ( 1 . 25 , 1 . 03–1 . 50 ) were independent risk factors . An increase of US$1 per day in per capita household income was associated with an 11% ( 95% CI 5%–18% ) decrease in infection risk . Deficiencies in the sanitation infrastructure where slum inhabitants reside were found to be environmental sources of Leptospira transmission . Even after controlling for environmental factors , differences in socioeconomic status contributed to the risk of Leptospira infection , indicating that effective prevention of leptospirosis may need to address the social factors that produce unequal health outcomes among slum residents , in addition to improving sanitation .
At present , one billion of the world's population resides in slum settlements [1] . This number is expected to double in the next 25 years [1] . The growth of large urban populations which are marginalized from basic services has created a new set of global health challenges [2] , [3] . As part of the Millennium Development Goals [4] , a major priority has been to address the underlying poor sanitation and environmental degradation in slum communities which in turn , are the cause of a spectrum of neglected diseases which affect these populations [2] , [3] , [5] . Leptospirosis is a paradigm for an urban health problem that has emerged due to recent growth of slums [6] , [7] . The disease , caused by the Leptospira spirochete , produces life-threatening manifestations , such as Weil's disease and severe pulmonary hemorrhage syndrome for which fatality is more than 10% and 50% , respectively [7]–[9] . Leptospirosis is transmitted during direct contact with animal reservoirs or water and soil contaminated with their urine [8] , [9] . Changes in the urban environment due to expanding slum communities has produced conditions for rodent-borne transmission [6] , [10] . Urban epidemics of leptospirosis now occur in cities throughout the developing world during seasonal heavy rainfall and flooding [6] , [11]–[18] . There is scarce data on the burden of specific diseases that affect slum populations [2] , however leptospirosis appears to have become a major infectious disease problem in this population . In Brazil alone , more than 10 , 000 cases of severe leptospirosis are reported each year due to outbreaks in urban centers [19] , whereas roughly 3 , 000 , 8 , 000 and 1 , 500 cases are reported annually for meningococcal disease , visceral leishmaniasis and dengue hemorrhagic fever , respectively , which are other infectious disease associated with urban poverty [20]–[22] . Case fatality ( 10% ) from leptospirosis [19] is comparable to that observed for meningococcal disease , visceral leishmaniasis and dengue hemorrhagic fever ( 20% , 8% and 10% , respectively ) in this setting [20] , [23] , [24] . Furthermore , leptospirosis is associated with extreme weather events , as exemplified by the El Niño-associated outbreak in Guayaquil in 1998 [25] . Leptospirosis is therefore expected to become an increasingly important slum health problem as predicted global climate change [26] , [27] and growth of the world's slum population [1] evolves . Urban leptospirosis is a disease of poor environments since it disproportionately affects communities that lack adequate sewage systems and refuse collection services [6] , [10] , [11] . In this setting , outbreaks are often due to transmission of a single serovar , L . interrogans serovar Copenhageni , which is associated with the Rattus norvegicus reservoir [6] , [28]–[30] . Elucidation of the specific determinants of poverty which have led to the emergence of urban leptospirosis is essential in guiding community-based interventions which , to date , have been uniformly unsuccessful . Herein , we report the findings of a large seroprevalence survey performed in a Brazilian slum community ( favela ) . Geographical Information System ( GIS ) methods were used to identify sources for Leptospira transmission in the slum environment . Furthermore , we evaluated whether relative differences in socioeconomic status among slum residents contributed to the risk of Leptospira infection , in addition to the attributes of the environment in which they reside .
The study was conducted in the Pau da Lima community ( Figure 1A ) which is situated in the periphery of Salvador , a city of 2 , 443 , 107 inhabitants [31] in Northeast Brazil . Pau da Lima is a region of hills and valleys , which was a sparsely inhabited area of Atlantic rain forest in the 1970s and subsequently transformed into a densely-populated slum settlement ( Figure 1B ) due to in-migration of squatters . In total , 67% of the population of Salvador and 37% of the urban population in Brazil reside in slum communities with equal or greater levels of poverty as that found in Pau da Lima [32] , [33] . A study site was established which comprised of four valleys in an area of 0 . 46 km2 ( Figure 1A ) . Active hospital-based surveillance found that the mean annual incidence of severe leptospirosis was 57 . 8 cases per 100 , 000 population at the study site between 1996 and 2001 ( unpublished data ) . The study team conducted a census during visits to 3 , 689 households within the site in 2003 and identified 14 , 122 inhabitants . Households were assigned sequential numbers . A computer-based random number generator was used to select a list of 1 , 079 sample households from a database of all enumerated households . Eligible subjects who resided in sample households and had five or more years of age were invited to be a study participant . Subjects were enrolled into the study between April 2003 and May 2004 according to written informed consent approved by the Institutional Review Boards of the Oswaldo Cruz Foundation , Brazilian National Commission for Ethics in Research , and Weill Medical College of Cornell University . The study team of community health workers , nurses and physicians conducted interviews during house visits and administered a standardized questionnaire to obtain information on demographic and socioeconomic indicators , employment and occupation , and exposures to sources of environmental contamination and potential reservoirs in the household and workplace . Responses reported by subjects were used to obtain information on race . The study team evaluated literacy according to the ability to read standardized sentences and interpret their meaning . Informal work was defined as work-related activities for which the subject did not have legal working documents . The head-of-household , defined as the member who earned the highest monthly income , was interviewed to determine sources and amounts of income for the household . Subjects were asked to report the highest number of rats sighted within the household property and the site of work-related activities . The study team surveyed the area within the household property to determine the presence of dogs , cats and chickens . An ArcView version 8 . 3 software system ( Environmental Systems Research Institute ) database was constructed with georeferenced aerial photographs and topographic maps provided by the Company for Urban Development of the State of Bahia ( CONDER ) . Photographs of the study site , which had a scale of 1∶2 , 000 and spatial resolution of 16cm , were taken in 2002 . During the census , the study team identified households within the study site and marked their positions onto hard copy 1∶1 , 500 scale maps ( Figure 1C ) , which were then entered into the ArcView database . A survey was conducted during the seasonal period of heavy rainfall between April and August 2003 to geocodify the location of open sewage and rainwater drainage systems . During three time points within this period , the study team mapped the sites of open accumulated refuse and measured the area of these deposits . Mean values for areas of refuse deposits were calculated and used for the analyses . Sera were processed from blood samples collected from subjects during house visits . The microscopic agglutination test ( MAT ) was performed to evaluate for serologic evidence of a prior Leptospira infection [34] . A panel of five reference strains ( WHO Collaborative Laboratory for Leptospirosis , Royal Tropical Institute , Holland ) and two clinical isolates [6] were used which included L . interrogans serovars Autumnalis , Canicola and Copenhageni , L . borgspetersenii serovar Ballum , and L . kirschneri serovar Grippotyphosa . The use of this panel had the same performance in identifying MAT-confirmed cases of leptospirosis during surveillance in Salvador [6] , [16] as did the WHO recommended battery of 19 reference serovars [34] . Screening was performed with serum dilutions of 1∶25 , 1∶50 and 1∶100 . When agglutination was observed at a dilution of 1∶100 , the sample was titrated to determine the highest titer . Information for subjects was double entered into an EpiInfo version 3 . 3 . 2 software system ( Centers for Diseases Control and Prevention ) database . Chi-square and Wilcoxon rank sum tests were used to compare categorical and continuous data , respectively , for eligible subjects who were and were not enrolled in the study . A P value ≤0 . 05 in two sided testing was used as the criterion for a significant difference . Preliminary analyses evaluated a range of MAT titers as criteria for prior Leptospira infection and found that the use of different cut-off values ( 1∶25–1∶100 ) identified similar associations with respect to the spatial distribution of seropositive subjects and risk factors for acquiring Leptospira antibodies . A titer greater or equal to 1∶25 was therefore used to define the presence of Leptospira antibodies in the final analyses . The presumptive infecting serovar was defined as the serovar against which the highest agglutination titre was directed [34] . Crude prevalence rates were reported since age and gender-adjusted values did not differ significantly from crude values . Ninety-five percent confidence intervals ( CI ) were adjusted for the cluster sampling of households . Kernel density estimation analysis was performed with a range of bandwidths ( 10–120 meters ) to evaluate smoothed spatial distributions of subjects with Leptospira antibodies and all subjects . The R version 2 . 4 . 1 statistical package ( R Foundation for Statistical Computing ) was used to obtain estimates which were adjusted for boundary effects . The ratio of the Kernel density estimators for subjects with Leptospira antibodies and all subjects was measured to determine the smoothed population-adjusted risk distribution . A digital terrain model of topographic data was used ( ArcGIS 3D Analyst Extension software ) to obtain continuous estimates of altitude for the study area . The distances , calculated in three-dimensional space , of households to nearest open drainage systems and refuse deposits were evaluated as proxies of exposure to these sources of environmental attributes . Elevation of households with respect to the lowest point in the valley in which they were situated was used as a surrogate for flood risk . Generalized additive models ( GAM ) [35] were used to evaluate the functional form of the association between continuous variables and the risk of acquiring Leptospira antibodies . When indicated , continuous variables were categorized in multivariate analyses according to the x-intercept value observed in the plots of fitted smoothed values . We used Poisson regression [36] to estimate the effect of demographic , socioeconomic , household and workplace-related factors on the prevalence of Leptospira antibodies . A Bayesian inference approach was used which incorporated two random effects in order to account for overdispersion and cluster sampling within households . This approach has been used to estimate parameters in complex models [37] and is less sensitive to sparse data [38] . Standard non-informative prior distributions were used in models which were fitted with WinBUGS version 1 . 4 . 2 ( MRC Biostatistics Unit ) . In multivariate analysis , all variables which had a P value below 0 . 10 in univariate analyses were included in the initial model . In order to address co-linearity among variables , we identified sets of covariates with the high Spearman correlation coefficients ( >0 . 3 or <−0 . 3 ) . Highly correlated variables were aggregated in a single variable when indicated , and evaluated in the model . The final model was obtained which used backward variable selection with an inclusion rule of P value <0 . 05 .
Among 3 , 797 eligible residents from the slum community site , 3 , 171 ( 84% ) were enrolled in the study . Study subjects had a higher proportion of females ( 56% of 3 , 171 subjects versus 37% of 626 subjects , respectively; P<0 . 05 ) and younger mean age ( 25 . 8±15 . 2 versus 28 . 1±14 . 6 years , respectively; P<0 . 05 ) than eligible residents who did not participate in the study . The kernel distribution of enrolled subjects according to place of residence was similar on visual inspection to that of residents who did not participate ( data not shown ) . The majority ( 85% ) of subjects were squatters who did not have legal title to their domiciles . Subjects belonged to mostly mixed ( pardo , 66% ) or black ( 28% ) racial groups . Median household per capita income for study subjects was US$ 1 . 30 per day . Among the subjects , 76% had not completed elementary school education and 23% were illiterate . Among 2 , 077 subjects ≥18 years of age , 77% did not have formal employment and 35% engaged in informal work . Among the 3 , 171 subjects , 489 had Leptospira agglutinating antibodies , as determined by the presence of MAT titer ≥1∶25 ( Figure 2 ) . Highest titers were directed against L . interrogans serovar Copenhageni in 436 ( 89 . 2% ) of the 489 subjects with Leptospira antibodies . For the 22 subjects ( 4 . 5% ) who had highest titers against two or more serovars , agglutination reactions recognized Copenhageni as one of the serovars . Copenhageni was the predominant serovar ( 88–100% ) recognized for the range of highest reciprocal titers ( Figure 2 ) . The overall prevalence of Leptospira antibodies was 15 . 4% ( 95% CI 14 . 0–16 . 8 ) . The crude prevalence among enrolled subjects was not significantly different from the prevalence ( 15 . 9% , 95% CI 14 . 6–17 . 1 ) which was adjusted for the age and gender distribution of eligible subjects in the study population . Prevalence was highest among adolescents and adults ( 16 . 2% and 21 . 2% for age groups 15–24 and >44 years , respectively ) . However , 8 . 3% ( 95% CI 6 . 2–10 . 5 ) of children 5–14 years of age had evidence for a prior exposure to Leptospira . The prevalence was higher in males than females ( 17 . 8 versus 13 . 6% , respectively; PR 1 . 32 , 95% CI 1 . 10–1 . 57 ) ( Table 1 ) . Similar associations with age and gender were observed when MAT titers of ≥1∶50 and ≥1∶100 were used to define subjects with Leptospira antibodies . Panels A and B in Figure 3 show smoothed spatial distributions of subjects with Leptospira antibodies and all subjects , respectively , according to place of residence . The population-adjusted distribution ( Figure 3C ) showed that risk of acquiring Leptospira antibodies clustered in areas occupied by squatters at the bottom of valleys ( Figure 3D ) . Similar spatial distributions were observed in analyses that used higher titer values to define subjects with Leptospira antibodies ( Figure S1 ) . Univariate analysis found the risk of acquiring Leptospira antibodies to be associated with increasing age , male gender , indicators of low socioeconomic level , occupations that entail contact with contaminated environments , informal work , time of residence in the study household , and environmental attributes and the presence of reservoirs in the household ( Table 1 ) . Significant risk associations were not found for formal employment and reported sighting of rats in the workplace environment . Open rainwater drainage structures and refuse deposits were distributed throughout the site; yet open sewers were more frequently encountered at the bottom of valleys ( Figure 3 ) . The distance of household to the nearest open sewer was a risk factor , whereas a significant association was not observed for distance to an open rainwater drainage system . GAM analysis showed that the risk of acquiring Leptospira antibodies had an inverse linear association with the distance of the subject's household to an open sewer and elevation of the household from the lowest point in the valley , a proxy for flood risk ( Figure 4 ) . Increased risk was observed among subjects who resided less than a threshold distance of 20 meters to these attributes ( Figure 4B and C ) . The risk of acquiring Leptospira antibodies had an inverse non-linear association with distance of the subject's household to an open refuse deposit ( results not shown ) . We explored a range of dichotomization criteria and found significant risk associations when subjects resided less than 20 meters from an open refuse deposit ( Table 1 ) . This association was not influenced by the size of the refuse deposit . Subjects who reported sighting two or more rats in the household environment had increased risk of acquiring Leptospira antibodies ( Figure 4D ) . Household per capita income had an inverse linear association with the presence of Leptospira antibodies ( Figure 4A ) . Of note , the distance of the household to an open sewer was highly correlated ( Spearmen correlation coefficient = 0 . 71 ) with household elevation ( Figure S2A ) since open sewers drain into the bottom of valleys . An aggregate variable , distance of household located less than 20 meters from an open sewer and lowest point in a valley , was therefore used to examine the association between open sewer and flood-related exposure and infection risk ( Table 1 ) . In contrast household per capita income was not highly correlated ( Spearmen correlation coefficient = 0 . 16 ) with the elevation of the household ( Figure S2B ) . Multivariate analyses found that the risk for acquiring Leptospira antibodies was associated with exposures in the household environment and not in the workplace setting ( Table 1 ) . Subjects who resided less than 20 meters from an open sewer and the lowest point in the valley had a 1 . 42 times ( 95% CI 1 . 14–1 . 75 ) increased risk for acquiring Leptospira antibodies than those who lived 20 meters or more from these attributes . Residence less than 20 meters from accumulated refuse was associated with a 1 . 43 times ( 95% CI 1 . 04–1 . 88 ) increased risk . Sighting of two or more rats and presence of chickens , a marker for rat infestation , in the household were significant reservoir-associated risk factors . After controlling for age , gender and significant environmental exposures , indicators of low socioeconomic level , household per capita income ( PR 0 . 89 for an increase of US$1 . 00 per day , 95% CI 0 . 82–0 . 95 ) and black race ( PR 1 . 25 , 95% CI 1 . 03–1 . 50 ) were risk factors for acquiring Leptospira antibodies ( Table 1 ) .
Efforts to identify interventions for urban leptospirosis have been hampered by the lack of population-based information on transmission determinants . In this large community-based survey of a slum settlement in Brazil , we found that 15% of the residents had serologic evidence for a prior Leptospira infection . The prevalence rate of Leptospira antibodies in the study slum community was similar to that ( 12% ) found in a city-wide survey performed in Salvador [39] . Risk factors for acquiring Leptospira antibodies were associated with exposures in the household environment . Interventions therefore need to target the environmental sources of transmission - open sewers , flooding , open refuse deposits and animal reservoirs - in the places where slum inhabitants reside . After controlling for the influence of poor environment , indicators of low socioeconomic status were found to be independently associated with the risk of acquiring Leptospira antibodies . This finding suggests that in slum communities with overall high levels of absolute poverty , relative differences in socioeconomic level contribute to unequal outcomes for leptospirosis . Leptospirosis has been traditionally considered an occupational disease , since work-related activities are frequently identified as risk exposures [9] . However slum inhabitants reside in close proximity to animal reservoirs and environmental surface waters which contain Leptospira [10] . We previously found that Leptospira infection clusters within households in slum communities in Salvador [40] . In this study , we found that after controlling for confounding , significant risk exposures were those associated with the household environment rather than workplace . As a caveat , interview-elicited responses were used to evaluate work-related exposures since GIS surveys were not performed at the sites where subjects worked . It is possible that slum residents may have had work-related risk exposures which were not detected by our survey . Nevertheless , our findings support the conclusion that the slum household is an important site for Leptospira transmission and provides the rationale for interventions that target risk exposures in this environment . The study's findings indicate that the domestic rat was the principal reservoir for Leptospira transmission in the study community . Highest agglutination titers among 89% of the subjects were directed against L . interrogans serovar Copenhageni , the serovar associated with the R . norvegicus reservoir . Reported sighting of rats is considered to be an unreliable marker of rat infestation . However we found that the number of rats sighted by residents was correlated with their risk of acquiring Leptospira antibodies ( Figure 4D ) , indicating that rat sightings may be a useful marker of infection risk in slum communities where inhabitants are accustomed to the presence of rats . Although dogs were not found to be a risk factor , detailed investigations of Leptospira carriage in urban reservoirs need to be performed . Of note , the presence of chickens in households was a risk factor , although they in of themselves are not reservoirs . This association may reflect a rat-related exposure not accounted for by reported sightings , since rats are attracted to chicken feed and waste . Raising chickens is a widespread practice in slum communities-48% ( 519 ) of the 1079 study households raised chickens . Control of rodent reservoir populations may therefore need to incorporate measures that directly address this practice . Our findings confirm hypotheses raised by previous ecologic studies [6] , [10] , [11] that infrastructure deficiencies related to open sewers , flooding and open refuse deposits are transmission sources for leptospirosis in the slum environment . Furthermore , there appears to be defined areas of risk associated with open sewers and refuse deposits , which serve as habitats and sources of food for rats . Home range radius of the domestic rat varies from 30–150 meters [41] , [42] , but home range use decreases from the centre to the edge . GAM analysis demonstrated that slum residents had a positive risk for acquiring Leptospira antibodies when households were situated within 20 meters from open sewers and refuse deposits . In addition , infection risk increased as distances from an open sewer or refuse deposit decreased , suggesting that households which are situated closer to these foci have a higher degree of environmental contamination with Leptospira and inhabitants of these households are exposed to higher inoculum doses during infection . Molecular approaches to quantify Leptospira in environmental samples [10] will be useful in answering this question and guiding recommendations for environmental decontamination and barrier control measures which can be implemented in slum communities . In addition , GAM analysis found that residents had positive risk for Leptospira infection when their households were situated within 20 meters from the lowest point in the valley ( Figure 4B ) . In Salvador [6] , [12] , [16] , [40] and other urban centers [11] , [13] , [15] , [17] , [18] , outbreaks of leptospirosis occur during heavy rainfall and flooding events . Slum communities are built on the poor land quality and often in areas susceptible to frequent flooding . At the study site and other slum settlements in Salvador , the water table rises up to one meter during flooding events because of inadequate rainwater drainage and blockage of drainage systems with silt and refuse . The finding that subjects had increased infection risk when their households were located within 20 meters from the lowest point in the valley suggests that this distance was a proxy for the degree of contact which residents encounter flood-related exposures in the peri-domiciliary environment . We found that in addition to attributes of the environment where slum inhabitants reside , low per capita household income and black race , an indicator of health inequality in Brazil [43] , [44] , were independent risk factors for Leptospira infection . The social gradient in health is a widespread phenomenon [45] , [46] . Our findings , although not unexpected , are noteworthy since they suggest that differences in status contribute to unequal health outcomes in a slum community where the household per capita income was less than US$1 per day for 44% of the inhabitants . Although errors in the measurement of risk exposures and residual confounding were a possibility , the strength of the association indicates a role for social determinants in Leptospira transmission . These factors may relate to risky behaviors , such as cleaning open sewers after flooding events , or limited use of protective clothing which reduce the risk of abrasions that facilitate entry of the Leptospira spirochete [47] . Low status and lack of access to amenities and social support are features of disadvantaged communities [45] which conceivably influence risk behaviors for leptospirosis . Further research is needed to evaluate the role of social factors such that effective interventions , including health education , can be implemented at the community level . A limitation of our study was the cross-sectional design which used serologic evidence for a prior Leptospira infection as the outcome . The MAT is the standard assay used in prevalence surveys [9] , yet there is not an established titer criterion for defining seropositive reactions . We previously found that a MAT titer of ≥1∶25 was a specific marker for prior Leptospira infection among slum residents from Salvador and when applied , identified household clustering of infection risk [40] . In this study , cutoff titers from 1∶25 and above identified similar risk associations . In Salvador , leptospirosis is due to transmission of a single agent , L . interrogans serovar Copenhageni [6] , [28] . Titers of 1∶25 , as well as higher titers , were directed against this serovar ( Figure 1 ) , indicating that this cutoff was a specific and more sensitive criteria for identifying prior infections in a region where a single serovar agent is circulating . In the study , there were more men and younger subjects among non-participating subjects than participating subjects . Crude prevalence was not different from the prevalence of Leptospira antibodies which was adjusted by the age and gender distribution of the overall study population , indicating that differences between participating and non-participating subjects may not have introduced a significant bias in the estimates . Infections may have occurred up to five years prior to the survey since agglutinating antibodies may persist for this period [48] , [49] . Major interventions to improve basic sanitation were not implemented in the study community , yet the possibility that environmental exposures were modified over time can not be excluded . Migration may have affected our ability to estimate prevalence and risk associations . An on-going cohort investigation of subjects enrolled in this study found that the annual out-migration rate is approximately 12% ( unpublished data ) . The study's findings therefore need to be confirmed in prospective studies . We found that Leptospira transmission was due to the interaction of factors associated with climate , geography and urban poverty . Since the study was performed in a single community in Salvador , Brazil , our findings may not be generalizable to other slum settings . However , a large proportion of the world's slum population resides in tropical climates similar to that in Salvador . Moreover , similar conditions of poverty and environmental degradation encountered at the study site ( Figure 1B ) are found in many slum settlements . In Brazil , 37% of the urban population resides in slums with equal or greater levels of poverty as found in the study community [33] . Our findings may therefore be relevant to other slum communities where leptospirosis is endemic and have increasing significance as global climate change [26] , [27] and growth of the world's slum population occur in the future [1] , [33] . The infrastructure deficiencies which were found to be transmission factors for Leptospira in this study can be readily addressed by improving sanitation in slum communities . Investment in sanitation is a cost-effective health intervention [50] , [51] . In Salvador , a city-wide sanitation program ( Bahia Azul ) was recently shown to have a major beneficial impact for diarrheal disease [52] . However , as frequently encountered with large-scale sanitation projects , the Bahia Azul program did not provide coverage to the study community and many of the slum settlements in the city's periphery . Equitable access to improved sanitation is therefore essential in reducing the burden of the large number of environmentally-transmitted infectious diseases , including leptospirosis , which affects slum populations . Furthermore , the finding that the social gradient within slum communities , in addition to the unhealthy environment , contributes to the risk of Leptospira infection suggests that prevention of urban leptospirosis will need to combine approaches for improving sanitation with approaches that identify and address the social determinants which produce unequal health outcomes . | Leptospirosis , a life-threatening zoonotic disease , has become an important urban slum health problem . Epidemics of leptospirosis now occur in cities throughout the developing world , as the growth of slum settlements has produced conditions for rat-borne transmission of this disease . In this prevalence survey of more than 3 , 000 residents from a favela slum community in Brazil , Geographical Information System ( GIS ) and modeling approaches identified specific deficiencies in the sanitation infrastructure of slum environments—open sewers , refuse , and inadequate floodwater drainage—that serve as sources for Leptospira transmission . In addition to the environmental attributes of the slum environment , low socioeconomic status was found to independently contribute to the risk of infection . These findings indicate that effective prevention of leptospirosis will need to address the social factors that produce unequal health outcomes among slum residents , in addition to improving sanitation . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/bacterial",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] | 2008 | Impact of Environment and Social Gradient on Leptospira Infection in Urban Slums |
Increasing emphasis on integrated control of neglected tropical diseases ( NTDs ) requires identification of co-endemic areas . Integrated surveys for lymphatic filariasis ( LF ) , schistosomiasis and soil-transmitted helminth ( STH ) infection have been recommended for this purpose . Integrated survey designs inevitably involve balancing the costs of surveys against accuracy of classifying areas for treatment , so-called implementation units ( IUs ) . This requires an understanding of the main cost drivers and of how operating procedures may affect both cost and accuracy of surveys . Here we report a detailed cost analysis of the first round of integrated NTD surveys in Southern Sudan . Financial and economic costs were estimated from financial expenditure records and interviews with survey staff using an ingredients approach . The main outcome was cost per IU surveyed . Uncertain variables were subjected to univariate sensitivity analysis and the effects of modifying standard operating procedures were explored . The average economic cost per IU surveyed was USD 40 , 206 or USD 9 , 573 , depending on the size of the IU . The major cost drivers were two key categories of recurrent costs: i ) survey consumables , and ii ) personnel . The cost of integrated surveys in Southern Sudan could be reduced by surveying larger administrative areas for LF . If this approach was taken , the estimated economic cost of completing LF , schistosomiasis and STH mapping in Southern Sudan would amount to USD 1 . 6 million . The methodological detail and costing template provided here could be used to generate cost estimates in other settings and readily compare these to the present study , and may help budget for integrated and single NTDs surveys elsewhere .
Health intervention needs generally exceed available funds . Managers of disease control programmes therefore need to decide how to allocate their resources most efficiently . This is particularly so for the control and elimination of neglected tropical diseases ( NTDs ) , which have been chronically underfunded [1] , [2] . In an attempt to increase the efficiency of NTD programmes , the co-administration of preventive chemotherapy ( PCT ) for lymphatic filariasis ( LF ) , onchocerciasis , schistosomiasis , soil-transmitted helminth ( STH ) infections and trachoma is widely advocated , the rationale being that in sub-Saharan Africa ( SSA ) the distributions of these diseases often overlap [3] . In areas of NTD co-endemicity one delivery structure , instead of several , could therefore be used for mass drug administration ( MDA ) of PCT [4] , [5] . Co-endemicity , the prerequisite for the anticipated efficiency gains of integrated control , applies not at the country level , but at sub-national levels where climatic and other determinants are suitable for the transmission of more than one of above diseases [6] , [7] . For example , onchocerciasis is associated with fast flowing rivers [8] , schistosomiasis occurs near calmer or stagnant waterbodies [9] , whilst STH infection , trachoma and LF occur over relatively large areas [10]–[12] . Because of these differing transmission ecologies , prevalence data on each disease are required to identify areas of overlap and to target these with integrated PCT delivery [13] . For large areas of SSA there are either no data to assess the potential for integration , or they are incomplete or out of date . Only the distribution of onchocerciasis has been comprehensively mapped , while LF mapping is ongoing in several countries and has not commenced in Chad and Eritrea [14] . In 2000 , prevalence data for schistosomiasis and STH infections were only available for a third of all districts in SSA [15] and , in spite of increased resources for mapping , survey coverage remains patchy or absent in many areas [16] . No trachoma data were available for seven countries in the Africa region in 2005 , while only a few countries had undertaken national surveys [17] . Although more NTD data are now being collected , the gaps in the known distributions are still considerable . Given that funds are not just limited for intervention , but are particularly hard to mobilize for apparent ‘research’ , some control programmes have started to survey multiple NTDs simultaneously instead of mapping the different diseases separately [18] , [19] . The underlying rationale for such integrated surveys is the same as that for co-administration of PCT; reaching communities in SSA is often challenging and associated with considerable costs , so avoiding repeated access to conduct similar activities is likely to minimize the investment required to achieve a desired outcome , be it classification of an administrative area ( e . g . district ) for intervention or curing people from NTD infection . As yet there is limited operational experience with integrated NTD surveys , but it is clear that their design needs to balance cost against the precision and accuracy with which administrative areas are classified according to treatment needs . To improve on current designs it is important to understand the main drivers of survey costs and investigate potential effects of modifying standard operating procedures . One such modification is altering the number of sites sampled , which is likely to affect both the cost and accuracy of determining whether a geographical area needs to be targeted with interventions . Such cost analyses should be undertaken using an approach that is ‘generalisable’ , hence allowing comparison between settings and use of results to plan and budget for similar undertakings elsewhere [20] , [21] . Southern Sudan , along with the Democratic Republic of Congo and Central African Republic , possibly has the largest unmapped NTD burden in SSA and hence the greatest need for up-to-date data . In 2008 , based on available information [6] , [22] , Southern Sudan developed a national strategy for the integrated control of onchocerciasis , LF , schistosomiasis , STH infection and trachoma . An essential component of this strategy is to generate data on the distribution and co-distribution of the targeted NTDs ( LF , schistosomiasis , STH infection nationwide , and trachoma in remaining regions ) . In 2009 , an integrated NTD survey was conducted in Northern Bahr-el-Ghazal State , the first of ten states in the country [19] , with completion of mapping in the remaining nine states planned for 2010/11 . The aims of this paper are to analyse the costs of the 2009 survey , to identify the main cost drivers , and to estimate the resources required to expand surveys to the whole of Southern Sudan . In addition , to help compare our results to those from other settings and estimate the costs of integrated surveys elsewhere , we present a standardised approach to costing integrated NTD surveys .
During 2007 , the Ministry of Health , Government of Southern Sudan ( MoH-GoSS ) , conducted a situation analysis of NTDs and their control in order to inform planning for NTD control and elimination . This analysis indicated that 12 NTDs were endemic , including all of the diseases for which MDA of PCT forms an important component of control , namely onchocerciasis , LF , schistosomiasis , STH infection and trachoma [22] . At the time , only onchocerciasis and trachoma had benefitted from regular MDA in some endemic areas , while STH infections in children had been treated through a number of deworming rounds alongside national immunization days . Although the need to control all NTDs endemic to Southern Sudan's was highlighted , an opportunity was identified to combine those diseases suitable for MDA-based control under an umbrella National Integrated NTD Control Programme . The aim of this undertaking was to increase geographical PCT coverage and the number of diseases treated in each location by expanding the scope of existing community-based delivery mechanisms , be it for NTDs or other interventions . Two delivery mechanisms were identified for initial integration , the volunteer networks for community-directed treatment with ivermectin ( CDTI , covering parts of the onchocerciasis endemic areas ) and similar networks for Guinea worm eradication ( which in some areas also delivers trachoma interventions ) . Given the lack of experience of expanding these networks into delivery platforms for PCT packages , the approach was to be piloted and gradually scaled up building on implementation experience gained along the way . In areas where neither CDTI nor the Guinea worm volunteer network are present , other existing delivery structures may need to be supported to take on integrated NTD control . Alternatively a new platform may need to be established to deliver MDA amongst other public health interventions [22] . Southern Sudan has four administrative tiers: state ( 1st ) , county ( 2nd ) , payam ( 3rd ) and boma ( 4th ) . A county is the administrative unit most comparable to a district in other African countries . The majority of counties include five or more payams . In 2005 , Southern Sudan had 49 counties , the majority of which were <10 , 000 km2 in size ( median of 8 , 033 km2 ) and had a population of 100 , 000 to 500 , 000 inhabitants . At the same time these counties were divided into a total of approximately 308 payams that varied greatly in population and size . The overall median population of payams was 31 , 607 , but this ranged from 2 , 000 to 120 , 000 inhabitants . The median payam size was 1 , 876 km2 , ranging from 126 km2 ( Cueibet in Lakes State ) to 58 , 210 km2 ( Raja in Western Bahr-el-Ghazal State ) . The district , or an area of equivalent size , is the recommended implementation unit ( IU ) for LF elimination [23] . An area this size , however , may be too large for co-administration of a standardized drug package , because schistosomiasis and onchocerciasis are unlikely to be endemic throughout . The MoH-GoSS therefore decided to consider both the payam and the county as IUs of the Integrated NTD Control Programme . A large-scale survey of LF , schistosomiasis , STH infection and loiasis was conducted in Northern Bahr-el-Ghazal State between February and May 2009 . Details of the study area and survey protocol are provided elsewhere [19] , [24] . In summary , quasi-random two-stage cluster sampling was used to select communities on the basis of potential risk of LF and schistosomiasis and to randomly select households within these communities . Each household head was requested to provide written consent , and all children aged 5 to 16 years were asked to give verbal consent before providing stool and urine samples for examination of schistosome and STH infection; adults were only sampled for LF testing using immunochromatographic tests ( ICT , BinaxNOW Filariasis , Inverness Medical ) . For LF , up to three communities and 250 individuals per payam were sampled . Sites selected for LF were also sampled for schistosomiasis and STH , as well as loiasis , a disease whose presence complicates LF elimination [25] . In each payam , between three to four additional communities were sampled for STH and schistosomiasis , with the actual number depending on payam size and estimated population . A total of 43 communities were sampled for LF infections and 73 communities for schistosomiasis and STH infections . Communities were selected from each of the five counties of Northern Bahr-e-Ghazal State , with only one out of 22 payams not surveyed . Due to delays in the supply of ICTs for LF , the survey had to be conducted in two phases , lasting 22 and 31 days respectively , with a gap of three weeks in-between . Because of this delay , a total of 13 sites had to be revisited for LF data collection . As this delay was not part of standard operating procedures , we excluded the cost associated with re-visiting these communities , though the test-specific costs were included . The composition of survey teams varied depending on the number of NTDs to be surveyed in each location . Where all four NTDs were surveyed , teams included two drivers , one supervisor , one interviewer/translator , and up to four technicians , travelling in two vehicles . One or two technicians undertook blood sampling and LF testing while the other two to three prepared and read stool and urine samples . For those sites where schistosomiasis and STH infection or just LF were surveyed , the team consisted out of one driver , one supervisor , one interviewer/translator , and at least two laboratory technicians , travelling in one vehicle . Due to severe human resource constraints affecting Southern Sudan , only two national laboratory technicians per team could be recruited , with additional technicians recruited as short-term consultants from the Vector Control Division , MoH , Uganda . Village guides were recruited locally . Owing to the poor infrastructure in Northern Bahr-el-Ghazal State and fluctuating security , teams established camps in locations that were centrally located between study communities . In many cases , space to pitch tents was provided by non-governmental or faith-based organizations in their compounds , but a charge was levied . Otherwise teams stayed in local guesthouses . For camping , sleeping and cooking equipment , food , fuel and small generators were procured . Three Toyota Land Cruisers were used during the surveys . To prepare for the surveys , ten days were needed to arrange supplies and develop the database , a total of four days were required to move vehicles from Juba to Northern Bahr-el-Ghazal State and back , and two days were required to train surveyors on the study protocol . At the end of the survey , one day was allocated to take stock and clean and store supplies and equipment . After completion of all survey activities , five days were needed to clean the data and undertake preliminary analysis . Both financial and economic costs were estimated from the perspective of the provider [26] , in this case Malaria Consortium and the MoH-GoSS . Financial costs were the cash expenditures made to enable implementation of the survey . For capital items , these were estimated for the total number of survey days by means of straight-line depreciation followed by calculation of an average financial daily cost . Economic costs captured the value of all resources consumed by the survey , including opportunity costs of volunteers and equipment that were used in the survey but not paid for , as well as appropriate treatment of costs of capital items with a value of >USD 100 and an expected useful life of more than one year [27] . Costs of MoH staff were based on the GoSS pay scale for 2009 while the time of international volunteers was valued using equivalent Malaria Consortium salaries for this setting . Capital items were discounted over their estimated useful life using the recommended discount rate of 3% [28] , [29] . Daily economic costs were calculated for all capital items and multiplied by the appropriate number of days in use during the survey . Based on our experience of working in the harsh climatic environment of Southern Sudan we estimated the useful life to be four years for vehicles and high frequency radios ( fitted to vehicles ) and two years for all other items , including laptop computers and laboratory field equipment . All resources used for research activities , such as this costing study , were excluded from the analysis as were the costs associated with the development of the survey protocol . Cost data were collated from financial expenditure records of Malaria Consortium during the survey and shortly afterwards . To accommodate considerable fluctuations in currency conversion rates , two time periods were used . For payments for survey supplies , which started in October 2008 and continued through May 2009 , we used average exchange rates of 1 United States Dollars ( USD ) = 0 . 67 British Pounds ( GBP ) or 1 USD = 2024 Ugandan Shillings ( UGX ) . For costs associated with the actual survey activities between February and May 2009 we used a rate of 1 USD = 2 . 3 Sudanese Pounds ( SDG ) ( http://www . oanda . com/convert/fxhistory ) . We assumed that these exchange rates and the wages paid reflected competitive foreign exchange markets , and therefore did not use shadow prices to adjust for possible distortions [30] . Costs were identified using an ingredients approach , whereby the total value of each of the services and goods employed in implementing the survey was estimated by identifying the number of units consumed and multiplying these by their unit price [26] , [31] . Following the structure of the survey , cost and services were organized into capital and recurrent costs , both of which were divided into cost categories [Box 1 , Dataset S1] . It is common practice to include overheads in cost analyses [32] . To capture the indirect costs of project management and administration we applied an overhead of 25% to the financial cost estimate for all budget lines . This rate is higher than those applicable to operations in more stable settings of eastern Africa [e . g . [33] , but in our experience provides an accurate reflection of the cost involved in implementing programmes in a land-locked country undergoing post-conflict reconstruction . The overall purpose of the integrated survey was to generate data by which IUs could be classified according to intervention thresholds recommended by the WHO [3] . The present analysis considered both the county and the payam as IUs , because a two-tier system may be needed to account for inherent differences in the geographical distribution of the targeted NTDs . While for LF elimination the county is likely to be an appropriate IU , control of onchocerciasis is already more geographically focused and the first round of integrated NTD surveys indicated that a similar approach is also more appropriate for schistosomiasis [19] . The outcomes used for our costing study were thus the county and the payam surveyed , to allow their classification for intervention according to WHO recommended thresholds . Calculation of costs and outcomes involves a number of assumptions; the ones underlying the present study have been outlined above and in table 1 . One-way sensitivity analysis was conducted to explore the effects of key assumptions on the results . We varied the discount rate ( reduced to 0% or increased to 10% ) and increased the assumed lifespan of vehicles to 7 . 5 years . In addition we investigated the effect of modifying the standard operating procedures . Instead of using Hemastix reagent strips validated by urine filtration for diagnosis of urinary schistosomiasis , as used in the first survey round [34] , we assumed that urinary schistosomiasis could be adequately diagnosed using reagent strips only and excluded the cost of urine filtration , a method that is relatively costly because it requires isopore membrane filters at a price of nearly USD 2 per filter . The rationale for cutting out urine filtration was that the use of reagent strips alone is a reliable diagnostic procedure in some African countries [35] and may , after validation , be applicable in high schistosomiasis transmission settings in Southern Sudan . We also explored the effect on cost of classifying counties , rather than payams , for LF elimination . Instead of sampling 250 individuals from up to three sites per payam , the analysis investigated the cost implications of applying exactly the same procedure but at county level . This procedure would have yielded the same classification of IUs for LF elimination in Northern Bahr-el-Ghazal State [19] and is likely to be appropriate throughout Southern Sudan . The present study involved collection of data on cost and non-financial inputs through analysis of Malaria Consortium expenditures and activities during an epidemiological survey . Data collection was conducted from the ‘provider perspective’ , rather than the all encompassing ‘societal perspective’ [26] , since participation in the survey incurred minimal time commitment from the study communities . The epidemiological survey itself received ethical approval from the Directorate of Research , Planning and Health System Development , MoH-GoSS , and from the Ethics Committee of the London School of Hygiene and Tropical Medicine , UK . Collection of the data presented here did not involve human subjects and therefore did not require ethical approval .
The total financial costs for the survey amounted to USD 182 , 067 , the majority of which was spent on recurrent items . The principal cost drivers were personnel ( 34 . 4% ) followed by survey consumables ( 30 . 4% ) ( Table 2 ) . Economic costs amounted to a total of USD 201 , 030 and were arrived at by including the imputed value of the time provided by MoH employees and other non-cash inputs , as well as the opportunity cost of capital items used in the survey . Because all assets above USD 100 in value and with a lifespan >1 year were annualised and only their time in use during the survey was included , capital cost amounted to only 4 . 3% of the total survey costs and were largely comprised of vehicle costs . As for the financial costs , most economic costs were taken up by survey consumables and personnel , accounting for 27% and 38% of the total , respectively ( Table 2 ) . The outcome from this investment was that 21 payams in the five counties of Northern Bahr-el-Ghazal State , an area the size of Belgium , were surveyed for LF , schistosomiasis , STH infection and loiasis . These five counties and 21 payams were , for the first time ever , categorised as requiring MDA delivery or not , according to WHO recommended thresholds [3] . The average economic cost per IU classified was USD 40 , 206 per county or USD 9 , 573 per payam . These estimates changed very little when the two key assumptions , the discount rate or lifespan of vehicles , were varied ( Table 3 ) . Survey costs could be lowered if the standard operating procedure were modified . Testing for urinary schistosomiasis infection could , for example , be conducted only with Hemastix reagent sticks ( Bayer Diagnostics , Basingstoke , UK ) , instead of the current procedure whereby all samples with blood in urine and a sub-sample of negatives were cross checked using urine filtration . By cutting out the cost of these and other supplies for urine filtration , the cost per implementation unit surveyed would have decreased by 2% ( Table 3 ) . This small reduction in economic cost is explained by the fact that urine filtration can be easily conducted alongside other operations; personnel requirements ( the major cost driver ) and survey duration would thus remain unchanged . In Northern Bahr-el-Ghazal State , where LF was found not to be endemic , sampling of up to 250 individuals from up to three sites per county would have resulted in the same classification of IUs . We therefore proposed that testing for LF infection is from now on conducted at the county instead of payam level throughout Southern Sudan , given that LF tends to be endemic over relatively large areas [12] . Sensitivity and specificity of LF detection , and hence correctly classifying IUs for interventions , is unlikely to be affected by surveying areas larger than the payam . This approach would also be more consistent with WHO recommendations for establishment and sampling of IUs for LF [23] , as size and administrative level of the county in Southern Sudan is equivalent to that of the district in other African countries . If the standard operating procedures of the survey were thus modified , quantities of ICT tests , safety lancets and other supplies required to safely take blood samples would be reduced by approximately 80% , incurring substantial savings on survey consumables , particularly ICTs that cost >USD 4 . 0 per test . Furthermore , existing vehicles and staff could be deployed in three teams , which would decrease the survey duration by about one day per payam and reduced the majority of running costs accordingly . The cost per IU classified could thus be reduced by about 25% without any apparent risk of missing LF foci . Based on the data presented here we estimate that completion of NTD mapping in Southern Sudan by using an integrated NTD survey design would amount to an economic cost of approximately USD 1 . 6 million . This calculation is based on the assumption that LF surveys would only be conducted at county level and that the administrative divisions used to conduct National Immunization Days ( NIDs ) in 2008 would be used to identify survey areas . At the time of the 2008 NIDs there were a total of 58 counties , of which 53 remain to be surveyed for NTDs today .
The major cost drivers of an integrated NTD survey in Southern Sudan were personnel and consumables . It is likely that this observation is applicable to most other settings , and should be taken into account during the budgeting process in countries wishing to apply this survey approach . Overall costs may be lower in settings that have not recently experienced conflict , because there will be less need to draw on international staff for technical support and overall operating costs may be lower . To confirm or refute this assumption we encourage further costing work on integrated , as well as single NTD surveys , using the approach outlined here . Most importantly , such analysis should help to identify areas of potential cost savings and be able to provide an estimate of the overall investment required to complete NTD mapping for specific areas . | Control of neglected tropical diseases ( NTDs ) is suggested to be more cost-effective when drugs are co-administered through a single integrated delivery system rather than separate systems . An essential prerequisite for such efficiency gains is sufficient geographical overlap of the targeted diseases – lymphatic filariasis ( LF ) , onchocerciasis , schistosomiasis , soil-transmitted helminth infection and trachoma . Lack of data on geographical NTD distribution currently hampers the implementation of integrated control in many African countries . To generate the required data quickly and efficiently , integrated surveys of several NTDs simultaneously have been recommended . However , experience with integrated surveys is limited and requires additional research on cost and effectiveness to inform improvements in methodology and to guide scale-up . Here we analyse costs of the first integrated NTD survey round in Southern Sudan , generating average costs per implementation unit surveyed . Cost estimates are presented for use of the existing survey method and for modified versions . Key cost drivers were survey consumables and personnel , both of which are recurrent costs . These inputs could be reduced or put to more efficient use by modifying sampling for LF . To generate comparable cost estimates and identify key cost drivers in other settings we provide detailed cost data and guidance on how to replicate this work . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/helminth",
"infections",
"evidence-based",
"healthcare/health",
"services",
"research",
"and",
"economics",
"public",
"health",
"and",
"epidemiology/health",
"services",
"research",
"and",
"economics",
"infectious",
"diseases/neglected",
"tropical",
"d... | 2010 | Integrated Surveys of Neglected Tropical Diseases in Southern Sudan: How Much Do They Cost and Can They Be Refined? |
Regulation of transforming growth factor-β ( TGF-β ) signaling is critical in vertebrate development , as several members of the TGF-β family have been shown to act as morphogens , controlling a variety of cell fate decisions depending on concentration . Little is known about the role of intracellular regulation of the TGF-β pathway in development . E3 ubiquitin ligases target specific protein substrates for proteasome-mediated degradation , and several are implicated in signaling . We have shown that Arkadia , a nuclear RING-domain E3 ubiquitin ligase , is essential for a subset of Nodal functions in the embryo , but the molecular mechanism of its action in embryonic cells had not been addressed . Here , we find that Arkadia facilitates Nodal signaling broadly in the embryo , and that it is indispensable for cell fates that depend on maximum signaling . Loss of Arkadia in embryonic cells causes nuclear accumulation of phospho-Smad2/3 ( P-Smad2/3 ) , the effectors of Nodal signaling; however , these must be repressed or hypoactive as the expression of their direct target genes is reduced or lost . Molecular and functional analysis shows that Arkadia interacts with and ubiquitinates P-Smad2/3 causing their degradation , and that this is via the same domains required for enhancing their activity . Consistent with this dual function , introduction of Arkadia in homozygous null ( −/− ) embryonic stem cells activates the accumulated and hypoactive P-Smad2/3 at the expense of their abundance . Arkadia−/− cells , like Smad2−/− cells , cannot form foregut and prechordal plate in chimeras , confirming this functional interaction in vivo . As Arkadia overexpression never represses , and in some cells enhances signaling , the degradation of P-Smad2/3 by Arkadia cannot occur prior to their activation in the nucleus . Therefore , Arkadia provides a mechanism for signaling termination at the end of the cascade by coupling degradation of P-Smad2/3 with the activation of target gene transcription . This mechanism can account for achieving efficient and maximum Nodal signaling during embryogenesis and for rapid resetting of target gene promoters allowing cells to respond to dynamic changes in extracellular signals .
Transforming growth factor-β ( TGF-β ) signaling controls a diverse set of cellular processes , including cell proliferation , differentiation , apoptosis , and specification of fate in vertebrate and invertebrate species . Disruption of signaling leads to developmental abnormalities and disease , including cancer . Activin and Nodal TGF-β ligands have been shown to act as morphogens in vertebrate development [1–4] . For example , in the mouse , Nodal is required for gastrulation , including development of the anterior primitive streak and the formation of the germ layers , endoderm and mesoderm [5 , 6]; for maintenance of pluripotency in the epiblast [7 , 8]; and for the specification of the anterior-posterior [9 , 10] and left-right axes [11] . Loss-of-function mutations in the Nodal gene , including enhancer deletions , lead to a reduction of Nodal RNA [12] and reveal that the highest level of Nodal signaling is required during gastrulation for the induction of the anterior primitive streak . This contains the precursors of the mammalian equivalent of the amphibian Spemann's organizer , and it gives rise to the anterior endoderm , the node , and the mesendoderm ( notochord and prechordal plate ) , all of which are required for subsequent patterning of the vertebrate embryo [6] . Complementary experiments in Xenopus embryos , where increasing amounts of Nodal RNA are injected , show that it functions as a dose-dependent inducer and that the highest level induces Spemann's organizer [13] . The dynamic changes in the concentration of ligands , to elicit different cellular responses , demand that the responding cells have rapid turnover of the signaling-effectors and frequent refreshing of target gene promoters . Therefore , how TGF-β is regulated , and particularly , how signaling is terminated in the nucleus after gene transcription , is key in understanding cell fate decisions and patterning in vertebrate development . TGF-β signals bind to cognate serine/threonine kinase receptors leading to phosphorylation and activation of the Smad family of signal transducers . Two different Smad signaling branches have been described . Ligands , like Activin , Nodal , Gdf1 , Vg1 , and TGFβ1 are transduced by the receptor-activated Smad2 and Smad3 ( Smad2/3 ) [14 , 15] . The phosphorylated form of Smads ( phospho-Smads [P-Smads] ) complex with Smad4 and together translocate into the nucleus , where they function as transcription factors in association with DNA-binding partners such as FoxH1 , Mixer , Jun/Fos , Runx , ATF3 , and E2F4/5 , etc . , which provide target gene specificity [15] . In the mouse , loss-of-function mutations affecting core components of the Nodal signal transduction pathway give patterning and cell fate defects similar to that of Nodal itself [16–20] . Extracellular cofactors [21] , antagonists [22 , 23] , and proteases [24 , 25] have been shown to regulate Nodal activity during mouse development . However , little is known about the role of intracellular regulation in cells receiving Nodal . Intracellular regulators of the pathway include negative regulators such as inhibitory Smads ( Smad6/7 ) that block TGF-β signaling by competing with Smads for association with the receptors or by targeting receptors for ubiquitin-mediated degradation [26–28]; in the nucleus , Ski and SnoN incorporate in the Smad DNA-binding complex to prevent them from binding to the transcriptional coactivator p300/CBP and repress transcription by recruiting histone deacetylase [29 , 30] . More recently , a phosphatase , PPM1A/PP2Ca , has been identified and is shown to de-phosphorylate P-Smad2/3 [31] and abrogate their signaling activity . Furthermore , proteasome-mediated degradation of ubiquitin-modified core components of the TGF-β signaling cascade has been shown to play a major role in controlling signaling output [32] . Poly-ubiquitination and proteasome-dependent degradation of proteins is one of the most prominent turnover mechanisms in the cell . Ubiquitination of protein substrates involves a cascade of enzymatic reactions . E3 ubiquitin ligases are the critical components responsible for the recognition of specific substrates for ubiquitination [33] . They are generally classified into the HECT- and RING-domain classes and exhibit substrate specificity [34 , 35] . Several ubiquitin ligases are known to reduce signaling by mediating the degradation of individual components of the pathway [28 , 36] . However , ligases that terminate signaling by degrading activated Smads ( P-Smads ) have not been identified . One of the important unanswered questions is how long the activated Smads transcribe target genes and how the promoters are refreshed to allow rapid intracellular responses to dynamic changes in concentration of ligands . Both de-phosphorylation of P-Smads followed by cytoplasmic recycling [37] and proteasome-mediated degradation of P-Smads have been proposed to terminate signaling in the nucleus [38] . However , there was no explanation for how “used” versus “unused” activated effectors could be distinguished by these mechanisms . An obvious mechanism to limit the time that an effector works is to link its turnover with its ability to drive transcription . We have shown previously that Arkadia , a nuclear RING-domain ubiquitin ligase , enhances Nodal signaling and is essential for the induction of the organizer/node [39 , 40] . In somatic tumor cell lines , Arkadia has been shown to enhance TGF-β signaling by ubiquitin-mediated degradation of Smad6/7 [41 , 42] . However , we show here that Arkadia functions by a different mechanism in embryonic cells . Specifically , we find that Arkadia directly ubiquitinates and degrades P-Smad2/3 and that this is coupled with their high activity . The link between activity and degradation provides a mechanism to ensure that only “used” P-Smad2/3 effectors are degraded and that signaling is terminated at the end of the cascade and not before . Therefore , Arkadia can account for rapid resetting of actively transcribed promoters , forcing transcription to rely only on fresh P-Smads and allowing cells to respond to dynamic changes in signaling . Similar mechanisms may operate in other signaling pathways in development to achieve peak efficiency and dynamic responses of cells during development .
The phenotype of Arkadia−/− embryos consists of loss of anterior primitive streak derivatives ( node , notochord , prechordal plate , and anterior definitive endoderm [ADE]/foregut ) leading to anterior patterning defects including head truncations [39 , 40] . We have shown previously that while Arkadia or Nodal heterozygous ( +/− ) mice are normal , a small number of double heterozygotes for Arkadia and Nodal recapitulate the Arkadia−/− phenotype [39 , 40] . This suggested a functional interaction between Arkadia and Nodal . To investigate the extent of Arkadia's role in Nodal signaling , and in additional Nodal-dependent developmental events , we generated Arkadia−/− embryos with only one wild-type copy of the Nodal gene ( Akd−/− , Nodal+/− ) . The majority of Akd−/− , Nodal+/− embryos that were analyzed ( n = 19/33 ) exhibited phenotypes never observed in Arkadia−/− embryos . Using whole mount in situ hybridization , we performed marker analysis to define whether these phenotypes are Nodal-dependent ( Figure 1 ) . Before gastrulation , Nodal signaling is responsible for anterior-posterior axis specification via the induction of the anterior visceral endoderm ( AVE ) domain at the distal tip of the mouse embryo and its migration to the prospective anterior [2 , 43] . Expression analysis of the AVE markers Hex [44] and Cerl [45] demonstrates that the Arkadia−/− embryos are always able to induce an AVE , which correctly migrates to the prospective anterior of the embryo ( compare Figure 1A and 1D to 1B and 1E ) , while the primitive streak marker Brachyury is normally expressed posteriorly ( Figure 1A and 1B ) . However , Akd−/− , Nodal+/− embryos have incomplete AVE-specific gene expression , as they are rarely able to induce an AVE ( one out of seven ) that expresses Hex ( none out of two; Figure 1C ) , Cerl ( one out of three; Figure 1F ) , or Lefty1 ( none out of two; unpublished data ) . Furthermore , in the Akd−/− , Nodal+/− embryos , the Cerl-expressing embryo AVE domain remains distal ( Figure 1F ) and the Brachyury domain remains proximal ( Figure 1C ) , indicating that when the AVE is induced it cannot migrate to the anterior . In addition , more than 50% of the Akd−/− , Nodal+/− embryos examined at mid-streak stage have a constriction between the embryonic and extraembryonic compartments . This phenotype , which is not observed in Arkadia−/− embryos , is thought to be due to a failure of the AVE to migrate and define the anterior-posterior axis , leading to failure of primitive streak elongation and mesoderm formation along the embryonic-extraembryonic boundary [46] . These new phenotypes in Arkadia−/− embryos carrying only one Nodal wild-type allele indicates that Arkadia is also involved in AVE formation and suggests that it enhances Nodal signaling at pre-gastrulation stages . During gastrulation , Nodal is responsible for the formation and patterning of endoderm and mesoderm . Cardiac mesoderm is considered an anterior mesodermal tissue [47] , and Arkadia−/− embryos form heart [39] . In contrast , all Akd−/− , Nodal+/− embryos fail to form morphologically visible heart ( n = 4; Figure 1L ) , as it is also shown by loss of Nkx2 . 5 expression; one of the earliest markers of myocardial differentiation [48] ( Figure 1G−1I ) , Shh expression ( Figure 1J ) marks the mesendoderm , the midline of the neural tube ( floor plate ) , and the gut endoderm [49] . In Arkadia−/− embryos , Shh expression is reduced [39] due to the absence of mesendoderm and foregut , but it is present in the midgut and hindgut ( Figure 1K ) . Akd−/− , Nodal+/− embryos , however , have a severe reduction in Shh expression indicating not only loss of mesendoderm but also of all endoderm ( n = 2; Figure 1L ) , as confirmed by histological analysis ( unpublished data ) . These new phenotypes seen in Arkadia−/− embryos carrying only one Nodal wild-type allele indicate that during gastrulation , Arkadia is involved in the formation of the entire endoderm and anterior mesoderm and suggest that this is mediated by its ability to enhance Nodal . As Arkadia facilitates Nodal signaling broadly , before and during gastrulation , it is likely to be a regular partner factor of the Nodal signal transduction pathway . To find at what position within the Nodal signaling cascade Arkadia functions , we compared the level of the receptor-activated ( phosphorylated ) signaling effector P-Smad2 in embryos and embryonic stem ( ES ) cells by Western blotting ( Figure 2 ) . We examined 20 wild-type and 20 Arkadia−/− embryos at 8 . 5 days post-coitum ( dpc ) ( Figure 2A and unpublished data ) and three wild-type and three Arkadia−/− blastocyst-derived ES cell lines ( Figure 2G ) . We found that while the total levels of Smad2 protein remain the same , P-Smad2 was always at least two times higher in all Arkadia−/− embryos and ES cells compared to the wild-type samples . Similarly , the other Nodal signaling effector P-Smad3 was found to be elevated in Arkadia−/−ES cell lines ( Figure S1 ) . Therefore , in embryonic cells , P-Smad2/3 are more abundant in the absence of Arkadia than in its presence . As the phosphorylation of Smad2/3 depends on the kinase activity of the ligand-activated receptors , the data suggest that in the absence of Arkadia the receptors are more active or that P-Smad2/3 are more stable after their phosphorylation . We examined the activity of the receptors by stimulating Arkadia−/− and wild-type ES cells with Activin A ligand ( Activin ) and comparing Smad2 phosphorylation over time ( Figure 2B–2D ) . Although Arkadia−/−cells start with higher basal levels of P-Smad2 compared to wild-type , Smad2 phosphorylation peaks 1 h after Activin addition in all ES cell lines ( Figure 2B–2D ) , indicating normal receptor kinase activity in the presence or absence of Arkadia . Interestingly , after peak stimulation in wild-type ES cells , P-Smad2 decreases to basal levels within 2 h ( Figure 2B–2D ) , but in Arkadia−/− ES cells P-Smad2 is maintained at peak levels ( >90% ) for at least 6 more h ( Figure 2B ) . The data suggest that after receptor phosphorylation , Smad2 is more stable or maintains the phosphorylation longer in the absence of Arkadia . The total Smad2 levels do not change during the course of the experiment , indicating that the increased stability is associated with only the phosphorylated fraction ( Figure 2C and 2D ) . To examine the possibility that other factors , such as a receptor inhibitor , is induced specifically in wild-type cells causing reduction of Smad2 phosphorulation , we repeated the above experiment in the presence of a protein synthesis inhibitor ( cycloheximide ) . The decay of P-Smad2 was found , as before , to be slower in Arkadia−/− cells ( Figure S2 ) , suggesting that differences in protein stability rather than synthesis account for the increase in P-Smad2 levels . To exclude the possibility that the receptors generate the differences in P-Smad2 levels , we blocked all the TGF-β receptors with the serine/threonine kinase inhibitor , H7 ( Figure 2E and 2F ) or used SB431542 selective inhibitor of Alk receptors ( SB ) , which blocks the receptors that specifically phosphorylate Smad2/3 ( Figure S1A ) . We found that in wild-type cells treated with two different concentrations ( 5 or 25 μM of H7 ) , P-Smad2 levels declined 40% and 60% , respectively , within 30 min and they diminished to 30% after 90 min . In Arkadia−/− ES cells , however , even after 90 min , with the highest amount of inhibitor , P-Smad2 levels were not significantly changed ( Figure 2F ) . The data indicate that the receptors are not responsible for generating the increase of P-Smad2 in Arkadia−/− cells and suggest that this is caused by P-Smad2 stabilization . The total Smad2 protein levels do not change during the course of the experiment ( Figure 2E ) , but as the fraction of P-Smad2 is most likely small , differences within this fraction may not be visible when the total levels are examined . All of the above experiments were reproducible in three different Arkadia−/− ES cell lines ( unpublished data ) and the same results were obtained for the other Nodal/Activin effector , P-Smad3 ( Figure S1B ) . Collectively , the data suggest that Arkadia acts downstream of the receptors and destabilizes the phosphorylated forms of Smad2/3 . P-Smad2/3 complex with Smad4 and translocate to the nucleus where they activate target genes and are subjected to different mechanisms of turnover and signaling termination such as ubiquitination/proteasome-mediated degradation [38] or de-phosphorylation [31] and nuclear export [50 , 51] . Cytoplasmic retention of P-Smad2/3 can result in both inability to activate target genes and increased stability . As in the absence of Arkadia , P-Smad2/3 are more stable and less transcriptionally efficient; it is possible that they are cytoplasmic and Arkadia regulates their nuclear localization . We therefore examined in three different wild-type and three Arkadia−/− ES cell lines the localization of P-Smads using Western blots of nuclear and cytoplasmic fractions ( Figure 2G ) or by immunofluorescence ( Figure 2H ) with antibodies against P-Smad2 or Smad2/3 . We did not find evidence that P-Smads are cytoplasmic in the absence of Arkadia . On the contrary , the Western blots revealed that P-Smad2 accumulates in the nucleus of Arkadia−/−cells . Furthermore , as P-Smad2/3 complex with Smad4 to translocate to the nucleus [15] , an increase of Smad4 in the nucleus of Arkadia−/− ES cells was observed ( Figure 2G ) . Therefore , we conclude that Arkadia destabilizes P-Smad2/3 without affecting Smad4 complex formation and nuclear localization . Furthermore , as Arkadia is nuclear , the data suggest that Arkadia regulates P-Smad2/3 stability in the nucleus . Arkadia is an E3 ubiquitin ligase and could be destabilizing P-Smad2/3 directly by poly-ubiquitinating them , leading to their proteasome-dependent degradation . To test this hypothesis , we examined whether Arkadia interacts specifically with phosphorylated Smads . We used HEK293T ( 293T ) cells stably expressing moderate levels of full-length Arkadia , tagged either with Flag on the N-terminus and Myc on the C-terminus , or with green fluorescent protein ( GFP ) fused to the N-terminus ( GAkd ) . We performed immunoprecipitation ( IP ) with anti-Flag ( Figure 3A and 3B ) , -Myc , or -GFP ( unpublished data ) antibodies and Western blotted with anti-P-Smad2 , -Smad2 ( Figure 3A ) , or -P-Smad3 ( Figure 3B ) . The results show that with Activin stimulation , Arkadia coIPs with the phosphorylated endogenous Smad2 ( Figure 3A ) and Smad3 ( Figure 3B ) . However , when the cells are treated with the SB receptor inhibitor , which eliminates Smad2/3 phosphorylation , Arkadia does not coIP unphosphorylated Smad2/3 ( Figure 3A and unpublished data ) . Furthermore , we examined the interaction of Arkadia with other phosphorylated Smads ( Smad1/5/8; Figure S3A ) or with Smad4 ( unpublished data ) and found no evidence of interaction . We therefore conclude that Arkadia interacts specifically with the phosphorylated forms of Smad2/3 . To test how direct this interaction is , we performed the IP in vitro ( Figure 3C ) using in vitro transcribed/translated ( recombinant ) Arkadia protein labeled with S35 and phosphorylated Flag-tagged Smad2 isolated by IP ( Figure S3B ) from 293T cells stimulated with constitutive active Alk4 ( Alk4* ) . The data show that phosphorylated flag-Smad2 protein can IP recombinant full-length Arkadia but not the N-terminal portion ( 1–510 amino acids [aa] ) , or the luciferase control ( Figure 3C ) . The data suggest that Arkadia interacts directly with P-Smad2/3 via a C-terminal domain . Arkadia is a 989-aa protein and its C-terminal half ( 516–989 aa ) contains a highly conserved domain of 100 aa ( 889–989 aa ) , which includes the RING-ubiquitin ligase activity-domain at the C-terminal region ( 947–965 aa ) , a nuclear localization signal ( NLS ) ( 903–909 aa ) , and a conserved domain ( 889–903 aa ) , termed here NRG , of unknown function ( Figure S3C ) . To understand the interaction of Arkadia with P-Smad2/3 , we mapped further the responsible domain . We used transient transfections of 293T cells to test the ability of various deletions of Arkadia ( all GFP-tagged; Figure 3D ) to IP P-Smad2/3 . The data show that the last 100 aa of Arkadia containing the NRG , the NLS , and the RING ( G-NRG-RING , 889–989 aa; Figure 3D ) are sufficient for the interaction ( Figure 3D ) . Deletion of the C-terminal end of Arkadia ( GAkdR* ) that eliminates one of the Zinc-binding fingers of the RING domain ( 965–989 aa ) does not affect the interaction with P-Smad2/3 ( Figure 3F ) . However , deletion of the NRG completely abrogates the interaction with P-Smad2 ( Figure 3E ) . To confirm that the NRG domain is necessary for the interaction with P-Smad2/3 within the context of the full-length Arkadia protein , we generated an internal partial deletion of only the first eight residues of the NRG ( GAkdNRG* ) and showed that it diminishes the interaction with P-Smad2/3 ( Figure 3F ) . As judged by fluorescence from the GFP tag , all of the above mutant Arkadia proteins are localized in the nucleus and are expressed at comparable levels to full-length Arkadia ( Figure S4 ) , indicating that loss of the interaction of the various Arkadia deletion constructs is not due to instability or differential localization . Collectively , the above data indicate that Arkadia interacts directly with P-Smad2/3 via its 100-aa C-terminal portion and that within this domain , a 14-aa NRG motif is essential for this interaction . As the 293T cells that we used for the IPs do not express FoxH1 ( unpublished data ) , one of the major P-Smad2/3 transcription partners in early embryogenesis , we conclude that the interaction of Arkadia with P-Smad2/3 may not depend on a particular partner . As P-Smad2/3 interact with Arkadia , it is possible that they are substrates of Arkadia ubiquitination . To address this , we examined the ubiquitination status of P-Smad2 in the presence or absence of Arkadia expression . We used Arkadia−/− mouse embryonic fibroblasts lines ( MEFs ) to exclude any endogenous Arkadia activity and introduced Flag-Smad2 and Alk4* to obtain phosphorylated Flag-Smad2 , in the presence of full-length or mutant forms of Arkadia . Western blot analysis of the IPs with anti-P-Smad2 antibodies showed the existence of higher molecular weight forms of P-Smad2 associated specifically with the presence of full-length Arkadia ( Figures 4A and 4B and S4 ) suggesting poly-ubiquitination . Probing with ubiquitin antibodies confirmed that these modifications contain ubiquitin chains ( Figures 4A and 4B and S4 ) . Furthermore , these blots show that P-Smad2 is not ubiquitinated in the presence of mutant Arkadia proteins lacking either ubiquitin ligase activity ( GAkdR* ) or the P-Smad2 interaction domain ( GAkdNRG*; Figure 4A and 4B ) . Therefore , Arkadia ubiquitinates P-Smad2 in vivo , and this depends on both its ubiquitin ligase activity and the P-Smad2/3 interaction domain , suggesting that Arkadia ubiquitinates them directly . To verify that P-Smad2/3 ubiquitination is directly dependent on Arkadia , we performed the assay in vitro by adding all the components of ubiquitination separately along with recombinant full-length Arkadia , the C-terminal part of Arkadia containing the RING and the NRG , or N-terminal Arkadia . In this reaction we added the phosphorylated Flag-Smad2 obtained by IP as shown before ( Figure S3B ) , and examined by Western blot , with anti-P-Smad2 antibody , whether or not it becomes modified by the ubiquitination reaction . We found that the full-length and C-terminal Arkadia are capable of poly-ubiquitinating in vitro Flag-P-Smad2 , only when all the ubiquitination components were present , while the N-terminal Arkadia does not ( Figures 4C and S5B ) . In addition , the Flag-P-Smad2 substrate does not become ubiquitinated without the addition of recombinant Arkadia , indicating that the IP is not contaminated with ubiquitin ligases from the cells . Collectively , the data show that P-Smad2/3 are ubiquitinated directly by Arkadia in vivo and in vitro , and therefore , they are qualified substrates of Arkadia ubiquitination . Poly-ubiquitination of proteins usually leads to degradation via the proteasome [34 , 35] . Consistent with this , our data show that P-Smad2/3 are unstable in the presence of Arkadia ( Figure 2 ) , but it was unknown whether this instability is mediated by the proteasome . To address this , we transfected Arkadia−/− MEFs with full-length Arkadia or Arkadia lacking ubiquitin ligase activity under ligand stimulation and examined with an anti-P-Smad2 antibody the stability of transfected P-Flag-Smad2 ( Figure 4D ) or endogenous P-Smad2 ( Figure 4E ) in the presence or absence of MG132 proteasome inhibitor . We found that P-Smad2 levels are reduced specifically in the presence of full-length Arkadia and that MG132 can inhibit this . Therefore , Arkadia , via its ubiquitin ligase activity , is sufficient to induce P-Smad2 proteasome-dependent degradation . In the above experiments , the degradation of P-Smad2 was achieved by the transfection of exogenous Arkadia in −/− MEFs . To examine whether endogenous Arkadia is necessary for proteasome-dependent degradation of P-Smad2 , we compared its decay in Arkadia−/− and wild-type ES cells in the presence or absence of MG132 . For this we first stimulated the ES cells with Activin ( 1 h ) , then added SB inhibitor to prevent further phosphorylation of Smad2 and examined its decay at different time points ( Figure 4F ) . We found that MG132 protects P-Smad2 in wild-type ES cells but has very little effect in Arkadia−/− ( Figure 4F ) . Therefore , Arkadia , by direct poly-ubiquitination , most likely mediates degradation of P-Smad2 by the proteasome . We showed above that loss of Arkadia leads to the stabilization and nuclear accumulation of P-Smad2/3 . But do these higher levels correspond to an increase in target gene transcription ? Analysis of Arkadia−/− and compound Arkadia−/− , Nodal+/− embryos showed that Nodal signaling is defective ( Figure 1 and [39 , 40] ) suggesting that Smad2/3 target gene transcription is compromised . We examined the transcriptional activity of P-Smad2/3 in three Arkadia−/− and three wild-type ES cell lines . To estimate the relative levels of P-Smad2/3 transcriptional activity , we used two different target gene luciferase reporters , 0 . 9-P1 , ( hereafter termed Pitx2-luc ) regulated by P-Smad2/3 and its partner factor FoxH1 , and 9xCAGA–luc , a Smad3 specific reporter [52 , 53] . Although the Arkadia−/− ES cell lines always have a higher amount of P-Smad2 protein compared to wild-type , they have on average 30% ( Figure 5A ) lower luciferase from the wild-type cell line with the lowest activity ( WT 3 designated as reference = zero ) . Stimulation with Activin did not change significantly the luciferase reporter expression ( unpublished data ) ; indicating that under standard culture conditions ES cells exhibit ligand-saturated signaling ( autocrine signaling ) . In addition , real-time PCR showed that the expression of the endogenous Nodal gene , which like the Pitx2-luc luciferase reporter is regulated by FoxH1/Smad2/3 binding sites ( known as ASE ) [12] , is reduced by about 70% in Arkadia−/− ES cells ( Figure 5B ) . Together , these observations suggest that Arkadia is necessary for efficient target gene expression and suggest that in its absence , the stable and high levels of P-Smad2/3 are hypoactive or prevented from activating their target genes . Therefore , although Arkadia degrades P-Smad2/3 , it is necessary for efficient P-Smad2/3 transcriptional activity . To address whether Arkadia is sufficient to activate P-Smad2/3 , we performed gain-of-function experiments in ES cells . We introduced full-length Arkadia ( GFP-tagged; GAkd ) in three −/− ES cell lines and showed that it enhances the expression of the luciferase reporters on average 100% ( Figure 5C ) and 230% ( Figure 5D ) above the level of a GFP-expressing plasmid . Interestingly , in wild-type ES cells , Arkadia does not significantly change the reporter activity ( Figure 5C and 5D ) even after Activin stimulation ( unpublished data ) . This indicates that endogenous Arkadia is adequate to activate all P-Smad2/3 generated by the receptors . Together , these results suggest that Arkadia is necessary and sufficient to enhance P-Smad2/3 target gene transcription . According to the above data , Arkadia has two functions: to destabilize P-Smad2/3 and enhance their activity . As it is possible that different domains mediate the two opposing functions , we used the −/− ES cell functional assay to identify domains that are essential for Arkadia to enhance reporter activity . We found that Arkadia constructs with ubiquitin ligase domain mutations ( GAkdR* and GAkdR2* ) or Arkadia without the P-Smad2/3 interaction domain ( NRG deletion , GAkdNRG ) fail to enhance ( Figure 5E ) ; suggesting that like the degradation of P-Smad2/3 , activation also requires a direct interaction with Arkadia and its ubiquitin ligase activity . Examination of several N-terminal deletions ( Figure 3D ) showed that the C-terminal half ( 516–989 aa ) of Arkadia is the minimum region sufficient to enhance the reporter efficiently ( unpublished data ) . Furthermore , loss of the P-Smad2/3 activation , but not the degradation properties of Arkadia , is expected to convert it to a repressor of signaling . However , none of the deletions and mutations of Arkadia separated the two functions . Together , the above data suggest that Arkadia activates and degrades P-Smad2/3 via the same domains and that the two functions are most likely coupled . To test whether the activation of the hypoactive P-Smad2/3 by Arkadia occurs at the expense of their levels , we examined the relationship between levels of endogenous P-Smad2 and the degree of enhancement after expression of Arkadia in −/− ES cells . To visualize levels of endogenous P-Smad2 , we isolated by fluorescence-activated cell sorting ( FACS ) pure populations of cells transfected with GAkd or enzymatically ( ubiquitin ligase ) inactive Arkadia ( GAkdR* ) or control GFP constructs . A portion of the cells was used for luciferase assays and the rest was used to examine endogenous P-Smad2 and total Smad2 levels in Western blots . As before , Arkadia enhances signaling only in −/− ES cells in a ubiquitin-ligase-dependent manner ( Figure 6A ) ; and this phenomenon was accompanied by an 80% reduction in the level of P-Smad2 ( Figure 6B ) , confirming that activation of the hypoactive P-Smad2 is followed by its degradation . We performed the same experiment in wild-type ES cells and found that overexpression of Arkadia does not change the levels of endogenous P-Smad2 or the reporter activity ( Figure 6A and 6B ) . We conclude that in wild-type cells endogenous Arkadia must be in excess and sufficient to activate all available P-Smad2/3 in the nucleus . The fact that P-Smad2 is not eliminated and signaling is never repressed by the overexpression of Arkadia suggests that only a fraction of P-Smad2 interacts with and gets degraded by Arkadia , i . e . , nuclear P-Smads and perhaps those engaged in transcription . The above data confirm the dual role and coexisting functions of Arkadia , suggesting that Arkadia enhances P-Smad2/3 activity at the expense of their levels . An interesting observation is that in −/− ES cells , Arkadia expression not only restores the transcriptional deficit but it enhances reporter activity on average 50% ( line in Figure 5C ) or 100% ( Figure 5D ) above the maximum level that can be achieved in wild-type ES cells . The simplest explanation for this phenomenon is that the extra activity most likely reflects the accumulated levels of the hypoactive P-Smad2/3 that is being simultaneously activated by the expression of Arkadia . As this enhancement exceeds the maximum that can be achieved in wild-type ES cells even under ligand stimulation conditions , we termed it super-activation . According to this hypothesis , Arkadia expression in −/− ES cells will eventually “consume” ( activate and degrade ) the accumulated P-Smad2/3 , releasing their activity to produce a transient super-activation of target genes . Subsequently , target gene transcription will be reduced to basal levels similar to that of wild-type cells . We tested this prediction by comparing the percentage of enhancement by GAkd over that of the GFP control at different time points after transfection in Arkadia−/− ES cells . The results indicate that maximum super-activation occurs as early as 9 h post-transfection and coincides with the appearance of GFP fluorescence , declines after 15–18 h , and disappears after 30 h ( Figure 5F ) . GFP fluorescence remained high throughout the experiment and past 48 h ( unpublished data ) and does not account for the loss of super-activation at 30 h . The above data suggest that in −/− ES cells Arkadia releases the activity of the hypoactive and stable P-Smad2/3 pool causing target gene transcription above wild-type levels ( super-activation ) , and as this is a transient phenomenon , it occurs at the expense of P-Smad2/3 abundance . Therefore , Arkadia functions by a mechanism that consumes P-Smad2/3 as it activates them . All the above analysis shows that in ES cells Arkadia functions as a coactivator of P-Smad2/3 transcription . To address whether this also occurs in other embryonic cells and if this is the underlying cause of the Arkadia−/− phenotype in the embryo , we examined the expression of known Smad2 target genes in Arkadia−/− embryos . The FoxH1/P-Smad2 complex directly upregulates the Nodal gene and is responsible for its tissue-specific expression in the visceral endoderm ( VE ) at pre-gastrulation stages [10 , 12 , 54] . Whole mount in situ hybridization , as expected , revealed that in Arkadia−/− embryos ( n = 10 ) Nodal expression is dramatically reduced in the epiblast and almost lost in the VE ( Figure 7A and 7B ) . Later in development , FoxH1/Smad2/3 regulates directly Nodal [12] , Pitx2 , and Lefty2 [55 , 56] expression , in the left lateral plate mesoderm ( LPM ) . However , this expression cannot be assessed in Arkadia−/− embryos because they lack a node and mesendoderm , which are essential for establishing left-right asymmetry . Using tetraploid chimeras ( TC ) , we have previously shown that restoration of Arkadia expression in the extraembryonic lineages is sufficient to rescue the node and notochord formation in an embryo that consists entirely of Arkadia−/− cells [39] . We therefore generated TC embryos by injecting Arkadia−/− ES cells in tetraploid wild-type blastocysts . The rescue of node formation in the Arkadia−/− TC embryos is shown with the appearance of Nodal expression around the node ( Figure 7D ) , which is absent in Arkadia−/− embryos ( Figure 7E ) . We used these rescued Arkadia−/− embryos to examine target gene expression in the left LPM . In the left LPM , expression of Nodal , Lefty2 , and Pitx2 is present in wild-type ( Figure 7G , 7I , and 7K ) and absent in Arkadia−/− embryos ( Figure 7F ) , while in the TC embryos ( n = 9 ) , Nodal and Pitx2 expression is severely reduced ( Figure 7H and 7L ) and Lefty2 is undetectable ( Figure 7J ) , indicating that Arkadia is required for Nodal target gene expression in the LPM . Furthermore , the expression of Nodal in the node of the TC embryos is not reduced ( compare Figure 7C and 7D ) and is consistent with previous findings that this expression does not depend on Smad2/3 [12 , 57] . This confirms that Arkadia is required specifically for the expression of Smad2/3 target genes in vivo . Therefore , in the absence of Arkadia , Nodal target gene expression is reduced broadly in many cells and tissues throughout early embryogenesis , before and during gastrulation , and can account for the developmental defects observed in the −/− embryos . In addition , the effect of Arkadia on Nodal gene expression can explain its non-cell autonomous functions in the Arkadia−/− TC embryos where wild-type VE ( expressing Nodal ) can rescue node formation , and those reported in the literature for Xenopus assays [39 , 40] . According to the biochemical and functional data in ES cells , Arkadia interacts with P-Smad2 , and this interaction is essential for the full expression of target genes . It is therefore expected that Arkadia−/− embryos and ES cells will have similar defects and exhibit the same phenotypes with those of Smad2−/− cells . The phenotype of mice with conditional FoxH1 or Smad2 deletion exclusively in the epiblast shows that they cannot form ADE/future foregut or prechordal plate ( the most anterior mesendoderm ) . Furthermore , injection of FoxH1−/− or Smad2−/− ES cells in wild-type blastocysts shows that the −/− cells cannot colonize tissues such as the gut and the prechordal plate , indicating a cell autonomous requirement for these factors in these chimeras ( mosaic embryos ) [58 , 59] . It is therefore expected that Arkadia−/− embryos and ES cells will have the same defect and exhibit the same phenotypes with those of Smad2−/− cells . Expression of Hex , an ADE/foregut marker [44] ( Figure 7M and 7N ) and Shh , a prechordal plate marker ( Figure S6 ) , show that although Arkadia−/− TC embryos develop node and mesendoderm , they exhibit a deficit in these tissues . Consistent with the role of the ADE and the prechordal plate in maintaining and patterning the head folds , we observed that the Arkadia−/− TC exhibit reduction of the head folds ( Figure 7J , 7L , and 7N ) that can also account for the reduction of Pitx2 expression in the forebrain ( Figure 7L ) . The requirement for Arkadia expression within the cells that form the ADE/foregut and the prechordal plate was addressed in mosaic chimeras generated either by injection of wild-type ES cells into Arkadia−/− blastocysts ( Figure 7O ) or by Arkadia−/− ES cells into wild-type blastocysts ( Figure 7P ) [39] . In both types of chimeras , the embryo consists of a mixture of wild-type ( unstained ) and Arkadia−/− cells ( β-galactosidase stained ) [39] . We found that mosaic chimeras ( n > 100 ) exhibit normal morphology , as long as wild-type cells colonize the foregut and the prechordal plate ( Figure 7O and 7P ) . Therefore , like Smad2 , Arkadia is required cell autonomously for ADE/foregut and prechordal plate formation . The above data show that Arkadia loss-of-function phenocopies that of Smad2 in embryonic cells , supporting a functional interaction between the two factors .
Experiments using somatic tissue culture cell lines showed that Arkadia might enhance TGF-β signaling by degrading the Smad6/7 , which are known to inhibit Smad2/3 phosphorylation mainly by mediating the degradation of the receptors [41 , 42] . If this is the case , then embryos and ES cells should have reduced levels of P-Smad2/3 in the absence of Arkadia . Contrary to this prediction , all Arkadia−/− embryos and ES cells have at least 2-fold higher P-Smad2/3 levels compared to wild-type . Therefore , Arkadia must enhance signaling via a different mechanism during early development . In vivo , in ES cells , MEFs , and embryos , loss of Arkadia causes P-Smad2/3 stabilization and accumulation in the nucleus and makes them resistant to proteasome degradation ( Figure 4 ) , suggesting that Arkadia directly or indirectly is associated with their stability and degradation . Destruction of P-Smad2 had been shown previously to occur in the nucleus [38] , but a nuclear ubiquitin ligase that interacts specifically with the phosphorylated form of Smad2/3 , as well as the role of proteasome-mediated turnover of P-Smads during development , remained unknown . We present here a number of experiments indicating that Arkadia , a nuclear E3 ubiquitin ligase , interacts with P-Smad2/3 and directly ubiquitinates them leading to their degradation by the proteasome . These include IP experiments showing that Arkadia interacts specifically with the phosphorylated forms of Smad2/3 via the conserved NRG domain ( Figure 3A and 3B ) , and in vitro ubiquitination assays showing that Arkadia directly poly-ubiquitinates P-Smad2 ( Figures 4A–4C and S5 ) . Furthermore , introduction of Arkadia in −/− cells causes P-Smad2 poly-ubiquitination and decreases its abundance in a proteasome-dependent manner ( Figure 4D–4E ) . However , as Arkadia is nuclear , it most likely degrades nuclear P-Smads . One of the major questions is how effector activity is terminated after target genes have been activated and not before . This should involve a mechanism to distinguish between effectors actively engaged in transcription and those that are fresh and unused . A modification or a change of conformation of the effectors when they interact with the target-gene promoter complex may allow recognition by a critical component of the termination mechanism ( referred to subsequently as the terminator ) . In this case , absence of the terminator should lead to a prolonged , persistent response to signaling , as the unhindered effectors will continue to transcribe . On the other hand , excess terminator could cause repression of signaling if it degrades and deactivates the effectors prematurely , i . e . , as soon as they bind to the promoter . A more precise and efficient mechanism would involve degradation of the effectors after they activate at least one round of transcription . This requires a terminator directly involved in transcription by the effectors or an even more stringent mechanism , where initiation of transcription is linked with the destruction of the transcription factor: a “suicide” model . In this case , absence of the terminator will pause transcription , while its overexpression will not degrade/deactivate the effectors prematurely to repress target gene transcription . We show here that Arkadia fulfils the criteria of such a termination mechanism as it degrades P-Smad2/3 and enhances their target gene transcription via the same domains . Does Arkadia directly activate P-Smad2/3 ? In the absence of Arkadia , P-Smad2/3 are stable and accumulate in the nucleus , but they are repressed or hypoactive as target gene transcription is reduced and in some cases lost ( Figure 7 ) . Introduction of Arkadia in −/− ES cells not only restores signaling but super-activates P-Smads , meaning that the expression of the reporter reaches levels higher than those that can be achieved in wild-type ES cells under maximum ligand stimulation conditions . The above findings suggest that the extra activity is not generated by receptors or ligands , but by the release of activity from the pools of the accumulated hypoactive P-Smad2/3 . Consistent with this hypothesis , super-activation by Arkadia in −/− ES cells is a transient phenomenon , as it depends on the excess P-Smad2 that is used up . Therefore , Arkadia degrades and spends P-Smad2/3 to activate their transcriptional activity . These observations suggest that P-Smad2/3 is reminiscent of fuel , which in the absence of Arkadia is stable and can be stored , while in its presence it is “ignited” and “burns , ” releasing activity . These observations suggest that phosphorylation of Smad2/3 by the receptors is not sufficient for their transcriptional activity and that they require an additional step mediated by the function of Arkadia as a coactivator in the nucleus . We have shown previously that in Xenopus animal cap assay , injection of Arkadia RNA enhances the ability of Nodal but not that of Smad2 RNA to induce mesendoderm [40] . This appears inconsistent with our current data , however; as in the animal cap there is no ligand to phosphorylate Smad2 , it remains a mystery how Smad2 RNA injection alone induces mesoderm and mesendoderm . Furthermore , the amount of Smad2 RNA required for this is very high and most likely does not represent physiological signaling conditions . More importantly , we show here that Arkadia recognizes P-Smad2 and not Smad2 , and therefore it is not surprising that in the Smad2-injected animal caps Arkadia does not induce mesendoderm . While the exact mechanism of how P-Smad2/3 are activated by Arkadia remains unknown , it is important to point out that Arkadia does not degrade P-Smads directly , it only modifies them by ubiquitination and is the proteasome that degrades them . Although the role of ubiquitin modification in mediating proteasome degradation is well established , it has become apparent that addition of ubiquitin onto proteins may also affect their properties . Recent studies have suggested a direct role of ubiquitin modifications and of the proteasome in transcriptional activation ( reviewed in [60–63] ) . It is possible , however , that Arkadia interacts with , and degrades simultaneously , P-Smads and a closely linked repressor . Future studies will determine the mechanism of P-Smad2/3 activation and how this is coupled with immediate degradation . In conclusion , by linking P-Smad2/3 activation with degradation , Arkadia provides a way to achieve rapid resetting of P-Smad2/3 target gene transcription in development and allows efficient and dynamic responses to Nodal/TGF-β signaling events . We had shown previously that Arkadia−/− embryos do not develop anterior primitive streak and its derivative tissues: ADE/future foregut , mesendoderm ( prechordal plate and notochord ) , and node [39] . However , the molecular mechanisms that underlie these abnormalities were not understood . Specifically , we had shown that the formation of the node is restored in chimeric embryos consisting of tetraploid wild-type extraembryonic lineages ( such as VE ) and exclusively Arkadia−/− embryonic tissues derived from −/− ES cells [39] . This indicates that the development of the node in the embryonic lineages requires Arkadia expression and function within a different lineage ( extraembryonic ) . As Arkadia is nuclear and unlikely to be secreted , this effect was presumed to be mediated by a downstream-secreted factor . Here we show that Arkadia increases directly the transcription of P-Smad2 target genes , which include the Nodal gene itself , particularly in the VE and early epiblast when node precursors are induced ( Figure 7 ) . This suggests that in the Arkadia−/− TC , the wild-type VE provides sufficient Nodal to partially restore the overall level of Nodal in the Arkadia−/− epiblast , thus bringing it up to the threshold required for node formation ( Figure 7D and unpublished data ) . Therefore , organizer/node induction requires high levels of Nodal , which can be achieved with a considerable contribution from the VE . While the node may be rescued , to what extent expression of Arkadia in the extraembryonic lineages is sufficient to restore normal development in the Arkadia−/− embryo of the TC ? We show here that these embryos cannot form the most anterior derivatives of the primitive streak , such as ADE/foregut and prechordal plate ( Figure 7 ) , and exhibit laterality defects , including delayed turning and heart looping ( [39] and unpublished data ) . This is consistent with our findings that Arkadia is required for efficient P-Smad2/3 target gene transcription . Specifically , the reduction of Nodal expression in the left LPM along with its known target genes ( Lefty2 and Pitx2 ) can account for the left-right axis defects . The Nodal target genes that are responsible for foregut and prechordal plate development remain unknown . However , the development of these tissues must depend on the expression of Arkadia within these cells or their precursors . This was shown by the analysis of chimeric embryos consisting of a mixture of wild-type and Arkadia−/− cells , which revealed that the −/− cells could not contribute to these tissues ( Figure 7O and 7P ) . Furthermore , as Smad2−/− cells [59] and FoxH1−/− cells [58] behave similarly in chimeras , our data provide in vivo evidence for the functional interaction between Arkadia and Smad2 , and reveal that the three factors together regulate target genes essential for foregut endoderm and prechordal plate formation . Our data show that Arkadia is essential for the development of tissues that depend on very high Nodal signaling such as ADE/foregut , and that its introduction in −/− ES cells can boost signaling above levels obtained by just high concentrations of ligand “super-activation . ” This latter phenomenon is presumably generated by the simultaneous activation of the accumulated P-Smad2/3 , which has been stabilized and reached higher than normal levels in the absence of Arkadia . It is possible that such a mechanism may be responsible for maximizing Nodal signaling in the embryo . However , it would require the transient absence of Arkadia in the precursors of the ADE/foregut . Although Arkadia RNA is present broadly in the embryo , protein analysis is needed to reveal whether super-activation occurs in the embryo . In conclusion , we reveal a novel ubiquitin-mediated mechanism of TGF-β signaling regulation that depends on Arkadia , involves the activation of P-Smad2/3 signaling effectors downstream of receptor-phosphorylation , and couples their high activity with turnover and signaling termination . As the TGF-β/Nodal signaling pathway has been linked to cancers and genetic diseases , the identification of key molecular players such as Arkadia would be useful for the development of new drug targets and therapeutic intervention .
Chimeras were generated as described before [39] . Both adult mice and embryos were genotyped using allele-specific PCR amplification of genomic DNA . The oligonucleotide primer pairs TGAGGTAGGCATACCTAGAG and TGACTTAAGCCCTGCAATCC; TGAGGTAGGCATACCTAGAG and CTGAGTGATTGACTACCCGT; and TCTGGATTCATCGACTGTGG and CTGGATGTAGGCATGGTTGGTAG were used to give diagnostic amplification products of 313 bp for the wild-type Arkadia allele , 293 bp for the disrupted Arkadia allele , and 925 bp for the disrupted Nodal allele , respectively . Histology and in situ hybridizations and marker plasmids used are as previously described [39] . Total RNA was extracted from ES cells using Trizol ( Invitrogen , http://www . invitrogen . com ) followed by digestion with RQ1 RNase-Free DNase ( Promega , http://www . promega . com ) to remove DNA contamination . Synthesis of cDNA from total RNA was performed with SuperScript II Reverse Transcriptase ( Invitrogen ) . Experiments were performed in quadruplicates using the DNA Engine Opticon Real-Time PCR Detection System ( MJ Research , http://www . bio-rad . com ) and SYBRGreen PCR Master Mix ( Applied Biosystems , http://appliedbiosystems . com ) . The Nodal 375-bp amplicon was produced by the forward/reverse primer pair AAGACCAAGCCACTGAGCAT and GCCTTTGCACACAATTTCAA . Nodal expression was quantified by normalizing against endogenous controls GAPDH ( primer pair TGCACCACCAACTGCTTAGC and GGCATGGACTGTGGTCATGAG ) or YWHAZ ( primer pair CGTTGTAGGAGCCCGTAGGTCAT and TCTGGTTGCGAAGCATTGGG ) using the delta Ct method . Tagged Arkadia: Mus musculus ring finger 111 ( Rnf111 ) constructs were generated by eliminating the first 9 aa of Arkadia and fusing in frame to GFP ( pEGFP-cI; Clontech , http://www . clontech . com ) or Flag-tag ( synthetic oligonucleotide ) . A single Myc-tag ( synthetic oligonucleotide ) was fused to the carboxy-terminus of Arkadia . The various tagged Arkadia sequences and the GFP control gene were subcloned into SmaI/XhoI of the pTriEx2-hygro ( Novagen , http://www . novagen . com ) vector . The Arkadia RING domain mutations were constructed by either deleting the last 24 aa that included the second Zinc finger ( GAkdR* ) or by point mutation leading to amino acid substitutions ( C2–A2 ) that disrupt both Zinc fingers ( GAkdR2* ) . GAkdNRG* consists of an internal deletion of the first 7 aa of the NRG domain by PCR . The truncated Arkadia containing only the C-terminal region of Arkadia was constructed by fusing GFP in frame with the N-terminus of the truncated Arkadias . G-NRG-RING includes aa 889–989; G-NLS-RING , 903–989; GRING 947–989 . Pitx2-luc and 123-luc ( gifts of H . Hamada , Osaka , Japan ) and 9xCAGA-luc; 6 Myc-Smad3; constitutively active forms of Alk4 and Alk6 ( gifts of K . Miyazono , Tokyo , Japan ) ; HA-Ub ( gift of Y . Fujita , London , United Kingdom ) ; and the Flag-Smad2 construct ( gift of J . Smith , Cambridge , United Kingdom ) . The ES cell lines used in this study were derived directly from blastocysts as described previously [9] . Subsequently , they were maintained feeder-free in 20% FCS in DMEM and LIF ( Invitrogen ) . Cells were transiently transfected with Lipofectamine 2000 ( Invitrogen ) , Lipofectamine Plus ( Invitrogen ) , or TransIT-293 ( Mirus , http://www . mirusbio . com ) . Arkadia null primary mouse fibroblasts of mixed 129Sv/MF1 genetic backgrounds were isolated from embryos at 9 . 5 dpc and immortalized with pBabe-puro-SV40-TA ( gift from Parmjit Jat , Ludwig Institute , London , United Kingdom ) , selected with 0 . 3 μg/ml puromycin and maintained in F12 ( Invitrogen ) medium with 20% FCS . Signaling stimulation was achieved using 50 μng/ml BMP2 or 10 μng/ml Activin A ( Sigma , http://sigmaaldrich . com ) ; signaling inhibition using serine/threonine kinase inhibitor H7 ( Calbiochem , http://www . calbiochem . com ) or SB431542 ( Sigma and gift from GSK ) ; and translation inhibition using 100 μg/ml cycloheximide ( Sigma ) . Cells were cotransfected with the Firefly reporter plasmids , a Renilla luciferase transfection control pRL-SV40 ( Promega ) , and various GFP-tagged Arkadia constructs in a 5:1:5 ratio . Dual luciferase assays were performed according to manufacturer's protocols ( Promega ) 18 h after transfection , unless stated otherwise . After normalizing the Firefly values to Renilla , the data was presented as relative luciferase values or as percent increase . For IP studies , cells in 10-cm culture dishes were treated for 5 h with 50 μM of the proteasome inhibitor MG-132 ( Calbiochem ) before lysis . The cells were lysed by adding 1 ml per 10-cm dish of RIPA ( 150 mM NaCl , 50 mM Tris [pH 8 . 0] , 0 . 5% DOC , 0 . 1% SDS , and 1% NP-40 ) or NP-40 buffer ( 20 mM Tris [pH 7 . 5] , 150 mM NaCl , 0 . 5% NP-40 ) , which was specifically used for the IPs shown in Figure S5A . Both IP solutions contained 10 mM N-ethylmaleimide ( cysteine protease inhibitor , Sigma ) , 100 μM MG-132 ( Calbiochem ) , 100 μM Epoxomicin ( Calbiochem ) , 100 μM clasto-Lactacystin β-Lactone ( Calbiochem ) , and protease/phosphatase inhibitor cocktails ( Sigma ) at a final concentration of 1% each . After centrifugation at 100 , 000 g for 30 min , the supernatants were immunoprecipitated for 1 h with the appropriate antibody and 50 μl of protein agarose beads ( Sigma ) . 2 μg of an anti-GFP mouse monoclonal antibody ( clone 7 . 1 and 13 . 1; Roche , http://www . roche . com ) or 2 μg of the anti-FLAG M2 antibody ( Stratagene , http://www . stratagene . com ) with 50 μl of Protein G Sepharose beads ( Amersham Biosciences , http://www . gehealthcare . com ) was used for the IP . Samples were eluted off the beads by boiling in 2× Laemelli buffer followed by standard SDS-ployacrylamide gel electrophoresis ( SDS-PAGE ) and Western blot analysis . For nuclear and cytoplasmic fractionation , cells were lysed in hypotonic buffer ( 20 mM Hepes [pH7 . 5] , 10 mM NaCl , 0 . 2 mM EDTA , 20% glycerol , 1 . 5 mM MgCl2 , 0 . 1% Triton X-100 ) containing protease and phosphatase inhibitors as above . Nuclei were pelleted by centrifugation at 1 , 000 rpm for 10 min , and the cytoplasmic fraction obtained by retaining the supernatant . Nuclear extracts were obtained by rocking the nuclear pellet in five times the volume of hypertonic solution ( hypotonic buffer + 500 mM NaCl ) for 1 h at 4° C and subsequent centrifugation at 13 , 000 rpm for 5 min . Nuclear fraction was obtained in the supernatant , and 30 μg of each protein sample was loaded on each lane . When required to inhibit protein degradation , MG-132 ( 50 μM ) was added to both the hypotonic and hypertonic buffers . For Western blotting , cells and embryos were lysed with the RIPA buffer as for IPs . 30 μg of each protein sample was loaded on each lane for SDS-PAGE . The following antibodies were obtained and used according to the manufacturer's instructions: rabbit anti-P-Smad2 ( Calbiochem ) , rabbit anti-P-Smad3 ( Cell Signaling Technologies , http://www . cellsignal . com and also gift from E . Leof , Mayo Clinic , United States ) , rabbit anti-P-Smad1/5/8 ( Cell Signaling Technologies ) , rabbit anti- Smad2 ( Zymed Laboratories , http://www . invitrogen . com ) , rabbit anti-GFP ( Abcam , http://www . abcam . com ) , rabbit anti- cyclin D2 ( H-295; Santa Cruz Biotechnology , http://www . scbt . com ) , as well as mouse monoclonal antibodies against Smad4 ( B-8; Santa Cruz ) ; actin ( Santa Cruz ) ; PCNA ( PC-10; Santa Cruz ) ; tubulin ( B-5-1-2; Sigma ) ; FLAG ( M2; Stratagene ) ; histone H3 ( Upstate , http://www . upstate . com ) ; ubiquitin ( Covance and Bethyl Laboratories , http://www . bethyl . com ) and HA ( Roche ) . [35S]-methionine-labeled or unlabeled recombinant proteins full-length Arkadia , N-Akd ( 1–516 aa ) , C-Akd ( 510–989 ) , and luciferase were generated by the TNT Quick Coupled in vitro transcription/translation kit ( Promega ) . Flag-P-Smad2 was obtained with IP from 293T cells . Beads bound to Flag-P-Smad2 protein were washed with RIPA buffer and incubated for 1 h by constant rotation at 4 °C with radiolabeled recombinant proteins in 1 ml of RIPA buffer . Protein-bead complexes were then washed four times with RIPA buffer ( 200 × bed volume ) and re-suspended in 2 × Laemelli buffer . The presence of radiolabeled Arkadia protein in pull-downs was detected by SDS-PAGE and autoradiography . Assays were carried out in 20 μl of ubiquitination assay buffer ( 20 mM Tris-HCl [pH7 . 7] , 100 mM KCl , 0 . 1 mM CaCl2 , 1 mM MgCl2 , 1 mM DTT ) containing 15 μg of GST-ubiquitin ( Boston Biochem , http://www . bostonbiochem . com ) , 1 μg of E1 ( Boston Biochem ) , 1 . 5 μg of E2 ( GSTUbcH5b and GST-UbcH5c; Boston Biochem ) , 0 . 6 μl of Energy Regeneration Solution ( Boston Biochem ) , and Flag-P-Smad2 immunoprecipitated from 293T cells . In vitro translated Arkadia and Arkadia mutants ( N-terminal corresponding to aa 1–516 and C-terminal truncation corresponding to aa 510–989 ) were generated using the TNT Quick Coupled in vitro transcription/translation kit ( Promega ) . Reactions were incubated for 1 h under constant rotation at 37 °C . The reaction was terminated by the addition of 2 × Laemelli buffer , and the presence of poly-ubiquitinated substrates was detected by Western blotting . Immunofluorescence was performed using the following primary antibodies: anti-P-Smad2 antibody ( 1:50; Calbiochem ) ; anti-Smad2/3 ( 1:100; BD Biosciences , http://www . bdbiosciences . com ) ; Alexa Fluor 568 anti-rabbit secondary ( 1:400; Molecular Probes , http://www . invitrogen . com ) ; and mounting medium with DAPI ( Vectashield , http://www . vectorlabs . com ) . Arkadia−/− and wild-type ES cells were grown on cover slips and rinsed in PBS for 5 min prior to fixation with 4% ( w/v ) paraformaldehyde in PBS for 10 min . Following fixation , cells were permeabilized with 0 . 5% Triton-X 100 in PBS for 2 min on ice and rinsed with PBS for 5 min and incubated in 10% FCS/PBS for 30 min to block nonspecific binding of antibodies . Subsequently , the cells were incubated with the appropriate primary antibody diluted in 10% FCS/PBS for 1 h at room temperature . A no primary control was included to verify antibody specificity . Cover slips were then washed several times with 10% FCS/PBS and incubated with secondary antibody for a further 1 h . This was followed by washes in PBS and cover slips mounted for fluorescence in medium containing DAPI ( Vectashield ) . The cells were visualized on a Leica TCS SP2 ( http://www . leica . com ) confocal microscope at 100× magnification .
The GenBank ( http://www . ncbi . nlm . nih . gov/genbank ) accession numbers for the entities from the discussed in this paper are Arkadia ( NM_033604 ) , GAPDH ( NM_001001303 ) , Nodal ( NM_013611 ) , and YWHAZ ( NM_011740 ) . | In development , cells respond to secreted signals ( called morphogens ) by turning on or off sets of target genes . How does gene activity adjust quickly in response to rapidly changing extracellular signals ? This should require efficient removal of old/used signaling effectors ( signal-activated transcription factors ) from the promoters of target genes to allow new ones to assume control . We previously discovered Arkadia , an E3 ubiquitin ligase , and showed that it is an essential factor for normal development . ( Ubiquitin ligases trigger the addition of ubiquitin residues to proteins , typically marking them for degradation . ) Here , we show that Arkadia is required for high activity of the major signaling pathway , TGF-β/Nodal . Arkadia has a dual role to degrade Smads , the TGF-β signaling effectors , and enhance their transcriptional activity . This coupling of degradation with activation provides a mechanism to ensure that only effectors “in use” are degraded , allowing the new ones to proceed . It is possible that very similar mechanisms operate in other pathways to establish dynamic regulation and efficient signaling , while their failure may be associated with developmental abnormalities and disease , including cancer . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"biochemistry",
"developmental",
"biology",
"mus",
"(mouse)",
"mammals"
] | 2007 | Arkadia Enhances Nodal/TGF-β Signaling by Coupling Phospho-Smad2/3 Activity and Turnover |
Plasmodium vivax is highly endemic in the lowlands of Papua New Guinea and accounts for a large proportion of the malaria cases in children less than 5 years of age . We collected 2117 blood samples at 2-monthly intervals from a cohort of 268 children aged 1 to 4 . 5 years and estimated the diversity and multiplicity of P . vivax infection . All P . vivax clones were genotyped using the merozoite surface protein 1 F3 fragment ( msp1F3 ) and the microsatellite MS16 as molecular markers . High diversity was observed with msp1F3 ( HE = 88 . 1% ) and MS16 ( HE = 97 . 8% ) . Of the 1162 P . vivax positive samples , 74% harbored multi-clone infections with a mean multiplicity of 2 . 7 ( IQR = 1–3 ) . The multiplicity of P . vivax infection increased slightly with age ( P = 0 . 02 ) , with the strongest increase in very young children . Intensified efforts to control malaria can benefit from knowledge of the diversity and MOI both for assessing the endemic situation and monitoring the effects of interventions .
Malaria caused by Plasmodium vivax infection is increasingly recognized as a public health burden . The worldwide population at risk is estimated to be 2 . 85 billion , with high prevalences observed in locations throughout Southeast Asia and the Pacific [1] . Even though P . vivax epidemiology is less well studied and understood compared to that of P . falciparum , it is thought that P . vivax will present a greater challenge on the way to elimination of malaria outside Africa . Papua New Guinea ( PNG ) presents a variety of different climatic and ecological zones which have differing levels of malaria transmission [2] with a high burden of P . vivax in the tropical lowlands . The occurrence of high prevalence and morbidity marks locations in PNG as suitable field sites for P . vivax drug trials [3] , [4] and potentially also for future vaccine trials [5] . In Maprik in northern PNG , both P . vivax and P . falciparum are highly prevalent [6] . The incidence of P . vivax clinical episodes has been shown to peak in the second year of life , while that of P . falciparum increases until the fourth year [7] , [8] . Children between one and five years are considered to be a target age-group for P . vivax vaccine trials [5] . While previous studies have provided baseline data on clinical incidence rates , epidemiological patterns and the age-distribution of disease [8] , [9] , genotyping data on individual P . vivax clones describing their diversity and molecular epidemiological parameters are scarce . For P . falciparum , the mean number of concurrent infections per patient ( multiplicity of infection , MOI ) has been used as one of several measures of the impact of interventions . MOI is crucial to assess the risk that an individual carries a drug resistant parasite and to evaluate levels of inbreeding [10] . Here we describe the genetic diversity and multiplicity of infection of a P . vivax population in an area of high malaria prevalence in the Maprik District in PNG . Data were obtained from children aged 1 to 4 . 5 years who were followed-up over 16 months [8] . We make use of two markers to genotype individual P . vivax clones , one a microsatellite and the other a region of the msp1 gene , encoding the Merozoite Surface Protein 1 . While the microsatellite MS16 ( located on NCBI contig XM_001615468 . 1 ) is considered a neutral marker that has been used in a number of population genetics studies [11] , [12] , msp1 ( XM_001614792 ) encodes a potential vaccine candidate ( reviewed in [13] ) and its diversity has been studied in different settings [14] , . In a previous study , the two markers showed robust PCR amplification and high diversity with a risk of less than 1% that two clones share the same two-loci haplotype [17] .
The cohort study was approved by institutional review boards of the PNG Medical Research Advisory Committee ( approvals 05 . 19 and 09 . 24 ) , University Hospitals Case Medical Center ( Cleveland , Ohio USA ) , and the Ethikkommission beider Basel ( approval 03/06 ) . Informed written consent was provided by the parents or legal guardians of each child . The cohort study was conducted in the Ilaita area , Maprik District , East Sepik Province , PNG between April 2006 and August 2007 . The study area has hyper- to holoendemic perennial transmission with moderate seasonal variation [8] , [18] , [19] , [20] . P . vivax infections are the most prevalent infection in young children and remain frequent into adulthood , while P . falciparum is the predominant infection in children over 4 years of age [6] , [21] . 268 children aged 1 to 3 years at enrolment were followed-up over a period of 16 months . As part of the cohort study , the children were visited every two months with blood samples taken at least for one and , for some surveys , on two consecutive days . In the analysis presented here , only blood samples taken on the first day were included . The prevalence of Plasmodium species by microscopy in the study population was 44 . 3% for P . vivax , 32 . 6% for P . falciparum and 4 . 2% for P . malariae [8] . Defining clinical episodes as the presence of fever >37 . 5°C ( axillary temperature measured twice and a third time if the difference was above 0 . 3° ) or history of fever during the last 48 hours together with parasitaemia observed by light microscopy , the clinical incidence rates were 2 . 56 P . falciparum and 2 . 46 P . vivax episodes per child per year [8] . Children presenting with parasitologically confirmed malaria ( i . e . positive blood slide or RDT ) were treated with Coartem . Details of the study have been published previously [8] , as well as genotyping data of the population of P . falciparum clones in this cohort [22] . Finger prick and venous blood samples were collected and DNA was extracted as previously described [8] . The presence of P . falciparum , P . vivax , P . ovale and P . malariae was detected by light microscopy as well as by post-PCR Ligase Detection Reaction ( LDR ) [8] , a molecular method for the detection and species identification of malaria parasites [23] . All samples which were P . vivax positive by light microscopy or LDR plus 88 negative baseline samples were selected for genotyping . The selection of highly diverse molecular markers is crucial for assessing MOI in molecular epidemiology studies . Based on our previous results [17] , we selected the polymorphic marker gene msp1F3 and the microsatellite MS16 for genotyping . In contrast to population genetic studies , where large numbers of neutral markers of moderate to high diversity ( e . g . microsatellites ) are generally analyzed , a small number of highly polymorphic markers are ideal for tracking clones in epidemiological studies for two reasons Where concurrent infections with several clones are common construction of haplotypes combining data from several PCR amplified markers is difficult . In addtion , when MOI is defined as the maximum number of clones by any of the markers typed , the risk of overestimating MOI due to PCR artefacts increases with the number of markers analyzed , in particular for microsatellite amplification where PCR artefacts caused by polymerase slippage are of concern [24] ( Tables A and B in Text S1 ) . Any highly polymorphic marker is suitable for studying the epidemiology of multiple infections . As long as high diversity is maintained , there is no need for selective neutrality . In this molecular epidemiological study , the coding sequence msp1F3 was chosen because it harbours a more complex repeat structure than microsatellites . This has the advantage that PCR artefacts due to slippage are rare for msp1F3 . Our second marker was MS16 , a highly polymorphic microsatellite that lacks dominant alleles . PCR and capillary electrophoresis were performed with slight modifications of the published protocol [17] to save costs and labour time: a multiplex primary PCR was done with the primers for the 2 markers msp1F3 and MS16 followed by individual nested PCRs for msp1F3 and MS16 . The primary PCR was done in a volume of 20 µl containing 1 µl template DNA , 0 . 25 µM of each primer ( Eurofins MWG Operon ) , 0 . 3 mM dNTPs ( Solis BioDyne ) , 2 mM MgCl2 , 2 µl Buffer B ( Solis BioDyne ) and 5 U TaqFIREPol ( Solis BioDyne ) . As we expected low parasitemia in samples negative by microscopy , 2 µl DNA instead of 1 µl were used for the primary PCR . 1 µl primary PCR product was used as the template for the nested PCR , which was performed in a volume of 20 µl containing 0 . 25 µM of each primer ( Applied Biosystems ) , 0 . 2 mM dNTPs ( Solis BioDyne ) , 2 mM MgCl2 , 2 µl Buffer B ( Solis BioDyne ) and 1 . 5 U TaqFIREPol ( Solis BioDyne ) . The forward primers for the nested PCR were labelled with fluorescent dyes: 6-FAM for msp1F3 , NED for MS16 . Cycling conditions were as follows: initial denaturation 95°C for 1 minute , then 30 cycles ( primary PCR ) or 25 cycles ( nested PCR ) with 15 seconds denaturation at 95°C , 30 seconds annealing at 59°C and 30 seconds elongation at 72°C plus a final elongation of 5 minutes at 72°C . Subsequently , capillary electrophoresis was performed as described [17] . The PCR data was analysed using the GeneMarker® programme version 1 . 85 ( SoftGenetics ) . Based on experience from preliminary studies , peaks above a cut off of 1000 units relative fluorescent intensity ( RFU ) were considered true amplification products , all peaks below this cut off were considered background noise as well as lesser peaks in the vicinity of strong peaks reaching 40% ( msp1F3 ) and 70% ( MS16 ) of their height . Occasionally the fluorescence intensity differed between plates or samples as indicated by varying signal intensities of the commercial size standard . To compensate for this technical shortfall , the standard cut off value was lowered from 1000 to 300 RFU if signal intensities of both sample peaks and size standard peaks were low ( generally below 1000 RFU ) . This practice was justified by a greater agreement in positivity at both loci . As a consequence , the proportion of samples positive only for a single marker dropped from 12% to 10% . All samples were checked visually ( after blinding of samples ) for stutter peaks thereby excluding one msp1F3 and 25 MS16 samples from further analyses . The genotyping method was validated by typing a subset of 28 samples for both markers in duplicate . 80% of msp1F3 clones and 88% of MS16 clones were detected in both replicates ( Tables C and D in Text S1 ) . An important reason for the imperfect detection of clones is the low concentration of template DNA in samples with scanty parasitemia , where partial amplification of all templates seems to be governed by chance . In serial dilutions of DNA in field samples , we have demonstrated this effect by performing PCR amplification in triplicate for each dilution ( Tables E and F and Figures A and B in Text S1 ) . At low DNA concentrations , the allelic composition of a blood sample differed between replicates with individual clones detected in an apparently random fashion ( out of several clones detected in undiluted DNA ) . Alleles were grouped into bins of 3 base pairs , defined by the expected size differences in the two markers: 3 base pairs ( bp ) for the coding region of Pvmsp1 as well as for microsatellite MS16 harbouring a 3 bp repeat unit . When a single genotype was observed with both markers , a blood sample was defined as single clone infection . In multiple clone infections , the highest number of clones observed for either marker defined the combined MOI of a blood sample . We used the kappa statistic to describe agreement between the molecular markers after correcting for chance agreement . The determination of MOI for single and double clone infections was validated by genotyping a subset of samples with 12 additional markers . MOI = 1 was confirmed in 67/92 ( 72 . 8% ) samples and MOI = 2 in 31/32 ( 96 . 9% ) samples ( Tables A and B in Text S1 ) . Although the distributions of MOI are skewed , we present the mean MOI , a common measure , to allow comparisons with other studies . We estimated the effect of age ( at the time of the survey in 6 months age groups ) and season on prevalence and MOI using regression models . To account for multiple visits per child , we included a random effect for child . The models were implemented in STATA version 10 [25] and WinBUGS version 1 . 4 [26] . The genetic diversity of a given locus in a population is expressed by the virtual heterozygosity HE , i . e . the probability that two clones taken at random from the population carry different alleles . HE was calculated using the formulawhere n is the number of clones analysed and p is the frequency of allele i . HE of msp1F3 and MS16 were determined by using only the first P . vivax positive sample of each study participant; HE of msp1F3/MS16 haplotypes by using the first single clone infection per patient . This procedure prevents potential sampling bias due to repetition of persisting clones from the same individual . Linkage between markers was assessed using LIAN 3 . 5 [27] . Linkage disequilibrium measured from only two markers cannot provide information on inbreeding and was used only to provide evidence that the markers occur independently of each other .
The overall prevalence of P . vivax based on positivity by PCR was 55% . The prevalence was lowest in children under 1 . 5 years at 44% and reached 62% in children aged 3 to 3 . 5 years . The increase of prevalence by age at the time of the survey was significant ( P = 0 . 005 ) largely driven by the lower prevalence in children less than 1 . 5 years . Without this youngest age group no evidence of a trend was observed . No major seasonal trend in P . vivax prevalence was observed with the exception of a slight peak in September ( P = 0 . 17 ) . In 1162 samples positive for P . vivax , 57 different msp1F3 and 103 different MS16 alleles were detected ( Figures 1A and 1B ) . Virtual heterozygocity HE was 97 . 8% for MS16 , 88 . 1% for msp1F3 and 99 . 1% for msp1F3-MS16 haplotypes determined in single-clone infections ( Table 1 ) . In 219 single clone infections from 148 patients , a total of 154 different haplotypes were observed with the most common haplotype detected in only six individuals ( Figure 1C ) . We tested this data set for independence of the two molecular markers . No linkage disequilibrium was observed ( IAS = −0 . 0001 , P = 0 . 53 ) . The MOI was determined for each marker separately , as well as for both markers combined . In P . vivax positive samples , the mean MOI was 2 . 27 for each marker individually and 2 . 69 when calculated from the maximal number of clones per sample by any marker . Among samples for which positive results were obtained from both markers , MOI was concordant in 38% ( 397/1050 ) ( kappa = 0 . 17 ) . A difference of one clone was observed in 38 . 5% ( 404/1050 ) of samples . The frequency distribution of MOI plotted separately for msp1F3 and MS16 was compared to the combined MOI ( Figure 2 ) . A single marker slightly underestimated MOI . Multiple clone infections were observed in 63% of all positive samples by msp1F3 and in 61% by MS16 . When results of both markers were combined , the proportion of multiple clone infections increased to 74% ( Table S1 ) . Likewise , the proportion of samples with a MOI of 3 or higher was underestimated based on a single marker . Of the 531 samples with a combined MOI between 3 and 9 , the combined MOI result was reproduced only in 279 samples ( 52% ) by msp1F3 alone and in 324 samples ( 61% ) by MS16 alone . The mean MOI of P . vivax was associated with age . In children up to 1 . 5 years the mean MOI was 2 . 4 increasing slightly up to 2 . 8 in children 3 . 5 to 4 . 5 years of age ( Figure 3 , P = 0 . 02 ) . If the youngest children below 1 . 5 years were excluded , no significant trend was observed ( P = 0 . 23 ) . The increase of the proportion of children bearing more than two clones was more pronounced ( Figure 4 ) . In the 18 youngest children aged 300 to 400 days , we observed a low MOI of 1 . 67 and only two children ( 11% ) carried more than 2 clones . There was no significant seasonal variation in MOI ( P = 0 . 50 ) .
We have genotyped P . vivax parasites in a cohort of 268 children from an area of Papua New Guinea with sympatric P . falciparum and P . vivax with prevalences of 49 . 6% and 53 . 0% respectively in this cohort at enrolment [8] . The ecology and epidemiology of P . vivax differs from that of P . falciparum in several aspects . Parasite densities are generally lower , which is likely to affect prevalence data generated by both microscopically or molecular diagnosis [28] , [29] . Gametocytes appear shortly after an infection is established in a host [30] , with implications for transmission and the frequency of sexual recombination , and the occurrence of relapses leads to appearance of new genotypes in the blood stream independent of mosquito transmission . Our genotyping enables the distinction of individual parasites within the human host and thus the assessment of MOI . The mean MOI for P . vivax in our cohort was 2 . 7 and 73 . 6% of samples carried multiple clones ( combined results from two independent markers ) . This compares to a substantially lower P . falciparum MOI of 1 . 5 in the same cohort determined by the marker msp2 and only 35 . 2% of samples carrying multiple clone infections [31] . The multiplicity of Plasmodium infections depends on a number of factors including transmission intensity or the duration of infection as the result of loss of infection and antimalarial treatment . An additional factor , unique to P . vivax , contributes to the number of blood stage infections circulating in the blood: relapses of semi dormant liver stages . Previous studies have shown that relapses often genetically differ from already present blood stage parasites [32] , [33] and thus lead to increased MOI . As Coartem does not clear liver stage parasites and the level of treatment in our cohort was high , the combined effect of treatment and relapses is likely to enhance differences between P . vivax and P . falciparum MOI . In addition , as mosquitoes biting people harbouring multi-clone infections are more likely to transmit several clones concurrently [34] , [35] , the higher MOI among P . vivax blood-stage parasites will increase the likelihood that multi-clone P . vivax infections are transmitted in a single mosquito bite . Under intense transmission such as found in lowland PNG , the MOI of P . vivax species increases with age , with the increase most pronounced in children below 1 . 5 years . In an earlier study , no evidence of differences between children aged 4 to 14 were found [36] . This increase in early childhood may at least in part be related to the increased exposed body surface with child growth thus leading to higher rates of mosquito bites and consequently risk of infection [37] . In addition , with rapidly increasing immunity fewer P . vivax infections may reach high densities which are associated with febrile illness and antimalarial treatment [8] and the average duration of a P . vivax infection may increase with age . Both the mean MOI and prevalence using genotyping data showed no pronounced seasonality , concurring with previous findings using light microscopy and LDR detection [8] . In contrast , the incidence of clinical disease increased in the wet season [8] . The lack of annual fluctuations in P . vivax prevalence and MOI observed in PNG is likely to be caused by relapses during periods where there is less mosquito transmission . The proportion of multiple-clone infections clearly differs from observations from countries of lower P . vivax endemicity . We observed polyclonal infections in three out of four samples . Even if numbers cannot be compared directly to other studies using other genotyping protocols ( higher numbers of markers increase the chance of observing several clones at least in one marker ) , this proportion only reaches 11 to 49% in the Amazon [12] , [38] , 55% in Sri Lanka [11] but 73% in Myanmar [11] . The molecular markers msp1F3 and MS16 showed a high degree of genetic diversity . While three msp1F3 alleles reached frequencies above 10% , the distribution of MS16 alleles is more homogenous with the highest frequency of 5% . The selected markers are suitable for studies where high resolution discrimination between P . vivax clones is required , such as longitudinal tracking of clones or discrimination between existing and incoming infections . The probability of two individual clones sharing the same 2-loci haplotype was below 1% . Analysing additional polymorphic markers would lead only to a minimal improvement of discrimination . As the mean MOI increases , the chance that two clones within a host share the same haplotype increases . Simulations indicate that mean MOI would be unlikely to be substantially underestimated with either marker , unless the mean MOI was greater than 6 ( Amanda Ross , manuscript in preparation ) . The diversity observed in P . vivax compares well to the genetic diversity of two P . falciparum markers , msp1 and msp2 , previously determined in the same cohort [22] . MS16 was highly diverse in our cohort , and has been shown to be almost as diverse in countries of lower P . vivax endemicity such as Peru [12] and Vietnam [39] but was lower in Sri Lanka and Ethiopia [11] . In more than half of the samples , the number of clones detected was discordant for the two markers . Three factors contribute to such discrepancies , namely ( i ) differences in the discrimination power of the two markers , ( ii ) imperfect detection of clones in samples with low parasite densities and ( iii ) mutation of one of the markers occurring within a host ( Text S1 ) . Given the high diversity of both markers , we expect that limited discrimination power only accounts for a small fraction of observed discrepancies . More likely , low parasite densities around the detection limit will cause imperfect detection of clones . In a previously published analysis of clone detectability in the same set of samples , we have determined the contribution of an additional blood sample collected 24 hours later from the same children . This analysis showed that detection of genotypes by PCR is equally imperfect for both species , P . falciparum and P . vivax . Overall , 17 to 31% of all clones were missed on a single day , and detection of clones was imperfect especially in samples harboring a high number of concurrent clones [31] . In addition , in serial dilutions of parasite DNA from field samples we now show that at very low concentrations , MOI and allelic composition differed between replicates . In particular minority clones were lost . It is therefore very likely that such stochastic amplification of genotypes also occurred in our samples . Due to its antigenicity and surface exposed location , MSP1 is considered a candidate for a malaria vaccine ( reviewed in [13] ) . We identified 57 MSP1F3 alleles with the 3 most abundant alleles adding up to a frequency of 53% . The predominant alleles maintained stable allelic frequencies throughout the study . In a similar number of Papua New Guinean children , 27 haplotypes were observed for another potential vaccine candidate , the Duffy Binding Protein II ( DBPII , XM_001615397 ) . The 3 most frequent DBPII alleles were present in 57% of infections [40] . These results indicate high diversity of P . vivax antigens in PNG . With respect to vaccine development based on PvMSP1 , the allelic frequencies generated in our genotyping study provide useful information on genetic diversity of this antigen . In summary , this study provides one of the first large data sets of P . vivax genotypes from a highly endemic area . Our high resolution typing technique accurately determined allelic frequencies and clone multiplicity . This adds to the knowledge about P . vivax epidemiology and may serve as reference data for high endemicity P . vivax populations . The molecular parameters established could be utilized as one of several measures for effective monitoring of intervention and control of P . vivax , for surveillance and to parameterize mathematical models of transmission dynamics [40] . | The parasite Plasmodium vivax is the second most frequent cause of malaria in humans . In the Maprik area in lowland Papua New Guinea , P . vivax and P . falciparum are sympatric each with a prevalence of around 50% . Longitudinal samples from 268 children aged 1 to 4 . 5 years over 16 months were collected . The 1162 blood samples positive for P . vivax were genotyped for two size-polymorphic molecular markers . A very high parasite diversity was observed . The number of co-infecting parasite clones per carrier ( multiplicity ) was nearly twice as high for P . vivax as for P . falciparum despite the similar prevalences of the species . The P . vivax multiplicity increased with age , with the strongest increase in young children below 1 . 5 . This is likely to be a consequence of fast acquisition of immunity against P . vivax malaria and also of relapses , the release of long-lasting , silent liver stages to the blood stream . This is the first dataset from a highly endemic setting that presents data on a large number of individual P . vivax clones genotyped with highly diverse markers . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"biology",
"microbiology",
"population",
"biology",
"parasitology"
] | 2011 | Multiplicity and Diversity of Plasmodium vivax Infections in a Highly Endemic Region in Papua New Guinea |
Theoretical models of infection spread on networks predict that targeting vaccination at individuals with a very large number of contacts ( superspreaders ) can reduce infection incidence by a significant margin . These models generally assume that superspreaders will always agree to be vaccinated . Hence , they cannot capture unintended consequences such as policy resistance , where the behavioral response induced by a new vaccine policy tends to reduce the expected benefits of the policy . Here , we couple a model of influenza transmission on an empirically-based contact network with a psychologically structured model of influenza vaccinating behavior , where individual vaccinating decisions depend on social learning and past experiences of perceived infections , vaccine complications and vaccine failures . We find that policy resistance almost completely undermines the effectiveness of superspreader strategies: the most commonly explored approaches that target a randomly chosen neighbor of an individual , or that preferentially choose neighbors with many contacts , provide at best a relative improvement over their non-targeted counterpart as compared to when behavioral feedbacks are ignored . Increased vaccine coverage in super spreaders is offset by decreased coverage in non-superspreaders , and superspreaders also have a higher rate of perceived vaccine failures on account of being infected more often . Including incentives for vaccination provides modest improvements in outcomes . We conclude that the design of influenza vaccine strategies involving widespread incentive use and/or targeting of superspreaders should account for policy resistance , and mitigate it whenever possible .
Seasonal influenza imposes a significant health burden: in the United States alone there are an estimated 25–50 million cases per year , with 30 , 000 deaths and numerous hospitalizations , especially among the elderly and individuals with severe medical conditions [1] , [2] . Vaccination generally commences in September prior to the influenza season and is non-mandatory for the general public [1]–[4] . The question of whether to focus efforts on increasing vaccine coverage in children ( who spread a disproportionate amount of infection ) or the elderly ( who suffer the greatest health burden from infection ) has received significant attention in the transmission modelling literature [5]–[7] . Much of this work indicates that immunizing children might be a more effective way to reduce overall disease burden in the population . However , vaccine coverage has not significantly expanded in children . This leaves room for considering alternative strategies . Many infectious diseases exhibit a highly heterogeneous form of transmission known as “superspreading” , wherein a minority of individuals are responsible for the majority of secondary infections [8]–[12] , and it is possible that influenza also exhibits this property [13] . In a contact network , a superspreader can be represented as an individual with a very large number of contacts . Network-based infectious disease transmission models show how targeting superspreaders can be a highly effective ( and efficient ) form of infection control [14]–[20] . This suggests there may be value in exploring the possibility of immunizing influenza superspreaders , hence the focus of our analysis in this paper . Transmission models generally treat vaccine coverage as a fixed control parameter [21] , requiring the implicit assumption that desired vaccine coverage can always be achieved . However , public health authorities do not decide influenza vaccine coverage because they do not control individual vaccinating decisions . Instead , they control decisions such as where to set up immunization clinics , how to disseminate information , and whether to offer incentives to get vaccinated . Using a theoretical model to address factors that public health actually controls requires incorporating individual vaccinating behavior into the model . However , models of superspreader vaccination strategies usually assume that targeted individuals will always agree to be vaccinated to an arbitrarily specified level of vaccine uptake [14]–[16] , [18] . Incorporating behavior into transmission models is increasingly important in an age of vaccine exemption , especially for influenza vaccine , for which coverage is typically suboptimal [22] , [23] . Combining incentive use with targeting of influenza superspreaders could potentially be very effective , but behavioral feedbacks need to be considered in program design . Previous research has integrated behavioral modelling with transmission modelling to explore aspects of vaccinating behavior for various vaccine-preventable infections ( see Refs . [24]–[26] for reviews ) . Earlier approaches used compartmental models , but more recently , researchers are also using simulation models where transmission of infection and/or information occurs through social contact networks [27]–[31] . Transmission patterns can change significantly–even qualitatively–when transmission occurs through a network rather than in a homogeneously mixing population , and vaccinating behavior can change accordingly . The proliferation of recent papers considering vaccinating behavior on social contact networks prevents our providing a comprehensive survey here . However , to our knowledge , these previous models have not analyzed how behavioral feedbacks influence the effectiveness of vaccination strategies that target superspreaders , nor have most of them explored how incentive use influences vaccination behavior . Here , we analyze an agent-based simulation model that couples seasonal influenza transmission on an empirically-based contact network with a psychologically realistic model of individual vaccinating decisions . We explore the effectiveness of incentive programs and targeted superspreader vaccination strategies . Our objectives are to understand: 1 ) whether superspreader vaccination strategies remain effective when behavior is accounted for; 2 ) whether economic incentives improve the effectiveness of such strategies; and 3 ) how perceived vaccine efficacy and the resulting vaccinating decisions are determined by interactions between network structure , transmission heterogeneity , and vaccine-disease dynamics .
For our baseline analysis we generated ten contact networks of 10 , 000 nodes each , by sampling subnetworks from a large contact network derived from empirical contact patterns in Portland , Oregon [32]–[34] . We ensured that the resulting node degree distribution and clustering coefficient matched that of the full empirical network ( see Text S1 ) . For influenza , susceptible individuals are recruited primarily through immunity waning , hence we assumed that the networks remained static , with no immigration or emigration . In our sensitivity analysis we explored hypothetical networks with exponential and Poisson node degree distributions . The contact network contains individuals representing the full spectrum of neighborhood sizes and does not impose a dichotomy between superspreaders and others . However , to assist with interpreting the output of our simulations , we defined a superspreader as an individual who infected more than the percentile from a Poisson distribution with mean , where is the basic reproduction number for the “null” deterministic model's approximation . Approximately of individuals in the empirically-based network met this definition of superspreaders ( see Text S1 for details ) [11] . We assumed a Susceptible - Infected - Recovered - Vaccinated - Susceptible ( SIRVS ) natural history . An infectious individual transmits influenza to a susceptible contact with probability per day , where varies seasonally . An infectious individual moves to the recovered state after a number of days sampled from a Poisson distribution with mean days . A recovered individual becomes susceptible with probability per season ( natural waning immunity ) . A vaccinated individual becomes susceptible with probability per season ( vaccine waning immunity ) . Vaccination has no impact on individuals who are in the naturally immune state and the vaccine efficacy is . Symptomatic infection occurs with probability . In our sensitivity analysis , we also allowed for heterogeneity with respect to the infectious period and the infectivity . This creates additional sources of heterogeneity that may cause some individuals to become superspreaders . More details appear in Table 1 , Table S1 and Text S1 . We structured the vaccination decision-making submodel according to known determinants of influenza vaccine acceptance . Empirical studies have identified that perceived vaccine effectiveness , previous acceptance of vaccine , past experiences with infection and vaccine complications , social influence , and perceived susceptibility are correlates of vaccine acceptance [35] , [36] . Although the data in these studies are not detailed enough to favor particular functional forms governing these effects , it is nonetheless possible to construct functional forms that are qualitatively consistent with them , which is also the approach adopted in some other models [37] . The payoffs for strategy choices are given by ( 1 ) ( 2 ) where is the payoff to vaccinate for season , is the payoff not to vaccinate , is the baseline payoff ( a state of perfect health ) , is the perceived vaccine efficacy , is the cost of vaccinating and is the perceived infection cost [37] . The perceived infection cost incorporates perceived susceptibility and past infection experiences . Perceived susceptibility is expressed through past influenza incidence in the population . Past infection experience is expressed through the time since the individual's last perceived infection , . Hence ( 3 ) where is the influenza incidence in season , is the penalty for being infected , controls the relative importance of personal history versus population history , and is the memory decay rate . Severe outcomes are implicitly accounted for in , which represents the combined foreseen risk of infection and any resulting complications . Thus , this equation captures how individuals use past experiences to guide future vaccinating decisions . The perceived vaccine efficacy for an individual in season generally differs from actual efficacy and is given by ( 4 ) where controls how quickly perceived vaccine efficacy drops upon a perceived vaccine failure , is the maximum perceived vaccine efficacy , and is a decay factor which causes memory of a previously ineffective vaccination to fade at a slower rate than a successful vaccination , since they have less information with which to update their impression [37] . The asymmetry between an event where individuals vaccinate and become infected versus an event where they did not vaccinate arises because of the distinction between “evidence of absence” and “absence of evidence” . Only symptomatically infected individuals update their values of and . The cost of vaccination also incorporates past experience: ( 5 ) where represents time and economic costs , is the time since the last perceived vaccine complication , and is the perceived cost of a vaccine complication . The probability an individual perceived a complication upon vaccinating is . We incorporate social influence through a learning process . Empirical studies of determinants of vaccine acceptance suggest that individuals form opinions through communicating with their peers and sharing personal experiences [35] , [36] . Hence , before each vaccination season , an individual engages in a social learning process with probability , sampling another individual at random and replacing their , and with the average of their pre-existing value and that of the sampled individual , weighted by and respectively . This captures both the tendency to personalize the experiences of others , as well as habit , since strategies change more slowly as . This mechanism of social learning is similar to that used in a previous behavior-incidence model that was validated against empirical data [38] . We also account for the impact of non-influenzal influenza-like-illness ( niILI ) on decision making , since niILI can be mistaken for true influenza and thus alter and . The probability an individual experiences niILI each day is , where is sampled from a log-normal distribution parameterized from empirical data on niILI incidence . An individual mistakes niILI for true influenza with probability , in which case and are updated accordingly ( see Text S1 for details ) . Passive Vaccination ( PV ) is the baseline strategy corresponding to how most influenza vaccination programs are designed: vaccines are made available ( e . g . at drug stores , public health clinics , doctors' offices ) , opening times are widely disseminated , and individuals seek out vaccination on their own , without being individually recruited by public health . To capture this , we assume an individual decides to get vaccinated for the current season with probability , which is a sigmoidal function of such that , , and ( see Text S1 ) . We assume individuals can be vaccinated only between September ( ) and December [4] . Those who choose to vaccinate have their times of vaccination distributed throughout this period according to a process described in Text S1 . If an individual perceives having been infected by influenza before it is their time to become vaccinated , they do not seek vaccination . In addition to making PV available , public health may also implement one of four pro-active strategies: 1 ) random vaccination ( RV ) which targets a randomly chosen individual; 2 ) nearest neighbor vaccination ( NN ) , which targets a randomly chosen individual and one of their neighbors ( i . e . contacts ) 3 ) chain vaccination ( CV ) , which either targets a randomly chosen individual or a neighbor of an individual targeted the previous day; and 4 ) improved nearest neighbor vaccination ( INN ) which targets a randomly chosen individual and one of their most popular neighbors . Under INN , “popular” means having the highest degree , and we assume imperfect knowledge of a neighbor's neighborhood size . We refer to NN , CV and INN as superspreader strategies , as their objective is to target individuals with a large number of contacts [15] , [16] , [18] , [39] . The number of individuals targeted by public health each day is held constant at for all strategies . In each case , if the targeted individual did not already decide to vaccinate under PV , they reconsider: they undergo the social learning process again and agree to be recruited for vaccination with probability . In our sensitivity analysis we also ran simulations where each targeted person was automatically recruited , corresponding to a situation where behavior is neglected ( NB ) . More details on the pro-active strategies appear in Text S1 . We allowed for the use of economic vaccination incentives under the pro-active strategies . Each time an individual is targeted they receive an incentive of value if they get vaccinated during the current season . An individual can receive multiple incentives . Hence superspreaders should receive more incentives , since they are likely to be targeted multiple times under NN , CV and INN . We considered ( baseline ) , , and . With incentives , the probability of vaccinating becomes a function of ( where is the number of times they have been targeted ) instead of . In order to express , and in the same payoff currency , , , and were expressed in quality-adjusted life years ( QALYs ) ( see Text S1 ) . For each of the ten networks , the transmission probability and amplitude of seasonality were calibrated so that the average seasonal incidence of influenza in the absence of vaccination was [40]–[42] , and prevalence peaked between January and February . was based upon utility scores derived from patient surveys [43] . was based on published vaccine costs [41] , [44]–[46] . and were calibrated such that the average annual vaccine coverage was . For a efficacious vaccine , vaccine coverage of reduces seasonal influenza incidence by about ( Table 2 ) , in line with what is expected for an imperfect vaccine covering one-third of the population . Examples of calibrated time series of annual coverage and weekly incidence appear in Figure S1 . For each network we generated 400 realizations of 150 years each , discarding the first 125 years to avoid transient effects .
In the absence of incentives , the improved nearest neighbor strategy ( INN ) is the most effective in reducing influenza incidence , followed by chain vaccination ( CV ) , nearest neighbor ( NN ) , random vaccination ( RV ) , and the baseline strategy of passive vaccination alone ( PV ) ( Table 2 ) . This relative ordering is to be expected , since previous research shows the advantages of targeting individuals with many contacts [14] , [16]–[20] . However , feedbacks due to the dependence of vaccinating decisions on infection history generates some surprises [26] , [38] , [47] , [48] . In this system , the feedbacks manifest as policy resistance [48] , where the response of the population to an intervention ( in this case , pro-active strategies and incentives ) tends to reduce the effectiveness of the intervention . In our model simulations , policy resistance arises because increased vaccine coverage in one season reduces incidence due to both direct and indirect ( herd ) protection , which in turn disincentives vaccination in future seasons , since decisions are based partly on infection history and perceived vaccine failure/complications . An additional source of policy resistance in this system is the tendency for pro-active strategies to waste recruitments on individuals who already decided to get vaccinated under passive vaccination , or who have already been infected ( this is a problem especially among superspreaders , who are both targeted more often and tend to get infected earlier in the season ) ( Table S2 ) . On average , only of the population was vaccinated through being contacted through a nearest neighbor under INN; whereas under NN and CV the percentage was . Policy resistance almost completely undermines the benefits of using pro-active strategies: passive vaccination ( PV ) reduces seasonal influenza incidence from to , but implementing improved nearest neighbour vaccination ( INN ) on top of that provides only slight additional reductions , down to . The other pro-active strategies ( NN , CV , RV ) are even less effective , reducing incidence to or ( Table 2 ) . Moreover , among pro-active strategies , superspreader strategies are only marginally more effective than random vaccination ( RV ) ( Table 2 ) . As expected , the superspreader strategies improve vaccine coverage among superspreaders . However , this is offset by lower coverage among non-superspreaders . As a result , the average vaccine coverage under superspreader strategies is the same as under random vaccination ( Figure 1 , Table 2 ) . The impact of policy resistance is made clear by considering the case where vaccinating behavior is neglected ( by assuming that targeted individuals are automatically recruited for vaccination ) . Neglecting behavior ( NB ) significantly overestimates both effectiveness and vaccine coverage for the pro-active strategies , both in superspreaders and non-superspreaders ( Table 2 ) . Hence , without accounting for behavior , we might have concluded that superspreader vaccination strategies can be significantly more effective than their non-targeted counterpart , but if we take behavior into account , their impact is greatly diminished . We note that the slightly higher effectiveness of the improved nearest neighbor strategy also arises because by preferentially immunizing those individuals with a large number of contacts , susceptible individuals tend to be clustered together on the network , reducing the opportunities for the susceptible-infected contacts necessary for transmission ( Table S3 ) . Individuals with more neighbors were more likely to be infected ( Figure 1a ) –d ) ) . This resulted in a higher probability of them getting vaccinated ( Figure 1e ) –h ) ) , but it also caused them to perceive the vaccine to be less effective ( Figure 1i ) –l ) ) , on account of higher infection rates causing higher rates of perceived vaccine failure . The effect of adding vaccinating incentives is likewise blunted by policy resistance ( Table 2 ) . Any increase in vaccine coverage due to use of incentives reduces incidence , which in turn disincentivizes future vaccine uptake ( especially among superspreaders under passive vaccination , Table S4 ) . Also , incentives often reach individuals who are already prone to get vaccinated ( Table S2 ) . However , modest improvements in program effectiveness due to the use of incentives are still possible . For example , adding a incentive to the improved nearest neighbor strategy reduces influenza incidence from to ( Table 2 ) . We estimated the net per capita costs ( total vaccine costs plus total infection treatment costs per member of the population ) for each strategy . The least expensive strategy was the improved nearest neighbor strategy ( INN ) without incentives , at a cost of per capita . In contrast , passive vaccination on its own ( PV ) costs per individual because infection costs are higher under PV than INN . These results assume the administrative costs of vaccination are the same for passive versus pro-active strategies , although in reality the marginal cost per vaccinated person may be higher under pro-active strategies , especially if they involve targeting superspreaders . We ran additional simulations where there was an additional marginal cost for recruiting contacts ( as under NN , CV and INN ) , finding that the marginal cost for recruiting contacts under INN would have to be at least per recruited individual before INN becomes more costly than PV . Using vaccinating incentives increased the total cost of all strategies , but not always significantly . Further details appear in Text S1 and Table S5 . On average , most individuals received few incentives and only a few individuals received many incentives ( Figure S2 ) . Superspreaders tended to receive more incentives by virtue of having more contacts , but the benefit of this was partly mitigated by the fact that they are likely to be infected and/or seek vaccination earlier in the season than individuals with few contacts , and hence have less time to accumulate incentives . Our baseline assumption was that superspreading is driven only by heterogeneity in neighborhood size ( node degree ) . Incorporating heterogeneity into the infectious period , transmission rate , or both did not significantly impact the results ( the superspreader strategies become slightly less effective in the absence of incentives ) . We suspect these forms of heterogeneity did not make a difference because an individual's infectious period and infectiousness were not correlated to their node degree , meaning that superspreader strategies on average do not target individuals with higher infectiousness or longer infectious period . This causes differences in effectiveness between the various strategies to be averaged out . Were correlations to exist between node degree on the one hand , and infectious period or transmission rate on the other hand , then we speculate the results could change qualitatively , either in the direction of greater effectiveness of superspreader strategies , or lesser effectiveness , depending on whether the correlation was positive or negative , respectively . For simulations on the hypothetical Poisson or exponential networks instead of the empirically-based network , we found that the pro-active strategies continued to provide very small reductions in incidence compared to passive vaccination alone ( Table S6 and Table S7 ) . Results were also qualitatively unchanged when incentives were distributed only to the recruited neighbor of individuals targeted under NN and INN ( Tables S2 , S5 ) ; however , the cost of the policies was reduced due to fewer incentives being distributed ( Table S5 ) .
Previous models of superspreader vaccination strategies have shown how targeting individuals with a very large number of contacts can be a very effective way to control infection [14]–[20] , [39] . These models have generally assumed that targeted individuals will always agree to be vaccinated . For voluntary influenza vaccination , this assumption may introduce inaccuracies , since individual choice is a major determinant of influenza vaccine uptake [22] , [35] , [36] . Here , we developed a psychologically structured model of influenza vaccinating behavior and coupled it to a model of seasonal influenza transmission through an empirically-based contact network . Our assumptions about vaccinating behavior were based on empirical studies exploring determinants of vaccine uptake [35] , [36] . We found that three of the most commonly investigated superspreader vaccination strategies ( nearest neighbor , chain vaccination , and improved nearest neighbor ) provided little or no improvements over their non-targeted counterpart ( random vaccination ) . This surprisingly strong policy resistance is driven by multiple mechanisms: individuals are less likely to get vaccinated if their most recent influenza infection was a long time ago , if they perceive low susceptibility to infection ( which can emerge from herd immunity generated by vaccination ) , or if they perceive recent vaccine complications or low vaccine efficacy . The presence of non-influenzal influenza-like illness ( niILI ) reinforces this because it can create the perception of vaccine failure . Moreover , superspreader strategies tend to reach individuals who are already more prone to get vaccinated without the need for active recruitment on account of their history of more frequent infections . Contact-based recruiting was also stymied by the fact that neighbors tended to share similar experiences and information , which led to neighbors more often than not sharing the same opinion regarding vaccination [27] , [37] . Providing vaccination incentives boosted effectiveness somewhat , including for superspreader strategies , although these gains were again partly mitigated by policy resistance . The most effective overall strategy was the improved nearest neighbor strategy ( INN ) with incentive: this strategy reduced the average annual influenza incidence to ( compared to under passive vaccination ) . The improved nearest neighbor strategy may also be cost-saving compared to passive vaccination . A few empirical studies have found that incentives can increase vaccine uptake [49] , [50] . However , these studies focused on incentivization for small groups of elderly individuals over the course of a single season , not widespread incentivization for individuals with a large number of contacts . Individuals with more contacts were more likely to be infected , which agrees with results from past models [51] , [52] . This resulted in an individual's perceived vaccine efficacy declining with their number of contacts . This is a potential barrier in superspreader vaccination compliance . More generally , how perceived vaccine efficacy evolves over time and in response to disease dynamics and the collective effects of individual vaccinating decisions merits further study . As with any model , our model made simplifying assumptions . For example , we assumed a targeted individual who is asked to recommend a contact for vaccination would always comply . We could address this limitation by introducing a parameter for compliance failure . This would result in a similar strategy to the chain vaccination strategy , where occasionally the recruitment process jumps to another individual rather than continuing along the chain of contacts . We also neglected age structure in the model , which did not allow us to address issues such as age-related heterogeneity in infection severity , and correlations between infection severity and neighborhood size . Incorporating greater heterogeneity into the model by stratifying individuals with respect to age would increase model realism . It would also allow us to address other objectives such as how incentives can be designed to boost vaccine coverage in children . However , this would also increase model complexity , and given that current model already required dealing with the extra complexity of incorporating behavior , we opted for the incremental approach of first developing a model without age structure . Another aspect of model development that requires greater attention is the functional forms used to capture psychological effects such as the role of past experiences and social influences , since typically more than one functional form is qualitatively consistent with existing data on determinants of vaccine uptake . These areas suggest potential for further work at the interface of theoretical modeling and empirical surveys . In particular , surveys of determinants of vaccination behavior can be designed to better meet the needs of models that couple disease dynamic models to vaccinating behavior models , for instance by helping to determine which functional forms best capture psychological effects . This will require collecting new psychological data from study populations . In other cases , these models suggest predictions which can be tested . For example , our model predicted that superspreaders would perceive a slightly lower vaccine efficacy than non-superspreaders , and it would be interesting to see whether this effect holds true for any populations . Thoroughly validated “behavior-incidence” models of influenza vaccinating behavior will help public health authorities to optimize influenza vaccine programs . However , as we have found here , vaccination strategies that target superspreaders and/or provide vaccination incentives must be carefully designed to mitigate the potentially strong effects of behavioral feedbacks and policy resistance . | Superspreaders are the small number of individuals responsible for the majority of infections . Theoretical models have shown how vaccinating superspreaders can be a highly efficient way to control disease . However , these models neglect behavior by assuming that superspreaders will always agree to be vaccinated . This is a problematic assumption for influenza vaccination , which is voluntary in most populations , and for which vaccine coverage is often suboptimal . We developed a model of seasonal influenza transmission on a network of individuals who make decisions about whether or not to get vaccinated based on known determinants of vaccine uptake , such as personal infection history , perceived vaccine risks , and social influences . We found that , because of feedbacks between disease spread and individual vaccinating behavior , attempts to boost vaccine coverage in superspreaders through the use of incentives or recruiting by social contacts are almost completely undermined by such feedbacks . For example , higher vaccine uptake in superspreaders reduces influenza incidence , which in the next season reduces the perceived need for vaccination among non-superspreaders , who then do not become vaccinated as much . Our results suggest that the design of potential strategies to reach influenza superspreaders should account for behavioral feedbacks , since they may blunt policy effectiveness . | [
"Abstract",
"Introduction",
"Model",
"Results",
"Discussion"
] | [
"medicine",
"public",
"health",
"and",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"population",
"dynamics",
"applied",
"mathematics",
"spatial",
"epidemiology",
"mathematics",
"population",
"modeling",
"population",
"biology",
"infectious",
"disease",
"control... | 2013 | Policy Resistance Undermines Superspreader Vaccination Strategies for Influenza |
Duplications play a significant role in both extremes of the phenotypic spectrum of newly arising mutations: they can have severe deleterious effects ( e . g . duplications underlie a variety of diseases ) but can also be highly advantageous . The phenotypic potential of newly arisen duplications has stimulated wide interest in both the mutational and selective processes shaping these variants in the genome . Here we take advantage of the Drosophila simulans–Drosophila melanogaster genetic system to further our understanding of both processes . Regarding mutational processes , the study of two closely related species allows investigation of the potential existence of shared duplication hotspots , and the similarities and differences between the two genomes can be used to dissect its underlying causes . Regarding selection , the difference in the effective population size between the two species can be leveraged to ask questions about the strength of selection acting on different classes of duplications . In this study , we conducted a survey of duplication polymorphisms in 14 different lines of D . simulans using tiling microarrays and combined it with an analogous survey for the D . melanogaster genome . By integrating the two datasets , we identified duplication hotspots conserved between the two species . However , unlike the duplication hotspots identified in mammalian genomes , Drosophila duplication hotspots are not associated with sequences of high sequence identity capable of mediating non-allelic homologous recombination . Instead , Drosophila duplication hotspots are associated with late-replicating regions of the genome , suggesting a link between DNA replication and duplication rates . We also found evidence supporting a higher effectiveness of selection on duplications in D . simulans than in D . melanogaster . This is also true for duplications segregating at high frequency , where we find evidence in D . simulans that a sizeable fraction of these mutations is being driven to fixation by positive selection .
In 2004 , two pioneering studies showing that copy number variants ( CNVs ) are abundant in healthy human individuals [1] , [2] accelerated research on this class of variation . The focus on these variants was well motivated because duplications and deletions of DNA regions have long been known to underlie a variety of genomic disorders [3] , [4] . The discovery of the abundance of CNVs in otherwise healthy individuals made them good candidates to underlie common and rare diseases as well as other physiological traits . In just a few years , CNVs were implicated in a variety of diseases such as autism [5] , schizophrenia [6] , Crohn's disease [7] , psoriasis [8] and other traits such as body weight [9] and starch consumption [10] . Duplications and deletions also have a long history of being implicated in adaptation and of being a major source of genetic innovation [11]–[14] . In domesticated animals , for example , they are responsible for white coat color in horses ( duplication within an intron leading to cis-regulatory changes [15] ) , reduced comb and wattle size in chickens ( duplication within an intron leading to expression changes [16] ) and short-legged dogs ( new retrogene [17] ) . Although much has been learned about CNVs , recent research raises more questions than it answers . Two independent avenues of research focus on studying the roles played by mutation and selection on copy number variation . Understanding the mutational processes underlying the formation of CNVs is important from both a medical and an evolutionary perspective . Duplications and deletions can result from the imperfect repair of DNA double strand breaks generated by both exogenous ( e . g . ionizing radiation ) and endogenous ( e . g . reactive oxygen species ) agents as a consequence of the normal cellular metabolism [18] , [19] . DNA replication errors can also generate CNVs , with or without the formation of DNA double strand breaks [4] , [19] . Replication-based repair processes have been proposed to explain complex CNVs ( i . e . CNVs with multiple breakpoints ) [20]–[22] but evidence suggests they underlie the formation of simple CNVs as well [23]–[25] . Several lines of evidence suggest that CNV mutation rates vary throughout the genome [26] , [27] and CNV hotspots have been identified in the human [27]–[29] , chimpanzee [28] , [30] , mouse [31] and fly [32]–[34] genomes . Mammalian CNV hotspots are significantly enriched with segmental duplications , which have been proposed to promote the occurrence of CNVs by facilitating non-allelic homologous recombination ( NAHR ) [3] , [4] . Following this observation , Sharp and colleagues specifically targeted genomic regions associated with segmental duplications in the human genome and were able to identify CNVs associated with previously unidentified genomic disorders [35] . But not all mammalian hotspots are associated with segmental duplications [28] , [30] and Drosophila hotspots are likely not associated with them at all [33] . As such , a priority of the field is to identify the genomic feature ( s ) , other than segmental duplications , that are associated with regions with increased numbers of CNVs . Understanding the evolutionary forces shaping the evolution of CNVs is also important from a medical and evolutionary perspective . Despite their pervasiveness , analyses of the genomic distribution of CNVs among different functional regions clearly indicate that a large fraction is under purifying selection . Population genetic models that address both demographic and selection processes have been used to estimate the strength of selection acting on different classes of CNVs . In both flies [36] and humans [37] coding CNVs are under the strongest purifying selection followed by intronic CNVs and finally intergenic CNVs . Evidence for positive selection has been less clear . There are examples of CNVs under positive selection in humans , such as the copy number variation of the amylase [10] and CCL3L1 [38] genes , and in flies ( e . g . duplication of the Cyp6G1 locus ) [36] , [39] . However , on a genome-wide scale , the over-representation of certain classes of genes in CNVs , namely “environmental” genes , is best explained by reduced purifying selection acting on these variants than by positive selection [40] . Although genome-scale studies of CNVs have only recently become technically feasible [41] , the study of gene duplication can be traced back to as early as 1911 [12] , [42] . An important problem is to determine the relative roles of positive selection and genetic drift in the fixation of new gene duplicates [43] . Most population genetic models assume that gene duplicates are fixed by genetic drift and that their subsequent fate in genomes ( being retained or lost ) is determined by ensuing mutations in one or both copies [43] , [44] . An alternative hypothesis is that gene duplications are fixed by positive selection . Assessing the roles of drift and selection requires the study of young duplications that still bear the hallmarks of the evolutionary process responsible for their fixation [11] , [14] , [43] . The aim of this work is to investigate the roles played by mutation and selection on duplication polymorphisms . We take advantage of the genetic model system composed by the sibling species D . melanogaster and D . simulans , which have been used extensively to conduct population and evolutionary genetic studies [45] . While they share a recent common ancestor and are morphologically very similar , at an average of 4% DNA sequence divergence , they are sufficiently diverged to provide many evolutionary insights [46] . Hence , the structural differences ( and similarities ) of their genomes can be leveraged to dissect the genomic features responsible for the variation in CNV density along the genome and elucidate the existence of duplication hotspots . For example , while the D . melanogaster genome is rich in inversion polymorphisms these are rare in D . simulans [47] . Similarly , the fraction of repetitive sequence is considerably larger in the D . melanogaster genome [46] , [48] and transposable elements are differentially distributed in the two species [46] . Another useful distinction between the two species , and one that can be used to investigate the role of selection , is the difference in their effective population sizes . D . simulans has a ten-fold larger effective population size than D . melanogaster , which is predicted to translate into a greater effectiveness of selection in D . simulans [49] , [50] . Thus this species is expected to be more efficient both at purging deleterious mutations and fixing those that are beneficial [51] . The differences in population size and genome structure between D . melanogaster and D . simulans provide us with a powerful genetic model in which to study how mutation and selection processes shape patterns of copy number variation . Duplication and deletion polymorphisms have previously been surveyed in 15 lines of D . melanogaster using tiling microarrays [36] . Here , we use the same approach to identify and characterize duplication polymorphisms in 14 lines of D . simulans . By integrating this new dataset of duplications in D . simulans with the previous dataset of duplications in D . melanogaster , we identified duplication hotspots shared between the two species . Significantly , we found that these hotspots are not associated with segmental duplications or transposable elements but are instead associated with regions of the genome that are late-replicating . We also show a higher effectiveness of selection acting on D . simulans duplications than on D . melanogaster duplications , and suggest an important role for positive selection in driving a sizeable fraction of D . simulans duplications to fixation .
We identified polymorphic duplications in the D . simulans genome using a similar strategy to the one previously used to identify CNVs in the D . melanogaster genome [36] . Briefly , we hybridized DNA from 14 D . simulans lines ( see Materials and Methods ) onto DNA tiling arrays ( three replicates per line ) . Because the tiling arrays were designed based on the D . melanogaster genome , of the ∼3 million probes available on the arrays , only the ∼900 , 000 that had a perfect ( and unique ) match to the D . simulans genome were used in this study ( see Materials and Methods ) . The hybridization intensities were then decoded into probabilities of copy number gains and losses ( and absence of changes in copy number ) using a Hidden Markov Model . A given region was called as a putative duplication if at least two consecutive probes gave hybridization signals decoded as having at least a 40% posterior probability of being duplicated . We identified 830 duplications segregating in the 14 D . simulans lines ( Table S1 ) . The duplications are on average 1 . 8 kb in size ( median 424 bp ) , with the smallest being 28 bp and the largest ∼127 kb . We evaluated the quality of the CNV calls by PCR . Our PCR assay assumed duplications occur in tandem , such that a pair of divergent primers placed within the region predicted to be duplicated would lead to the amplification of a band only in the presence of a tandem duplication . Sequencing of that band would provide the exact duplication breakpoints . Out of 24 putative duplications , 18 produced the expected band , yielding a confirmation rate of 75% . The remaining 6 duplication candidates yielded no band , which could suggest: 1 ) the duplication is a false positive; 2 ) the duplication is not in tandem; or 3 ) the PCR reaction failed . To exclude the third possibility , we designed pairs of convergent primers ( outside the putative duplication ) for the 6 unconfirmed duplications , such that lines predicted to have a duplication would produce larger bands than the lines without it . This strategy confirmed one of the 6 remaining duplications , increasing the confirmation rate to 79% . A survey of recently fixed gene duplications in Drosophila [52] , found that 82% of duplications in D . melanogaster and 78% in D . yakuba occur in tandem , with the remaining being dispersed in the genome . Our confirmation rate is , therefore , in good agreement with the expectation for the proportion of tandem duplications , and further supports the view that the majority ( ∼80% ) of newly generated duplications occur in tandem , with the remaining being dispersed throughout the genome ( including at a certain distance from each other within the same chromosome ) . It is important to note that since our confirmation strategy involved designing primers within the predicted duplication , we only attempted to confirm duplications larger than 300 bp ( see Materials and Methods ) . We also detected 379 deletions . However , out of 32 deletions assayed by PCR ( primers located outside the putative deletion ) , only 13 were confirmed , yielding a false positive rate of almost 60% . The Sanger sequencing of the false positives revealed the presence of small indels and SNPs overlapping with the probes that were called as being deleted . Because the probes in the tiling arrays are only 25 bp , any variant that occurs within them knocks out the hybridization signal in a manner similar to a deletion . This same problem was encountered when characterizing deletions in D . melanogaster ( false positive rate of 47% [36] ) . Because most of our deletion calls are likely to be false positives , we did not include these variants in our study . Distinguishing novel duplications segregating in the lines examined ( derived duplications ) from ancestral duplications is important . The latter correspond to situations where the duplication is fixed in D . simulans but because it is present in the reference genome sequence as a single copy ( due to deletion or genome mis-assembly ) it appears in our survey as a duplication segregating at high frequency . Distinguishing derived from ancestral duplications can be accomplished by determining the duplication status of these regions in the D . melanogaster genome . Derived duplications would appear as single copy regions in both the D . melanogaster and D . simulans reference genomes , whereas ancestral duplications would appear as duplications in D . melanogaster . We found only one event for which there was evidence of a duplicate in the D . melanogaster genome reference sequence . However , we found three duplications in D . simulans that appear in the D . melanogaster genome reference sequence as single-copy , but are detected as polymorphic duplications in all ( or most ) of the 15 lines previously used to identify CNVs in the D . melanogaster genome [36] . This result is not entirely surprising given the nature of the ascertainment bias when inferences are made from arrays designed from a single reference sequence [36] , [53] . Two of these duplications are predicted to have identical breakpoints in the two species and are detected in all lines in both species . The third duplication , completely encompasses two genes involved in drug metabolism ( Ugt86Dj and Ugt86Dh ) , and was detected in all 14 D . simulans lines and in 13 out of the 15 D . melanogaster lines . We sequenced the breakpoints of this duplication in both species and they are identical , suggesting they derive from the same mutation . The most unusual aspect of these events is their apparent absence in the genome references , which should be unlikely if the duplication is ancestral to both D . melanogaster and D . simulans . The most likely explanation is that the duplicates became fixed before the split of D . melanogaster and D . simulans and were either collapsed during the genome assemblies or the sequenced genome strains contain deletions of one of the copies . For the third duplication mentioned , given that a cost of resistance can be associated with insecticide resistance [e . g . 54] it is perhaps not surprising that strains shielded from the selective pressure of insecticides may preferentially lose such mutations under laboratory cultivation . Although duplications are found throughout the whole genome , they are distinctly less frequent in functional elements: even though 41% of the D . simulans genome is annotated as coding sequence [55] , only 28% of duplications overlap with these regions . The majority of duplications are restricted to intergenic ( 50% ) and intronic ( 22% ) regions , implying that a large fraction of these mutations are deleterious and are quickly removed from the populations by purifying selection . Overall , duplications are kept at very low frequencies in the lines surveyed , with 83% of them being detected in only one of the 14 lines ( Table S1 ) . Figure 1 illustrates that the distribution of duplications among genomic contexts varies dramatically between those that are kept at very low frequencies , e . g . singletons ( 1 out of 14 lines ) , and those that are segregating at high frequencies ( found in at least 6 of the 14 lines ) . Counterintuitively , while only 25% of duplications segregating at very low frequencies overlap coding sequence ( i . e . partial and complete gene duplications ) , 70% of duplications segregating at high frequencies encompass coding sequence ( Fisher exact test , p = 0 . 0001 ) . If genetic drift was responsible for high frequency derived alleles , one would not predict such duplications to overlap coding sequence , because these mutations are less likely to be neutral [36] . This apparent contradiction can be resolved if we instead posit that positive selection plays an important role in driving these mutations to fixation . In support of this hypothesis , we found that while complete gene duplications represent 3 . 6% of duplications segregating in only one line , they represent 35% of duplications segregating in 6 or more lines ( Fisher exact test , p = 9 . 95×10−6 ) . Although there is also an increase in the proportion of partial gene duplications ( from 22% of all duplications segregating in only one line to 35% of duplications segregating in 6 or more lines ) , this increase is not statistically significant ( Fisher exact test , p = 0 . 2 ) . This means that of duplications overlapping exonic sequence , only complete gene duplications are over-represented among high-frequency variants . There is no Gene Ontology category over-represented in the set of genes present in high-frequency duplications ( p>0 . 01 ) . In a previous study , Emerson and colleagues used the same microarray platform and a similar strategy to detect duplication polymorphisms in 15 D . melanogaster lines [36] . In that study all ∼3 million probes present in the tiling arrays were used to make the duplication calls ( as opposed to the ∼900 , 000 available for this study ) which provided more power to detect them ( especially the smaller ones ) and better breakpoint resolution . As expected , more duplications were detected in D . melanogaster than in D . simulans ( 2016 vs . 830 ) , and the former were also , on average , shorter ( 1 . 2 kb vs . 1 . 8 kb ) although the difference is not statistically significant ( Wilcoxon rank sum test , p = 0 . 8 ) . The set of D . melanogaster duplications used here differs from the set originally published by the exclusion of those duplications detected by only one probe because these were not included in the set of D . simulans duplications . Figure S1 shows the genomic location of the duplications detected in both species . Although a higher proportion of the D . simulans genome is annotated as coding ( 41% vs . 33% in D . melanogaster ) [55] , we found that D . simulans has a significantly lower proportion of coding duplications than D . melanogaster ( 28% in D . simulans vs . 39% in D . melanogaster , Fisher exact test , p = 2×10−9 ) , suggesting purifying selection acts more strongly on D . simulans duplications . Figure 2 shows , for the two species , the proportion of the different classes of duplications partitioned by their frequency in the lines surveyed . While D . simulans has a smaller proportion of partial and complete gene duplications segregating in low and medium frequencies than D . melanogaster ( consistent with stronger purifying selection ) , the opposite pattern is observed for high frequency duplications . In this latter class , D . simulans has a significantly higher proportion of complete gene duplications than D . melanogaster ( 35% vs . 15% , respectively; Fisher exact test , p = 0 . 0001 ) . Because complete gene duplications are more likely to be advantageous than all other classes of duplications , this result can be interpreted as supporting a more pervasive role for positive selection in driving the fixation of duplications in D . simulans than in D . melanogaster . If duplications of complete genes often have only small positive effects on fitness , they will be detected and favored more readily in D . simulans than in D . melanogaster because of the former's larger effective population size ( see Discussion ) . There are two other notable differences in the patterns of duplication polymorphism found between D . simulans and D . melanogaster that could support the hypothesis that both purifying and positive selection are stronger in the D . simulans . First , there is a significantly higher proportion of low frequency duplications ( i . e . those present in only one of the lines ) segregating in D . simulans than in D . melanogaster ( 83% vs . 74% , respectively; Fisher's exact test , p = 1 . 7e−07 ) . Because purifying selection is expected to lead to an excess of rare variants , the higher proportion of duplications kept at low frequencies in D . simulans could suggest stronger purifying selection . Second , there is an excess of high frequency duplications segregating on the X chromosome when compared to the autosomes in D . simulans ( Fisher's exact test , p = 0 . 03 ) but not in D . melanogaster ( Fisher's exact test , p = 0 . 5 ) . Given the different biology and population genetics of the X chromosome , differences found between the X and the autosomes could be due to multiple ( and non mutually exclusive ) factors [56] . However , if one assumes that most beneficial mutations are recessive or partially recessive , then positive selection is expected to be more efficient on the X than on autosomes ( faster X evolution [56] ) , which would lead to a higher proportion of high-frequency duplications on the X than on autosomes . There is , however , an alternative explanation for both the overall higher proportion of low frequency duplications in D . simulans , and the excess of high frequency duplications in the X chromosome of this species . Demographic processes , such as population expansion , bottlenecks and population structure , can also generate these patterns of polymorphism [57] . The two species have different demographic histories , which could easily generate differences in genome-wide patterns of polymorphism between them . Demographic processes cannot , however , explain the differences between the two species in the proportion of coding vs . non-coding duplications for low and high frequency variants . This is because unlike selection , demography cannot discriminate between functional and non-functional regions of the genome , instead affecting equally the genome as a whole . Perry and colleagues compared global maps of copy number variation for the human and chimpanzee genomes [28] , [30] , finding a significant excess of overlap between CNVs of the two species . They proposed that these segments correspond to CNV hotspots , regions of recurrent CNV mutations in both genomes . To examine this question in flies , we compared the distribution of duplications in the genomes of the two Drosophila species . Of the 830 duplications detected in D . simulans , 769 ( 93% ) were mapped onto the D . melanogaster genome ( see Materials and Methods ) . Most of the D . simulans duplications that failed to map onto the D . melanogaster genome are located close to the pericentromeric regions ( which are also regions poorly represented on the tiling arrays due to their repetitive nature ) . Out of the 769 D . simulans duplications mapped onto the D . melanogaster genome , 96 ( 12% ) overlap with polymorphic duplications in D . melanogaster . Figure S1 shows the location of the overlapping duplications in the genome . The number of duplications that overlap between the two species is significantly higher than what is expected by chance: randomly shuffling the coordinates of D . simulans and D . melanogaster duplications 1 , 000 times within each chromosome yielded at most 53 ( 7% ) duplications showing overlap , with a median of 32 ( 4% ) duplications overlapping by chance in the two species ( see Materials and Methods ) . The clear excess of duplications overlapping between the two Drosophila species could be due to either shared ancestral polymorphisms or to recurrent mutation at mutational hotspots . For 67 of 96 overlapping duplications , we can directly exclude the shared ancestral polymorphism hypothesis because the size of the duplicated regions varies considerably between the two species . For the remaining 29 duplications the microarray resolution is insufficient to determine if the breakpoints are the same or not . However , the proposition that these 29 duplications represent ancestral shared polymorphisms is unlikely . Neutral polymorphisms are not expected to be retained for the 2–3 million years that have already passed since these two species split . Only 1% of neutral polymorphisms are expected to be retained after 5 . 3N generations [58] , which assuming a population size ( N ) of 106 [59] and 10 generations a year , means 99% of shared polymorphisms should be resolved within ∼530 , 000 years after the two species split . Selection could , in principle , maintain shared polymorphisms for much longer [58] but most of these 29 duplications are either intergenic or intronic , which argues against this hypothesis . Overall , the set of overlapping duplications has a higher fraction of non-coding duplications ( i . e . intergenic and intronic duplications ) than the general dataset ( 80% vs . 64% , respectively , Fisher exact test , p = 0 . 0005 ) . There is no difference in the proportion of partial and complete gene duplications between the two datasets ( Fisher exact test , p = 0 . 3 ) . Table S1 has the location and genomic annotation of all overlapping duplications . A more likely explanation for the observed excess of overlap between the duplications identified in the two species is that there are orthologous regions in the two Drosophila genomes that experience higher rates of duplication . This is also the explanation favored by Perry and colleagues to explain the excess of overlap found between human and chimpanzee CNVs [28] , [30] . CNV hotspot regions shared between human and chimpanzees are strongly enriched with segmental duplications [28] , [30] . Segmental duplications are known to facilitate the occurrence of further duplications ( and deletions ) by mediating non-allelic homologous recombination [3] , [4] and are responsible for the high mutation rates observed at some loci associated with genomic disorders [27] . To investigate the causes for the Drosophila duplication hotspots , we tested for an enrichment of segmental duplications and transposable elements ( also capable of mediating non-allelic homologous recombination ) in these regions . We found that the duplications showing overlap between D . simulans and D . melanogaster were not enriched with either ( Fisher exact test , p = 0 . 7 for segmental duplications and p = 0 . 9 for transposable elements ) . Despite the previous observation linking human/chimpanzee segmental duplications with CNV hotspots , this result was not surprising . Segmental duplications are less abundant in fly than in mammalian genomes , and in flies are mainly restricted to pericentromeric regions [60] where none of the duplication hotspots identified here is located ( these regions are under-represented in the microarrays because of their repetitive nature ) . Transposable elements are also mostly kept to pericentromeric regions [61] , and those that are not have different distributions in the two Drosophila species [46] . In D . melanogaster , polymorphic duplications are not distributed uniformly throughout the genome . There are regions of the genome that show unusually high levels of duplication [33] . Importantly , these regions were shown to be significantly associated with regions of the genome that are late-replicating [33] . Hence , we hypothesized that the duplication hotspots identified between the two Drosophila species were also associated with these late-replicating regions of the genome . There are several high-resolution replication timing maps available for the D . melanogaster genome ( e . g . [62] , [63] ) . Here , we use the replication timing profile generated by Schwaiger and colleagues for the D . melanogaster Kc cell line , where a Hidden Markov Model was applied to classify the genome into early- , mid- and late-replicating regions [62] . Additional replication timing maps for D . melanogaster were generated as part of the modENCODE project for three cell lines ( Kc , S2 and Bg3 [63] ) . Our results were robust to the choice of the replication timing dataset ( Figures S2 and S3 ) . Replication timing varies between -4 and 4 with the former indicating late-replicating regions and the latter early-replicating regions . Figure 3 compares the replication timing profile of duplications that do not overlap between D . simulans and D . melanogaster ( grey ) and those that do ( salmon ) . Consistent with the observation that regions of the D . melanogaster genome that are rich in duplications tend to be late-replicating [33] , we found that duplications that overlap between the two species are also significantly enriched in late-replicating regions ( Wilcoxon rank sum test , p = 0 . 002 ) . This result is strengthened if we restrict our analysis to those duplications that are smaller than 5 kb and therefore are less likely to show overlap due to chance alone . In this latter case , the median replication timing observed for duplications that overlap between the two species is -1 . 5 ( p = 0 . 001 ) , which is also the median replication timing observed for late-replicating regions of the genome as a whole . Figures S2 and S3 show this same analysis using the replication timing profiles of the three cell lines generated as part of the modENCODE project ( Kc , S2 and Bg3 cell lines ) . The results are qualitatively similar . Late-replicating regions of the genome have lower gene density than early- and mid-replicating regions , which means they have larger intergenic regions . We also determined that genes located in late-replicating regions have longer introns than genes located in early- and mid-replicating regions ( median of 200 bp , 77 bp and 88 bp , respectively ) . Given that we observe strong purifying selection against duplications encompassing coding regions , the association between duplications that overlap between the two species and late-replicating regions could be due to a higher proportion of non-coding sequences in these late-replicating regions . Three independent observations do not support this possibility . First , there is not an overall increase of D . simulans duplications in late-replicating regions as would be expected if they were accumulating in these regions due to a lower selective constraint . However , there is a significant excess of overlapping duplications in late-replicating regions when compared to non-overlapping duplications ( Fisher exact test , p = 0 . 009 ) . The same holds true for D . melanogaster duplications . Although there is an overall excess of D . melanogaster duplications in late-replicating regions ( binomial test , p = 0 . 028 ) , the proportion of overlapping duplications in late-replicating regions is significantly higher than the proportion of non-overlapping duplications ( Fisher's exact test , p = 0 . 03 ) . Second , we compared the observed number of late-replicating duplications that overlap between the two species with what would be expected by chance alone . Although there are 47 duplications located in late-replicating regions that overlap between the two species , when we shuffled the coordinates of duplications in late-replicating regions 1 , 000 times exclusively within late-replicating regions we observed at most 12 duplications showing overlap , with a median of 4 ( i . e . only 9% of the actual observed number of duplications showing overlap ) . Third , late-replicating duplications do not show evidence of lower constraint in their site-frequency spectra when compared to duplications located in either early- or mid-replicating regions ( as measured by comparing the proportion of duplications segregating in only one line ) . We therefore conclude that selection is not responsible either for the excess of overlap found between duplications in D . melanogaster and D . simulans or for the enrichment of these duplications in late-replicating regions . Instead our data provides strong evidence for the hypothesis proposed previously [33] that replication timing impacts the genomic distribution of duplication rates . Our data further suggests that the existence of duplication hotspots within late-replicating regions is not simply a consequence of the accumulation of duplications in these regions . Late-replicating regions are probably acting synergistically with other factors , such as particular types of sequences ( e . g . more prone to breakage ) or higher-order chromatin features ( e . g . chromatin condensation ) , to generate the duplication hotspots . Hence , late-replicating regions do not act homogeneously as duplication hotspots . Instead , duplication hotspots correspond to discrete regions that tend to be located within late-replicating regions .
The density of duplications has been shown to vary throughout the human [26] , [27] and fly genomes [33] , and the existence of duplication hotspots has been suggested for these and other species [28] , [31] . By comparing the distribution of polymorphic duplications along two Drosophila genomes , we found a significant excess of duplications overlapping between the two species , suggesting the existence of shared duplication hotspots . In mammalian genomes duplication hotspots are associated with genome regions enriched in segmental duplications [27] , [28] , [31] . We did not find an enrichment of these sequences in Drosophila duplications hotspots . Rather , we found that duplication hotspots are significantly associated with late-replicating regions of the genome , further supporting the hypothesis that some regions within late-replicating regions of the genome experience increased rates of duplication [33] . Prior observations support a link between replication timing and the formation of duplications . For example , in yeast , large spontaneous duplications are associated with replication termination sites [64] . Fragile sites in both humans and Drosophila have been proposed to represent sequences that are late-replicating [65]–[67] and , at least in humans , fragile sites are hotspots for chromosomal rearrangements in cancer [65] , [68] and are also likely to mediate structural variation in the germline [65] . A recent study [69] suggested that fragile sites occur in regions of the genome showing a paucity of replication initiation events . Sparseness of initiation sites would force replication forks to cover longer distances to finish replication , thereby creating the association between fragile sites and late-replication . We tested this hypothesis by determining whether duplications overlapping between the two Drosophila species tended to be , on average , located further away from known origins of replication than the remaining duplications ( and a randomly generated set of sequences ) . We found no significant difference between the two sets of duplications in their distance to known origins of replication ( for origins of replication identified in the Kc , S2 and Bg3 cell lines as part of the modENCODE project ( data not shown [63] ) ) . It is important to note , however , that the location of origins of replication ( and replication timing ) can vary among cell types [62] , [65] , [69] and all analysis reported here were conducted using data obtained from cell lines instead of germline cells . Because replication-associated repair is proposed to be responsible for the formation of both simple and complex CNVs in the human genome [20]–[22] , [24] , [70] , the presence of Drosophila duplication hotspots in late-replicating regions of the genome could be interpreted as supporting an important role for replication-associated repair in the formation of CNVs in these species . However , the association between late-replicating regions and duplication hotspots does not necessarily imply that the latter arise as a direct consequence of replication-associated repair . An increase in DNA double-strand breaks and/or stalled replication forks in particular regions within late-replicating regions that are ( incorrectly ) repaired by the canonical DNA repair pathways ( i . e . non-homologous end-joining or homologous recombination ) would also generate this association . Similarly , duplication hotspots could be associated with regions within late-replicating regions that , while experiencing normal rates of DNA double strand breaks , have a higher rate of incorrect repair ( for example , because of higher chromatin condensation ) . It is also important to note that in D . melanogaster , only duplication-rich regions of the genome were found to be associated with late-replicating regions . Deletion-rich regions were associated instead with early-replicating regions [33] . This is in apparent contradiction with the concept that fragile sites are associated with duplication hotspots because fragile sites are expected to be associated with both types of rearrangements , not only with duplications [65] , [68] . However , deletions tend to be more deleterious than duplications [36] , [37] and so purifying selection preferentially removes them from the population . As a result , even though similar numbers of deletions and duplications are created at hotspots , because of stronger purifying selection acting on deletions , an excess of duplications is instead observed . The existence in D . melanogaster's early-replicating regions of deletion-rich regions , but not of duplication-rich regions , can be explained by the fact that non-homologous end joining is the preferred repair pathway in early S phase ( i . e . in early-replicating regions ) and that it mostly creates deletions [3] , [4] , [18] . In late S phase ( i . e . in late-replicating regions ) homologous recombination is the preferred DNA repair pathway and it generates both duplications and deletions [3] , [4] , [18] . Because our work suggests that duplication hotspots are enriched within late-replicating regions of the genome , we asked if there are particular classes of genes enriched in these regions . A Gene Ontology analysis of the genes located in late-replicating regions of the D . melanogaster genome revealed that these regions are significantly enriched with sensory genes , both olfactory genes ( Holm-Bonferroni correction , p = 5 . 3×10−5 ) and gustatory genes ( Holm-Bonferroni correction , p = 2 . 3×10−4 ) . We confirmed this result by determining the replication timing for all olfactory receptor and gustatory receptor genes ( as defined by [71] ) . Although only ∼20% of the genes in the genome are located in late-replicating regions , more than 40% of gustatory receptor genes and more than 50% of olfactory receptor genes are late-replicating . Figure S4 compares the distribution of replication timing of olfactory receptor and gustatory receptor genes and all genes in the genome . Both classes of sensory genes tend to be late-replicating ( p = 0 . 003 for gustatory receptor genes and p = 2 . 5×10−8 for olfactory receptor genes ) , but olfactory receptor genes tend to replicate later than gustatory receptor genes ( median replication timing of 0 . 21 and -1 . 6 , respectively ) . In the set of D . melanogaster duplications , 5 overlap with olfactory receptor genes and 4 with gustatory receptor genes and in the set of D . simulans duplications , 2 overlap with olfactory receptor genes . There is further evidence in the literature of copy number variation in sensory genes in D . melanogaster ( e . g . [34] , [72] ) . The number of observed duplication polymorphisms encompassing sensory genes is , however , likely to be an under-estimation of the actual number of duplication polymorphisms associated with this class of genes . Microarray probes have to map to unique regions of the genome , which excludes regions with recent gene duplications , such as some of the regions that harbor sensory genes . For this reason , the abundance of sensory genes among copy number variants in Drosophila should be re-examined using next generation sequencing technology , which should not be affected by the existence of recent duplicates ( for an encouraging first step see [34] ) . What would be the predicted dynamics of sensory genes in Drosophila ? Our data suggests that duplication hotspots are enriched within late-replicating regions , but that does not mean that sensory genes are enriched in the late-replicating regions hotspots . Additionally , our data suggests an important role for selection in the fixation of duplications in Drosophila . Thus , even if sensory genes experience , on average , higher duplication rates , this may not necessarily translate into increased numbers of fixed differences in the number of sensory genes between species . Accordingly , McBride and Arguello found little variation in the number of olfactory and gustatory genes in the D . melanogaster subgroup of species ( the exception being a high rate of loss of gustatory receptor genes in the two Drosophila specialists: D . sechellia and D . erecta [71] caused by nonsense mutations , not by deletions ) . High rates of duplication for sensory genes would predict instead increased levels of within-species duplication polymorphisms , which could be translated into increased levels of variation in gene expression . Testing this hypothesis awaits an appropriate dataset describing population-level variability in levels of gene expression for olfactory and gustatory receptor genes in either of the two Drosophila species . Several lines of evidence suggest that selection plays a major role in shaping patterns of duplication polymorphism in D . simulans . The action of purifying selection can be seen in the skew of duplications toward low frequency variants ( i . e . 83% of duplications are present in only 1 of the 14 lines ) but more robustly ( with regards to the alternative hypothesis of demography ) in the strong depletion of coding duplications . A role for positive selection can also be inferred . There is a significant over-representation among high-frequency duplications ( segregating in at least 6 of the 14 lines ) , of complete gene duplications ( 35% of all high-frequency duplications ) . Although there are many ways in which duplications can generate novel phenotypes ( e . g . [15] , [16] , a large fraction are expected to be complete gene duplications [11] , [13] , like the ones segregating at high-frequency in D . simulans . A comparison of the patterns of duplication polymorphism between D . simulans and D . melanogaster suggests stronger selection in the former . The dearth of duplicates overlapping coding sequence is significantly stronger in D . simulans than D . melanogaster , as is the skew of duplications toward low frequency variants . While this latter difference can also be explained by the different demographic histories of the two species ( and of the lines used for each species ) , the difference in the duplication density in coding sequence can only be explained by stronger purifying selection acting on D . simulans duplications . On the other side of the frequency spectrum , positive selection also seems stronger in D . simulans . Although in D . melanogaster there is also a significant increase in the proportion of complete gene duplications among those duplications segregating at high-frequency , there is a significantly higher proportion of complete gene duplications segregating among high-frequency duplications in D . simulans than in D . melanogaster . The hypothesis of stronger selection in D . simulans than D . melanogaster is consistent with previous data suggesting that D . melanogaster has experienced a reduction in its effective population size [49] , [50] . Because the effectiveness of selection is determined by the product of the effective population size and the intensity of selection [73] , the larger the effective population size , the more effective both purifying and positive selection are expected to be . Several observations support this notion for the two Drosophila species . For example , D . simulans has a higher codon bias than D . melanogaster [74] , there are higher levels of amino acid polymorphism in D . melanogaster than D . simulans [49] and there are stronger signatures of purifying selection at synonymous sites in D . simulans than D . melanogaster [51] . Most population genetic models that attempt to describe the early evolutionary trajectories of new duplications ( i . e . gene duplications ) assume that the force responsible for the fixation of the duplication is genetic drift [43] , [44] . These models assume that the ultimate fate of the duplication is dictated by subsequent mutations that occur in one or both copies , which can lead to the permanent preservation of the duplication in the genome or , alternatively , allow its loss [14] , [43] , [44] . D . simulans' duplication polymorphism data suggests instead an important role for selection in the fixation of a significant fraction of duplications . A study of a small number of recently fixed gene duplications in the Arabidopsis thaliana genome also suggested an important role for positive selection in driving these variants to fixation [75] . If the observation made here for D . simulans , that selection plays an important role in the fixation of duplications , holds true , then population genetic models will have to include positive selection when modeling the early stages of the evolution of this class of mutations ( for an example see [76] ) . The observation that a large fraction of duplications are fixed not by drift but by positive selection should not be surprising in light of the overwhelming evidence that between 40-50% of amino acid substitutions in Drosophila species are adaptive [77] .
We generated the dataset of D . simulans duplications by hybridizing the DNA of 14 natural lines to Affymetrix D . melanogaster tiling arrays ( three replicates per line ) . Each tiling array was hybridized with DNA pooled from 30 female virgin flies . Among the 14 lines , 9 were from three different locations in Madagascar ( MD01 , MD04 , MD72 , MD105 , MD197 , MD210 , MD222 , MD236 and MD239 ) , one was from Israel ( SFSR2IIST ) , one from Reunion Island ( W74 ) , one from New Guinea , one from Kenya ( Impala 6 ) and one from Indiana ( Valparaiso ) . The D . simulans lines were selected with the goal of maximizing levels of variability and so were mostly sampled from the known diversity center of the species [46] . The protocol used to prepare the DNA samples for the microarray experiments was the same one used by Emerson and colleagues to detect CNVs in the D . melanogaster genome [36] . The hybridization intensities were decoded into differences in copy number using a Hidden Markov Model . The Hidden Markov Model used here is the same one used by Emerson and colleagues to detect CNVs in the D . melanogaster genome [36] . The only difference between the two genomic surveys lied in the number of probes used . Since we only wanted to use those probes in the array that had a unique and perfect match to the D . simulans genome , we used MegaBlast to blast all ∼3 , 000 , 000 probes present in the array against this genome ( droSim1 ) and kept only those that met our criteria [78] . We ended up with ∼900 , 000 probes with which to survey the D . simulans genome . The raw microarray data and the results from the Hidden Markov Model are deposited in GEO under the accession ( GSE29260 ) . We classified D . simulans duplications as intergenic if they encompass exclusively intergenic sequence , as intronic if they encompass exclusively intronic sequence , as a partial gene duplication if they encompass exonic sequence or exonic and intronic sequence , and finally as a complete gene duplication if they encompass the complete gene structure of a gene ( protein-coding or non-protein-coding ) . Table S1 contains the location of each duplication and its annotation . We looked for the presence of noncoding genes within our dataset using the current D . simulans genome annotation . There is only one non-protein coding gene that overlaps with one duplication: a small nucleolar RNA ( snoRNA , FBgn0256493 ) , completely duplicated and present in 1 of the 14 lines . In D . melanogaster there are 11 duplications that overlap with noncoding genes . We used BEDTools ( v2 . 10 . 1 ) [79] to compare the coordinates of the duplications with the genomic coordinates of all gene structures annotated as part of the Release 3 . 1 of the D . simulans genome . We evaluated the quality of the duplication calls by attempting to confirm a subset of 24 by PCR ( and long-range PCR ) . We used two different strategies . The first was to design a pair of divergent primers within the predicted boundaries of the duplication so that there would only be DNA amplification in the presence of a tandem duplication . Using this strategy we confirmed 18 duplications . Some of the duplications required long-range PCR instead of regular PCR because the amplified bands were larger than 5 kb . We performed long-range PCR using the TaKaRa La Taq system and the recommended protocol . The second strategy was to design a pair of convergent primers outside the predicted duplication boundaries . The presence of a tandem duplication creates a band larger than expected . This second strategy required the use of long-range PCRs and confirmed one additional duplication . We sequenced some of the duplication breakpoints identified using the first strategy . There was a good agreement between the predicted and the actual breakpoints ( Figure S5 ) . The final PCR validation rates for the D . melanogaster and D . simulans duplications were 64/74 ( 86% ) [36] and 19/24 ( 79% ) respectively , and were not significantly different from each other ( Fisher's exact test , p = 0 . 86 ) . The strategy of designing divergent pairs of primers within the putative duplications imposed a limit on the size of the duplications assayed . We limited our confirmations to duplications larger than 300 bp . The confirmation dataset has a mean size of 2 . 6 kb ( vs . 1 . 8 kb in the general dataset ) and the smallest duplication confirmed was 332 bp . The duplications present in the D . melanogaster confirmation dataset were , on average , 5 kb [36] . The duplications in the confirmation dataset were chosen blindly regarding their posterior probabilities of duplication ( the output of the Hidden Markov Model ) and number of probes suggesting the duplication ( the smallest duplication was covered by 5 probes ) . There were no differences between the confirmation dataset and the general dataset in terms of frequency ( i . e . the proportion of duplications detected in only one line vs . multiple lines ) and genomic annotation . Included in the confirmations are 3 D . simulans duplications showing overlap . Within the confirmation dataset there were no apparent differences between the set of duplications confirmed and those that were not . However , given that only 5 duplications were not confirmed there would be little power to detect any differences , even if they existed . Table S1 has the location ( and characterization ) of the duplications confirmed and those not confirmed . Although for a duplication to be called in D . simulans , two consecutive probes had to have hybridization intensities decoded by the Hidden Markov Model as being duplicated , in the original D . melanogaster dataset only one probe was required . Thus , we removed from the set of D . melanogaster duplications all those that were predicted by only one probe . This resulted in excluding 195 duplications . We also converted the D . melanogaster duplication coordinates from release 4 to release 5 using FlyBase's coordinate converter ( http://flybase . org/static_pages/downloads/COORD . html ) , and updated the genome annotation to release R5 . 33 . We mapped the duplications identified in D . simulans onto the D . melanogaster genome ( release 5 ) with BLAT [80] by selecting the reciprocal best hit between the two genomes . Of the 830 duplications , 769 were unequivocally mapped . Most duplications that failed to map were located close to pericentromeric regions in D . simulans and either had no good hit in D . melanogaster or mapped to multiple locations . We required at least 90% of the region duplicated in D . simulans to be unambiguously mapped to the D . melanogaster genome and the difference between the region duplicated in D . simulans and its ortholog in D . melanogaster not to exceed 30% of the size of the duplication in D . simulans . Duplications were considered to overlap when at least 1 bp of a duplication in D . simulans overlapped with 1 bp of a duplication in D . melanogaster . In order to evaluate the significance of the observed number of duplications that overlap between the two species , we compared it with what was observed for 1 , 000 sets of randomly generated coordinates created using BEDTools ( i . e . BEDshuffle ) [79] . For each species , we generated 1 , 000 datasets , perfectly matching the duplication datasets by shuffling the coordinates within each chromosome . Then , for each of the 1 , 000 datasets in each species we determined their overlap . We also did a similar analysis focusing only on late-replication regions . For this analysis we generated 1 , 000 matching sets for the duplications located in late-replicating regions and shuffled the coordinates exclusively within these regions . We used the map of segmental duplications identified by Fiston-Lavier and colleagues [60] and the map of transposable elements identified by Bergman and colleagues [61] for the D . melanogaster genome to evaluate the association between these elements and duplications overlapping between the two species . We considered a duplication to be associated with either a segmental duplication or a transposable element if the distance between them was smaller than 2 kb ( including direct overlap with the duplication ) . For both datasets we updated the coordinates from release 4 to release 5 using the tool Coordinate Converter on Flybase . Schwaiger and colleagues [62] generated the replication timing data described along the text . They generated five replication timing profiles for Kc cells using Affymetrix tiling arrays , which were averaged and then smoothed to generate the replication timing profile for this cell line . Then , using a Hidden Markov Model , they classified genomic regions into early- , mid- and late-replicating [62] . The other replication timing datasets ( for Kc , Bg3 and S2 cell lines ) were downloaded directly from the modENCODE webpage ( http://www . modencode . org/ ) . A very small number of duplications overlapped with more than one replication timing environment ( e . g . early- and mid-replicating regions ) . For these duplications , the replication timing corresponded to the mean replication timing of the two environments . In order to determine if duplications showing overlap between the two species are located , on average , further away from origins of replication than the remaining duplications , we calculated the distance between the two sets of duplications to the origins of replication identified in the Kc , Bg3 and S2 cell lines as part of the modENCODE project ( data downloaded directly from the modENCODE webpage ) . We also compared these results with the distribution of median distances to the three sets of origins of replication generated for 1 , 000 random sets of coordinates matching the duplication datasets . In order to determine if there are particular classes of genes enriched in late-replicating regions we first classified all genes in the genome as early- , mid- or late-replicating . Some genes overlapped with more than one replication timing environment . For these genes we selected the replication timing environment closest to the start of the gene . The results did not change if we chose instead the replication timing environment closest to the end of the gene or if we excluded genes overlapping more than one replication timing environment . We used the Gene Ontology tool on FlyMine [81] ( using the Holm-Bonferroni correction for multiple testing ) to see if there were any classes of genes enriched in the set of genes classified as late-replicating . We performed this same analysis using the Gene Ontology tool Gorilla [82] , which gave similar results ( i . e . enrichment in olfactory and gustatory genes ) . We used the complete list of olfactory and gustatory receptor genes identified by McBride and Arguello [71] to ascribe for each gene their replication timing . If a gene overlapped with more than one replication timing environment we ascribed that gene the mean replication time for the two environments . We also used this list of genes and the list of olfactory and gustatory receptor genes identified in the D . simulans genome by the same authors to identify the polymorphic duplications in both Drosophila species encompassing these genes . All statistical analyses were done using the statistical package R [83] and the application Rstudio ( http://www . rstudio . org/ ) . | DNA duplications are important contributors to the phenotypic differences observed between individuals . These mutations can disrupt the normal functioning of genes and so are often associated with disease . But because they can add genetic information they can also lead to evolutionary change . Understanding how selection and non-random mutation processes shape the distribution of duplications throughout the genome is important to elucidate both the medical and evolutionary impacts of these mutations . Here , we examined the roles of selection and mutation in shaping patterns of duplication polymorphisms across the genomes of the fruit fly Drosophila melanogaster and its sister species , D . simulans . We found that selection is pervasive in both genomes but is more efficient in D . simulans than in D . melanogaster . We also found that these two species have shared duplication hotspots , i . e . orthologous regions experiencing high rates of duplication in the two genomes . After excluding the hypothesis that Drosophila duplication hotspots are associated with regions of the genome rich in segmental duplications ( as observed for mammalian genomes ) , we show that they are associated with late-replicating regions of the genome . Our work therefore proposes a link between DNA replication and rates of duplication across the genome . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetic",
"mutation",
"population",
"genetics",
"mutational",
"hypotheses",
"animal",
"models",
"mutation",
"drosophila",
"melanogaster",
"model",
"organisms",
"mutation",
"types",
"effective",
"population",
"size",
"genomics",
"genetic",
"polymorphism",
"comparative",
"... | 2011 | Drosophila Duplication Hotspots Are Associated with Late-Replicating Regions of the Genome |
Leprosy is an infectious disease affecting skin and peripheral nerves resulting in increased morbidity and physical deformities . Early diagnosis provides opportune treatment and reduces its complications , relying fundamentally on the demonstration of impaired sensation in suggestive cutaneous lesions . The loss of tactile sensitivity in the lesions is preceded by the loss of thermal sensitivity , stressing the importance of the thermal test in the suspicious lesions approach . The gold-standard method for the assessment of thermal sensitivity is the quantitative sensory test ( QST ) . Morphological study may be an alternative approach to access the thin nerve fibers responsible for thermal sensitivity transduction . The few studies reported in leprosy patients pointed out a rarefaction of thin dermo-epidermal fibers in lesions , but used semi-quantitative evaluation methods . This work aimed to study the correlation between the degree of thermal sensitivity impairment measured by QST and the degree of denervation in leprosy skin lesions , evaluated by immunohistochemistry anti-PGP 9 . 5 and morphometry . Twenty-two patients were included . There were significant differences in skin thermal thresholds among lesions and contralateral skin ( cold , warm , cold induced pain and heat induced pain ) . The mean reduction in the density of intraepidermal and subepidermal fibers in lesions was 79 . 5% ( SD = 19 . 6 ) and 80 . 8% ( SD = 24 . 9 ) , respectively . We observed a good correlation between intraepidermal and subepidermal fibers deficit , but no correlation between these variables and those accounting for the degree of impairment in thermal thresholds , since the thin fibers rarefaction was homogeneously intense in all patients , regardless of the degree of sensory deficit . We believe that the homogeneously intense denervation in leprosy lesions should be objective of further investigations focused on its diagnostic applicability , particularly in selected cases with only discrete sensory impairment , patients unable to perform the sensory test and especially those with nonspecific histopathological finds .
Leprosy is an infectious disease affecting skin and peripheral nerves [1] , [2] , [3] , [4] . The neural impairment results in increased morbidity and , sometimes , disabling permanent physical deformities . Prompt diagnosis during the incipient stages is important to avoid these complications . The World Health Organization and the Brazilian Ministry of Health guidelines propose leprosy diagnosis based on the detection of skin lesions with impaired sensation , thickened peripheral nerves or a positive skin smear [5] , [6] , [7] . In the initial phase , the presence of hypochromic macule occurs without neural thickening and with a negative dermal smear , making the sensitivity test of a suspicious lesion an important criterion to establish the early diagnosis . The tactile sensitivity test with Semmes-Weinstein monofilaments is the most common and applicable test among the available sensitivity tests for outpatient setting [8] . In leprosy lesions , loss of tactile sensitivity is preceded by loss of thermal sensitivity , since tactile sensitivity , mediated by thick myelinated A-beta type nerve fibers , can be preserved , even if the loss of thermal sensitivity , mediated by thin myelinated A-delta type fibers and thin unmyelinated C type fibers , has already occurred [9] , [10] . Thus , the assessment of thermal sensitivity is of fundamental importance to establish an early diagnosis . The use of test tubes with hot and cold water is hampered in ambulatory practice since it is time consuming and the water temperature control may be difficult to standardize [11] . The quantitative sensory test ( QST ) performed by electronic equipment is considered the gold standard method to assess thermal sensitivity [12] , [13] , [14] , [15] , [16] , despite its large size , high cost and need for an experienced professional . Morphological study of skin biopsy is an alternative to assess thin nerve fibers structure and densities related to the thermal sensitivity function [17] . The protein gene product ( PGP 9 . 5 ) is a neuronal pan-axonal marker widely used for intraepidermal and dermal nerve endings analysis and quantification . Current guidelines recommend skin biopsy rather than peripheral nerve biopsy in the diagnosis of thin fiber neuropathies [18] since the methods to evaluate peripheral nerve conduction , which assess thick nerve fibers , may fail to detect nerve impairment [19] . In leprosy patients a few morphological studies showed a decrease of thin cutaneous nerve fibers density associated to the worsening of heat and cold detection thresholds . However , such studies have used non-systematic semi-quantitative methods to evaluate the nerve density and the thermal thresholds , and did not exclude treated patients [20] , [21] , [22] , [23] , [24] .
This work aimed to study the correlation between the degree of thermal sensitivity impairment , measured by QST , and the degree of denervation in leprosy skin lesions , evaluated by immunohistochemistry anti-PGP 9 . 5 and morphometry . According to the Brazilian guidelines ( 27 ) , written in the form of ordinances , a case of leprosy is defined if a patient fulfills one of the criteria: skin lesion with decreased sensitivity , positive skin smear or enlarged peripheral nerve . Although the diagnosis can be based on only one of these criteria , only patients who had skin lesions could be included in our work . In our reference center , at the Eduardo de Menezes Hospital of the Minas Gerais State Hospital Foundation , the tactile sensitivity test of suspicious lesions is performed with dry cotton wool and , for research purposes , monofilaments . The thermal sensitivity test is performed with hot and cold water tubes and , for research purposes , sensory thermo-analyzer . The histamine test is performed in suspicious lesions that showed no significant sensitivity impairment . Skin smears are routinely performed on all suspected or confirmed cases . The pathological examination of the lesions is routinely performed for diagnostic investigation in suspected cases and research purposes in previously confirmed cases . Clinical examination of peripheral nerves is done routinely in suspected and confirmed cases . Leprosy patients with at least one skin lesion with a minimum diameter of three centimeters to offer an adequate docking area for the thermal stimulator were included if they had been in treatment for maximum of 30 days . Patients were excluded if lesions were located in body regions impaired by leprosy peripheral neuropathy , clinically evaluated by peripheral nerves palpation , tactile sensitivity test of the palms and soles , evaluation of muscular trophy and palms and soles hydration . We also excluded patients with other diseases known to cause peripheral neuropathy , such as alcoholism , diabetes , HIV infection , thyroidopathy , metabolic disturbances or systemic vasculitis as well as patients with limited cognitive ability , unable to respond adequately to QST , or whose scars from previous biopsies compromised the skin areas to be studied . The study was approved by the Research Ethics Committee of the Federal University of Minas Gerais and all participants gave their written informed consent . The Statistical Package for Social Sciences ( SPSS ) 15 . 0 ( SPSS , USA ) was used in the descriptive analysis measures of central tendency ( mean and median ) , variability ( the standard deviation and coefficient of variation ) and percentages . For comparative analysis between the paired groups “lesions” and “contralateral skin” , nonparametric continuous variables were studied with Wilcoxon test , Kruskal-Wallis test and Spearman correlation coefficient . Taking as reference the work of Cohen [26] , the sample size of twenty-two patients was sufficient to study the correlation aiming a power of 0 . 8 and accepting as parameters a significance of 0 . 1 and an effect size of 0 . 5 .
Between January and December 2009 twenty two new leprosy cases were included in the study . All patients had lesions with thermal hypoesthesia and/or tactile deficit . At least 19 patients showed typical histology of the disease . Three patients had only mild perineural , perivascular and periannexal lymphocytic infiltrate . These three had hypochromic macules with indubitable thermal and/or tactile hypoesthesia . All patients showed improvement of skin lesions after treatment initiation . Eleven patients were men . The mean age was 42 years ( ranging between 10 and 73 years ) . Ten patients ( 45 . 5% ) presented with six or more lesions and four ( 18 . 2% ) had more than one clinically impaired peripheral nerve . According to the WHO operational classification ten patients ( 45 . 5% ) were considered multibacillary . Under the Ridley-Jopling classification four patients ( 18 . 2% ) were considered to have the indeterminate form of leprosy , two ( 9 . 1% ) had the tuberculoid form , five ( 22 . 7% ) had the borderline tuberculoid form , nine ( 40 . 9% ) had the borderline lepromatous form and two ( 9 . 1% ) had the lepromatous form . Tactile sensitivity was preserved to the 0 . 05 g , 0 . 2 g , 2 . 0 g , 4 . 0 g and 10 . 0 g monofilament in 30% , 30% , 20% , 15% and 5% of the lesions , respectively . The lesions with preserved or mostly preserved tactile sensitivity tended to present a longer evolution time until diagnosis than lesions with a more pronounced loss of tactile sensitivity ( Fig . 1A ) . QST results were different in “lesions” and “contralateral skin” groups for each of the thermal thresholds evaluated: CPT , WPT , CPPT and HPPT ( Table 1 ) . We noticed a trend towards an association between worsening WPT ( Fig . 1B ) and CPT ( Fig . 1C ) and the degree of tactile sensitivity loss . The loss of thermal sensitivity , interpreted as the absolute and the percentage difference between lesions and contralateral skin is presented in Table 1 and is stratified according to the tactile sensitivity of the lesions in Table 2 . The subepidermal nerve fibers in contralateral biopsies were visualized as thicker brown structures with variable diameter and density of staining . These fibers were linear or sometimes grouped in small clumps , placed parallel to the basal layer and sometimes positioned to cross it towards the surface ( Fig . 2A ) . The contralateral skin biopsies also showed intraepidermal nerve fibers as discontinuous linear structures stained in brown , resembling “beads” that vertically crossed the basal , spinous and granular layers toward the skin surface , then bent and continued parallel to the corneum stratum , inside of which they could eventually be seen ( Fig . 2C and D ) . However , not all intraepidermal fibers were visualized throughout their entire course . Often , the fibers crossed the section obliquely and were only seen in portions of its course . Lesion biopsies showed a substantial decrease in subepidermal fibers ( Fig . 2B , arrowheads ) and intraepidermal fibers ( Fig . 2E ) . Evaluation of intraepidermal and subepidermal denervation was performed in each patient by determining the percentage difference between the values obtained in the lesion and the contralateral skin . The lesions presented a mean rarefaction of subepidermal fibers of 79 . 5% ( SD = 19 . 6; coefficient of variation = 0 . 24 ) and a mean reduction of intraepidermal fibers of 80 . 8% ( SD = 24 . 9; coefficient of variation = 0 . 30 ) compared to the contralateral areas ( Table 1 ) . Good correlation was detected between the deficit of intraepidermal fibers and the deficit of subepidermal fibers ( Spearman coefficient: 0 . 60; p = 0 . 004 ) . However , the bivariate analysis showed lack of statistical correlation between intraepidermal or subepidermal fibers deficit and thermal thresholds deficit .
The lack of correlation between cutaneous denervation and thermal sensitivity impairment may arise from the fact that variations in thermal thresholds are also related to changes in the innervation of deeper skin layers that we did not approach in the current study . However , it should be considered that we have not yet conducted the qualitative ultrastructural analysis of nerve endings , the investigation of the immunocytochemical expression of the several neurotransmitters involved in thermal sensitivity control , or the study of inflammatory cytokines expression , factors that may be related to the functional impairment of cutaneous nerve endings [48] , [49] . | Our study has addressed objectively the rarefaction of cutaneous thin nerve fibers density and its correlation with quantitative thermal sensory test in leprosy patients . Thermal sensitivity evaluation is crucial to the early diagnosis of leprosy , since it is the first type of cutaneous sensitivity lost in the lesions . However , some patients are unable to perform thermal tests , like children and patients with cognitive impairment . The pathological study of those lesions is also fundamental and it is the gold standard method to diagnose and classify leprosy patients . However , it may also be unspecific , especially in the indeterminate form of leprosy . Our findings show that even in lesions with slight thermal sensitivity impairment , there is a homogeneously intense denervation , specifically in the superficial skin layers . We believe that our findings pave the way to future studies focused on the diagnostic applicability of the cutaneous thin nerve fibers density quantification in leprosy suspected lesions . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"dermatology",
"infectious",
"diseases",
"bacterial",
"diseases",
"public",
"health",
"and",
"epidemiology",
"skin",
"infections",
"leprosy",
"public",
"health"
] | 2012 | Degree of Skin Denervation and Its Correlation to Objective Thermal Sensory Test in Leprosy Patients |
The major histocompatibility complex ( MHC ) contains the most polymorphic genetic system in humans , the human leukocyte antigen ( HLA ) genes of the adaptive immune system . High allelic diversity in HLA is argued to be maintained by balancing selection , such as negative frequency-dependent selection or heterozygote advantage . Selective pressure against immune escape by pathogens can maintain appreciable frequencies of many different HLA alleles . The selection pressures operating on combinations of HLA alleles across loci , or haplotypes , have not been extensively evaluated since the high HLA polymorphism necessitates very large sample sizes , which have not been available until recently . We aimed to evaluate the effect of selection operating at the HLA haplotype level by analyzing HLA A~C~B~DRB1~DQB1 haplotype frequencies derived from over six million individuals genotyped by the National Marrow Donor Program registry . In contrast with alleles , HLA haplotype diversity patterns suggest purifying selection , as certain HLA allele combinations co-occur in high linkage disequilibrium . Linkage disequilibrium is positive ( Dij'>0 ) among frequent haplotypes and negative ( Dij'<0 ) among rare haplotypes . Fitting the haplotype frequency distribution to several population dynamics models , we found that the best fit was obtained when significant positive frequency-dependent selection ( FDS ) was incorporated . Finally , the Ewens-Watterson test of homozygosity showed excess homozygosity for 5-locus haplotypes within 23 US populations studied , with an average Fnd of 28 . 43 . Haplotype diversity is most consistent with purifying selection for HLA Class I haplotypes ( HLA-A , -B , -C ) , and was not inferred for HLA Class II haplotypes ( -DRB1 and—DQB1 ) . We discuss our empirical results in the context of evolutionary theory , exploring potential mechanisms of selection that maintain high linkage disequilibrium in MHC haplotype blocks .
Human leukocyte antigen ( HLA ) genes in the major histocompatibility complex ( MHC ) on Chromosome 6 provide the core function of antigen presentation for the adaptive immune system . Each HLA allele can present a restricted repertoire of peptides from either self or non-self proteins to T cell receptors . HLA loci are among the most polymorphic in the human genome [1] [2] , as are their close MHC homologs in other organisms [3] . HLA allele homozygotes have been suggested to be at a significant disadvantage in that their peptide repertoire is more limited than heterozygous individuals [4] . Other genetic systems that have comparable levels of polymorphism to HLA in humans include the olfactory receptors and killer immunoglobulin-like receptors ( KIRs ) [5 , 6] . The direction and magnitude of selective pressure on genes can be estimated through the analysis of allelic variation within populations . For most genes , the nonsynonymous substitutions that alter the amino acid sequence of a protein are typically neutral or deleterious , with very few advantageous variants appearing . However , for HLA genes amino acid variation in the antigen recognition domain of HLA proteins determines the repertoire of peptides loaded onto the HLA protein and presented to the T cell receptor . Compared to other genes nonsynonymous variants in HLA genes were proposed to be often advantageous because a novel peptide repertoire may improve control of evolving pathogen strains [7] . Comparing patterns of variation among different genes , HLA has among the highest ratios of nonsynonymous substitutions relative to synonymous substitutions , which is a hallmark of balancing selection ( i . e . selection maintaining a larger number of alleles than expected from genetic drift ) . Furthermore , the Ewens-Watterson test for neutrality also shows that observed HLA allele homozygosity is less than expected , another indicator consistent with balancing selection [8] . Host-pathogen co-evolution has also been proposed to lead to balancing selection that maintains high levels of HLA allelic diversity within populations . Viral escape mutations may be positively selected even if they have a fitness cost to the virus [9–13] , reducing the number of epitopes from existing viral variants that can be presented by frequent HLA alleles . This process of immune evasion could produce a fitness advantage to hosts carrying rare HLA alleles . As pathogens maintain escape mutations in epitopes presented by the more frequent HLA alleles within a population , negative frequency-dependent selection favors less frequent HLA alleles [14] . While the forces shaping the HLA allele frequency distribution have been extensively discussed , the forces affecting co-occurrence of alleles across HLA loci and the resulting haplotype ( allele combination ) frequency distribution have not yet been thoroughly examined . HLA haplotype dynamics add another layer of complexity since HLA alleles are in clear linkage disequilibrium [15] . Some sets of HLA alleles co-occur on haplotypes more often than expected given the allele frequency distribution and other sets of alleles co-occur much less often than expected . A large number of new haplotypes emerges in the human population in every generation through recombination when compared to less frequently occurring mutation events . For example , the Southeast Asian population and the European populations have been estimated to diverge about 23 , 000 years ago [10] , yet 64% of the Asian HLA haplotypes are not represented in the European population . HLA haplotype diversity across loci is thus far greater than the allelic diversity at a single locus . Viral escape mutations typically alter recognition by a single HLA allele . However , specific haplotypes may provide better control of pathogens , or ensure proper activation of natural killer ( NK ) cells [16] . The innate immune response can be modulated by engagement of inhibitory KIR receptors on NK cells with HLA Class I ligands on target cells . The KIR ligand status is dependent on epitopes present on a subset of HLA alleles . Different HLA alleles will engage with different KIRs , or no KIRs at all , and these interactions influence the degree to which NK cells are licensed to kill target cells in which HLA expression is disrupted . Beyond immune modulation from HLA polymorphisms there is also genomic copy number variation for the number of inhibitory and activating KIR receptors . Allele combinations at different loci on the same haplotype that present multiple epitopes from frequent pathogens could be preferred because they could have better redundancy to prevent immune escape . Interestingly , new results using the ratio of synonymous to non-synonymous mutations suggest selection favors heterozygotes with more divergent allele sequences [17] . This same mechanism preferring divergence in sequence ( and therefore function ) could also apply across HLA loci on haplotypes . Epistatic selection has been argued to affect immune and autoimmune responses [18] . Within the HLA locus , epistatic effects have been observed in the class II region [19] . In viruses , combination of epistasis and balancing selection has been shown to affect the genomes of viral populations [20] . While the high linkage disequilibrium ( LD ) between HLA alleles has long been known [21] , possible models for the selection inducing this LD and its effect on haplotype frequency distribution have never been studied empirically . Such an analysis has been hampered by the need for extremely large samples to ascertain the shape of the haplotype frequency distribution . To meet this challenge we employed a large dataset of HLA haplotype frequency estimates for 23 United States populations derived from over six million volunteer stem cell donors recruited by the National Marrow Donor Program ( NMDP ) registry [22] . We developed a set of population genetic models to attempt to infer the selection pressures that operate on this large haplotype frequency distribution . We here present multiple pieces of evidence indicating that the HLA haplotype frequency distribution deviates from expectations under neutral evolution [23] , and conclude that selection favoring existing frequent haplotypes best explains the distribution of HLA haplotypes observed .
High resolution HLA haplotype frequencies were previously estimated using the expectation-maximization ( EM ) algorithm based on NMDP registry HLA genotypes [24] . Accuracy of HLA haplotype frequency estimates are limited by the resolution of the input genotypes [25] . HLA typing assays historically could not distinguish between all known HLA alleles , due to either lack of ability to phase polymorphisms within the gene , or lack of complete sequencing of the gene , or both . The resulting HLA typing results were considered “low resolution” and are represented by a list of possible allele pairs , or genotypes , that might be present at that locus . “High resolution” genotypes are generated when all known alleles that differ in the exons coding for the antigen recognition site ( exons 2 and 3 for class I and exon 2 for class II ) were distinguished experimentally . The NMDP haplotype frequencies were estimated at high resolution and utilized as input both low resolution genotypes and high resolution HLA genotypes . Low resolution typing within a locus can result in misidentification of the high resolution allele in the frequency data , while a lack of experimental data on haplotype phase between HLA loci can lead to the construction of incorrect arrangements of high resolution alleles into haplotypes . In order to assure that EM-based estimates of HLA haplotype frequency distributions are suitable for estimating selection , we performed several validation procedures . To assess the effect of incomplete HLA typing , we analyzed 38 , 715 donors with both high and low resolution typing and measured concordance between the predicted high resolution haplotype constructed from the EM algorithm using the low resolution typing and the experimentally measured high resolution typing ( Fig 1A ) . The discordance rate ( most probable predicted high-resolution haplotype different than actual ) varied from 14% for very rare haplotypes ( frequency<1 . e-7 ) to less than 1% for frequent haplotypes ( f>1 . e-2 ) . However , this discordance does not lead to an error in the overall shape of the frequency distribution , since low frequency haplotypes are typically replaced by other low frequency haplotypes and not by high frequency haplotypes , as can be seen by the Quantile-Quantile ( QQ ) plot ( position of percentiles of one distribution vs the other ) of the low and high resolution haplotype distributions ( Fig 1B ) . To further validate that the correct haplotypes are replaced by discordant haplotypes of similar frequency , we divided all haplotypes based on their frequency in the EM based haplotype distribution . We then computed the average frequency of the appropriate high resolution haplotypes . As can be seen in Fig 1C , the distributions are similar . To validate that phasing errors do not affect the shape of the expected haplotype distribution , we compared haplotype phasing of 4 , 000 cord-mother pairs using EM versus direct counting using pedigree analysis ( see Methods ) . We found that the shape of the haplotype frequency distribution was not appreciably affected by EM phasing errors . ( Fig 1D ) . In order to correlate haplotype and allele frequencies with possible selection models , we define here in detail the terms we use describing selection forces for alleles and haplotypes along with the many underlying evolutionary mechanisms that have been proposed to contribute to these selection forces . We use “balancing selection” as an umbrella term for all those selection pressures that lead to a greater diversity in HLA allele frequency distributions than what would be expected under a neutral evolutionary model . Underlying this balancing selection are several distinct evolutionary mechanisms that together may combine to form the allele frequency distributions we observe . The model of negative frequency-dependent selection , or rare allele advantage , suggests that continual evolution of viral strains to evade common HLA variants maintains high diversity in HLA alleles in a population . Heterozygote advantage , or overdominance , is a model where heterozygote genotypes have higher fitness than homozygote genotypes . Heterozygotes are capable of presenting a wider peptide repertoire than homozygotes , which would confer improved likelihood of immune detection of pathogens . Takahata and Nei found that “Minority advantage considered here produces essentially the same pattern of genetic polymorphism as that for overdominant selection” , and many other researchers have since encountered similar challenges in teasing apart the mechanisms behind balancing selection [26] . We here introduce "purifying selection" as a parallel umbrella term for all those selection pressures that lead to less diversity in HLA haplotype frequency distributions than what would be expected under a neutral model . Several evolutionary mechanisms may underlie purifying selection . Under purifying selection , rare deleterious mutations are continually purged from populations because they contribute to lower fitness . Purifying selection can occur at the haploid level as HLA alleles are co-dominant or at the diploid level in the case of recessive deleterious mutations in the MHC . Positive frequency-dependent selection is another potential mechanism that would favor more frequent HLA haplotypes that most effectively modulate the immune response . As was the case with alleles , both frequency-dependent and non-frequency dependent mechanisms can induce the same haplotype frequency distribution we observe . HLA allele frequencies have long been shown to be consistent with balancing selection , [8 , 15] . An important indicator for the action of balancing selection is the Ewens-Watterson homozygosity test [27] . In order to test for balancing selection on single-locus allele distributions within the studied populations , we performed Ewens-Watterson tests on random subsamples of 1 , 200 alleles for each US subpopulation [28] ( S1 Table ) We also computed the observed and expected homozygosity ( as predicted from the sum of squares of haplotype frequencies ) , and normalized deviate of homozygosity Fnd [29] of the subsamples . Negative Fnd indicates observed homozygosity below expected homozygosity . The Ewens-Watterson test shows that many , but not all , populations exhibit homozygosity values significantly lower than the expectation from neutral evolution in a fixed population , suggesting balancing selection ( See Methods for multiple measurement corrections method ) . Further , negative Fnd values were observed for the HLA alleles in all subpopulations ( Fig 2 and S1 Appendix showing lower homozygosity than expected in alleles ) in agreement with previous reports [8] . The largest difference between expected and observed homozygosity occurred in HLA-C and DQB1 ( average Fnd values of -1 . 23 and -1 . 33 ) , and on average over all populations , balancing selection was observed in all loci . Some populations show balancing selection more clearly than others , with Koreans and Caribbean Hispanics displaying the strongest deviation from neutrality . The only notable exception to this observation is the DRB1 locus in the Filipino population , which did not exhibit balancing selection . Detecting selection on haplotypes is more complex than for alleles . Thus , we applied a set of different tests—all showing clear signs of deviation from neutral drift toward a lower than expected diversity , consistent with models , where existing frequent haplotypes are favored over new rare haplotypes . We computed the normalized linkage disequilibrium value Dij' for all 2-locus HLA haplotypes , as defined by Lewontin [30] , and estimated its value as a function of the haplotype frequency . Lewontin’s Dij' is a normalized coefficient of linkage disequilibrium between two specific alleles ranging between -1 and 1 , with positive values indicating that the specific 2-locus haplotype is more frequent than expected by the marginal frequencies of the two alleles . A 2-locus haplotype is defined as a combination of two alleles ( e . g . one allele of HLA-B and one allele of HLA-C ) or two haplotypes of multiple loci in the same class cluster ( one haplotype of Class I and one haplotype of Class II where each haplotype is treated as if it were a single-locus allele ) . In a neutral model , rare haplotypes would have a low Dij value , while frequent haplotype would have a high Dij ( See S2 Text for simulation results ) . Positive FDS or similar purifying selection mechanisms would push more haplotypes from intermediate to high frequency values ( for selectively favored haplotypes ) and to low values ( for haplotypes selected against ) . Positive FDS may thus produce negative Dij' values for intermediate frequencies and positive values for low and high frequencies . An analysis of the patterns of linkage disequilibrium in all four relevant pairs of loci: A~B , A~C , B~C , and DRB1~DQB1 , across all populations uniformly shows this bimodal pattern with Dij' of rare haplotypes close to zero , negative linkage disequilibrium for intermediate frequency haplotypes , and positive LD for the most common haplotypes ( Fig 3 ) . The main effect was observed for the A~B and A~C Class I haplotypes , in agreement with the results from the Ewens-Watterson tests discussed below . In order to test that this linkage disequilibrium pattern is not an artifact of the population dynamics , we have performed simulations of neutral evolution ( See S1 Text for description of simulations ) , using realistic parameters for the HLA loci ( See S2 Table for parameter estimates ) , and show that the neutral simulations do not produce the observed LD patterns . ( S2 Text . Section 1 ) . These simulations were performed for either constant or growing populations and with or without population substructure . We have further produced model simulations of populations with positive FDS to show that such populations can display a minimal value of Dij in intermediate frequency ranges ( S2 Text Section 3 ) . This is obviously not a proof that positive FDS is the model driving the observed dynamics , only that positive FDS is a possible mechanism for the purifying selection of HLA haplotypes . In order to further test deviation from the null model of neutral evolution , we performed the Ewens-Watterson test on the five-locus haplotype frequency distribution . In contrast with allele frequency , in haplotype frequency distributions a clear positive and significant Fnd is observed for the 5-locus haplotype frequencies of all populations ( average Fnd = 28 . 43 ) . The resulting positive Fnd values are consistent with positive selection . ( Fig 2 and S1 Appendix ) . These results are robust to sampling and to changes in sample size ( S2 Text Section 4 ) . While the Ewens-Watterson test was developed specifically for non-recombining loci , its efficacy in the detection of positive selection in the case of haplotypes has previously been established [31] . Moreover , the deviation from neutrality attributed to recombination generally decrease the haplotype homozygosity [32 , 33] and is thus not expected to be interpreted as positive selection . In order to test that the presented results are not the result of recent population growth , population substructure , the high recombination and mutation rate of the HLA loci , the balancing selection on alleles or sampling effect , we simulated such scenarios ( see detailed list of scenarios studied in S1 Text ) . None of the simulated scenarios produced deviation from neutrality that approached what was observed in our haplotype frequency distributions ( S2 Text Section 5 ) . Moreover , none of these scenarios led to a combination of positive Fnd values for haplotypes and negative Fnd values for alleles . However , when balancing selection on alleles is combined with purifying selection on haplotypes the opposite deviations from neutrality can be easily obtained ( S2 Text Section 6 ) . In order to test the robustness of these results to sampling , we have repeated the analysis 100 times for each subpopulation , and obtained a very limited variance over all populations studied ( S2 Text Section 5 ) . As was the case for alleles , large variation in Fnd values was observed among all populations . Among the broad race groups , the Fnd statistic was significant at the p<0 . 001 level after multiple measurement corrections for all populations . Fnd was largest in the European population , and comparably large for the rest , while among the detailed race groups the Vietnamese population showed the strongest effect ( S1 Appendix ) . Note that all populations show a very clear deviation of excess homozygosity , in opposition to the observation in the allele frequency distributions . In order to test which haplotypes affect the deviation from the null hypothesis of neutral evolution , we performed the Fnd test described above on two-locus and three-locus HLA Class I and two-locus Class II haplotypes . The results show positive Fnd across most populations for the HLA Class I haplotypes , and especially for haplotypes containing the HLA-A locus . Meanwhile , Fnd values were mostly negative for the DRB1~DQB1 HLA Class II haplotypes ( Fig 2 ) . The Fnd statistic and p-values from the Ewens-Watterson test are measures of deviation from neutrality and equilibrium and not directly measures of selection . Other factors may affect the observed deviations , as has been explored by Akey et al [35] . The Dij' measure may also be affected , in theory , by other elements that impact LD . In order to directly test whether purifying selection models are better consistent with the observed distribution than neutral models , we directly analyzed the haplotype frequency distribution—examining the relation between the frequency of a haplotype in the sample and the number of unique haplotypes at that frequency . By very general arguments , one may describe the allele and haplotype frequencies by a birth-death process . If birth and death are balanced , the population can be studied in equilibrium , while if birth exceeds death , an out-of-equilibrium model must be constructed . If one assumes no selection , two models can be considered: One can model purifying and balancing selection through a frequency-dependent selection process . If such a model is invoked , a third type of stochastic model known as a Birth , Death and Innovation Model ( BDIM ) can be used to fit the observed frequency distribution . BDIM models admit the possibility of density-dependent growth and death rates , which can be interpreted as a non-neutral evolution [38] . Specifically , the total growth and death rates of each sub population are proportional to the population size plus a constant . If these constants are 0 , the model is neutral . The details of all models are explained in S2 Text Section 7 . Note that in this context density-dependent selection can be used as a rough proxy for other types of selection in the sense that the resulting haplotype frequency distribution can be compared with the one expected in neutral evolution . In order to identify which model best represents the observed distributions , we fit both the Yule-Simon and BDIM models to the frequency distributions of haplotypes using a maximum likelihood approach . The Ewens model was not suitable for fitting because it has no free parameter ( except for a normalization constant ) , and the distribution did not fit the observed distribution . The functional forms of all distributions are described in the methods section . The BDIM model contains a selection parameter , which determines whether small populations have a higher or lower net growth rate than large populations . A positive value implies that small populations are selected against ( positive FDS favoring existing frequent haplotypes ) , while a negative value implies that small populations are selected for ( balancing selection ) . Model selection was performed using the Bayesian Information Criterion ( BIC ) measure which incorporates both quality of fit and number of parameters—more parameters have a higher penalty to avoid overfitting . Among the models above , the BDIM model had a significantly better fit with the HLA haplotype frequency distributions , even when accounting for its extra parameter by using BIC for model selection [39] ( Fig 4A ) . The positive ΔBIC values for BDIM minus Yule ( which is better than Ewens ) indicate that BDIM is produces a significantly better model than any of the neutral models . The selection parameters in the BDIM model , which can be interpreted as the net fitness disadvantage of rare haplotypes , were significantly larger than 0 ( Fig 4B ) . No significant advantage existed for either model in fitting the allele frequency distributions . In order to estimate absolute fit of the BDIM model to the data , the allele and haplotype frequency distribution was binned into 20 logarithmic bins , and the R^2 value of the comparison between the predicted and observed distribution was computed for all populations and all allele and allele combinations , as in Fig 4 . The average coefficient of determination ( R^2 ) value over populations , alleles and haplotypes ( A , B , C , DR , DQ and 5 locus haplotypes ) was 0 . 75 . and 0 . 68 for the Yule model .
Applying several population genetic models , we find that frequent HLA haplotypes occur more often than would be expected under a neutral evolutionary model in all studied US populations , which suggests purifying selection . At the same time , we corroborated numerous previous studies showing that balancing selection may be operating at each individual HLA locus . Taken together , the multiple complementary analyses , fitting the frequency distributions to different evolutionary models , the Fnd measure of homozygosity deviation from a null model , and linkage disequilibrium analysis , all suggest purifying selection at the haplotype level , and that positive FDS provides a good fit for the haplotype frequency distribution . While we do not explicitly simulate the diploid fitness-based selection models , theoretically both frequency-dependent and non-frequency-dependent mechanisms could produce the same type of distribution . To our knowledge , this is the first time that explicit evolutionary population dynamics models have been compared across such diverse populations at a scale of millions of individuals . While these models utilize the mechanism of frequency-dependent selection , they may be a proxy for other purifying selection mechanisms . Indeed , many other researchers have found that multiple disparate evolutionary mechanisms each capable of producing the same shape for observed frequency distributions [40] . While we are unable to tease apart the exact mechanisms involved , the main contribution of this paper is the identification of empirical data suggesting that selection has an opposite impact on allele frequency versus haplotype frequency distributions . Note that we have not explicitly modeled population structures . We will now develop models combining population structure and selection to test their combined effect . Multiple previous results suggest that HLA haplotype frequencies are shaped by selection . The high levels of linkage disequilibrium observed among HLA alleles serve to limit the amount of diversity in HLA haplotypes and multi-locus genotypes . Several different HLA haplotypes have been maintained at high frequency in different populations over long periods of time and have been termed “conserved extended haplotypes” or “ancestral haplotypes” [41] . If the HLA system had lower linkage disequilibrium , more combinations of alleles at different loci would be observed at higher frequency in populations . Finally , the amount of genetic recombination between HLA loci does not correlate directly with genetic distance in the MHC [42] , indicating that selection may be shaping patterns of human MHC haplotype variation . As the recombination rate between HLA loci is faster than the allele formation rate through mutation or gene conversion , these two processes can be interpreted as two different time scales in our evolutionary models . Over long time scales , new alleles are introduced within a single locus . At shorter time scales , certain haplotypes are generated from existing alleles by recombination . While new haplotypes continuously arise , the number of highly successful haplotypes would be limited compared to the space of all possible combinations . These successful conserved extended haplotypes may be maintained in populations . The fitness of specific haplotypes and multi-locus genotypes thereof may differ over time and among populations [14] . Within the HLA region , the most significant deviation from neutrality was observed in haplotypes composed of Class I haplotypes , while no such deviation was observed in the Class II haplotypes . This evidence for Class I selection is in good agreement with the effect of Class I variation on survival in the presence of different pathogens [43 , 44] , and also may be correlated with the interaction between HLA Class I and KIR [45] . Note that the functions of Class I and Class II alleles and haplotypes can have epistatic effects that may impact how selection operates on the overall HLA system . Moreover , differences in migration rate in class I and class II could explain the difference between the two regions , However , such differences cannot explain the difference between alleles and haplotypes . A possible caveat is the sample used . In theory , the NMDP cohort may not be representative of the general population . However there are no obvious recruitment practices that would lead significant systematic HLA genetic bias within donor populations . Thus it is common practice for registry and blood bank donors to be used as controls in disease association studies ( e . g . the Wellcome Trust Case Control Consortium used blood bank donors ) . Moreover , there are no obvious reasons that such a misrepresentation would affect the difference between alleles and haplotypes . Multiple evolutionary mechanisms may explain purifying selection for HLA Class I haplotypes . HLA alleles found along the same haplotype may have complementary peptide repertoires across loci to present multiple epitopes from a single viral protein simultaneously . In the case of SIV in monkeys , there has been selection for certain combinations of HLA Class I alleles across loci that control SIV and its escape variants [46] . If pathogens require multiple mutations to achieve immune escape from all HLA alleles in an individual , the likelihood of escape is minimized . Haplotypes containing alleles with redundant recognition capabilities may be preferentially selected for fitness in individuals , while haplotypes without complementary repertoires would be eliminated . Alternatively , haplotypes may need to present epitopes from multiple different viruses . We have shown that the number of epitopes presented by different HLA alleles can vary over many orders of magnitude [11 , 47] . Haplotypes with more limited epitope repertoires may be detrimental , and selected against . Finally , as mentioned , interactions with KIR proteins on natural killer cells may determine the capacity of the immune system to mount a response [48] , requiring specific HLA allele combinations to ensure adequate response . A possible unified selection model for the patterns of diversity observed in both HLA alleles and haplotypes has been proposed by van Oosterhout called Associative Balancing Complex ( ABC ) selection provides an explanation for how linkage disequilibrium between HLA alleles could be maintained by epistasis in the MHC region [49] . Under this model mutations in MHC haplotype blocks accumulate under a sheltered load near HLA genes . Recombination in HLA haplotypes would expose low fitness homozygous genotypes . Epistatic selection operates against this recombination and increase linkage disequilibrium . Purifying selection against deleterious recessive mutations is weak because recombination is low . Frequent HLA haplotypes are maintained and increase divergence from one another over time . While we do not model ABC selection explicitly , our data is consistent with this model of balancing selection on HLA alleles , epistatic selection that limits recombination , and purifying selection in HLA haplotypes . HLA genes are distributed throughout the MHC throughout a large ~4-megabase region of Chromosome 6 . Because the distance between HLA loci can be as much as 1 megabase , HLA haplotype phase cannot be experimentally determined with current classic HLA typing methods . HLA alleles were phased into haplotypes computationally using the expectation-maximization ( EM ) algorithm rather than experimentally . The EM algorithm attempts to find a maximum likelihood estimate wherein all HLA unphased genotypes are explained using a minimal set of HLA haplotypes . We have here shown that while the EM produces a non-negligible level of allele classification and phasing errors , these errors have a minimal effect on the shape of the resulting haplotype frequency distribution . Population substructure in HLA-typed cohorts can cause excess homozygosity which would confound the selection model results , however we find that our consistent results across populations , along with differing forces at Class I versus Class II rules out the possibility that population substructure could explain our findings . The HLA-DP locus has not been included in our analysis due to the historical absence of typing information in the registry and subsequent haplotype frequency datasets . We recognize that extending this analysis to include this locus will be challenging due to the relatively higher rate of recombination between HLA-DP and the other HLA loci in this study [50–52] . The here described opposite selection types and alleles and haplotype levels may be a general evolutionary principle combining the introduction and novelty and the maintenance of high fitness combination . We now plan to further study theoretically and experimentally the presence of such mechanisms in other genomic regions and organisms .
Five-locus high resolution HLA A~C~B~DRB1~DQB1 haplotype frequencies were estimated using the expectation-maximization ( EM ) algorithm for over six million donor HLA typings from the National Marrow Donor Program registry ( USA ) published by Gragert et al [14 , 53] . Given typing ambiguities , a large number of very low probability haplotypes emerge from the EM . The frequency distribution of such rare haplotypes and genotypes is a mere artifact of the EM . We removed low probability haplotypes by assigning each person in the sample a single pair of their most probable haplotypes with the remainder of their haplotype pair probability distribution discarded . The population haplotype frequencies were thus recalculated with a single haplotype pair assigned for each individual . Allele frequencies were derived as marginal sums of the haplotype frequencies . In order to confirm the accuracy of these haplotype frequency estimates , we performed two validation experiments in which the most likely high resolution HLA genotype was imputed for individuals that had HLA typing ambiguity [25] . The first validation dataset , intended to confirm the accuracy of high resolution HLA allele identification , consisted of 38 , 715 registry donors who had high resolution confirmatory typing performed on behalf of a patient , allowing for comparison of the imputed high resolution allele with the true high resolution allele determined experimentally . The second validation set , intended to confirm accuracy in haplotype phase assignment , consisted of a cohort of 4 , 235 cord blood units where the cord mother was also HLA typed , and therefore haplotype phase was known by pedigree and could be compared with the results of imputation . We computed frequency distributions for two-locus and three-locus haplotypes by summing over the frequencies at the other loci ( i . e . produce the marginal distributions ) . For example , to compute the two-locus A~B haplotype frequencies , we merged all extended A~C~B~DRB1~DQB1 haplotypes with a given A~B combination into a single reduced A~B haplotype . We computed haplotype distributions for five different combinations of loci ( A~B , A~C , B~C , A~B~C , DRB1~DQB1 ) in this manner . In the Ewens-Watterson test , one calculates the hypothetical F1 homozygosity following random mating ( the sum of squared allele frequencies ) of the sample , and compares it to the expected homozygosity of a sample with the same attributes ( sample size and total number of alleles ) from the Ewens sampling formula . The expected homozygosity and the p-value of the test are usually obtained using Slatkin's method [54] . The test parameters are limited by the numerical calculation of Ewens' sampling formula so that the maximal number of alleles in a sample is currently limited to a thousand and the maximal sample size to a few thousand . In order to perform such tests for extremely large samples such as ours , a representative subsample must be taken . We randomly sampled 1 , 200 haplotypes from each US subpopulation and performed the EW test on the subsample , using either the Arlequin [55] or PyPop [56] software packages with similar results . We calculated Fnd values for each sample ( Fnd = ( Fobs − Fexp ) /Var ( Fexp ) 1/2 ) , where Fobs is the observed homozygosity of the sample , and Fexp is the expected homozygosity for a population with the same parameters calculated using random samples , which also provided an estimate for the variance . We fit the Yule model and a Birth , Death and Innovation Model ( BDIM ) detailed in Table 1 to the relative frequency distributions of haplotype/alleles populations by maximum-likelihood estimation , using a global optimization algorithm for the numerical maximization [57] . The normalizing constant C was determined by the equality:C=∑i=1Nmaxp ( i ) , where Nmax is the absolute frequency of the most abundant haplotype . This normalization is equivalent to fitting the model conditioned on the event that the maximal number of haplotypes is Nmax . We computed the linkage disequilibrium using the normalized approach proposed by Lewontin [58] . In short , the value of Dij' for each pair of alleles is normalized by the theoretical maximum for the observed allele frequencies . The Dij' value was computed for each pair of alleles i and j ( e . g . given HLA A and HLA-B alleles ) , and binned across all haplotypes with similar frequency . Note that in this context a haplotype is either treated as a pair of alleles , or as a pair of haplotypes treated as alleles ( e . g . a class I haplotype and a class II haplotype ) . A two-sided T-Test was used for comparing Fnd values and Dij' values for each population size bin to the neutral drift , and also testing evolutionary model fit . For the Fnd values , we treated each sample as an independent observation . In the Dij' analysis , each haplotype was treated as an independent sample . For the evolutionary model fit , each population was treated as an independent sample . Where relevant ( e . g . where values were computed separately for each population ) a Benjamini—Hochberg procedure to adjust for multiple tests using false discovery rate ( FDR ) was performed . | The adaptive immune system presents antigens derived from pathogenic and normal self proteins on the cell surface using human leukocyte antigen ( HLA ) molecules . The HLA loci coding for these molecules are found in major histocompatibility complex ( MHC ) region , the most polymorphic region in the human genome , with over 15 , 000 HLA alleles observed so far in the world population . A high frequency of many different HLA alleles is thought be sustained by balancing selection . New HLA alleles may have an advantage over existing frequent alleles since immune escape mutations in pathogens within a population are maintained primarily in epitopes presented on frequent HLA alleles . Host immune function is not determined by single HLA alleles , but by both copies of autosomal HLA genes together ( genotypes ) . Complementarity in function across the two potentially-variant copies of HLA at each locus can result in overdominance and heterozygote advantage at the genotype level . Less explored are selection mechanisms that may be operating across combinations of HLA alleles across loci ( haplotypes ) . Indeed , in addition to high allelic diversity , HLA also has distinctive patterns of haplotype diversity , as certain HLA alleles co-occur in high linkage disequilibrium across five classical HLA loci ( HLA-A , -B , -C , -DRB1 , -DQB1 ) . We applied multiple population genetic models to a dataset of HLA haplotype frequencies derived from over six million individuals with the goal of determining what type of selection may impact HLA haplotype diversity . We found frequent haplotypes were preferentially maintained in the population across 23 US populations studied . Thus , balancing selection at the allele level and purifying selection at the haplotype level may together affect HLA diversity in human populations . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"population",
"genetics",
"immunology",
"alleles",
"genetic",
"mapping",
"linkage",
"disequilibrium",
"clinical",
"medicine",
"population",
"biology",
"major",
"histocompatibility",
"complex",
"homozygosity",
"genetic",
"loci",
"ha... | 2017 | HLA class I haplotype diversity is consistent with selection for frequent existing haplotypes |
Pseudomonas entomophila is an entomopathogenic bacterium that infects and kills Drosophila . P . entomophila pathogenicity is linked to its ability to cause irreversible damages to the Drosophila gut , preventing epithelium renewal and repair . Here we report the identification of a novel pore-forming toxin ( PFT ) , Monalysin , which contributes to the virulence of P . entomophila against Drosophila . Our data show that Monalysin requires N-terminal cleavage to become fully active , forms oligomers in vitro , and induces pore-formation in artificial lipid membranes . The prediction of the secondary structure of the membrane-spanning domain indicates that Monalysin is a PFT of the ß-type . The expression of Monalysin is regulated by both the GacS/GacA two-component system and the Pvf regulator , two signaling systems that control P . entomophila pathogenicity . In addition , AprA , a metallo-protease secreted by P . entomophila , can induce the rapid cleavage of pro-Monalysin into its active form . Reduced cell death is observed upon infection with a mutant deficient in Monalysin production showing that Monalysin plays a role in P . entomophila ability to induce intestinal cell damages , which is consistent with its activity as a PFT . Our study together with the well-established action of Bacillus thuringiensis Cry toxins suggests that production of PFTs is a common strategy of entomopathogens to disrupt insect gut homeostasis .
The intestinal epithelium has a role in defining the barrier between the host and the external environment [1] . This barrier protects the host against invasion and systemic dissemination of both pathogenic and commensal microorganisms . Both resistance and tolerance mechanisms contribute to maintain the gut integrity from the assault of infectious bacteria [2] . Resistance mechanisms involve the activation of various local immune responses that directly target pathogens . In contrast , tolerance mechanisms involve the activation of repair and stress pathways that quickly seal damages caused by infectious agents . Pathogenic bacteria have the capacity to overcome gut defenses and impede the return to homeostasis [3] . To study how pathogenic bacteria disrupt gut homeostasis , we chose to investigate the interactions between Drosophila and a newly identified entomopathogen , Pseudomonas entomophila . P . entomophila is closely related to the saprophytic soil bacterium Pseudomonas putida [4] , [5] . It was originally isolated from a fly sampled in Guadeloupe and subsequently shown to be lethal to Drosophila larvae and adults after ingestion . P . entomophila can also effectively kill members of other insect orders ( e . g . Bombyx mori , Anopheles gambiae , Galleria mellonella ) . After ingestion , P . entomophila is able to persist in the Drosophila gut . It induces the expression of antimicrobial peptide genes via the Imd pathway , both locally in the intestinal epithelium and systemically in the fat body , an organ analog to the mammalian liver [4] . It was shown that P . entomophila virulence is under the control of two global regulatory systems: the well known GacS/GacA two component system , and a second system involving a secreted secondary metabolite synthesized by the pvf gene products [4] , [6] . The Gac system also controls the production of a secreted protease , AprA , which is important for P . entomophila to counteract the local immune response of Drosophila [7] . Recent studies revealed that upon bacterial infection , homeostasis in the gut is restored only when bacterial clearance is coordinated with the repair of infection-induced damage through epithelium renewal [8]–[10] . Epithelium renewal of the Drosophila gut is stimulated by the release of the secreted ligand Upd3 from damaged enterocytes , which then activates the JAK/STAT pathway in intestinal stem cells to promote both their division and differentiation , establishing a homeostatic regulatory loop [8] , [9] . In contrast to infection with non-lethal bacteria , P . entomophila infection inflicts strong damage to its host without triggering an epithelial renewal [8] , [11] . This suggests that the damages inflicted by P . entomophila are too severe to be repaired . How damages are inflicted however remains unknown . One hypothesis was that P . entomophila produces cytotoxic factors that damage the intestinal epithelium . In this study , we identified a secreted protein that plays an important role in the damage inflicted by P . entomophila to the Drosophila gut . We showed that this protein is a pore-forming toxin ( PFT ) that we called Monalysin . Our work indicates that production of PFTs is a strategy used by entomopathogenic bacteria to disrupt gut homeostasis .
We previously showed that P . entomophila secretes large amount of the metalloprotease , AprA , which can degrade antimicrobial peptides [7] . The production of this protease is regulated by the GacS-GasA system , known to control secondary metabolite production , protein secretion , and virulence determinants in γ-proteobacteria [12] . To identify additional factors responsible for P . entomophila virulence , we analyzed the culture supernatant of the wild-type bacterium and a gacA mutant by SDS-PAGE ( Figure 1A ) . Bands corresponding to major secreted proteins in the wild type strain , but not in the gacA mutant were submitted to analysis by mass spectrometry . This allowed us to confirm that one of the major bands corresponds to the 51 kDa AprA . Three bands contained Hcp , Vgr and Rhs , proteins known to be secreted by the Type VI Secretion System ( T6SS ) . T6SS are bacterial needle-like structure involved in the injection of effectors into the cytoplasm of eukaryotic but also prokaryotic cells [13] . We also identified a band with an apparent molecular weight of 30 kDa , containing a protein encoded by the uncharacterized gene pseen3174 that we named Monalysin . In order to investigate the role of these secreted proteins , we made a T6SS mutant ( affecting the ORF pseen0535 ) and a monalysin ( mnl ) mutant ( Δmnl ) , and tested their virulence in Drosophila . No difference could be observed between the wild type strain and the T6SS mutant . Interestingly , the mnl mutant presented a decreased pathogenicity . Indeed , survival analysis of Drosophila adults after oral infection with the wild-type strain , the gacA mutant , and the mnl mutant showed that only 40% of the flies infected with the mnl mutant succumbed within 3 days , while 70% of the flies died after infection with wild-type P . entomophila ( Figure 1B and Figure S6 ) . As previously shown [4] , a gacA deficient mutant did not show any pathogenicity using this assay . The attenuated virulence of the mnl mutant was fully rescued by complementation with a wild-type copy of the monalysin gene . It was previously shown that P . entomophila virulence towards Drosophila is associated to its ability to persist in the gut and the transcription of antibacterial peptide genes both locally and systematically [4] . In order to better characterize the role of the monalysin gene in the infectious process , we next compared the ability of the mnl mutant ( Δmnl ) to persist to that of the wild type strain or a gacA mutant . Flies were infected by feeding and bacterial loads were quantified at two time points ( Figure 1C ) . While bacterial loads were indistinguishable after 3 hrs , persistence of the mnl mutant and the gacA mutant were significantly decreased when compared to wild type bacteria [4] . We then compared the activation of the Imd pathway after infection by the wild type , the gacA , and the mnl mutant . We used reverse transcriptase quantitative PCR ( RT-qPCR ) to measure the expression of the Diptericin gene , a target of the Imd pathway , specifically in the gut ( local response ) or in whole flies ( reflecting mostly the systemic expression of Diptericin by the fat body ) ( Figure 1D and E ) . As previously shown [7] , Diptericin expression increased already 4 h after infection by P . entomophila and even more after 16 h , both in the gut and the fat body , an increase that was not observed for the gacA mutant [4] . The mnl mutant leads to an increase in Diptericin expression in the gut similar to that observed for the wild-type bacterium ( Figure 1D ) . However , while Diptericin expression increased to wild-type levels in the fat body 4 h after infection , no further increase was observed ( 16 h ) in flies infected with the mnl mutant ( Figure 1E ) . We next investigated the contribution of Monalysin in the damage caused by P . entomophila to the Drosophila gut . We first monitor the induction of cell death upon bacterial ingestion using an acridine orange staining . A high number of dead cells were detected in guts from flies infected by wild type P . entomophila , but not in guts from flies infected by a gacA mutant as previously reported [6] . Interestingly , a reduced level of cell death was observed in the mnl mutant ( Figure 1F and S7 ) . Oral infection with P . entomophila resulted in a decrease of the adherens junction marker Cadherin-GFP ( Figure 2A ) and to morphologically altered guts , with regions devoid of enterocytes indicative of a disruption of tissue integrity ( see the lack of nuclear DAPI staining due to the loss of cell in Figure 2A3 ) . Interestingly , gut collected 16 hrs after oral infection with gacA and mnl mutants did not show any Cadherin-GFP signal decreases or a rupture of the gut integrity ( Figure 2A4 to 2A6 ) . Previous studies showed that ingestion of P . entomophila activates both JAK-STAT and the Jun N-terminal kinase ( JNK ) pathway in the Drosophila gut [8] that participate in the repair and stress responses , respectively [8] , [14]–[16] . The activation of both pathways can be monitored by measuring by RT-qPCR the expression of puckered ( puc ) ( a direct downstream target of JNK signaling ) or upd3 ( a target of JAK-STAT signaling ) and Socs36E ( a target of JAK-STAT signaling that encodes a negative regulator of this pathway ) . Figure 2B , 2C and 2D shows that the mnl mutant was less efficient than wild type P . entomophila to activate the JNK and JAK-STAT pathways , yet more efficient than a gacA mutant . Consistent with the RT-qPCR analysis , expression of the upd3-GFP reporter gene ( upd3-Gal4 , UAS-GFP ) was strongly induced in the gut of flies orally infected with a sub-lethal dose of P . entomophila but not the mnl mutant ( Figure 2E ) . Altogether , these data show that even though the mnl mutant retains some ability to cause intestinal damage , this ability is strongly diminished compared to wild type P . entomophila . This suggested a specific role of the Monalysin protein in P . entomophila cytotoxicity towards Drosophila . In order to characterize the activity of the secreted protein encoded by monalysin , we produced and purified a his-tag version of it in E . coli . Ingestion of the recombinant protein at high dose had no impact on fly survival . However , the Monalysin protein was highly toxic when directly injected in the body cavity ( Figure 3A ) . This dose-dependent lethal activity suggests that Monalysin function as a bacterial toxin . Results shown in Figures 3B , 3C and Table S2 indicate that Monalysin has a strong cytotoxicity towards S2 cells ( derived from Drosophila embryonic hemocytes ) and SF9 cells ( from the Lepidoptera Spodoptera frugiperda ) . Moreover , Monalysin treated S2 cells showed DNA fragmentation and condensation that are sign of apoptosis ( Figure 3D and E ) . Along the same line , Figure 4A shows that the recombinant toxin rapidly induced hemolysis in a dose dependant manner as measured by the loss of sample turbidity . In addition we found that two mammalian culture cell lines – Hela and RPE1– were also sensitive to Monalysin ( Figure 4B and Table S2 ) . Altogether these observations show that Monalysin is a secreted cytotoxic factor of P . entomophila with a broad range of activity . P . entomophila virulence is controlled by several regulatory systems: i ) the two component system GacS/GacA that functions at the post-transcriptional level , ii ) a secreted signaling molecule produced by the pvf genes , and iii ) AlgR that is known to control alginate production in other bacteria [5] , [6] , [7] . To determine which of these mechanisms regulate Monalysin production , crude cell extracts or filtered supernatants from wild-type and mutant P . entomophila were analyzed using a specific antiserum ( Figure 5A , Figure S5 ) . Monalysin was undetectable in both cells and medium of P . entomophila lacking the two-component system GacS/GacA and the pvf signaling molecule . Interestingly , this analysis revealed that the supernatant of P . entomophila contained a shorter form of Monalysin when compared to the form detected in cell extracts . A N-terminal Edman sequencing of the shorter form found in culture supernatant ( extracted from SDS gels ) was performed , which revealed that the size shift was due to a cleavage taking place before Asparagine 34 ( indicated in Figure 6A ) . Many toxins require a proteolytic activation , which can be performed by proteases produced by the bacterium itself or by enzymes of the host digestive tract [17] . Interestingly P . entomophila secretes large amounts of the metallo-protease AprA . To test whether AprA could be responsible for maturation of pro-Monalysin to Monalysin , we analyzed the supernatants of AprA-deficient and wild-type P . entomophila by Western blotting . Figure 5B shows that pro-Monalysin was found in supernatant derived from the aprA mutant while the mature form predominates in supernatant from wild-type P . entomophila . Collectively , our data indicate that Monalysin production is controlled by the global regulatory systems Gac and Pvf and that its N-terminus is cleaved by AprA upon secretion into the extracellular medium . The Monalysin amino acid sequence does not show any homology to other sequences using P Blast , except for two uncharacterized orthologs found in Pseudomonas putida F1 strain ( Figure S1 ) . Neither the P . entomophila nor the P . putida gene products displayed any obvious protein domains . Nevertheless , the use of the HHpred software ( Homology detection & structure prediction by HMM-HMM comparison ) revealed the presence of an internal region with alternating polar and hydrophobic residues flanked by stretch of serine- and threonine residues , a hallmark of the membrane-spanning region of ß-barrel pore-forming toxins ( Figure 6A ) . PFTs can be classified according to the secondary structure of their membrane-spanning region as α- and ß-PFTs . Far-UV circular dichroism analysis of Monalysin revealed a spectrum typical of structured proteins ( Figure S2 ) . The content of α-helixes and β-sheets was estimated to be 13% and 40% , respectively in agreement with the secondary structure prediction obtained with the program JPRED giving 17% of α-helixes and 35% of β-sheets as indicated in Figure 6A . This program also indicated that the putative membrane-spanning region of Monalysin was formed of a ß-sheet . This sequence analysis suggests that Monalysin is related to PFT of the ß-type . ß-PFTs are synthesized as soluble proteins and have the ability to multimerize into circular polymers at high concentration , a step that for certain toxins , such as Aerolysin , requires proteolytic activation [18] . We next investigated whether Monalysin shared these properties with PFTs . SDS-PAGE analysis of a fresh recombinant Monalysin solution revealed a major band at the expected size ( 30 kDa ) as well as several high molecular weight bands corresponding to oligomers that were resistant to SDS ( see below ) . Interestingly , a shorter form of the protein was observed upon storage of samples at 4°C ( Figure S3A ) . This together with the observation that Monalysin is matured by AprA indicates the existence of a protease sensitive site in the N-terminus part of Monalysin ( Figure 5B ) . The cleavage of the recombinant pro-Monalysin into its shorter form could also be induced by a limited trypsinolysis ( Figure S3B ) . This processed form has a molecular weight of 26 . 5 kDa as determined by MALDI-TOF analysis , as opposed to 30 . 2 kDa for the full-length pro-toxin . Interestingly , Monalysin had stronger hemolytic activity than pro-monalysin ( Figure 4A ) , indicating that the removal of the N-terminal fragment constitutes a maturation step that enhances the cytotoxic activity . Processing of pro-monalysin to its mature form was accompanied by an increase and change in the higher order SDS-resistant complexes ( Figure S3B ) . While multiple size oligomers were observed by SDS-PAGE , a single species was observed by native PAGE analysis of Monalysin ( Figure S3C ) . Multi-Angle Light Scattering analysis ( MALS/UV/RI ) confirmed the presence of a single species with a molecular mass of 546 kDa and hydrodynamic radius of 7 . 5 nm hence a diameter of 15 nm , which would correspond to about 18 monomers ( Figure S4 ) . This was further confirmed by electron microscopy of negatively stained recombinant Monalysin which showed circular like ( top view ) and barrel like structures ( side view ) similar to that observed with other ß-PFT ( Figure 6B ) . Sequence analysis of Monalysin , its ability to form ring like high order structures combined with its hemolytic activity strongly indicate that the toxin is a PFT . To address this issue directly , we analyzed its ability to form channels in planar lipid bilayers , an extremely sensitive electrophysiological method that enables the study of single-channel events . Addition of Monalysin led to a stepwise increase in membrane current , reflecting the formation of pores ( Figure 6C ) . Collectively , our data show that P . entomophila Monalysin is a bona fide pore-forming toxin of the ß type . Many bacterial pathogens , both Gram-positive and Gram-negative , produce PFTs that contribute to their virulence [17] . Here we report the identification of a novel PFT that contributes to the virulence of P . entomophila against Drosophila . Our data show that Monalysin requires N-terminal cleavage to become fully active , forms oligomers in vitro , and induces pore formation in artificial lipid membranes . The prediction of the secondary structure of the membrane-spanning domain indicates that Monalysin is a PFT of the ß-type . Outside of this domain , Monalysin does not show any homology to any other PFT and appears rather different from previously identified insecticidal PFTs such as B . thuringiensis Cry toxins . Nevertheless , Monalysin has two homologs in the closely related P . putida F1 strain . These proteins could participate to the interaction of some Pseudomonas species with eukaryotic cells , defining a new family of PFTs . We previously showed that P . entomophila virulence is multi-factorial and regulated by multiple signaling modules . Taking advantage of the genetic amenability of both the host and the pathogen , we aimed to identify P . entomophila and Drosophila pathways and effectors involved in the infectious process . Using this integrated approach , we previously proposed a role for the AprA metalloprotease in protection against antimicrobial peptides [7] . We now identify a second virulence factor , the ß-PFT Monalysin . Like AprA , a mnl mutant is affected in several , but not all , aspects of P . entomophila virulence . This attenuated virulence of the mnl mutant is clearly shown by survival analysis , which monitors the global outcome of infection . Our study indicates that Monalysin significantly contributes to the damage inflicted to intestinal cells by the bacterium , which is fully consistent with its activity as a pore-forming toxin . Supporting this notion , we observe that a mnl-deficient mutant induced less cell damage and a lower level of stress and repair pathway activity . The mnl mutant still induced a local immune response but the systemic immune response is drastically attenuated . This is also consistent with a role of Monalysin as a cytotoxin since activation of a systemic immune response is probably linked to damage of intestinal tract rendering possible the translocation of peptidoglycan , the bacteria elicitor activating the Imd pathway , from the lumen to the hemolymph compartment . We also show that Monalysin production is regulated both by the GacS/GacA two component-system and the pvf genes . However , a mnl mutant still causes higher levels of stress and damage to the intestinal epithelium than gacA or pvf mutants . This indicates that these signaling modules regulate additional virulence factors contributing to P . entomophila cytotoxicity . Alternatively , it is possible that the overall cytotoxicity is caused by a synergy between the metalloprotease AprA and the PFT Monalysin , both of them being regulated by GacA . Along this line , we observed that AprA promotes the rapid cleavage of the pro-Monalysin into its active form . Since Monalysin can also be processed by trypsin , it is likely that AprA is not essential for PFT function of Monalysin in the Drosophila gut , as this toxin could also be processed by host enzymes . Both Monalysin and AprA are expressed in the algR mutant that affects a transcriptional regulator involved in alginate production as well as genes that are often associated to virulence ( ie: pili biosynthesis , cyanide production… ) [19] . The observation that an algR mutant is still avirulent ( Vodovar 2005 ) , although expressing both Monalysin and AprA , indicates the existence of additional virulence factors . Future studies should investigate at which level Pvf and GacS/GacA affect Monalysin production as well as identify other potential virulence factors regulated by the Pvf , Gac or AlgR . Recent studies have shown that cells respond to PFTs by inducing repair and stress signal-transduction pathways to repair damage . Studies in C . elegans and mammalian cells have revealed a role for the P38 pathway , the unfolded protein response , and hypoxia in cellular resistance to the action of PFT [20]–[22] . The reduced expression of JAK-STAT and JNK pathway activities in guts infected with the mnl mutant indicate that Drosophila epithelial cells respond to PFT by activating stress and repair pathways . Thus , the P . entomophila/Drosophila interaction provides an interesting model to dissect the host response to PFTs in a natural infectious context . Insects are potential reservoirs for microbes and are ideal vectors for their transmission due to their motility and their capacity to live in bacteria-rich environments [23] . This is exemplified by fruit flies that live in rotting fruits and are capable of transmitting phytopathogenic bacteria [24] . Insects are notably resistant to microbial infection allowing them to colonize these microbe-rich environments . This is largely due to the existence of very efficient physical barriers that block entry of microbes in the body cavity . As an illustration , injection of less than 10 cells of P . aeruginosa or Serratia marcescens in the body cavity rapidly kills flies , while high doses of these bacteria have only modest effects on survival when ingested [25] . In contrast to mammals , the gut of insects is lined with a chitinous matrix , the peritrophic matrix [26] , that blocks the direct interaction between bacteria and epithelia cells and prevents the use of virulence devices such as type III and VI secretion systems that allow the injection of virulence factors directly into target cells . Rare bacterial species such as Photorhabdus luminescens can bypass this physical barrier since there are transported by symbiotic nematodes that can pierce the insect cuticle [27] . Other entomopathogens that enter through the oral route have to escape the local immune response and breach the gut barrier [23] . Despite the characterization of several virulence factors in few species , the mechanisms by which enteric pathogens kill insects remain poorly understood . This paper together with the well-characterized action of Bacillus thuringiensis cytotoxin Cry suggests that PFTs efficiently promote bacterial colonization of the insect gut [28]–[30] . This heavy artillery strategy does not require a direct contact between bacteria and host cells since PFTs can cross the pores of the peritrophic matrix and reach intestinal cells . PFTs can induce gut damage and rupture of intestinal homeostasis that in fine will lead to a weakening of the gut barrier and an inhibition of gut peristaltism promoting bacteria persistence . Gut damage and food uptake blockage are two symptoms of insect pathogenesis and could reflect the action of PFTs [23] . It would be interesting to know if other entomopathogens such as Serratia marcescens and Serratia entomophila also used PFT to colonize their insect host . In conclusion , this and other studies using different bacteria species contribute to uncovering strategies used by entomopathogens to breach insect barriers . A better knowledge of these strategies could also open the route to new methods of insect pests control .
P . entomophila L48 [4] was grown in LB for all experiments . P . entomophila mutated for the gacA , aprA , algR , and the pvf gene are described elsewhere [4] , [5] , [6] , [7] , [11] . The mnl deletion construct was generated by amplifying flanking regions of the monalysin gene ( pseen3174 or mnl ) by PCR . The resulting PCR product was cloned into the plasmid pEXG2 . This plasmid was then used to create the strain Δ3174 ( alternatively Δmnl ) , containing a deletion of the gene pseen3174 . Complementation construct were made by cloning into the plasmid pPSV35 of PCR-amplified DNA fragments from P . entomophila containing the mutated genes . Pseudomonas Isolation agar ( PIA , Difco ) was used for selection after conjugations and persistence experiments . When E . coli was grown , antibiotics were used when necessary at the following concentrations: G418 , 25 µg/ml and tetracycline , 5 µg/ml . When P . entomophila was grown , antibiotics were used when necessary at the following concentrations: gentamicin , 50 µg/ml for liquid cultures and 150 µg/ml for solid media , tetracycline 40 µg/ml and rifampicin , 30 µg/ml . The bacterial strains used in this study and the culture conditions are presented in Table S1 . All primer sequences are available upon request . Insertion constructs were generated as previously described [6] , [11] . DNA sequence searches and analysis were performed using the Pseudomonas genome database ( www . pseudomonas . com ) . The monalysin gene ( ORF PSEEN3174 ) corresponds to the accession number YP_608728 . 1 . Monalysin putative orthologs in Pseudomonas putida Pput_1063 and Pput_1064 correspond to the accessions numbers YP_001266408 . 1 , YP_001266409 . 1 respectively . The ORF PSEEN0535 involved in the production of the type VI secretion system corresponds to the accession number YP_606298 . 1 Monalysin amino-acids sequence analysis was performed using the HHpred software ( Homology detection & structure prediction by HMM-HMM comparison http://toolkit . tuebingen . mpg . de/hhpred ) . Oregon R flies were used as a standard wild-type strain and were maintained at 25°C . Adherens junctions were visualized using ubi-DE-cadherin-GFP flies [14] , [31] . Upd3 expression in unchallenged gut and following infection , was monitored using upd3-Gal4 , UAS-GFP flies ( Buchon et al . , 2009 ) . Fly natural infections were carried out at 29°C on 4- to 8- day-old adult females as previously described . All the infections , except when specified , were carried out with bacterial preparation adjusted to an OD600 = 100 which correspond to 6 . 5E10 colony forming units per ml [11] . Monalysin was injected in the body cavity of fly using a Nanodrop microinjector ( Nanoject ) . Virulence assays were performed at least three times in triplicate . Total RNA was extracted from whole flies ( 5 for each assay ) or from dissected guts without Malpighian tubules ( 14 for each assay ) using TRIzol ( Invitrogen ) . RT-qPCR was performed using SYBR Green I ( Roche ) on a Lightcycler 2 . 0 ( Roche ) as previously described [32] . Data represent ratio of the amount of mRNA detected normalized to the amount of the control rpl32 mRNA . Experiments were performed at least three times independently . Averages of more than three experiments are shown . The macrophage-like lineage S2 cells derived from D . melanogaster embryos where grown in Schneider's medium ( Invitrogen ) . The Sf9 cells ( Invitrogen ) derived from Spodoptera frugiperda ( Lepidoptera ) pupal ovarian tissue were cultured in complete TNM-FH ( Invitrogen ) . The mammalian cell lines Hela and the Retinal Pigmented Epithelial ( RPE1 ) were grown in a humidified incubator with 5% CO2 at 37°C . Hela cells were cultured in MEM media supplemented with 10% fetal calf serum , 1% penicillin-streptomycin , 1% glutamine and 1% NEAA ( Gibco ) . RPE1 cells were cultured in DMEM media supplemented with 10% fetal calf serum , 1% penicillin-streptomycin and 1% glutamine ( Gibco ) . Cell viability was observed using the LIVE/DEAD Viability/Cytotoxicity Assay Kit ( Invitrogen ) according to the provider instruction . Briefly cells are simultaneously labeled with calcein AM that reveals intracellular esterase activity in live cells and ethidium homodimer ( EthD-1 ) that reveals plasma membrane damages . LDH release from damaged cells was measured following the instructions of the CytoTox-One Homogeneous Membrane integrity Assay kit ( Promega ) . In Situ Nick Translation was performed to detect fragmented DNA in nuclei . In situ DNA synthesis was performed by a DNA polymerase I ( 150 units/ml ) ( Takara ) in the presence of a dNTP mix in which dUTP is tetramethylrhodamine-conjugated ( Roche ) . The reaction was carried out for 90 min at room temperature . Live imaging and immunofluorescence were performed as previously described [8] . After treatment , cells were recovered , fixed with 4% PFA and permeabilized with 0 . 3% Triton X-100 . Dead cells were detected using acridine orange staining ( Invitrogen ) . Dead cells quantification was performed as follows: 16 hours after infection , guts were dissected and stained with acridine orange and DAPI . Pictures were taken using a fluorescent microscope . From these pictures , groups of 100 hundred DAPI stained nuclei were randomly defined and the number of acridine orange positive nuclei ( ie dead cells ) was determined . Three parcels per guts were analyzed . The results are the mean of four independent experiments . Nuclei were stained by DAPI ( Sigma ) . All the images were performed using a Zeiss Axioimager Z1 . All cloning steps were performed as described earlier [33] . The sequence of Monalysin ( from residue 1 to 271 , access number pseen3174 ) was PCR-amplified from genomic DNA ( isolated from P . entomophila ) and cloned into pDONR201 ( Invitrogen ) . The ORF was then subcloned into the pETG-20A E . coli ( a generous gift from Dr A . Geerlof , EMBL ) destination vector to generate a constructs encoding Monalysin with an N-terminal fusion composed of the thioredoxin ( TRX ) protein , followed by a 6xHis-tag and a Tobacco Etch Virus ( TEV ) protease cleavage site . The construct was sequenced verified . The production and purification were performed as described earlier [34] . Briefly , the pETG-20A-Monalysin was transformed into Rosetta ( DE3 ) pLysS E . coli cells ( Novagen ) . An overnight LB pre-culture ( with 100 mg mL-1 ampicillin and 34 mg mL-1 chloramphenicol ) was used to inoculate large cultures in ZYP-5052 auto-induction media [35] supplemented with the same antibiotics and incubated with vigorous shaking ( 250 rpm ) at 37°C during 4 h . At this stage , the temperature was decreased to 17°C , and the cultures were allowed to grow for an additional 18 h with vigorous shaking ( 250 rpm ) . After 18 h , cells were harvested by centrifugation ( 4000g for 10 min ) and the pellet was homogenized and frozen in lysis buffer ( 50 mM Tris-HCl; 500 mM NaCl; 0 . 5 mM lysozyme; 10 mM imidazole and 1 mM phenylmethylsulfonyl fluoride ( PMSF ) , pH 8 ) . The cell pellets were thawed and lysed with a sonicator after the addition of DNase I at 20 mg mL−1 and 1 mM MgSO4 . The pellet and soluble fraction were separated by centrifugation ( 30 min at 16 , 000g ) of an early stationary phase culture . The supernatants were filtered through a 0 . 22 µm filter and were concentrated 50-fold by using 5 kDa cutoff Centricon membranes ( Biorad ) . The cell pellets were washed in PBS , resuspended in PBS containing protease inhibitors and lysed by sonication . The soluble fraction was purified by immobilized metal ion affinity chromatography using a 5 mL HisTrap crude ( GE Healthcare ) Ni2+-chelating column equilibrated in buffer A ( 500 mM NaCl; 50 mM Tris–HCl; 10 mM imidazole; pH 8 ) . After the loading of the soluble fraction and a column wash ( buffer A with 50 mM Imidazole ) , the protein was eluted with buffer A supplemented with 250 mM imidazole . The eluted fraction was desalted in buffer A ( Hiprep 25/10 Desalting column , GE ) and the protein concentration of the TRX-His6-TEV-Monalysin determined . After a 4°C overnight cleavage of the protein with 1∶20 w:w His-TEV protease , the TRX-His6-TEV and the His-TEV were separated from the pure Pro-Monalysin by collecting the Flow Through ( FT ) of a second Nickel purification . The final purification and the characterization of the oligomeric state of the monalysin were achieved by the separation of the FT on a size exclusion chromatography ( HiLoad 16/60 Superdex 200 prep grade , GE ) , equilibrated in Tris 10mM , NaCl 500mM pH8 . The pure Pro-Monalysin was used for the functional characterizations . For the MultiAngle Light Scattering analysis , size exclusion chromatography was carried out on an Alliance 2695 HPLC system ( Waters ) using a Silica Gel KW804 column ( Shodex ) equilibrated in 10 mM Tris and 150 mM NaCl at pH 7 . 5 at a flow of 0 . 5 ml/min . Detection was performed using a triple-angle light scattering detector ( Mini-DAWN TREOS , Wyatt Technology ) , a quasi-elastic light scattering instrument ( Dynapro , Wyatt Technology ) , and a differential refractometer ( OptilabrEX , Wyatt Technology ) . Proteins were analyzed by SDS-PAGE . Native PAGE was performed to determine the oligomeric state of Monalysin . To generate Monalysin , pro-monalysin samples were submitted to limited trypsinolysis by adding trypsin ( 1∶100 w:w ) . The reaction was stopped by using a trypsin-chymotrypsin inhibitor ( Invitrogen ) . Pro-monalysin and Monalysin were detected by Western-blot using a specific serum that recognized better the pro-monalysin than the Monalysin ( see Figure S5 ) . The Guinea pig antibody anti-Monalysin was provided by Eurogentec . Far-UV Circular Dichroism ( CD ) spectra ( Figure S2 ) were recorded with a JASCO J-810 spectropolarimeter ( JASCO Corporation ) equipped with a Peltier temperature control and using 1-mm thick quartz cells . The molecular weight of recombinant pro-monalysin and Monalysin was determined by MALDI-TOF/TOF . Molecular weight and hydrodynamic radius determination was performed by the ASTRA V software ( Wyatt Technology ) . Proteins were loaded at a final concentration of 0 . 02 mM . After SDS-PAGE electrophoresis and Coomassie blue staining , protein bands were excised . Proteins were extracted from gel and blotted onto polyvinylidene difluoride membranes with the ProSob system ( Applied Biosystems ) . The N-terminal sequences of proteins were determined by automated Edman degradation by introducing the blots into a Procise P494 automated protein sequencer ( Applied Biosystems ) . The sequences obtained were compared to sequences in public protein sequence databases . Planar lipid bilayer experiments were performed as previously described [36] . The bilayer was formed by painting a solution of 50% PC ( egg lecithin ) / 50% DOPE ( w:w ) in n -decane ( 40 mg ml −1 ) on an aperture ( d = 150 µm , pretreated with the same solution ) in a delrin cuvette separating two chambers , each containing 1 ml of 1 M NaCl , 5 mM CaCl2 10 mM HEPES , pH 7 and agar bridge connection ( 1 M KCl ) to Ag/AgCl electrodes ( Warner Instrument Corp . Hamden , CT ) . Monalysin was added to the cis chamber at room temperature . Survival assays have been performed at least three times in triplicate . The Kaplan-Meier log rank test was used to determine statistical significance . Dashed brackets represent the significance between the different infections ( *: p<0 . 05 , **: p<0 . 01 , ***: p<0 . 001 , ns = not significant ) . RT-qPCR analysis and cell death quantification using acridine orange staining are averages of at least 4 independent experiments . Error bars indicate standard errors . Statistical analysis was performed using a Wilcoxon test , and letters indicate significantly different values ( P<0 . 05 ) . | Insects are potential reservoirs for microbes and ideal vectors for their transmission due to their motility and capacity to live in bacteria-rich environments . This is exemplified by fruit flies that live in rotting fruits and are capable of transmitting phytopathogenic bacteria . Insects are notably resistant to microbial infection allowing them to colonize these microbe-rich environments . To study how pathogenic bacteria disrupt gut homeostasis , we investigated the interactions between Drosophila and a newly identified entomopathogen , Pseudomonas entomophila . Ingestion of P . entomophila inflicts severe damage to the Drosophila intestine . How damages are inflicted , however , remains unknown . In this study , we identified a secreted protein that plays an important role in the damage inflicted by P . entomophila to the Drosophila gut . We showed that this protein is a pore-forming toxin ( PFT ) that we named Monalysin . Our study reveals that Monalysin oligomerizes into ring-like structures that form pores into the plasma membrane of target cells leading to the disruption of membrane permeability and cell death . Our work together with studies on the insecticidal Cry toxins produced by Bacillus thuringiensis suggests that production of PFTs is a common strategy of entomopathogenic bacteria to interfere with insect gut homeostasis . | [
"Abstract",
"Introduction",
"Results",
"Materials",
"and",
"Methods"
] | [
"toxicology",
"immunology",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"immune",
"response"
] | 2011 | Monalysin, a Novel ß-Pore-Forming Toxin from the Drosophila Pathogen Pseudomonas entomophila, Contributes to Host Intestinal Damage and Lethality |
Toll-like receptor ( TLR ) ligands have been explored as vaccine adjuvants for tumor and virus immunotherapy , but few TLR ligands affecting schistosoma vaccines have been characterized . Previously , we developed a partially protective DNA vaccine encoding the 26-kDa glutathione S-transferase of Schistosoma japonicum ( pVAX1-Sj26GST ) . In this study , we evaluated a TLR7/8 ligand ( R848 ) and a TLR9 ligand ( CpG oligodeoxynucleotides , or CpG ) as adjuvants for pVAX1-Sj26GST and assessed their effects on the immune system and protection against S . japonicum . We show that combining CpG and R848 with pVAX1-Sj26GST immunization significantly increases splenocyte proliferation and IgG and IgG2a levels , decreases CD4+CD25+Foxp3+ regulatory T cells ( Treg ) frequency in vivo , and enhances protection against S . japonicum . CpG and R848 inhibited Treg-mediated immunosuppression , upregulated the production of interferon ( IFN ) -γ , tumor necrosis factor ( TNF ) -α , interleukin ( IL ) -4 , IL-10 , IL-2 , and IL-6 , and decreased Foxp3 expression in vitro , which may contribute to prevent Treg suppression and conversion during vaccination and allow expansion of antigen-specific T cells against pathogens . Our data shows that selective TLR ligands can increase the protective efficacy of DNA vaccines against schistosomiasis , potentially through combined antagonism of Treg-mediated immunosuppression and conversion .
Schistosomiasis is regarded as one of the most neglected tropical diseases ( NTDs ) of high importance , and remains a major problem in public health in endemic countries [1] , [2] . Although schistosomiasis can be treated with the drug praziquantel [3] , high reinfection rates limit its overall success , where repeated administering is often necessary multiple times during the first two decades [4] , [5] . Therefore , the development of a safe and effective vaccine would improve the long-term treatment of schistosomiasis and should improve the efficacy of therapeutic interventions [6] , [7] . Despite decades of effort developing vaccines against schistosoma , including Schistosoma japonicum ( S . japonicum ) , the current schistosoma vaccine induces only limited protection for reasons that remain unclear . A potential issue limiting the immune response to vaccination is the presence of regulatory T cells ( Tregs ) that suppress T cell activation [8] , [9] . Multiple studies in mice have shown that Tregs dampen the immune response against pathogens , including S . japonicum [10] , [11] . Increased levels of Tregs have been documented in the peripheral blood of schistosoma-infected patients [12] . Furthermore , naturally occurring CD4+CD25+ Tregs as well as adaptive CD25+Foxp3+ Tregs , Tr1 cells , and Th3 cells have all been detected in schistosoma-infected mice [13] , [14] . Treg depletion improves the efficacy of vaccines against pathogens in mice [15] , [16] . Therefore , vaccine strategies that target both the innate and adaptive immune systems for the generation/upregulation of potent anti-pathogen immune responses and simultaneously overcome Treg-mediated immune inhibition are more likely to succeed . Toll-like receptors ( TLRs ) are mediators of innate immune responses that detect conserved pathogen-associated molecules . Binding of TLRs with their specific ligands induces a signaling cascade resulting in the induction of type I IFNs and other cytokines , which drive an inflammatory response and activate the adaptive immune systems [17] . As TLRs provide an important link between innate and adaptive immunity , TLR ligands are increasingly being used in the development of pathogen vaccines [18] . In addition to activating effector T cells ( Teffs ) , TLR agonists can indirectly or directly modulate the function of Treg cells . There is evidence that TLR activation can block Treg cellular responses , thereby breaking tolerance to self-antigens . For instance , the TLR9 ligand CpG can synergize with anti-CD3 to partially abrogate the suppressive activity of Tregs [19] , [20] . Synthetic and natural ligands for human TLR8 can also reverse Tregs function independently of dendritic cells ( DCs ) [21] . Given that certain TLR ligands , in particular the TLR7/8 ligand resiquimod ( R848 ) , and the TLR9 ligand CpG , can elicit a strong immune response , these ligands may be used as adjuvants during treatment of virus-infected or cancer patients with high numbers of Tregs [22] , [23] . To date , few TLR ligands affecting schistosoma vaccines have been characterized [24] . We recently developed a partially protective DNA vaccine encoding the 26-kDa glutathione S-transferase of S . japonicum ( pVAX1-Sj26GST ) and demonstrated that an abundance of CD4+CD25+ Tregs induced after pVAX1-Sj26GST vaccination may explain the limited protection conferred by this vaccine [25] . Given that R848 and CpG can elicit strong immune responses and potentially inhibit the suppression of Tregs during vaccination [23] , [26] , we tested whether combining CpG and/or R848 with the pVAX1-Sj26GST vaccine would obtain better immune efficacy against S . japomicum and assessed the impact of these TLR ligands on Treg function in vitro . Following pVAX1-Sj26GST vaccination , we found that combined use CpG of R848 significantly enhances splenocyte proliferation and IgG and IgG2a levels , increases IFN-γ and TNF-α levels in the supernatant of splenocytes , and improves immune protection against S . japonicum . The combination of CpG and R848 inhibited Treg-mediated immunosuppression , upregulated the production of IFN-γ , TNF-α , IL-4 , IL-10 , IL-2 , and IL-6 , and decreased Foxp3 expression in vitro , which collectively may contribute to prevent Treg suppression and conversion during vaccination and allow expansion of schistosome antigen-specific T cells . The combination of an S . japonicum DNA vaccine with the TLR ligands CpG and R848 thus represents a promising new approach for the design and implementation of schistosome vaccination .
Animal experiments were performed in strict accordance with the Regulations for the Administration of Affairs Concerning Experimental Animals ( 1988 . 11 . 1 ) , and all efforts were made to minimize suffering . All animal procedures were approved by the Institutional Animal Care and Use Committee ( IACUC ) of Nanjing Medical University for the use of laboratory animals ( Permit Number: NJMU 09-1107 ) . Six-week-old C57BL/6 female mice were provided by the Center of Experimental Animals ( Nanjing University , Nanjing , China ) and bred in university facilities . The experimental protocol was approved by the Institutional Animal Care and Use Committee ( IACUC ) as previously described [25] . Oncomelania hupensis harboring S . japonicum cercariae ( Chinese mainland snail strain ) were purchased from the Jiangsu Institute of Parasitic Diseases ( Wuxi , China ) . The TLR9 ligand CpG oligodeoxynucleotides ( ODN ) 1826 ( CpG; 5′-TCCATGACGTTCCTGACGTT-3′ ) , with a nuclease-resistant phosphorothioate backbone and no detectable endotoxin , was purchased from Coley Pharmaceutical Group ( Wellesley , MA , USA ) . The TLR7/8 ligand resiquimod ( R848 ) was purchased from Invivogen ( Toulouse , France ) . Soluble schistosome worm antigen ( SWA ) was prepared as previously described [25] , [27] . The construction , expression , and purification of pVAX1-Sj26GST and pVAX1 have previously been described [25] , [28] . Sj26GST sequence: GenBank accession no . M14654 . 1 ( http://www . ncbi . nlm . nih . gov/nuccore/160926 ) ; UniProtKB/Swiss-Prot accession no . P08515 . 3 ( http://www . ncbi . nlm . nih . gov/protein/P08515 . 3 ) . For in vivo experiments , all plasmids were prepared using the Qiagen Endo-Free Plasmid Kit ( Qiagen , Valencia , CA ) . The Limulus Amebocyte Lysate QCL-1000 Kit ( Cambrex , Charles City , IA , USA ) was used to confirm that the endotoxin concentrations were below 0 . 1 endotoxin units ( EU ) per dose . For the characterization of immune responses , three independent experiments were performed . In each experiment , C57BL/6 mice ( 6 mice per group ) were injected with pVAX1-Sj26GST ( 50 µg ) , with or without 25 µg CpG , 25 µg R848 , or both CpG ( 25 µg ) and R848 ( 25 µg ) . The DNA vaccine was delivered intramuscularly into the left tibialis anterior muscle in a total volume of 50 µL PBS . CpG and/or R848 were injected subcutaneously in 100 µL PBS at the base of the tail . As negative controls , mice were treated with pVAX1 , R848 , or CpG only . The immunization was repeated 3 times at 14-day intervals . One week after the final vaccination , mice were sacrificed for characterization of cellular and humoral immune responses . CpG and R848 dosing and subcutaneous injection method were based on previous publications [29] , [30] . For the vaccination challenge trial , 2 independent experiments were carried out . In each experiment , C57BL/6 mice were divided into 7 groups of 8 mice each . Each mouse was injected with pVAX1-Sj26GST ( 50 µg ) , with or without CpG ( 25 µg ) , R848 ( 25 µg ) , or both CpG ( 25 µg ) and R848 ( 25 µg ) as above . Immunization was repeated 3 times at 14-day intervals . Two weeks after the final vaccination , all mice from each group were challenged percutaneously with 40±1 S . japonicum cercariae . After 6 weeks , mice were sacrificed and perfused to determine the adult worm burdens and liver egg burdens . Reductions in worms/liver egg burdens are expressed as a percentage of the burden recorded in the pVAX1 control group . For antibody detection , serum samples were collected 7 days after the last immunization . Standard ELISAs were performed using SWA as the antigen source , which was prepared as previously described [25] , [27] , [28] . Antibody detection in the sera of immunized mice was performed as previously described [25] , [31] . In brief , ELISA plates ( Titertek Immuno Assay-Plate , ICN Biomedicals Inc . , Costa Mesa , CA , USA ) were coated with SWA ( 15 µg/mL ) in 50 mM carbonate buffer ( pH 9 . 6 ) and stored overnight at 4°C . Plates were washed and developed using tetramethylbenzidine ( TMB ) substrate ( BD Biosciences Pharmigen , San Diego , CA ) . The enzymatic reaction was stopped with 1N H2SO4 and plates were read at a 450-nm wavelength . To analyze IgG , IgG1 , IgG2a , and IgM , mouse-specific secondary antibodies ( Bio-Rad , Hercules , CA , USA ) were used at a dilution of 1∶1000 . All samples were assayed in triplicate . To determine the titers of antibodies after the last immunization , the sera from mouse within a group were pooled , serially diluted , and analyzed by ELISA as described above . All samples were assayed in triplicate . End-point titers were defined as the highest plasma dilution that resulted in an absorbance value ( OD 450 nm ) two times greater than that of non-immune plasma with a cut-off value of 0 . 05 . [3H] thymidine ( 3H-TdR ) incorporation was used to measure splenocyte proliferation . Seven days after the last immunization , 6 mice from each group were sacrificed and splenocytes were harvested . In 96-well plates , 2×105 cells per well were incubated for 72 h in 200 µL of complete media in the presence of SWA ( 15 µg/mL ) . After 56 h in culture , 0 . 5 µCi [3H] thymidine ( Amersham , Burkinghamshire , UK ) was added to each well . At the end of the incubation period , the cells were harvested on filters and the incorporated [3H] thymidine was counted . To evaluate cytokine production , single-cell suspensions of splenocytes were cultured in the presence of 15 µg/mL SWA or control medium at 2×105 cells/well in round bottom 96-well plates . After 3 days , culture supernatants were collected and assayed for IFN-γ , TNF-α , IL-4 , and IL-10 using the FlowCytomix Mouse Cytokine Kit ( Bender MedSystems , Vienna , Austria ) according to the manufacturer's instructions . Single cell suspensions were prepared by teasing apart spleens and inguinal and mesenteric lymph nodes ( LNs ) from 6 mice per group in PBS containing 1% FCS and 1% EDTA followed by red blood cell ( RBC ) lysis with Tris ammonium chloride buffer . CD4+ T cells were purified from single cell suspensions with a CD4+ T cell negative-isolation kit ( Miltenyi Biotec , Auburn , CA ) and a magnetic activated cell sorter ( MACS ) according to the manufacturer's recommendations ( >97% CD4+ T cells by flow cytometric analysis ) . CD4+CD25+ and CD4+CD25− cell populations were separated from purified CD4+ T cells using a mouse regulatory T cell isolation kit ( Miltenyi Biotec ) following the manufacturer's protocol . The CD25+ populations were >95% CD4+CD25+ , and the CD4+CD25− populations were 98% pure as determined by flow cytometry . Antigen presenting cells ( APCs ) were obtained from single cell suspensions by depleting T cells using a mixture of magnetic beads conjugated with either anti-CD8 or anti-CD4 monoclonal antibodies ( mAb ) ( Miltenyi Biotec ) followed by irradiation ( 30 Gy ) . For suppression assays , 1×105 CD4+CD25− T cells/well , 5×104 CD4+CD25+ T cells/well or both were cultured in 96-well U-bottom plates with 1×105 APCs/well in triplicate for 72 h at 37°C in complete RPMI 1640 medium ( 0 . 2 mL/well ) . Cultures were stimulated with 1 µg/mL soluble anti-CD3 ( BD Pharmingen , San Diego , CA ) in the presence or absence of 3 µg/mL CpG , 3 µg/mL R848 , or a combination of both ( 3 µg/mL each of CpG and R848 ) . Proliferation was measured by incubating cells with 0 . 5 µCi/well 3H thymidine and measuring 3H thymidine incorporation during the final 16 h of a 3-d culturing period . The supernatants were collected to quantitatively measure IFN-γ , TNF-α , IL-4 , IL-10 , IL-2 , and IL-6 production using the FlowCytomix Mouse Cytokine Kit ( Bender MedSystems , Vienna , Austria ) according to the manufacturer's instructions . For analysis of CD4+CD25+Foxp3+ T cells , the Mouse Regulatory T Cell Staining Kit ( eBioscience , San Diego , CA ) was used . Splenocytes from immunized mice or naïve mice incubated in the presence or absence of CpG ( 3 µg/mL ) , R848 ( 3 µg/mL ) , or a combination of both ( 3 µg/mL each of CpG and R848 ) for 48 h , were surface-stained with PerCP anti-CD3 monoclonal antibody ( mAb ) ; ( eBioscience , San Diego , CA ) , FITC–anti-CD4 mAb , and APC–anti-CD25 mAb , followed by fixation and permeabilization with Cytofix/Cytoperm and intracellular staining with phycoerythrin ( PE ) -mouse anti-Foxp3 or PE–IgG2a rat IgG control antibody according to the manufacturer's protocol . Data were collected on a FACSCalibur flow cytometer using CellQuest ( BD Biosciences , Franklin Lakes , NJ ) and analyzed with FlowJo software ( Tree Star , San Carlos , CA ) . Statistical analyses were performed using SPSS version 10 . 1 ( Statistical Package for Social Sciences statistical software , Chicago , IL ) . Results were expressed as the mean ± standard error of the mean ( SEM ) . The Mann-Whitney U test was used to calculate the significance between the different groups and a P<0 . 05 ( two-tailed ) was considered statistically significant .
TLR ligands are potent inducers of immune activation and have been shown to augment vaccine efficacy . To assess the effect of TLR ligands on the protective efficacy of the pVAX1-Sj26GST vaccine , C57BL/6 mice were immunized 3 times at 2-week intervals with pVAX1 , CpG , R848 , pVAX1-Sj26GST , or pVAX1-Sj26GST plus CpG and/or R848 . The degree of protection induced by vaccination was measured by the reduction in adult worm and egg burden . As shown in Figure 1 , compared with the pVAX1-inoculated control group , mice inoculated with pVAX1-Sj26GST experienced a 29 . 04% decrease in worm burden and a 25 . 28% decrease in eggs in the liver ( P<0 . 05 ) ( Figure 1; Table 1 ) . Meanwhile , addition of CpG or R848 during pVAX1-Sj26GST vaccination led to a 31 . 82% and 33 . 33% decrease in worms and a 29 . 75% and 26 . 83% decrease in eggs in the liver , respectively ( Figure 1; Table 1 ) . However , the combination of CpG and R848 during pVAX1-Sj26GST vaccination resulted in a higher decrease in worm burden ( 53 . 69% ) and liver eggs ( 49 . 65% ) compared with pVAX1-Sj26GST vaccination alone or with a single ligand , whereas CpG or R848 alone provided almost no reduction in worm burden or liver egg burden ( Figure 1; Table 1 ) . These results suggest that the combination of CpG and R848 significantly improves the protective efficacy of pVAX1-Sj26GST vaccination . The results described above demonstrated that the combination of CpG and R848 improved the protection of pVAX1-Sj26GST vaccination . Thus , we investigated whether adjuvant CpG and R848 allows a more robust induction of immune responses after pVAX1-Sj26GST vaccination . To determine how these TLR ligands influence the immune response following schistosome antigen-specific stimulation , splenocyte proliferation and antibody production were assessed . Splenocytes were isolated from mice vaccinated with pVAX1 , CpG , R848 , pVAX1-Sj26GST , or pVAX1-Sj26GST plus CpG and/or R848 and stimulated with soluble worm antigen ( SWA ) . To exclude the possibility that CpG and/or R848 induced splenomegaly by increasing splenocyte proliferation , mice were subcutaneously injected with CpG and/or R848 three times at 14-day intervals , after which the spleens were weighed and spleen cells counted . Injection of CpG and/or R848 did not induce splenomegaly ( Figure S1 ) , as the spleen weight and cell number from CpG and/or R848-injected mice were not significantly different that those of PBS-injected mice . As shown in Figure 2A , in vitro SWA stimulation significantly increased the proliferation of splenocytes isolated from pVAX1-Sj26GST-vaccinated mice . However , vaccination in combination with CpG and R848 resulted in higher splenocyte proliferation than vaccination with pVAX1-Sj26GST alone or together with CpG , and CpG or R848 alone led to almost no improvement in splenocyte proliferation . These results suggest that the combination of CpG and R848 enhanced antigen-specific T-cell proliferation during pVAX1-Sj26GST immunization . To examine whether adjuvant CpG and R848 influences antibody production , the levels of specific SWA antibodies in the serum of vaccinated mice were examined . As shown in Figure 2B , pVAX1-Sj26GST vaccination causes a significant increase in antigen-specific IgG levels ( P<0 . 01 ) compared with pVAX1 control inoculation ( Figure 2B ) . However , vaccination in combination with CpG and R848 increased IgG levels more than vaccination with pVAX1-Sj26GST alone or with single ligands , and CpG or R848 alone provided almost no improvement in IgG levels . Furthermore , the combination of CpG and R848 induced a small but statistically significant increase in IgG2a level compared with pVAX1-Sj26GST alone or single ligands . No IgG1 response was observed in immunized mice , regardless of whether TLR ligands were used ( Figure 2B ) . The end-point antibody titers are shown in Table 2 . High levels of SWA-specific antibody titers were obtained from the total IgG from mice immunized with pVAX1-Sj26GST or pVAX1-Sj26GST plus CpG and/or R848 . A robust antibody titer for IgM was also observed in the sera of the aforementioned vaccinated mice . However , the highest titers of IgG and IgM were observed in the group in which pVAX1-Sj26GST was combined with CpG and R848 . No detectable levels of SWA-specific antibodies ( IgG and IgM ) were detected in mice immunized with pVAX1 , CpG , or R848 alone . Taken together , these results indicate that the combination of CpG and R848 specifically enhances both splenocyte proliferation and IgG and IgG2a production during pVAX1-Sj26GST vaccination . To further investigate the influence of TLR ligands on the immune response , the levels of cytokines in splenocytes isolated from mice vaccinated with pVAX1 , CpG , R848 , pVAX1-Sj26GST , or pVAX1-Sj26GST plus CpG and/or R848 after SWA stimulation were examined . Compared with the pVAX1 control , pVAX1-Sj26GST vaccination significantly increased the production of IFN-γ ( Figure 3A ) , whereas IL-4 levels decreased and TNF-α and IL-10 levels were not significantly changed ( Figure 3B , 3C , and 3D ) . However , vaccination along with the combination of CpG and R848 resulted in higher IFN-γ and TNF-α levels than vaccination with pVAX1-Sj26GST alone or with single ligands ( Figure 3A and 3B ) . Compared with pVAX1-Sj26GST plus R848 , IL-4 levels were significantly elevated ( P = 0 . 016 ) while IL-10 levels were slightly increased without statistical significance ( P = 0 . 423 ) in mice vaccinated with pVAX1-Sj26GST plus CpG and R848 ( Figure 3C and 3D ) . However , IL-4 and IL-10 levels in splenocytes from mice vaccinated with pVAX1-Sj26GST plus CpG and R848 were lower than in splenocytes from mice vaccinated with pVAX1 control ( Figure 3C and 3D ) . Overall , these results demonstrate that the combination of CpG and R848 during pVAX1-Sj26GST vaccination causes the upregulation of proinflammatory cytokines . It has been reported that , while stimulating antigen-specific effector T cells , Tregs may be expanded to regulate effector T cells [32] , [33] . Furthermore , our previous study demonstrated that pVAX1-Sj26GST immunization induces a significant increase of CD4+CD25+Foxp3+ Tregs that may be involved in the limited protection the vaccine confers [25] . We examined whether the TLR ligands enhanced the protection of pVAX1-Sj26GST by decreasing the number of Tregs during vaccination . To determine the impact of pVAX1-Sj26GST immunization with adjuvant CpG and R848 on CD4+CD25+ Treg frequency , splenocytes isolated from vaccinated mice were stained pre- and postimmunization for the Treg markers CD4 , CD25 , and Foxp3 . As shown in Figure 4 , in addition to the pVAX1-Sj26GST plus CpG and R848 group , other vaccinated mice showed an increase in CD4+CD25+ Tregs as judged by the fold change of Treg proportions after vaccination . Consistent with the results we described previously [25] , both pVAX1- and pVAX1-Sj26GST–immunized mice induced an increase in the percentage of CD4+CD25+ Tregs after vaccination; however , there was no difference in the percentage of Tregs between the pVAX1 and pVAX1-Sj26GST groups . Inclusion of CpG or R848 alone in the vaccination did not affect Treg proportion; however , the combination of CpG and R848 significantly decreased the number of Tregs in immunized mice compared with mice vaccinated with pVAX1-Sj26GST alone , pVAX1-Sj26GST together with single ligands , or the control groups . Single CpG or R848 also induced an increase in the CD4+CD25+ Treg population after vaccination ( Figure 4 ) . These results suggest that the combination of CpG and R848 might prevent the expansion of CD4+CD25+ Tregs and thereby improve the immune response and protection of pVAX1-Sj26GST vaccination . TLR8 and TLR9 ligands have been shown to directly impair Treg function in humans or rats [20] , [21] . In order to examine further the effects of CpG and R848 on Treg activity in our system , CD4+CD25− T cells ( responder cells ) were sorted and cocultured with CD4+CD25+ T cells from naïve mice . Figure 5 shows that following stimulation with anti-CD3 antibody , CD4+CD25+ T cells were highly effective at suppressing CD4+CD25− T-cell proliferation . Conversely , adding CpG moderately reduced the inhibition of T-cell proliferation . However , adding either R848 or both CpG and R848 significantly inhibited Treg-mediated suppression of T-cell proliferation . Compared to the combination of CpG and R848 , CpG alone induced lower level in inhibiting T-cell proliferation , whereas R848 alone showed no statistical significant reduction in Treg-mediated inhibition ( Figure 5 ) . These results suggest that the combination of CpG and R848 not only reduces the Treg population in vaccinated mice in vivo ( Figure 4 ) but also inhibits Treg function in vitro . It has been reported that cytokines secreted by APCs in response to TLR ligands are important to counteract the immunosuppressive effects of CD4+CD25+ Tregs [34] , [35] . Several reports have also shown that a variety of proinflammatory cytokines can overcome Treg suppression during infection or in an inflamed environment , including IL-2 [36] , IL-4 [37] , IL-6 [21] , [38] , and TNF-α [39] , [40] . To investigate the impact of CpG and R848 on cytokine production in vitro , CD4+CD25− cells from naïve mice were cocultured with naïve murine CD4+CD25+ cells , irradiated APCs , and anti-CD3 in the presence or absence of CpG , R848 , or both and cytokine production in the supernatants was evaluated . As shown in Figure 6 , consistent with previously reported observations that CD4+CD25+ Tregs suppress the production of IFN-γ , TNF-α , IL-4 , IL-10 , IL-2 , and IL-6 [41] , [42] , CpG or R848 significantly enhanced the production of these cytokines . Compared to the combination of CpG and R848 , CpG alone induced lower levels of IFN-γ , TNF-α , IL-2 , and IL-6 , whereas R848 alone induced almost equal levels of the above-mentioned cytokines ( Figure 6 ) . Compared to IL-10 levels , IFN-γ , TNF-α , IL-4 , IL-2 , and IL-6 levels were remarkably high in a conventional in vitro suppression assay after adding CpG and/or R848 ( Figure 6 ) . The elevated levels of proinflammatory cytokines in the presence of CpG and/or R848 correlates with inhibition of Treg function as described in Figure 5 . Thus , these results suggest that R848 or its combination with CpG induces higher levels of proinflammatory cytokines , which may help break the immunosuppression of CD4+CD25+ Tregs in a conventional in vitro suppression assay . It has been reported that the transcription factor Foxp3 is required for the suppressive activity of Tregs , and its expression in non-regulatory cells converts them into immunosuppressive cells [43] . Furthermore , it has been reported that IFN-γ [44] , IL-4 [44] , IL-6 [45] , and TNF-α [40] can inhibit Foxp3 expression . The elevated amounts of proinflammatory cytokines conferred by the combination of CpG and R848 in a conventional in vitro suppression assay may inhibit the expression of Foxp3 and further affect Treg function and conversion . To test whether the inhibition of Treg function by CpG and R848 was related to the reduction of Foxp3 expression , splenocytes from naïve mice were isolated and exposed to CpG and/or R848 for 48 h in vitro . Foxp3 expression was then analyzed by flow cytometry ( FCM ) . FCM showed that the population of Foxp3-expressing splenocytes was significantly reduced in the presence of CpG and R848 . Single CpG or R848 induced a reduction in Foxp3 expression compared with medium alone that was not statistically significant ( Figure 7 ) . These results are consistent with the findings that the combination of CpG and R848 decreased the population of CD4+CD25+ Tregs in pVAX1-Sj26GST–vaccinated mice and inhibited Treg function in vitro .
TLR ligands stimulate innate , adaptive , and regulatory immune responses and as vaccine adjuvants represent a promising approach to stimulating strong immune responses and enhancing vaccine-induced protection [46] . Engagement of TLR9 by CpG enhances immune responses to co-delivered antigens in animal models and are now being developed for clinical use as either vaccine adjuvants or immune therapeutics by Coley Pharmaceuticals ( Pfizer ) and Dynavax Technologies , among others [46] . The TLR7/8 ligand R848 has been approved by the U . S . Food and Drug Administration for use as a stand-alone entity [47] , and was proven to enhance the immune response to co-administered antigens as a vaccine adjuvant [48] , [49] . However , the impact of CpG and R848 on vaccines against schistosomiasis , a disease that poses a significant public health concern in many tropical countries , is unknown , and was the subject of this investigation . In the present study , we demonstrated that immunization with pVAX1-Sj26GST combined with CpG and R848 as adjuvants induces a stronger protection compared with pVAX1-Sj26GST alone or with either single ligand . It has been reported that the use of TLR ligands as adjuvants can elicit more vigorous immune responses against infection and cancer [46] , [50] . Indeed , immunization with pVAX1-Sj26GST combined with CpG and R848 resulted in a significant increase in vaccine-induced splenocyte proliferation and IgG and IgG2a levels . These results are consistent with another study demonstrating that CpG and R848 are the strong Th1-biased adjuvants [51] , because Th1-associated IgG2a was significantly increase in the group in which pVAX1-Sj26GST was combined with CpG and R848 . However , CpG or R848 alone did not enhance protection and immune responses conferred by the vaccine . This finding is consistent with a recent publication showing that TLR ligands as combination adjuvants induce qualitative changes in T-cell responses needed for antiviral or antiparasite protection in mice [52] , [53] . Furthermore , consistent with the current work , Ahmad and colleagues reported that use of R848 as an immunopotentiating agent slightly boosted the protective effects of Sm-p80 , now considered a leading putative vaccine candidate antigen from Schistosoma mansoni ( S . mansoni ) , in both the “DNA prime-protein boost” and “recombinant protein” immunization approaches in mice [24] . The quantification of cytokines in splenocyte culture supernatants indicated that pVAX1-Sj26GST vaccination induces significantly increased IFN-γ levels and decreased IL-4 levels compared with vaccinated control pVAX1-treated mice . However , combining CpG and R848 with pVAX1-Sj26GST augmented the production of IFN-γ and TNF-α in vaccinated mice . Elevation of IFN-γ and TNF-α in response to the combination of CpG and R848 may contribute to the enhanced protection conferred by pVAX1-Sj26GST vaccination . Because the protection induced by many schistosoma vaccines was associated with elevated production of IFN-γ and TNF-α [54] , [55] , our data also suggest that the activation of more than one TLR could be an effective approach to optimize immune responses in vaccination . Consistent with these findings , Lanzavecchia and colleagues [56] reported that synergistic TLR stimulation mimics pathogens that contain several TLR ligands and induces enhanced and sustained T helper type 1 responses in DCs . Furthermore , several studies have shown that certain TLRs enhance T cell-mediated immune responses through synergistic activation of DCs when their ligands are detected in pairs [56] , [57] or through induction of high levels of proinflammatory cytokines by simultaneously activating different signaling pathways [34] , [52] . The enhancement of T-cell responses and proinflammatory cytokine secretion may therefore improve the protection conferred by pVAX1-Sj26GST vaccination . Apart from enhancing effector T-cell proliferation and cytokine production , TLR ligands could be involved in the modulation of adaptive immunity , including Treg-mediated immune suppression in vaccination [58] , [59] . Furthermore , our previous study demonstrated that induction of CD4+CD25+ Tregs after pVAX1-Sj26GST vaccination may explain the limited protection conferred by this vaccine [25] . We thus hypothesized that enhancement of immune responses and protection conferred by the combination of CpG and R848 may be related to the inhibition of Treg induction after pVAX1-Sj26GST vaccination . Indeed , we did see a small but significant decrease in CD4+CD25+ Tregs after vaccination with pVAX1-Sj26GST plus CpG and R848 . However , use of CpG or R848 alone only slightly affected the Treg population , which is consistent with other viral studies suggesting that combination of TLR ligands prevents expansion of Foxp3+ Tregs and thereby improves T cell responses [53] . However , Hoelzinger and colleagues report that intratumoral delivery of CpG ODN strongly reduces the levels of Tregs within the tumor [60] . Hackl and coworkers demonstrated that TLR7 ligands , e . g . , R848 , reduce the number of Tregs generated de novo from naïve murine T cells in vitro and in vivo [45] . Whether these differences are due to different host systems , different disease models , or different vaccine formulations remains to be investigated in future studies . The decreased population of CD4+CD25+ Tregs in response to the combination of CpG and R848 may therefore improve immune responses and protection in pVAX1-Sj26GST-vaccinated mice . There is evidence that TLR signaling can modulate the suppressive functions of Tregs [35] , [39] . Based on these studies , we investigated the effects of CpG and R848 on Treg-mediated suppression in vitro . In contrast to the human TLR8 ligands CpG-A and poly ( G10 ) [21] and the TLR9 ligand CpG ODN [20] , which abrogate or reverse the immunosuppressive function of CD4+CD25+ Tregs , we found that CpG did not inhibit Treg suppression , whereas R848 alone or in combination with CpG significantly inhibited the function of Tregs . This finding support the notion that , despite several structural and functional properties shared by all the members of the TLR family , the signaling through various TLRs elicits qualitatively and quantitatively diverse immune responses [61] , such as their impact on Treg function . It has been suggested that exposure to inflammatory cytokines released by APCs can render Tregs defective in immunosuppression [41] . For example , IL-6 production by TLR-activated DCs can inhibit the suppressive function of Treg cells [38] . Furthermore , exposure to TNF-α can inhibit the function of Tregs by signaling through TNF receptor II [40] . Consistent with a previous study reporting that CD4+CD25+ Tregs are able to suppress T-cell proliferation and cytokine production [62] , our study demonstrated that the presence of CpG and/or R848 in a conventional in vitro suppression assay induces a panel of inflammatory cytokines , including IFN-γ , TNF-α , IL-4 , IL-10 , IL-2 , and IL-6 , that may inhibit Treg suppression . Although IL-10 is a major anti-inflammatory cytokine induced by TLR signaling and functions to inhibit production of TLR-induced proinflammatory mediators , such as TNF [63] , this study shows that elevated levels of IL-10 in the presence of CpG and/or R848 in an in vitro suppression assay is insufficient to overcome the strong inflammatory context caused by other cytokines . Furthermore , CpG and R848 reduce the expression of Foxp3 in CD4+ T cells in vitro , which is indispensable in Treg development and function [44] , [64] . Consistent with the above-mentioned cytokine production observed in an in vitro suppression assay , a variety of cytokines has been reported to inhibit Treg function by inhibiting Foxp3 expression , including IFN-γ [44] , TNF-α [40] , IL-4 [44] , and IL-6 [45] . Although it remains to be determined whether the increased production of inflammatory cytokines induced by CpG and R848 directly stimulates CD4+CD25− effector T cells or indirectly acts on APCs , these results strongly suggest that the combination of CpG and R848 contributes to the activation and expansion of effector T cells , increases cytokine secretion , and interferes with Treg function by downregulating Foxp3 expression . Although the in vitro assays of TLR ligands on Tregs fail to completely mimic the in vivo milieu , they lead us to speculate that the downregulation of Foxp3 expression not only affects Treg function in vitro , but also may impair the generation of Tregs after vaccination in vivo , thereby reducing the number of CD4+CD25+ Tregs in mice vaccinated with pVAX1-Sj26GST together with CpG and R848 . These results are consistent with a previous study demonstrating that activation of DCs by TLR7 ligands leads to downregulation of Foxp3 expression after initial induction and consequently lowers Treg numbers in DC–T-cell cocultures in vitro [45] . Furthermore , single TLR ligands less potently decrease CD4+Foxp3+ T cells , whereas combined TLR ligands might prevent Foxp3+ Treg expansion and thereby improve T-cell responses [53] . However , R848 induced higher cytokine production than CpG in a conventional in vitro suppression assay . This is in contrast to other studies on cytokine secretion by splenocytes stimulated with CpG or R848 in which CpG was a greater stimulant of IL-6 and IL-12 secretion than R848 and R848 was superior to CpG in promoting IL-10 secretion [65] . The differences between CpG and R848 in inducing cytokine production might be related to the differences in their respective TLR signaling pathways , differences in the stability of CpG and R848 interactions with the ligands , and/or differences in the stability of the molecules in cells in the conventional in vitro suppression assay . Although there is no direct evidence that R848 is superior to CpG in induction of inflammatory cytokines , Martín-Fontecha and colleagues reported that R848 , but not CpG1826 , can recruit NK cells to produce IFN-γ and prime T cells for the induction of TH1 cells [29] . Further analysis is needed to determine why R848 induced more cytokines than CpG and which cells produced these cytokines in a conventional in vitro suppression assay . A greater understanding of the cellular events triggered by single or combinations of TLRs will be valuable in the rational design of more successful TLR-based immunotherapies and vaccination strategies . However , It should be noted that although cooperation among TLRs during infection and vaccination may result in more robust immune responses and protection , if not properly controlled , these strong responses can result in immunopathologies such as autoimmunity [66] . Indeed , the use of the TLR7 and TLR8 agonist imiquimod in patients with cancer exacerbates psoriasis [67] . Thus , TLR-regulated Treg activity and conversion could enhance pathogen clearance but also increase the risk of autoimmune reactions . Future studies using a TLR-based vaccine strategy are required to evaluate this possibility . In conclusion , this work demonstrates that the combination of CpG and R848 increases the proliferation of splenocytes and IgG levels and improves disease protection after immunization with the S . japonicum vaccine pVAX1-Sj26GST . This enhancement of protection may be related to the inhibition of Treg expansion and function , as the combination of CpG and R848 may impair the Treg development and function by upregulating the secretion of proinflammatory cytokines and decreasing Foxp3 expression . In combination with the vaccine , TLR ligands may protect the effector response from Treg-mediated suppression , thereby eliciting the appropriate immune response to improve vaccine efficacy . Our findings support the notion that , similar to an infection , vaccination also may allow Tregs to expand concurrently with T cells [68] . However , the addition of paired TLR ligands as adjuvants induces an proinflammatory setting which acts either by direct inhibition of Treg suppression or rescue of Teffs from Treg-mediated suppression to allow expansion of antigen-specific T cells against S . japonicum . Therefore , modulation of Tregs by adjuvant TLR ligand combinations may represent an attractive strategy to enhance the efficacy of vaccination against pathogens . | There is evidence that TLR activation can block Treg cell responses and thereby break tolerance to self-antigens . It is expected that the use of TLR ligands as vaccine adjuvants will induce potent anti-pathogen immune responses and simultaneously overcome immune inhibition mediated by Tregs . However , the impact of TLR ligands on schistosomiasis vaccines is unclear . Here , we demonstrate that the use of a TLR7/8 ligand ( R848 ) and a TLR9 ligand ( CpG ) as adjuvants in combination with the S . japonicum vaccine pVAX1-Sj26GST improves disease protection . The combination of CpG and R848 administered after vaccination causes an immune response marked by an upregulation of splenocyte proliferation and IgG and IgG2a levels that also coincides with a decreased proportion of CD4+CD25+ Tregs in mice . We also show that combined adjuvant use of CpG and R848 may impair Treg development and function by promoting the secretion of proinflammatory cytokines and reducing Foxp3 expression . Our findings suggest that in combination with the vaccine , TLR ligands may protect the effector response from Treg-mediated suppression , thereby eliciting the appropriate immune response to improve vaccine efficacy . Immunization combined with the TLR ligands CpG and R848 thus represents a promising new approach for the design of schistosoma vaccines . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"vaccines",
"medicine",
"vaccination",
"infectious",
"diseases",
"schistosomiasis",
"clinical",
"immunology",
"immunity",
"immunology",
"parasitic",
"diseases",
"immunomodulation"
] | 2013 | Combined TLR7/8 and TLR9 Ligands Potentiate the Activity of a Schistosoma japonicum DNA Vaccine |
Modified vaccinia virus Ankara ( MVA ) is the leading poxvirus vector for development of vaccines against diverse infectious diseases . This distinction is based on high expression of proteins and good immunogenicity despite an inability to assemble infectious progeny in human cells , which together promote efficacy and safety . Nevertheless , the basis for the host-range restriction is unknown despite past systematic attempts to identify the relevant missing viral gene ( s ) . The search for host-range factors is exacerbated by the large number of deletions , truncations and mutations that occurred during the long passage history of MVA in chicken embryo fibroblasts . By whole genome sequencing of a panel of recombinant host-range extended ( HRE ) MVAs generated by marker rescue with 40 kbp segments of vaccinia virus DNA , we identified serine protease inhibitor 1 ( SPI-1 ) as one of several candidate host-range factors present in those viruses that gained the ability to replicate in human cells . Electron microscopy revealed that the interruption of morphogenesis in human cells infected with MVA occurred at a similar stage as that of a vaccinia virus strain WR SPI-1 deletion mutant . Moreover , the introduction of the SPI-1 gene into the MVA genome led to more than a 2-log enhancement of virus spread in human diploid MRC-5 cells , whereas deletion of the gene diminished the spread of HRE viruses by similar extents . Furthermore , MRC-5 cells stably expressing SPI-1 also enhanced replication of MVA . A role for additional host range genes was suggested by the restoration of MVA replication to a lower level relative to HRE viruses , particularly in other human cell lines . Although multiple sequence alignments revealed genetic changes in addition to SPI-1 common to the HRE MVAs , no evidence for their host-range function was found by analysis thus far . Our finding that SPI-1 is host range factor for MVA should simplify use of high throughput RNAi or CRISPR/Cas single gene methods to identify additional viral and human restriction elements .
Vaccinia virus ( VACV ) has been developed as a live recombinant expression vector that is widely used for making candidate vaccines against unrelated pathogens [1–5] . Although VACV was successfully used as a smallpox vaccine , concerns regarding safety with regard to the creation of new vaccines led to interest in more attenuated poxvirus vectors including fowlpox virus [6] , canarypox virus [7 , 8] , and recombinant VACV strains in which one or multiple genes were deleted selectively [9 , 10] or by blind passaging [11 , 12] . One such attenuated strain , modified vaccinia virus Ankara ( MVA ) , was produced by passaging the parental chorioallantois vaccinia virus ( CVA ) strain more than 500 times in chicken embryo fibroblasts ( CEF ) for the purpose of producing a safe smallpox vaccine [11] . Initial analysis of the MVA genome revealed six major deletions compared to the parent virus [13] . These large deletions as well as numerous additional genetic changes were confirmed by genome sequencing [14] . Notwithstanding the loss of considerable genetic material and the consequent inability to efficiently produce infectious virus in most mammalian cells [13 , 15–17] , MVA retains the ability to express viral as well as recombinant proteins regulated by VACV promoters in non-permissive cells at levels comparable to replicating VACV and to induce both humoral and cellular immune responses [18 , 19] . These beneficial features propelled the use of MVA for development of numerous candidate vaccines , some of which are in clinical trials [20] . Despite extensive testing of candidate MVA vaccines in humans , the basis for the host-restriction of MVA , which is important to fully understand its attenuation , remains unknown . The large number of deletions , truncations and mutations that occurred during the long passage history of MVA in CEF severely complicates efforts to determine those changes important for its host-range defect . Indeed , a comparison of MVA with its parent CVA revealed 71 orthologous ORFs predicted to encode identical gene products , whereas the remaining 124 ORFs encode gene products with amino acid changes , insertions or deletions [21] . One attempt to investigate the genetic changes responsible for the replication defect consisted of deleting DNA sequences corresponding to the six major deletions of MVA from the genome of the parental CVA [22] . Remarkably , the loss or truncation of 31 open reading frames ( ORFS ) totaling ~ 25 kbp of DNA from the parental virus was insufficient to produce the host-range phenotype of MVA , leading to the conclusion that the major determinants lie outside of these deletions . In a related approach , the large deletions of MVA were introduced into the Lister strain of VACV [23] . Loss of the genes corresponding to those missing from deletion I of MVA , located near the left end of the genome , reduced replication in HeLa cells by 4- to 5-fold . No additional effect was observed upon introducing the additional deletions of MVA into the Lister strain . A totally different approach entailed a marker rescue scheme in which recombinant viruses were produced by transfecting MVA-infected cells with cosmids containing DNA segments of ~40 kbp spanning the genome of a replicating strain of VACV [24] . When the infected cell lysates were plated on BS-C-1 cells , large plaques were observed in samples that had been transfected with DNA derived from the left end of the genome . Following clonal isolation , six of eight independently isolated recombinant viruses ( v44 . 1 , v44 . 2 , v44/47 . 1 , v44/47 . 2 , v51 . 1 , v51 . 2 ) , named after the cosmids used for their rescue , were found to also replicate to high titers in human MRC-5 , HeLa and A549 cells . We refer to these recombinant viruses as host-range extended ( HRE ) MVAs . A subsequent study demonstrated that v44/47 . 1 and v51 . 1 replicate well in monkey Vero cells , which are frequently used for vaccine manufacturing , while still exhibiting severe attenuation in immunocompetent and immunodeficient mice [25] . Thus far , there has been only a limited investigation of how the HRE MVAs overcome the host range restriction . Dobson and Tscharke [26] found that the F5L gene , which was restored in v44 . 1 , was important for plaque morphology but did not enhance replication of MVA . A comparison of the whole genome sequences of MVA and its parent CVA revealed that only two known host range genes , C12L and K1L , located near the left end of the genome , are specifically missing or truncated in MVA [21] . However , introduction of the K1L gene into MVA did not reverse the human host-range defect [27] . A corresponding study of the effects of insertion of the C12L gene had not been reported , even though polymerase chain reactions ( PCR ) of the HRE MVA genomes revealed a correlation of the acquisition of C12L DNA with replication in human cells [24] . The protein encoded by C12L belongs to the serine protease inhibitor superfamily known as serpins and is called serine protease inhibitor-1 ( SPI-1 ) [28] . SPI-1 is conserved in orthopoxviruses and expressed as an intracellular non-glycosylated 40-kDa species [29] . Deletion of the SPI-1 ORF from rabbitpox virus ( RPXV ) or VACV strain WR causes diminished replication in human A549 and pig kidney 15 cells but not in several avian and monkey cell lines [29 , 30] . A recent human genome-wide RNAi screen implicated three genes ( IRF2 , FAM111A and RFC3 ) in the restriction of SPI-1 deletion mutants in human A549 cells , although the mode of their action remains to be determined [31] . The primary intent of the present study was to identify specific genes lost during the passage history of MVA that contribute to its host-range defect . The whole genome sequences of five independently isolated HRE MVAs revealed the presence of an intact C12L open reading frame ( ORF ) in those viruses that gained the ability to replicate in human cells . Additionally , we observed a similar assembly block in MVA and a SPI-1 mutant derived from VACV WR . Most importantly , we demonstrated that insertion of the C12L ORF into MVA enhanced replication by more than 2 logs in human MRC-5 cells whereas deletion of the C12L ORF from the HRE MVAs diminished replication by similar amounts . Although multiple sequence alignments revealed additional genetic changes common to the HRE MVAs , no evidence for their host-range function has been found by mutational analysis thus far .
Although numerous genes were deleted or truncated during the long passage history of MVA , the only ones with known human host-range function are C12L encoding SPI-1 and K1L . Table 1 summarizes previous PCR data [24] confirming the absence of C12L DNA and truncation of K1L in MVA . Strikingly , C12L was detected by PCR in all HRE MVAs that were able to replicate in human cells , whereas the presence of full-length K1L correlated with replication only in rabbit kidney cells ( Table 1 ) . The correlation of C12L and replication in human cells focused our attention on SPI-1 as a missing host-range factor for MVA . Unlike most orthopoxvirus host-range mutants , which exhibit blocks in gene expression , the earliest recognized defect in MVA replication occurs during morphogenesis [15 , 18 , 32] . Interestingly , the second exception to the general rule for the predominance of impaired gene expression is the morphogenesis defect of SPI-1 deletion mutants of rabbitpox virus and the WR strain of VACV in non-permissive cells [30 , 33] . The possibility that the absence of the SPI-1 gene contributes to the host-range defect of MVA persuaded us to compare their morphogenesis blocks . Human A549 cells that had been infected for 24 h with VACV WR , a WR SPI-1 deletion mutant ( WRΔSPI-1 ) or MVA were prepared for transmission electron microscopy . In the cells infected with WR ( Fig 1A ) , there was a predominance of brick-shaped mature virions ( MVs ) and some wrapped or partially wrapped virions ( WVs ) as well as crescents ( C ) and immature virions ( IVs ) . The cells infected with WRΔSPI-1 ( Fig 1C ) had many aberrant particles with the spherical shape and dimensions of IVs but with dense unstructured interiors , which are referred to as dense virions ( DVs ) . Many spherical DVs were also present in the cells infected with MVA ( Fig 1E ) . Higher magnification confirmed the similar appearances of the DVs in the cells infected with MVA and the WRΔSPI-1 ( Fig 1D and 1F ) and the more mature morphology of MVs in the cells infected with WR ( Fig 1B ) . Thus , the impairment in morphogenesis occurred at a similar stage in non-permissive cells infected with MVA and WRΔSPI-1 . Nevertheless , this similarity only provided suggestive evidence of related defects . The presence of the C12L ORF in HRE MVAs that replicate in human cells and the similarity in the morphogenesis block of MVA and other SPI-1 deletion mutants led us to investigate whether the introduction of the SPI-1 ORF into the MVA genome would have a discernible effect on replication in human cells . To facilitate the construction of the recombinant MVA , we used a transfer plasmid vector in which the C12L ORF was regulated by the well characterized modified mH5 promoter that has strong early and moderate late activities [34 , 35] . Recombination occurred into the site of deletion III located near the right end of MVA so as not to interrupt or alter additional genes . Permissive CEF were used for infection and transfection and the recombinant virus , named MVA-SPI-1 , was clonally isolated by several rounds of picking foci that fluoresce due to co-expression of the green fluorescent protein ( GFP ) . PCR and Sanger sequencing were performed to confirm insertion of the complete C12L ORF . The effect of the C12L ORF addition was assessed by virus spread in several cell lines with a range of virus multiplicities of MVA or MVA-SPI-1 . MVA is less cytopathic than other strains of VACV and does not form regular shaped plaques under semisolid medium that can be easily discerned by staining with crystal violet or neutral red . Consequently , the irregular foci formed were visualized by immunostaining with antibody to VACV . We used a range of virus multiplicities because of cytopathic effects that occur during the 48 h incubations . In permissive CEF , the two viruses formed foci of similar number , size and staining intensity best seen at a MOI of 0 . 001 before confluence occurred ( Fig 2A ) . In monkey BS-C-1 and human MRC-5 cells , MVA-SPI-1 foci that were larger and exhibited more intense staining relative to MVA were best seen at MOI of 0 . 001 and 0 . 01 ( Fig 2A ) . In HeLa and A549 cells , MVA-SPI-1 also exhibited an increase in staining relative to MVA , but the effect was less than in BS-C-1 and MRC-5 cells and was best discerned at a MOI of 0 . 1 ( Fig 2A ) . To complement the results of addition of the SPI-1 gene to MVA , we deleted the gene from the HRE MVAs . This was accomplished by homologous recombination with DNA containing the GFP ORF regulated by the p11 late promoter within C12 flanking sequences . Recombination was carried out in CEF and the virus in fluorescent foci were clonally purified by repeated isolations . The loss of the C12L gene was confirmed by PCR and Sanger sequencing . The effect of the gene deletion from v51 . 2 ( v51 . 2ΔSPI-1 ) was determined by infecting cells with 0 . 001 to 0 . 1 PFU per cell . There was no discernible effect of the gene deletion in CEF and only a slight effect in BS-C-1 cells , whereas in HeLa , A549 and MRC-5 cells the foci formed by v51 . 2ΔSPI-1 stained less intensely than those formed by the parent virus v51 . 2 ( Fig 2B ) . A comparison of Fig 2A and 2B suggested that loss of SPI-1 by v51 . 2 had a greater impact than gain of SPI-1 by MVA particularly in A549 cells . We also compared the effects of SPI-1 deletions on the other independently isolated HRE MVAs ( v51 . 1 , v44 . 1 and v44/47 . 1 ) . Although deletion of the C12L ORF had no discernible effect in CEF , in each case the staining intensity of foci was reduced in MRC-5 cells , which was most clearly seen at the lowest MOI ( Fig 3 ) . Virus yields were determined at 48 h after inoculating MRC-5 and A549 cells with viruses at a multiplicity of infection ( MOI ) of 0 . 001 or 0 . 01 in order to quantify the effects of SPI-1 on replication and spread . In MRC-5 cells , expression of SPI-1 increased the virus yield relative to MVA by 160-fold ( p = 0 . 0006 ) and deletion of SPI-1 from v51 . 2 , v51 . 1 , v44 . 1 and v44/47 . 1 reduced the yields by approximately 400-fold in each case ( p<0 . 01 ) ( Fig 4A ) . In A549 cells , there was also an increase in yield produced by expression of SPI-1 relative to MVA and a decrease in yield of HRE viruses due to deletion of SPI-1 ( Fig 4B ) . However , both effects were much smaller than in MRC-5 cells . However , even in MRC-5 cells , addition of SPI-1 to MVA did not increase the yield to the levels of the HRE viruses , which all have C12L , nor did deletion of C12L from the latter viruses reduce the yield to the level of MVA . Therefore , we concluded that absence of SPI-1 strongly contributes to the host-range defect of MVA but is not the sole factor responsible . Serine protease inhibitor activity of RPXV SPI-1 was suggested by the formation of a stable complex with cathepsin G in vitro , which was prevented by mutation of the phenylalanine to alanine in the putative reactive loop [36] . Furthermore , when the phenylalanine to alanine mutation was introduced into the RPXV genome , the host range was similar to that of an SPI-1 deletion mutant in A549 cells . However , we found that , recombinant MVAs containing SPI-1 with or without the reactive loop mutation ( F322A ) or a control mutation outside of the loop ( T309R ) enhanced MVA spread similarly in MRC-5 cells ( Fig 4C ) suggesting that SPI-1 may have more than one mode of function . Thus far , we have shown that expression of SPI-1 by MVA and HRE MVAs enhanced replication in human cells . However , viral genome alterations could have unanticipated effects . To circumvent this potential problem , we determined the effect of trans-expression of SPI-1 on MVA replication . The C12L ORF with a 2xMyc tag regulated by the CMV promoter was introduced into A549 and MRC-5 cells by transduction with a retroviral vector . Expression of SPI-1 was demonstrated by Western blotting ( Fig 5A and 5C ) . We would expect that ectopic expression of SPI-1 would have little or no effect on the replication of MVA-SPI-1 and v51 . 2 as they already express SPI-1 , whereas there would be enhancement of MVA and v51 . 2ΔSPI-1 . This is precisely what occurred in MRC-5 cells , with increases of approximately 6- and 11-fold for MVA and v51 . 2ΔSPI-1 , respectively ( Fig 5B ) . Ectopic expression of SPI-1 in A549 cells also increased replication of v51 . 2ΔSPI-1 ( Fig 5D ) but had little or no effect on replication of MVA . The latter result was consistent with the small enhancement of MVA-SPI-1 replication compared to MVA in A549 cells ( Fig 4B ) . Thus , both trans- and cis-expression of SPI-1 have similar effects on host range . We were curious whether the greater effects of addition and deletion of SPI-1 in MRC-5 cells compared to A549 cells was specific for MVA . To our knowledge , A549 is the only human cell line in which the effect of SPI-1 deletion had been tested for either RPXV or VACV WR [29–31] . For comparison , A549 and MRC-5 cells were infected with VACV WR and RPXV SPI-1 deletion mutants and the parental viruses . In A549 cells , deletion of SPI-1 reduced the spread of VACV WR and RPXV by 15-and 640-fold respectively , whereas in MRC-5 cells the reductions were 39- and 12 , 000 respectively ( Fig 5E ) . Thus , not only was RPXV more dependent than VACV WR on SPI-1 , but the requirement was greatly increased in MRC-5 cells compared to A549 cells for both viruses . We conclude that viral as well as cellular genetic backgrounds determine the degree of dependency on SPI-1 for replication . The v51 . 2 , v51 . 1 , v44 . 1 and v44/47 . 1 HRE MVAs each replicated more than 3 logs higher than the parental MVA in MRC-5 cells ( Fig 4A ) . However , even after deletion of C12L they still replicated at least one log higher than MVA suggesting the presence of one or more additional host range genes . The entire genomes of the recombinant HRE MVAs were sequenced in order to identify additional genes that might contribute to the alleviation of the host-range defect and were deposited in GenBank . In the multiple alignments depicted in Fig 6A , the ORFs derived from the partially overlapping cosmids used for marker rescue and have at least one nucleotide polymorphism are colored green and the ORFs identical to those in MVA are colored yellow . Inserted DNA was detected near the left ends of v51 . 1 , v51 . 2 , v44 . 1 and v44/47 . 1 , consistent with the cosmids used for their generation . Although no DNA insertions were found anywhere in the v47 . 1 genome , there were some sequence differences presumably arising spontaneously that enabled replication in monkey but not human cells . The left end deletions I , V and II in MVA are annotated . Repair of deletion I , which included insertion of C12L , occurred in each of the recombinant viruses able to replicate in human cells , whereas repair of deletions II and V only occurred in v51 . 1 and v44/47 . 1 and therefore were not essential for replication although the F5L gene affects plaque morphology [26] . Furthermore , the genetic changes in v51 . 2 , which were concentrated around deletion I , were present in each of the HRE MVAs capable of replicating in human cells ( Fig 6A and 6B ) , suggesting that they included the minimal set of potential host range genes . Our strategy was to delete these genes from v51 . 2ΔSPI-1 to see if that further reduced virus spread in MRC-5 and A549 cells but not CEF . We focused on the C15L , C16L and C17L ORFs because they are absent in MVA but present in the HRE viruses . Therefore , these ORFs in the HRE viruses were individually deleted by replacement with mCherry regulated by the p11 promoter . MVA , v51 . 2 , v51 . 2ΔSPI-1 , v51 . 2ΔSPI-1ΔC15 , v51 . 2ΔSPI-1ΔC16 and v51 . 2ΔSPI-1ΔC17 viruses replicated equally well in permissive CEF . In A549 and MRC-5 cells , the replication of v51 . 2 was diminished to the same extent by deletion of SPI-1 alone and deletion of both SPI-1 and either C15L , C16L or C17L ( Fig 7A ) , suggesting that the latter genes are not involved in host range . Although C10L and C11R are present in MVA , there are sequence differences in the homologs of the HRE MVAs that potentially could affect host range . However , deletion of C10L or C11R from v51 . 2 or v51 . 2ΔSPI had no effect on virus spread in CEF or MRC-5 cells ( Fig 7B ) , even though C11R is a growth factor [37 , 38] and has been shown to enhance VACV spread under some conditions [39–41] . Thus , we did not identify an additional gene in v51 . 2 that significantly impaired replication in MRC-5 cells .
A previous study [24] showed that the host-range defect of MVA could be rescued by insertion of large DNA segments from the left end of the genome of a replication-competent virus . This result contrasted with the failure of the opposite approach: an attempt to generate the host-range defect in the CVA strain of VACV , which is the parent of MVA , by deleting DNA corresponding to the six major MVA deletions [22] . The failure of the latter strategy is somewhat perplexing as we showed here by genome sequencing that independently isolated HRE MVAs all had repaired deletion I and that one of these ( v51 . 2 ) acquired no additional DNA . Moreover , deletion of SPI-1 located within the repaired deletion I segment of v51 . 2 restricted the host range . Therefore , if repair of deletion I can extend the host range of MVA and deletion of SPI-1 reduced the replication of HRE MVAs , why didn’t the introduction of deletion I into CVA restrict host-range ? Synthetic lethality , a recognized genetic event in which deficiencies in the expression of two or more genes results in cell or organismal death whereas a deficiency in any one gene does not [42] , could explain this discrepancy . Although this phenomenon is not usually considered for viruses because of their relatively small genomes , poxviruses have genomes encoding ~200 genes and evidence for synthetic lethality has been obtained in other contexts for VACV [43 , 44] . Thus , a second gene that allows VACV replication in human cells could be present outside of the 6 deletions in CVA but not in MVA . The importance of the virus background received support from the fact that deletion of SPI-1 from RPXV is more debilitating than the same deletion from VACV WR . Also , the introduction of deletion I into the Lister strain of VACV reduced replication by 5-fold in HeLa cells ( the reduction inMRC-5 cells was not presented ) [23] . Therefore , deletion of SPI-1 causes no , moderate or severe host range defects depending on the orthopoxvirus strain . Just as differences in the genetic background of VACV strains impact the effect of SPI-1 deletion , so do differences in human cell lines . Introduction of the SPI-1 ORF into MVA increased replication more in MRC-5 cells than HeLa and A549 cells . Nevertheless , deletion of the SPI-1 gene from an HRE MVA had a strong negative effect in A549 as well as in HeLa and MRC-5 cells . The cell-dependent differences in addition and deletion of SPI-1 suggests that there are two distinct mechanisms of host range restriction and that both are highly active in A549 cells but only one in MRC-5 cells . Thus , expression of SPI-1 is necessary for enhanced replication of MVA in MRC-5 and A549 cells but is sufficient only in the former . The host-range function of SPI-1 was first revealed by the failure of spontaneous RPXV mutants to replicate in pig kidney and human A549 cells [29] . The major defect appeared as a block in virus particle maturation accompanied by some features of apoptosis [33] . We showed here that the aberrant particles that accumulate in cells infected with MVA and a VACV with a specific SPI-1 deletion look remarkably similar by electron microscopy . However , the host-range function of RPXV SPI-1 in A549 cells is abolished by mutation of the tryptophan to alanine in the predicted reactive loop [36] , whereas the mutated SPI-1 retains host-range function for MVA in MRC-5 cells . This difference could be due to the presence of additional host factors that inhibit MVA or to differences in the mechanism of inhibition of RPXV and MVA SPI-1 deletion mutants . Taken together , these data also suggest that SPI-1 has more than one host range function depending on the virus and host cell backgrounds . The present results are an important step towards the goal of understanding the basis for the human host range restriction and attenuation of a vector that forms the basis of numerous vaccines in clinical trials . In addition , the information could help to design new and improved vectors . Furthermore , the demonstration that human cells expressing SPI-1 support MVA replication may lead to the development of non-avian cell lines for propagation of candidate MVA vaccines . Finally , our finding that SPI-1 is host range factor for MVA can simplify use of high throughput RNAi or CRISPR/Cas single gene methods to identify additional viral and human restriction elements .
A549 cells ( ATCC CCL-185 ) were grown in Dulbecco’s modified Eagle’s medium/F-12 ( Life Technologies ) supplemented with 10% fetal bovine serum ( FBS , Sigma-Aldrich ) , 2 mM L-glutamine , 100 units of penicillin , and 100 μg of streptomycin per ml ( Quality Biologicals , Inc . ) . Primary CEF prepared from 10-day old fertile eggs ( Charles River ) and BS-C-1 ( ATCC CCL-26 ) and MRC-5 ( ATCC CCL-171 ) cells were grown in minimum essential medium with Earle’s balanced salts ( EMEM ) supplemented with 10% FBS , 2 mM L-glutamine , 100 units of penicillin , and 100 μg of streptomycin per ml ( Quality Bologicals ) . HeLa cells ( ATCC CCL-2 ) were grown in Dulbecco’s modified eagle’s medium ( DMEM ) supplemented with 10% FBS , 2 mM L-glutamine , 100 units of penicillin , and 100 μg of streptomycin per ml ( Quality Biologicals ) . Rabbit antibody to VACV strain WR was described previously [45]; c-Myc antibody ( 9E10 ) conjugated to horse radish peroxidase ( HRP ) ( catalog number sc-40 HRP ) was from Santa Cruz Biotechnology; and rabbit anti-actin ( catalog number A2066 ) was from Sigma-Aldrich . WRΔSPI-1 was derived from the Western Reserve ( WR ) strain of VACV ( ATCC VR-1354 ) and was described previously [31] . A panel of human replication-competent recombinant MVAs ( v51 . 1 , v51 . 2 , v44 . 1 , and v44/47 . 1 ) with segments of added VACV DNA of various lengths was described [24] . RPXV and VACV WR SPI-1 deletion mutants were described previously [31] . Modified viruses were constructed by homologous recombination using fluorescent reporter genes for selection . To generate MVA-SPI-1 , a C12L DNA segment was introduced into the genome of MVA at the deletion III site by inserting the DNA fragment downstream of the mH5 promoter in pLW44-derived vector which also contains the P11 VACV promoter driven GFP [35] . The MVA-SPI-1 F322A and MVA-SPI-1 T309R were constructed by mutating the Phe322 into Ala and Thr309 into Arg using Q5 Site-Directed Mutagenesis Kit ( New England Biolabs , Inc . ) . C12L genes from v51 . 1 , v51 . 2 , v44 . 1 , and v44/47 . 1 were deleted by homologous recombination with a PCR product containing the P11 VACV promoter-driven GFP gene flanked by sequences on either side of C12L . Fluorescent plaques were identified and cloned by repeated plaque isolation . Similarly , C10L and C11R were deleted by replacing the corresponding gene with P11 promoter-driven mCherry . Red plaques were picked and purified by repeated isolation . To generate vΔC12/C11 and vΔC12/C10 , fluorescent foci that expressed both GFP and mCherry were picked and plaque purified . A similar strategy was adopted to delete the C15L , C16L , and C17L from v51 . 2ΔC12 . The recombinant viruses were PCR amplified and sequenced to confirm the identities . Homologous recombination was carried out by infecting CEF with 1 PFU/cell of virus , followed by transfection with assembled PCR products using Lipofectamine 2000 ( Thermo Fisher ) . After 24 h , cells were harvested and lysed by three freeze-thaw cycles . The lysates were diluted 10-fold and used to infect CEF monolayers . Fluorescent recombinant plaques were distinguished from the parental plaques and clonally purified five times . The purities of the recombinant viruses were confirmed by PCR amplification and sequencing of the modified region . MVA and recombinant viruses were propagated in CEF . CEF were grown in 12-well plates and infected with 0 . 001 or 0 . 01 PFU/cell of virus in MEM supplemented with 2 . 5% FBS for 2 h . The cells were washed extensively with the same medium , incubated at 37°C , and harvested at 48 h after infection . Harvested cells were lysed by 3 freeze-thaw cycles , and virus titers were determined by plaque assay on CEF monolayers . Virus samples were dispersed in a chilled water bath sonicator with two 30-s periods of vibration , followed by 10-fold serial dilutions in EMEM supplemented with 2 . 5% FBS . Diluted viruses were distributed onto CEF monolayers . After adsorption for 2 h , the medium was aspirated and replaced with medium containing 2 . 5% FBS and 0 . 5% methylcellulose . After 48 or 72 h , infected cells were fixed with methanol-acetone ( 1:1 ) , washed with water , and incubated with rabbit anti-VACV antibody ( 1:2 , 000 dilution ) for 1 h . The cells were washed again with water and incubated with a 1:3 , 000 dilution of protein A conjugated with peroxidase ( Thermo Scientific ) for 1 h . The cells were washed and incubated with the substrate dianisidine saturated in ethanol for 5 min . After color formation , the dianisidine solution was removed , and the cells were washed in water . A549 and MRC-5 cells expressing the 2xMyc tagged SPI-1 protein were created using retroviral transduction . A eukaryotic codon-optimized SPI-1 ORF with an N-terminal 2xMyc tag ( 2xMyc-SPI-1 ) was inserted into pQCXIP ( Clontech ) to generate pQCXIP-2xMyc-SPI-1 . Retrovirus particles were produced by co-transfecting pQCXIP or pQCXIP-2xMyc-SPI-1 ( transfer plasmid ) , pMLV-Gag-Pol ( packaging plasmid ) , and pVSV-G ( VSV-G envelope plasmid ) into 293T cells using Lipofectamine 2000 . A549 and MRC-5 cells were infected with the retroviruses in the presence of 5 μg/ml polybrene ( Sigma-Aldrich ) . The cells were subcultured and passaged several times in selection medium containing 1 μg/ml of puromycin ( Sigma-Aldrich ) . The expression of SPI-1 protein was determined by Western blotting using HRP-conjugated anti-Myc antibody ( 9E10 ) . Cells were harvested , washed , and lysed in Lysis buffer ( 20 mM Tris ( pH 7 . 4 ) , 150 mM NaCl , 2 mM EDTA , 1% Triton X-100 and protease inhibitor ) on wet ice for 30 min with frequent agitation . Cell lysates were cleared by centrifugation at 13 , 000 xg for 10 min at 4°C; the proteins were resolved on 4 to 12% NuPAGE Bis-Tris gels ( Thermo Fisher ) and transferred to a nitrocellulose membrane with an iBlot2 system ( Thermo Fisher ) . The membrane was blocked with 5% nonfat milk in Tris-buffered saline ( TBS ) for 1 h , washed with TBS with 0 . 1% Tween 20 ( TBST ) , and then incubated with the primary antibody in 5% nonfat milk in TBST overnight at 4°C . The membrane was washed with TBST and incubated with the secondary antibody conjugated with horseradish peroxidase ( Jackson ImmunoResearch ) in TBST with 5% nonfat milk for 1 h . After the membrane was washed , the bound proteins were detected with SuperSignal West Dura substrates ( Thermo Scientific ) . The cells were fixed , dehydrated and embedded in Embed 812 resin ( Electron Microscopy Sciences , Hatfield , PA ) as described previously [46] . Specimens were viewed with a FEI Tecnai Spirit transmission electron microscope ( FEI , Hillsboro , OR ) . Libraries for 454 pyrosequencing were made using Rapid Library Preparation Method Manual ( October 2009 ) GS FLX Titanium Series ( Roche , Branford , CT ) and Paired End Library Preparation Method Manual , 3kb Span ( October 2009 ) GS FLX Titanium Series . Each library was processed using emPCR Method , Manual Lib-L MV ( October 2009 ) in separate emulsion reactions . The paired-end sample was loaded on a single lane and the fragment sample was loaded in two lanes of an 8-region 454 GS FLX Titanium sequencing run . Genome assembly and gap closure was performed as previously described [44] . Sanger sequencing was used to correct errors due to runs of identical nucleotides . Genome sequences of MVA 44 . 1 , 44/47 . 1 , 47 . 1 and 51 . 2 were deposited as GenBank Accession Numbers MK314710 , MK314711 , MK314712 and MK314713 , respectively . The GenBank Accession Number MG663594 for MVA 51 . 1 was previously published [44] . | Poxvirus vectors have outstanding properties for development of vaccines against a myriad of infectious agents due to their ability to retain long segments of foreign DNA and high-level gene expression . Safety concerns led to a preference for attenuated poxviruses that lost the ability to produce infectious progeny in human cells . The most widely used poxvirus vector is modified vaccinia virus Ankara ( MVA ) , which exhibits an extreme host-range restriction in most mammalian cells . MVA was attenuated by passaging more than 500 times in chicken embryo fibroblasts during which large deletions and numerous additional genetic changes occurred . Despite ongoing clinical testing of MVA-vectored vaccines , the basis for its host-range restriction remained unknown . Here we show that re-introduction of the SPI-1 gene into MVA or host cells increased virus spread by more than 100-fold in a human diploid cell line , providing an important insight into the mechanism responsible for the host-range restriction . This information could help design improved vectors and develop non-avian cell lines for propagation of candidate MVA vaccines . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"poxviruses",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"viruses",
"virus",
"effects",
"on",
"host",
"gene",
"expression",
"dna",
"replication",
"dna",
"viruses",
"genome",
"analysis",
"dna",
... | 2019 | SPI-1 is a missing host-range factor required for replication of the attenuated modified vaccinia Ankara (MVA) vaccine vector in human cells |
Understanding the computations performed by neuronal circuits requires characterizing the strength and dynamics of the connections between individual neurons . This characterization is typically achieved by measuring the correlation in the activity of two neurons . We have developed a new measure for studying connectivity in neuronal circuits based on information theory , the incremental mutual information ( IMI ) . By conditioning out the temporal dependencies in the responses of individual neurons before measuring the dependency between them , IMI improves on standard correlation-based measures in several important ways: 1 ) it has the potential to disambiguate statistical dependencies that reflect the connection between neurons from those caused by other sources ( e . g . shared inputs or intrinsic cellular or network mechanisms ) provided that the dependencies have appropriate timescales , 2 ) for the study of early sensory systems , it does not require responses to repeated trials of identical stimulation , and 3 ) it does not assume that the connection between neurons is linear . We describe the theory and implementation of IMI in detail and demonstrate its utility on experimental recordings from the primate visual system .
To understand the function of neuronal circuits and systems , it is essential to characterize the connections between individual neurons . The major connections between and within many brain areas have been mapped through anatomical studies , but these maps specify only the existence of connections , not their strength or dynamics ( temporal properties ) . Measuring the strength and dynamics of the connection between two neurons requires physiological experiments in which the activity of both neurons is measured . The most direct of these experiments involves intracellular recordings , which allow the connection between the two neurons to be directly investigated . However , intracellular recordings are difficult to perform in vivo and impossible to obtain from more than a few cells at a time . Instead , most physiological studies of connectivity rely on extracellular recordings from multi-electrode arrays ( or , more recently , imaging of calcium activity ) . In these experiments , it is not usually possible to explicitly verify anatomical connectivity , nor to directly characterize the connections . Instead , the strength and dynamics of ‘functional’ connectivity must be inferred through statistical methods . The traditional method for characterizing the strength and dynamics of the connection between two neurons is the cross correlation function ( CXY ) , which measures the linear correlation between two signals over a range of specified delays [1] . While CXY and its variants have been used successfully in a number of studies ( see , for example , Usrey and Reid [2] for a review of many such studies in the visual system ) , it has limitations that must be considered when studying the connection between neurons [3]–[5] . The limitations of CXY arise from the fact that it is a measure of the total ( linear ) dependency between two signals and , thus , implicitly assumes that all dependencies between them are due to their connection . In the case of neurons , there are in fact many potential sources of dependency – shared external stimuli , intrinsic cellular and network properties , etc . – and CXY cannot disambiguate these dependencies from those due to the actual connection . Several modified versions of CXY have been proposed to address these drawbacks . For example , if neuronal activity in response to repeated trials of the same external stimulus is available for analysis , as is often the case in early sensory systems , the ‘shift-predictor’ can be used to remove some of the correlations due to the stimulus [1] . Further modifications to CXY have also been proposed to remove the correlations due to stimulus-driven covariations in activity [6] and background activity [7] . While these modified approaches have certainly improved upon the standard CXY , the confound of dependencies due to the connection and those arising from other sources remains a general problem . In addition to correlation-based methods , there are several other approaches to characterizing the dependency between two signals that can be used to study the connection between two neurons . These methods can be generally divided into two classes: model-based and model-free . The most common model-based approach to characterizing dependency is Granger causality ( GC ) [8] . With GC , one signal is predicted in two different ways: 1 ) using an autoregressive model based on its own past and 2 ) using a multivariate autoregressive model based on its own past and the past of the second signal . The strength of the dependency is given by the difference in the predictive power of the two models and the dynamics of the dependency are reflected in the regression parameters that correspond to the influence of the second signal . The power of model-based approaches such as GC is dependent on the validity of the underlying model; if the dependency between the two signals is approximately linear , then the characterization provided by GC will be accurate , but in situations where the properties of the dependency are complex or unknown , as is often the case with neurons , a model-free approach may be more appropriate . The most common model-free approach to characterizing dependency is transfer entropy ( TE ) , the information-theoretic analog of GC [9] . TE measures the reduction in the entropy of one signal that is achieved by conditioning on its own past and the past of the second signal relative to the reduction in entropy achieved by conditioning on its own past alone . TE is a powerful tool for measuring the overall strength of a dependency , but is not suitable for characterizing its dynamics . In this paper , we detail a new model-free approach for characterizing both the strength and dynamics of a dependency by ‘conditioning out’ the temporal correlations in both signals before assessing the strength of the dependency at different delays . This approach can overcome some of the confounds that are common in studies of neuronal connectivity [10]–[12] , as it has the potential to disambiguate statistical dependencies that reflect the connection between neurons from those caused by other sources ( e . g . shared inputs or intrinsic cellular or network mechanisms ) provided that the dependencies have appropriate timescales . In the following sections , we outline the theory behind our measure , which we call incremental mutual information , illustrate its usage on simulated neuronal activity and experimental recordings from the primate visual system , and consider its relationship to other common measures of dependence . Matlab code for measuring incremental mutual information is available for download at http://www . ucl . ac . uk/ear/research/lesicalab
In order to characterize the strength and dynamics of the connection between two signals , it is necessary to quantify how much one signal at one point in time influences the other signal at nearby points in time . Most measures of dependence between two signals X and Y seek to quantify the difference between the joint distribution p ( X , Y ) and the product of the marginal distributions p ( X ) p ( Y ) . For example , the cross correlation function measures the difference between the mean of the joint distribution and the product of the means of the marginal distributions ( the covariance ) , normalized by the product of the standard deviations for a given delay δ: ( 1 ) where CXY[δ] is the correlation coefficient between X[n] and Y[n] , which are assumed to be discretized signals , at integer delay δ . As described in the Introduction , CXY has limitations that are important to consider when studying neuronal connectivity . Most importantly , CXY , as with all dependency measures that operate only on the joint distribution p ( X , Y ) and the marginal distributions p ( X ) and p ( Y ) , cannot differentiate between the dynamics of the connection between the neurons and the temporal correlations in their activity that are due to other sources . It is possible to overcome this limitation by conditioning out the temporal correlations in each signal before measuring the dependency between them , i . e . rather than operate on p ( X , Y ) , p ( X ) , and p ( Y ) , operate on p ( X , Y| ) , p ( X| ) , and p ( Y| ) , where is a vector containing the past and future of X[n] and Y[n] relative to the delay of interest ( 2 ) as shown in the schematic diagram in figure 1 . The analog of CXY for conditional distributions is the partial cross correlation: ( 3 ) While CXY|Z overcomes the major limitation of CXY , it is still a linear measure and may not accurately characterize nonlinear dependencies . The idea of partial correlation can be generalized for the study of any dependency by formulating the information-theoretic analog of CXY|Z as a partial mutual information [13]: First , the entropy of X is measured after conditioning on its own past and future , as well as the past and future activity of Y relative to the delay of interest . Then , the strength of the influence of Y on X at the delay of interest can be measured as the additional reduction in entropy that occurs after observing Y at that delay: ( 4 ) Because this quantity , which we call the incremental mutual information ( IMI ) , reduces the uncertainty of X as much as possible before measuring the influence of Y at each delay , it has the potential to provide an accurate description of both the strength and dynamics of their dependency . In this form , ΔIXY is similar to a partial covariance in that its value is dependent on the properties of the individual signals ( e . g . the total entropy of X , the strength of the temporal correlations in X , etc . ) . In some cases , it may be preferable to use a normalized measure that is more similar to a partial correlation coefficient , i . e . a measure that expresses the incremental mutual information as a fraction of its maximum possible value: ( 5 ) To determine whether IMI is appropriate for use in any particular context , it is important to consider the relative timescales of the dependency between the signals and the other dependencies to be conditioned out . At any particular delay , the effects of dependencies with durations that are long relative to the time bins used for discretization will be predictable from the past and future values of the signals , so their contribution to the IMI will be small , i . e . dependencies with a slow timescale will make a relatively large contribution to initial reduction in the entropy of X based on past and future values of X and Y , , but not to the additional reduction in the entropy of X based on the present value of Y , . Conversely , the effects of dependencies that have a duration that is similar to the time bins used for discretization will not be predictable from the past and future values of the signals , so their contribution to the IMI will be large , i . e . dependencies with a fast timescale will make a small contribution to the initial reduction in entropy , but will make a large contribution to the additional reduction in entropy . Thus , IMI will be most useful when the duration of the dependency between the signals is similar to the size of the time bins used for discretization and the durations of the other dependencies to be conditioned out are longer . Fortunately , this is often the case for neurons in sensory systems , as will be illustrated in the examples in the Results . As with any measure based on entropies , the calculation of IMI requires careful consideration . Because IMI is a model-free approach , the number of samples required to produce a result of a given precision are likely to significantly exceed those of model-based approaches . The bias and variability of the entropy estimates that underlie the computation of IMI can vary substantially depending on the size of the data sample , the number of possible values that a signal can take on , and the signal dimensionality . Fortunately , neuronal activity typically has only a few possible values ( e . g . the number of spikes in each time bin ) . However , the terms , , , and representing the past and future of the signals are vectors . In practice , these vectors must be limited to some finite length , which we term ω , and this length will determine their dimensionality: ( 6 ) Thus , the calculation of IMI requires a tradeoff: increasing the value of ω allows the entropy of the first signal to be reduced as much as possible before measuring the influence of the second signal , but also increases the chances that the entropy estimates may be biased or highly variable . There are a number of bias correction techniques available that may be useful in mitigating problems related to sample size [14] . For the examples below , we corrected the entropy estimates using ‘quadratic extrapolation’ bias correction via the information toolbox software available at http://www . ibtb . org [15] . Also , for all of the examples below , time is discretized into sufficiently small bins such that each bin contains no more than one spike , limiting the possible values of X and Y to 0 and 1 . Because the bias and variability of entropy estimates are dependent on sample size , it is critical to establish the validity and precision of any calculation of IMI using statistical methods . In the experimental examples presented below , we use two different bootstrap procedures with random sampling to establish 95% confidence intervals and to determine whether the observed values are significantly different from zero . To establish 95% confidence intervals , we calculated IMI from 100 random samples of the same size drawn with replacement from the original sample . To preserve the temporal dependencies in the data , sampling was performed after the vectors were formed and the three vectors were sampled together . Confidence intervals were defined as the mean ± 2 standard deviations of the values calculated from the random samples . To establish the significance of the observed values , the same procedure was followed , but Y was sampled separately from . This sampling preserved the dependencies between , but removed the dependencies between X and Y ( and , thus , in theory , removed any IMI between them ) . The observed values were considered significantly different from zero if they were greater than 2 standard deviations above the mean of the values calculated from the random samples .
IMI is designed to give accurate measures of the strength and dynamics of the connections between neurons even in cases when the correlation may not , i . e . when the activities of individual neurons contain temporal correlations unrelated to the connection between them . In these cases , the cross correlation function can be ambiguous – its shape can reflect either the true dynamics of the connection , temporal correlations in the activities of the individual neurons , or a combination of both . A simple example of this ambiguity is illustrated in figure 2a . We first simulated a pair of neurons X and Y with independent , uncorrelated inputs and a dynamic connection , i . e . a spike from neuron Y caused a prolonged increase in the spiking probability of neuron X . We simulated the activity of neuron Y as a dichotomized Gaussian noise and the activity of neuron X as the dichotomized sum of a Gaussian noise and the filtered activity of Y: ( 7 ) where and are uncorrelated , ε = 0 . 5 is a scaling factor determining the overall strength of the connection , θ = 1 is the spiking threshold , and the input from Y to X , , is the convolution of the activity of Y with a Gaussian filter g[n] with a peak delay of 4 samples and a half width of 3 samples ( note that , the filtered version of Y , is unobserved ) . From the simulated activity of this pair of neurons ( with a sample size of 220 ) , we estimated the cross correlation function CXY and normalized incremental mutual information ( with ω = 2 ) at delays ranging from δ = −10 to 10 samples . Both CXY and for this pair were broad , reflecting the dynamics of the connection . We next simulated another pair of neurons that was similar to the first one , except that Υ received input with temporal correlations and the connection between Υ and X was static with a delay of 4 samples: ( 8 ) where is the convolution of sy with a Gaussian filter g[n] with a peak at zero delay and a half width of 3 samples . While CXY for this pair was also broad because of the temporal correlations in the activity of Y , was sharp , reflecting the static connection . Thus , while IMI captures the differences in the connections between these two pairs of neurons , correlation conflates connection dynamics with temporal correlations in individual activities and yields ambiguous results . This example can also be used to illustrate the necessity of conditioning out the both past and future activities of the neurons . A modified version of IMI can be formulated in which only the past activities of the two neurons are conditioned out: ( 9 ) In this formulation , the IMI is related to transfer entropy ( see Discussion ) . As shown in figure 2b , correctly conditions out the effects of the temporal correlations in the activity of Y for delays that are smaller than that of the actual connection , but not for delays that are larger than that of the actual connection . This reason for this asymmetry is as follows: Because of the temporal correlations in the activity of Y , its value will be similar for neighboring samples . When the delay of interest δ is smaller than the delay corresponding to the actual connection δ* , Y[n−δ*] is included in the vector of past activity and , since Y[n−δ] carries no information about X beyond that which is carried by Y[n−δ*] , Y[n−δ] makes no contribution to the IMI . However , when Y[n−δ*] is not included in the vector of past activities , Y[n−δ] , which is similar to Y[n−δ*] because of the temporal correlations in Y , will carry additional information about the activity of X and , thus , will contribute to the IMI . As a further consequence of the ambiguity in the cross correlation function illustrated in the example above , temporal correlations in individual activities may mask weak connections between neurons entirely . A simple example of this problem is shown in figure 3a . We simulated a pair of neurons that received a shared input with temporal correlations and had a weak static connection between them with a delay of 3 samples: ( 10 ) where and are the convolution of Gaussian noise with a Gaussian filter as described above with a correlation coefficient of 0 . 5 between them , and ε = 0 . 25 ( other parameter values are as described above ) . CXY for this pair of neurons was broad , with no discernable increase at the delay corresponding to the connection ( black arrow ) , while exhibits a sharp peak at the appropriate delay . Thus , by conditioning out dependencies due to shared input , IMI is able to reveal connections that may not be evident in the cross correlation function . A slight modification of the previous example can be used to illustrate a situation where shared inputs cannot be conditioned out and contaminate the IMI . As described above , IMI will be most useful when the duration of the dependency between the signals is similar to the size of the time bins used for discretization and the durations of the other dependencies to be conditioned out are longer , as is the case in example 2 . If the simulation in example 2 is modified so that the shared input is uncorrelated over time , the dependency resulting from the shared input can no longer be conditioned out , as the past and future activities of the neurons can no longer be used to infer the effects of the input at the delay of interest . As a result , has two peaks , one with no delay reflecting the shared input , and another with a delay reflecting the actual connection , as shown in figure 3b . It should be noted that this type of contamination can potentially arise both from shared external sources such as sensory stimuli as well as from other unobserved neurons . To test the utility of IMI on experimental data , we analyzed the activity of two pairs of thalamic relay neurons and their retinal ganglion cell ( RGC ) inputs recorded in the lateral geniculate nucleus ( LGN ) of an anesthetized monkey as shown in figure 4a . The details of the experimental procedures can be found in Carandini et al . [16] . During the recordings , visual stimulation was presented via an LED that illuminated the receptive field center with an intensity that varied naturally ( i . e . with temporal correlations typical of the natural environment ) . In this example , the stimulus was approximately 12 min in duration and did not repeat . The histograms in figure 4b show the basic relationship between the activity of the retinal and thalamic neurons in each pair . For the first pair , less than half of the RGC postsynaptic potentials ( PSPs ) evoked immediate LGN spikes , while the connection between the second pair was stronger , with nearly 75% of PSPs evoking immediate spikes . We calculated the cross correlation function and incremental mutual information for these pairs after binarizing the spike trains in 2 ms time bins . CXY for these pairs has a complex shape with 3 components: a broad positive peak with a half width of approximately 20 ms reflecting the temporal correlations in the visual stimulus , two sharp negative peaks reflecting refractory effects , and a sharp positive peak reflecting the actual connection between the cells . In contrast , for these pairs had one main peak reflecting the connection between the neurons - the effects of statistical dependencies arising from the stimulus correlations have been completely removed and the refractory effects have been largely conditioned out . For the first pair , had a relatively long tail , reflecting temporal summation of RGC PSPs that failed to evoke an immediate LGN spike . For the second pair , was sharper , reflecting the stronger connection between the cells . In early sensory systems , experiments are often designed such that the activity in response to repeated trials of an identical stimulus are observed so that the correlation between neurons can be separated into two distinct parts known as signal correlation and noise correlation . The signal correlation , which reflects both correlation in the stimulus itself and similarities in neurons' preferred stimulus features , will capture the correlation in the fraction of the response that is repeatable from trial to trial , i . e . the correlation that remains after the trial order has been randomized: ( 11 ) where Xi[n] is the response of neuron X on trial i and indicates the expectation over all possible combinations of trials i and j in which their values are not equal . In studies of neuronal connectivity , is often referred to as the ‘shift-predictor’ . The noise correlation , which results from network and intrinsic cellular mechanisms , will capture the remaining correlation in the fraction of the response that is variable from trial to trial ( 12 ) and , thus , captures the dependencies between the neurons that are not locked to the external stimulus . However , while may provide a better measure of the strength and dynamics of the connection between two neurons than CXY , it still confounds connection dynamics and temporal correlations that are independent of the stimulus , e . g . refractory effects or coupled oscillations . For comparison with and , the signal and noise IMI between X and Y can be formulated in an analogous fashion . The signal IMI is the reduction in the entropy of the response of X on trial i that results from observing the response of Y on trial j at the delay of interest , beyond that which results from observing the past and future responses of both neurons on trial i: ( 13 ) where . The noise IMI is the difference between the total IMI and the signal IMI , i . e . the reduction in the entropy of the response of X on trial i that results from observing the response of Y on trial i at the delay of interest and the past and future responses of both neurons on trial i , beyond that which results from observing the response of Y at the delay of interest on trial j and the past and future responses of both neurons on trial i: ( 14 ) We estimated the signal and noise correlations and signal and noise IMI for the same two retinogeniculate pairs that were analyzed in experimental example 1 using a different set of responses to 140 repeated trials of identical stimulation in which each trial was 5 seconds in duration . As shown in figure 5 , for both pairs was broad , reflecting the temporal correlations in the visual stimulus . In contrast , was nearly zero at all delays – because the temporal correlations in the visual stimulus were slow relative to the bin size used for discretization , there was little information about stimulus-induced dependencies to be gained by observing the RGC activity at any particular delay on a different trial when RGC and LGN activity at surrounding delays on the current trial were already known . While the effects of the stimulus correlations were removed from for both pairs , these functions still had a complex shape , with two negative peaks reflecting refractory effects , and one positive peak reflecting the actual connection between the neurons . Thus , while shuffling removed some of the confounding correlations in CXY , others still remained , while in , which has one main peak reflecting the connection between the neurons , most of the confounding dependencies have been conditioned out . This example illustrates an important property of IMI . as shown in figure 5 is nearly identical to for the same two pairs shown in figure 4 . Thus , unlike the cross correlation function , IMI does not require multiple trials in order to differentiate the temporal correlations in the responses of individual neurons from the dynamics of the connection between them .
The major determinant of the ability of IMI to differentiate connection dynamics from other dependencies is the relative timescale of the other dependencies . If the other dependencies have a long duration relative to the time bins used for discretization , then their effects can be conditioned out through observation of past and future neuronal activity , as demonstrated in the experimental examples presented above . If the other dependencies have a duration that is similar to the bin size , then their effects cannot be conditioned out without explicit observation of their source . As formulated here , IMI is designed to analyze the connection between a pair of neurons . However , in many brain areas , each neuron receives input from a large population , and correlations between these other inputs and the input under study could contaminate the IMI . If the other inputs are unobserved , it will be difficult to account for their effects with a model-free approach , though recent work with model-based approaches has demonstrated some success [17]–[20] . If the other inputs are observed ( which is becoming increasingly common with recent advances in recording and imaging technology that allow for simultaneous recording of the activity complete or nearly complete local populations of neurons ) , there is no reason that , in principle , IMI cannot be extended to condition out dependencies due to the activity of the other neurons . However , adding the activity of additional neurons to the conditioning vector will increase its dimensionality , and , thus , the bias and variability of the entropy estimates that underlie the computation of IMI . While this may not be a problem for a small number of neurons , it is certain to be a problem for large populations . Thus , for large populations , it may be more appropriate to use a model-based approach such as Granger causality within a generalized linear model framework [21] . Of the existing approaches to characterizing dependencies between signals , IMI is most similar to transfer entropy [9] . TE measures the dependency between two signals as the difference in the entropy of one signal after conditioning on its own past and conditioning on its own past and the past of the other signal , or , in the terminology used to define IMI , . From this definition , it is clear that TE and IMI are designed for different purposes: TE measures the overall causal strength of the dependency between two signals by first conditioning out the past of one signal and then measuring how much can be learned about the present value of that signal based on the past of the second signal , while IMI measures the strength and dynamics of the dependency between two signals by first conditioning out past and future of both signals and then measuring how much can be learned about the present value of one signal from the present value of the other relative to some delay . The key difference between TE and IMI , as illustrated in the simulated example presented above , is that , even if computed at a range of delays , TE is not suitable to assess the dynamics of a dependency because it considers only past activity and , as a result , conditions out temporal correlations appropriately for delays that are shorter than that of the actual dependency , but not for delays that are longer than that of the actual dependency . The most effective model-based approach for studying the functional connectivity in a neuronal circuit is the generalized linear model ( GLM ) [22]–[24] . The GLM attempts to predict a neuron's activity based not only on its own activity and the activity of other neurons , but also on external inputs . Because all of the filters in the model are fit simultaneously , the influence of the external inputs on the activity of each neuron , as well as those of its own past activity , are separated from the influence of other neurons . The power of the GLM lies in the fact that once the filters have been estimated , the model can be used to predict the activity of the entire group of neurons to any external input , but this power comes at the expense of assuming a particular parametric structure . Relative to IMI , which makes no assumptions about the connections between neurons , the drawback of the GLM is that the interactions between neurons are assumed to be of a particular nature ( usually additive ) . However , this assumption also allows the GLM to be readily applied to large populations . | The root of our brain's computational power lies in its trillions of connections . With our increasing ability to study these connections experimentally comes the need for analytical tools that can be used to develop meaningful quantitative characterizations . In this manuscript , we present a new such tool , incremental mutual information ( IMI ) , that enables the characterization of the strength and dynamics of the connection between a pair of neurons based on the statistical dependencies in their spiking activity . IMI is an important step forward from existing approaches , as it has the potential to disambiguate dependencies due to the connection between two neurons from those due to other sources , such as shared external inputs , provided that the dependencies have appropriate timescales . We demonstrate the utility of IMI through the analysis of simulated neuronal activity as well as activity recorded in the primate visual system . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"neuroscience/theoretical",
"neuroscience",
"neuroscience/sensory",
"systems"
] | 2010 | Incremental Mutual Information: A New Method for Characterizing the Strength and Dynamics of Connections in Neuronal Circuits |
During meiosis , chromosomes undergo a homology search in order to locate their homolog to form stable pairs and exchange genetic material . Early in prophase , chromosomes associate in mostly non-homologous pairs , tethered only at their centromeres . This phenomenon , conserved through higher eukaryotes , is termed centromere coupling in budding yeast . Both initiation of recombination and the presence of homologs are dispensable for centromere coupling ( occurring in spo11 mutants and haploids induced to undergo meiosis ) but the presence of the synaptonemal complex ( SC ) protein Zip1 is required . The nature and mechanism of coupling have yet to be elucidated . Here we present the first pairwise analysis of centromere coupling in an effort to uncover underlying rules that may exist within these non-homologous interactions . We designed a novel chromosome conformation capture ( 3C ) -based assay to detect all possible interactions between non-homologous yeast centromeres during early meiosis . Using this variant of 3C-qPCR , we found a size-dependent interaction pattern , in which chromosomes assort preferentially with chromosomes of similar sizes , in haploid and diploid spo11 cells , but not in a coupling-defective mutant ( spo11 zip1 haploid and diploid yeast ) . This pattern is also observed in wild-type diploids early in meiosis but disappears as meiosis progresses and homologous chromosomes pair . We found no evidence to support the notion that ancestral centromere homology plays a role in pattern establishment in S . cerevisiae post-genome duplication . Moreover , we found a role for the meiotic bouquet in establishing the size dependence of centromere coupling , as abolishing bouquet ( using the bouquet-defective spo11 ndj1 mutant ) reduces it . Coupling in spo11 ndj1 rather follows telomere clustering preferences . We propose that a chromosome size preference for centromere coupling helps establish efficient homolog recognition .
Processes in meiosis are geared to recombine homologous chromosomes to both increase genetic diversity , and segregate them efficiently thus producing viable gametes for sexual reproduction . In the absence of recombination ( as in a spo11 diploid cell [1] ) , chromosomes fail to homologously align , yet the two chromosomal divisions still occur generating highly aneuploid spores . Homologous pairing and recombination between chromosomes favor the formation of stable pairs [2 , 3] , which are secured by the proteinaceous synaptonemal complex ( SC ) , containing ZMM proteins such as Zip1 [4] . In addition to holding homologs in alignment during meiotic prophase I , the SC is also implicated in crossover formation [5] . Two dynamic homology-independent events precede homolog pairing: the meiotic bouquet and non-homologous centromere coupling . The meiotic bouquet is formed through clustering of telomeres , when they become embedded in a small section of the nuclear envelope [6 , 7] . The bouquet persists when meiotic cohesin Rec8 is absent [8] . The bouquet represents a transition from a Rabl configuration , with clustered centromeres close to spindle pole body , to a reverse Rabl configuration during the bouquet stage . The bouquet undergoes rapid telomere-led movements requiring Ndj1 [9 , 10] , as well as Csm4 , Mps3 , and actin [11–13] . Bringing telomeres to the nuclear envelope is accomplished mostly by Ndj1 [14] , while clustering and rapid movements are more Csm4-dependent [11 , 14] . Rapid prophase movements have been shown to destabilize recombination [11] and to contribute to the generation of heterologous and homologous collisions between centromeres for pairing [15] . During the second homology-independent event prior to homolog pairing , “centromere couples” are formed by the transient association of non-homologous chromosomes at their centromeres [16 , 17] . Couples are dispersed throughout the nucleus at this stage [16] , and an uncoupling mechanism must exist to ensure homolog pairing ensues; a likely candidate for such mechanism is the phosphorylation state of the SC protein Zip1 [18] . The non-homologous centromere associations are proposed to provide a path for a chromosome to find its homolog , as transient non-homologous couples are replaced by stable homologous pairs as pairing , recombination and SC formation progress in a timely fashion [16] . Meiotic non-homologous centromere associations have been described in many model organisms , including yeasts , flies , plants and mammals [19] . In mice , the inability to observe complete coupling suggests that it might be either very short-lived or partial [20 , 21] . Studies of centromere coupling in Saccharomyces cerevisiae have demonstrated its independence on recombination ( as in a spo11 diploid ) and on the presence of homologous chromosomes ( as in spo11 haploids undergoing a forced meiotic induction ) [16] . Centromere coupling is dependent on the SC component Zip1 [16 , 17] and some requirements regarding the regulation of complete centromere coupling have started to emerge , such as roles for the meiotic cohesin Rec8 [22] , for the SC component Zip3 in coupling and tethering [16 , 23] , and for the phosphorylation of Zip1 by ATM/ATR DSB checkpoint kinases [18] . However , the underlying architecture of centromere coupling remains to be understood . In particular , the presence of an interaction pattern of centromeres , if any , might point towards an intrinsic mechanism for coupling . So far previous studies have relied on low-scale , traditional approaches not amenable to testing this hypothesis on a larger level . The budding yeast genome , despite its compact size , exhibits a high level of inter-chromosomal contacts and long-range cis interactions between distant loci [24] . Chromosome Conformation Capture ( 3C ) enables the detection of DNA regions in close nuclear proximity through formaldehyde crosslinking of such interactions followed by restriction enzyme digestion , dilute ligation to favor intra-molecular products that are crosslinked , and PCR detection [25] . 3C was first developed in budding yeast to study chromosome dynamics during meiosis and higher-order chromatin organization [25] , and has since been applied the investigation of diverse biological processes such as silencing [26] , organization of the pericentric chromatin [27] , and gene looping [28 , 29] . 3C has yielded several related techniques that have enabled the characterization of long-range genome associations in mammals [30–34] . One such variant , Taqman-based 3C-qPCR , is well suited for focused studies , with high sensitivity and dynamic range , low background and quantitative detection of interacting fragments [32] . Here we present the first multiple pairwise characterization of centromere coupling . We modified and combined the yeast 3C protocol [35 , 36] with Taqman-based real-time detection of 3C ligation products ( 3C-qPCR ) [32] to quantify all possible non-homologous interactions between the 16 centromeres ( CENs ) of S . cerevisiae during meiosis . We observed a non-random CEN interaction pattern based on similarity of chromosome sizes in strains capable of coupling ( spo11 diploids and haploids ) , which is absent in coupling-deficient strains ( spo11 zip1 diploids and haploids ) . Importantly , these size-dependent preferential contacts are present at early time points in normal meiosis ( WT diploids ) , prior to pachytene and full homolog pairing . We also found a role for the meiotic bouquet in pattern establishment , with bouquet absence ( spo11 ndj1 ) associated with decreased size dependence . From our results , we propose that centromere coupling , with its preference for chromosomes of similar size , helps chromosomes find their homolog .
We used a modified 3C-qPCR assay to specifically look at interactions between non-homologous centromeres . Each of the sixteen similarly-sized centromere regions are defined by restriction enzyme sites . Two primers were designed for each centromere region , one on each side of the restriction fragment oriented towards the enzyme recognition site ( Fig 1A ) . Taqman probes , which allow quantitative detection by real-time qPCR , were synthesized on each side of the CEN fragment , closer to the restriction enzyme cutting site than the primer annealing site ( Fig 1A ) . High-Fidelity EcoRI ( EcoRI-HF ) was used as a suitable restriction enzyme for 3C . However , many EcoRI restriction sites were far from the CEN ( Fig 1B , left panel; 15/32 sites > 2 kb away from CEN ) , generating fragments which varied in size considerably . Large variations in size might create biases during intra-molecular ligation , favoring the preferential recovery of certain interacting pairs over others [37] . To circumvent this potential issue , we incorporated a double digestion step ( 3C2D ) with the high-fidelity MfeI restriction enzyme ( MfeI-HF ) , generating compatible cohesive ends with EcoRI while recognizing a different consensus site . The 3C2D modification resulted in a more even distribution of restriction site distances from the CENs ( Fig 1B , right panel; 2/32 sites > 2 kb away from CEN ) and centromeric fragments of less variable size ( S1 Fig ) . This experimental design enables the quantification of 480 distinct centromeric interactions , or all 120 possible combinations of non-homologous couples . To test our 3C2D-qPCR protocol , we isolated genomic DNA from haploid and diploid yeast cells to generate control libraries for 3C , which consist of non-crosslinked , EcoRI-MfeI digested genomic DNA that is randomly ligated [35 , 36] . These control samples aim to contain all possible ligation products in near equimolar ratios [25] and serve to test PCR efficiencies of different combinations of primers and Taqman probes [32 , 38] . All 480 interactions were compared in haploid and diploid control libraries , looking at the average enrichment ( average number of qPCR cycles ) across multiple dilutions and replicates for each combination . A majority of combinations have enrichments within 1 qPCR cycle ( ~2-fold ) of the average enrichment across all possible combinations for haploid controls and diploid controls ( 56% for haploids , 54% for diploids; S2 Fig ) . For the same combination , there is a high correlation when comparing enrichments between diploid and haploid controls , and 71% lie within 1 qPCR cycle ( ~2-fold ) ( Pearson’s r = 0 . 72 , p<10−15; S3 Fig ) . Overall , we find that different combinations of primers and Taqman probes perform similarly . We next analyzed all possible non-homologous CEN interactions in a coupling-proficient strain by generating a 3C experimental sample from a spo11 diploid [16] , which consists of EcoRI-MfeI digested , crosslinked chromatin that is ligated in dilute conditions . Non homologous couples in the recombination deficient spo11 mutants are stable , since homologous pairing does not occur . As a negative control , interactions were also characterized in coupling-deficient spo11 zip11 diploid [16] . Cells were harvested 14h after meiotic induction , a time point where most spo11 cells have dispersed the centromere cluster into 16 distinct CEN foci ( from 32 chromosomes marked by kinetochore component Ctf19 ) as determined by immunofluorescence microscopy on meiotic chromosome spreads [16 , 39]; in spo11 zip1 , the centromere cluster gives rise to 32 CEN foci . In the case of non-homologous centromere coupling , if certain inter-chromosomal centromeric fragments couple more frequently than other combinations , then they would become crosslinked and subsequently ligated at higher frequencies than less-interacting CENs . As a control to ensure that the 3C experimental libraries are enriched for fragments with spatial proximity , we compared amplification of intra-chromosomal proximal fragments ( 10 kb away ) and distal fragments ( 80 kb away ) on the same chromosome ( S4A Fig ) . In the un-crosslinked control libraries , proximal and distal fragments have similar interaction frequencies ( randomly-ligated genomic DNA in equimolar proportions ) ( S4B Fig ) . For the 3C experimental samples , a higher interaction frequency between proximal fragments than distal fragments is observed ( S4B Fig ) , confirming that we can detect preferential crosslinking and ligation of restriction fragments closer in the nucleus . We analyzed 480 non-homologous combinations in a spo11 diploid and in a spo11 zip1 diploid using 3C2D-qPCR . Interaction frequencies between non-homologous centromeres were plotted on a heatmap after normalization ( Fig 2A for spo11 diploid and Fig 2B for spo11 zip1 diploid ) . For each chromosome , the 15 non-homologous chromosomes were ranked according to the strength of their CEN interaction ( S5 Fig for spo11 diploid and S6 Fig for spo11 zip1 diploid ) . In the case of the spo11 diploid library , we observed a non-random interaction pattern during centromere coupling , with centromeres of smaller chromosomes interacting preferentially with those from small chromosomes ( Fig 2A and S5 Fig ) . In brief , centromeres interact with centromeres from liked-size chromosomes more frequently . To test the significance of this relationship , we asked the following: do the top three CENs with the highest interacting frequencies happen to be the three closest chromosomes in length more often than random ? By performing a non-parametric permutation test to generate a randomized matrix , representing what level of association is expected by chance alone , we found that this chromosome size interaction pattern was present in coupling-proficient spo11 diploids ( p < 0 . 01 ) , but not in coupling-defective spo11 zip1 diploids ( p > 0 . 10 ) . We plotted a normalized interaction score of all possible interaction frequencies , binned in 5 categories according to chromosome size similarity ( bin 1–3 for the 3 chromosomes most similar in size , … , bin 13–15 for the 3 chromosomes most dissimilar in size ) . A positive value indicates an increased frequency of interactions compared to the average level of interaction for that particular genotype , and a negative value indicates fewer interactions than average . Couples between one chromosome and its three chromosomes most similar in size ( bin 1–3 ) are overrepresented in spo11 diploids , but not in spo11 zip1 diploids ( Fig 2C ) . Examples for coupling partners most similar or dissimilar in size to a short , medium and large chromosome are presented in Fig 2D . In spo11 diploids , the interaction pattern was identical for the four shortest chromosomes and the four largest chromosomes . We performed a sensitivity analysis for our model , considering a ) the three most interacting CENs and the five closest chromosomes in length , b ) the five most interacting CENs and the three closest chromosomes in length , and c ) the five most interacting CENs and the five closest chromosomes in length . In all cases , the pattern was statistically significant in spo11 diploids ( p < 0 . 05 ) , but not in spo11 zip1 diploids ( p > 0 . 10 ) . We compared the mean raw cycle numbers ( +/- standard error of the mean ( SEM ) ) between spo11 and spo11 zip1 diploids as an estimate of the total number of interactions , with a smaller cycle value representing a quicker qPCR amplification to a detectable level above background , which is directly related to the abundance of a particular couple among all 3C DNA ligation products . We observed a ~ 24-fold difference between spo11 diploids ( 32 . 64 +/- 0 . 30 ) and spo11 zip1 diploids ( 37 . 21 +/- 0 . 34 ) ( enrichment = difference of 4 . 57 on a log2 scale ) . Differences in raw interaction frequencies between spo11 and spo11 zip1 diploids are plotted as a heatmap in S7 Fig . These results suggest that coupling is absent in spo11 zip1 , consistent with previous imaging studies [16 , 17] . Although a highly dynamic process , where couples form but are not maintained , cannot be ruled out from these genomic data , it is unlikely . Since DNA sequence homology on chromosome arms has a role in homolog pairing [3] , preferential interactions between non-homologous couples might derive from large blocks of homologous regions on arms . For all 16 chromosomes of Saccharomyces cerevisiae , we considered the extent of homologous regions between pairs of S . cerevisiae chromosomes that had apparently derived from a single chromosome of the reconstructed yeast ancestor prior to whole genome duplication [40] . Following a similar non-parametric testing procedure to generate a randomized matrix of interaction frequencies , we found that sequence homology on chromosome arms cannot explain the pairwise coupling pattern observed in spo11 diploids ( p > 0 . 05 ) . It is also possible to align pairs of centromeres with optimal homology based on ancestry by identifying the two Saccharomyces cerevisiae centromere regions related to a single centromere region from Kluyveromyces waltii , a budding yeast that diverged from the Saccharomyces lineage prior to whole-genome duplication [41 , 42] . We asked whether ancestral centromere homology was the mechanism for centromere coupling , with the strongest interactions between the two ancestral centromeres ( CEN1-CEN7 , CEN2-CEN4 , CEN3-CEN14 , CEN5-CEN9 , CEN6-CEN16 , CEN8-CEN11 , CEN10-CEN12 , CEN13-CEN15 ) [42] . We found that , for each centromere , the strongest interacting partner was never its ancestral sister ( Fig 2E ) . In budding yeast , haploid cells , lacking any homologues , can be forced to undergo a meiotic induction by expressing the opposite mating type cassette from an ectopic locus [6] . Haploids exhibit centromere coupling , forming 8 CEN couples that are de facto non-homologous , from the 16 chromosomes [16] . As in diploids , coupling is abolished in the absence of Zip1 [16] . We wondered whether preferential interactions during centromere coupling also follow a chromosome size-dependent pattern in the absence of potential interactions with homologous chromosomes . We repeated the multiple pairwise 3C2D-qPCR analysis to detect all possible centromeric interactions in coupling-proficient spo11 haploids and in coupling-deficient spo11 zip1 haploids . Cells were harvested 20h after meiotic induction , a time point where most cells contain 8 CEN foci ( from 16 chromosomes marked by kinetochore component Ctf19 ) as determined by immunofluorescence microscopy on meiotic chromosome spreads [16 , 39] . Interaction frequencies between non-homologous centromeres were plotted on a heatmap after normalization ( Fig 3A for spo11 haploid and Fig 3B for spo11 zip1 haploid ) . Again all 15 chromosomes were ranked by the strength of their CEN interaction for any given chromosome ( S8 Fig for spo11 haploid and S9 Fig for spo11 zip1 haploid ) . Similarly to spo11 diploids , spo11 haploids have preferential interactions based on comparable chromosome sizes ( Fig 3A and S9 Fig ) . For each chromosome , there is a strong , non-random , bias towards preferential interactions with chromosomes of similar sizes in spo11 haploids ( top three CENs closest in length ) ( p < 0 . 01 ) . Again this pattern was absent in coupling-deficient spo11 zip1 haploids ( p > 0 . 10 ) and ancestral centromere homology does not play a role in haploids to establish the interaction pattern ( Fig 3E ) , with only CEN8 having its strongest interacting partner as its ancestral sister CEN11 . Note that Chromosome 11 is one of the chromosomes most similar in size to chromosome 8 ( 667 kb vs . 563 kb , respectively ) . Using a normalized interaction score of all possible interaction frequencies binned in 5 categories according to chromosome size similarity , couples between one chromosome and its three chromosomes most similar in size ( bin 1–3 ) are overrepresented in spo11 haploids , but not in spo11 zip1 haploids ( Fig 3C ) . Examples for coupling partners most similar or dissimilar in size to a short , medium and large chromosome are presented in Fig 3D . Of note , we observed a strong level of interaction with chromosome 11 for both spo11 and spo11 zip1 haploids ( horizontal line ) . We do not have a biological explanation for such a pattern . However , when looking at the raw interaction frequency differences between spo11 and spo11 zip1 haploids ( S10 Fig ) , this horizontal line became attenuated . Moreover , using a randomization approach , we determined that , in spo11 haploids , only the four shortest chromosomes had preferential interactions based on size similarity ( p < 0 . 05 ) , but not the four largest chromosomes ( p > 0 . 10 ) . Similarly to diploids , using the mean raw cycle numbers ( +/- SEM ) as an estimate of the total amount of interactions , spo11 haploids had on average ~17-fold more interactions than spo11 zip1 haploids ( 30 . 25 +/- 0 . 22 for spo11 haploids vs . 34 . 30 +/- 0 . 22 for spo11 zip1 haploids ) ( enrichment = difference of 4 . 05 on a log2 scale ) . Moreover , haploids displayed more coupling interactions than diploids ( ~ 5–8 fold difference in cycle numbers ) , consistent with obligate non-homologous interactions in haploids . Both spo11 haploids and diploids have preferential coupling interactions based on similarity in chromosome size . Overall 3C2D-qPCR profiles show a statistically significant agreement , whether using raw interaction frequencies ( R2 = 0 . 22; p < 0 . 001 ) or ranked data ( ρ = 0 . 27; p < 10−4 ) . This agreement between both datasets tends to be greater in the upper third of the lists , for the top 5 interacting pairs ( 44% overlap , p = 0 . 01 , non-parametric resampling test ) . We then asked whether a strong coupling interaction determined by our genomic approach could be confirmed using immunofluorescence microscopy . According to our 3C2D-qPCR data , two centromeres from small chromosomes , CEN1 and CEN3 , would form non-homologous couples preferentially . We tested this possibility by constructing a haploid spo11 strain with a lacO array at CEN3 , a TetO array at CEN1 , LacI-GFP and TetR-mCherry . Meiotic spreads from this strain were used to assess CEN1/CEN3 colocalization by immunofluorescence . A control chromosome pair was chosen from our genomic data . CEN3 , from a short chromosome , and CEN5 , from a medium-sized chromosome , do not appear to interact preferentially above a low-medium background level . We constructed an isogenic strain but with the TetO array at CEN5 and quantified CEN3/CEN5 couples by immunofluorescence . Chromosome spreads were first screened with DAPI to ensure proper spreading [39] . Among those spreads , the kinetochore arrangement was assessed with Ctf19 immunostaining for ~8 Ctf19 foci ( centromeres ) ( Material and Methods ) [16] . As predicted from our 3C2D-qPCR genomic analyses , CEN3 associates more often with CEN1 than with CEN5 on individual meiotic spreads ( Fig 4B; p < 10−7 , Fisher’s Exact Test ) . Indeed , CEN1 and CEN3 were coupled on average in 40% of meiotic spreads ( range = 38–42% ) , while CEN3 and CEN5 were coupled on average in 12 . 7% of spreads ( range = 10–16% ) , a more than 3-fold increase for the predicted interaction between CEN1 and CEN3 based on the similarity of their chromosome lengths . In a wild-type ( WT ) diploid yeast , non-homologous centromere couples are transient , present early in meiosis and gradually replaced by stable homologous pairs by pachytene [16] . Using our genomic assay , we next asked whether coupling interactions occurring in a wild-type yeast show the same size preference for partners . WT cells were harvested at multiple time points after meiotic induction ( 8h , 9h , 10h , 11h , and 14h ( pachytene ) ) . Immunofluorescence was performed to monitor meiotic progression from centromere organization ( Ctf19 ) and the appearance of SC components ( Zip1 and Red1 ) ( S11 Fig ) . Interaction frequencies between centromeres were plotted on a heatmap after normalization ( S12A Fig ) . For each chromosome , the 15 non-homologous chromosomes were ranked according to the strength of their CEN interaction ( S12B Fig ) . Early in meiosis ( 8h , 9h time points ) , centromeres interact preferentially based on similarity of chromosome sizes , as found in spo11 diploids and haploids ( S12 Fig; top three chromosomes closest in length: p < 0 . 01 ) . Later in meiosis this pattern becomes less significant at 10h and 11h ( S12 Fig; p > 0 . 10 for 10h and p = 0 . 064 for 11h ) and achieves its lowest value at 14h/pachytene ( S12 Fig; p > 0 . 10 ) . In budding yeast , the shortest chromosomes are the last ones to pair with their homolog and synapse [15] . Extrapolating this result to non-homologous coupling , one would then expect that interactions involving larger chromosomes would decrease and that the proportion of interactions between small chromosomes would increase . As meiosis progresses ( from 8h to 14h ) , our genomic data suggest that this is the case . Indeed , using a randomization procedure , we observed that , for the final time point in our WT yeast ( late; 14h ) , the interaction pattern based on chromosome size is very strong for the four shortest chromosomes ( p < 0 . 01 ) , and completely absent for the four largest chromosomes . For example , interaction frequencies between chromosomes 1 and 3 ( short chromosomes ) are increased while those between chromosomes 4 and 12 ( largest chromosomes ) decrease ( Fig 5A ) . For simplifying a comprehensive analysis , we combined meiotic time points 8h and 9h as “early” , and 10h and 11h as “mid” . Time point 14h is considered “late” . Differences of normalized interaction frequencies between non-homologous centromeres were plotted on a heatmap to compare their relative progression ( Fig 5B , 5C and 5D for early→mid , mid→late , and early→late; red = relative increase , blue = relative decrease ) . When considering heatmaps involving the late time point , there is a relative increase in the interactions involving smaller chromosomes , most obvious with chromosome 1 , and a relative decrease when longer chromosomes are involved , especially for chromosome 12 , albeit less striking in general . Additionally , by ranking interactions frequencies for all 120 unique combinations of couples , i . e . chr1-2…chr1-16 , chr2-3 …chr2-16 , … , chr15-16 ( couple with most interactions = 1 , couple with least interactions = 120 ) , we observed that the average rank of all 15 interactions involving chromosome 1 ( smallest chromosome ) became closer to the first ranks , from 34 ( early ) to 28 ( mid ) to 16 ( late ) ( out of 120 ) , while that of all interactions involving chromosome 12 ( largest chromosome ) became closer to the last ranks , from 77 to 80 to 83 ( out of 120 ) ( Fig 5E ) . This trend towards the top ranked couples for non-homologous interactions with CEN1 signifies that CEN1 continues to undergo active coupling . Moreover , we typically observe more coupling interactions for CEN1 than CEN12 , with a difference in their mean enrichment ratios approximately 8-fold for early and mid time points ( Fig 5F ) . This difference in mean enrichment ratios jumps from 8-fold to 78-fold at the late time point ( Fig 5F ) , revealing that coupling interactions with smaller chromosomes become more prevalent as meiosis proceeds while non-homologous contacts between large chromosomes decrease . This supports the observation that longer homologous chromosomes pair first [15] , presumably because of larger blocks of homology on chromosome arms , or more numerous potential pairing blocks . As such , long chromosomes would be taken out of the coupling pool earlier , leaving mostly short chromosomes engaged in non-homologous interactions , in search for their homolog . A unique limitation of this experiment , during which coupling is analyzed in wild-type yeast cells , is their asynchronous entry into meiosis . Wild-type yeasts from the BR1919-8B background enter meiosis fairly asynchronously [43] . Our cytological analysis in wild-type cells ( S11 Fig ) revealed that no more than ~ 75% of the cells at each time point , but more than ~ 50% , are in the same stage . This heterogeneity of cells in different stages likely contributes to noise present in the interaction frequency heat maps , and demand caution in interpreting the results . However , despite the noise resulting from the isolation of cells at different steps in meiosis , we were able to identify an interaction pattern based on size and confirm previous findings about earlier pairing of larger chromosomes [15] . In contrast , asynchronous entry into meiosis is not an issue for the remainder of the experiments performed in a BR1919-8B spo11 background . In diploid and haploid strains lacking Spo11 , centromere coupling persists through prophase for several hours [16 , 17] . Studies performed in the same BR1919-8B spo11 background , at similar time points for cell collection than this study , found that centromeres formed distinct foci in ~ 95% of diploid spo11 cells and haploid spo11 cells ( ~5% of cells with clustered centromeres ) [22 , 44] . Similarly , aliquots taken as cells were harvested from our large cultures of various spo11 strains showed that centromeres formed multiple distinct foci ( separated/coupled ) in > 80% of cells ( median 91 . 4% ) ( S13 Fig ) . Thus , in contrast to wild-type BR1919-8B cells , spo11 BR1919-8B are minimally influenced by asynchronous entry into meiosis , as they remain in a state with centromeres forming distinct foci for an extended period of time . Given the chromosome size-dependent preferential interactions we observed , a possible mechanism to help in establishing this interaction pattern could be bouquet formation . Early in zygotene , chromosomes associate non-homologously at their telomeres in a small region of the nuclear envelope , forming the meiotic bouquet [6 , 7] . Bouquet formation is disrupted in ndj1 mutants [7 , 9 , 10] and persists in rec8 mutants [8] . Centromere coupling has been previously assessed by microscopy approaches in strains with altered bouquet formation . Bouquet formation was found to be dispensable for centromere coupling , given that spo11 ndj1 diploids form no bouquet but still had ~16 CEN foci , as did coupling-proficient spo11 diploids [16] . On the other hand , immunofluorescence data suggest that only 23% of spo11 rec8 diploid cells undergo non-homologous coupling ( 16–20 CEN foci ) [22] , arguing that spo11 rec8 diploids display at most partial coupling . The coupling defect observed in spo11 rec8 diploids is likely due to a reduction in Zip1 loading around centromeres , in particular on cohesin-rich pericentromeric regions [22] . Using the high sensitivity of our 3C2D-qPCR method for assessing specifically non-homologous centromeric interactions , we first tested the hypothesis that the size-dependent pairwise pattern would be absent ( or decreased ) in bouquet-deficient spo11 ndj1 diploids . Interaction frequencies between non-homologous centromeres were plotted on a heatmap after normalization ( Fig 6A for spo11 ndj1 diploids ) . For each chromosome , the 15 non-homologous chromosomes were ranked according to the strength of their CEN interaction ( S14 Fig for spo11 ndj1 diploids ) . Consistent with a role for bouquets in size establishment , the chromosome size-dependent pattern was absent when the bouquet was abolished in spo11 ndj1 diploids ( Fig 6A and S14 Fig; top three chromosomes closest in length: p > 0 . 10 ) . In normalized interaction score plots , spo11 ndj1 diploids do not show preference for liked-size chromosomes ( Fig 6C ) . In spo11 ndj1 diploids , there is a ~38-fold increase in the raw interaction levels estimated by raw cycle numbers compared to spo11 zip1 diploids ( 31 . 95 +/- 0 . 35 for spo11 ndj1 vs . 37 . 21 +/- 0 . 34 for spo11 zip1 ) ( enrichment = difference of 5 . 26 on a log2 scale ) , a similar increase as observed in spo11 diploids , which is consistent with robust coupling [16] . Despite the fact that spo11 rec8 diploids undergo at most partial coupling , i . e . coupling in a minority of cells [22] , we asked whether we could detect non-homologous coupling interactions in those cells , taking advantage of the sensitivity and specificity of our 3C2D-qPCR assay . In spo11 rec8 diploids , interactions are reduced by ~6–9 fold compared to coupling-proficient strains ( 35 . 13 +/- 0 . 31 for spo11 rec8 vs . 31 . 95 +/- 0 . 35 for spo11 ndj1 or 32 . 64 +/- 0 . 30 for spo11 ) ( enrichments = differences of 3 . 18 and 2 . 49 on a log2 scale ) , but are increased ~4-fold compared to spo11 zip1 ( 37 . 21 +/- 0 . 34 ) ( enrichment = difference of 2 . 08 on a log2 scale ) . This is in accordance with previous data showing a defect in coupling in spo11 rec8 diploids [22] . Similar to spo11 diploids , spo11 rec8 diploids show a significant bias towards interactions between chromosomes of similar length ( Fig 6B and S15 Fig; top three chromosomes closest in length: p < 0 . 01 ) . In normalized interaction score plots , looking at bins 1…3 and 4…6 , spo11 rec8 diploids display a robust chromosome size-dependent pattern ( Fig 6C ) . This suggests that the size-dependent pairwise pattern is not disrupted in bouquet-persisting spo11 rec8 diploids . Uniquely , for spo11 rec8 diploids , a significant decrease in CEN interactions between chromosomes of most dissimilar length ( e . g . small vs . large ) is seen . To test the significance of this relationship based on dissimilarity of chromosome lengths , we performed a non-parametric permutation test similar to the one previously used for similarity of sizes: do the last three CENs with the lowest interaction frequencies happen to be the three chromosomes most dissimilar in chromosome lengths more often than expected by chance ? This avoidance of coupling interactions between chromosomes of most dissimilar lengths was found in spo11 rec8 diploids ( p < 0 . 01 ) , but not in spo11 , spo11 ndj1 or spo11 zip1 diploids ( p > 0 . 10 ) . Accordingly , normalized interaction score plots depict a strong underrepresentation of interactions between chromosomes of most dissimilar length in spo11 rec8 ( Fig 6C ) . This trend held true for small , medium-sized and large chromosomes ( Fig 6D ) . Even compared to spo11 diploids and haploids , spo11 rec8 diploids show a greater decrease in normalized interaction score across all 16 chromosomes between the three partners most similar in size to a particular chromosome and the three most dissimilar in size ( Figs 2C , 3C and 6C; bin 1–3 vs . bin 13–15 ) . However , caution should be exercised in interpreting these results , in light of reduced levels of coupling in spo11 rec8 diploids ( [22] , and confirmation by the lower raw interaction frequencies , in this study ) . Overall , these results suggest that the meiotic bouquet might create a favorable architecture for assorting chromosomes by length , thus helping to establish non-homologous coupling contacts based on chromosome size . Recent in silico simulations have demonstrated that the bouquet sorts chromosomes based on their size [45] . The tightness of the bouquet ( i . e . clustering opposite telomeres on a narrower section of the nuclear envelope ) plays a greater role for associations between shorter chromosomes , with these chromosomes arranged in a shorter U-shaped structure [45] . In contrast , the levels of chromosomal rigidity/flexibility and of periodic juxtaposition have a greater influence on interactions between longer chromosomes . Absence of the bouquet , as in a spo11 ndj1 diploid , disrupts the interaction pattern . On the other hand , persistence of the bouquet , as in a spo11 rec8 strain , does not disrupt the interaction pattern in the minority of cells that undergo coupling in this genotype , and , additionally , we observed avoidance of interactions between CENs from chromosomes of most dissimilar sizes . In the meiotic bouquet , with telomeres confined to a section of the nuclear envelope , the centromeres of chromosomes likely project towards the center of the nucleus , into a reverse Rabl-like configuration . Since most of the centromeres do not sit precisely at the midpoint of the 16 yeast chromosomes , the length of the shorter arm of the chromosome ( centromere to telomere ) would limit the distance from the base of the bouquet . As such , centromeres from chromosomes with similarly-sized short arms might be closer than even more similarly-sized chromosomes , hence engaging in coupling interactions more often . For example , in some long chromosomes ( 12 and 2 ) the centromeres are subtelocentric and thus might associate more often with very small chromosomes ( such as 3 , 5 and 6 ) . We repeated our analysis for spo11 and spo11 zip1 diploids and haploids , but did not observe any association between the level of interaction frequencies and the similarity of short arm sizes ( p > 0 . 05 ) . Thus physical constraints based on chromosome size , such as 3D conformation , chromosomal condensation and bending rigidity in the arms , probably play a greater role in the establishment of couples than the maximum linear distance from the centromere to its closest telomere . Another contribution to the size-dependent pattern might come from the interplay between centromeres , telomeres and the spindle pole body . In fission yeast , the telomere bouquet is necessary for accurate chromosome segregation through interactions with the spindle pole body and spindle assembly , independent of recombination [46] . Centromeres need to interact with the telomere-spindle pole body microenvironment for full assembly during meiosis [47] . Of note , in the absence of bouquet formation , centromeres have the ability to interact with the spindle pole body to mediate spindle assembly instead of telomeres , keeping chromosomes close to an interphase Rabl configuration [48] . It is possible that a more complex size-dependent pattern is propagated at the spindle pole body from transitioning between the Rabl configuration in interphase , the bouquet , and then centromere coupling . Our findings from WT diploids and bouquet mutants guide us to update a previous coupling model [16] , where centromeres are randomly paired to a revised model ( Fig 6E ) where bouquet formation would first help to establish chromosomal interactions based on chromosome size . The bouquet appears to serve as a chromosome size sorter , not only for homologous chromosomes as previously postulated [45] but also for non-homologous coupling . This sorting mechanism would rely on the degree of clustering forces and on the biophysical properties of chromosomes [45] , as well as the overall chromosomal configuration away from telomeres . Specifically our results suggest the bouquet’s role in the mechanism for homolog pairing: this configuration sets up the chromosomes in a size-dependent alignment for coupling , as a first step to homolog recognition . As meiotically-programmed DSBs occur , and recombination-based homology searches begin , Zip1 becomes phosphorylated , releasing the couples [18] , and repeated pairing partner switching ensues ( speed-dating model ) [16] . As chromosomes find their homologs , and begin to synapse , they are effectively removed from the coupling pool , incrementally restricting the possible couples . Longer chromosomes tend to become paired with their homologs earlier [15] and locked in through SC formation and recombination , whereas small chromosomes continue their non-homologous contacts . This late pairing phase is in concordance with data obtained on a smaller scale using electron microscopy [15] . While we found a preference for centromere coupling interactions based on chromosome size similarities , our data do not perfectly fit this pattern . Closer inspection of heatmaps reveals the presence of “cold” orthogonal diagonals , with non-homologous couples interacting less frequently . This brings the possibility that there are likely cold and hot spots for coupling interactions . In budding yeast , the 32 telomeres appear as 3–8 clusters in interphase [49 , 50] . Could telomere clusters , present prior to the formation of the meiotic bouquet , play a role in establishing the interaction patterns observed in centromere coupling ? We asked whether chromosomes found in the same telomere cluster are strong interacting partners in coupling . Telomere cluster assignments differed whether they were determined by genetics and chromosomal tagging methods [51] , or derived from 4C genomic data [24 , 52] . A coupling interaction pattern based on telomere clusters from Schober et al . [51] is highly significant in spo11 ndj1 diploids ( Randomization matrix test; p < 0 . 01 ) , but not in spo11 , spo11 zip1 , or spo11 rec8 diploid strains . Since spo11 ndj1 yeast do not form the meiotic bouquet , preferential interactions based on prior telomere clusters might be favored during centromere coupling , compared to the chromosome size-dependent pattern observed in spo11 and WT diploid yeast that go through the bouquet stage . The role of centromere coupling remains unknown . It might be an initial step for chromosomes to query whether one chromosome is its homologous match , but since zip1 ( coupling and SC defective ) mutants are capable of robust homolog pairing [53] , coupling must be a redundant path for homolog pairing . Another function of coupling might be to block the deleterious establishment of recombination at centromeres of homologous chromosomes [17] . Centromeres likely constitute a special region for these interactions , providing a cis regulatory center for each chromosome where conditions must be met before SC formation is permitted . Although SC formation can initiate at sites other than the centromere , centromere synapsis generally occurs earlier [54] . Indeed Zip1 has been shown to deflect deleterious crossing over in the immediate centromere vicinity [55] . Identification of additional functional requirements for centromere coupling will likely provide more clues into its role in early meiosis . In this study , we have proven the advantages of genomics approaches to characterize a biological phenomenon . While more technically challenging , expansive and time consuming than standard methods , only such a strategy would have been able to identify pairwise trends systematically with this higher level of confidence .
Yeast strains are isogenic with BR1919-8B ( S1 Table ) [56] . Strain growth was performed as described [39] . Cultures were grown first for 24 h in YPADU at 30°C to early stationary phase . Then cultures were resuspended in 2% sporulation media ( 2% potassium acetate ) to a final cell concentration of 2–4 X 107 cells/mL , as determined by OD600 on a spectrophotometer . About 4 X 109 cells were needed per 3C sample [35] . 200 mL cultures were grown in 2 L flasks to promote good sporulation by providing adequate oxygenation ( 6 L flasks for 5-time points in WT diploid cells ) . Haploid cells were grown for 20 h in sporulation media while diploid cells were grown for 14 h . Meiotic chromosome spreading was performed as previously described [39] . Staining was performed using antibodies against Red1[57] and against the Myc epitope ( 9E10 ) to detect Ctf19-Myc [16] . Cy3-conjugated anti-mouse ( Ctf19 ) and FITC-conjugated anti-rabbit ( Red1 ) secondary antibodies were used for detection . Slides were also stained with DAPI in the mounting media to observe compact spreading and core formation . Meiotic chromosome spreads were visualized on a Nikon E800 microscope and images were visualized on the IPLab software , as previously described [58] . Immunofluorescence was performed on WT diploids at various times throughout meiosis to determine centromere organization ( Ctf19 ) and the appearance of SC components ( Zip1 and Red1 ) . For the tetO-tetR-mCherry/lacO-lacI-GFP experiment , primary antibodies against c-MYC ( 9E10 mouse ) , GFP ( chicken ) and m-Cherry ( rabbit ) were visualized with the following secondary antibodies: anti-mouse CY5 , anti-chicken FITC and anti-rabbit Texas Red . Nuclear spreads were screened by DAPI for compact spreading and core formation . Among those spreads , the kinetochore arrangement ( Ctf19-Myc staining ) was scored for phase and spreading . Early spreads contain a single focus of kinetochores or a loosened bundle and are not useful for determining coupling . Other spreads may contain greater than 8 Ctf19 foci and represent disruption of coupling from physical spreading . Only those nuclear spreads that met criteria for being coupled ( ~8 Ctf foci ) were scored for pairing . Three independent biological replicates were carried out for each strain and fifty nuclei were scored for each experiment . The 3C procedure followed previously described protocols from the Dekker lab with few modifications [25 , 35 , 36] . An extra chloroform extraction step and a post-precipitation wash with 70% ethanol were added for increased purity of 3C libraries . For the double digestion , EcoRI-HF and MfeI-HF ( NEB ) generated cohesive ends that were compatible , yet recognizing slightly different sequences ( GAATTC and CAATTG respectively ) . They were selected due to their extremely low star activity , extended enzymatic half-life suitable for overnight digestion and stronger activity in samples of lower purity , making them appropriate for digestion of crosslinked chromatin . 50 U of each enzyme were used per tube ( each sample is divided into 40 individual tubes at this step; see [35] ) . For quality control , we ensured that formaldehyde crosslinking enabled preferential ligations of fragments in close proximity in the nucleus [25] . On our 3C2D libraries , we amplified fragments with primer pairs located 10 kb and 80 kb away on chromosome 8 . We used a control representing randomly-ligated , not crosslinked genomic DNA , in which proximal and distal fragments are present in similar abundance [25] . Several dilutions of templates were used for PCR amplification . PCR products were detected by gel electrophoresis on a 2% agarose gel . Band intensity was determined using ImageJ ( NIH ) . 3C2D libraries that did not show a formaldehyde-mediated enrichment for proximal intra-chromosomal interactions over distal ones were discarded . Primers were designed 100–150 bp away from the closest CEN-proximal MfeI or EcoRI site on each side of the CEN , with length between 18 and 24 bp optimally and a melting temperature around 56–58°C ( S2 Table ) . Each primer is oriented towards the restriction site in order to amplify fragments from different chromosomes ligated together . Taqman probes were designed as recommended for mammalian 3C-qPCR [32] on the opposite strand from the primer and around 10–60 bp away from the restriction site . Taqman 5’FAM-3’TAMRA probes ( QuickProbes , Operon ) were designed for annealing at 65–68°C , with the following constraints: 20–36 bp in length , more Cs than Gs , no G at the 5’ end to prevent quenching , no stretch of 4 identical nucleotides and no more than 2 C’s and/or G’s in the last 5 nucleotides of the 3’ end ( S3 Table ) . All 480 possible interactions ( CEN1 left vs . CEN2 left , CEN1 left vs . CEN2 right , CEN1 left vs . CEN3 left , … , CEN1 right vs . CEN2 left , CEN1 right vs . CEN2 right , … ) were analyzed by performing qPCR reactions in triplicates and diluted 2- , 4- and 8-fold . Individual reactions were set up following previously established guidelines [32]: 2 μL of each primer ( 2 . 5 μM ) , 1 μL Taqman probe ( 1 . 5 μM ) , 5 μL QuantiTect Probe PCR master mix ( Qiagen ) , and 1 μL diluted 3C2D or control sample . qPCR reactions were run on a LightCycler 480 ( Roche ) with the following parameters: 1 ) enzyme activation for 15 min . at 95°C , and 2 ) 55 cycles of amplification with a 15 s denaturation at 95°C , a 60 s amplification at 60°C and a single fluorescence acquisition . The “Second derivative maximum” analytical tool in the LightCycler480 was used to obtain Crossing point values ( Cp ) . For each individual reaction , the amplification curve was visually inspected to ensure that the reported Cp value was plausible for the exponential phase of the curve . Cp values were adjusted according to the sample dilution in individual qPCR reactions , including a correction for successful completion when certain dilutions failed to amplify . Values were averaged for each of 120 combinations ( CEN1 vs . CEN2 , CEN1 vs . CEN3 , etc . ) to generate a raw interaction frequency matrix . Each combination is formed by the following interactions: CENX Left-CENY Left , CENX Left-CENY Right , CENX Right-CENY Left , and CENX Right-CENY Right . The resulting interaction frequencies were normalized as described in the analytical section of the original 3C-qPCR protocol [32] . First , they were normalized to the amplification efficiencies obtained from control samples ( randomly-ligated genomic DNA ) . Second , they were normalized according to the DNA concentration of the 3C2D library ( internal loading control ) . In this case , we performed RT-qPCR with the SYBR Green system on a LightCycler480 ( Roche ) to quantify an internal product amplified from a primer pair which does not amplify across MfeI or EcoRI sites , using dilutions of the 3C2D library ( 1:12 . 5 , 1:25 , 1:50 , 1:100 , 1:200 ) [32] . Finally , since we are looking only at inter-chromosomal interactions between non-homologous centromeres , our experimental design protects from artifactual peaks that can arise when local conformation of chromatin influences the detection of intra-chromosomal interactions [38] . For individual genotypes , heat maps were generated in R using interaction frequencies calculated as described above . We used the color scheme “YlOrRd” from package RColorBrewer . Darker shades of red indicate a higher level of interaction . For visualization purposes , we separated chromosomes in three groups based on chromosome size , using k-means clustering . For the multiple time points in a WT diploid , differences of normalized interaction frequencies were plotted on a heatmap to compare their relative progression ( early→mid , mid→late , and early→late ) . In this case , heatmaps were unscaled and generated using the color scheme “RdBu” , with white meaning no changes , red for increases , and blue for decreases . To test the significance of the pairwise pattern based on similarities of chromosome lengths , we used a non-parametric testing procedure . For each chromosome ( out of 16 ) , we identify the number of times one of three closest chromosomes in length to that particular chromosome would be also amongst the three chromosomes with the highest interaction frequencies with that particular chromosome . We summed the number of times we had such a case across all 16 chromosome . We next simulated the null distribution by randomizing the matrix of interaction frequencies , then selecting the three strongest interacting partners ( out of 15 ) for each particular chromosome and asking whether they would be one of the three closest chromosomes in length as well . We repeated this 15 more times ( 16 total ) and summed to obtain the grand total of top 3 interactions with top 3 chromosomes closest in size across all 16 chromosomes . This constitutes one iteration . We performed this procedure 100 , 000 times . The p-value is given by the fraction of random iterations with greater or equal association between IF strength and chromosome size similarities ( grand total ) than found experimentally for each genotype . A similar randomization approach was used on a subset of the matrix when comparing the four chromosomes of shortest size , and the four chromosomes of largest size . For arm homology , a similar non-parametric procedure was performed , except that we used , for each chromosome , the three chromosomes with the highest level of arm homology as determined from the figures and from the raw data of ORF homology [40] . For the agreement between haploid and diploid spo11 strains at the top of the interaction list , we used a similar non-parametric strategy , except that , for each chromosome , we randomly selected 5 chromosomes for the haploid strain and 5 chromosomes for the diploid strain , asking how many chromosomes overlap . Then we summed across all 16 chromosomes and performed 100 , 000 iterations . Data to generate all heatmaps and graphs are available from the Dryad Digital Repository: http://dx . doi . org/10 . 5061/dryad . 71425 . Additional data and a few sample codes are deposited in GitHub: https://github . com/plefrancois/CENcoupling . | Meiosis enables sexual reproduction in eukaryotes by producing gametes . In the process , it increases genetic diversity through recombination of homologous chromosomes from the parents . Genetic diversity constitutes an evolutionary advantage . Prior to finding their unique pairing partner ( homolog ) , chromosomes associate non-homologously with other chromosomes through their centromeres , a process termed centromere coupling . Little is known about the nature and mechanism of centromere coupling . In this study , we present the first pairwise characterization of this process conserved amongst eukaryotes , using the budding yeast as a model . We quantitatively analyzed the interactions between centromeres for each pair of chromosomes . We observed an interaction pattern based on chromosome size , where centromeres from smaller chromosomes frequently associated with those from other small chromosomes , and a similar association for large chromosomes . This pattern appears ubiquitous , since recombination-defective diploid cells , haploid cells forced to undergo meiosis , and wild-type yeast early in meiosis , until homologous chromosomes become paired , all undergo non-homologous centromere coupling . Centromeres derived from a common ancestor , prior to genome duplication , do not associate more often , excluding ancestral homology as the mechanism . Data from mutants affecting the meiotic bouquet , where all chromosome ends become embedded and clustered in the nuclear envelope prior to coupling , suggest a potential mechanism to generate interactions . Deciphering the mechanisms for proper pairing of homologous chromosomes helps us to understand and prevent chromosomal abnormalities in pregnancy . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"meiosis",
"homologous",
"chromosomes",
"chromosome",
"structure",
"and",
"function",
"centromeres",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"telomeres",
"fungi",
"model",
"organisms",
"dna",
"homologous",
"recombination",
"research",
"and",
"... | 2016 | Multiple Pairwise Analysis of Non-homologous Centromere Coupling Reveals Preferential Chromosome Size-Dependent Interactions and a Role for Bouquet Formation in Establishing the Interaction Pattern |
Genome-wide association studies ( GWASs ) have identified many disease associated loci , the majority of which have unknown biological functions . Understanding the mechanism underlying trait associations requires identifying trait-relevant tissues and investigating associations in a trait-specific fashion . Here , we extend the widely used linear mixed model to incorporate multiple SNP functional annotations from omics studies with GWAS summary statistics to facilitate the identification of trait-relevant tissues , with which to further construct powerful association tests . Specifically , we rely on a generalized estimating equation based algorithm for parameter inference , a mixture modeling framework for trait-tissue relevance classification , and a weighted sequence kernel association test constructed based on the identified trait-relevant tissues for powerful association analysis . We refer to our analytic procedure as the Scalable Multiple Annotation integration for trait-Relevant Tissue identification and usage ( SMART ) . With extensive simulations , we show how our method can make use of multiple complementary annotations to improve the accuracy for identifying trait-relevant tissues . In addition , our procedure allows us to make use of the inferred trait-relevant tissues , for the first time , to construct more powerful SNP set tests . We apply our method for an in-depth analysis of 43 traits from 28 GWASs using tissue-specific annotations in 105 tissues derived from ENCODE and Roadmap . Our results reveal new trait-tissue relevance , pinpoint important annotations that are informative of trait-tissue relationship , and illustrate how we can use the inferred trait-relevant tissues to construct more powerful association tests in the Wellcome trust case control consortium study .
Genome-wide association studies ( GWASs ) have identified thousands of genetic loci associated with complex traits and common diseases . However , the majority ( ~90% ) of these associated loci reside in noncoding regions and have unknown biological functions [1] . Systematic characterization of the biological function of genetic variants thus represents an important step for further investigating the molecular mechanisms underlying the identified disease associations . Functional characterization of genetic variants is challenging because the function and genetic effects of variants on most traits are likely acted through a tissue-specific fashion , despite their tissue-wide presence ( certainly with the notable exception of somatic mutations ) [2–4] . For example , it is well recognized that many psychiatric disorders , such as bipolar and schizophrenia , are consequences of dysfunctions of various genes , pathways as well as regulatory elements in neuronal and glia cells , resulting from brain-specific genetic effects of polymorphisms [5–9] . Therefore , characterizing the function of variants in various brain regions can help elucidate the biology of psychiatric disorders . For most complex traits , however , their trait-relevant tissues are often obscure . As a result , identifying trait-relevant tissues and characterizing the functions of genetic variants within relevant tissues holds the key for furthering our understanding of disease etiology and the genetic basis of phenotypic variation [10–16] . Both experimental and computational studies have recently produced a rich resource of variant annotations that can help characterize the function of genetic variants in a tissue-dependent fashion [17–21] . For example , the ENCODE and Roadmap epigenomics projects collect various measurements of histone modification , open chromatin , and methylation from more than a hundred different tissue and cell types , where each epigenetic mark characterizes a specific aspect of variant function [22 , 23] . Similarly , the GTEx project produces gene expression measurements from multiple tissues and quantifies variant function in terms of its ability to regulate gene expression levels in the given tissue [24] . Besides the experimental efforts , many computational methods have also been developed to create synthetic functional annotations for variants in a tissue-dependent manner . For example , the chromHMM converts measurements of multiple histone modifications in each tissue into 15 chromatin states that have more biologically interpretable functions than the original histone occupancy based annotations [25 , 26] . Similarly , several other methods provide different ways to summarize multiple SNP annotations into a single , potentially more interpretable annotation for various tissues [27 , 28] . With the large and growing number of tissue-specific variant annotations , however , one naturally wonders how these different annotations can be incorporated together to facilitate the identification of trait-relevant tissues . Several statistical methods have been recently developed to test the role of various functional annotations in predicting the variant effect sizes or causality for GWAS traits [10–16 , 29–31] . These methods often test one annotation at a time and produce a test statistic signifying the importance of the given annotation for a GWAS trait . By comparing the test statistics of univariate annotation from different tissues and ordering tissues by the magnitude of the test statistics , existing methods can be used to identify tissue relevance for a given trait [10 , 16] . However , examining one variant annotation at a time can be inefficient as it may fail to incorporate the rich information contained in various other annotations that likely characterize other aspects of variant function [27 , 28 , 32] . For example , some annotations are designed to evaluate evolutionary conservation of a variant , while some other annotations are designed to quantify its biochemical functionality [29 , 33] . Even annotations that belong to the same general category may characterize substantially different functions of a variant . For example , different histone modifications are used to annotate variants by different functional genomic regions: H3K4me3 annotates promoter; H3K4me1 annotates enhancer; H3K36me3 annotates transcribed regions; H3K27me3 annotates polycomb repression; H3K9me3 annotates heterochromatin; and H3K9ac annotates both enhancer and promoter [22 , 34] . Therefore , testing one annotation at a time may be suboptimal , and it would be ideal to incorporate multiple sources of information together to identify trait-relevant tissues . Besides the potential loss of inference efficiency , examining one annotation at a time can sometimes lead to incoherent results on the identification of trait-relevant tissues: partly due to a lack of statistical power , the trait-relevant tissues inferred by different SNP annotations may not always agree , and it is often not straightforward to consolidate results from using different annotations [27 , 28] . Despite the potential inefficiency due to the use of univariate annotation tests , several studies have explored the feasibility of inferring trait-relevant tissues for various complex traits using SNP functional annotations [16 , 31 , 35 , 36] . While the inferred trait-tissue relevance often makes biological sense , it is unclear how to further make use of these inferred trait-relevant tissues to benefit future association studies [11 , 16] . In principal , levering the information learned from the identified trait-relevant tissues could enable powerful association tests , as the functional annotations in the trait-relevant tissues could contain important SNP causality and effect size information . In practice , however , incorporating trait-relevant tissue information into association tests is challenging , partly because the existing statistical methods for identifying trait-relevant tissues mainly rely on polygenic models [16 , 31 , 35 , 36] while the existing statistical methods for association tests mostly rely on univariate tests or sparse regression models [37–40] . The disparity between the methods used for trait-tissue relevance inference and the methods used for association tests make it hard to share information across the two different tasks . Here , we develop a simple method to address the above two challenges . First , we incorporate multiple binary and/or continuous annotations to facilitate the identification of trait-relevant tissues for GWAS traits . To do so , we modify the commonly used linear mixed model [38 , 41–45] to relate variant effect sizes to variant annotations by introducing variant specific variance components that are functions of multiple annotations . We quantify and evaluate the joint contribution of multiple annotations to genetic effect sizes by performing parameter inference using the widely used generalized estimation equation ( GEE ) [46] . Our GEE-based algorithm is closely related to the recent LDSC [16] , ployGEE [47] , and MQS methods [48] , allows for the use of summary statistics , and naturally accounts for the correlation among summary statistics due to linkage disequilibrium . With GEE statistics , we further apply mixture models to classify tissues into two categories—those that are relevant to the trait and those that are not—thus formulating the task of identifying trait-relevant tissues into a classification problem . Second , our method is closely related to the sequence kernel association test ( SKAT ) [49–51] , and this relationship allows us to apply parameter estimates from the inferred trait-relevant tissues as SNP weights to construct SNP set test and power new association studies . We refer to our overall analytic procedure as the Scalable Multiple Annotation integration for trait-Relevant Tissue identification ( SMART ) . We provide an overview of our method in the Materials and Methods section , with details described in the S1 Text . With simulations , we show that , compared with analyzing one annotation at a time , analyzing multiple annotations jointly can improve power for the identification of trait-relevant tissues . In addition , we show that , using parameter estimates from inferred trait-relevant tissues as SNP weights leads to more powerful SNP set tests than the standard SKAT [49–51] . We apply our method for an in-depth analysis of 43 GWAS traits with multiple functional annotations in more than one hundred tissues derived from ENCODE and Roadmap . We show how our method and analysis can help provide biological insights for the genetic basis of complex traits and benefit future association studies . The SMART method is implemented as an R package , freely available at http://www . xzlab . org/software . html .
We consider a simple extension of the linear mixed model to evaluate jointly the contribution of multiple SNP annotations . To do so , we first consider the following multiple linear regression model that relates genotypes to phenotypes y=Xβ+ϵ , ϵi∼N ( 0 , σe2 ) , ( 1 ) where y is an n-vector of phenotypes; X is an n by m matrix of genotypes; β is an m-vector of effect sizes; and ϵ is an n-vector of residual errors; each element ϵi is independent and identically distributed from a normal distribution with variance σe2 . We center the phenotype y and standardize each column of the genotype matrix X to have zero mean and unit variance , allowing us to ignore the intercept in the model . Because p ≫ n , we have to make further modeling assumptions on the SNP effect sizes β to make the model identifiable . We do so by incorporating SNP annotation information and making the effect size βj of j-th SNP depending on its annotations . Specifically , we assume that all SNPs are characterized by a same set of c annotations . For the j-th SNP , we denote Aj = ( 1 , Cj1 , Cj2 , ⋯ , Cjc ) T as a ( c+1 ) -vector of realized annotation values including a value of one that corresponds to the intercept . These annotations can be either discrete or continuous . For example , one annotation could be a binary indicator on whether the SNP resides in exonic regions , while another annotation could be a continuous value of the CADD score [18] of the SNP . Because our model includes an intercept ( more details in the next paragraph ) , we require that any linear combination of these annotations does not sum to a vector of one’s across SNPs in order to avoid identifiability issues–a requirement holds for standard linear regression models . For example , we cannot include two non-overlapping annotations that form a partition of the genome ( i . e . Cj1 = 1 when Cj2 = 0 , and Cj1 = 0 when Cj2 = 1 ) . Our coding scheme is conventionally referred to as the reference coding scheme . To simplify presentation , we assemble the annotation vectors across all SNPs into an m by ( c+1 ) annotation matrix A , where each row contains the annotation vector for the corresponding SNP . We assume that the annotations for a given SNP influence its effect size . In particular , we assume that each effect size βj is independent and follows a normal distribution with mean zero and a variance σj2 that is SNP specific . The SNP-specific variance assumption generalizes the standard LMM assumption where every SNP is assumed to share a common variance [37 , 52] . We further impose an extra layer of hierarchy by assuming that the SNP specific variance is a function of the annotation vector , or βj∼N ( 0 , σj2/m ) , σj2=Ajα* , ( 2 ) where α*= ( α0α ) is a ( c+1 ) -vector of coefficients that include an intercept α0 and a c-vector of annotation coefficients α . Each annotation coefficient is large when the corresponding annotation is predictive of the SNP effect size . Therefore , the annotation coefficients can be used to evaluate the importance of annotations . Above , we center the 2nd to the ( c+1 ) -th columns of the annotation matrix A to have mean zero across SNPs . After centering , the ratio mα0/ ( mα0+σe2 ) has the natural interpretation of SNP heritability , which is defined as h2=E ( ∑βj2 ) / ( E ( ∑βj2 ) +σe2 ) , roughly following [16] , where E represents prior expectation . The intercept α0 effectively determines how large a typical SNP effect size would be , while the other annotation coefficients determine how the SNP effect size variance would vary around the average depending on what annotations the SNP has . Note that the assumed linear relationship between the SNP specific variance and annotations also naturally extends the modeling assumptions made in LDSC [16] and MQS [48] , both of which examine one annotation at a time in the presence of multiple binary annotations , though LDSC has also been recently extended to examine one annotation at a time in the presence of continuous annotations [31] . In addition , our polygenic modeling assumption complements alternative approaches in using sparse models for integrating functional annotations [11 , 12 , 39 , 53] . For inference on the annotation coefficients ( α ) , we follow the main idea of LDSC and MQS in using the marginal χ2 statistics . Using marginal statistics allows our method to be applied to data where only summary statistics are available . Unlike the detailed algorithms of LDSC or MQS that were initially designed to examine one annotation at a time in the presence of multiple binary annotations , however , we applied the generalized estimating equation ( GEE ) [46 , 54] inference method that allows for the joint inference of multiple binary and continuous annotations ( details in S1 Text ) . GEE is widely used for modeling correlated data and is particularly suitable here to account for the correlation among the marginal χ2 statistics due to linkage disequilibrium . In the case of binary annotations , the results of our GEE on each annotation by using a diagonal matrix as the working covariance matrix can reduce to that of LDSC and MQS , while the results of our GEE by using an LD based general working covariance matrix can reduce to that of polyGEE [47] . Importantly , just like other summary statistics based methods , GEE inference can be carried out using summary statistics that include marginal χ2 statistics and the m by m SNP correlation matrix . The SNP correlation matrix can be obtained from a reference panel , by using , for example , the genotypes from the 1 , 000 Genomes Project [55] . To facilitate both computation and memory storage , we further approximate the SNP correlation matrix by a block diagonal matrix ( details in S1 Text ) , allowing us to capture the main block-wise linkage disequilibrium pattern commonly observed in the human genome [39 , 56–58] . Finally , with GEE , we obtain both point estimates α^ and their variance V ( α^ ) for all annotation coefficients in a closed form . We can then compute the multivariate Wald statistic α^TV ( α^ ) −1α^ which can be used as a measurement of trait-tissue relevance . Previous approaches to identify trait-relevant tissues examines one ( univariate ) Wald statistics at a time , and uses an asymptotic normal test to obtain a p-value to identify significant trait-tissue pairs . Because one annotation in one tissue is often highly correlated with the same annotation in other tissues as well as other annotations in other tissues , the p-values for even the trait-irrelevant tissues are often significant due to the annotation correlation across annotations and tissues . Indeed , as previous studies have shown , even in simple simulations , trait-irrelevant tissues can be falsely identified as trait-relevant in 20% of the simulation replicates [16] . As a consequence , previous studies have to use a set of baseline annotations as covariates to reduce the cross-tissue correlation among annotations , thus reducing false positives . However , it is often unclear how many and what types of baseline variables one should include for a given data set: using a small number of baseline covariates may not control for false positives well , while using a large number of covariates may reduce the power to detect the true trait-relevant tissues . Indeed , the use of baseline variables seems to be highly dependent on data sets ( with varying sample sizes and SNP numbers ) , and needs adjustment in different data sets to achieve sensible results [16] . Here , we present an alternative strategy for identifying trait-relevant tissues . Specifically , for each trait in turn , we model the multivariate Wald statistics across tissues with a mixture of two non-central chi-squared distributions to classify tissues into two groups . The two non-central chi-squared distributions have the same degrees of freedom that equals to the number of annotations fitted in GEE ( i . e . c ) , but different noncentrality parameters . The chi-squared distribution with the small noncentrality parameter represents the empirical null distribution that contains tissues irrelevant to the trait . The small , nonzero , noncentrality parameter characterizes the fact that these irrelevant tissues tend to have Wald statistics larger than what would be expected under the theoretical null distribution ( i . e . central chi-squared ) simply due to annotation correlation across tissues . In contrast , the chi-squared distribution with the large non-central parameter represents the alternative model that contains tissues relevant to the trait . The large noncentrality parameter characterizes the fact that these relevant tissues tend to have Wald statistics larger than those from the irrelevant tissues . By classifying tissues into two groups , we can identify tissues with strong trait-relevance without the need to explicitly model the empirical null distribution using a data generative model . Therefore , our strategy effectively formulates the task of identifying trait-relevant tissues as a classification problem instead of a testing problem . By modeling the empirical null distribution directly , we can reduce false discoveries and potentially gain power at a given false discovery rate ( FDR ) . We also note that this classification strategy follows closely recent applications of mixture models to estimate the empirical null distribution in other genomics settings [59 , 60] . Technically , we use the expectation-maximization ( EM ) algorithm to fit the mixture model and infer the two noncentrality parameters as well as the proportion of alternatives from data at hand ( details in the S1 Text ) . For each tissue in turn , we then obtain the inferred posterior probability ( PP ) of it being in the alternative model as its evidence for trait-relevance . We use these inferred posterior probabilities ( ranging between 0 and 1 ) for all following analyses . Note that while our linear mixed model itself does not explicitly model the correlation structure among annotations across tissues by incorporating all annotations from all tissues into a single model , our mixture model and classification strategy can implicitly account for the annotation correlation across tissues . Finally , we ask the question of how to make use of the inferred trait-relevant tissues to enable more powerful future association studies . We note that our model defined in equations ( 1 ) and ( 2 ) is closely related to the sequence kernel association test ( SKAT ) model [49–51] for SNP set test . In particular , the SNP specific variance σj2 in our model can be naturally interpreted as the SNP specific weight in the SKAT [49–51] framework . Because of this close relationship between our model and SKAT [49–51] , we propose to use the estimated SNP specific variance σ^j2 in the top trait-relevant tissue from our model as SNP weights to perform SKAT [49–51] analysis in new association studies . Intuitively , if a SNP tends to have a large effect size , then weighting it high in the subsequent SNP set analysis can help achieve greater association mapping power . We examine this intuition with both simulations and real data applications . Note that our weighted SKAT [49–51] approach is related to a recent method , FST , which also extends SKAT to accommodate multiple functional annotations [32] . However , our method borrows information across all SNPs to infer trait-relevant tissues and estimate annotation coefficients , and further relies on the estimated annotation coefficients in the trait-relevant tissues to construct SNP specific weights for SKAT analysis . In contrast , FST [32] examines one gene at a time , calculates a test statistic for each annotation in turn , and effectively chooses the minimal test statistics among all these annotations as the final statistics for testing . In addition , while our method is polygenic in nature , the idea of using SNP specific weights to construct test statistics is also related to a recent study that uses functional annotations to design SNP specific weights in sparse regression models to improve disease risk prediction performance [61] . We used tissue-specific SNP annotations from the ENCODE [23] and the Roadmap [22] projects in the present study . Specifically , we downloaded the broadPeak files from the Roadmap Epigenomics web portal ( http://egg2 . wustl . edu/roadmap/web_portal/ ) . The broadPeak files contain peak regions for four histone marks ( H3K27me3 , H3K36me3 , H3K4me1 , H3K4me3 ) from 16 cell lines in the ENCODE project and 111 tissues from the Roadmap project ( release 9 ) . Among the 127 tissue/cell types , we excluded ESC , IPSC , and ES-derived cell lines to focus on the remaining 105 tissue/cell types ( S1 Table ) . Following previous studies [16 , 22 , 27] , we further classified 105 tissues into 10 tissue groups based on anatomy ( BloodImmune , Adipose , AdrenalPancreas , BoneConnective , Cardiovascular , CNS , Gastrointestinal , Liver , Muscle , Other; S1 Table ) . For each tissue and each histone mark in turn , we created a binary histone mark annotation indicating whether the SNP resides inside the peak regions of the histone mark . The average proportions of SNPs residing in each of the four mark labeled regions across the 105 tissues are 25 . 75% for H3K27me3 , 18 . 51% for H3K36me3 , 17 . 98% for H3K4me1 , 10 . 69% for H3K4me3 ( S1 Table ) . In addition to the binary annotations , for each tissue group and each histone mark in turn , we averaged the binary annotation indicator across all tissue types within the tissue group and used the average value as a new , continuous , tissue group level histone mark annotation . Therefore , we obtained both tissue-specific binary histone mark annotations and tissue-group-specific continuous histone mark annotations . Besides the above histone mark annotations , we also obtained SNP annotations based on 15 chromatin states ( TssA , TssAFlnk , TxFlnk , Tx , TxWk , EnhG , Enh , ZNFRpts , Het , TssBiv , BivFlnk , EnhBiv , ReprPC , ReprPCWk and Quies ) inferred from ChromHMM [62] in the 105 tissues . In particular , we downloaded the posterior probabilities of each of the 15 states for each genomic location in different tissues from the Roadmap Epigenomics web portal . For each tissue group and each posterior probabilities in turn , we then averaged the posterior probabilities across all tissue types within the tissue group and used the average value tissue as tissue group specific continuous ChromHMM annotation . We performed two sets of simulations to illustrate the benefits of our method in terms of ( 1 ) using multiple SNP annotations and ( 2 ) enabling more powerful SNP set tests . For all simulations , we used real genotypes from the Genetic Epidemiology Research Study on Adult Health and Aging ( GERA; dbGaP accession number phs000674 . v2 . p2 ) [63 , 64] . The original genotype data of the GERA study consists of 675 , 367 SNPs on 62 , 313 individuals . We removed SNPs with a missingness percentage above 0 . 05 , a minor allele frequency ( MAF ) below 0 . 05 , and a Hardy-Weinberg equilibrium test p-value below 10−4 . We then randomly selected 10 , 000 individuals with European ancestry , and obtained the first 27 , 640 ( or 10 , 000 ) SNPs on chromosome one to perform the first ( or the second ) set of simulations . For the first set of simulations , we obtained two histone marks ( H3K4me1 and H3K4me3 ) from ten different tissue groups from the ENCODE and Roadmap projects , and used them as SNP annotations ( details in the previous subsection ) . Among the ten tissue groups , we randomly selected one as the trait-relevant tissue group in each simulation replicate . We designated all SNPs to be causal , and simulated the causal SNP effects independently from a normal distribution with a SNP-specific variance determined by annotations in the trait-relevant tissue . In particular , we set the variance intercept ( i . e . α0 ) to be 0 . 1 and we varied each of the two annotation coefficients ( i . e . α1 , α2 ) from -0 . 1 to 0 . 5 ( -0 . 1/0/0 . 05/0 . 1/0 . 25/0 . 5 ) to cover a range of possible values estimated from real data ( details in Real Data Applications ) . We performed 1 , 000 simulation replicates for each combination of the two annotation coefficients ( α1 , α2 ) . Note that the median estimates of the two annotations across 43 GWAS traits ( details in Real Data Applications ) is close to ( α1 , α2 ) m = ( 0 . 1 , 0 . 05 ) . We simulated the residual errors from a normal distribution with variance 0 . 9 , so that the resulting trait has a SNP heritability of 0 . 1 , which corresponds to the median SNP heritability estimate across 43 traits in the real data analysis . We then summed all genetic effects and the residual errors together to form the simulated phenotypes . With genotypes and simulated phenotypes , we fitted a linear regression model for one SNP at a time and computed marginal χ2 statistics . We further paired these marginal statistics with a SNP correlation matrix estimated using 503 individuals of European ancestry from the 1 , 000 Genomes Project [55] . We then examined the ten candidate tissues in turn using either two annotations together or one annotation at a time . For additional comparisons at the median setting ( α1 , α2 ) m = ( 0 . 1 , 0 . 05 ) , we also included LDSC [16] , which in default includes 75 baseline annotations as covariates . We used this first set of simulations for two purposes . In the S1 Text , we used simulations to illustrate the benefits of using mixture models to post-process the Wald statistics in order to address correlations among annotations and reduce false positives ( S1 Fig ) . In the main text , we used simulations to illustrate the benefits of modeling multiple annotations jointly . For the second set of simulations , we used 10 , 000 SNPs and divided them into 100 blocks with 100 SNPs inside each block . For the null simulations , we set the effect sizes of all SNPs to be zero and performed 50 , 000 simulation replicates . For the alternative simulations , we randomly selected 10 non-adjacent blocks as causal blocks and we randomly selected 20% SNPs inside these causal blocks to be causal SNPs ( i . e . a total of 200 causal SNPs ) . We then simulated ten tissue-specific annotation sets , each with two annotations , which are simulated to correlate with SNP causality [30] . Specifically , the annotation values for the non-causal SNPs are sampled from a normal distribution with mean 0 and variance 1 . The causal SNPs are randomly divided into three groups: for the first annotation , its annotations values for the first group are sampled from a normal distribution with mean 0 and variance 1 while its annotations values for the second and third groups are sampled from a normal distribution with mean 10 and variance 1; for the second annotation , its annotations values for the second group are sampled from a normal distribution with mean 0 and variance 1 while its annotations values for the first and third groups are sampled from a normal distribution with mean 10 and variance 1 . The proportion of the three groups of causal SNPs are set to be either ( 1/2 , 1/2 , 0 ) , ( 1/3 , 1/3 , 1/3 ) or ( 0 , 0 , 1 ) . Because two annotations share similar annotation values in the third group of causal SNPs , the proportion of the third group determines the correlation between the two annotations for causal SNPs within the annotation set . Therefore , the selected proportion of the third group SNPs being 0 , 33% and 100% correspond to low , median and high correlation between the two annotations in causal SNPs , respectively . Once we had the annotations , we simulated the effect sizes for causal SNPs independently from a normal distribution with a SNP-specific variance determined by the designated annotation set . Specifically , we set α0 = 0 . 5 and chose either ( α1 , α2 ) = ( 0 . 4 , 0 . 4 ) ( in the case of two informative annotations ) or ( α1 , α2 ) = ( 0 . 4 , 0 ) ( in the case of one informative annotation ) . These parameters were selected to ensure that the 10 causal blocks explain a large proportion of variance in phenotypes ( per-block PVE > 0 . 01; S2 Fig ) so that we will have reasonable power to detect them . Certainly , power is a continuous function of per-block PVE and is non-zero even for small values of per-block PVE . We simulated the residual errors from a normal distribution with variance 0 . 5 . We summed all genetic effects and the residual errors together to form the simulated phenotypes . We then randomly divided the 10 , 000 individuals into two sets: a training set with 7 , 000 individuals and a test set of 3 , 000 individuals . In the training set , we followed the same procedure described in the previous paragraph to obtain marginal χ2 statistics in the data and SNP correlation matrix from a reference panel to fit our model . We applied the parameter estimates from the best trait-relevant tissue determined in the training set to compute the SNP specific variance σ^j2 as SNP weights . For the computed variance , we subtracted from them the minimal variance across SNPs and added a small constant ( 10−15 ) to ensure that all weights are positive . We then multiplied the resulting SNP weights with the posterior probability ( PP ) of the best trait-relevant tissue and further added a value of 1-PP to all SNPs , thus effectively obtaining a set of averaged SNP weights by using both the constructed SNP weights in the identified trait-relevant tissue and the equal SNP weights . Averaging the constructed weights and the equal weights using PP ensures the robustness of our method and guards against the special case where none of the tissues are trait-relevant: in this case , the resulting SNP weights would equal to the equal weights due to a small PP value and would thus still be effective in the subsequent SNP set analysis . We finally applied the SNP weights constructed in the training data to the test data to perform SNP set analysis . We performed 1 , 000 simulation replicates for each alternative simulation setting . We divided these replicates into 10 sets , each with 100 replicates , and computed the power to detect the causal blocks in each set . We report the mean and variance of these power values across 10 sets . We obtained summary statistics in the form of marginal z-scores for 43 traits from 28 GWAS studies . Details are provided in S2 Table . These studies collect a wide range of complex traits and diseases that can be classified into six phenotype categories [28 , 65]: anthropometric traits ( e . g . height and BMI ) , hematological traits ( e . g . MCHC and RBC ) , autoimmune diseases ( e . g . CD and IBD ) , neurological diseases ( e . g . Alzheimer's disease and Schizophrenia ) , metabolic traits ( e . g . FG and HDL ) , and social traits ( e . g . ever smoked and college completion ) . We removed SNPs within the major histocompatibility complex ( MHC ) region ( Chr6: 25Mb-34Mb ) following [16] . We then intersected the SNPs from all the studies and retained a common set of 622 , 026 SNPs for analysis . We paired the marginal z-scores from these studies with the SNP correlation matrix estimated using 503 individuals of European ancestry from the 1 , 000 Genomes Project [55] for inference . Finally , after the analysis , we computed correlation among traits in terms of their tissue relevance and used the Bayesian information criterion ( BIC ) implemented in the clustering package mclust [66] in R with the standard option EEI to classify traits . Clustering with BIC automatically inferred a total of five main trait clusters . We rely on previous literature to partially validate the inferred trait-relevant tissue results in real data . We reasoned that , if a tissue is indeed relevant to a given trait , then there would be extensive prior biomedical researches carried out on the tissue for the trait . Therefore , the number of previous publications on a trait-tissue pair can serve as a useful indicator on the potential relevance and importance of the tissue for the trait . To estimate the number of previous publications on trait-tissue pairs , we conducted a literature search on PubMed ( https://www . ncbi . nlm . nih . gov/pubmed/ ) . Specifically , for each trait-tissue pair , we used the names of trait and tissue as input and counted the number of publications that contain the input values either in the abstract or in the title . For traits , we used trait names directly . For tissues , we excluded the “Other” tissue group and focused on the nine remaining tissue groups . For these remaining tissue groups , we used the following key words in addition to the tissue group name in the PubMed search: ( 1 ) CNS: brain , central nervous system , neuron , glia and CNS; ( 2 ) BloodImmune: blood , T-cell , B-cell , thymus and immune system; ( 3 ) Adipose: adipose . ( 4 ) AdrenalPancreas: adrenal , pancreas; ( 5 ) BoneConnective: bone , fibroblast and connective tissue; ( 6 ) Cardiovascular: heart , cardiovascular; ( 7 ) Gastrointestinal: gastrointestinal , esophagus , stomach , intestine and rectum; ( 8 ) Liver: liver; ( 9 ) Muscle: muscle . For example , for the schizophrenia-CNS trait-tissue pair , we conducted the search by using “schizophrenia [Title/Abstract] AND ( CNS [Title/Abstract] OR brain [Title/Abstract] OR central nervous system [Title/Abstract] OR neuron [Title/Abstract] OR glia [Title/Abstract] ) ” , which yielded 17 , 720 hits . The number of publications on each trait-tissue pair from the PubMed search is listed in S3 Table ( the search was carried out on June 23 , 2017 ) . For each trait in turn , we further normalized the data by dividing the number of publications for a tissue by the total number of publications across all tissues for the trait . We used the resulting proportion for the final analysis . Because of the close relationship between our method and the sequence kernel association test ( SKAT [49–51] ) ( details in Method Overview ) , we propose to use the estimated SNP specific variance in the top trait-relevant tissue from our method as SNP weights in SKAT [49–51] to perform SNP set test in new association studies . To examine the utility of this association mapping strategy in real data , we estimated annotation coefficients in consortium studies , applied them to construct SNP weights ( details in the above Simulations subsection ) , with which we performed SKAT [49–51] for the corresponding traits in the Wellcome Trust Case Control Consortium ( WTCCC ) study [67] . The WTCCC data consists of about 14 , 000 cases from seven common diseases and 2 , 938 shared controls . The cases include 1 , 963 individuals with type 1 diabetes ( T1D ) , 1 , 748 individuals with Crohn's disease ( CD ) , 1 , 860 individuals with rheumatoid arthritis ( RA ) , 1 , 868 individuals with bipolar disorder ( BD ) , 1 , 924 individuals with type 2 diabetes ( T2D ) , 1 , 926 individuals with coronary artery disease ( CAD ) , and 1 , 952 individuals with hypertension ( HT ) . We excluded HT and considered the remaining six diseases for which we had summary statistics in other larger consortium studies . We obtained quality controlled genotypes from WTCCC [67] and imputed missing genotypes using BIMBAM [68] to obtain a total of 458 , 868 SNPs that are shared across all individuals . The genotypes were further imputed by SHAPEIT [69 , 70] and IMPUTE2 [71] with the 1 , 000 Genomes Project [55] as a reference . We removed SNPs with a Hardy-Weinberg equilibrium p-value < 10−4 or a minor allele frequency < 0 . 05 , and intersected SNPs from WTCCC with the consortium data to obtain a final set of 335 , 170 overlapped SNPs . Meanwhile , we obtained genome locations for a set of 19 , 189 protein coding genes from GENCODE project [72] . We intersected SNPs with these genes and identified gene-harboring SNPs that reside within 10 kb upstream of the transcription start site and 10 kb downstream of the transcription end site . To perform gene-set test , we focused on 5 , 588 genes that have at least 10 SNPs , with an average of 29 . 6 SNPs inside each gene and a total of 153 , 813 gene-harboring SNPs . For each gene in turn , we computed SNP-specific variance using annotation coefficient estimates from the best trait-relevant tissue inferred with consortium study summary statistics for the corresponding trait . We used the SNP-specific variance as SNP weights . As in simulations , for these weights , we subtracted from them the minimal weight across SNPs and added a small constant ( 10−15 ) to ensure that all weights are positive . We then multiplied the resulting SNP weights with the posterior probability ( PP ) of the best trait-relevant tissue and further added a value of 1-PP to all SNPs as the final SNP weights to perform SKAT [49–51] analysis . The SNP-weights for the 43 traits can be downloaded from http://www . xzlab . org/ .
Our first set of simulations are used to illustrate the benefits of using multiple annotations to identify trait-relevant tissues . Details of simulations are provided in Materials and Methods . Briefly , we obtained 27 , 640 SNPs from 10 , 000 randomly selected individuals in the GERA study [63 , 64] and simulated phenotypes . We considered two histone annotations ( H3K4me1 , H3K4me3 ) from ten tissue groups ( S1 Table ) , among which we randomly designated one as the trait-relevant tissue . We then simulated SNP effect sizes under a polygenic model based on the two annotations in the trait-relevant tissue . We added genetic effects with residual errors to form simulated phenotypes . We obtained summary statistics from the data and considered three different approaches to identify trait-relevant tissues: Above , we have included two versions of univariate tests: Uni and UniMax . While the Uni approach is widely applied in previous studies [16 , 27 , 28 , 48] , the UniMax approach can be statistically more appropriate than Uni for summarizing tissue-level evidence . We considered a range of realistic annotation coefficient combinations ( i . e . ( α1 , α2 ) ) . For each combination , we performed 1 , 000 simulation replicates . For each method , we computed the power of various methods in detecting the trait-relevant tissue at a false discovery rate ( FDR ) of 0 . 1 ( Fig 1A ) . In the majority of settings , analyzing multiple annotations jointly also improves power compared with analyzing one annotation at a time . For example , based on power at 10% FDR , SMART is ranked as the best method in 15 out of 25 simulation settings where both annotations have non-zero effects , while UniMax is the best in 10 settings ( Fig 1A ) . While the performance of SMART is often followed by UniMax , the power improvement of SMART compared with UniMax can be large ( median improvement = 9 . 2% ) . Certainly , in the special cases where one annotation coefficient is exactly zero or close to zero , then SMART is often outperformed by UniMax , presumably due to its smaller degree of freedom there . For example , among the 11 settings where at least one annotation has zero effects ( grey area , Fig 1A ) , SMART is ranked as the best method only 4 times , while UniMax is ranked as the best method 7 times . Finally , to further explore the characteristics of annotations that can influence the power of SMART in identifying trait-relevant tissues , we simulated annotations that have various genome-occupancy characteristics and that have various annotation effect sizes and signs ( S1 Text ) . We show that the power of SMART increases with increasing annotation coefficients , is not sensitive to the signs of annotations , and is relatively stable with respect to the genome-occupancy of the annotations as we have standardized the annotations to have mean zero and standard deviation one across the genome ( S3 Fig ) . We examine in detail a simulation setting where ( α1 , α2 ) are chosen to be close to the median estimates ( 0 . 1 , 0 . 05 ) from the real data sets ( i . e . gold shade in Figs 1A and S1 ) . Note that even though these parameters are chosen based on real data , we have much less SNPs or samples in the simulations than in real data and are thus underpowered in simulations . In any case , we first obtained annotation coefficient estimates ( α^1 , α^2 ) across simulation replicates in this setting . We found that the estimates are centered around the truth as one would expect ( Fig 1B ) , suggesting accurate parameter estimation by our approach . Next , in addition to the six approaches listed above , we also included a UniMax_LDSC approach into comparison . In the UniMax_LDSC approach , we applied LDSC to analyze one trait at a time and used the maximum Wald statistics among the two to measure trait-tissue relevance . Different from the UniMax_Wald , however , UniMax_LDSC used a set of 75 baseline annotations to address correlation among annotations . As a result , UniMax_LDSC performs similarly as UniMax in terms of power to detect trait relevant tissues at different FDR thresholds ( Fig 1C ) , suggesting that using mixture modeling is competitive compared to using covariates to control for annotation correlation across tissues . Because both UniMax_LDSC and UniMax use only one annotation , they are often less powerful compared to SMART that models two annotations together ( Fig 1C ) . Our second set of simulations is intended to illustrate the benefits of our method in using inferred trait-relevant tissue to enable more powerful SNP set tests . Here , we ask the question of how to make use of the inferred trait-relevant tissues to enable more powerful future association studies . As explained in the Method Overview section , our model is closely related to the sequence kernel association test ( SKAT ) [49–51] for SNP set test . In particular , the SNP specific variance in our model can be naturally interpreted as the SNP specific weight in the SKAT [49–51] framework . Because of this close relationship between our model and SKAT [49–51] , we propose to use the estimated SNP specific variance in the top trait-relevant tissue from our model as SNP weights to perform analysis in new association studies . Intuitively , if a SNP tends to have a large effect size , then weighting it high in the subsequent SNP set analysis can help achieve greater association mapping power . To explore the possibility of using the inferred SNP-specific variance estimate σ^j2 as a priori weight to construct SNP set test in the SKAT framework [49–51] , we obtained 10 , 000 SNPs from the same set of 10 , 000 individuals in the GERA study [63 , 64] and simulated phenotypes ( Materials and Methods ) . To do so , we divided these SNPs evenly into 100 blocks and randomly selected 10 blocks to be causal blocks . In each casual block , we further selected 20 SNPs to be causal SNPs . We then simulated ten tissue-specific annotation sets with two annotations within each set and designated one set as the trait-relevant tissue . We simulated causal SNP effect sizes based on the two annotations from the trait-relevant tissue and added residual errors to form the simulated phenotypes . Afterwards , we divided individuals randomly into two sets: a training set of 7 , 000 individuals and a test set of 3 , 000 individuals . We applied SMART_EM and UniMax_EM in the training set to identify trait-relevant tissues and to estimate annotation coefficients . We then applied the following weighting options to perform SKAT [49–51] analysis in the test set: We first simulated 50 , 000 replicates under the null where there is no causal SNP so that both α0 and ( α1 , α2 ) are 0 . We used the null simulations to examine the type I error control for various methods and we found that all these methods are well behaved ( Fig 2A ) . Next , we simulated 1 , 000 replicates under the alternative where we have non-zero α0 and ( α1 , α2 ) . We divided SNPs into 100 blocks , among which 10 are causal . We compared different methods in terms of their power to identify the causal blocks . In the simulations , we generated two annotations whose values in the trait-relevant tissue are predictive of SNP causality . The annotation values for the two annotations are almost identical in a certain proportion of causal SNPs ( chosen to be 0% , 33% , or 100% ) so that the two annotations can contain complementary information ( in the case of 0% ) or overlapping information ( in the case of 100% ) . Intuitively , information overlapping in the two annotations would reduce the relative power gain of using multiple annotations versus using a single annotation in constructing SNP set tests . As one would expect , in all settings , constructing SNP weights using the true coefficients from the correct trait-relevant tissue ( i . e . TissueWeight_Oracle; red bars in Fig 2B and 2C ) achieves the greatest power compared with the other methods . For example , compared with using equal weights , using the oracle weights improves power by 14 . 1% on average ( median = 13 . 7% ) across all settings . Importantly , constructing SNP weights using the estimated coefficients from the inferred trait-relevant tissue weight ( i . e . TissueWeight_SMART; green bars in Fig 2B and 2C ) can often achieve almost identical power as TissueWeight_Oracle . Comparing between TissueWeight_SMART and TissueWeight_UniMax , when both annotations have non-zero coefficients , we found that using multiple annotations often leads to greater power gain than using a single annotation . However , as one would expect , when the two annotations contain overlapping information ( e . g . in the case of 100% ) , then using one annotation yields similar power as using two annotations ( green vs blue in Fig 2B ) . In the special case where only one annotation has a non-zero coefficient , then using multiple annotations also has similar power compared with using a single annotation , even when the two annotations contain complementary information ( green vs blue in Fig 2C ) . Next , we explore how various simulation parameters influence the weighted SKAT power ( S1 Text ) . Here , we also include TissueWeight_UniMaxLDSC , where we applied LDSC with the UniMax procedure in the training data and used the coefficient estimate for the annotation with the larger Wald statistics in the inferred trait-relevant tissue to construct SNP weights . In this set of simulations , we varied the annotation coefficients and varied the number of causal blocks . The results are presented in Supplementary S4 Fig . With simulations , we show that the power of SNP set test primarily depends on the phenotype variance explained ( PVE ) by each causal block ( i . e . per-block PVE ) as well as annotation coefficients , and increases with increasing per-block PVE or annotation coefficients . In contrast , power is not influenced by the number of blocks when per-block PVE is fixed to be a constant , though it would decrease with increasing number of blocks when the total PVE is fixed ( as per-block PVE is negatively correlated with number of blocks in this case ) . Importantly , TissueWeight_SMART outperforms TissueWeight_UniMax and TissueWeight_UniMaxLDSC in most scenarios and outperforms EqualWeight in all scenarios . Besides applying our method to infer trait relevant tissues , we first rely on the knowledge gained from previous biomedical literature to measure trait-tissue relevance . To do so , we conducted a PubMed literature search and counted the number of publications existed for each trait-tissue pair . We reasoned that , if a tissue is indeed relevant to a given trait , then there would be a fair number of studies performed on the tissue for the given trait . Therefore , the number of publications carried out on a trait-tissue pair would be a good indicator on their relevance . Next , for each trait in turn , we normalized the count data by computing the proportion of previous publications performed on each of the nine tissue groups ( i . e . the ten tissue groups minus the “Other” group ) , resulting in for each trait a vector of nine proportion values that sum to one ( details in Materials and Methods ) . The number of total publications and proportion values for all traits are summarized in S4 Table . The PubMed literature search results are generally consistent with what we would expect . For example , for schizophrenia ( SCZ ) , 63 . 8% of the previous literatures are focused on CNS , with the rest of the literature scattered across other tissues . The proportion of literature carried out on each trait-tissue pair obtained in PubMed thus provides a reasonable a priori measure of trait-tissue relevance . We use these measurements to validate some of our analysis results using tissue specific annotations . We then applied our method to jointly analyze multiple annotations for each of the three sets of tissue-specific annotations described above . We denote the analysis on annotation set ( 1 ) as HB ( i . e . histone marks , binary ) , analysis on annotation set ( 2 ) as HC ( i . e . histone marks , continuous ) , and analysis on annotation set ( 3 ) as cHMM ( i . e . ChromHMM annotation ) . For HC and cHMM , we obtained the posterior probability values ( PPs ) from each of the 10 tissue groups for each trait . For HB , we first obtained PPs from each of the 105 tissues for each trait . For each trait in turn , we then averaged these PP values within each tissue group and used the averaged tissue group level PP values for the following analyses–this way we can perform comparisons at the tissue group level across different annotation sets . As a comparison , we also applied our method to each of the three annotation sets and performed univariate analysis that corresponds to the UniMax procedure explained in the simulation section . These univariate analyses include HBuMax , which is a univariate analysis of the annotation set ( 1 ) ; HCuMax , which is a univariate analysis of the annotation set ( 2 ) ; cHMMuMax , which is a univariate analysis of the annotation set ( 3 ) ; and HBuMaxLDSC , which is a univariate analysis on the binary annotation set ( 1 ) using LDSC . In these univariate analyses , for each trait in turn , we first selected the annotation with the maximum Wald statistic in every tissue ( or tissue group ) . We then computed the PPs of the selected mark in all tissues ( or tissue groups ) . When necessary , we further averaged PPs ( for HBuMax ) or Wald statistics ( for HBuMaxLDSC ) within each tissue group to allow for comparison at the tissue group level . Overall , we obtained 10 tissue group level PP values or Wald statistics for every trait from each of the five different approaches . We list the top trait-relevant tissue groups with the largest PP value or Wald statistics identified by each of the above approaches in S3 Table , with the corresponding tissue group PPs listed in the same table . The tissue group PPs from HC are also plotted in S5 Fig . The results from those different approaches are largely consistent with the PubMed search results , though notable deviations exist . For example , PubMed search identified CNS to be the most relevant tissue to five neurological traits ( ADD , BIP , SCZ , Autism and Alzheimer ) . Approaches using annotations also identified CNS as the most trait-relevant tissue for four of the five neurological traits ( ADD , BIP , SCZ and Autism ) . However , for Alzheimer’s disease , using tissue-specific annotations revealed BloodImmune tissue as a trait-relevant tissue , which is consistent with recent discoveries that inflammation and microglia are important for Alzheimer’s disease etiology [73 , 74] . As another example , PubMed search identified liver to be the most relevant tissue for hematological traits ( MCHC , MCH , MCV , MPV , PLT and RBC ) , presumably because of liver’s important role in producing extrarenal erythropoietin [75] . In contrast , using tissue-specific annotation highlighted BloodImmune tissue as the most relevant tissue for hematological traits . The similarity and difference between SMART and PubMed search results highlight the importance of using different information to infer trait tissue relevance . We further quantify the similarity between various approaches and PubMed results . To do so , we compare the tissue group level PP values from the annotation integration approaches to the proportion of publications on each tissue group obtained from PubMed search . For each trait and each approach in turn , we computed the correlation between the PP vector for the nine tissue groups and the corresponding proportion values from PubMed search ( Fig 3C ) . We reasoned that , if an approach makes good use of the annotation information , then the trait tissue relevance inferred by this approach would be consistent with the trait tissue relevance measured by PubMed search . Indeed , the Spearman’s rank correlations between different integrative approaches and the PubMed search are reasonable , with an median ( average ) value of 0 . 474 ( 0 . 420 ) , 0 . 417 ( 0 . 379 ) , 0 . 283 ( 0 . 242 ) , 0 . 433 ( 0 . 397 ) , 0 . 433 ( 0 . 376 ) , 0 . 417 ( 0 . 360 ) and 0 . 417 ( 0 . 344 ) for HB , HBuMax , HBuMaxLDSC , HC , HCuMax , cHMM and cHMMuMax , respectively . The correlation results also suggest that , for the same annotation set , using multiple annotations often yields better performance than using one annotation alone ( i . e . HB vs HBuMax or HBuMaxLDSC , HC vs HCuMax , and cHMM vs cHMMuMax ) . Finally , comparing different annotation sets , we found that using 15 chromatin states ( i . e . cHMM and cHMMuMax ) often result in lower correlation than using the annotations based on histone occupancy , suggesting that post-processing histone occupancy data into chromatin states may lose important trait-tissue relevance information , dovetailing earlier findings [76] . To characterize trait-tissue relevance at the tissue group level , we focused on the results from the HC approach in more details and examined the annotation coefficients for the four histone marks inferred from the most trait-relevant tissue . We show the estimates and their significance for individual traits in Fig 3A and then grouped coefficients across 43 traits in Fig 3B . Overall , among the four histone marks , two of them have positive coefficient estimates: the coefficient estimates for H3K4me1 are positive for 42 out of 43 traits , while the coefficient estimates for H3K4me3 are positive for 32 traits ( Fig 3B ) . The positive coefficients for the two histone marks are consistent with their role in marking promotors and enhancers , which are enriched in or near association loci identified by multiple GWASs [11 , 45] . In contrast , the coefficient estimates of H3K27me3 and H3K36me3 are mostly estimated to be close to zero , with approximately random positive or negative signs ( positive signs in 22 traits for H3K27me3 , and in 25 traits for H3K36me3; Fig 3A ) . The near-zero estimates of H3K27me3 and H3K36me3 suggests that SNP effect sizes in polycomb repression regions and transcribed regions often do not differ much from the rest of the genome . In terms of the magnitude of the estimated coefficients , two of the four marks ( H3K4me3 and H3K4me1 ) have large effect estimates , as well as high estimation variation , across all examined traits ( Fig 3B ) . The large coefficient estimates for H3K4me3 and H3K4me1 suggest that both promotor regions and enhancer regions are highly predictive of SNP effect sizes and are often the most informative for inferring trait-tissue relevance . The results with the HC approach are also consistent with what we see in the simulations . In particular , while our extended linear mixed model itself does not explicitly model the correlation among annotations across tissues , our mixture modeling and classification strategy implicitly accounts for annotation correlation across tissues and allows us to identify multiple trait-relevant tissues for a given trait when they are present . Indeed , examining the tissue group PPs from HC ( S5 Fig ) , we found that , among the 43 GWAS traits , more than half of them have one trait-relevant tissue with PP>0 . 5 , while 4 of them have two or more trait-relevant tissues with PPs>0 . 5 . For example , consistent with [16] , the CNS tissue group was identified as the trait-relevant tissue for SCZ , BIP , YE and Ever smoked . Consistent with [16] , the blood immune tissue group was identified as the trait-relevant tissues for CD , RA and UC . While consistent with [16 , 27] , multiple tissue groups , including bone connective , muscle , cardiovascular and adipose , were identified as relevant for height . To further characterize trait-tissue relevance at the tissue level , we examined the results from the HB approach in details . The annotation set ( 1 ) contains binary annotations for 105 tissues that belong to 10 tissue groups . We have only focused on examining group-level results from this set of annotations so far . Here , we focus instead on the PP values for the 105 tissues directly; thus we have a 105-vector of PP values for every trait . We relied on the PP values to rank tissues for every trait . The tissue rank list for each trait represents the tissue footprint of each trait: the trait-relevant tissues are ranked high in the list while the trait-irrelevant tissues are ranked low in the list . With the tissue rank list , we assess the similarity between GWAS traits in terms of their tissue relevance by hierarchical clustering ( Fig 4 ) . We also computed pair-wise spearman correlation between traits based on the tissue rank list ( Fig 5 ) . Overall , applying the Bayesian information criterion ( BIC ) to the correlation plot using mclust [66] revealed five main trait clusters . The first and second trait clusters contain psychiatric disorders ( i . e . SCZ , BIP , Autism , DS , Neuroticism and ADD ) and neurological related traits ( e . g . College , YE , Menarche , Child_Obes , Child_BMI , and EverSmoked ) . For these two clusters , the CNS tissue tends to be identified as the most trait-relevant , consist with previous studies [16 , 27] . Among these traits , BIP , Menarche , College , YE , BIPSCZ , SCZ , DS , Neuroticism are highly correlated with each other and all belong to the first trait cluster; while ADD , Autism , EverSmoked , Child_BMI , Child_Obes are all grouped into the second trait cluster . Among the psychiatric trait pairs in the first two clusters , SCZ and BIP pair has the highest correlation ( spearman correlation = 0 . 561; versus median/mean correlation = 0 . 098/0 . 167; BIPSCZ are excluded ) , suggesting that SCZ and BIP are more similar to each other in terms of tissue relevance footprint compared with the other trait pairs . The third trait cluster contains traits that are often related to several tissue groups . Specifically , the anthropometric traits ( i . e . BL , BW , FNBMD , LNBMD and Height ) are related to the bone , connective and gastrointestinal tissues; while the metabolic traits ( i . e . FG , T2D and HR ) are related to the gastrointestinal , liver and adipose tissues . The fourth trait cluster mainly contains two categories of traits that include immune diseases ( e . g . RA , Lupus , T1D , UC , PBC , CD and IBD ) and hematological traits ( e . g . MCHC , MCH , MCV , PLT , RBC and MPV ) . Both these two categories are related to blood immune tissues . However , the fourth cluster also contains Alzheimer’s disease . The classification of Alzheimer’s disease in the fourth cluster rather than in the first two clusters presumably reflects the close relationship of the disease to both BloodImmune and CNS [28 , 74] . Finally , the fifth trait cluster mainly contains metabolic traits ( TC , TG , LDL , HDL and CAD ) that are related to gastrointestinal and blood immune tissues . Note that traits from the last three clusters tend to have positive correlations among each other , while have negative correlations with traits from the first two clusters ( Fig 5 ) . Finally , we explored the use of annotation coefficient estimates from the inferred trait-relevant tissues to construct SNP set tests in a separate data , the Wellcome trust cast control consortium ( WTCCC ) study . WTCCC contains the six common diseases that include T1D , T2D , CD , BIP , RA and CAD . We focused on a set of 5 , 588 genes and used 153 , 813 SNPs inside these genes to perform SNP set test using SKAT [49–51] ( details in Materials and Methods ) . As in simulations , we considered three different SNP weights for SKAT test: ( 1 ) SNP weights constructed by the multivariate analysis approaches of SMART ( i . e . HC and HB ) ; ( 2 ) SNP weights constructed by the univariate maximal statistics approach ( HCuMax , HBuMaxLDSC and HBuMax ) ; and ( 3 ) equal SNP weights ( EqualWeight ) . We apply different weights to each of the six diseases . We first display the quantile-quantile ( QQ ) plot of–log10 p-values from SKAT in Fig 6 ( for CD ) and S6 Fig ( for T1D , T2D , BIP , RA and CAD ) , which , consistent with simulations , suggests proper control of type I error . In the analysis , different approaches identified a different number of associated genes that pass the Bonferroni corrected genome-wide significance threshold ( 8 . 95x10-6 ) , and these numbers range from 12–15 ( the union of them contains 17 genes ) . These genes are associated with either CD , RA , T1D or T2D , and have all been validated either in the original WTCCC study or in other GWASs of the same trait ( Table 1 ) . Consistent with simulations , we found that SNP set tests using weights constructed from the trait-relevant tissue achieves higher power compared with using equal weights . For example , the HC approach or the HB approach identified 15 genes , 3 more than that identified using equal weights ( Table 1 ) . In particular , the HC approach identified C1orf141 [77] and BSN [78] for CD as well as FTO [79] for T2D , and neither of these were identified by the equal weights approach . While the HB approach identified BSN [78] and SLC22A5 [78] for CD as well as FTO [79] for T2D , and neither of these were identified by the equal weights approach . Finally , within each annotation set , using multiple annotations identified slightly more genes than using one annotation at a time ( i . e . HB vs . HBuMax or HBuMaxLDSC and HC vs . HCuMax ) , again consistent with the simulation results .
We have presented a simple modification of the commonly used linear mixed model to integrate multiple SNP annotations with GWAS traits to facilitate the identification of trait-relevant tissues . We have described an accompanying GEE based parameter inference algorithm that makes use of summary statistics and naturally accounts for genetic correlation due to linkage disequilibrium . We have shown how the task of identifying trait-relevant tissues can be formulated into a classification problem and how mixture modeling can facilitate the inference of trait-relevant tissues in the presence of annotation correlation across tissues . We have also illustrated how the link between our extended linear mixed model and the commonly used SKAT can enable more powerful SNP set tests in new association studies . With both simulations and an in-depth analysis of 43 GWAS traits , we have illustrated the benefits of modeling multiple annotations jointly . Our approach complements several recently developed methods that aim to derive a single , interpretable synthetic annotation by combining information of multiple annotations in a tissue specific fashion [17 , 20 , 27 , 28] . Most of these methods rely on multiple annotation information and use clustering algorithms to cluster SNPs into two categories . The posterior probability of SNPs in the category with the smaller number of SNPs thus becomes a synthetic annotation and is often interpreted as the posterior probability of being a “causal” or “functional” SNP . While these synthetic annotations have the benefits of simplicity and potential interpretability , they often have the drawback of being derived without taking into account the GWAS trait of interest . Arguably , functions of genetic variants depend on traits and clustering SNPs without considering the trait of interest may be suboptimal . Our approach complements these previous methods in that we effectively derive a synthetic annotation by taking GWAS traits into account . In particular , our method can be viewed as a supervised approach to combine multiple annotations into a single annotation , where the single annotation is represented as a weighted summation of the multiple annotations , with the weights being the estimated annotation coefficients inferred directly using the GWAS trait . Therefore , the approach we develop effectively takes the trait of interest into account . Certainly , both our approach and these previous approaches make a key modeling assumption that multiple annotations in the trait-relevant tissue are more relevant to SNP effect sizes or causality as compared with annotations in trait-irrelevant tissues . While it is a reasonable assumption for histone occupancy based annotations we examine here , this assumption may not hold well for certain annotations and for certain complex traits . For example , it is possible the classification of trait-relevant tissue depends on what annotation one examines: the SNP effect sizes can be predicted well by using one annotation in one tissue , or by using another annotation in a different tissue . To briefly explore the utility of our method in the case of multiple trait-relevant tissues , we performed a simulation study that is similar to the polygenic scenario presented in the results section but with two trait-relevant tissues: we used one annotation from one tissue and another annotation from another tissue to simulate SNP effect sizes . In this setting , as one might expect , the difference between the multivariate approach and the univariate approach is small ( S7 Fig ) . Therefore , developing method for the case of multiple trait-relevant tissues is an interesting future direction . We rely on a polygenic model to evaluate the contribution of annotations to SNP effect sizes and infer trait-relevant tissues . Our polygenic model assumes that all SNP effect sizes are non-zero and follow a normal distribution with SNP-specific variance that is a function of multiple annotations . Therefore , our approach is different from several previous approaches that rely on a sparse model to evaluate the contribution of annotations to SNP causality [11 , 12 , 27 , 28 , 39 , 53]We have focused primarily on using tissue-specific annotations based on histone occupancy from the ENCODE and ROADMAP projects . Other tissue-specific annotations are nowadays readily available . For example , the GTEx project measures expression quantitative trait loci ( eQTL ) information in 53 tissues , many of which overlap with that in ROADMAP . Our method can easily incorporate multiple annotations from different data sources and include both eQTL annotations from GTEx and histone annotations from ROADMAP , though caution needs to be made to account for accuracy difference in these eQTL annotations from different tissues due to sample size variation . In any case , jointly analyzing multiple sources of annotations will likely improve power further in identifying trait-relevant tissues in the future . In the real data application , we have attempted to infer fine-scale trait-tissue relevance by using 105 tissues instead of the 10 tissue groups using the HB approach . The inferred top-ranking tissue types from the HB approach for each of the 43 GWAS traits are listed in Supplementary S5 Table . As expected , most of these top-ranking tissues belong to the top-ranking tissue group ( median = 70 . 5% across traits ) , suggesting relatively stable inference results whether tissues or tissue groups were used in the analysis . For example , all of the top-ranking tissues ( with PP > 0 . 5 ) for ever smoked and YE belong to the CNS tissue group , and 28 of the 39 top-ranking tissues for CD belong to the blood immune tissue group . We have attempted to further quantify the tissue-level relevance results by comparing them to the corresponding PubMed search results , as we have done in the main text for the tissue group analysis . However , we found that PubMed search results are unable to distinguish fine-scale tissue types for most traits . Therefore , we had to rely on prior biology knowledge obtained in various other studies to validate our tissue relevance analysis . In many cases , the top-ranking tissue fits our prior expectation . However , we also acknowledge that identifying relevant tissues from >100 tissue types is indeed a challenging task . Specifically , for 34 out of 43 traits , the PPs for more than half of the tissues within the corresponding top-ranking tissue group are greater than 0 . 5 , suggesting that it is often difficult to identify a single trait-relevant tissue within the tissue group . Alternative approaches to explore fine-scale trait-tissue relevance have been suggested before . For example , a two-step analysis procedure was proposed to first identify trait-relevant tissue group and then identify trait-relevant tissue within the tissue group [80] . In addition , using synthetic annotations generated from Genoskyline [28] or FUN-LDA [81] could be particularly useful for identifying fine scale trait-relevant tissues . Our method can be easily adapted to incorporate a two-step analysis procedure and/or accommodate synthetic annotations , and has the potential to yield better trait-tissue relevance resolution in the future . Finally , while we have mainly focused on inferring trait-relevant tissues , we have also explored the feasibility of using inferred trait-relevant tissues and the estimated annotation coefficients to enable more powerful SNP set test in future GWASs . In practice , multiple annotation sets can be used to construct SNP set tests ( e . g . HC and HB annotations sets as used in our real data application ) . It is often difficult a priori to determine which annotation set would yield the best results . Therefore , we recommend analyzing all of these annotation sets separately and choose the one that yields the highest power , as we have done in the real data application . In addition , sometimes the trait of interest may have multiple relevant tissues . In this case , we can apply the PPs from the identified trait-relevant tissues ( with PP>0 . 5 ) to weight the corresponding estimated annotation coefficients from these tissues to form a set of weighted annotation coefficients , in line with the Bayesian model averaging idea . Finally , while incorporating annotation information does increase SNP set test power , we also found that such power improvement in realistic settings depends on traits and can be variable . The variability in power improvement of our method is consistent with many previous studies that have shown similar variability in power improvement by integrating SNP annotations into single SNP association tests [11 , 39] . However , increasing sample size and the development of better SNP annotations will likely facilitate the adaption of various annotation integration methods in the near future . | Identifying trait-relevant tissues is an important step towards understanding disease etiology . Computational methods have been recently developed to integrate SNP functional annotations generated from omics studies to genome-wide association studies ( GWASs ) to infer trait-relevant tissues . However , two important questions remain to be answered . First , with the increasing number and types of functional annotations nowadays , how do we integrate multiple annotations jointly into GWASs in a trait-specific fashion ? Doing so would allow us to take advantage of the complementary information contained in these annotations to optimize the performance of trait-relevant tissue inference . Second , what to do with the inferred trait-relevant tissues ? Here , we develop a new statistical method and software to make progress on both fronts . For the first question , we extend the commonly used linear mixed model , with new algorithms and inference strategies , to incorporate multiple annotations in a trait-specific fashion to improve trait-relevant tissue inference accuracy . For the second question , we rely on the close relationship between our proposed method and the widely-used sequence kernel association test , and use the inferred trait-relevant tissues , for the first time , to construct more powerful association tests . We illustrate the benefits of our method through extensive simulations and applications to a wide range of real data sets . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"genome-wide",
"association",
"studies",
"sociology",
"dna-binding",
"proteins",
"social",
"sciences",
"simulation",
"and",
"modeling",
"probability",
"distribution",
"mathematics",
"statistics",
"(mathematics)",
"test",
"statistics",
"genome",
"analysis",
"molecular",
"gen... | 2018 | Identifying and exploiting trait-relevant tissues with multiple functional annotations in genome-wide association studies |
Genetic risk prediction is an important goal in human genetics research and precision medicine . Accurate prediction models will have great impacts on both disease prevention and early treatment strategies . Despite the identification of thousands of disease-associated genetic variants through genome wide association studies ( GWAS ) , genetic risk prediction accuracy remains moderate for most diseases , which is largely due to the challenges in both identifying all the functionally relevant variants and accurately estimating their effect sizes in the presence of linkage disequilibrium . In this paper , we introduce AnnoPred , a principled framework that leverages diverse types of genomic and epigenomic functional annotations in genetic risk prediction for complex diseases . AnnoPred is trained using GWAS summary statistics in a Bayesian framework in which we explicitly model various functional annotations and allow for linkage disequilibrium estimated from reference genotype data . Compared with state-of-the-art risk prediction methods , AnnoPred achieves consistently improved prediction accuracy in both extensive simulations and real data .
Achieving accurate disease risk prediction using genetic information is a major goal in human genetics research and precision medicine . Accurate prediction models will have great impacts on disease prevention and early treatment strategies [1] . Advancements in high-throughput genotyping technologies and imputation techniques have greatly accelerated discoveries in genome-wide association studies ( GWAS ) [2] . Various approaches that utilize genome-wide data in genetic risk prediction have been proposed , including machine-learning models trained on individual-level genotype and phenotype data [3–8] , and polygenic risk scores ( PRS ) estimated using GWAS summary statistics [9 , 10] . Despite the potential information loss in summary data , PRS-based approaches have been widely adopted in practice since the summary statistics for large-scale association studies are often easily accessible [11 , 12] while individual-level data are more difficult to acquire , deposit , and process . However , prediction accuracies for most complex diseases remain moderate , which is largely due to the challenges in both identifying all the functionally relevant variants and accurately estimating their effect sizes in the presence of linkage disequilibrium ( LD ) [13] . Explicit modeling and incorporation of external information , e . g . pleiotropy [7 , 8] and LD [10] , has been shown to effectively improve risk prediction accuracy . Recent advancements in integrative genomic functional annotation , coupled with the rich collection of summary statistics from GWAS , have enabled increase of statistical power in several different settings [14–16] . To our knowledge , the impact of functional annotations on performance of genetic risk prediction has not been systematically studied . Here , we introduce AnnoPred ( available at https://github . com/yiminghu/AnnoPred ) , a principled framework that integrates GWAS summary statistics with various types of annotation data to improve risk prediction accuracy . We compare AnnoPred with state-of-the-art PRS-based approaches and demonstrate its consistent improvement in risk prediction performance using both simulations and real data of multiple complex diseases . AnnoPred risk prediction framework has three main stages ( Methods ) . First , we estimate GWAS signal enrichment in 61 different annotation categories , including functional genome predicted by GenoCanyon scores [17] , GenoSkyline tissue-specific functionality scores of 7 tissue types [14] , and 53 baseline annotations for diverse genomic features [18] for each trait analyzed . Second , we propose an empirical prior of SNP effect size based on annotation assignment and signal enrichment . In general , SNPs located in annotation categories that are highly enriched for GWAS signals receive a higher effect size prior . Finally , the empirical prior is adopted in a Bayesian framework in which marginal summary statistics and LD matrix estimated from a reference panel are jointly modeled to infer the posterior effect size of each SNP . AnnoPred PRS is defined by PRS=∑j=1MXjEA ( βj|β^ , D^ ) where Xj and βj are the standardized genotype and effect size of the jth SNP , respectively , β^ is the marginal estimate of β , D^ is the sample LD matrix , and EA ( βj|β^ , D^ ) denotes the posterior expectation of effect sizes under an empirical prior based on annotation assignment for all SNPs when adjusting for LD matrix estimated from a reference panel ( Methods ) .
We first performed simulations to demonstrate AnnoPred’s ability to improve risk prediction accuracy . We compared AnnoPred with four popular PRS approaches ( Methods ) , including PRS based on genome-wide significant SNPs ( PRSsig ) , PRS based on all SNPs in the dataset ( PRSall ) , PRS based on tuned cutoffs for p-values and LD pruning ( PRSP+T ) , and recently proposed LDpred [10] . Mean correlations between simulated and predicted traits were calculated from 100 replicates under different simulation settings ( Methods ) . AnnoPred showed the best prediction performance in all settings when the causal SNPs are highly enriched in annotated regions ( Table 1 , S2 Table and S2 Fig ) . In general , performance of PRSsig , PRSP+T , LDpred , and AnnoPred all improved under a sparser genetic model and higher trait heritability . PRSall showed comparable performance between sparse and polygenic models but its prediction accuracy was consistently worse than other methods . Sample size in the training set was also crucial for risk prediction accuracy . Increasing sample size could lead to continuous improvement in prediction accuracy under different settings ( Fig 1 ) . To illustrate the improved risk prediction performance in real data , we applied AnnoPred to five human complex diseases—Crohn’s disease ( CD ) , breast cancer ( BC ) , rheumatoid arthritis ( RA ) , type-II diabetes ( T2D ) , and celiac disease ( CEL ) . We first estimated GWAS signal enrichment in different annotation categories ( Methods ) . Enrichment pattern varies greatly across diseases ( Fig 2A; S1 Table ) , reflecting the genetic basis of these complex phenotypes . Functional genome predicted by GenoCanyon was consistently and significantly enriched for all five diseases . Blood was strongly enriched for three immune diseases , namely CD ( P = 8 . 9×10−12 ) , CEL ( P = 7 . 0×10−15 ) , and RA ( P = 9 . 9×10−6 ) , while gastrointestinal ( GI ) tract was enriched in CD ( P = 2 . 6×10−5 ) and CEL ( P = 1 . 4×10−4 ) , both of which have a known GI component . For BC , epithelium ( P = 7 . 4×10−4 ) , GI ( P = 5 . 9×10−3 ) , and muscle ( P = 6 . 1×10−3 ) were significantly enriched . A few studies have shown that breast cancer could arise from epithelial cells [19 , 20] . The connections between breast cancer and muscle as well as GI tract have also been previously suggested [21 , 22] . In addition , studies have suggested that GI can be used as diagnostic and treatment target for type-II diabetes , Crohn’s disease , and celiac disease [23–25] . Furthermore , the connection between immune system and Crohn’s disease , celiac disease and rheumatoid arthritis have been extensively studied in literature [26–28] . Next , we evaluated the effectiveness of proposed empirical effect size prior in three diseases ( i . e . CD , CEL , and RA ) with well-powered testing cohorts ( N>2 , 000 ) . Interestingly , despite the highly variable enrichment results in training datasets , integrative effect size prior could effectively identify SNPs with large effect sizes and consistent effect directions in independent validation cohorts ( Fig 2B and 2C ) . Correlations between the calculated PRS and disease status ( COR ) for different approaches are summarized in Table 2 . AnnoPred showed consistently improved prediction accuracy compared with all other methods across five diseases . Notably , PRSsig and PRSall showed suboptimal performance in these datasets , reaffirming the importance of modeling LD and other external information . A likelihood ratio test was used to test for the difference in the prediction accuracy between models comparing the likelihood of a logistic regression fitting PRS of one method to that of a logistic regression fitting PRS of two methods jointly ( S11 Table ) . From the test , AnnoPred with 61 annotations performed significantly better than LDpred ( p = 1 . 2E-22 for CD , p = 0 . 045 for BC , p = 4 . 2E-7 for RA , p = 3 . 3E-4 for T2D and p = 1 . 3E-3 for CEL ) . Reversing the order of test ( that is , comparing the likelihood of model using annotations with model using and not using annotations jointly ) results in non-significant p-values for most tests ( S11 Table ) , which further demonstrates that PRS incorporating functional annotations mostly encompasses the information of PRS without annotations . To test different methods’ ability to stratify individuals with high risk , we compared the proportion of cases among testing samples with high PRS . AnnoPred outperformed all other methods in CD , CEL , RA , and T2D ( S1 Fig ) . Next , we tested AnnoPred’s performance using only the 53 baseline annotations and observed a substantial drop in prediction accuracy for all diseases ( S3 Table ) . AnnoPred with GenoCanyon and GenoSkyline annotations only ( nine annotation tracks in total ) yields better performance than the 53 baseline annotations ( S10 Table ) . For CD and T2D , by using these 9 categories AnnoPred even achieved higher accuracy than the model with all 61 annotation tracks added . These results highlight the importance of annotation quality in genetic risk prediction , and also demonstrate GenoCanyon and GenoSkyline’s ability to accurately identify functionality in the human genome . Since different diseases have various enrichment patterns , we also run AnnoPred with significantly enriched annotations ( enrichment test p value less than 0 . 05 ) for each disease ( S10 Table ) . In general , using only the significantly enriched annotations indeed improved the performance in most diseases . Tissue specificity plays an important role in genetic risk prediction . Integrating more functional annotations with higher tissue and cell type specificity may further increase risk prediction accuracy , especially when the tissue type that is biologically relevant to the disease is not well characterized by the seven available tissue tracks in our current analyses . To explore how these factors will affect the AnnoPred model , we performed a few follow-up analyses . We have recently expanded our GenoSkyline annotations to more than 100 tissue and cell types from the Roadmap Epigenomics Project [29] . We investigated the performance of AnnoPred after integrating 66 annotation tracks representing a spectrum of adult tissue and cell types . As shown in S10 Table , incorporating more annotations into the model does not always further improve risk prediction accuracy compared with AnnoPred with fewer annotations in the model . This may be due to the overlap between functional regions ( e . g . functional annotations for slightly different brain regions ) when incorporating too many annotation tracks into the model , which will cause numerically unstable heritability estimates . This is because annotation-stratified LD score regression , the method we used to empirically estimate the informative prior for SNPs’ effect sizes , is a multiple linear regression model that regresses SNP-level summary statistics against annotation-stratified LD scores . When two functional annotation tracks are similar , the corresponding LD scores will also be correlated by definition . It is well understood that if multi-collinearity ( i . e . correlation among covariates ) in multiple regression leads to numerically unstable estimates for regression coefficients [30] ( the heritability parameters in our case ) . In order to study the effect of highly associated SNPs ( e . g . SNPs in MHC regions for immune traits ) , we repeated the analysis on CD , RA , BC and T2D after removing the SNPs in MHC region ( chr6: 28 , 477 , 797–33 , 448 , 354 bp ) . Re-analysis of CEL was unnecessary since the training summary statistics of CEL does not contain any SNP in the MHC region . After removing SNPs in MHC regions , the prediction accuracies for RA drops dramatically for all methods and AnnoPred remained to be the method with the best performance ( S9 Table ) . For the rest diseases , results varied little from the original analysis . Besides COR , we also included AUCs for all the analysis performed ( S2 , S6 , S9 and S10 Tables ) , all of which showed consistent patterns . Due to distinct allele frequencies and LD structures across populations , risk prediction accuracy usually drops when the training and testing samples are from different populations . In order to investigate the robustness of AnnoPred against population heterogeneity , we applied AnnoPred to three non-European cohorts for breast cancer and type-II diabetes while training the model using summary statistics from European-based studies . The CORs and AUCs are summarized in S6 and S7 Tables . As expected , we observed a drop in prediction accuracy for all methods . However , AnnoPred still performed the best in all three trans-ethnic validation datasets .
Our work demonstrates that functional annotations can effectively improve performance of genetic risk prediction . AnnoPred jointly analyzes diverse types of annotation data and GWAS summary statistics to upweight SNPs with a higher likelihood of functionality , which lead to consistently better prediction accuracy for multiple complex diseases . Our method is not without limitation . First , despite the consistent improvement compared with existing PRS-based methods , accuracies for most diseases remain moderate . In order to effectively stratify risk groups for clinical usage , our model remains to be further calibrated using large cohorts with measured environmental and clinical risk factors [1] . Second , accurate estimation of GWAS signal enrichment and SNP effect sizes requires a large sample size for the training dataset . This could potentially be improved by new estimators for annotation-stratified heritability [19] . A few Bayesian models combining GWAS summary statistics with functional annotations have been proposed for the purpose of fine-mapping functional variants [16 , 20 , 21] . Whether these models could be adapted to benefit risk prediction accuracy remains to be investigated in the future . Importantly , the rich collection of publicly available integrative annotation data , in conjunction with the increasing accessibility of GWAS summary statistics , makes AnnoPred a customizable and powerful tool . As GWAS sample size continues to grow , AnnoPred has the potential to achieve even better prediction accuracy and become widely adopted as a summary of genetic contribution in clinical applications of risk prediction .
GenoCanyon is a statistical framework to predict functional regions in the human genome through integrative analysis of ENCODE epigenomic data and multiple conservation metrics [17] . Later we have further extended the model and developed GenoSkyline , which aimed to predict tissue-specific functionality [14] . In the AnnoPred model , we incorporated GenoCanyon general functionality scores , GenoSkyline tissue-specific functionality scores for seven tissue types ( brain , gastrointestinal tract , lung , heart , blood , muscle , and epithelium ) , and 53 LDSC baseline annotations that covered a variety of genomic features [18] ( S1 Table ) . We smoothed GenoCanyon scores by a 10Kb window , a strategy previously shown to improve robustness of functionality prediction [22] . The smoothed GenoCanyon annotation and raw GenoSkyline annotations of seven tissue types were dichotomized based on a cutoff of 0 . 5 . The regions with GenoCanyon or GenoSkyline scores greater than the cutoff were interpreted as non-tissue-specific or tissue-specific functional regions in the human genome . Such dichotomization has been previously shown to be robust against the cutoff choice [14] . Notably , the AnnoPred framework allows users to specify their own choice of annotations . We assume throughout the paper that both the phenotype YN×1 and the genotypes XN×M are standardized with mean zero and variance one . We assume a linear model YN×1=XN×MβM×1+εN×1 X , β and ε are mutually independent . We also assume that β is a random effect and effects of different SNPs are independent . A key idea in the AnnoPred framework is to utilize functional annotation information to accurately estimate SNPs’ effect sizes . In order to achieve that , we first partition trait heritability by annotations using LD score regression [18] . Since genotypes are standardized , per-SNP heritability is defined as the variance of βi for the ith SNP , and is used to quantify SNP effect sizes . More specifically , assume there are K + 1 pre-defined annotation categories , denoted as S0 , S1 , … , SK with S0 representing the entire genome . Under an additive assumption for heritability in overlapped annotations , we have βi∼N ( 0 , ∑j:i∈Sjτj ) , where τ0 , τ1 , … , τK , quantify the contribution to per-SNP heritability from each annotation category . Denote the estimated marginal effect size of the ith SNP as β^i=XiTYN , then we have the following approximation E ( Nβ^i2 ) ≈ ( N−1 ) ∑kτkl ( i , k ) +1 where l ( i , k ) is the annotation-stratified LD score and N denotes the total sample size . Regression coefficients τk are estimated through weighted least squares . The estimated heritability of the ith SNP is then Var^ ( βi ) =∑j:i∈Sjτ^j . Based on per-SNP heritability estimates , we propose two different priors for SNP effect sizes to add flexibility against different genetic architecture . For the first prior , we assume that SNP effect size follows a spike-and-slab distribution βi∼p0N ( 0 , σ^i2p0 ) + ( 1−p0 ) δ0 where p0 is the proportion of causal SNPs in the dataset , and δ0 is a Dirac function representing a point mass at zero . The empirical variance of each SNP , i . e . σ^i2 , is determined by the annotation categories it falls in . More specifically , we assume σ^i2=c ( ∑j:i∈Sjτ^j ) , where c is a constant calculated from the following equation ∑iσ^i2=H^2 . We do not directly use ∑j:i∈Sjτ^j as the empirical variance prior because it is estimated in the context where all SNPs in the 1000 Genomes Project database are included in the model [18] . Such per-SNP heritability estimates cannot be extrapolated to the risk prediction context where many fewer SNPs are analyzed [23] . Therefore , we rescale the heritability estimates to better quantify each SNP’s contribution toward chip heritability . Following [24] , we use a summary statistics-based heritability estimator that approximates the Haseman-Elston estimator: H^2= ( χ¯2−1 ) Nl− where χ¯2 and l¯ denote the mean Nβ^i2 and mean non-stratified LD score , respectively . In the first prior , we assumed the same proportion of causal SNPs but different effect sizes across annotation categories . We now describe the second prior that assumes different proportions of causal SNPs but the same effect size across annotation categories . To be specific , we assume the causal effect size to be Var ( βcausal ) = V , the total number of SNPs to be M0 , and the overall proportion of causal SNPs to be p0 . The total heritability H02 can then be written as H02=p0M0V . For the ith SNP , use Ti= ( ⋂j:i∈SjSj ) ∩ ( ⋂k:i∉SkSkc ) to denote the collection of SNPs that share the same annotation assignment with the ith SNP , and let MTi=|Ti| , i . e . the number of SNPs in the set . Then , the total heritability of SNPs in Ti is HTi2=pTiMTiV , with pTi denoting the proportion of causal SNPs in Ti . Following these notations , we have βi∼pTiN ( 0 , V ) + ( 1−pTi ) δ0 where V=H0p0N0 and pTi=p0M0HTi2MTiH02 . We use H^2 to estimate H02 , and the following formula to estimate HTi2 . Finally , p0 is treated as a tuning parameter for both prior functions in our analysis . By Bayes’ rule , the posterior distribution of β is: f ( β|β^ , D^ ) ∝f ( β^|β , D^ ) f ( β ) where D^=1NXTX is the sample correlation matrix and β^=1NXTY is the marginal effect size estimates . Given β and D^ , β^ follows a multivariate normal distribution asymptotically with the following mean and variance E ( β^|β , D^ ) =1N[E ( XTXβ|β , D^ ) +E ( XTε|β , D^ ) ]=D^β Var ( β^|β , D^ ) =Var ( 1NXTε|β , D^ ) =1N ( 1−hg2 ) D^ . However , D^ is usually non-invertible and has very high dimensions . We thus study the posterior distribution of a small chunk of β^ instead . Let β^b be the estimated marginal effect size of SNPs in a region b ( e . g . a LD block ) and the corresponding genotype matrix is Xb and sample correlation matrix is D^b . Then the conditional mean and variance of β^b are E ( β^b|βb , D^b ) =1N[E ( XbTXβ|βb , D^b ) +E ( XbTε|βb , D^b ) ]=D^bβb Var ( β^b|βb , D^b ) =1N2var ( XbTXbβb+XbT ( X−bβ−b+ε ) |βb , D^b ) =1N2var ( XbT ( X−bβ−b+ε ) |βb , D^b ) =1N2XbTvar ( X−bβ−b+ε|βb , D^b ) Xb=1N ( 1−hb2 ) D^b where hb2=∑i∈bσi2 is the heritability of SNPs in region b , and X−b and β−b denote the genotype matrix and effect sizes of SNPs not in region b . The conditional distribution of βb is: f ( βb|β^b , D^b ) ∝N ( D^bβb , 1N ( 1−hb2 ) D^b ) ∏i∈bf ( βi ) ∝{N ( D^bβb , 1N ( 1−hb2 ) D^b ) ∏i∈b[ p0N ( 0 , σi2p0 ) + ( 1−p0 ) δ0 ] , underthefirstpriorN ( D^bβb , 1N ( 1−hb2 ) D^b ) ∏i∈b[ pTiN ( 0 , V ) + ( 1−pTi ) δ0 ] , underthesecondprior Although it is difficult to derive E ( βb|β^b , D^b ) from the joint conditional distribution of βb , each element of βb follows a mixed normal distribution conditioning on β^b , D^b , and all other elements in βb . Therefore , we apply a Gibbs sampler to draw samples from f ( βb|β^b , D^b ) and use the sample mean as an approximation for E ( βb|β^b , D^b ) . We further performed a sensitivity analysis on the choice of the size of block b ( S6 Fig ) . Specifically , we ran AnnoPred on the data of Crohn’s disease with different sizes of block and found that the results were robust to the sizes . In practice , the size of block b is specified by the total number of variants divided by 3 , 000 . PRS is calculated using the following formula PRS=∑j=1MXjEA ( βj|β^ , D^ ) , where EA denotes the posterior expectation as described above . In practice , the individual-level genotype matrix is not available and we use the LD matrix estimated from a reference panel or the validation samples to substitute D^ . We apply the same standard of choosing the size of b as described in [10] . Choices of prior and p0 can be tuned in an independent cohort . For the data analysis described in this work , we adopted a cross-validation scheme to select tuning parameter due to the challenge in finding multiple independent cohorts without overlapping with the training GWAS summary statistics . The training datasets in our real data analyses and simulations are always fixed , i . e . GWAS summary statistics . We did not perform a classical cross-validation by using different subsets of the complete data to train and test our prediction model . The purpose of cross-validation in our study is purely parameter tuning . To select a suitable tuning parameter , we divide the independent testing dataset ( individual level genotype and phenotype data ) into two equal parts ( A and B ) , and select the tuning parameters by optimizing prediction accuracy on dataset A . We then evaluate prediction accuracy using the remaining half of testing data , i . e . dataset B . Finally , we repeat the analysis one more time by choosing the tuning parameter on dataset B while evaluating the prediction accuracy on dataset A . Results from these two separate analyses are averaged to quantify model performance . For T2D where multiple independent cohorts are available ( phs000237 and phs000388 ) , we used an independent cohort for parameter tuning and the other for evaluating performance ( S12 Table ) . The results are consistent with the cross-validation . We compared AnnoPred with several commonly used risk prediction methods based on summary data of association studies . PRSsig and PRSall were both calculated as the inner product of marginal effect size estimates and the corresponding genotypes . PRSall used all the SNPs that are shared between training and testing datasets while PRSsig only used SNPs with p-values below 5 × 10−8 in the training set . PRSP+T used SNPs passing both LD pruning and p-value thresholding . The thresholds are tuned in an independent dataset over a grid ( 0 , 0 . 1 , 0 . 2 , … 0 . 9 for LD; 1 , 0 . 3 , 0 . 1 , 0 . 03 , 0 . 01 , 3E-3 , 1E-3 , 3E-4 , 1E-4 , 3E-5 , 1E-5 , 1E-6 , 1E-7 , 5E-8 , 1E-8 for p-value ) . LDpred can be viewed as a special case of AnnoPred , assuming the whole genome as the only functional annotation . This is because when enrichment is constant ( i . e . causal variants are uniformly distributed across the genome ) , per-SNP heritability estimates would be nearly constant and therefore results in similar performance to LDpred . We have performed an additional simulation to demonstrate this using WTCCC genotype data with ~15K individuals and ~330K variants . We randomly divided the genome into two parts ( two annotations ) and uniformly selected causal SNPs . Then the traits were simulated in a similar way as other simulations in this paper . We estimated per-SNP heritability using LDSC in the two annotation categories , respectively . We ran the procedure for 100 times and the distributions of estimated per-SNP heritability in both regions are summarized in the figure below ( the dashed line denotes the true per-SNP heritability , added as S4 Fig in the manuscript ) , which indicates that the per-SNP heritability estimates are uniform across the genome under constant enrichment . Therefore , AnnoPred would be mathematically equivalent with LDpred with enrichment is constant . We downloaded python code for PRSP+T and LDpred from Github ( https://github . com/bvilhjal/ldpred ) . All the tuning parameters were tuned through cross-validation as we did for AnnoPred . Besides all these PRSs , we also compared AnnoPred with a evaluating method used in [5] , which uses 1E-1 , 1E-2 , … , 1E-5 as p-value threshold to select SNPs and report the accuracy for the best performed threshold ( S4 and S5 Tables ) . Given that many large-scale GWAS summary statistics have included almost all available cohorts for a disease of interest , it is challenging to find independent datasets with individual-level genotype and phenotype information and sufficient sample sizes . We were able to identify ideal validation datasets for the five diseases we analyzed in this paper . The performance of different methods on more traits shall be evaluated when we get access to more data in the future . We simulated traits from WTCCC genotype data , which contain 15 , 918 individuals genotyped for 393 , 273 SNPs after filtering variants with missing rate above 1% and individuals with genetic relatedness above 0 . 05 . We first generated two annotations and each annotation was simulated by randomly selecting 10% of the genome , denoted as A1 and A2 , which we assume are known when applying AnnoPred . Denote the heritability of the trait as hg2 ( 25% or 50% ) and the number of causal variants as m ( 300 or 3 , 000 ) . Causal variants were generated as follows: m3 causal variants were selected from A1 , m3 from A2 and the rest from ( A1UA2 ) C corresponding to a high enrichment of signals in A1 and A2 . Effect sizes of causal variants were sampled from N ( 0 , hg2m ) . For each simulation , we used 70% of the data to calculate the training summary statistics and randomly divided the rest 30% into two parts for parameter tuning . We also randomly selected half of the training data to calculate summary statistics in order to study the effect of sample size on prediction accuracy . In order to evaluate the improvement in accuracy , we performed a permutation test to compare the CORs of AnnoPred and LDpred . Suppose the CORs of LDpred and AnnoPred in simulations are x1 , x2 , … , xn and y1 , y2 , … , yn , respectively . And the hypothesis we want to test is H0:μx=μy↔H1:μx≠μy where μx and μy represent the population mean of accuracies of LDpred and AnnoPred . We used |x¯−y¯| as the test statistics and the p value can be calculated as p=Pr ( |x¯−y¯ ) >|x¯obs−y¯obs||H0 ) , in which x¯−y¯ represents the random variable and x¯obs−y¯obs represents the actually observed values . We used permutation to approximate the distribution of ( x¯−y¯ ) when H0 is true . Specifically , we first pooled xi′s and yi′s together . Then x˜1 , x˜2 , … , x˜n and y˜1 , y˜2 , … , y˜n were sampled from the pooled data for N = 106 times and we calculated ( x˜¯−y˜− ) for each x˜i′s and y˜i′s sampled , which formed the empirical distribution of ( x¯−y¯ ) under H0 . And the p value could be approximated by p^=∑k=1NI{|x˜¯k−y˜¯k|>|x¯obs−y¯obs|}N , in which x˜¯k−y˜¯k represents the sampled test statistic of the kth permutation . To further study the effect of sample size on prediction performance , we simulated traits using SNPs of chromosome 1 , chromosomes 1 and 2 , chromosomes 1 to 4 and the whole genome while keeping the same proportion of causal variants and heritability to mimic the situation of increasing sample size . The corresponding relative sample sizes ( NMMs , where N is the number of individuals , M is the total number of variants and Ms is the number of variants used in simulation ) for the four scenarios are ~135K , ~67K , 37K and ~11K . For each effective sample size , we simulated traits under four settings: h2 = 0 . 25 , p = 0 . 001; h2 = 0 . 25 , p = 0 . 01; h2 = 0 . 5 , p = 0 . 001; h2 = 0 . 5 , p = 0 . 01 , where p represents the proportion of causal variants ( Fig 1 ) .
The study was approved by YALE UNIVERSITY HUMAN INVESTIGATION COMMITTEE with approval number 100 FR1 and 100 FR27 .
We trained AnnoPred using publicly accessible GWAS summary statistics and evaluated risk prediction performance using individual-level genotype and phenotype data from cohorts independent from the training samples . Only SNPs shared between training and testing datasets were kept in our analyses . Details for each training and testing dataset are provided in S1 Text and S8 Table . For Crohn’s disease , we trained the model using summary statistics from International Inflammatory Bowel Disease Genetics Consortium ( IIBDGC; Ncase = 6 , 333 and Ncontrol = 15 , 056 ) [25] . Samples from the Wellcome Trust Case Control Consortium ( WTCCC ) were removed from the meta-analysis and used as the validation dataset ( Ncase = 1 , 689 and Ncontrol = 2 , 891 ) [26] . For breast cancer , we trained the model using summary statistics from Genetic Associations and Mechanisms in Oncology ( GAME-ON ) study ( Ncase = 16 , 003 and Ncontrol = 41 , 335 ) [27] , and tested the performance using samples from the Cancer Genetic Markers of Susceptibility ( CGEMS ) study ( Ncase = 966 and Ncontrol = 70 ) [28] . Shared samples between CGEMS and GAME-ON were removed . We used samples from the CIDR-GWAS of breast cancer for trans-ethnic analysis ( Ncase = 1 , 666 and Ncontrol = 2 , 038 ) [29] . For rheumatoid arthritis , we used summary statistics from a meta-analysis with 5 , 539 cases and 20 , 169 controls to train the model [30] . WTCCC samples were removed from the meta-analysis and used for validation ( Ncase = 1 , 829 and Ncontrol = 2 , 892 ) [26] . For type-II diabetes , the training dataset is Diabetes Genetics Replication and Meta-analysis ( DIAGRAM ) consortium GWAS with 12 , 171 cases and 56 , 862 controls [31] . We used samples from Northwestern NUgene Project for validation ( Ncase = 662 and Ncontrol = 517 ) [32] . Samples from Institute for Personalized Medicine ( IPM ) eMERGE project are used for trans-ethnic analysis ( African American: Ncase = 517 and Ncontrol = 213; Hispanic: Ncase = 477 and Ncontrol = 102 ) [33] . The training dataset for celiac disease is from a GWAS with 4 , 533 cases and 10 , 750 controls [34] . Samples in the National Institute of Diabetes and Digestive and Kidney Diseases ( NIDDK ) celiac disease study were used for validation ( Ncase = 1 , 716 and Ncontrol = 530 ) [35] . AnnoPred software and source code are freely available online at https://github . com/yiminghu/AnnoPred . | Genetic risk prediction plays a significant role in precision medicine . Accurate prediction models could have great impact on disease prevention and early treatment strategies . For example , mutations in BRCA1 and BRCA2 have been used to evaluate women’s breast cancer risk and as a guideline for early screening . However , genetic risk prediction models also present important challenges , including extreme high-dimensionality , limited access to and efficient computational methods for individual-level genotype data . To make use of rich GWAS summary statistics , we propose a novel method to address these challenges by integrating genomic functional annotations , which have been successfully applied in GWAS to generate biological insights . We demonstrate the improvement in accuracy in both extensive simulation studies and real data analysis of breast cancer , Crohn’s disease , celiac disease , rheumatoid arthritis and type-II diabetes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods",
"Ethics",
"statement",
"Data",
"access"
] | [
"genome-wide",
"association",
"studies",
"rheumatology",
"medicine",
"and",
"health",
"sciences",
"breast",
"tumors",
"functional",
"genomics",
"immunology",
"cancers",
"and",
"neoplasms",
"oncology",
"diabetes",
"mellitus",
"rheumatoid",
"arthritis",
"clinical",
"medicin... | 2017 | Leveraging functional annotations in genetic risk prediction for human complex diseases |
An extraordinarily precise regulation of chlorophyll biosynthesis is essential for plant growth and development . However , our knowledge on the complex regulatory mechanisms of chlorophyll biosynthesis is very limited . Previous studies have demonstrated that miR171-targeted scarecrow-like proteins ( SCL6/22/27 ) negatively regulate chlorophyll biosynthesis via an unknown mechanism . Here we showed that SCLs inhibit the expression of the key gene encoding protochlorophyllide oxidoreductase ( POR ) in light-grown plants , but have no significant effect on protochlorophyllide biosynthesis in etiolated seedlings . Histochemical analysis of β-glucuronidase ( GUS ) activity in transgenic plants expressing pSCL27::rSCL27-GUS revealed that SCL27-GUS accumulates at high levels and suppresses chlorophyll biosynthesis at the leaf basal proliferation region during leaf development . Transient gene expression assays showed that the promoter activity of PORC is indeed regulated by SCL27 . Consistently , chromatin immunoprecipitation and quantitative PCR assays showed that SCL27 binds to the promoter region of PORC in vivo . An electrophoretic mobility shift assay revealed that SCL27 is directly interacted with G ( A/G ) ( A/T ) AA ( A/T ) GT cis-elements of the PORC promoter . Furthermore , genetic analysis showed that gibberellin ( GA ) -regulated chlorophyll biosynthesis is mediated , at least in part , by SCLs . We demonstrated that SCL27 interacts with DELLA proteins in vitro and in vivo by yeast-two-hybrid and coimmunoprecipitation analysis and found that their interaction reduces the binding activity of SCL27 to the PORC promoter . Additionally , we showed that SCL27 activates MIR171 gene expression , forming a feedback regulatory loop . Taken together , our data suggest that the miR171-SCL module is critical for mediating GA-DELLA signaling in the coordinate regulation of chlorophyll biosynthesis and leaf growth in light .
Chlorophylls are complexed with their binding proteins and serve two primary functions in photosynthesis: they trap light energy and transfer it to the reaction centers of photosystems [1] , [2] . During light absorption and energy transfer , chlorophylls inevitably generate highly reactive singlet oxygen , particularly under strong light , leading to the inhibition of photosynthesis , plant growth and even to cell death [3] , [4] . In addition , many chlorophyll precursors present in their free state are strong photosensitizers that produce reactive oxygen species upon light illumination . Therefore , the chlorophyll biosynthetic pathway is strictly regulated in response to developmental and environmental cues . It has been well documented that chlorophyll biosynthesis is finely regulated at the multiple steps in the pathway and at both transcriptional and post-transcriptional levels [5] . For example , protochlorophyllide ( Pchlide ) levels of etiolated seedlings are negatively regulated by the phytochrome-interacting factors PIF1 and PIF3-5 [6]–[9] , but positively regulated by two transposase-derived transcription factors , FAR1 ( far-red impaired response 1 ) and FHY3 ( far-red elongated hypocotyls 3 ) [10]; the activity of the key enzyme glutamyl-tRNA reductase ( HEMA1 ) is inhibited directly by heme and the FLU protein via a feedback mechanism [2] , [11]–[15] , while Mg-chelatase is stimulated by the binding of genomes uncoupled 4 ( GUN4 ) to the ChlH subunit ( GUN5 ) of Mg-chelatase , protoporphyrin IX ( PPIX ) and Mg-PPIX [16]–[19] . It is worth emphasizing that the activity of the key enzyme Pchlide oxidoreductase ( POR ) is primarily subject to the transcriptional regulation [8] , [20]–[21] . The Arabidopsis genome contains three differentially regulated POR genes . It has been shown that PORA is expressed in etiolated seedlings and its mRNA level drops sharply in light; PORB is expressed in both etiolated seedlings and light-grown plants; PORC expression is activated by light in a fluence rate-dependent manner [22]–[24] . Available evidence revealed that the expression of PORA and PORB is regulated by the transcription factors ethylene insensitive 3 ( EIN3 ) and its homolog EIN3-like1 ( EIL1 ) via directly binding to the EBS cis-elements in the promoter region [21] . Although PORC expression was reported to be directly induced by PIF1 [8] , it remains unclear how PORC is regulated in light where PIF proteins are degraded . Gibberellic acid ( GA ) is an important phytohormone that controls many aspects of plant development and growth via the GA-GID-DELLA signaling module in Arabidopsis [25]–[30] . With regard to the chlorophyll biosynthetic pathway , DELLA stabilization in the GA-deficient ga1-3 mutant leads to increased accumulation of Pchlide and PORs in etiolated seedlings , which are substantially more resistant to photo-oxidative damage after transferred from darkness to light [20] . DELLAs promote Pchlide biosynthesis by repressing the transcriptional activity of PIFs in the dark [20] , [31] , [32] . In contrast , the molecular mechanism underlying the DELLA-regulated POR expression is not fully understood . Recently , the miR171-targeted scarecrow-like ( SCL ) transcription factors SCL6/SCL6-IV , SCL22/SCL6-III and SCL27/SCL6-II ( also known as hairy meristems [HAM] and lost meristems [LOM] ) have been demonstrated to play an important role in the proliferation of meristematic cells , polar organization and chlorophyll synthesis [33]–[38] . However , it remains unknown how these SCL proteins control chlorophyll synthesis . Here , we provide convincing evidence that DELLA-regulated POR expression is , at least in part , mediated by miR171-targeted SCLs in light .
As previously reported [38] , both MIR171c over-expressors ( MIR171c-OX ) and scl6 scl22 scl27 triple mutants produce dark green leaves ( Figure 1A and Figure S1A ) , which contain approximately 40% more chlorophyll than wild type ( WT ) leaves ( Figure 1B ) . In contrast , the over-expression of miR171-resistant LUC-rSCL27 ( fused to the luciferase gene ) results in leaf yellowing ( Figure 1A and Figure S1A ) and a significant decrease in chlorophyll content ( Figure 1B ) . These results indicate that miR171-targeted SCLs are negative regulators of chlorophyll biosynthesis . To explore the physiological role of SCL proteins in the regulation of chlorophyll biosynthesis , we constructed transgenic plants expressing rSCL27 fused to the β-glucuronidase ( GUS ) gene driven by the SCL27 native promoter , designated pSCL27::rSCL27-GUS . We examined the pattern of GUS expression in the 3- to 11-day-old seedlings . The results of GUS staining clearly showed that the SCL27-GUS fusion protein started to accumulate in the newly developed leaves ( Figure S2 ) . In the first pair of leaves , the GUS signal was first observed in the 3-day-old seedlings through the whole leaves , and maintained at a relative stable level at the basal region until to the 7-day-old seedlings , and suddenly disappeared in the 8-day-old seedlings ( Figure 2A–2B and Figure S2 ) . Consistent with this observation , the scl6 scl22 scl27 mutant exhibited more intense chlorophyll fluorescence at the base of leaves than did the WT , whereas chlorophyll fluorescence intensity at the leaf apical region was identical to that in the WT ( Figure 2C and 2D ) , suggesting that SCL proteins play an important role in inhibiting chloroplast development before cell expansion . This result is consistent with a previous report that leaf greening and cell expansion initiate at the leaf tip and proceed in a basipetal direction [39] . We further evaluated the role of SCLs in plant adaptation to high light stress by measuring the ratio of variable fluorescence to maximum fluorescence ( Fv/Fm ) , which reflects the maximal photochemical efficiency of photosystem II ( PSII ) photochemistry ( PSII activity ) . Compared to WT plants , PSII activity decreased more slowly in MIR171c-OX and scl6 scl22 scl27 plants but decreased more rapidly in LUC-rSCL27-OX plants ( Figure 1C ) in light stress , indicating that miR171-targeted SCLs are also involved in plant adaptation to excess light . We also investigated the role of SCLs in the growth of etiolated seedlings and chloroplast development . As shown in Figure S3A–S3C , manipulation of SCL gene expression slightly but not significantly affected greening ratio , Pchlide content and etioplast ultrastructure of the 5-day-old dark-grown seedlings . However , stacked and stromal thylakoid membranes were thicker in chloroplasts from MIR171c-OX and scl6 scl22 scl27 mature leaves while was thinner in those from LUC-rSCL27-OX leaves , compared to WT ( Figure S4A–S4B ) . Consistently , immunoblotting analysis showed that the levels of light-harvesting complex subunits including LHCB1 , LHCB2 , LHCB5 , and LHCA1 were higher in MIR171c-OX and scl6 scl22 scl27 than in WT but lower in LUC-rSCL27 ( Figure S4C ) . However , changes in SCL expression had no effect on the accumulation of PsaD ( PSI subunit ) and AtpB ( ATP synthase beta subunit ) in mature leaves ( Figure S4C ) . Taken together , these results indicate that SCLs are involved in chlorophyll biosynthesis mainly in light but not in the dark . To elucidate the molecular mechanism underlying SCL-regulated chlorophyll synthesis , we analyzed the transcriptional levels of several key genes in the pathway , including the genes encoding HEMA1 , GUN4 , GUN5 , PORs and chlorophyll a oxygenase ( CAO ) . Quantitative PCR ( qPCR ) and northern blotting assays showed that among the inspected genes the levels of PORs and CAO transcripts were higher in MIR171c-OX and scl6 scl22 scl27 while were lower in LUC-rSCL27-OX , compared to those in the WT ( Figure 1D and Figure S1B ) . The expression levels of PORs and CAO were correlated well with chlorophyll content in the leaves of MIR171c-OX , scl triple mutant and LUC-rSCL27-OX plants , suggesting that the expression of PORs and CAO is regulated by SCLs . Immunoblotting analysis using a POR antibody that can recognize all three isoforms of POR showed that MIR171c-OX and scl6 scl22 scl27 plants accumulated higher levels of PORC and PORB proteins than did WT and LUC-rSCL27-OX plants ( Figure S1C ) . Thus , the data obtained indicate that the expression of POR , the key enzyme in the chlorophyll biosynthetic pathway , is negatively regulated by SCLs . To verify the role of POR in SCL-regulated chlorophyll synthesis , we down-regulated the expression of POR in WT and scl6 scl22 scl27 mutant plants using an artificial microRNA that was designed to specifically target the three POR genes . Transgenic plants ( por-amiR ) with substantially reduced levels of POR expression were identified using qPCR ( Figure S1D and S1E ) . Knocking down POR expression in WT , MIR171c-OX and scl triple mutant plants led to a pale-green phenotype and a lower level of chlorophyll and PSII activity than in the corresponding controls ( Figure 1A–1C ) . Taken together , these data indicate that miR171-targeted SCLs regulate chlorophyll biosynthesis via the key enzyme POR . The important role of PORs in SCL-regulated chlorophyll biosynthesis prompted us to investigate whether SCLs can directly control the promoter activity of POR genes . Because both PORC and MIR171 are regulated by light but not by the circadian clock [22]–[24] , [40] , we hypothesized that PORC was a direct target of SCLs . To test this hypothesis , we co-expressed the LUC reporter gene under the control of the PORC promoter ( a 1685-bp genomic fragment upstream of the start codon ) together with 6xMYC-rSCL27 in Nicotiana benthamiana leaves using a transient expression system . The expression of LUC was much lower in the leaves transformed with 6xMYC-rSCL27 than in leaves transformed with the empty vector and rSCL27-VP16 ( a transcriptional activator ) ( Figure 3A ) , suggesting that the PORC promoter is a direct target of SCL27 . To identify the PORC promoter region bound by SCL27 , three fragments extending from the PORC start codon to −1685 , −861 and −455 bp upstream were fused to the LUC reporter gene . LUC expression under the control of either pPORC-1685 or pPORC-861 was significantly reduced by 6xMYC-rSCL27 but not by rSCL27-VP16 , whereas LUC expression driven by pPORC-455 was low and unaffected by 6xMYC-rSCL27 or rSCL27-VP16 ( Figure 3A ) . Consistently , LUC expression under the control of pPORC-1685 or pPORC-861 was higher in MIR171c-OX and scl6 scl22 scl27 than in WT ( Col ) , whereas LUC expression driven by pPORC-455 did not significantly differ between WT and MIR171c-OX or between WT and scl6 scl22 scl27 ( Figure S5A ) . These data suggest that the promoter region between −861 bp and −455 bp is required for SCL27 binding to the PORC promoter . We then performed chromatin immuno-precipitation ( ChIP ) and qPCR assays to further define the SCL27-binding region within the PORC promoter ( Figure 3B ) . Our results showed that fragments II ( −778 bp to −598 bp ) and III ( −572 bp to −372 bp ) were enriched in immuno-precipitates from the transgenic plants over-expressing 6xMYC-rSCL27 but not in those from WT plants ( Figure 3C ) , whereas fragments I ( −1524 bp to −1324 bp ) and IV ( 1144 bp to 1246 bp of the coding sequence , used as a negative control ) were not enriched ( Figure 3C ) , indicating that fragments II and III contain SCL27-binding cis-elements . We next performed electrophoretic mobility shift assays ( EMSAs ) to confirm whether SCL27 can directly bind to fragments II and III of the PORC promoter . Consistent with the ChIP-qPCR results , shifted bands were observed when purified recombinant SCL27 protein ( Figure S5B ) was incubated with DNA fragments II or III , and the intensity of the bands gradually increased with increasing concentrations of SCL27 ( Figure 3D ) . However , no shifted band was detected when SCL27 was incubated with fragment I ( Figure 3D ) . Taken together , our in vivo and in vitro data suggest that SCL27 inhibits PORC expression via directly binding to the PORC promoter . GT elements have been reported to be important for light-regulated gene expression , and DNA fragments II and III contain these cis-elements [41] . To test whether GT-elements are important for SCL27 binding to the PORC promoter , we chose the 62-bp DNA fragment from −500 bp to −438 bp , which contains three G ( A/G ) ( A/T ) AA ( A/T ) GT element repeats [41] ( Figure 3E ) . The EMSA results showed that purified recombinant SCL27 bound to the W fragment but not to the M fragment ( Figure 3F ) . The formation of the SCL27-DNA complex was suppressed by a 100- , 200- or 400-fold excess of unlabeled W fragment , but not by the unlabeled M fragment ( Figure 3G ) . Thus , we conclude that GT elements are required for SCL27 to bind to the PORC promoter . DELLA proteins up-regulate the expression of PORs either in a PIF-dependent or PIF-independent manner [20] . We therefore tested whether miR171-targeted SCLs mediated DELLA-regulated POR expression . For this purpose , we generated LUC-rSCL27-OX/ga1-3 or pSCL27::rSCL27/pRGA::RGAd17 ( the GA-insensitive form of RGA ) plants via sexual crossing . Over-expressing LUC-rSCL27 in the WT or ga1-3 genetic background led to pale-green phenotypes and significantly decreased chlorophyll content ( Figure 4A , 4B and Figure S6A ) . Likewise , expressing pSCR27::rSCL27 in the pRGA::RGAd17 plants also resulted in a pale-green phenotype and decreased chlorophyll content ( Figure S6B and S6C ) . qPCR analysis showed that over-expressing LUC-rSCL27 in WT and ga1-3 plants led to a dramatic decrease in PORC expression ( Figure 4C ) . To confirm the epistasis of SCLs to DELLAs in the regulation of chlorophyll biosynthesis , we over-expressed MIR171c in WT ( Ler ) and della pentuple mutants . Indeed , over-expressing MIR171c in these plants resulted in dark green leaves and increased chlorophyll content ( Figure 4D and 4E ) . Consistently , the level of PORC expression was higher in MIR171c over-expressors than in the corresponding control WT ( Ler ) and della pentuple plants ( Figure 4F ) . These data indicate that DELLA-promoted chlorophyll biosynthesis and PORC expression are dependent on SCLs . To examine the role of the DELLA-SCL module in chlorophyll biosynthesis in the dark , we measured the greening ratio and Pchlide content of 5-day-old etiolated WT , MIR171c-OX , scl6 scl22 scl27 and LUC-rSCL27-OX seedlings in the presence of paclobutrazol ( PAC ) , which increases the levels of DELLA proteins . Our results showed that changes in SCL expression did not affect the greening ratio and Pchlide content in the absence or presence of PAC ( Figure S7A and S7B ) , indicating that SCLs are not involved in DELLA-promoted Pchlide biosynthesis in the dark . We then tested whether DELLA proteins directly interact with SCLs in vitro and in vivo . Yeast two-hybrid assays showed that the DELLA protein RGA interacted with SCL27 via the N-terminal domain ( Figure 4G ) . In addition , SCL22 bound to all DELLA proteins in yeast ( Figure S8A ) , indicating that the interaction between SCLs and DELLAs is universal . The in vivo interaction between SCLs and DELLAs was examined by bimolecular fluorescence complementation ( BiFC ) and co-immunoprecipitation ( Co-IP ) assays using a transient expression system . A strong YFP signal was observed when either rSCL27-nYFP and RGA-cYFP or rSCL27-Nter-nYFP and RGA-cYFP were co-expressed in leaves ( Figure 4H and Figure S8B ) . Co-IP results also showed that RGAd17-3xHA bound to 6xMYC-rSCL27 but not to the control MYC-YFP ( Figure 4I ) . Furthermore , MYC-SCL27 was precipitated by the antibody against RGA in total proteins extracted from transgenic plants over-expressing 6XMYC-rSCL27 treated with PAC but not treated with GA3 ( Figure 4J ) . These results demonstrate that RGA interacts directly with SCL27 both in vitro and in vivo . In addition , qPCR assays showed that the accumulation of SCL transcripts was not altered in plants treated with GA3 ( Figure S9A ) or PAC ( Figure S9B ) , or in GA mutants , including ga1-3 , gai-2 or rga gal1 rgl2 rgl3 plants ( Figure S9C ) . Likewise , RGA and GAI expression was not apparently affected by SCL levels ( Figure S9D ) . Thus , these results exclude the possibility that DELLAs and SCLs are mutually regulated at the transcriptional level . DELLAs have been shown to regulate various biological processes by preventing transcription factors from binding to DNA [31]–[32] , [42]–[46] . The antagonistic role of DELLAs and SCLs in the regulation of chlorophyll biosynthesis raises the possibility that DELLAs might inhibit SCL binding to DNA . To test this hypothesis , we analyzed the promoter activity of PORC using a dual-luciferase reporter assay by transforming RGAd17-3xHA and/or 6xMYC-rSCL27 into N . benthamiana leaves . The results showed that pPORC-1685::LUC reporter activity was significantly suppressed by 6xMYC-rSCL27 but was not affected by RGAd17-3xHA . The degree of inhibition of pPORC-1685::LUC reporter activity by SCL27 was partially mitigated by the co-expression of RGAd17-3xHA ( Figure 4K ) . Consistent with these results , EMSA analysis showed that the interaction between RGA and SCL27 decreased the binding of SCL27 to DNA ( Figure 4L and Figure S10A , S10B ) . Furthermore , ChIP-qPCR analysis also showed that the enrichment of fragments II and III containing GT elements in the PORC promoter ( shown in Figure 2B ) was higher in MYC antibody pulled-down precipitates from GA-treated 6xMYC-rSCL27-OX plants but lower in those from PAC-treated 6xMYC-rSCL27-OX plants than in those from plants given the mock treatment ( Figure 4M and Figure S10C ) . Thus , our data demonstrate that the RGA-SCL27 interaction decreases SCL27 DNA-binding activity . In general , the level of miRNA expression is inversely correlated with the level of target gene expression . However , miR171 accumulation was reported to peak 6 hours earlier than that of SCL6 [40] . Recent studies have shown that the expression of miRNAs can be controlled by their target genes in a feedback manner [47] . Consistent with this idea , GT elements have been found in the promoters of MIR171s . qPCR assays showed that the expression levels of all MIR171 genes are much higher in LUC-rSCL27-OX plants than in WT plants , whereas the expression levels of these genes are lower in scl6 scl22 scl27 plants than in WT plants ( Figure 5A ) , indicating that SCLs are positive regulators of MIR171 expression . However , the extent to which SCL27 regulated MIR171 expression differed among the MIR171 genes ( Figure 5A ) . For example , SCL27 had the greatest effect on the level of MIR171a expression but had lower , similar effects on the expression levels of MIR171b and MIR171c ( Figure 5A ) . Additionally , we generated transgenic plants expressing 6xMYC-rSCL27 fused to the rat glucocorticoid receptor ( GR ) under the control of the 35S regulatory sequence in the scl triple mutant background; these plants were designated 35S::6xMYC-rSCL27-GR/scl6 scl22 scl27 . Compared to mock ( dimethyl sulfoxide , DMSO ) -treated plants , transgenic plants treated with 10 µM dexamethasone ( DEX ) were pale green and accumulated less chlorophyll ( Figure 5B and 5C ) . qPCR analysis showed that the level of PORC mRNA was rapidly decreased in the transgenic plants treated with DEX for 4 hours ( Figure 5D ) . Using this inducible expression system , we found that MIR171a transcripts accumulated to levels more than 3-fold higher in DEX-treated plants than in the control , whereas two other MIR171 genes were only slightly up-regulated by SCL27 ( Figure 5D ) . To confirm the observation that SCL27 activates MIR171 gene expression , the LUC reporter gene driven by the MIR171a promoter ( pMIR171a::LUC ) was co-transformed with or without 6xMYC-rSCL27 into N . benthamiana leaves . As shown in Figure 5E , pMIR171a::LUC activity was significantly increased by 6xMYC-rSCL27 . These results indicate that SCLs can up-regulate MIR171 gene expression . To confirm that SCL27 directly regulates MIR171 gene expression , ChIP-qPCR was performed using three fragments: V ( −726 bp to −495 bp , without GT elements ) , VI ( −260 bp to −71 bp , containing GT elements ) and VII ( the precursor of MIR171a ) . Indeed , only fragment VI was enriched in MYC antibody pulled-down precipitates obtained from the 6xMYC-rSCL27 over-expressing plants but not in those obtained from the WT plants ( Figure 5F and 5G ) . Taken together , these data clearly indicate that miR171 and its target SCLs form a feedback loop to finely regulate chlorophyll biosynthesis .
miR171 and its target SCL proteins have been reported to play an important role in plant development and growth [33]–[38] . However , little is known about the molecular mechanisms by which the miR171-SCL module functions . In this study , we found compelling evidence showing that SCLs are GT element-binding transcriptional factors that can suppress or promote gene expression in Arabidopsis . Given that GT elements are widely distributed in tandem repeats within the promoter regions of many photosynthetic and plastid ribosomal genes [41] , it is reasonable to assume that the miR171-SCL module can regulate the expression of other genes in a manner similar to that used for the PORC gene . In higher plants , light and GA are important signals that antagonistically regulate chloroplast biogenesis , which is a complicated process including chloroplast division and the formation of the photosynthetically active chloroplast [31] , [32] . It is well established that PIFs , which are negative regulators of chlorophyll biosynthesis , are critical downstream effectors in light and GA signal transduction pathways [20] , [31] , [32] . PIFs bind directly to the conserved DNA G-box motif of gene promoters and regulate the chlorophyll biosynthetic pathway by inhibiting Pchlide accumulation and inducing POR gene expression in an additive , redundant or specific manner [6]–[9] , [20] . This regulatory mechanism involving PIFs is apparently important for the prevention of free Pchlide accumulation and the subsequent greening of etiolated seedlings upon light exposure [20] , [31] , [32] . Based on the results derived from this study , we suggest that SCLs play an important role in regulating chlorophyll biosynthesis under light conditions ( Figure 6 ) , in which PIFs are rapidly degraded . In addition , PIF proteins can be sequestered by DELLAs , the levels of which are elevated in light and decreased in the dark , blocking the ability of PIFs to bind to their target gene promoters [20] . Thus , the SCLs and PIFs control chlorophyll biosynthesis in different yet cooperative manners , and PIFs are replaced by miR171-targeted SCLs to inhibit chlorophyll biosynthesis in light . Since both the levels of DELLA proteins and miR171 expression are elevated in light , the inhibition of SCLs on chlorophyll biosynthesis is coordinately relieved at transcriptional and post-translational levels , while the positive feedback regulation pathway in which SCLs activate miR171 expression might be important for auto-regulating the homeostasis of SCL proteins in light . In addition to environmental cues , chloroplast development is regulated by developmental signals . Early leaf growth is divided into two sequential cellular processes after primordium initiation: cell proliferation and cell expansion [39] . Usually , chloroplast development is suppressed in the cell proliferation region at the leaf base , then remains relatively stable over a certain period , and finally is abolished abruptly . Once a cell has stopped proliferating , it enters the stage of cell expansion , which is triggered by chloroplast differentiation [39] . Thus , chloroplast differentiation plays an important role in the timing of the transition from cell proliferation to cell expansion . However , blocking chloroplast differentiation and retrograde signaling from chloroplasts to the nuclei using norflurazon cannot completely stop cell expansion , suggesting that other mechanisms are also involved in the phase shift [39] . Our data showed that a negative regulator of chloroplast development , SCL27 , is highly expressed at the base of growing leaves , and chloroplast development proceeds more rapidly in a scl triple mutant than in WT . These results suggest that miR171-targeted SCLs play an important role in suppressing chloroplast development in dividing cells during early leaf growth . Interestingly , leaf size is apparently altered in SCL27 over-expressors and the scl triple mutant compared to that in WT ( Figure 1A ) . One explanation is that SCLs function as coordinators that simultaneously regulate chloroplast development and cell proliferation; another possibility is that the onset of SCL-regulated chloroplast development leads to a change in the timing of cell proliferation exit . Further investigation is required to elucidate the molecular mechanism by which SCLs coordinately regulate leaf size and chloroplast development . Chloroplast biogenesis is also coordinated with cell expansion during leaf growth to achieve optimal photosynthesis rates . For example , leaf greening accompanies cell expansion , which initiates at the leaf tip and proceeds in a basipetal direction in Arabidopsis [39] . It has been demonstrated that GA plays a critical role in controlling cell expansion and chloroplast biogenesis through DELLA proteins in both dicot and monocot plant species [48] . The number of thylakoid membranes per granum and the chloroplast density per cell are increased in the ga1-3 mutant , indicating that more chlorophyll is synthesized in the mutant chloroplasts . It is likely that DELLA proteins , which are stabilized in the ga1-3 mutant , promote chlorophyll biosynthesis by suppressing the inhibitory transcriptional activity of SCLs . Thus , it appears that the DELLA-SCL module functions to balance chloroplast development and cell expansion , which is accompanied by a dramatic increase in photosynthesis . Furthermore , we observed another complex phenomenon: over-expression of SCLs did not completely rescue the dark-green phenotype of the ga1-3 mutant , indicating that DELLAs transmit signals that affect chlorophyll biosynthesis by regulating other interacting proteins . A number of DELLA-interacting transcriptional factors have been identified thus far [31]–[32] , [42]–[46] , including EIN3 and EIL1 , which are downstream effectors of ethylene signaling . DELLAs de-repress EIN3 and EIL1 function during apical hook formation in etiolated seedlings [43] . Interestingly , EIN3 and EIL1 were also shown to regulate chlorophyll biosynthesis by repressing the accumulation of Pchlide and by activating the expression of POR genes ( PORA and PORB ) [21] . It is likely that DELLAs regulate PORA and PORB expression directly via EIN3 and EIL1 . DELLAs might also indirectly regulate chlorophyll biosynthesis through other interacting transcriptional factors , including brassinosteroid-resistant 1 ( BZR1 ) and the jasmonic acid ZIM-domain proteins ( JAZs ) [42] , [49] , [50] . Taken together , the findings described here indicate that DELLAs are critical factors integrating various signaling pathways to dynamically regulate chlorophyll biosynthesis .
MIR171c-OX , scl6 scl22 scl27 triple mutants , 35S::LUC-rSCL27 , ga1-3 , gai-2 ( SAIL_587_CO2 ) , rga rgl1 rgl2 rgl3 mutants , and pRGA::RGAd17 are in the Arabidopsis thaliana Columbia ecotype ( Col ) background [27] , [38] , [44]; the ga1-3 mutant was backcrossed with the wild type ( Col ) for six generations; the por-amiR and pSCL27::SCL27-GUS were transformed in Col background; the por-amiR/MIR171c-OX , LUC-rSCL27-OX/ga1-3 , and rSCL27/RGAd17 plants were generated by crossing; the della pentuple is in the Ler ecotype [44] . Seeds were germinated and grown on the half Murashige and Skoog ( MS ) media containing 1% sucrose and 0 . 7% phytoagar . All plants were grown at 21°C under light ( 110 µmol . m−2 . s−1 ) in long days ( 16-h light/8-h dark ) . About 1 . 7- and 1 . 2-kb promoter fragments at the upstream of the start codon were amplified from PORC and MIR171a genes in the Col genome , respectively , with primers listed in Table S1 . The amplified fragments were inserted in the XhoΙ/BamHI sites of the pGREEN0800LUC vector [51] , [52] to produce pPORC-1685::LUC and pMIR171a::LUC vectors . The pPORC-861::LUC and pPORC-455::LUC vectors were constructed in a similar way . To make the POR amiRNA vector , the amiRNA target sequences for POR genes and primers including POR I miR-s , POR II miR-a , POR III miR*s and POR IV miR*a were designed using the WMD3 Web microRNA Designer ( http://wmd3 . weigelworld . org/cgi-bin/webapp . cgi ) and listed in Table S1 . The amiRNA precursor was amplified by overlapping PCR from the pRS300 template to produce the fragment containing the POR target amiRNA foldback . DNA fragments were gel-purified and cloned into the Gateway cloning vector pENTR-SD/D/TOPO ( Invitrogen ) according to the manufacturer's instructions . After sequencing confirmation , the cloned DNA fragments were transferred to the 35S over-expression vector ( pGWB2 ) ( Invitrogen ) using LR clonase ( Invitrogen ) . For yeast two-hybrid analysis , SCL22 cDNA was cloned into the pGBKT7 vector ( Clontech ) . RGA , GAI , RGL1 , RGL2 and RGL3 cDNAs were cloned into the pGADT7 vector ( Clontech ) . SCL27 and RGA cDNAs were cloned into pDEST22 ( Invitrogen ) ; cDNAs encoding SCL27 and its N-terminal ( 1–267 amino acids ) and GRAS domain ( 268–640 amino acids ) were cloned into pDEST32 ( Invitrogen ) . The primers used for these constructs are given in Table S1 . For BiFC analysis , SCL27 , SCL27 N-terminal , and SCL27-GRAS sequences were cloned into pCAMBIA1300 ( nYFP ) , whereas RGA was cloned into pCAMBIA1300 ( cYFP ) . For in vitro protein-DNA binding analysis , SCL27 and RGA was cloned into the pET28b and pGEX6p-3 vectors , respectively . The constructs were transformed into the expression strain BL21 for protein expression . For co-IP analysis , RGAd17 and miR171-resistent SCL27 ( rSCL27 ) were cloned into the binary vector with 3xHA or 6xMYC . Transgenic plants were generated by the floral dipping method [53] and were screened with 50 mg/mL of kanamycin sulfate or 50 mg/mL of hygromycin . Seedling greening was analyzed by exposing 5-day-old dark-grown seedlings to white light ( 16 h-light/8 h-dark ) for 2 days . Chlorophyll autofluorescence was analyzed using a confocal laser scanning microscope ( Olympus , FV10-ASW ) . In the PAC-treated etiolated seedlings , 0 . 01 µM of PAC was used . Greening ratio was determined by counting the percentage of green cotyledons of each genotype . Pchlide was extracted from 5-day-old etiolated seedlings with 1 mL of ice-cold 80% acetone in the dark . The samples were centrifuged at 13000 rpm for 10 min , and fluorescence was excited by the wavelength of 440 nm and scanned from 600 nm to 700 nm using a fluorescence spectrophotometer ( Hitachi ) at room temperature [54] . The results were presented by relative fluorescence per seedling . Chlorophyll was measured as described previously [55] . The Fv/Fm parameter was measured using light-stressed leaf discs after 15-min adaptation to darkness [56] . For electron microscopy observation , cotyledons of 5-day-old etiolated and 25-day-old seedlings were fixed and processed as previously described [57] , and examined with an H-7650 transmission electron microscope ( Hitachi ) . Plasmids were transformed into yeast strain AH109 by the lithium chloride–polyethyleneglycol method according to the manufacturer's manual ( Clontech ) . The transformants were selected on SD-Leu-Trp plates . The protein-protein interactions were tested on SD-Trp-Leu -His-Ade plates with or without 3-amino-1 , 2 , 4-triazole . The A . tumefaciens strain GV3101 transformed with each of the two constructs for BiFC analysis was cultured in the solution containing 10 mM MES , 10 mM MgCl2 , and 100 µM acetosyringone to an optical density ( OD600 ) of 0 . 6 to 0 . 8 . Then , two strains were mixed and incubated at the room temperature for at least 2 h . The YFP fluorescence was analyzed using a confocal laser scanning microscope ( Olympus , FV10-ASW ) 48 to 96 h after N . benthamiana leaves were infiltrated with the mixture . Plant materials were submerged in 90% acetone for 15 min , and then transferred into 0 . 5 mg/mL X-Gluc solution ( 0 . 1 M sodium phosphate buffer , pH 7 . 0 , 10 mM EDTA , 0 . 1% Triton X-100 , 0 . 5 mM potassium ferrocyanide , 0 . 5 mM potassium ferricyanide ) . Plant materials were vacuumized , kept at 37°C and decolorized in 70% ethanol . Agrobacteria-infiltrated N . benthamiana leaves and transgenic plants over-expressing 6xMYC-rSCL27 were used for Co-IP analyses . The soluble proteins were extracted with the extraction buffer ( 50 mM Heps [pH 7 . 5] , 150 mM NaCl , 10 mM EDTA [pH 8 . 0] , 0 . 2% Nonidet P-40 , 10% glycerol , 1% PVPP , 2 mM DTT , 1× Complete Protease Inhibitor Cocktail [Sigma] ) . The beads were washed with the buffer ( 50 mM Heps [pH 7 . 5] , 200 mM NaCl , 10 mM EDTA [pH 8 . 0] , 0 . 1% Nonidet P-40 , 10% glycerol ) . Immunoprecipitation was performed with the anti-MYC antibody using N . benthamiana leaves . For Arabidopsis samples , immunoprecipitation was performed with the anti-RGA antibody . RGAd17-3xHA and 6xMYC-SCL27 fusion proteins were detected by immunoblotting with anti-HA ( Sigma ) and anti-MYC antibodies ( Santa Cruz ) . To analyze POR , LHCB1 , LHCB2 , LHCB5 , LHCA1 , PsaD , and AtpB protein levels in vivo , samples ( 0 . 1 g ) were ground in liquid nitrogen and suspended with 200 µL extraction buffer ( 125 mM Tris [pH 8 . 8] , 4% SDS , 20% glycerol , 5% β-Me ) . Total protein was extracted by incubating the samples in boiled water for 5 min , and then centrifuged at 13 000 rpm for 10 min . Proteins were detected with the anti-POR , LHCB1 , LHCB2 , LHCB5 , LHCA1 , PsaD , and AtpB antibodies ( Agrisera ) after total proteins were separated onto a SDS-PAGE gel and transferred to Hybond-ECL Nitrocellulose membrane ( Amersham Biosciences ) . The transient expression assay ( Dual-LUC ) was carried out as described previously [52] . Agrobacteria-infiltrated N . benthamiana leaves were used for LUC/REN analyses . Leaf samples were collected for the transient expression assay using commercial Dual-LUC reaction ( DLR ) reagents , according to the manufacturer's instruction ( Promega ) . One µg of total RNAs was used for reverse transcription in a 20 µL reaction system using the RNA PCR ( AMV ) kit ( Promega ) . Quantitative PCR was performed with SYBR-Green PCR Mastermix ( Takara ) , and amplification was real-time monitored on stepone and steponeplus real-time PCR system ( Applied Biosystems ) . ACTIN2 was used as an internal control for normalization . The primers are listed in Table S1 . Northern blot analysis was carried out as described [58] . ChIP experiments were performed according to published protocols [59] . Briefly , about 3 g tissues of 3-week-old 6xMYC-rSCL27-OX transgenic plants were harvested . For GA3 or PAC treatment , samples were harvested from the plants treated with 10 µM GA3 and 0 . 1 µM PAC for 2 day . After fixation , the materials were resuspended in extraction buffer followed by sonification . One third of the solution was saved as input total DNA without precipitation; another one-third was mixed with the MYC-fused agarose ( Sigma ) ; and the remaining one-third was precipitated in parallel with HA-fused agarose as a negative control . The resulting DNA samples were purified using a PCR purification kit ( Qiagen ) . The relative concentrations of the DNA fragments were analyzed by qPCR , using the β-TUBULIN2 gene promoter as the reference . The EMSA was performed as reported previously [60] . The primers used were shown in Table S1 . The Cy5 fluorescence-labeled DNA ( 1 nM ) was incubated with the indicated amount of the purified His-SCL27 protein in 20 µL of the binding buffer . The concentration of the proteins used for the competitive assay in Figure 4L was 1000 nM . After incubation at 30°C for 20 min , the reaction mixture was electrophoresed at 4°C on a 6% native polyacrylamide gel in 0 . 5×Tris-borate-EDTA for 2 h ( about 200-bp ) or 1 h ( 62-bp ) at 100 V . Fluorescence-labeled DNA on the gel was then detected with the Starion FLA-9000 ( FujiFilm , Japan ) . SCL27 ( At2G45160 ) , SCL22 ( At3G60630 ) , SCL6 ( At4G00150 ) , MIR171A ( At3G51375 ) , MIR171B ( At1G11735 ) , MIR171C ( At1G62035 ) , β-TUBULIN-2 ( At5G62690 ) , HEMA1 ( AT1G58290 ) , GUN4 ( AT3G59400 ) , GUN5 ( AT5G13630 ) , PORA ( AT5G54190 ) , PORB ( AT4G27440 ) , PORC ( AT1G03630 ) , GAI ( AT1G14920 ) , RGA ( AT2G01570 ) , RGL1 ( AT1G66350 ) , RGL2 ( AT3G03450 ) , and RGL3 ( AT5G17490 ) . | Chlorophyll biosynthesis is essential for plant growth and development . To date , the regulatory mechanisms of chlorophyll biosynthesis have been well understood only in dark conditions . Previous reports showed that miR171-targeted SCL6/22/27 proteins were involved in chlorophyll biosynthesis . However , the molecular mechanism of SCL action remains unclear . In this study , we found that SCLs negatively regulated chlorophyll biosynthesis though suppressing the expression of the key gene PROTOCHLOROPHYLLIDE OXIDOREDUCTASE ( POR ) . SCL27 is highly expressed at the basal cell proliferation region of young leaves , suggesting an important role of SCLs in inhibiting chloroplast development before cell expansion . In addition , GT-cis elements were required for SCL27 directly binding to the PORC promoter . Furthermore , we showed that SCLs mediated GA-regulated chlorophyll biosynthesis through direct interaction with DELLA proteins . The interaction between SCLs and DELLAs reduced the DNA binding activity of SCL27 . Our uncovered GA-DELLA-SCL module and its DNA binding targets provide new insights into molecular mechanisms by which chlorophyll biosynthesis and cell proliferation are coordinately regulated during leaf development in response to developmental and environmental cues . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"anatomy",
"plant",
"science",
"plant",
"physiology",
"leaves",
"cell",
"biology",
"plant",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"chloroplasts",
"cellular",
"structures",
"and",
"organelles"
] | 2014 | Arabidopsis miR171-Targeted Scarecrow-Like Proteins Bind to GT cis-Elements and Mediate Gibberellin-Regulated Chlorophyll Biosynthesis under Light Conditions |
Ciliopathies are a group of human disorders caused by dysfunction of primary cilia , ubiquitous microtubule-based organelles involved in transduction of extra-cellular signals to the cell . This function requires the concentration of receptors and channels in the ciliary membrane , which is achieved by complex trafficking mechanisms , in part controlled by the small GTPase RAB8 , and by sorting at the transition zone located at the entrance of the ciliary compartment . Mutations in the transition zone gene CC2D2A cause the related Joubert and Meckel syndromes , two typical ciliopathies characterized by central nervous system malformations , and result in loss of ciliary localization of multiple proteins in various models . The precise mechanisms by which CC2D2A and other transition zone proteins control protein entrance into the cilium and how they are linked to vesicular trafficking of incoming cargo remain largely unknown . In this work , we identify the centrosomal protein NINL as a physical interaction partner of CC2D2A . NINL partially co-localizes with CC2D2A at the base of cilia and ninl knockdown in zebrafish leads to photoreceptor outer segment loss , mislocalization of opsins and vesicle accumulation , similar to cc2d2a-/- phenotypes . Moreover , partial ninl knockdown in cc2d2a-/- embryos enhances the retinal phenotype of the mutants , indicating a genetic interaction in vivo , for which an illustration is found in patients from a Joubert Syndrome cohort . Similar to zebrafish cc2d2a mutants , ninl morphants display altered Rab8a localization . Further exploration of the NINL-associated interactome identifies MICAL3 , a protein known to interact with Rab8 and to play an important role in vesicle docking and fusion . Together , these data support a model where CC2D2A associates with NINL to provide a docking point for cilia-directed cargo vesicles , suggesting a mechanism by which transition zone proteins can control the protein content of the ciliary compartment .
Primary cilia are microtubule-based organelles protruding from the apical surface of most differentiated vertebrate cell types where they play a crucial role in transduction of extra-cellular signals to the cell [1] . Cilia achieve this function by concentrating and regulating receptors and channels that are required for sensing these signals in their membrane domain . Consequently , the ciliary membrane has a distinct composition from that of the adjacent plasma membrane , despite them being continuous with each other [2] . The tight regulation required to maintain the specificity of the ciliary membrane composition is achieved by complex trafficking and sorting mechanisms at the entry point to the ciliary compartment , as well as by a diffusion barrier present at the base of the cilium [3 , 4] . The transition zone , at the base of the ciliary axoneme , plays a crucial role in this sorting mechanism [5 , 6] . Indeed , dysfunction of proteins normally localized at the transition zone leads to both abnormal access to the ciliary compartment for proteins that should not localize there and loss of normal localization for ciliary proteins [5 , 7] . The actual mechanism , by which these transition zone proteins contribute to this sorting of ciliary proteins , remains however largely unknown . Mutations in transition zone proteins in humans lead to several ciliopathies such as Joubert syndrome . Ciliopathies are a group of human disorders caused by dysfunction of primary cilia and characterized by overlapping genetics and phenotypes [8] . As cilia are present on most vertebrate cells , their dysfunction can manifest as a wide array of phenotypic features affecting most organs systems [9] . Retinal dystrophy is a common finding in ciliopathies given that retinal photoreceptor outer segments , which are the site of the phototransduction cascade , are highly specialized primary cilia [10] . Joubert syndrome ( JBTS ) ( OMIM 213300 ) is a prototypical ciliopathy with a phenotypic spectrum that can encompass most of the typical ciliopathy phenotypes [11 , 12] . It is characterized by a specific hindbrain malformation termed the molar tooth sign ( MTS ) , in addition to which affected individuals may have retinal dystrophy , tubulo-interstitial kidney disease , liver fibrosis , skeletal dysplasia and polydactyly [13–15] . To date , mutations in over 27 different genes have been reported as an underlying cause for JBTS [12 , 16–20] . Most of these genes encode proteins associated in multi-protein complexes localized at the transition zone of the primary cilium [7 , 21] . Mutations in CC2D2A ( Coiled-coil and C2-domains containing protein 2A ) are the second most common genetic cause for JBTS , accounting for almost 9% of affected individuals [12 , 22] . Moreover , mutations in CC2D2A can also result in the genetically related and more severe Meckel syndrome , which is a perinatal-lethal disorder characterized by encephalocele , polydactyly , cystic kidneys and liver fibrosis [23] . CC2D2A is part of one of the ciliary transition zone complexes with several other JBTS proteins [7 , 21] . Two Cc2d2a mouse mutants have been described , presenting with severe brain malformation ( holoprosencephaly ) , microphthalmia , curved body axis and randomized left-right axis , all typical ciliopathy-associated phenotypes [7 , 24] . Interestingly , mouse embryonic fibroblasts from one of the reported Cc2d2a-/- mice appear to lack cilia entirely [24] whereas disruption of CC2D2A function in the other reported mutant does not compromise ciliogenesis in mouse embryonic fibroblasts [7] . Instead , the ciliary localization of several proteins ( including ARL13B , Adenylyl Cyclase III , Smoothened and Polycystin2 ) is lost , suggesting that presence of CC2D2A at the transition zone is required for appropriate targeting of proteins to the ciliary compartment [7] . The zebrafish cc2d2a mutant sentinel demonstrates a curved body axis , pronephric cysts and a striking retinal phenotype with short and dysmorphic photoreceptor outer segments [25] . In addition , the photoreceptors of cc2d2a mutants also show mislocalization of opsins in the cell body and cytoplasmic accumulation of vesicles in the apical portion of the cells and around the connecting cilium ( equivalent of the transition zone in photoreceptors ) , suggesting a defect in opsin trafficking . Opsins are the photosensitive pigment molecules concentrated at high levels in the outer segments and required for sensing the light signal . Trafficking of opsins from the cell body towards the ciliary compartment is ( at least in part ) controlled by the small GTPase Rab8 , which coats rhodopsin-carrier vesicles allowing their docking and fusion at the ciliary base [26] . Expression of a dominant-negative form of Rab8a leads to accumulation of rhodopsin-containing vesicles in photoreceptors [27] . In addition , RAB8A is also involved in ciliary membrane biogenesis in other cell types and thus appears to play a general role in orchestrating trafficking towards the ciliary compartment [28 , 29] . The trafficking defect observed in cc2d2a-/- photoreceptors appears to be mediated by loss of normal Rab8 localization [25] but the precise mechanism by which loss of this transition zone protein affects the localization of Rab8 and the trafficking of ciliary-directed opsins remains unclear . In the current work , we identify a chain of physical interactions linking CC2D2A to RAB8A through NINL and MICAL3 . Using a zebrafish model , we demonstrate that loss of Ninl function leads to a similar retinal phenotype as loss of Cc2d2a , including short outer segments , mislocalization of opsins and accumulation of vesicles . Based on the physical and genetic interactions that we identify , we propose a model in which CC2D2A provides a docking point at the photoreceptor ciliary base , allowing RAB8A-positive vesicles to bind through a series of interactions involving CC2D2A-NINL-MICAL3-RAB8A .
In order to shed light on the function of CC2D2A we performed a dedicated ciliary yeast two-hybrid assay with different fragments of CC2D2A against a panel of 164 proteins , containing most of the ciliopathy-associated proteins [30] . A direct binary interaction between CC2D2A and both isoforms ( A and B ) of the centrosome- and basal body-associated protein NINL ( Ninein-like protein ) was identified ( Fig 1a ) . NINL isoforms A and B are distinguished by the fact that isoform B is 349 amino acids shorter due to skipping of the large exon 17 ( S1a Fig ) . Both isoforms share predicted EF-hand domains in the N-terminal region as well as coiled-coil domains in the more C-terminal portion . Both isoforms display similar broad expression patterns , with the strongest expression patterns in cochlea , brain , testis , kidney and retina [31] . By generating deletion constructs for CC2D2A and subsequent evaluation of the interaction with NINL , we could pinpoint the interaction to the two predicted coiled-coil domains ( 433-637aa ) present in CC2D2A ( Fig 1a ) . Since CC2D2A and NINL isoform B demonstrated the strongest interaction ( Fig 1a ) , we focused on NINL isoform B ( NINLisoB ) for confirmation and further investigation of this interaction . Co-immunoprecipitation assays performed using full-length tagged-constructs for NINLisoB and CC2D2A , showed co-precipitation of the two proteins . FLAG-tagged LRRK2 that was used as a negative control did not co-precipitate with HA-tagged NINLisoB , which confirmed the specificity of the interaction between NINLisoB and CC2D2A in this assay ( Fig 1b ) . A reciprocal co-immunoprecipitation experiment confirmed the interaction between NINLisoB and CC2D2A ( Fig 1c ) . To further validate the interaction between CC2D2A and NINLisoB in ciliated mammalian cells , we transfected hTERT-RPE1 cells ( human telomerase reverse transcriptase retinal pigment epithelium cells ) with expression-constructs of wild-type mRFP-tagged NINLisoB , eCFP-tagged CC2D2A or a combination of both . When expressed alone , eCFP-tagged CC2D2A localizes to the ciliary base ( basal body , accessory centriole ) and also ( partly ) to the ciliary transition zone , which was visualized using anti-RPGRIP1L as a marker ( Fig 2a and 2b ) . mRFP-tagged NINL isoform B was localized at the ciliary base adjacent to the ciliary transition zone ( Fig 2c and 2d ) . When co-expressed , NINLisoB and CC2D2A co-localized at the base of the primary cilium ( Fig 2e-e” ) . To investigate the localization and function of endogenous Ninl , we turned to the zebrafish model . The zebrafish genome harbors a single ninl orthologue ( Genbank NP_001268727 ) that has 45% similarity with human NINL . Conserved domains include the predicted EF-hand domains and multiple coiled-coil domains . Cloning of zebrafish ninl from whole embryo mRNA at 5dpf revealed that all identified zebrafish transcripts lack the large exon 17 which is present only in human NINL isoform A but not in isoform B ( S1A Fig ) . Therefore , zebrafish ninl is most similar to the shorter human NINL isoform B . RNA in situ hybridization with two different antisense probes derived from the 5’-end and the 3’-end of the zebrafish ninl transcript revealed broad expression at 14–18 somites in neural tube , inner ear , developing eye and pronephros . Expression persists in the retina at least up to 6 dpf ( days post fertilization; last developmental stage assessed ) ( S1B Fig ) . Antibody staining showed punctate localization of endogenous Ninl in zebrafish retina at the base of the cilium in 4dpf larvae ( Fig 2g ) . While co-staining with Cc2d2a antibodies was not possible due to different fixation conditions , co-staining of serial sections with anti-centrin and anti-Ninl or anti-Cc2d2a antibodies respectively revealed that both endogenous proteins partially co-localize at the base of the photoreceptor cilium of 4 dpf old zebrafish larvae ( Fig 2f and 2g ) . We observed that Cc2d2a localizes slightly more apically with respect to Centrin than Ninl , consistent with Cc2d2a localization at the connecting cilium , while Ninl localization is overlapping more broadly with the basal body ( Fig 2f and 2g , schematized in j ) . In order to determine whether Cc2d2a localization is dependent on the presence of Ninl , we performed morpholino-induced knockdown studies in zebrafish . Injection of 2 ng/nl of ninl translation-blocking morpholino ( atgMO ) led to efficient knockdown of Ninl , as demonstrated by substantially decreased antibody staining in cryosections through morphant retina ( S2a and S2b” Fig ) . On Western blots of whole 5dpf larval extracts , a single strong band of 80 kDa is present in wild-type fish ( S2d Fig ) , which is consistent with results from immunoprecipitation from retinal bovine extracts with a previously published antibody against human NINL ( S2e Fig [31] ) . This band is strongly reduced in ninl atgMO injected larvae ( S2d Fig ) , supporting the specificity of the morpholino and of the antibody . Ninl knockdown led to typical ciliopathy-associated phenotypes , including curved body shape , enlarged brain ventricle and pronephric cysts ( S3a–S3g Fig ) . The specificity of the observed phenotype was confirmed by rescue experiments , in which co-injection of 2 ng/nl ninl MO with capped MO-resistant human NINL-mRNA reduced the prevalence of the curved body phenotype in a dose-dependent manner ( curved body shape in 71% of ninl atgMO injected larvae ( n = 207 ) versus 36% in ninl atgMO + ninl mRNA injected larvae ( n = 203 ) , data pooled from 2 biological replicates , P<0 . 0001 , two-tailed Fisher’s exact test; S4a–S4d Fig ) . Finally , the specificity of the observed phenotypes was further confirmed by a second morpholino against ninl targeting the splice site at the intron14/exon15 junction and thus causing aberrant splicing with premature truncation ( S5c Fig ) . This splice morpholino led to similar phenotypes as the atgMO , including ventriculomegaly and abnormal photoreceptor outer segments ( S5a and S5b Fig ) . The body curvature phenotype was absent in the splice morphants , which may be explained either by rescue of this early phenotype by maternal ninl mRNA , which remains unaffected by splice morpholinos ( as seen in some ciliopathy zebrafish mutants such as talpid3 where only the maternal zygotic mutants have a curved body shape [32] ) , or by less efficient gene knockdown with this morpholino , as normal transcript persists in addition to the aberrant transcript ( S5c Fig ) . Indeed , using the anti-NINL antibody , we observed a milder decrease of Ninl protein on Western blots and on immuno-histochemistry of retinal cryosections at 5dpf for the ninl ex15 spMO as compared to the atgMO ( S2c and S2d Fig ) . Localization of Cc2d2a at the connecting cilium , shown by anti-Cc2d2a immunostaining , was unaffected by Ninl knockdown ( Fig 2h ) . Conversely , immunostainings using anti-Ninl antibodies revealed no clear mislocalization of Ninl in the retina of cc2d2a-/- larvae ( Fig 2i ) . Taken together , these data indicate that Cc2d2a and Ninl co-localize at the ciliary base independently of each other . Since cc2d2a-/- zebrafish have prominent retinal abnormalities [25] , we focused our phenotypic analysis on the retina of ninl morphants . Retinal lamination was unaffected in ninl morphants ( Fig 3a and 3b ) . In contrast , photoreceptors demonstrated shortened axonemes and abnormal outer segments , as seen on retinal cryosections at 4 dpf stained with boron-dipyrromethene ( bodipy ) to mark the outer segment membrane disks ( Fig 3c and 3d’ ) and anti-acetylated alpha-tubulin and anti-Ift88 antibodies to mark the axoneme ( Fig 3e and 3f ) . Measurement of outer segment ( OS ) length of early 4dpf larvae , performed in a blinded manner as to injection status , revealed a significant shortening ( mean OS length 1 . 6 +/- 0 . 26 μm in ninl atgMO morphants compared to 3 . 9 +/- 0 . 32 μm in wild-type , P<0 . 0001 , unpaired Student’s t-test , n>10 larvae from each group in each of 2 biological replicates; S4e Fig ) . This retinal phenotype was observed with both the ninl translation-blocking and the splice-blocking morpholinos ( S5b’ Fig ) . Little to no photoreceptor cell death was observed with TUNEL assay on cryosections of 4dpf morphant larvae ( S3i and S3j Fig ) compared to ift88-/- retinas that are known to display prominent photoreceptor cell death at the same stage ( and were thus used as positive controls; S3k Fig ) . In general , minimal cell death was observed at 4dpf throughout the embryo , including in brain of larvae with overt ventriculomegaly ( S3l Fig ) . Co-injection of 150pg of capped human NINL mRNA with 2ng/nl ninl atgMO restored normal outer segment length ( mean OS length in rescued larvae 3 . 8 +/- 0 . 25 μm , P<0 . 0001 , unpaired Student’s t-test , n = 10 larvae; S4a’–S4c’ and S4e Fig ) . Immuno-staining with anti-opsin antibodies ( 4D2 antibody ) demonstrated significant accumulation of opsins in the inner segment and throughout the cell body of ninl-depleted photoreceptors ( Fig 3g–3i; mean intracellular fluorescence was significantly increased in ninl morphants compared to controls , P<0 . 0001 , unpaired Student’s t-test , n = 15 morphant larvae and 7 control larvae , 2 replicate experiments ) . At the ultra-structural level , two types of abnormal membrane-bound structures were observed by transmission electron microscopy in ninl morphants: large vacuole-like structures were present in the cell body and small vesicular structures accumulated around the Golgi complex and below the connecting cilium ( Fig 3j–3m; vacuolar and/or vesicular structures were present in 45/112 photoreceptors from 6 morphant eyes compared to 13/192 photoreceptors from 4 uninjected and 4 Control Oligo injected eyes; P<0 . 0001 , Fisher’s exact test ) . These phenotypes are partially reminiscent of those observed in cc2d2a-/- embryos [25] , supporting a common or coordinated function for Cc2d2a and Ninl in the process of vesicular trafficking towards the ciliary compartment . To further delineate the relationship between cc2d2a and ninl , we tested whether a synergistic effect was detectable between the two genes by using partial ninl knockdown in the cc2d2a mutant background . We observed that injection of a sub-phenotypic dose of ninl MO ( 0 . 75 ng/nl ) , which causes no discernible phenotype in wild-type larvae , significantly increased the penetrance and severity of pronephric cysts in cc2d2a mutants: 89% of ninlMO-cc2d2a-/- zebrafish developed cysts compared to 40% of uninjected cc2d2a-/- larvae ( p<0 . 0001 , Fisher’s exact test ) and the size of these cysts was significantly increased ( as measured by the area of the dilated glomerulus and proximal tubules: 0 . 044 +/-0 . 004 mm2 for cc2d2a-/- + ninlMO ( n = 16 ) as compared to 0 . 016 +/-0 . 002 mm2 for uninjected cc2d2a-/- ( n = 8 , P<0 . 0001 , unpaired Student’s t-test ) ( Fig 4a–4d ) . Importantly , the cc2d2a+/- and cc2d2a+/+ siblings from the same injection clutch did not develop pronephric cysts at these sub-phenotypic ninl MO doses ( Fig 4a and 4d ) . In the retina , the opsin mislocalization phenotype in cc2d2a-/- larvae ( 4 dpf ) was enhanced by the addition of the same sub-phenotypic dose of ninlMO ( Fig 4e–4h; P<0 . 0001 , Student’s t-test , n = 19 cc2d2a-/- + ninlMO and n = 16 cc2d2a-/- uninjected , 2 replicates ) . These findings support a genetic interaction between cc2d2a and ninl and suggest that NINL could be a genetic modifier for CC2D2A-caused disorders or even contribute to the genetic spectrum underlying Joubert/Meckel syndrome . Following this rationale , we sequenced NINL in a cohort of 346 individuals with Joubert syndrome ( from 291 families ) using a molecular inversion probes ( MIPs ) capture method followed by next-generation sequencing [33] but did not identify any individuals carrying bi-allelic rare deleterious NINL variants . We did however find 3 individuals with heterozygous NINL mutations predicted to be deleterious . Individual UW48-3 carried the homozygous missense CC2D2A mutation c . 3364C>T ( p . P1122S ) , previously shown to be causal for Joubert syndrome , and a heterozygous NINL frameshift mutation leading to a stop codon after 43 amino acids ( c . 3020delC , p . P1007Lfs*43 ) ( Fig 4i ) ( and no other rare deleterious variant in any of the known JS genes ) . Phenotypically , this subject had a severe form of JBTS with retinal dystrophy , hearing loss , ventriculomegaly in addition to the MTS and renal failure leading to death at age 7 years . In comparison , subject UW 36–3 carried the same homozygous CC2D2A c . 3364C>T ( p . P1122S ) mutation but no additional NINL variants ( or rare deleterious variants in other JBTS genes ) and presented with the “pure JBTS” phenotype , consisting only of the MTS with associated ataxia , developmental delay and respiratory rhythm disturbance ( Fig 4j ) . Subject UW07-3 carried a heterozygous NINL nonsense mutation ( c . 2446 G>A , p . R816X ) in addition to causal , compound heterozygous C5ORF42 frameshift mutations ( c . 8726delG; p . A2909Qfs*4 and c . 493delA , p . I165Yfs*17 ) . This subject presented a classical Joubert phenotype without extra-neuronal manifestations , suggesting that the additional NINL frameshift had no effect on the clinical manifestations ( Fig 4k ) . Finally , subject UW57-3 carried a heterozygous NINL missense mutation ( c . 1631A>T , p . E544V ) , predicted to be deleterious by Polyphen2 , along with bi-allelic causal TMEM67 mutations ( c . 2825T>G , p . F942C and c . 978+3 A>G ) . This individual had Joubert syndrome with coloboma but no retinal , renal or hepatic involvement ( Fig 4l ) . Given the known association between TMEM67 mutations and coloboma [12 , 34] , this additional feature is most likely explained by the causal gene mutations , while the additional NINL variant appears to have no obvious effect on the phenotype in individual UW57-3 . While it remains possible that additional sequence variants in non-JBTS genes also contributed to the enhanced phenotype in individual UW48-3 and while our findings from a large human cohort remain of anecdotal nature given the rarity of this highly heterogeneous genetic disorder , taken together with the zebrafish experiments , they suggest that NINL may act as a genetic modifier specifically for CC2D2A-caused Joubert syndrome . Previous work on the cc2d2auw38 zebrafish mutant demonstrated that loss of Cc2d2a leads to abnormal Rab8a localization in retinal photoreceptors [25] . Given the opsin mislocalization and vesicle accumulation phenotypes observed in ninl morphants , the known role of Rab8a in opsin trafficking [27 , 35 , 36] and the interaction with cc2d2a demonstrated here , we next determined whether loss of Ninl function also had an effect on Rab8a localization . For this purpose , we used a transgenic construct that drives expression of mCherry-tagged Rab8a in wild-type zebrafish photoreceptors in a punctate manner [25] . When expressed in ninl morphants ( atgMO ) , mCherry-tagged Rab8a localized in significantly fewer puncta than when expressed in controls ( 42% of expressing photoreceptors displayed Rab8 puncta in ninl morphants ( n = 38/87 from 14 larvae ) compared to 73% in uninjected controls ( n = 48/66 from 13 larvae ) , p = 0 . 0005 , two-tailed Fisher’s exact test; Fig 5a–5c ) . Instead , expression of the transgene was mostly diffuse throughout the photoreceptor cell body of ninl morphants . A similar result was obtained using an anti-Rab8a antibody that recognizes endogenous small Rab8a puncta , which are found throughout the cell body , concentrated at the synapse and in the inner and outer segments in controls ( Fig 5d-d’ ) . In ninl-knockdown larvae , the number of endogenous Rab8a puncta was significantly reduced ( Fig 5e-e’ and quantification in f: the average number of puncta per μm2 was reduced to 0 . 04 +/- 0 . 01 ( or 1 puncta per 25 μm2 ) in ninl morphants as compared to 0 . 09 +/- 0 . 01 ( or 1 puncta per 11 μm2 ) in uninjected wild-type , P = 0 . 01 , unpaired Student’s t-test ) , supporting a role for Ninl in Rab8 localization . In order to unravel the underlying molecular cause of the observed vesicle accumulation and to identify proteins that interact with NINL , we next generated N-terminal Strep/FLAG-tagged fusion proteins of NINL isoA and isoB . A single-step affinity purification combined with quantification by stable isotope labeling of amino acids in cell culture ( SILAC ) and tandem affinity purification ( TAP ) [37] were applied to isolate the protein complexes in their native functional states from human embryonic kidney 293T ( HEK293T ) cells . The complexes were subsequently analyzed by liquid chromatography coupled with tandem mass spectrometry ( LC-MS/MS ) . The identified interactome consisted of 174 unique proteins ( Fig 6a , S1 table ) . An important association was found with multiple subunits of the cytoplasmic dynein 1-dynactin motor complex ( DYNC1H1 , DYNC1LI1 , DYNC1LI2 , DYNCI2 , DYNLRB1 , DCTN1-4 , and DCTN6 ) which is involved in minus end–directed microtubule-associated transport . In addition , six actin-binding proteins ( ARP1 , ARP1B , ARP10 , CAPZA1 , CAPZA2 and CAPZB ) and three subunits of Ca2+/calmodulin-dependent protein kinase II ( CaMKII ) ( CAMK2A , CAMK2D , and CAMK2G ) , involved in non-canonical Wnt5a signaling , synaptic plasticity and kidney development [38] , were found to associate with NINL . An additional relevant NINL interaction partner identified was MICAL3 ( Microtubule-associated Monooxygenase , Calponin and LIM domain containing 3 protein ) , which is known to participate in a protein complex with RAB6 and RAB8 that is involved in the fusion of exocytotic vesicles [39] , a process that appears to be deficient in the retina of cc2d2a mutants and ninl morphants . We validated the interaction between NINLisoB and MICAL3 by reciprocal co-immunoprecipitations ( Fig 6b ) and confirmed that endogenously expressed MICAL3 is present at the photoreceptor connecting cilium in rat retina ( P20 ) , partially overlapping with the cilium and basal body marker polyglutamylated tubulin ( Fig 7b–7d’ ) . In hTERT-RPE1 cells , mRFP-tagged NINLisoB ( Fig 7e-e” ) and eCFP-tagged CC2D2A ( Fig 7f-f” ) partially overlapped with tagged MICAL3 . To evaluate the role of NINL and CC2D2A in MICAL3 localization , we silenced the expression of NINL and CC2D2A in ciliated hTERT-RPE1 cells using siRNA , which was quantified by qPCR analysis ( Fig 7j and 7k ) . Subsequent immunohistochemical stainings showed predominant MICAL3 localization at the ciliary base in non-targeting siRNA-treated cells ( Fig 7g ) whereas silencing of NINL expression resulted in a dispersed distribution of MICAL3 throughout the cell body ( Fig 7h ) . Downregulation of CC2D2A expression in hTERT-RPE1 cells had a less pronounced effect on MICAL3 localization , resulting in partial mislocalization to the cell body ( Fig 7i ) . These findings support a link between CC2D2A and MICAL3-RAB8-mediated vesicle trafficking/fusion through NINL .
Dysfunction of transition zone proteins causes several ciliopathies such as Joubert syndrome , Meckel syndrome , nephronophthisis or Usher syndrome [16 , 21 , 40–43] . Previous work suggests that transition zone proteins in general , and CC2D2A in particular , are required for correct localization of transmembrane proteins to the ciliary membrane [7 , 25] , . The mechanism by which transition zone proteins exert this function and the link to upstream ciliary-directed vesicular trafficking mechanisms remain however largely unknown . In this work , we identify NINL as a novel physical interaction partner for the transition zone protein CC2D2A and propose a model linking CC2D2A to RAB8A-controlled vesicle trafficking through a dual role for NINL in microtubule-based vesicle transport ( Fig 8 ) . The association of NINL with both the cytoplasmic dynein 1-dynactin motor complex ( Dona et al , co-submitted manuscript ) and MICAL3 supports a role for NINL in the initial transport of trans-Golgi network-derived RAB8A-MICAL3 coated vesicles towards the base of the photoreceptor cilium , while the association of NINL with CC2D2A provides a docking point for these incoming vesicles at the entrance of the ciliary compartment . Ciliary transmembrane proteins are synthesized in the cell body and travel from the Golgi towards the cilium in vesicles which move along microtubules using a cytoplasmic dynein motor [44] . Once at the entrance of the ciliary compartment , these vesicles must dock and fuse with the periciliary membrane to deliver their cargo into the ciliary membrane [45] . This path has been particularly well studied in photoreceptors , where large quantities of opsins and membrane continuously have to replenish the disks which constitute this photo-sensitive structure [26 , 46 , 35] . Opsin trafficking is severely affected in both zebrafish cc2d2a mutants and ninl morphants , suggesting that both proteins play an important role in this transport which is crucial for the correct morphogenesis and homeostasis of the outer segments . Their co-localization at the base of the photoreceptor cilium could suggest that Cc2d2a and Ninl play a similar or combined role in opsin transport or that one protein is required to localize the other . However , since we found that each protein localizes independently of the other , and that ninl knockdown enhances the cc2d2a null mutant phenotype , the relationship between them is likely more complex than a simple linear pathway . At the ultrastructural level , the loss of function phenotypes of these two proteins also slightly diverge from each other: although vesicles accumulate in both cases in the affected photoreceptors , small vesiculo-tubular structures accumulate mostly apically around the connecting cilium in cc2d2a mutants [25] , while this work shows that small vesicles and larger vacuoles are also present more basally and closer to an abnormal Golgi apparatus in ninl morphants . This suggests that both proteins are important for vesicular trafficking but play different roles in this process . NINL has been previously shown to bind several other ciliopathy proteins present at the base of cilia , specifically LCA5 and USH2A [31] , suggesting that it may play a more pivotal role in vesicular trafficking in photoreceptors than CC2D2A . Given that the zebrafish ninl morphant phenotype is more severe than the cc2d2a mutant phenotype , this further suggests a more central role for Ninl than for Cc2d2a in cilium-directed trafficking . This hypothesis is also supported by the lack of bi-allelic NINL mutations in a large human cohort of Joubert syndrome . Indeed , this may be interpreted as lack of tolerance to loss-of-function mutations in NINL , as these would lead to more severe phenotypes or early embryonic lethality . The direct interaction between NINL and the dynein 1-dynactin complex [47] which we confirmed and expanded in the associated study by Dona et al , suggests that NINL might be involved in minus end–directed microtubule-associated transport of organelles and cargo towards the base of the cilium . An appealing model would thus propose that NINL functions both more upstream in ciliary-directed vesicular trafficking than CC2D2A as well as at the base of the cilium where it interacts with several different proteins including CC2D2A . While no bi-allelic rare deleterious NINL variants were identified in our JBTS cohort , we did find heterozygous NINL mutations in individuals with Joubert syndrome . Interestingly , only the individual with causal bi-allelic CC2D2A mutations and a heterozygous truncating NINL mutation had a severe phenotype with retinal and terminal renal disease . In comparison , the individuals with causal mutations in other JBTS genes and a heterozygous deleterious NINL mutation ( or with the same causal CC2D2A mutation alone ) had the classical “pure Joubert” phenotype without retinal or renal involvement . While bi-allelic CC2D2A mutations can result in a wide range of JBTS-associated phenotypes , the majority of individuals with causal CC2D2A mutations and JBTS display the “pure JBTS” phenotype [22] . The more severe phenotype only of the individual carrying causal CC2D2A mutations and an additional NINL truncating variant suggests that deleterious variants in NINL may act as genetic modifiers specifically of CC2D2A-caused ciliopathies such as Joubert syndrome . Unfortunately , the rarity of this disorder and its prominent genetic heterogeneity with over 27 associated genes prevent identification of multiple individuals sharing the same combination of causal and additional genetic variants , precluding identification of a statistically significant effect of rare variants as genetic modifiers using human genetics alone . Our findings from a large Joubert cohort therefore remain of anecdotal nature . However , the physical and genetic interaction in zebrafish identified in this work substantially strengthen the significance of this finding and suggest that deleterious variants in NINL may indeed enhance the retinal and renal phenotype in individuals with CC2D2A-associated Joubert syndrome . The effect on the retinal phenotype may be explained by the importance of NINL function in photoreceptors as highlighted in the present study . Enhancement of the renal phenotype by the additional NINL mutation may be explained by the association identified in this study between NINL and the PKD2-target CaMKII , which is important for renal development [38] . The identification of MICAL3 as an interaction partner for NINL is of particular relevance in the context of vesicular trafficking given that MICAL3 binds RAB8A and plays a role in exocytotic vesicle fusion [39] . MICAL3 is part of the MICAL family of flavoprotein monooxygenases which regulate the actin cytoskeleton by disassembling the actin filaments . The redox function of MICAL3 is required to promote vesicle fusion , possibly by destabilizing protein complexes and remodeling the docking-fusion complexes in which it is engaged [39] . The role of RAB8A in vesicle fusion at the ciliary base has been abundantly documented in various cell types including photoreceptors [27 , 28] . While RAB8A was found to bind several ciliopathy proteins directly including CEP290 and RGPR [48 , 49] , no direct interaction has been demonstrated between CC2D2A and RAB8A , despite a functional interaction in zebrafish photoreceptors and a requirement for CC2D2A in RAB8A localization in mouse embryonic fibroblasts [24 , 25] . Our findings now provide a model explaining the link between CC2D2A and RAB8A ( Fig 8 ) : RAB8A-coated vesicles destined to the ciliary compartment are bound by MICAL3 which in turn binds NINL that is associated to the cytoplasmic dynein 1 motor complex ( Dona et al , companion manuscript ) , allowing the movement along the microtubules . Once at the base of the cilium , NINL interacts with CC2D2A , providing the specificity of the docking point at the entrance to the ciliary compartment . Finally , the redox activity of MICAL3 promotes remodeling of the complex allowing fusion of the vesicle and release of cargo into the peri-ciliary membrane . A role for CC2D2A in promoting the assembly of ciliary subdistal appendages was recently suggested whereby CC2D2A would be required for docking of transport vesicles [24] . This is compatible with our model which also provides a possible mechanism to explain how transition zone proteins may regulate ciliary protein composition by providing specific docking points at the entry to the ciliary compartment . Dysfunction of transition zone proteins can lead to a variety of ciliopathies and it is likely that abnormal ciliary protein composition is at least in part responsible for the observed disease phenotypes even in the absence of ciliogenesis defects . This provides an opportunity for the development of pathway-specific therapies aiming at modulating trafficking routes and restoring normal ciliary protein content . In this perspective , unraveling the cell biological function of disease genes such as CC2D2A as presented in the current study is a prerequisite for the future development of pharmacological treatments for patients with ciliopathies .
All animal protocols were in compliance with Swiss legal ethical guidelines and the European Union Regulatory Agency guidelines for the use of fish in biomedical research and experiments and were approved by the local authorities ( Veterinäramt Zürich TV4206 ) . Human Subject Research Procedures were approved by the Institutional Review Boards at the University of Washington and Seattle Children’s Hospital ( IRB-UW # 28853 ) , and all participants or their legal representatives provided written informed consent . Zebrafish ( Danio Rerio ) were maintained as described [50] . The cc2d2aw38/sentinel mutant ( referred to as cc2d2a mutant or cc2d2a-/- ) was previously described [25 , 51 , 52] . The transgenic Tg ( wt1b:EGFP ) line was previously described [53] . Embryos were raised at 28°C in embryo medium and pigment development was inhibited by phenylthiourea as described in Westerfield [50] . ninl translation-blocking ( 5’-CATCCTCGTCCATCCCACCACATAC-3’ ) morpholino ( MO ) and splice blocking ( 5’-CCCAACACTAAAGAGATACACCAAT-3’ ) morpholinos were designed by Gene Tools Inc . ( USA ) and 1nl was injected into zebrafish embryos at the one-cell stage . After a titration curve , we established that 2ng/nl was the optimal phenotypic dose consistently causing the major phenotypes without significant cell death , while at the low dose of 0 . 75ng/nl , no phenotypes were observed ( therefore called the “sub-phenotypic dose” ) . For the splice morpholino , the optimal phenotypic dose was 4ng/nl . For rescue experiments , cDNA encoding full length human NINL isoform B was cloned into a pCS2+ vector made compatible with the Gateway system ( Invitrogen , USA ) , pCS2+/DEST , and subsequently transcribed with the SP6 Message Machine kit ( Ambion , USA ) according to manufacturer’s instructions . The cherry-Rab8a construct was previously described [25] . All quantifications were performed blinded as to injection status . All animal protocols were in compliance with internationally recognized guidelines for the use of fish in biomedical research and experiments and were approved by the local authorities ( Veterinäramt Zürich TV4206 ) . pDONR201 vectors containing cDNA encoding human NINL isoform A and B as well as aa 1–998 , aa 433–637 , aa 992–1177 of human CC2D2A were previously described [31 , 52] . Using Gateway cloning technology , cDNA fragments encoding aa 992–1620 and aa 1171–1620 of human CC2D2A ( NM_001080522 ) were cloned in pDONR201 according to manufacturer’s instructions . pEGFP-C1-MICAL3 was kindly provided by Dr . A . Akhmanova ( Utrecht University , The Netherlands ) . The direct interaction between CC2D2A and other ciliary proteins was tested using a GAL4-based yeast two-hybrid system ( Hybrizap , Stratagene , USA ) as previously described [30] . The DNA binding domain ( GAL4-BD ) fused to full length CC2D2A was used as a bait to test the interaction with previously described ciliopathy and cilium-associated proteins fused to an activation domain ( GAL4-AD ) . Constructs encoding GAL4-BD and GAL4-AD fusion proteins were co-transformed in yeast strain PJ69-4A . The direct interaction between baits and preys induced the activation of the reporter genes , resulting in the growth of yeast colonies on selective media ( deficient of histidine and adenine ) and induction of α-galactosidase and β-galactosidase colorimetric reactions [54] . HEK293T cells transiently expressing the SF-TAP tagged NINLisoB were grown in SILAC DMEM ( PAA ) supplemented with 3 mM l-glutamine ( PAA ) , 10% dialyzed fetal bovine serum ( PAA ) , 0 . 55 mM lysine , and 0 . 4 mM arginine . Light SILAC medium was supplemented with 12C6 , 14N2 lysine and 12C6 , 14N4 arginine . Heavy SILAC medium was supplemented with either 13C6 lysine and 13C6 , 15N4 arginine or 13C6 , 15N2 lysine and 13C6 , 15N4 arginine . 0 . 5 mM proline was added to all SILAC media to prevent arginine-to-proline conversion [55] . All amino acids were purchased from Silantes . For one-step Strep purifications , SF-TAP–tagged proteins and associated protein complexes were purified essentially as described previously [37 , 56] . HEK293T cells transiently expressing the SF-TAP tagged constructs were lysed in lysis buffer containing 0 . 5% Nonidet-P40 , protease inhibitor cocktail ( Roche ) , and phosphatase inhibitor cocktails I and II ( Sigma-Aldrich ) in TBS ( 30 mM Tris-HCl , pH 7 . 4 , and 150 mM NaCl ) for 20 minutes at 4°C . After sedimentation of nuclei at 10 , 000 g for 10 minutes , the cleared lysates were transferred to Strep-Tactin-Superflow beads ( IBA ) and incubated for 1 hour before the resin was washed 3 times with wash buffer ( TBS containing 0 . 1% NP-40 and phosphatase inhibitor cocktails I and II ) . The protein complexes were eluted by incubation for 10 minutes in Strep-elution buffer ( IBA ) . After purification , the samples were precipitated with chloroform and methanol and subjected to in-solution tryptic cleavage as described previously [57] . LC-MS/MS analysis was performed on an Ultimate3000 nano HPLC system ( Dionex ) coupled to a LTQ OrbitrapXL mass spectrometer ( Thermo Fisher Scientific ) by a nanospray ion source . The raw data were analyzed using Sequest ( Thermo Fisher Scientific ) or Mascot and Scaffold ( Proteome Software ) as described previously [57] . Proteins were considered to be specific protein complex components if they were not detected in the control and were detected at least twice with two or more peptides ( peptide probability >80% ) in three experiments . The protein probability threshold was set to 99% . Three Silencer Select siRNAs targeting NINL and CC2D2A were purchased from Life Technologies ( targeting sequences are listed in S2 Table ) . For transfection , a pool of three siRNAs per gene ( 45 nM final concentration ) were plated in MW12 plates with or without glass slides . Lipofectamine RNAiMax ( LifeTechnologies ) and Opti-MEM ( LifeTechnologies ) were added to the duplexes and incubated for 10–20 minutes according to manufacturer’s protocol to allow the formation of transfection complexes . Human telomerase reverse transcriptase-transformed retinal pigment epithelium ( hTERT-RPE1 ) cells from American Type Culture Collection ( ATCC ) were then plated in MW12 plates . Per plate , non-targeting Silencer Select duplexes ( LifeTechnologies ) were included as negative controls . After 24 hours of transfection , cells were serum-starved to induce ciliogenesis . After 72 hours of transfection , knockdown-efficiency was determined by isolating total RNA from one 12-well with Trizol ( Invitrogen , USA ) , followed by first-strand cDNA synthesis ( iScript; Bio-Rad , USA ) . Quantitative PCRs using GoTaq ( Promega ) , with validated NINL- , CC2D2A- and GUSB-specific primers ( sequences are listed in S3 Table ) , were performed as previously described [31] . The second 12-well of cells were fixed with 2% paraformaldehyde , permeabilized with 1% Triton-X-100/PBS and stained with anti-MICAL3 antibodies ( kindly provided by Dr . A . Akhmanova ) . Images were taken with an Axioplan2 Imaging fluorescence microscope ( Zeiss , Germany ) equipped with a DC350FX camera ( Zeiss , Germany ) . HA-tagged NINL isoform B was expressed by using the mammalian expression vector pcDNA3-HA/DEST , FLAG-tagged CC2D2A , LRRK2 and STRAD by using p3xFLAG-CMV/DEST and strep-FLAG-tagged NINL isoform B by using pNTAPe5/DEST from the Gateway cloning system ( Invitrogen , USA ) . eGFP and eGFP-tagged MICAL3 were expressed from pEGFP-C1 ( Clontech , USA ) . All plasmids contain a CMV promoter . HEK293T cells were co-transfected using Effectene ( Qiagen , USA ) according to manufacturer’s instructions . Twenty-four hours after transfection cells were washed with PBS and subsequently lysed on ice in lysis buffer ( 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 1% Triton-X-100 supplemented with complete protease inhibitor cocktail ( Roche , Germany ) ) . HA-tagged NINL isoform B was immunoprecipitated from cleared lysates overnight at 4°C by using rat monoclonal anti-HA-beads ( Roche , Germany ) , while FLAG-tagged CC2D2A , LRRK2 , STRAD and NINL isoform B were immunoprecipitated by using monoclonal anti-FLAG M2 Agarose beads ( Sigma , Germany ) and eGFP-tagged MICAL3 was immunoprecipitated using anti-GFP polyclonal antibodies ( Abcam ) coupled to ProtA/G beads ( Santa Cruz , USA ) . After 4 washes in lysis buffer , the protein complexes were analyzed on immunoblots using the Odyssey Infrared Imaging System ( LI-COR , USA ) . Tagged molecules were detected by anti-HA , anti-FLAG or anti-GFP mono- or polyclonal antibodies . As secondary antibody IRDye800 goat-anti-mouse IgG ( Rockland Antibodies and Assays ) and Alexa Fluor 680 goat-anti-rabbit IgG ( Life Technologies ) were used . Zebrafish larvae were fixed in 4% paraformaldehyde ( PFA ) overnight at 4°C , embedded in OCT and cryosectioned following standard protocols . Sections were blocked using PBDT ( PBS , 1% DMSO , 0 . 1% Triton X , 2mg/ml BSA ) with 10% goat serum for 30 minutes at RT before incubation with primary antibodies overnight . Primary antibodies were mouse monoclonal anti-acetylated alpha tubulin ( 1:500 , clone 6-11B-1 , Sigma ) , mouse monoclonal anti-polyglutamylated tubulin GT335 ( 1:500 , gift from C . Janke , Institut Curie , France ) , mouse anti-zebrafish Cc2d2a ( 1:20 , [25] ) , rabbit anti-NINL ( 1:100; LSBio Cat# LS-C201509 ) , mouse anti-pan centrin 20H5 ( 1:200 , clone 20H5 Millipore ) , mouse anti-Rab8a ( 1:100 , clone 3G1 Novus Biologicals ) , mouse anti-opsin 4D2 ( 1:100 , gift from R . Molday , University of British Columbia ) and rabbit anti-Ift88 ( gift from B . Perkins [58] , Cleveland Clinic ) , mouse monoclonal anti-FLAG ( 1:1000 , Sigma ) , rabbit polyclonal anti-human MICAL3 [39] . Secondary antibodies were Alexa Fluor goat anti-rabbit or goat anti-mouse IgG ( Life Technologies ) used at 1:300 . Bodipy ( 1:300 , Invitrogen ) was applied for 20 minutes after the secondary antibodies and nuclei were counterstained with DAPI . Rab8 puncta detected by immuno-staining using the mouse anti-Rab8 antibody were analyzed blinded as to injection status in ImageJ . A region of interest was manually determined on single confocal sections and was thresholded ( allways with the same parameters ) ; the “analyze particles” function of ImageJ was then used to determine the number of puncta per μm2 . For quantification of intracellular fluorescence after 4D2 ( opsin ) immuno-staining , a region of interest including 10–15 photoreceptor cell bodies was determined on single confocal sections using ImageJ and the mean grey value was measured . For quantification of the proximal pronephric area , the fluorescent region corresponding to the glomerulus and the proximal tubules up to the curved part of the tubule was outlined manually in ImageJ and the “measure” function was used to determine the area of the outlined region . All quantifications were performed blinded as to injection status . Confocal imaging was performed on a Leica HCS LSI . For paraffin sections , 4 dpf old ninl morphant larvae were fixed in 4% PFA overnight at 4°C , embedded in paraffin and sectioned following standard protocols . For Transmission Electron Microsopy , ninl morphant and control larvae were fixed overnight at 4°C in a freshly prepared mixture of 2 , 5% glutaraldehyde and 2% paraformaldehyde in 0 . 1 M sodiumcacodylate buffer ( pH 7 . 4 ) . After rinsing in buffer , specimens were post-fixed in a freshly prepared mixture , containing 1% osmiumtetroxide and 1% potassiumferrocyanide in 0 . 1 M sodiumcacodylate buffer ( pH 7 . 4 ) , during 2 h at room temperature . After rinsing , tissues were dehydrated through a graded series of ethanol and embedded in epon . Ultrathin ( rostrocaudally ) sections ( 70nm ) , comprising zebrafish eyes at the optic nerve level , were collected on formvar coated grids , subsequently stained with 2% uranyl acetate and Reynold’s lead citrate , and examined with a Jeol1010 electron microscope . 346 individuals ( from 291 families ) with Joubert syndrome ( JBTS ) from the University of Washington Joubert Research Center were examined for mutations in NINL . Minimal enrollment criteria included clinical findings of JBTS ( intellectual impairment , hypotonia , ataxia ) and diagnostic or supportive brain imaging findings , or presence of a sibling with JBTS along with supportive clinical or imaging features . Procedures were approved by the Institutional Review Boards at the UW and Seattle Children’s Hospital , and all participants or their legal representatives provided written informed consent . Genomic DNA from peripheral blood or saliva was extracted and all NINL exons were captured by Molecular Inversion Probes ( MIPS ) [33] . Captured DNA was PCR amplified and sequenced on either the Illumina HiSeq or MiSeq platform . Sequence reads were mapped using the Burrows-Wheeler Aligner ( BWA v . 0 . 5 . 9 ) . Variants were called using the Genome Analysis Tookit ( GATK v2 . 5–2 ) and annotated with SeattleSeq ( http://snp . gs . washington . edu/SeattleSeqAnnotation138/ ) . Minimal quality criteria for analyzed variants were DP ( Depth ) ≥ 8 , QD ( Quality by Depth ) > 5 , and ABHet ( Heterozygous Allele Balance ) <0 . 8 . The variant list was then filtered for rare and deleterious variants . Only variants with minor allele frequency of <1% were considered given the rarity of JBTS ( estimated prevalence 1/80’000 [11] ) . All non-sense , frameshift and canonical splice-site variants , as well as missense variants with Polyphen2 scores >0 . 8 were considered deleterious . Selected variants were Sanger confirmed . For all quantifications of zebrafish experiments , the Graphpad Prism6 software ( http://www . graphpad . com/scientific-software/prism/ ) was employed to generate scatter plots , calculate mean values and SEM values , and perform statistical tests . Continuous data was analyzed using two-tailed , unpaired Student’s t-test and categorical data was analyzed using Fisher’s exact test . | Ciliopathies are a group of disorders caused by dysfunction of primary cilia , ubiquitous organelles involved in signal transduction . Mutations in CC2D2A cause two ciliopathies , Joubert and Meckel syndromes , and result in loss of ciliary protein localization . The mechanism by which CC2D2A , located at the ciliary transition zone , controls ciliary protein composition and its link to vesicular trafficking of incoming cargo remain largely unknown . Here , we identify a series of physical interactions linking CC2D2A to vesicular trafficking controlled by the small GTPase RAB8 , suggesting a new model , whereby CC2D2A provides a specific docking point for ciliary-bound vesicles at the entrance to the ciliary compartment . We first identify NINL as a physical and genetic interaction partner of CC2D2A , show that both proteins co-localize at the entrance to the cilium and demonstrate that absence of Ninl or Cc2d2a result in similar retinal phenotypes in zebrafish , including mislocalization of Rab8 . We further identify MICAL3 , a protein known to bind RAB8 , as another NINL interaction partner , thus linking CC2D2A to RAB8A-controlled trafficking . Finally , we describe an individual with Joubert syndrome , in whom combined CC2D2A and NINL mutations result in an enhanced phenotype , illustrating the impact of the detected interaction on the disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | The Ciliopathy Protein CC2D2A Associates with NINL and Functions in RAB8-MICAL3-Regulated Vesicle Trafficking |
Skin invasion is the initial step in infection of the human host by schistosome blood flukes . Schistosome larvae have the remarkable ability to overcome the physical and biochemical barriers present in skin in the absence of any mechanical trauma . While a serine peptidase with activity against insoluble elastin appears to be essential for this process in one species of schistosomes , Schistosoma mansoni , it is unknown whether other schistosome species use the same peptidase to facilitate entry into their hosts . Recent genome sequencing projects , together with a number of biochemical studies , identified alternative peptidases that Schistosoma japonicum or Trichobilharzia regenti could use to facilitate migration through skin . In this study , we used comparative proteomic analysis of human skin treated with purified cercarial elastase , the known invasive peptidase of S . mansoni , or S . mansoni cathespin B2 , a close homolog of the putative invasive peptidase of S . japonicum , to identify substrates of either peptidase . Select skin proteins were then confirmed as substrates by in vitro digestion assays . This study demonstrates that an S . mansoni ortholog of the candidate invasive peptidase of S . japonicum and T . regenti , cathepsin B2 , is capable of efficiently cleaving many of the same host skin substrates as the invasive serine peptidase of S . mansoni , cercarial elastase . At the same time , identification of unique substrates and the broader species specificity of cathepsin B2 suggest that the cercarial elastase gene family amplified as an adaptation of schistosomes to human hosts .
Human skin is a formidable barrier for much of the microbial world . In addition to the mechanical barrier of structural proteins in the epidermis , basement membrane and dermal extracellular matrix , both the epidermis and dermis are bathed in plasma proteins , including early sentinels of the immune system [1] . In order to successfully breach this barrier , an invading pathogen must degrade protein matrices while minimizing the immune response that it elicits . To this end , many invading organisms utilize insect bites or other mechanical trauma to facilitate their entry into skin , but the multi-cellular larvae of the schistosome blood fluke—the causative agent of the disease schistosomiasis—have the remarkable ability to directly penetrate host skin and gain access to dermal blood vessels [2] , [3] . The invasive larva ( e ) —termed cercaria ( e ) —is 300 µm long , 70 µm wide and comprised of roughly 1000 cells [4] . Upon direct contact with the surface of human skin , cercariae begin to secrete vesicles containing a variety of proteins and an adhesive , mucin-like substance [5] . Proteomic studies identified the majority of proteins secreted by S . mansoni cercariae . These include histolytic peptidases [6] , [7] . The most abundant peptidase in S . mansoni secretions is an S1A serine peptidase , termed cercarial elastase ( SmCE ) ( GenBank: AAC46967 . 1 ) that has activity against insoluble elastin and other fibrillar macromolecules of skin [8] . Biochemical and immunolocalization studies have confirmed SmCE activity in cercarial secretions [9] , [10] . Moreover , applying an irreversible serine peptidase inhibitor to ex vivo skin before exposure to cercariae blocks the majority of larvae from invading , suggesting that this serine peptidase has an essential role in skin penetration [11] . While the serine peptidase , cercarial elastase , plays a key role in S . mansoni skin invasion , the zoonotic species S . japonicum has no serine peptidases in its larval secretions . S . japonicum , however , encodes a number of isoforms of cathepsin B2 ( SjCB2 ) ( GenBank: CAA50305 . 1 ) , a cysteine peptidase , which are secreted by the invading parasite [12] . Moreover , orthologs of SjCB2 have been identified in the cercarial secretions of other , non-human schistosome species , including members of the genus Trichobilharzia [13] . This led us to the hypothesis that the primary invasive peptidase differs between schistosome species , with S . mansoni , a human-specific schistosome species , utilizing cercarial elastase , and S . japonicum , a zoonotic schistosome species , utilizing cathepsin B2 . Given that T . regenti also appears to utilize cathepsin B2 for skin invasion , these observations suggest that the use of a serine peptidase in invasion is the exception , not the rule , among parasitic schistosomes . The use of cercarial elastase may reflect unique properties required by S . mansoni to preferentially infect human hosts . To confirm that cathepsin B2 is also capable of facilitating skin invasion , we used a proteomic approach to identify potential substrates in host skin , for both S . mansoni cercarial elastase and S . mansoni cathepsin B2 ( SmCB2 ) ( GenBank: CAC85211 . 2 ) , a close homolog of S . japonicum CB2 . Although RNAi has been developed as a tool in juvenile and adult schistosome worms , it is currently unavailable for the intramolluscan and cercarial stages of development [14] . We therefore chose to use a proteomic approach to validate the roles of these peptidases in skin invasion . We found that the vast majority of cleaved proteins resulting from human skin exposure to either purified SmCE or SmCB2 overlap , suggesting that both enzymes are capable of facilitating parasite migration through skin . However , we also identified several potential substrates in skin that appear to be cleaved by only one of the two enzymes . Candidate substrates were further validated by in vitro cleavage of purified human skin proteins with either peptidase . Together , these observations suggest that more than one mechanism of skin penetration may have evolved as an adaptation specific to the schistosome-host relationship .
To determine the number of cercarial elastase and cathepsin B2 protein isoforms in schistosome species , all full-length protein sequences ( i . e . , those possessing the full catalytic core of the peptidase ) were collected from both GenBank ( NCBI ) and S . japonicum and S . mansoni genome annotation websites ( Sanger Institute GeneDB ) . ClustalW ( DNA Databank of Japan ) , was then used to perform multiple sequence alignments and to construct phylogenetic trees . A Blosum protein weight matrix was used to score the alignment , with a gap open penalty of 10 , a gap extension penalty of 0 . 20 , and gap distance penalty of 5 . Bootstrapping values were calculated using the p-distance method , with a count of 100 . The resulting phylogenetic tree was visualized with the program Dendroscope . S . mansoni cercariae were shed from Biomphalaria glabrata using a light induction method as previously described [11] . SmCE activity was purified from lysate as previously described with the following modifications [15] . Cercariae shed from approximately 50 snails were pelleted by centrifugation at 100 rcf for 1 minute and stored at −20°C . One milliliter of pelleted cercariae was resuspended in 5 ml 300 mM sodium acetate , pH 6 . 5 , 0 . 1% Triton X-100 , 0 . 1% Tween-20 , 0 . 05% NP40 , and sonicated for 1 minute at 40% output . Soluble protein was harvested by centrifugation for 15 minutes at 7 , 500 rcf , followed by 0 . 2 µ filtration . Fractions were again measured for SmCE activity against AAPF-pNA ( Ala-Ala-Pro-Phe-p-nitroanilide ) , and active fractions were run on 10% bis-TRIS polyacrylamide gels ( Invitrogen , Carlsbad , CA ) according to the manufacturer's specifications , and silver stained [16] . For confirmation of protein identification , bands corresponding to the correct molecular weight of SmCE were excised from the gel , and subjected to in-gel trypsin digestion , followed by LC-MS/MS peptide sequencing , described below . Active site titration was performed using the synthetic peptide inhibitor AAPF-CMK ( Ala-Ala-Pro-Phe-chloromethylketone ) . Recombinant SmCB2 was expressed in Pichia pastoris as previously described [17] . Media containing secreted protein underwent 0 . 2 µ filtration and lyophilization . SmCB2 activity was purified as previously described [17] . Fractions were monitored for SmCB2 activity against 5 µM ZFR-AMC ( Z-Phe-Arg-7-amino-4-carbamoylmethylcoumarin ) in citrate-phosphate buffer , pH 5 . 3 supplemented with 4 mM DTT . Enzyme concentration was measured by active site titration using the cysteine peptidase inhibitor CAO74 ( N- ( L-3-trans-propylcarbamoyloxirane-2- carbonyl ) -L-isoleucyl-L-proline ) . The human skin sample was taken in compliance with protocols approved by the Committee on Human Research at the University of California , San Francisco . Written informed consent was obtained for the operation and use of tissues removed . Excised human skin was stored at −80°C . For digestion experiments , skin was thawed , dissected into eight 150–170 mg sections , and placed in 1 . 5 ml microfuge tubes . To each of these skin sections 100 µl of digestion solution containing either peptidase or inhibited peptidase at 1 . 8 µM was added , along with corresponding controls . SmCE reaction buffer consisted of 100 mM TRIS-HCl , pH 8; SmCB2 reaction buffer consisted of 100 mM sodium acetate , pH 5 . 5 , 4 mM DTT . Inhibited SmCE was prepared by incubating 1 . 8 µM SmCE with 2 µM AAPF-CMK for one hour at room temperature; inhibition was monitored against AAPF-pNA , prior to its addition to skin . Similarly , inhibited SmCB2 digestion solution was prepared by incubating 1 . 8 µM SmCB2 with 2 µM CAO74 for one hour at room temperature , with full inhibition monitored by activity against ZFR-AMC , prior to its addition to skin . Inhibitor alone digestion solutions were prepared to control for human skin peptidase activity using either 2 µM AAPF-CMK in 100 mM Tris , pH 8 . 0 , or 2 µM CAO74 in 100 mM sodium acetate , pH 5 . 5 , 4 mM DTT . After addition of digestion solution to skin samples , the reaction mix was vortexed briefly , and then incubated for 5 hours at 37°C . Following incubation , reactions were centrifuged for 20 minutes at 16 , 000 rcf at 4°C , and the resulting supernatant was saved as the soluble fraction . Fifteen microliters were removed for analysis on a bis-TRIS 4-20% acrylamide gel . Gels were silver-stained and stored at 4°C . Proteomic analysis of skin digestion samples was performed by LC-MS/MS on two independent preparations as follows . Representative preparative gels are shown in Figures S1 and S2 , and contain replicate lanes of approximately 20 µg total protein for each of the skin digestion solutions . Each pair of sample lanes was cut into ten protein bands , and diced into 1–2 mm cubes , then subjected to in-gel trypsin digestion , following a previously published protocol [6] . The resulting peptides were extracted and analyzed by on-line liquid chromatography/mass spectrometry using an Eksigent nanoflow pump and a Famos autosampler that were coupled to a quadrupole-orthogonal-acceleration-time-of-flight hybrid mass spectrometer ( QStar Pulsar or QStar Elite , Applied Biosystems , Foster City , CA ) . Peptides were fractionated on a reversed-phase column ( C18 , 0 . 75×150 mm ) and a 5–50% B gradient was developed in 35 min at a 350 nl/min flow rate . Solvent A was 0 . 1% formic acid in water , solvent B was 0 . 1% formic acid in acetonitrile . Data were acquired in information-dependent acquisition mode: 1 sec MS surveys were followed by 3 sec CID experiments on computer-selected multiply charged precursor ions . Peak lists were generated using Analyst 2 . 0 software ( Applied Biosystems ) with the Mascot script 1 . 6b20 ( Matrix Science , London , UK ) . Database searches were performed using ProteinProspector v . 5 . 7 . 1 ( http://prospector2 . ucsf . edu ) [18] . Searches were performed using the SwissProt databank ( August 10 , 2010 , 519 , 348 entries ) . For false discovery rate estimation , this database was concatenated with randomized sequences generated from the original database [19] . Search parameters included selecting trypsin as the digestion enzyme , allowing one missed cleavage but no non-specific cleavages . Peptide modifications that were searched included carbamidomethyl ( Cys ) as the only fixed modification , and up to two variable modifications from among the following: oxidation ( Met ) , acetyl ( N-term ) , oxidized acetyl ( N-term ) , pyroglutamate ( Gln ) , Met-loss ( N-term ) , and Met-loss+acetyl ( N-term ) . Mass accuracy settings were 200 ppm for precursor and 300 ppm for fragment masses . Data reported in Table S3 has a Protein Prospector minimum score cutoff of 22 ( protein ) , 15 ( peptide ) and maximum expectation values of 0 . 01 ( protein ) and 0 . 05 ( peptide ) , resulting in a 2% false discovery rate . Lyophilized type I human skin collagen ( Calbiochem ) was resuspended in 17 . 5 mM acetic acid for a final concentration of 1 mg/ml . Human complement C3 ( Calbiochem ) was purchased as a 1 . 2 mg/ml stock . For SmCB2 digestion , 180 nM enzyme was added to 50 µl collagen I or 25 µl complement C3 in 50 mM sodium acetate , pH 5 . 5 , 4 mM DTT and incubated at 37°C for 1–22 hours . For SmCE digestion , 180 nM enzyme was added to 50 µl collagen I or 25 µl Complement C3 in 50 mM Tris , pH 8 . 0 and incubated at 37°C for 1–22 hours . Both enzymes were also pre-incubated with 1 mM CAO74 ( SmCB2 ) or 1 mM AAPF-CMK ( SmCE ) for one hour at room temperature prior to their addition to collagen . As a control , collagen was incubated in 50 mM sodium acetate , pH 5 . 5 , 4 mM DTT or 50 mM Tris , pH 8 . 0 for 22 hours at 37°C . To stop the reaction , 15 µl reduced SDS-PAGE loading dye ( Invitrogen ) was added , and a sample of each reaction was run on a 4–20% Tris-Glycine SDS PAGE gel ( Invitrogen ) . Bands were then electroblotted onto PVDF membrane ( Biorad , Foster City , CA ) and visualized by Coomassie Blue staining . N-terminal sequence of selected bands was determined using Edman chemistry on an Applied Biosystems Procise liquid-pulse protein sequenator at the Protein and Nucleotide Facility , Stanford University .
To outline the molecular evolution of larval peptidases in schistosomes , all previously reported orthologs were re-examined ( Figure 1 ) . In addition to the previously identified full-length cercarial elastase isoforms in S . mansoni--SmCE1a ( GenBank: AAM43939 . 1 ) , SmCE1b ( GenBank: CAA94312 . 1 ) , SmCE1c ( GenBank: AAC46968 . 1 ) , SmCE2a ( AAM43941 . 1 ) and SmCE2b ( GenBank: AAM43942 . 1 ) and Schistosoma haematobium cercarial elastase ( GenBank: AAM4394 ) --sequencing and annotation of the full S . mansoni genome revealed three additional full-length genes [15] , [20] ( Figure 1A ) . In marked contrast , the S . japonicum genome contains only a single cercarial elastase isoform ( Sjp_0028090 ) . No cercarial elastase genes have been detected in any Trichobilharzia species . Both S . mansoni and S . japonicum encode a number of cathepsin B genes ( Figure S3 ) . We chose to focus on the cathepsin B2 isotype , since a proteomic analysis of S . japonicum cercarial secretions identified a peptide sequence common to this subset [12] ( Figure 1B ) . Notably , while the S . mansoni genome encodes only a single cathepsin B2 isoform , S . japonicum encodes four CB2 isoforms . In one of these isoforms , SjCB ( Y ) 2d ( GenBank: CAX71091 . 1 ) , the nucleophilic cysteine of the active site is mutated to tyrosine , which may diminish , if not eliminate , its catalytic activity . Three of the four SjCB2 isoforms ( SjCB2b ( GenBank: CAX71088 . 1 ) , SjCB2c ( GenBank: CAX71090 . 1 ) and SjCB ( Y ) 2d correspond to the peptide sequence identified in proteomic analysis of S . japonicum cercarial secretions [12] . A full list of schistosome cercarial elastase and cathepsin B isoforms is provided as supplementary material ( Tables S1 and S2 ) . A previous proteomic study generated a list of proteins that were released as soluble peptides from ex vivo human skin upon treatment with live S . mansoni cercariae , indicating that they are actively degraded during cercarial migration through skin [21] . These included many of the structural components of skin , including extracellular matrix proteins , proteins involved in cell-cell adhesion and multiple serum proteins . To identify specific substrates of CE in skin , and to compare these to potential substrates of cathepsin B2 , we treated ex vivo skin with either peptidase . Since active , recombinant S . japonicum cathepsin B2 is not currently available , and purifying sufficient amounts of native peptidase from S . japonicum was not feasible , we used S . mansoni cathepsin B2 as model peptidase in our analysis . S . mansoni cathepsin B2 has high homology to the S . japonicum cathepsin B2 ( 90% sequence identity and 94% sequence similarity for the mature peptidase , see Figure S4 ) , including the active site and substrate binding pocket , and therefore is likely to display highly similar biochemical properties and substrate specificity [17] . SmCE was purified directly from S . mansoni cercariae , and the protein composition of proteolytically active fractions was determined by mass spectrometric analysis as a mixture of SmCE1a , 1b and 2a isoforms , but not SmCE2b . This is consistent with the isoform composition of previous proteomic analysis of S . mansoni cercarial secretions [6] , [22] . Active SmCB2 was expressed in recombinant form in P . pastoris and purified as previously described [17] . To ensure that equimolar amounts of active enzyme were added to skin samples , an active site titration was first performed for both SmCE and SmCB2 with respective covalent inhibitors . In comparison to control samples treated with inhibited peptidase , multiple skin proteins migrated through an SDS-PAGE gel as smaller fragments , i . e . fragments less than the predicted molecular weight of the full-length protein , upon addition of active SmCE or SmCB2 . These were thus identified as substrates of the specific enzyme and included multiple extracellular matrix proteins ( Table 1 ) . Addition of both SmCE and SmCB2 to skin led to the cleavage of collagen VI , which is found in interstitial tissue , and collagen XII , a collagen located in the basement membrane of the epidermis [23] . Only samples incubated with active SmCB2 showed cleavage of collagens I , III and XVIII . In addition to collagen , several other components of the extracellular matrix were degraded upon treatment with either peptidase , including vitronectin , fibronectin , and galectin . Both vimentin and talin-1 , cytoskeletal proteins that are associated with desmosomes , were cleaved upon addition of either peptidase . Two additional extracellular matrix components , tenascin-X and thrombospondin-1 , were uniquely cleaved upon addition of SmCB2 . Another subset of extracellular proteins identified as substrates of SmCE and SmCB2 were derived from blood plasma that bathes the dermis . These included components of the coagulation cascade , e . g . fibrinogen , antithrombin-III , as well as proteins involved in the host immune response , e . g . complement C3 , complement factor D . Addition of either active SmCE or SmCB2 led to the digestion of gelsolin , an actin assembly protein that exists intracellularly and in plasma . Addition of active SmCB2 also led to the digestion of both kininogen-1 and fibrinogen , both of which are members of the coagulation cascade . Complement C3 , an integral component of both the classical and alternative complement activation pathways was cleaved upon addition of either SmCE and SmCB2; complement C4A and complement D proteins , respective members of the classical and alternative complement activation pathways , were cleaved by SmCB2 alone . In addition to the extracellular proteins identified , many cytosolic proteins were also cleaved by either SmCE or SmCB2 . A complete list of peptides identified is provided as a supplementary table ( Table S3 ) . To corroborate proteomic identification of substrates in skin , candidate substrates were selected for in vitro digestion with either SmCE or SmCB2 . Type I collagen was of particular interest , given that lower molecular weight peptides of the protein were only found in skin samples treated with SmCB2 , suggesting it is cleaved by SmCB2 but not SmCE . To test this with purified protein , type I human collagen was treated with either SmCB2 or SmCE for up to 22 hours at 37°C , and cleavage of the protein was determined by SDS-PAGE analysis ( Figure 2 ) . While the majority of collagen I was degraded after 5 hours with SmCB2 ( Figure 2A ) , SmCE treatment resulted in the appearance of discrete lower molecular weight bands only after 22 hours of enzyme treatment ( Figure 2B ) . This confirms that SmCE shows reduced activity against type I collagen relative to SmCB2 , even in vitro . To confirm that the two peptidases cleaved collagen at unique sites , candidate lower molecular weight bands resulting from peptidase treatment were submitted for N-terminal sequencing , and the resulting amino acid sequence was mapped onto the full protein to determine cleavage sites ( Figure 2C ) . Consistent with previous analysis of SmCE substrate specificity , in vitro digestion of collagen I revealed that peptide bond cleavage only occurred following a leucine residue ( VRGL/TGPI ) [15] , [24] . In comparison , SmCB2 cleavage occurred following an arginine residue ( GER/GGP ) , which is consistent with its reported activity , including a level of “promiscuity” in its amino acid preference in the P2 substrate binding pocket , relative to other types of cathepsins [17] , [25] . Complement C3 was also of particular interest as a potential substrate of both SmCE and SmCB2 , given its role in the host immune response against the parasite [26] . Purified complement C3 was treated with SmCB2 or SmCE . Discrete lower molecular weight bands were visible within 1 hour of treatment with either peptidase , in comparison to inhibited peptidase controls ( Figure 3A , B ) . N-terminal sequencing of selected fragments again revealed that both SmCE and SmCB2 digested the protein in a manner consistent with their known specificities , with an arginine in the P1 position ( RR/SVQ ) for SmCB2 and and a tyrosine in the P1 position ( TMY/HAK ) for SmCE ( Figure 3C ) .
In S . mansoni , the most abundant peptidase in cercarial secretions is a serine peptidase , termed cercarial elastase ( SmCE ) for its ability to degrade insoluble elastin [8] , [22] . In addition to proteomic analysis , biochemical and immunolocalization studies have detected SmCE activity in cercarial secretions and confirmed that the enzyme is able to cleave such substrates as type IV collagen ( basement membrane collagen ) , fibronectin , laminin and immunoglobulin in vitro [9] , [10] , [27] . Here , we have shown that SmCE cleaves additional substrates in skin , including several types of collagen , other extracellular matrix proteins , and components of the complement cascade . Recent sequencing and annotation of the S . mansoni genome suggests a unique role for cercarial elastase . An expanded gene family was identified with ten individual genes that encode multiple isoforms of the peptidase . Even without a complete genome , multiple orthologs of SmCE have been also been found in S . haematobium , a related human-specific species of schistosome common throughout North Africa and the Middle East [15] . This is not the case for the zoonotic S . japonicum , a schistosome species that infects humans and other mammals throughout southeast Asia . The S . japonicum genome contains only a single gene encoding cercarial elastase . This gene corresponds to the cercarial elastase “2b” isoform in S . mansoni , for which minimal transcript is made relative to other CE isoforms ( Ingram and McKerrow , unpublished ) . While one report suggested that CE was detected by immunofluorescence in S . japonicum secretions , no cercarial elastase protein was detected in a high resolution mass spectrometric proteomic analysis of S . japonicum acetabular secretions , and no cercarial elastase-like activity was identified by direct biochemical assays [12] , [20] . Trichobilharzia regenti , an avian schistosome that is capable of invading human skin , but not establishing a successful infection in humans , encodes a cysteine peptidase , cathepsin B2 ( TrCB2 ( GenBank: ABS57370 . 1 ) ) , which has elastinolytic properties and localizes to the acetabular glands of the parasite [13] . S . japonicum also encodes a cathepsin B2 ortholog , and transcript is expressed in the developing larval stage of the parasite . Moreover , proteomic analysis has identified cathepsin B2 as being present in S . japonicum cercarial secretions [20] . Notably , S . japonicum has 40- fold higher cathepsin B activity in its acetabular secretions , relative to S . mansoni secretions [12] . It is therefore likely that in S . japonicum cercariae , cathepsin B2 , not cercarial elastase , is the predominant invasive enzyme . The differential use of these two classes of peptidases raises the question of how their respective pH optima are achieved in schistosome secretions . SmCB2 is maximally active under acidic , reducing conditions [28] . Since the influence of S . japonicum cercarial secretions on the local environment of skin is unknown , SmCB2 incubations were performed under acidic conditions to ensure optimal peptidase activity . SmCE activity is optimal in a slightly alkaline environment , and S . mansoni secretions are also alkaline; therefore all SmCE incubations were performed at pH 8 [29] . Certainly , for S . mansoni , the evolutionary selection is most likely coordination of the pH of the acetabular gland secretions and the pH optimum of the peptidase . The pH optimum of the cercarial elastase is 8 , and the pH of the secretions is also alkaline [30] . As S . mansoni cercariae migrate through skin , a microenvironment is created by the secreted material , which allows for optimal activity of the peptidase . The situation is less clear for S . japonicum and the Trichobilharzia cercariae . While some activity of the cathepsin B2 is likely to continue at neutral , or even alkaline pH , the pH optimum is slightly acidic [17] . The situation is reminiscent of the secretion of cathepsin B by macrophages into tissue compartments of vertebrates . Secreted human cathepsin B is known to degrade extracellular matrix proteins in human tissue , where it has been reported to facilitate tumor invasion and metastasis [31] . The pH optimum of mammalian cathepsin B is also slightly acidic [32] . It is not known if the microenvironment around migrating macrophages is acidic or when that enzyme is released; however , it appears that there is sufficient cathepsin B activity to cause tissue degradation . Given the unavailability of active , recombinant SjCB2 or sufficient amounts of S . japonicum cercariae from which to purify the native enzyme , we chose to perform our proteomic study with SmCB2 , which displays high sequence homology ( 90% amino acid sequence identity for the mature peptidase ) to its S . japonicum ortholog . We therefore hypothesized that it is likely to display similar biochemical characteristics , including similar substrate specificity . While we cannot say conclusively that SjCB2 is the protease facilitating S . japonicum cercarial invasion , we believe that our study , along with previous work from other groups , supports the proposed role for cathepsin B2 in host skin protein degradation [12] , [13] . This conclusion , that S . japonicum uses a cathepsin B2 peptidase for skin invasion , while S . mansoni uses a serine peptidase ( SmCE ) , has implications for the evolution of the human host-parasite relationship in schistosomiasis . A plausible model is that the cathepsin B2 family first emerged as the functional cercarial peptidase during trematode evolution . In contrast , the “humanized” parasites such as S . mansoni appear to have switched to a serine peptidase for cercarial invasion . This model is supported by the notable expansion of the serine peptidase gene family from the single 2b gene found in S . japonicum to the multiple isoforms expressed in S . mansoni [20] , [33] . While the genome of the other “humanized” parasite , S . haematobium , has not been completed , it is already clear from EST analysis that more abundant serine peptidase isoforms are present in that genome [15] . What is the advantage of a larval serine peptidase for the “humanized” schistosomes ? It is interesting to note that by BLAST analysis , some of the proteins with highest homology to cercarial elastase are mammalian mast cell peptidases , which are present in skin [12] . It is therefore possible that cercarial elastase evolved by convergence to resemble a human peptidase , in order to evade detection by the host immune system . Previous work shows that S . mansoni cercariae migrate through skin at a much slower rate than their S . japonicum counterparts [34] . Despite this , an inflammatory response to S . japonicum cercariae occurs more frequently than to S . mansoni cercariae [34] , [35] . Cathepsin B2 is a likely target of the inflammatory response , given that many cysteine peptidases are allergenic [36] . Perhaps the rapid transit of non-humanized cercariae through skin precludes the need for an invasive enzyme that mimics a host peptidase . Other aspects of immune evasion , such as the elimination of complement factors and immunoglobulin , may be common to both species . C3 and C4 components bind to the tegument of schistosomes , but are degraded by both SmCE and SmCB2 [26] , [37] . The results reported here show that S . mansoni cathepsin B2 ( a model for S . japonicum cathepsin B2 ) and S . mansoni cercarial elastase are both capable of degrading proteins in skin that act as a barrier to cercarial invasion . Many skin proteins are substrates for both enzymes , but cathepsin B2 appears to cleave a broader range of substrates , and therefore may be a more effective invasive enzyme than cercarial elastase . | Schistosome parasites are a major cause of disease in the developing world , but the mechanism by which these parasites first infect their host has been studied at the molecular level only for S . mansoni . In this paper , we have mined recent genome annotations of S . mansoni and S . japonicum , a zoonotic schistosome species , to identify differential expansion of peptidase gene families that may be involved in parasite invasion and subsequent migration through skin . Having identified a serine peptidase gene family in S . mansoni and a cysteine peptidase gene family in S . japonicum , we then used a comparative proteomic approach to identify potential substrates of representative members of both classes of enzymes from S . mansoni in human skin . The results of this study suggest that while these species evolved to use different classes of peptidases in host invasion , both are capable of cleaving components of the epidermis and dermal extracellular matrix , as well as proteins involved in the host immune response against the migrating parasite . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biochemistry",
"biology",
"proteomics",
"zoology",
"parasitology"
] | 2011 | Proteomic Analysis of Human Skin Treated with Larval Schistosome Peptidases Reveals Distinct Invasion Strategies among Species of Blood Flukes |
Adenomatous polyposis coli ( APC ) inactivating mutations are present in most human colorectal cancers and some other cancers . The APC protein regulates the β-catenin protein pool that functions as a co-activator of T cell factor ( TCF ) -regulated transcription in Wnt pathway signaling . We studied effects of reduced dosage of the Ctnnb1 gene encoding β-catenin in Apc-mutation-induced colon and ovarian mouse tumorigenesis and cell culture models . Concurrent somatic inactivation of one Ctnnb1 allele , dramatically inhibited Apc mutation-induced colon polyposis and greatly extended Apc-mutant mouse survival . Ctnnb1 hemizygous dose markedly inhibited increases in β-catenin levels in the cytoplasm and nucleus following Apc inactivation in colon epithelium , with attenuated expression of key β-catenin/TCF-regulated target genes , including those encoding the EphB2/B3 receptors , the stem cell marker Lgr5 , and Myc , leading to maintenance of crypt compartmentalization and restriction of stem and proliferating cells to the crypt base . A critical threshold for β-catenin levels in TCF-regulated transcription was uncovered for Apc mutation-induced effects in colon epithelium , along with evidence of a feed-forward role for β-catenin in Ctnnb1 gene expression and CTNNB1 transcription . The active β-catenin protein pool was highly sensitive to CTNNB1 transcript levels in colon cancer cells . In mouse ovarian endometrioid adenocarcinomas ( OEAs ) arising from Apc- and Pten-inactivation , while Ctnnb1 hemizygous dose affected β-catenin levels and some β-catenin/TCF target genes , Myc induction was retained and OEAs arose in a fashion akin to that seen with intact Ctnnb1 gene dose . Our findings indicate Ctnnb1 gene dose exerts tissue-specific differences in Apc mutation-instigated tumorigenesis . Differential expression of selected β-catenin/TCF-regulated genes , such as Myc , likely underlies context-dependent effects of Ctnnb1 gene dosage in tumorigenesis .
Colorectal cancers ( CRCs ) harbor accumulated mutations in tumor suppressor genes and oncogenes along with epigenetic alterations . Many CRCs arise from precursor lesions , such as adenomatous polyps or serrated epithelial lesions with dysplasia . Inactivating mutations in the APC ( adenomatous polyposis coli ) and TP53 tumor suppressor genes are found in roughly 80% and 60% of CRCs , respectively [1] . Oncogenic mutations activating the functions of the KRAS and PI3KCA ( phosphoinositide-3-kinase , catalytic , alpha polypeptide ) proteins are found in about 40% and 20% of CRCs , respectively [1] . Constitutional mutations inactivating one APC allele underlie the familial adenomatous polyposis ( FAP ) syndrome , where affected individuals often develop hundreds to thousands of colon adenomas during their second or third decades of life . The wild type APC allele is somatically inactivated in adenomas arising in those with FAP [1 , 2] . Mice carrying certain heterozygous , constitutional mutations inactivating Apc , such as the ApcMin mutation , may develop 50–100 small intestinal tumors and occasional colon tumors by 140 days of age and nearly all of the tumors are adenomas . Similar to the situation in FAP tumors , intestinal tumors in ApcMin mice show somatic inactivation of the wild type Apc allele [3] . The best understood function of the roughly 300 kD APC protein is regulation of the pool of β-catenin protein that functions in the canonical ( β-catenin-dependent ) Wnt signaling pathway [4–6] . In the absence of an activating Wnt ligand signal , the β-catenin destruction complex—comprised by the APC , AXIN , casein kinase I , and glycogen synthase kinase-3β factors and other proteins—promotes phosphorylation of conserved serine/threonine residues in the β-catenin amino ( N ) -terminal region . The N-terminally phosphorylated β-catenin can then be β ubiquitinated and degraded by the proteasome . Activating Wnt ligands inhibit degradation of the “free” or Wnt signaling pool of β-catenin via binding at the cell surface to the frizzled and LRP5/6 ( low density lipoprotein-related proteins 5 and 6 ) cognate receptor complex , resulting in inhibition of β-catenin phosphorylation and/or ubiquitination by the destruction complex [4 , 6] . In colon adenomas and CRCs where both APC alíeles are defective , destruction of the free pool of β-catenin is impaired and active β-catenin accumulates in the cytoplasm and nucleus , where it can complex with DNA binding proteins of the TCF ( T-cell factor family ) /Lef ( lymphoid enhancer family ) family . β-catenin functions as a transcriptional co-activator for TCFs [7] . Normally , β-catenin/TCF transcriptional activation is restricted to the crypt base , especially in the so-called crypt base columnar stem cells characterized by expression of the Wnt-regulated Lgr5 presumptive stem cell marker protein [8] . Constitutive activation of β-catenin/TCF transcription in Wnt pathway-defective adenomas and CRCs may promote a stem or progenitor cell phenotype in epithelial cells independent of cell position in the crypt [9 , 10] . Activation of β-catenin/TCF-dependent transcription also alters crypt compartmentalization and coordinated migration of cells , apparently through increased expression of the EphB2 and EphB3 receptors and via inhibition of the expression of their ligands ephrin B1 and B2 [11 , 12] . The MYC gene has been highlighted as a potentially key target gene regulated by β-catenin/TCF in CRCs . Genes encoding negative-feedback inhibitor proteins functioning in the Wnt/β-catenin/TCF pathway , such as AXIN2 , DKK1 , and NKD1 , are also activated by β-catenin/TCF ( see http://www . stanford . edu/~rnusse/pathways/targets . html for a list of candidates ) . In APC-mutant neoplastic cells , the ability of these induced regulator proteins to inhibit the Wnt signaling pathway is abrogated because the factors function upstream of or at the level of the APC protein in the pathway [13] . Besides these findings , other evidence indicates that APC inactivation may promote cancer development through β-catenin dysregulation . For instance , while most CRCs harbor APC mutations , a subset of CRCs and other cancers lacking APC mutations have CTNNB1 gene mutations resulting in production of oncogenic β-catenin proteins that are resistant to regulation by the destruction complex and that activate β-catenin/TCF transcription [6 , 13] . Also , some prior studies have used genetic approaches to study effects of Ctnnb1 gene dosage on liver , small intestine , and mammary gland tumor phenotypes in mouse models as well as effects of Ctnnb1 hemizygous inactivation state ( Ctnnb1+/- ) in Apc-mutation induced mouse embryonic development phenotypes [14 , 15] . The prior studies indicated the Ctnnb1+/- constitutional state can inhibit intestinal and liver tumorigenesis in mice carrying mutations in the Apc gene ( Apc1638N , ApcMin , or Apcfl ) [14 , 15] . In contrast , mammary gland tumorigenesis was enhanced in Apc1638N Ctnnb1+/- mice , perhaps because Ctnnb1 functions as a tumor suppressor gene in the mammary gland tumors via β-catenin’s role in E-cadherin-dependent tumor suppression [14] . Nonetheless , while the prior studies yielded evidence that β-catenin signaling dosage impacts Apc mutation-induced tumorigenesis in some tissues , the prior work did not assess the role of Ctnnb1 dosage in Apc mutation-induced colon tumorigenesis , the chief site of APC mutation-dependent tumorigenesis in humans . Moreover , the work used mice constitutionally deficient in β-catenin , not just in Apc-mutant epithelial cells , and the findings did not highlight specific factors and mechanisms that might account for effects of Ctnnb1 dosage in Apc mutation-instigated tumorigenesis in different contexts . We report here on studies of the effects of Ctnnb1 gene dosage on β-catenin protein expression and β-catenin/TCF transcription in Apc mutation-induced colon and ovarian mouse tumors and cell culture models . We provide evidence that Apc mutation-induced tumorigenesis in the colon is inhibited by Ctnnb1 hemizygous gene status through marked effects on the free pool of β-catenin in the cytoplasm and nucleus and its ability to activate key β-catenin/TCF-regulated target genes , including those encoding key stem factors , such as Lgr5 , and regulators of crypt compartmentalization , such as the EphB2/B3 receptors . We also uncovered a novel feed-forward mechanism where β-catenin protein stabilization and β-catenin/TCF transcription appear critical in regulating Ctnnb1/CTNNB1 transcription in the setting of Apc inactivation in mouse colon and human colon cancer cells . Moreover , we found that differences in the ability to activate Myc expression may underlie colon versus ovary tissue-specific differences in Apc mutation-instigated tumorigenesis in the setting of Ctnnb1 hemizygous gene dosage .
We previously described CDX2P-G22Cre transgenic mice , in which human CDX2 regulatory sequences and an out-of-frame Cre transgene allele , carrying a 22-basepair guanine nucleotide repeat tract affecting the Cre open reading frame , manifest mosaic Cre recombinase expression in caudal embryonic tissues and in epithelium of the distal ileum , cecum , colon , and rectum during adult life [16] . We also previously described CDX2P-CreERT2 transgenic mice that express a tamoxifen ( TAM ) -regulated Cre protein ( CreERT2 ) under control of human CDX2 regulatory sequences , allowing for TAM-inducible targeting of loxP-containing alleles in adult terminal ileum , cecum , colon , and rectal epithelium [17] . Using the CDX2P-G22Cre or CDX2P-CreERT2 transgenic mice , we have described the phenotypic consequences in colon epithelium of somatic , bi-allelic , inactivating mutations in Apc [16 , 17] . Consistent with our prior studies , we found CDX2P-G22Cre Apcfl/fl mice lived only for 8–20 days after birth ( median survival = 13 d; Fig 1A ) . After three daily doses of TAM to inactivate both Apc alleles in distal intestinal epithelial tissues , CDX2P-CreERT2 Apcfl/fl adult mice lived on average for 22 days ( Fig 1B ) . In marked contrast , concurrent somatic inactivation of one Ctnnb1 allele along with both Apc alleles , using either the CDX2P-G22Cre or CDX2P-CreERT2 transgene for somatic gene targeting , led to a dramatically increased life span relative to that seen in mice with Apc bi-allelic targeting , with median survival of 168 d of age in CDX2P-G22Cre Apcfl/fl Ctnnb1fl/+ mice and for 134 d after TAM treatment in the CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice ( Fig 1A and 1B ) . Consistent with our prior reports [16 , 17] , the proximal colon and cecum of both CDX2P-G22Cre Apcfl/fl mice ( when moribund at 8–20 d of age ) and CDX2P-CreERT2 Apcfl/fl mice ( only 20 days after TAM induction ) were dramatically thickened and many polypoid lesions were seen ( S1 Fig ) . Histological analysis of proximal colon epithelial tissues from these mice showed significant hyperplastic and dysplastic ( adenomatous ) changes along with frequent crypt fission and branching ( Fig 1C ) . The dramatic polyposis in cecum and colon seen following bi-allelic Apc inactivation was significantly inhibited at both early and later time points by concurrent inactivation of one Ctnnb1 allele , with no grossly discernable epithelial phenotype seen in the proximal colon and only two to four polyps in the cecum per mouse as the CDX2P-G22Cre Apcfl/fl Ctnnb1fl/+ and CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice were aged ( S1 Fig ) . The cecal polyps arising in CDX2P-G22Cre Apcfl/fl Ctnnb1fl/+ and CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice may contribute to their premature mortality relative to control mice , as no other grossly detectable intestinal lesions or pathology were noted in the mice . The Cre-mediated somatic inactivation of both Apc alleles and one Ctnnb1 allele in proximal colon epithelium was confirmed by genotyping . The rare cecal adenomas arising in CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice were found to have significant fractions of cells that escaped Cre-mediated Ctnnb1 targeting , even though Cre-mediated somatic inactivation of both Apc alleles occurred to the same extent in the rare adenomas and proximal colon mucosa ( S1 Fig ) . Compared to the situation in CDX2P-CreERT2 Apcfl/fl mice , microscopic examination of proximal colon tissues of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice revealed modest hyperplastic changes and minimal crypt branching ( Fig 1C ) . Our efforts to inactivate both Ctnnb1 alleles in colon epithelium via either CDX2P-G22Cre- or CDXP-CreERT2-mediated targeting with or without Apc inactivation indicated that colon epithelial cells completely lacking β-catenin expression and function could not be generated . This likely reflects a required role for β-catenin function in colon epithelium , perhaps not limited to Wnt signaling , but also in cadherin-mediated adhesion , centrosome assembly or other functions . Immunohistological analysis of colon sections from CDX2P-CreERT2 Apcfl/fl mice showed strong cytoplasmic and nuclear β-catenin expression in many epithelial cells , compared to the nearly uniform membrane β-catenin staining in colon epithelium of wild-type mice ( Fig 2A ) . In CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice , we observed infrequent cells with elevated cytoplasmic and/or nuclear β-catenin expression ( Fig 2A ) . Paneth cells , a specialized secretory cell linage that expresses lysozyme and other markers , are found at the crypt base in normal mouse small intestinal epithelium , but are absent in normal mouse colon epithelium . Paneth cells have been proposed to have a key role in generation and/or maintenance of the intestinal crypt stem cell niche [18] . Bi-allelic Apc inactivation has been associated with the generation of many ectopic lysozyme-expressing Paneth-like cells throughout the crypts of small intestine and colon [12 , 17 , 19 , 20] . We confirmed this finding in CDX2P-CreERT2 Apcfl/fl mice ( Fig 2A ) . Whereas no Paneth-like cells were seen in normal mouse colon , modest numbers of lysozyme-expressing cells were seen in the colons of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice ( Fig 2A ) . We also used a transgenic mouse line carrying a Cre-activated enhanced yellow fluorescence protein ( EYFP ) reporter gene at the ubiquitously expressed Rosa26 locus to monitor colon epithelial cells and glands where Cre-mediated targeting had occurred . Ectopic lysozyme-expressing cells were found in nearly all of the EYFP-positive crypts in CDX2P-G22Cre Apcfl/fl Ctnnb1fl/+ and CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice ( Fig 2B and 2C ) . The rare occurrence of Paneth-like cells in crypts without EYFP expression likely reflects the possibility that Cre may more efficiently target the loxP sites at the Apc locus than at the Rosa26 locus . Prior studies from other groups and ours have shown that Apc bi-allelic inactivation increases both cell proliferation and apoptosis in intestine and colon epithelium [12 , 17 , 19 , 21] . Following TAM-induced Apc bi-allelic inactivation in proximal colon epithelium , we confirmed significantly elongated crypts and increased cell proliferation and apoptosis relative to control epithelial tissues ( Fig 2D and 2E ) . In contrast , only modestly increased crypt height , cell proliferation and apoptosis relative to control epithelium were seen in epithelium of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice following combined Apc and Ctnnb1 gene inactivation ( Fig 2D and 2E ) . Furthermore , although bi-allelic Apc inactivation induced extensive cell proliferation in the upper half of targeted colon crypts , cell proliferation following gene targeting in CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice was largely restricted to the bottom half of each crypt , with only a slight increase in cell proliferation compared to control colon epithelium ( Fig 2D and 2E ) . The cell proliferation and apoptosis results were well correlated with the immunohistochemical studies of β-catenin levels and localization ( Fig 2A ) , suggesting differences in the strength of β-catenin-dependent Wnt signaling in cells with bi-allelic Apc defects underlie the observed effects on colon epithelial morphology , cell fate and differentiation , and cell proliferation and apoptosis . The orientation of the mitotic spindle axis may impact on cell fate decisions in intestinal epithelium . At cytokinesis , the orientation of the spindle axis in a planar fashion ( i . e . , parallel to the crypt axis ) is thought to generate two daughter cells with equivalent luminal ( apical ) and basement ( extracellular matrix ) surfaces . If the spindle axis is not oriented parallel to the crypt axis , cytokinesis generates daughter cells with differences in luminal and basement membrane surfaces and the potential for resultant differences in the fates adopted by the two daughter cells . We previously reported significant increases in the percentage of epithelial cells where the mitotic spindle axis was oriented orthogonal to the planar axis in Apc-mutant mouse colon crypts relative to wild type crypts [17] . Consistent with our prior results , in colon epithelium of CDX2P-CreERT2 Apcfl/fl mice treated with TAM to inactivate both Apc alleles , roughly 50% of the cells in mitosis had mitotic spindle axes ≥30° degrees out of the planar axis , with nearly 20% showing spindle axes between 60° and 90° out of planar alignment . In contrast , in epithelium of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice and control mice , >75% of mitotic colon epithelial cells had their mitotic spindles aligned within 30° of the planar ( crypt ) axis ( S2 Fig ) . The findings indicate β-catenin levels have a key role in the altered mitotic spindle axis phenotype of Apc-mutant colon epithelium . As described above , bi-allelic Apc inactivation acutely induces hyperproliferation and dysplastic alterations in mouse proximal colon epithelium , with the altered epithelium arising in part from expansion of the crypt progenitor compartment at the expense of the differentiated compartment , along with frequent crypt fission/branching [12 , 17 , 21] . The EphB/ephrinB signaling axis has been implicated in control of intestinal epithelial cell compartmentalization along the crypt axis and in cell migration [11 , 22] . The EphB2 and EphB3 receptors are two key effectors of compartmentalization and cell migration in the crypt , and EphB2 and EphB3 are each encoded by a gene activated in intestinal tissues by β-catenin/Tcf transcription . The EphB2/B3 receptor ligands , ephrinB1 and ephrinB2 , show highest expression levels in differentiated cells at the crypt surface , and expression of ephrins B1 and B2 is negatively regulated by β-catenin/Tcf activity [11 , 23] . Of note , EphB-ephrinB interactions generate repulsive forces that separate and compartmentalize the EphB- and ephrinB-expressing cells to maintain crypt architecture [11 , 23] . In normal mouse colon epithelium , the EphB2 and EphB3 receptors were expressed only in progenitor cells at the crypt base ( Fig 3A and 3B ) . We found bi-allelic Apc inactivation in colon epithelium not only increased EphB2 and EphB3 expression , but also perturbed the gradient of EphB2 and B3 receptor expression along the crypt axis , with EphB2/B3 expression seen even at the crypt surface in Apc-mutant crypts ( Fig 3A and 3B ) . In the case of ephrin ligand expression , our studies demonstrated strong expression of ephrinB1 and B2 in normal colon surface epithelial cells and normal colon crypt cells other than the crypt base . The normal pattern of ephrinB1/B2 expression remained largely unaffected in Apc-mutant crypts with one Ctnnb1 allele inactivated ( Fig 3C ) . In contrast , in Apc-mutant crypts where Ctnnb1 dosage was intact , ephrinB1/B2 expression was markedly down-regulated in colon surface epithelial cells and throughout the crypt ( Fig 3C ) . Our findings are consistent with those in a prior study that showed increased expression of EphB2/B3 and loss of ephrinB1/B2 expression in colon adenomas of Apcmin/+ mice [23] . Although expression of EphB2/B3 was moderately elevated in some colon epithelial cells of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice compared to control mice , elevated EphB2/B3 expression remained restricted to the crypt base region , rather than spreading throughout the crypt as was seen in Apc-mutant crypts with intact Ctnnb1 gene dosage ( Fig 3A and 3B ) . This observation suggests the reduced β-catenin levels in CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice leads to a failure to induce enough β-catenin/TCF-regulated EphB2/B3 expression to overcome the repulsive effects of the retained expression of ephrinB1/B2 ligands in Apc-mutant crypts with reduced Ctnnb1 dosage . A similar expression pattern to that seen for EphB2 and EphB3 was also found for Sox9 , a transcription factor encoded by a β-catenin/Tcf target gene . Sox9 expression is restricted to stem/progenitor cells at the normal colon crypt base ( S3 Fig ) . Sox9 expression was only modestly increased and expanded following gene targeting in crypts of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice relative to the marked changes in Sox9 levels and the number of Sox9-expressing cells in crypts from CDX2P-CreERT2 Apcfl/fl mice ( S3 Fig ) . In spite of the reduced increase in β-catenin levels in colon crypts of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice relative to CDX2P-CreERT2 Apcfl/fl mice , the resultant signaling was still sufficient to generate some ectopic Paneth-like cells ( Fig 3A and 3B ) . In addition , the modest increase in β-catenin levels in targeted crypts of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice was sufficient to induce expression in targeted crypts of a β-galactosidase reporter gene integrated into the β-catenin/TCF-regulated Axin2 locus . However , β-galactosidase expression was reduced in crypts and few if any colon surface epithelial cells expressed β-galactosidase in Apc-mutant epithelium with hemizgyous Ctnnb1 dosage ( S3 Fig ) . In contrast , uniformly strong β-galactosidase expression was seen throughout Apc-mutant colon crypts and surface epithelial cells with intact Ctnnb1 dosage ( S3 Fig ) . Taken together , the findings indicate distinct β-catenin/Tcf target genes in colon epithelium display differing transcriptional responses to β-catenin levels , with Axin2 perhaps representing a target gene capable of being activated by modest to moderate levels of β-catenin in colon epithelium . The Sox9 , EphB2 , and EphB3 genes appear dependent on higher levels of β-catenin for transcriptional activation in colon epithelium . To address mechanisms underlying suppression of crypt fission and branching in Apc-deficient colon epithelium when one Ctnnb1 allele was inactivated , we compared expression of presumptive stem cell markers in Apc-deficient colon crypts where both Ctnnb1 alleles were intact or where only one Ctnnb1 allele was active . Consistent with our prior work [17] , 20 days after TAM-induced bi-allelic Apc inactivation , we detected strong induction of enhanced green fluorescent protein ( EGFP ) expressed from the Lgr5 locus ( Lgr5-EGFP ) ( Fig 4A ) in Apc-deficient colon epithelium generated by CDX2P-CreERT2 targeting . Lgr5 is a β-catenin/TCF-regulated gene and a marker of presumptive crypt base columnar stem cells in normal colon , and the Lgr5 allele that we used has a EGFP open reading frame integrated in the locus to allow for monitoring of endogenous Lgr5 expression [8] . In Apc-mutant epithelium , we also confirmed strong induction of the Msi1 RNA-binding protein ( Fig 4B ) , another presumptive intestinal stem cell marker [24 , 25] . In contrast to a prior study where it was reported that Lgr5-expressing cells were only expanded at the lower part of the crypts in colon epithelium following mutant β-catenin induction [21] , we detected EYFP-positive and Msi-positive cells essentially throughout the Apc-mutant dysplastic colon crypts when both Ctnnb1 alleles were active , though expression of EYFP was more prominent near the crypt base region , including in the de novo crypts . While the net number of EYFP- and Msi1-expressing cells per crypt were slightly increased ( e . g . from 3–4 to 5–8 Lgr5-positive cells per crypt ) in colon epithelium of TAM-treated CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice compared to the control mice , paralleling the subtle increase in crypt fission/budding seen , the expanded population of EYFP-positive cells remained restricted to the crypt base region ( Fig 4A ) , consistent with the EphB2 and EphB3 data described above . EYFP expression patterns in colon similar to those seen in TAM-treated CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice were also obtained when we used TAM-treatment to activate the Lgr5-driven CreERT2 transgene to target Apc and Ctnnb1 alleles and EYFP expression was used to mark Lgr5-expressing cells ( Fig 4A ) . Consistent with our studies of Lgr5 and Msi expression patterns in colon epithelium , the levels of transcripts encoding presumptive stem cell markers , including Lgr5 , CD44 , Msi1 , and Hopx , were also found to increase dramatically in the colon tissues of CDX2P-CreERT2 Apcfl/fl mice ( S4 Fig ) . The induction of genes encoding stem cell markers and other selective β-catenin/Tcf target genes ( such as Axin2 , Nkd1 , Ccnd1 and Irs1 ) observed in Apc-deficient colon epithelium was significantly suppressed in colon epithelium from CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice ( S4 Fig ) . Taken together , our data indicate that the robust induction of many β-catenin/Tcf-regulated genes that is seen response to Apc inactivation was variably inhibited in reduced Ctnnb1 gene dosage and β-catenin protein levels in mouse colon epithelium . In the setting of inactivation of one Ctnnb1 allele , the inability of Apc inactivation to substantially activate certain key β-catenin/TCF-regulated genes with functions in colon crypt compartmentalization and cell migration ( e . g . , EphB2 and EphB3 ) or stem cell fate ( e . g . , Lgr5 and Msi ) is likely to underlie the dramatic abrogation of adenoma formation in CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice . Myc is a well-known β-catenin/TCF-regulated target gene [10 , 26] , and we found that the strong induction of Myc gene expression in mouse colon epithelium seen following Apc bi-allelic inactivation was abrogated when Apc bi-allelic inactivation occurred concurrently with somatic inactivation of one Ctnnb1 allele ( Fig 5A ) . Another interesting observation was that Ctnnb1 transcripts were increased roughly 2-fold in proximal colon tissues following Apc bi-allelic inactivation in colon epithelium with wild type Ctnnb1 gene dosage , compared to the levels of Ctnnb1 transcript in untargeted colon tissues of Apcfl/fl mice ( Fig 5B ) . Hemizygous Ctnnb1 gene dosage was associated with an inability of Apc bi-allelic inactivation to activate Ctnnb1 transcript levels in proximal colon tissues ( Fig 5B ) . The effects of Apc inactivation and Ctnnb1 gene dosage in mouse colon epithelium on Myc and Ctnnb1 transcript levels did not appear to simply reflect a change in the epithelial cell numbers in Apc-mutant colon epithelium , because transcripts for the epithelial markers epithelial cell adhesion molecule ( Epcam ) and E-cadherin ( Cdh1 ) were similar in the mouse colon tissues independent of genotype ( Fig 5C and 5D ) . The findings indicate that not only does Apc inactivation lead to increased β-catenin protein levels in murine colon tissues , but Ctnnb1 transcript levels in the colon tissues are also increased by Apc inactivation , consistent with an apparent feed-forward mechanism for up-regulation of Ctnnb1 transcripts following Apc inactivation . Because the Apc mutation-dependent induction of Ctnnb1 transcripts in mouse colon epithelium was not seen in the setting of reduced Ctnnb1 gene dosage , the findings imply that the feed-forward mechanism for Ctnnb1 induction may require sufficient levels of β-catenin and β-catenin/TCF-dependent transcription . Of interest with regard to a role for β-catenin and β-catenin/TCF transcription in regulating Ctnnb1 transcription is that chromatin immunoprecipitation ( ChIP ) studies from the ENCODE project indicate that the TCF4 protein , encoded by the TCF7L2 gene , is bound in the proximal promoter and exon 1 region of the CTNNB1 gene in selected cell lines . The mouse Ctnnb1 and human CTNNB1 promoter regions lack known TCF family protein consensus binding elements . Nonetheless , to further explore the role of CTNNB1 transcript and β-catenin levels in regulating CTNNB1 transcription , we generated a reporter gene construct in which a 555 bp fragment of human CTNNB1 upstream and exon 1 sequences ( -336 to +219 relative to the transcription start site ) were cloned upstream of a firefly luciferase sequence ( Fig 6A ) . This region of the CTNNB1 gene corresponds to the region that the ENCODE ChIP data indicates is occupied by the TCF family member TCF4 in some cell lines . We used two different doxycycline-induced shRNAs against CTNNB1 to study effects of antagonizing CTNNB1 endogenous transcript levels in the DLD1 human colon cancer cell line , which harbor APC defects leading to constitutive activation of β-catenin/TCF signaling ( Fig 6B ) . The doxycycline-mediated shRNA-mediated inhibition of CTNNB1 transcript levels in DLD1 cells also led to marked inhibition of MYC expression ( Fig 6B ) . In addition , we found CTNNB1 reporter luciferase activity was inhibited by about 30–40% by the reduction in CTNNB1 transcripts in the cells , whereas , the prototypical Wnt/β-catenin/TCF reporter gene construct TOPflash was more robustly inhibited by the reduction in CTNNB1 levels ( Fig 6C ) . These studies and data complement the primary mouse colon tissue findings presented above by showing that reduction of CTNNB1 levels in colon cancer cells has a demonstrable effect on CTNNB1 transcriptional activity . Canonical ( β-catenin-dependent ) Wnt signaling is dependent on increases in the levels and localization of a hypo- or un-phosphorylated pool of β-catenin , which is often termed the “free” or “active” pool of β-catenin in the cytoplasm and nucleus of cells with Wnt pathway activation [27 , 28] . Active β-catenin can function as a co-activator for TCF-dependent transcription of endogenous Wnt/β-catenin/TCF target genes [27] . To further address how reduced CTNNB1 gene expression affects β-catenin protein levels and β-catenin/TCF-regulated target gene induction in colon epithelial cells , we used an shRNA approach to antagonize CTNNB1 transcript and β-catenin protein levels in colon cell lines . In the immortalized , non-neoplastic human colon epithelial cell ( HCEC ) line , through use of a doxycycline ( DOX ) -regulated shRNA against APC , we reduced endogenous APC gene and protein expression in the cells to less than 10% of control levels ( S5 Fig ) . Following APC shRNA induction , the levels of active β-catenin , as detected with a previously described antibody against the hypo-phosphorylated or active form of β-catenin , were significantly increased , whereas only a minor increase in total β-catenin levels was seen ( Fig 7A and S6 Fig ) . Concurrent DOX-mediated induction of the APC shRNA and either of the two independent CTNNB1 shRNAs , which reduced CTNNB1 transcript levels to about 20–30% of control levels in HCECs , led to dramatic inhibition of the APC inactivation-stimulated effects on active β-catenin protein levels , but only modest to moderate reduction in the levels of total β-catenin protein ( Fig 7A and S6 Fig ) . The marked effects of the APC and CTNNB1 shRNA approaches on the active β-catenin pool , with only more modest to moderate effects on total β-catenin levels in HCEC cells were reproducible ( S6 Fig ) . Following APC shRNA induction by DOX treatment , expression of multiple β-catenin/TCF-regulated target genes , such as AXIN2 , BMP4 , NKD1 and IRS1 , was significantly induced in HCECs ( Fig 7B–7E ) . These APC shRNA-mediated increases in β-catenin/TCF-regulated target gene expression were almost completely abolished by shRNA-mediated inhibition of β-catenin ( Fig 7B–7E ) . We also studied sub-cellular localization of β-catenin in the HCEC cells following APC shRNA induction and combined APC and CTNNB1 shRNA induction by DOX . Consistent with the marked increase in active β-catenin levels following APC shRNA induction , we found β-catenin protein mainly accumulated in the cytosol and nucleus of HCECs ( S7 Fig ) . Concurrent induction of both the APC and CTNNB1 shRNAs in HCECs dramatically reduced the levels of β-catenin protein in the nucleus and cytoplasm ( S7 Fig ) , consistent with the notion that the active , signaling pool of β-catenin in the cytoplasm and nucleus is highly sensitive to changes in Ctnnb1 transcript levels . The strong inhibitory effect on the active pool of β-catenin compared to that for total β-catenin when CTNNB1 transcript levels were reduced in HCECs was further studied in three human colon cancer cell lines stably transduced with the two DOX-inducible CTNNB1 shRNAs . These included a colon cancer cell line with a gain-of-function mutation in CTNNB1 ( HCT116 ) and two colon cancer cell lines with APC loss-of-function mutations ( DLD1 and SW480 ) . At 7 days after DOX-induction of the CTNNB1 shRNAs , in the three colon cancer cell lines , we found moderate ( HCT116 ) to dramatic ( DLD1 and SW480 ) decreases in the active pool of β-catenin protein with only modest changes in total β-catenin protein levels ( S8 Fig ) . Expression of the CTNNB1 shRNAs led to potent inhibition of the expression of Wnt/β-catenin/TCF-regulated target genes in the DLD1 and HCT116 cells , including AXIN2 , BMP4 , NKD1 , LGR5 , and CD44 ( S9 Fig ) . To assess the role of β-catenin function in another Apc mutation-dependent tumor model , we explored the role of Ctnnb1 gene dosage in a mouse model of ovarian endometrial adenocarcinoma ( OEA ) arising from bi-allelic inactivation of both the Apc and Pten genes [29] . Prior studies have shown that the Wnt/β-catenin/Tcf signaling pathway is deregulated by mutations in 16%–38% of human OEAs , and PTEN mutations are often seen in the OEAs with Wnt pathway mutations [29–32] . In the mouse OEA model , tumors are initiated by conditional inactivation of the Apc and Pten genes following injection of AdCre into the right ovarian bursa of Apcfl/fl Ptenfl/fl mice [29] . Interestingly , in both Apcfl/fl Ptenfl/fl mice and Apcfl/fl Ptenfl/fl Ctnnb1fl/+ mice , adenocarcinomas morphologically similar to human OEAs formed following AdCre injection , with 100% penetrance and no difference in tumor latency between mice with two wild type Ctnnb1 alleles or one wild type and one floxed Ctnnb1 allele ( Table 1 ) . In addition , no significant differences in survival rates , tumor volumes , and rates of liver metastasis were found between AdCre-injected Apcfl/fl Ptenfl/fl mice and Apcfl/fl Ptenfl/fl Ctnnb1fl/+ littermates ( Table 1 ) , and OEAs arising in both lines of mice shared similar histological features and immunohistochemical staining patterns for cytokeratin-8 ( CK8 ) , E-cadherin and α-inhibin ( Fig 8A and Table 1 ) . Efficient Cre-mediated deletion of Ctnnb1 and Apc was confirmed in tumors from these mice , and no OEAs arose in the AdCre-injected right ovaries in Apcfl/fl Ptenfl/fl Ctnnb1fl/fl mice , indicating OEAs could not arise from cells completely lacking β-catenin . The findings on the lack of a demonstrable effect of Ctnnb1 hemizygous gene dosage in the mouse OEA model contrast with the findings above , where Apc-mutation-dependent polyposis in colon epithelium was dramatically suppressed by Ctnnb1 hemizygous inactivation . Nonetheless , similar to the situation in mouse colon , based on immunohistochemical staining , the presumptive Wnt pathway signaling-competent pool of β-catenin in the nucleus and cytoplasm was significantly reduced in the OEAs in Apcfl/fl Ptenfl/fl Ctnnb1fl/+ mice compared to OEAs in the Apcfl/fl Ptenfl/fl mice ( Fig 8B ) . We also examined the β-catenin/TCF-mediated gene transcription in the OEAs arising in the Apcfl/fl Ptenfl/fl mice and Apcfl/fl Ptenfl/fl Ctnnb1fl/+ mice . Consistent with the β-catenin dosage-dependent effects of Ctnnb1 transcripts seen in mouse Apc-mutant colon tissues described above , Ctnnb1 transcripts were significantly reduced in the OEAs arising in Apcfl/fl Ptenfl/fl Ctnnb1fl/+ mice compared to the OEAs in Apcfl/fl Ptenfl/fl mice ( Fig 8C ) . Interestingly , although the Ctnnb1 hemizygous state in the Apc- and Pten-mutant OEAs markedly suppressed the induction of some β-catenin/TCF-regulated target genes , such as Axin2 and Nkd1 ( Fig 8C ) , hemizygous Ctnnb1 function did not abrogate induction of Myc transcripts in the OEAs ( Fig 8C ) . Therefore , our findings showing that Ctnnb1 hemizgyous state did not prevent development of Apc- and Pten-mutant OEAs even though there was a reduction in β-catenin levels and expression of some β-catenin/TCF-regulated genes suggest that retention of Myc induction in OEAs with hemizygous Ctnnb1 function , but not in Apc-deficient colon epithelium with hemizygous Ctnnb1 function , may be a contributing factor in the observed differences in tumor development in the two tissues . We also studied the consequences of shRNA-mediated inhibition of CTNNB1 on MYC gene expression in the TOV112D human ovarian endometrioid carcinoma cell line that harbors a CTNNB1 oncogenic mutation leading to β-catenin/TCF dysregulation [33] . We found that doxycycline-mediated induction of the two CTNNB1 shRNAs in TOV112D cells reduced CTNNB1 levels to about 50% of baseline ( Fig 8D ) , but no statistically significant effect on MYC transcript levels was seen .
Mutations inactivating the APC tumor suppressor gene are believed to be critical initiating lesions in the majority of colon adenomas and carcinomas [6 , 34 , 35] . APC mutations are likely key contributing factors in the development of some other cancer types , including a subset of human OEAs . The best understood function of the APC protein is to act as a component of a phosphorylation- and ubiquitination-dependent destruction complex that regulates the free or active pool of β-catenin . This pool of β-catenin functions as a regulator of TCF transcription in the Wnt pathway signaling [1] . In the studies described above , we assessed the effects of Ctnnb1 gene dosage on Apc mutation-instigated tumorigenesis in mouse genetically engineered colon and ovarian tumor models and in cultured cells . We found the florid polyposis phenotype resulting from somatic Apc bi-allelic inactivation in mouse colon epithelium is potently inhibited by concurrent somatic inactivation of one Ctnnb1 allele . The few polyps arising in CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice were found to have escaped Ctnnb1 targeting , though Cre-mediated somatic inactivation of both Apc alleles occurred in the lesions , likely reflecting strong positive selection for maintenance of wild type Ctnnb1 gene dosage for Apc-mutant colon adenomas to arise and persist . In contrast to the situation in colon epithelium , in a mouse model of the human OEAs that harbor inactivating mutations in the APC and PTEN genes , we found that , regardless of whether the mice had two wild type Ctnnb1 alleles or one wild type and one targeted Ctnnb1 allele , adenocarcinomas morphologically similar to human OEAs formed with 100% penetrance and no differences in latency , size , morphology , or metastatic potential of the lesions arising from AdCre-mediated targeting of the Apc and Pten genes . In-depth mutational analyses of the germline and somatic mutations in adenomas arising in patients with FAP led to the proposal there was strong biological selection for a “just-right” level of β-catenin signaling that would be optimal for tumor formation [36] . Some prior studies have used genetic approaches to study experimentally the effects of Ctnnb1 gene dosage on Apc mutation-dependent tumorigenesis in the small intestine , liver , and mammary gland in mouse models [14 , 15] . The earlier work indicated that the Ctnnb1+/- constitutional hemizygous state can inhibit intestinal and liver tumorigenesis in mice carrying Apc mutations . In mammary gland tumorigenesis , tumorigenesis was enhanced in Apc1638N Ctnnb1+/- mice relative to Apc1638N Ctnnb1+/+ mice , perhaps because Ctnnb1 functions as a tumor suppressor gene in Apc1638N mammary gland tumors via β-catenin’s role in E-cadherin-dependent tumor suppression [14] . While our findings in a mouse Apc mutation-dependent colon tumorigenesis model are consistent with the prior work on Apc mutation-instigated small intestine and liver tumorigenesis , some significant differences in the studies should be noted . The prior work emphasized models where the mice carried constitutional mutations in one Apc allele and tumors arose following stochastic loss or inactivation of the remaining wild type Apc allele . In addition , mice in the prior small intestine work were constitutionally hemizygous for Ctnnb1 . In our Apc mutation-dependent colon tumorigenesis model , both Apc alleles are somatically inactivated in colon epithelium by Cre-mediated targeting , and the Ctnnb1 hemizygous deficiency state was also somatically generated only in the colon epithelial cells by Cre-mediated targeting . In addition , our OEA model work contrasts with the prior published work , as it also relies on somatic targeting of Apc and Ctnnb1 . The OEA results also differ from our own colon tumorigenesis results , as the findings indicate Ctnnb1 hemizygous gene dosage had no demonstrable effect on cancer latency , size , or morphology or the metastatic potential of mouse OEAs arising from combined somatic inactivation of Apc and Pten . Besides highlighting tissue-specific differences for Ctnnb1 gene dosage in Apc mutation-instigated colon and ovarian tumorigenesis , our studies and data have provided several unique and in-depth insights into cell and tissue mechanisms by which Ctnnb1 gene dosage likely contributes to Apc mutation-dependent phenotypes in mouse colon epithelium . Inactivation of one Ctnnb1 allele markedly inhibited the increases in β-catenin cytoplasmic and nuclear levels that result from bi-allelic Apc inactivation in mouse colon epithelium . In turn , there was strikingly attenuated expression of key β-catenin/TCF-regulated target genes , including those encoding the EphB2/B3 receptors , and the stem cell markers Lgr5 , Msi1 , and Hopx . Of significant interest in terms of a likely key mechanisms through which Ctnnb1 gene dosage inhibits adenoma formation , the inability of the Apc-mutant colon epithelial cells to up-regulate and alter the crypt ( high ) -surface ( low ) gradient of EphB2/B3 expression appears to restrict high levels of EphB2/3 expression to the crypt base . Activated β-catenin/TCF transcription has been implicated in repression of ephrinB expression [11 , 23] . We found that the robust ephrinB1/B2 expression seen in the upper two-thirds of normal colon crypts as well as in the normal colon surface epithelium was maintained in Apc-mutant crypts when one Ctnnb1 allele was inactivated . In contrast , Apc-mutant crypts with intact Ctnnb1 dosage markedly down-regulated ephrinB1/B2 expression and dramatically upregulated and expanded EphB2/B3 expression throughout the colon crypts . Because the EphB and ephrin molecules mediate critical repulsive interactions in intestinal crypts , the maintenance of the normal inverse EphB/ephrinB gradient from crypt base to cell surface in Apc-mutant crypts where one Ctnnb1 allele is inactive restricts the expansion of the Lgr5-positive crypt stem cell pool and the crypt fission/branching that would result from unrestrained crypt stem cell expansion and altered migration [11 , 23] . As a result of the preservation of the inverse gradient of EphB/ephrinB expression in Apc mutant crypts with reduced Ctnnb1 dosage , stem cell expansion and the dysplastic and adenomatous changes induced by Apc inactivation in colon epithelium are potently inhibited , even though Lgr5 and some other stem cell marker genes are modestly increased in expression in the targeted crypts of CDX2P-CreERT2 Apcfl/fl Ctnnb1fl/+ mice relative to crypts in normal mice . Our findings showing that Ctnnb1 transcripts are up-regulated in Apc-mutant mouse colon epithelium as well as in Apc-mutant mouse OEAs , together with our findings that Ctnnb1 hemizygous gene dosage inhibited Apc mutation-dependent Ctnnb1 transcript induction in colon and ovarian tumor models imply that the Ctnnb1 gene is subject to feed-forward activation by β-catenin levels and β-catenin/TCF-regulated transcription . Of interest with regard to a role for β-catenin and β-catenin/TCF transcription in regulating Ctnnb1 transcription are chromatin immunoprecipitation ( ChIP ) studies from the ENCODE project reporting that the TCF4 protein , encoded by the TCF7L2 gene , is bound in the promoter region of the CTNNB1 gene in selected cell lines . Based on our mouse colon tissue studies and the ENCODE project findings , we generated a CTNNB1 reporter gene construct containing 555 bp of human CTNNB1 upstream and exon 1 sequences and found that shRNA-mediated inhibition of CTNNB1 endogenous gene expression in the APC-mutant DLD1 human colon cancer cell line led to inhibition of the activity of the CTNNB1 reporter gene . These data demonstrate that CTNNB1 transcript levels affected CTNNB1 transcription in colon cancer cells . The lack of known TCF consensus element binding sites in the mouse Ctnnb1 and human CTNNB1 promoter regions currently limits support for the argument that β-catenin/TCF transcription directly regulates Ctnnb1 transcription , though further studies to address the point will need to be pursued . In contrast to the near complete abrogation of Myc induction in Apc-mutant colon epithelium with one Ctnnb1 allele , Myc induction was retained in the Apc-mutant mouse OEAs with one functional Ctnnb1 allele . Of note , in prior studies , it has been shown that hemizyous inactivation of Myc dramatically inhibited Apc mutation-induced small intestine tumor phenotypes , but not Apc mutation-induced effects on liver cell proliferation and size [37 , 38] . Hence , the findings from our work and these prior studies [37–39] highlight Myc as perhaps one of the key β-catenin/TCF-regulated genes with tissue-specific differences in its regulation by β-catenin/TCF that may account for why Ctnnb1 hemizygous state abrogates Apc mutation-induced effects in some tissues ( e . g . , small intestine and colon epithelium ) but not in other tissues ( e . g . , liver and ovarian epithelium ) . The identification of a possible feed-forward mechanism for β-catenin and β-catenin/TCF transcription in regulating Ctnnb1 transcript levels following Apc inactivation are also potentially interesting with regard to Myc , because the ENCODE project work also indicates that the Myc protein is bound in the promoter and intron one regions of the Ctnnb1 gene in selected cell lines . As such , β-catenin/TCF transcription may cooperate in some fashion with Myc , itself encoded by a β-catenin/TCF target gene , in a more complex feed-forward loop to activate Ctnnb1 transcription in certain cell types when Apc is inactivated . Further studies are needed to better understand the details of the apparent feed-forward mechanisms through which β-catenin and β-catenin/TCF transcription may regulate Ctnnb1/CTNNB1 transcription in the setting of Apc/APC inactivation . Besides Myc , other β-catenin/TCF target genes may also be differentially regulated in a tissue- and context-dependent fashion , perhaps contributing in some fashion to the tissue-specific differences of Ctnnb1 hemizygous gene dosage on Apc mutation-instigated tumorigenesis observed . In addition , the basis for the dramatic changes in the free or active pool of β-catenin protein relative to the more modest effects on total β-catenin protein levels when Ctnnb1/CTNNB1 transcript levels are reduced in colon epithelial cells with Wnt pathway dysregulation remains to be elucidated . Nonetheless , our findings highlight the possibility that novel approaches and/or agents that can reduce CTNNB1 transcript levels and/or the free pool of β-catenin protein might have quite dramatic effects on the development and perhaps persistence of neoplastic cells with Wnt pathway defects .
To target Apc and/or Ctnnb1 alleles in colon tissues , CDX2P-G22Cre transgenic mice [16] , or CDX2P-CreERT2 transgenic mice [17] , or Lgr5-EGFP-IRES-CreERT2 ( B6 . 129P2-Lgr5tm1 ( cre/ERT2 ) Cle/J ) transgenic mice [8] ( The Jackson Laboratory , Bar Harbor , ME ) , were first intercrossed with mice homozygous for Apc targeted alleles ( Apcfl/fl , 580S ) [40] and Ctnnb1-targeted alleles ( Ctnnb fl/fl , B6 . 129-Ctnnb1tm2Kem/KnwJ ) [41] . The resulting Cre positive Apcfl/+ Ctnnb1 fl/+ mice were then crossed to Apcfl/fl mice in order to target two alleles of Apc and one allele of Ctnnb1 ( Apcfl/fl Ctnnb1 fl/+ ) or only alleles of Apc ( Apcfl/fl ) , respectively . The Cre positive Apcfl/fl Ctnnb1 fl/+ and Apcfl/fl littermates were compared and the Cre negative littermates served as normal control . The CDX2P-CreERT2 Lgr5-EGFP-IRES-CreERT2 Apcfl/fl compound mice or CDX2P-CreERT2 Lgr5-EGFP-IRES-CreERT2 Apcfl/fl Ctnnb1 fl/+ compound mice were constructed by crossing Lgr5-EGFP-IRES-CreERT2 Apcfl/fl mice to CDX2P-CreERT2 Apcfl/fl mice and CDX2P-CreERT2 Apcfl/fl Ctnnb fl/fl littermates , respectively . To assess Cre-mediated recombination or Wnt signaling in colon epithelium , mice carrying the Gt ( ROSA ) 26Sor tm1 ( EYFP ) Cos/J reporter allele ( EYFP ) [42] or the B6 . 129P2-Axin2tm1Wbm/J allele ( Axin2-LacZ ) [43] ( The Jackson Laboratory ) were bred into CDX2P-CreERT2 Apcfl/fl mice or CDX2P-CreERT2 Apcfl/fl Ctnnb1 fl/+ mice . To assess the role of β-catenin function in another Apc mutation-dependent mouse tumor model , we used the previously described mouse model of ovarian endometrioid adenocarcinoma ( OEA ) , arising from bi-allelic inactivation of both the Apc and Pten genes ( Apcfl/fl Ptenfl/fl ) [29] . To introduce the floxed Ctnnb1 allele , Ctnnb fl/fl mice were first crossed to Apcfl/fl Ptenfl/fl mice to generate Apcfl/+ Ptenfl/+ Ctnnb fl/+ mice , and then Apcfl/+ Ptenfl/+ Ctnnb fl/+ mice were bred to Apcfl/fl Ptenfl/fl mice to generate Apcfl/fl Ptenfl/fl , and Apcfl/fl Ptenfl/fl Ctnnb fl/+ mice . All mice were on a mixed C57BL/6 and 129 background , which were backcrossed to C57BL/6 mice for at least 10 generations , except the EYFP reporter mice and the CDX2P-CreERT2 transgenic mice , which were backcrossed for 7 and 3 generations , respectively . All experimental compound mice were on a mixed C57BL/6 and 129 background , and littermates with similar genetic background and different genotypes were used for comparison ( see breeding scheme above ) . Animal husbandry and experimental procedures were carried out under approval from the University Committee on Use and Care of Animals , University of Michigan and according to Michigan state and US federal regulations . All the mice were housed in specific-pathogen free ( SPF ) conditions . After weaning , rodent 5001 chow and automatically supplied water were provided ad libitum to mice . Animals were euthanized and analyzed at the specified time points , based on particular study design parameters or defined humane treatment and euthanasia guidelines . Human colonic epithelial cells ( HCEC ) [44] were kindly provided by Dr . Jerry Shay ( UT Southwestern Medical School , Dallas , TX ) and routinely grown on media made up with Dulbecco's modified Eagle's medium ( DMEM; Life Technologies , Grand Island , NY ) and medium 199 ( Thermo Scientific HyClone , Waltham , MA ) at the ratio of 4:1 , supplemented with EGF ( 25 ng/mL ) ( PeproTech , Inc , Rocky Hill , NJ ) , insulin ( 10 μg/mL , Life Technologies ) , hydrocortisone ( 1 μg/mL ) , transferrin ( 2 μg/mL ) , sodium selenite ( 5 nm ) ( all from Sigma-Aldrich , St Louis , MO ) , and 2% cosmic calf serum ( Thermo Scientific HyClone ) . Cells were cultured on Primaria dishes ( BD Biosciences , San Jose , CA ) or chamber slides ( Lab-Tek II , Vernon Hills , IL ) and grown in 2% oxygen and 7% carbon dioxide . HCEC cells were infected with a TRIPZ inducible lentiviral vector ( GE Dhamacon , Lafayette , CO ) carrying a shRNA against APC ( targeting sequence: 5’-CAAATCATATGGATGATAA-3’ ) or a non-silencing scramble shRNA ( Scrmbl ) . Cells were selected with 1μg/mL of puromycin ( Sigma-Aldrich ) for 5 days . The resulting stable cell lines ( HCEC/APC shRNA or HCEC/Scrmble ) were further transduced with TRIPZ lentiviruses driving expression of two different shRNAs targeting CTNNB1 ( CTNNB1-1 and CTNNB1-2; targeting sequence for CTNNB1-1: 5’-TGGGTGGTATAGAGGCTCT-3’; and targeting sequence for CTNNB1-2: 5’- AGCTGATATTGATGGACAG-3’ ) or a non-silencing scramble shRNA ( Scrmbl ) . Human colon cancer cell lines , HCT116 , SW480 , and DLD1 , and human OEA-derived cell line , TOV-112D , were grown in 5% CO2 with DMEM containing 10% fetal bovine serum and penicillin/streptomycin . HCT116 , SW480 , DLD1 and TOV-112D cells stably expressing the shRNAs targeting CTNNB1 ( CTNNB1-1 and CTNNB1-2 ) or a non-silencing scramble shRNA ( Scrmbl ) were made in the same way as HCEC cells . Expression of shRNAs was induced by incubation of cells with doxycycline ( DOX; Sigma-Aldrich ) at 2 μg/ml or a solvent control for 3 days ( for HCEC cells ) or 7 days ( for HCT116 , SW480 , and DLD1 cells ) . The degree of inhibition of the shRNAs on APC transcripts and protein and CTNNB1 transcripts and the respective β-catenin protein was assessed by qRT-PCR and Western blotting assays . DNA fragment containing human CTNNB1 sequences from −336 to +219 relative to the transcription start site was obtained by PCR amplification of genomic DNA , and was subcloned upstream from the luciferase reporter gene in the pGL3Basic reporter vector ( Promega , Madison , WI ) , using the MluI and XhoI sites . The forward primer for generating the CTNNB1 reporter construct was 5′-ACGCGTGCTGCTCTCCCGGTTCG -3′; the reverse primer for generating the CTNNB1 reporter construct was 5′- CTCGAGCAGGGGAACAGGCTCCTC-3′ . Mice with the CDX2P-CreERT2 transgene or Lgr5-EGFP-IRES-CreERT2 were injected intraperitoneally with TAM ( Sigma-Aldrich ) dissolved in corn oil ( Sigma-Aldrich ) . For two TAM daily dosing , we used 150mg/kg weight per dose; for three consecutive daily doses , we administered TAM at 100mg/kg weight per dose . Mice were injected with TAM at 2- to 3-months of age . For OEA induction , 5 x 107 plaque-forming units of replication-incompetent recombinant adenovirus expressing Cre recombinase ( AdCre , from the University of Michigan’s Vector Core ) were injected into the right ovarian bursal cavities of 6–10 week old female mice as previously described [29] . Mouse tissues were prepared for paraffin-embedding or cryosectioning as described previously [16] . For assessment of cell proliferation , mice were pulsed with 5-bromo-2-deoxyuridine ( BrdU; Sigma-Aldrich ) for 1 hr before euthanasia . Sections of paraffin-embedded human or mouse tissues were subjected to immunohistochemical analysis as previously described [45] . The following primary antibodies were used for immunohistochemical analysis with sections of paraffin-embedded tissues: mouse anti-BrdU ( 1:500; BD Biosciences ) ; rabbit anti-lysozyme ( 1:2000; Dako , Carpinteria , CA ) ; mouse anti-β-catenin ( 1:800; BD Biosciences ) ; rat anti-CK8 ( 1:100 , The Developmental Studies Hybridoma Bank , Iowa City , IA ) ; goat anti-E-cadherin ( 1:100 , R&D Systems , Minneapolis , MN ) ; mouse anti-α-inhibin ( 1:200 , Bio-Rad Laboratories , Inc . , Raleigh , NC ) . For BrdU staining , tissue sections were treated with 2N HCl at 37°C for 30 min after performing antigen retrieval with citrate buffer ( pH 6 . 0 , Biogenex , San Ramon , CA ) . For immunofluorescence using frozen sectioned tissues , mouse colon and intestinal tissues were fixed in 4% paraformaldehyde ( PFA ) overnight , cryo-protected and frozen in O . C . T . ( Fisher HealthCare , Houston , TX 77038 ) . Standard immunofluorescence staining was performed on 6-μm frozen sections with rabbit anti-lysozyme antibody ( 1:1000; Dako ) . For immunofluorescence using paraffin-embedded tissues , the following primary antibodies were used: rabbit anti-lysozyme ( 1:1000; Dako ) , rabbit anti-Sox9 ( 1:200; Millipore , Temecula , CA ) , rat anti-Msi1 ( 1:500; a gift from Dr . Hideyuki Okano [46 , 47] ) , goat anti-EphB2 ( 1:100; R&D Systems ) , goat anti-EphB3 ( 1:100; R&D Systems ) , goat anti-ephrinB1 ( 1:200; R&D Systems ) , goat anti-ephrinB2 ( 1:100; R&D Systems ) , mouse anti-α-tublin ( 1:1000; Sigma-Aldrich ) , and rabbit anti-Crb3 ( 1:1000; kindly provided by Dr . Benjamin Margolis at University of Michigan ) . The secondary antibodies used were Alexa fluor 488-conjugated donkey anti-goat , Alexa fluor 488-conjugated donkey anti-rabbit , Alexa fluor 488-conjugated goat anti-rabbit , Alexa fluor 594-conjugated goat anti-mouse , Alexa fluor 488-conjugated goat anti-mouse , Alexa fluor 594-conjugated goat anti-rabbit , and Alexa fluor 488-conjugated goat anti-rat ( Molecular Probes , Life Technologies , Carlsbad , CA ) , diluted at 1:1000 . DNA was labeled by Hoechst 33342 ( Molecular Probes , Life Technologies ) by adding to the washing buffer at 5 μg/ml . β-gal analysis for mouse with Axin2-LacZ reporter was performed as described previously [16] . To assess apoptosis , TUNEL assays were undertaken using 4-μm sections of formalin-fixed , paraffin-embedded mouse colon tissues , after the tissue sections were deparaffinized , rehydrated and treated with 20 μg/ml protease K ( Roche Applied Sciences , Indianapolis , IN ) at 37°C for 15 min . The nicked DNA was labeled by using terminal transferase ( TdT ) ( New England Biolabs , Ipswich , MA ) and Biotin-16-UTP ( Roche Applied Sciences ) according to the manufacturer’s recommendation . The signal was detected by using the Vectastain ABC kit ( Vector Laboratories , Burlingame , CA ) according to the manufacturer’s suggestion . The spindle angles were defined by the orientation of mitotic spindles , based on α-tubulin staining , relative to the most adjacent apical membrane , as indicated by Crb3 staining . The mitotic spindle axis angle relative to the planar axis of the cells ( defined by the most adjacent apical membrane ) was measured by ImageJ ( NIH ) . Western blot analyses on lysates from HCEC , HCT116 , SW480 , DLD1 and TOV-112D cells were performed as described [45] . The following antibodies were used: mouse anti-active β-catenin ( 1:2000; Millipore , Temecula , CA ) , mouse anti-total β-catenin ( 1:10 , 000; BD Biosciences ) , rabbit anti-APC ( clone C-20 , 1:1000; Santa Cruz Biotechnology , Santa Cruz , CA ) , mouse anti-APC ( clone Ab-5 , 1:1000; Millipore ) , and mouse anti-β-actin ( 1:10 , 000; Sigma ) . The density of Western blotting bands was quantified using AlphaImager HP system ( ProteinSimple , San Jose , CA ) . cDNA was synthesized using a high capacity cDNA reverse transcription kit ( Applied Biosystems , Foster City , CA ) . qRT-PCR was performed with an ABI Prism 7300 Sequence Analyzer using a SYBR green fluorescence protocol ( Applied Biosystems ) . See S1 Table for primer sequences used in qRT-PCR . DLD1 cells , stably expressing two different doxycycline-inducible shRNAs targeting CTNNB1 ( CTNNB1-1 and CTNNB1-2 ) or a non-silencing scramble shRNA ( Scrmbl ) , were treated for 4 days with DOX at 2 μg/ml or a solvent control . At second day during DOX treatment , cells were plated in 35-mm six-well plates . After 12—24h cells were then transfected with 0 . 5 μg of CTNNB1 reporter or TOPflash , 1 μg of pCDNA3 ( Invitrogen ) and 0 . 05 μg of PRL-CMV Renilla luciferase reporter vector ( Promega ) using Mirus TransIT-LT1 transfection reagent ( Mirus Bio , Madison , WI ) according to the manufacturer's protocol . Cells were harvested 45h later and luciferase activities were measured using a Dual-luciferase kit and GloMax-Multi Detection System from Promega . All data for qRT-PCR were evaluated by Student's t test and asterisks denote significance with P < 0 . 05 . Error bars denote standard deviations ( S . D . ) . Kaplan-Meier survival curves were compared by log-rank ( Mantel-Cox ) test . Chi-Square test was used to determine significance when mitotic spindle angles were compared among mice with different genotypes . P < 0 . 05 is considered statistically significant . Animal husbandry and experimental procedures were carried out under approval from the University Committee on Use and Care of Animals , University of Michigan ( PRO00005075 ) and according to Michigan state and US federal regulations . All the mice were housed in specific-pathogen free ( SPF ) conditions . After weaning , rodent 5001 chow and automatically supplied water were provided ad libitum to mice . Animals were euthanized and analyzed at the specified time points , based on particular study design parameters or defined humane treatment and euthanasia guidelines . | Enhanced Wnt signaling contributes to colorectal and other cancers . β-catenin functions in Wnt signaling as a T cell factor ( TCF ) transcriptional co-activator . Previous studies showed specific β-catenin dosage favors Wnt signaling-dependent tumorigenesis for some tumor types . However , earlier studies emphasized the role of constitutional Ctnnb1 and Apc gene variations , rather than somatic gene targeting , and the work focused on small intestine tumors and no effects on colon tumor phenotypes were described . Furthermore , definitive insights were lacking into how reduced Ctnnb1 gene dosage affected Apc mutation-dependent tumorigenesis . Here , we show somatic inactivation of one Ctnnb1 allele dramatically inhibits mouse colon adenomatous polyposis induced by somatic bi-allelic Apc inactivation . In contrast , Ctnnb1 hemizygous inactivation does not affect mouse ovarian endometrioid adenocarcinoma development arising from Apc- and Pten-inactivation . Ctnnb1 hemizygous gene dose dramatically reduces the active pool of β-catenin , leading to the significant inhibition of β-catenin/TCF-regulated target gene expression , including those encoding key stem cell regulatory and crypt compartmentalization factors in colon epithelium . Tissue-specific differences for expression of selected β-catenin/TCF-regulated genes , such as Myc , may contribute to the context-dependent effects of Ctnnb1 gene dosage in Apc mutation-driven colon and ovarian tumors . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Tissue-Specific Effects of Reduced β-catenin Expression on Adenomatous Polyposis Coli Mutation-Instigated Tumorigenesis in Mouse Colon and Ovarian Epithelium |
The intricate molecular details of protein-protein interactions ( PPIs ) are crucial for function . Therefore , measuring the same interacting protein pair again , we expect the same result . This work measured the similarity in the molecular details of interaction for the same and for homologous protein pairs between different experiments . All scores analyzed suggested that different experiments often find exceptions in the interfaces of similar PPIs: up to 22% of all comparisons revealed some differences even for sequence-identical pairs of proteins . The corresponding number for pairs of close homologs reached 68% . Conversely , the interfaces differed entirely for 12–29% of all comparisons . All these estimates were calculated after redundancy reduction . The magnitude of interface differences ranged from subtle to the extreme , as illustrated by a few examples . An extreme case was a change of the interacting domains between two observations of the same biological interaction . One reason for different interfaces was the number of copies of an interaction in the same complex: the probability of observing alternative binding modes increases with the number of copies . Even after removing the special cases with alternative hetero-interfaces to the same homomer , a substantial variability remained . Our results strongly support the surprising notion that there are many alternative solutions to make the intricate molecular details of PPIs crucial for function .
The study of high-resolution three-dimensional ( 3D ) structures of proteins as deposited in the PDB , the Protein Data Bank [1] , began with peptides [2] , [3] and has increasingly included larger complexes of interacting proteins [4] . These complexes , or PPIs ( Protein-Protein Interactions ) , capture the molecular details of interaction networks . The network view , in turn , has become increasingly important for , e . g . , the ranking of genes according to their probability of being causative for a particular disease [5]–[7] as needed for Genome-wide Association Studies ( GWAS ) . Despite this wealth of high-resolution interaction data , the set of interactions for which the exact molecular mechanisms are known remains immensely incomplete [8] and with it experimental and computational descriptions of binding positions and binding-induced conformational changes [9]–[11] . Nevertheless , studies of available structures have shown that related proteins have similar binding sites [12] , that permanent and transient interactions differ so substantially from each other [13] that PPI hotspots can be predicted from sequence [14] , [15] , and that we can accurately distinguish between specific and unspecific contacts [16] . Many others have addressed related tasks [16]–[23] , including even the contribution of water to the binding modes of PPIs [24] . An excellent recent work reviews various types of protein interactions [25] . We want to complement it with a quantitative analysis of the interface variability of external interactions , i . e . interactions between two protein chains coming from different genes . These typically correspond to the edges in a PPI network . The atomic structures of their interfaces often seem to cluster into particular architectures [16] , [26] and it has been suggested that they are conserved within and between organisms [26]–[31] . Many authors have also analyzed the molecular details of binding within and between their domain families [32]–[37] . For example , they found that two different SCOP domain families exhibit more than one orientation of binding about 24% of the time [33] . Beside this number , however , only few more details were given about the underlying biological variety and in particular the causes of differential interfaces . The problem we see with this approach is that members of a SCOP family only share similar 3D structures and that the observed variability in binding might simply be explained by sequence variation . In fact , the inference of similarity in structure ( homology modeling ) is much more accurate than the inference of protein-protein interactions [26] . So far no studies based on significantly sized data sets have addressed the question to which extent the interface between two different proteins is biologically conserved , i . e . excluding diversity due to sequential differences . Another challenge for the analysis of large-scale data sets have been crystal contacts and the difficulties of automated methods to correct such problems ( e . g . the PQS [38] or the PISA [39] service ) . Authors “addressed” these problems by either entirely excluding different interfaces suspecting that those originated from non-biological contacts , or by leaving it open to which extent their results might have been created by such contacts . Here , we address both issues . First , we realized that the number and quality of author-assigned biological assemblies in the PDB now suffices to enable a quantitative study like this one . For the large majority of entries , the PDB now provides biologically relevant structures from the crystallographers themselves . Similar to PQS or PISA , they describe a complex as it occurs in the living cell . At the same time , however , they are more accurate and easier to verify than de novo predictions . Therefore , we did not discard any high-resolution complex or interface therein . Secondly , put most extremely , we ask the question: if X-ray crystallographers measure the same interaction twice , do they get the same result ? The main focus is first on the variability of the interaction between identical variants of the same two proteins ( SameSeq ) . In other words , we look at external interactions corresponding to the same pair of protein sequences and estimate how often the interfaces are different ( Fig . 1A; Fig . 1B: the red arrow compares two sequence-identical interactions ) . We then extend our analysis by allowing minor sequence variations in corresponding interactors ( e . g . in the form of point mutations; SameProt ) . However , we still maintain the comparison between essentially the same proteins , because we make sure that a sequence change does not go hand-in-hand with a change of the original protein ( Fig . 1B: for the blue interface comparisons , the sequences have changed [S1/S3 vs . S2/S3] , but the original proteins [Px/Py] remained the same ) . Finally , we compared two external interactions corresponding to the same family pair , i . e . “interologs” ( Interolog ) . In a dimer-dimer comparison on this Interolog-level , corresponding interactors still had a similar 3D structure , but their sequences could be very different . ( Several authors have been using the term “interolog” [40] , [41]; it has the advantage over the term “homolog” that no evolutionary relation is implied in the definition; Fig . 1B , green: interfaces between proteins Px and Py are compared to those between Pz and Py ) .
Each node in a PPI network typically refers to a UniProt [42] entry . While UniProt stores information about proteins , its first layer of organization is genetic: every entry corresponds to a unique location on a genome . Hence , in order to find reliable structural evidence of PPI network edges , we mined the PDB [43] , [44] for interacting proteins which map to different Uniprot/Swiss-Prot [45] identifiers . We extracted such external protein-protein interactions ( i . e . interactions originating from two different genes ) in the following way: first , we downloaded all author assigned biological assemblies from the PDB . We then retained only X-ray structures that had a resolution <2 . 5 Å and mapped to at least two different UniProt/Swiss-Prot entries ( author assignment available for 99% of all such structures ) . We primarily used the PDB< = >Swiss-Prot mapping provided by the PDB and only performed the following step if this mapping was not available: we BLASTed [46] the PDB SEQRES sequence ( at least 30 residues long ) against the Swiss-Prot database , thresholding at E-Values <10-3 and requiring at least 90% of the PDB chain to be aligned . ( When we found more than one hit , we took the one with the lowest E-Value; when we had none , we discarded this complex . ) Having found those ‘interesting’ complexes , we extracted all interacting pairs of chains pointing to two different Swiss-Prot entries . At this early stage of our procedure , we only required one pair of atoms of the two chains to be closer than 0 . 6 nm ( 6 Å ) in order to consider them interacting . Note that in an earlier version of this work , we had exclusively used the PISA service [39] to obtain biologically relevant assemblies . In Section S3 . 3 in Text S1 , we give reasons why we switched to author assigned complexes , an accuracy estimation of PISA in the context of hetero-complexes and other results compiled with the PISA based data set . Having found all structures of external interactions , we annotated their interfaces . Given a hetero-dimer with chains X and Y ( X and Y come from different genes ) , we considered a pair of residues Rx and Ry as part of the interface if it contained at least one pair of atoms closer than 0 . 6 nm ( 6 Å ) or if it met all three conditions: ( i ) both residues changed their accessible surface area upon binding ( dASA: replacing the binding partner by water ) , ( ii ) Rx had no other interaction partner within 0 . 6 nm ( 6 Å ) , ( iii ) of all residues in protein chain Y that changed their accessible surface area ( ASA ) , Ry was the closest to Rx . The latter included interactions that fell slightly above the 0 . 6 nm ( 6 Å ) threshold but should still be considered interacting by their ASA change ( we present a brief analysis of the effect of including dASA in the interface annotations in Section S3 . 1 in Text S1 ) . We annotated each interface residue by two structural descriptors: dASA and d reflecting the distance ( in Ångstrøm of the closest binding residue ) . Having defined all interface residues , we removed each hetero-dimer with fewer than five interacting residues on either chain from our data set . Finally , we assigned each remaining hetero-dimer its “interface copy number” . To this end , we first determined the original complex a hetero-dimer was extracted from . Then we counted how many other hetero-dimers were also extracted from this complex and had exactly the same two SEQRES sequences as the hetero-dimer under consideration . This “interface copy number” was assigned to all these sequence-identical hetero-dimers of the complex ( Section S5 in for details ) . Overall , we applied nine different interface similarity measures to our data , covering various types of changes . They are defined in detail in Section S2 in Text S1 . The variety of these measures guaranteed that we captured as many aspects of “interaction similarity” as possible . We found significant differences between these measures , but with respect to our overall conclusions , we considered it more important to eschew obfuscation than to present all necessary details . Therefore , we used only the two most representative and intuitive measures in the main text , namely the Face Position Similarity and the L_rms . In the following , we refer to “interface” as all the residues that “touch each other” between two interacting proteins ( Fig . 1 ) , and as “face” as all the residues on one side of the interface . Also note that we always reduced hetero-dimers to common residues before comparing their interfaces . Please see Section S2 in Text S1 for details of this procedure . Before we could apply the interface similarity measures to our entire collection of external interactions , we needed to group the structures so that we could differentiate between ( and not mix ) different types of sequence divergence . This also addressed the redundancy immanent in the PDB in the form of , e . g . , overrepresented protein families . We hierarchically clustered the hetero-dimers over three levels , corresponding to increasing levels of sequence divergence ( a more technical description of the following procedure can be found in Section S1 . 1 in Text S1 ) First , we assigned two hetero-dimers to the same Level SameSeq group if they corresponded to same pair of SEQRES sequences ( Fig . 1B: we add interfaces 1 and 2 to the same Level SameSeq group; other interfaces become single member Level SameSeq groups ) . Next , we reduced the influence of over-represented proteins . This was achieved by adding Level SameSeq groups to the same Level SameProt group if they corresponded to the same pair of associated Swiss-Prot identifiers ( Fig . 1B: Level SameSeq groups S1-S3 and S2-S3 go into one Level SameProt group , S3-S4 to another ) . Clusters obtained in this way should represent the classical notion of edges and nodes in a PPI network . Our final Level Interolog addressed overrepresented families: we merged Level SameProt groups that pointed to the same pair of Pfam [48] families into one Level Interolog group ( Fig . 1B: both Level SameProt groups are merged into one Level Interolog cluster; Fig . S1 in Text S1 for a graphical illustration of the clustering ) . Without the grouping above , any distribution of pairwise interface similarities would have been highly dominated by large and well-studied complexes for which many structures are available . Avoiding this bias demanded to group PPIs differently ( Levels SameSeq to Interolog ) and also to embrace this alternative grouping when trying to infer biologically meaningful similarity distributions . While the following procedure successfully reduced the bias stemming from overrepresented sequences and sequence families , we deliberately left Level SameSeq clusters unchanged in the assumption that all binding modes are biologically meaningful and that eliminating this redundancy would remove more biology than noise ( please see Section S1 . 2 in Text S1 for a more mathematical description of the following procedures ) . The same proteins may aggregate to form a homo-oligomer and bind as such a complex to another protein . In this case , the other protein often “sees” different parts of the homomeric chain , resulting in very different external interfaces . For example , a homo-dimer with chains X1 and X2 might bind to another chain Y with two different interfaces ( Fig . 2 ) . Hence , we will have two hetero-dimers X1/Y and X2/Y with low interface similarity . As it can be argued whether both of these interfaces should be considered as one big interface or treated separately ( Discussion ) , we analyzed their influence on the distributions D-SameSeq to D-Interolog . To this end , we defined homo-oligomers in two different ways . Firstly , we used the classical notion , namely that all chains of a homomer have the same SEQRES sequence . Secondly , we introduced “structural homomers” as interacting chains from the same family . This corresponded to all complexes that look homo-oligomeric on a structural level ( low RMSD; Fig . S4B in Text S1 ) , but can differ in sequence . Consequently , when comparing two interfaces X/Y and X′/Y′ from two different PDB entries , it was checked whether or not one of the chains X′ and Y′ were part of homo-oligomers ( i . e . whether there were homomers X′/X′1/…/X′n or Y′/Y′1/…/Y′m ) and whether or not these homo-oligomers had other external interfaces with the same interaction partner as in X′/Y′ ( i . e . whether X′ had interfaces with Y′1/…/Y′m or Y′ interfaces with X′1/…/X′n ) . Having identified the set of all those sequence- or family-identical interfaces ( including X′/Y′ ) , they were compared to X/Y . Only if X/Y< = >X′/Y′ was the best match among all alternatives , the corresponding similarity was used . Otherwise , the entire comparison was discarded ( Fig . 2 . ) Eventually , the roles of X/Y and X′/Y′ are switched , and the procedure is repeated because all interfaces are compared with all others in the distributions D-SameSeq to D-Interolog . In this way asymmetries arising from only considering the oligomeric context of one side of the comparison were filtered out . We applied “structural homomerization” only in the context of D-Interolog . For the two other distributions , it would have led to comparisons of interfaces between different protein pairs , thereby invalidating the constraints of these distributions . Also note that the above definition only allowed for comparisons of interactions between two different families .
Our complete data set of external protein-protein interactions ( PPIs ) comprised 37 , 338 hetero-dimers . We grouped and filtered them on three different levels with decreasing sequence redundancy ( Methods ) . For instance , the first clustering level had 634 groups that contained sequence-identical hetero-dimers from at least two different high-resolution PDB entries . We compiled various statistics on this data set , including the distribution of cluster sizes on each clustering level , of oligomeric states , interface sizes and even of the conservation of protein function ( Section S3 . 2 in Text S1 ) . In order to capture different facets of interface similarity , we introduced and evaluated nine different similarity measures ( Sections S2 and S6 in Text S1 ) . In the following , we focused on the results from two measures ( Face Position Similarity and L_rms; Methods ) , and reported only qualitative findings for the others . The first measure ( Face Position Similarity ) was most representative for all other seven measures while the second ( L_rms ) enabled direct comparisons of our results to related work , e . g . to the CAPRI [47] experiments . For each measure , we used our clustering to calculate three different interface similarity distributions , corresponding to increasing levels of sequence divergence between interactions ( D-SameSeq to D-Interolog [Methods] ) . These distributions constitute the main result of this work and are presented in the following . They were calculated such that all proteins and families of our data set contributed equally , regardless of their respective over-representation in the PDB . Finally , we measured how the distributions change when excluding the interface variability introduced by a protein binding differently to the same homomer . We give a short summary of this after the presentation of the unmodified distributions . When two different experiments measured exactly the same external interaction ( distribution D-SameSeq; Methods ) , usually the interface between the two proteins was identical . Depending on the similarity measure , the number of largely conserved interfaces varied between 60 and 89% ( Fig . S3 and Fig . S9 in Text S1 ) . The most representative measure ( Face Position Similarity ) found the same interface in 75–79% of all cases ( Fig . 3A , D-SameSeq , rightmost bar ) . Conversely , interfaces largely differed between two observations in about 12% ( Fig . 3A , Face Position Similarity <0 . 5 ) . Other measures introduced in this work were , for instance , very sensitive to side-chain movements ( Convex Hull Overlap ) , or only roughly assessed the conservation of the interface location ( Sphere Radius Ratio ) . Taking into account two similarity measures simultaneously , small differences were observed in as many as 49% of all comparisons ( Section S6 . 3 in Text S1 ) . In contrast to our measures , the L_rms ( used by CAPRI ) returned distances in Å for the entire protein rather than for the interface alone . This perspective could capture conformational changes outside the binding regions that would be missed by other measures . L_rms found 64–69% of all “ligands” ( per definition the smaller of the two proteins in the interaction ) not to exhibit conformational changes and to bind to the larger proteins at the same positions ( RMSD <1 . 0 Å ) . Conversely , 10–14% of the interfaces differed very substantially between alternative experiments ( >9 . 0 Å ) . Considering Face Position Similarity and L_rms at the same time suggested that about 1% of all comparisons did not differ by the first but differed substantially ( >9 Å ) by L_rms ( Fig . S11 in Text S1 ) . In other words , at least 1% of all the changes between different experiments can be attributed to conformational changes outside the binding region . Another CAPRI measure , the I_rms , compared the shapes of the interface regions common to both protein pairs . We found these common regions to be very different in about 4% ( e . g . due to a rotation of one of the proteins ) and largely conserved in 80% ( Fig . S9 in Text S1 ) . We confirmed the surprising result of interface variability without sequential changes through a variety of additional analyses . The degree to which interfaces were mostly robust ( ratio between rightmost and leftmost bars in Fig . 3 ) was a function of the number of copies of a particular interaction in a complex ( i . e . , a function of the ‘interface copy number’; Methods; e . g . Fig . 1: S1/S3 observed once in C1 ) : the more copies , the relatively lower the bars on the right and the higher on the left ( Fig . S8 in Text S1 ) . But all of this also varied between families and particular complex subgroups ( Fig . S7 in Text S1 ) . For instance , MHC ( Major Histocompatibility Complex ) interactions were much less diverse than others . In fact , they contributed importantly to our overall results , although they constituted only a small fraction of all interactions . Like many before us , we also had to choose key parameters to define an interface ( Methods ) . As previous studies differed in these parameters , we also provided results for several alternative choices ( Section S3 . 1 in Text S1 ) . For instance , we included structures with a resolution >2 . 5 Å , used 4 Å instead of 6 Å as the minimal distance between two interacting residues or did not consider the change in solvent accessibility upon binding ( dASA ) when defining interface residues . These additional analyses demonstrated that some of our quantitative results depended crucially on the choice of critical parameters while the qualitative findings did not . Two hetero-dimers can differ by minor sequence variations and still correspond to the same external interaction . Comparing these slightly different pairs ( Fig . 3 , D-SameProt ) suggested considerably lower interface conservation than for the same pairs ( Fig . 3 , D-SameSeq ) : the most conserved bin ( 0 . 9–1 . 0 ) was reduced by about five percentage points for Face Position Similarity ( Fig . 3A black vs . dark gray ) and by nine percentage points for the L_rms measure ( Fig . 3B black vs . dark gray ) . These reductions were evenly distributed over the other similarity ranges . This result can be cast into two opposing views . On the one hand , it suggested that a PPI network accurately reflected the interactions because different protein variants only had a small effect on interfaces . On the other hand , there was a significant influence of small sequence changes . For instance , the range of very different interfaces ( 0 . 0–0 . 5 ) by the Face Position Similarity measure rose from 12% to 17% . In other words , about one interface pair in six differed substantially . When two experiments measured external interactions that did not correspond to the same protein pair , but to the same family pair ( D-Interolog ) , interface conservation dropped significantly by both measures ( Fig . 3 , D-Interolog , rightmost bars; Face Position Similarity: 28–36%; L_rms: 7–11% ) . For Face Position Similarity , these differences largely originated from a shift toward intermediate levels of conservation , suggesting that most changes partially preserved the approximate interface location . The Sphere Radius Ratio supported this interpretation ( Fig . S9 in Text S1 ) . Nevertheless , the interfaces with clear dissimilarity also increased from 13% ( D-SameSeq ) to almost 30% ( D-Interolog , Fig . 3 , cumulative black to light gray bin with <0 . 5 ) . This effect was stronger for L_rms: 33–40% of all comparisons were by this measure clearly dissimilar ( >9 Å; Fig . 3[B] , light gray vs . black ) . For these strong differences , the effects from conformational changes ( Fig . S5 in Text S1 ) and from local interfaces appeared to act synergistically . We hypothesized that families of interologs without alternative binding might have similar functions and that the same could be true for families with extreme binding diversity . Unfortunately , only for 18 Level Interolog clusters , interfaces were always maintained ( Face Position Similarity >0 . 9 at 100% ) , while only 17 others always used very different interfaces ( Face Position Similarity <0 . 5 at 100% ) . These numbers were too small to permit statistically significant analyses on the functional differences between those interactions . We still provided a list of those cases in Section S8 in Text S1 . The two most extreme findings of this analysis were that the Gene Ontology ( GO ) [49] term “tetrapyrrole binding” appeared over-represented in the interactions that differed , while “cytoskeletal protein binding” appeared over-represented in the interactions that did not change . With a special filter , we might remove all alternative binding of a protein to the same homomer from D-SameSeq to D-SameFam ( Methods ) . Obviously , filtering out diversity will reduce the signal of diversity observed . Nevertheless , we performed this analysis . As expected , the observed effects dropped significantly ( Table 1 ) , most extremely for D-SameSeq , i . e . for the same pairs . The differential behavior between D-SameSeq and D-Interolog might be explained by sequence divergence increasingly leading to very different interfaces for the same protein pair and ultimately to different quaternary states . Despite all the filtering for homomers , varying interfaces remain frequent and still almost one third of the change seen in interologous pairs ( D-Interolog ) is also observed between the same pairs ( D-SameSeq ) . Our finding that most interactions form identically when repeating the experiment might not be surprising . The observation that many interactions differed substantially , in contrast , appears much more counter-intuitive . Readers might attribute the difference to some mistake in the way we measure similarity or build our data sets . We addressed these concerns by expanding our analysis in many directions . On top , we analyzed ten case studies in more detail . Three are discussed in detail in the following ( Fig . 4 ) , the other seven in Section S7 in Text S1 . Our entire collection of interesting protein pairs is available in Dataset S1 .
Empirically , we found several reasons for the same two proteins to have different interfaces ( D-SameSeq ) . The simplest was merely technical: some experimental findings may not have been completely correct . We reduced this effect by excluding complexes with resolutions >2 . 5 Å , but even structures at 1 . 2 Å can contain errors [56] , [57] . Another reason was local flexibility or disorder: many proteins have local regions that are natively unstructured and these often form protein-protein interfaces [58]; such regions are difficult to track experimentally . Often , the N- and C-termini contributed to the observed interface variability . Another reason was environmental differences: despite all efforts , we could not entirely exclude artificial interfaces due to crystal packing . Different pH values could trigger conformational changes , as was the case for small ligands or other interaction partners . The presence of another protein changing the overall structure of a complex played a similar role . In all that , however , we still miss one important aspect: proteins often have evolved to interact in different ways . For such cases of biologically important alternatives , we might interpret the variety observed in a single PDB structure as an example of one protein binding to multiple copies of the same interaction partner . There were various reasons why variability in binding was higher between sequentially modified proteins than for identical proteins . The modifications that preserved the original protein ( D-SameProt ) were usually point mutations ( i . e . changes of single amino acids , e . g . by site directed mutagenesis or in the form of Nucleotide Polymorphisms [SNPs] ) . Others included protein tags at the N- or C-terminus ( e . g . to facilitate protein purification ) , post-translational modifications ( protein cleavage ) and alternative splicing . For interologs ( D-Interolog ) , finally , there was also evolutionary driven sequence divergence . As described before , however , the mere presence of insertions or deletions was not enough for low interface similarity: we reduced structures to common residues before comparing them . Thus , the increase in variability was actually the result of changes in the common parts of two structures . Using similar measures as we did , other groups [33] , [37] have found that many families interact in more than one way . Our analyses support this result . However , they also reveal that the differences in interfaces span the entire spectrum of the distribution , especially for D-Interolog . Only 18 of the 151 pairs of families completely conserved the binding modes . This finding suggests the model of a continuum of binding modes rather than clearly defined groupings , e . g . as obtained by clustering at predefined thresholds . Furthermore , in our results , about one third of the variability observed in a family-family interaction appeared to be protein-intrinsic in the sense that it was also observed between sequence-identical pairs ( D-SameSeq ) , i . e . did not originate from sequence variations ( as , e . g . for D-Interolog ) . As mentioned before , alternative interfaces might be due to the intrinsic capability of proteins to bind at different positions . This is often encountered among homo-oligomers [59] . In our case , however , it leads to a debatable scenario: a protein A can bind to multiple copies of protein B , all of which alone form a homo-oligomeric complex ( Fig . 2 ) . Do we then have to treat the various external interfaces between the same proteins as one interface , or are they indeed individual interfaces that ought to be differentiated ? We argue for the second case: first , considering the homomer a requirement for the hetero-interactions implies that by disabling the homomerization ( e . g . through site-directed mutagenesis ) , we also lose the interaction . This is not always the case [60] . Secondly , it is unclear why such a filtering should be limited to homo-complexes . Also the formation of a hetero-multimer could be a requirement for the interaction with another protein . Studying which interactions remain after disabling the potentially highly complex hetero-multimer is much beyond the currently available data . Finally , also the original publication of a complex usually describes different interfaces to the same homomer as separate interfaces . Our results raise the question whether the molecular details of protein-protein interactions ( PPIs ) are crucial for function . Protein crystallography captures static views on those molecular details along with some information about the dynamic nature of PPIs . If the details always had to be the same to guarantee function , different experiments would identify the same interfaces . We applied many reasonable ways of measuring interface similarity in order to analyze the consistency of the molecular details of protein-protein interactions between different experiments . For sequence-identical pairs of proteins , i . e . the same biological interaction , most interfaces were almost completely conserved by all measures . However , all measures also revealed an unexpected variety . Depending on how much detail we required to be similar in order to consider two experiments to yield the same results , we found 11–37% of all observations to have significant differences , and up to 10% to be completely different . One important result was that this was a significant fraction of the difference observed between homologous PPIs . Put differently , over a third of the differences in the interactions between pairs of homologous proteins are also observed between identical proteins . These numbers may challenge the notion that the maintenance of the molecular details is crucial for function . At least , our results suggest that there appear to be many alternative solutions to maintain or actually enable the intricate molecular details: change seems an extremely frequent exception for protein-protein interfaces . | The number of known protein-protein interactions ( PPIs ) grows rapidly , yet their molecular details remain largely unknown . Over the last years , structural biologists have addressed this issue with an increased output of structurally resolved hetero complexes . This wealth now enables statistically significant quantitative statements about interface properties . Here , we addressed the question how interfaces differ when observing the same proteinprotein interaction twice . A new dataset derived from the entire PDB was analyzed employing different definitions for the “same interaction” and a range of interface similarity measures . The hypothesis was that the interface between the same pair of proteins stays the same irrespectively of how often it is measured . Although the results mostly confirm this hypothesis , the surprising finding was how often it was not true: for many comparisons of interfaces , the molecular details of the interaction differed importantly , often without the slightest change of amino acids . In addition , no matter how much “special cases” were sieved out , the essential message remained: interfaces appear immensely plastic . Hand-selected sample structures largely support this view . In general , we complement a series of recent studies focusing either on family-family interactions or exploring other aspects of protein-protein complexes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"macromolecular",
"complex",
"analysis",
"biochemistry",
"protein",
"interactions",
"proteome",
"genetic",
"polymorphism",
"proteins",
"protein",
"structure",
"macromolecular",
"assemblies",
"population",
"genetics",
"biology",
"computational",
"biology",
"proteomics",
"macro... | 2012 | Alternative Protein-Protein Interfaces Are Frequent Exceptions |
Most phenotypic variation present in natural populations is under polygenic control , largely determined by genetic variation at quantitative trait loci ( QTLs ) . These genetic loci frequently interact with the environment , development , and each other , yet the importance of these interactions on the underlying genetic architecture of quantitative traits is not well characterized . To better study how epistasis and development may influence quantitative traits , we studied genetic variation in Arabidopsis glucosinolate activation using the moderately sized Bayreuth×Shahdara recombinant inbred population , in terms of number of lines . We identified QTLs for glucosinolate activation at three different developmental stages . Numerous QTLs showed developmental dependency , as well as a large epistatic network , centered on the previously cloned large-effect glucosinolate activation QTL , ESP . Analysis of Heterogeneous Inbred Families validated seven loci and all of the QTL×DPG ( days post-germination ) interactions tested , but was complicated by the extensive epistasis . A comparison of transcript accumulation data within 211 of these RILs showed an extensive overlap of gene expression QTLs for structural specifiers and their homologs with the identified glucosinolate activation loci . Finally , we were able to show that two of the QTLs are the result of whole-genome duplications of a glucosinolate activation gene cluster . These data reveal complex age-dependent regulation of structural outcomes and suggest that transcriptional regulation is associated with a significant portion of the underlying ontogenic variation and epistatic interactions in glucosinolate activation .
Most phenotypic variation present in natural populations is under polygenic control , largely determined by genetic variation at multiple quantitative trait loci ( QTLs ) , which has motivated considerable efforts to elucidate the genetic basis of these polygenic traits [1]–[3] . A complete understanding of quantitative traits necessitates identification of the underlying genes and their associated additive , dominance , and epistatic effects [2] . In addition , the underlying genetic architecture of many quantitative traits may vary across development and different environments . As such , a comprehensive description of a quantitative traits genetic architecture requires analysis in several developmental or environmental contexts to assess stability of the genetic architecture [2] , [4] , [5] . QTL mapping , which measure the association of genetic markers with phenotypic variation , is one of the most common approaches for identifying loci and epistatic interactions controlling polygenic inheritance [1] . Improved statistical models , marker technology , and genomic resources have facilitated QTL mapping experiments for a wide array of quantitative traits , ranging from development and morphology to metabolism and disease resistance [2] , [4] , [5]; However , QTL mapping experiments are often limited to a single stage in development and one or few environments . As a consequence , there is little information available to answer the question of how the underlying genetic architecture varies across developmental and environmental contexts . Accurate characterization of a quantitative trait's underlying genetic architecture is often limited by practical considerations that limit the number of progeny included in a mapping analysis . Small populations are especially problematic in the presence of epistasis between QTLs , as the pair wise comparisons required to detect these interactions rapidly exhausts the available genotypic variation , leading to an underestimation of numbers of loci and interactions , resulting in an incomplete picture of the genetic architecture [6] , [7] . One common type of epistasis occurs when a trait is controlled by one or few large effect loci and numerous modifier QTLs of smaller effect , a situation frequently observed in plant disease resistance [8]–[16] . In such systems , the effects of any modifiers are most detectable in those lines containing the appropriate allele at the large effect locus; however this reduces by half the population in which to detect these smaller effect loci , significantly reducing statistical power [5]–[7] . Thus , resolution of the underlying genetic basis of complex traits requires the analysis of large populations across different environments or developmental stages [17] , [18] . To investigate how development and epistasis can interact to control the variation in an adaptive trait , we studied the outcome of glucosinolate activation within Arabidopsis thaliana using a moderately sized recombinant inbred population . Glucosinolates are the inert storage form of a two part phytochemical defense system found throughout the Brassicaceae , where biologically active structures are catabolically produced by the enzyme myrosinase [19]–[21] ( Figure 1 ) . The particular structural outcome , as defined by the chemical structure of the end product , of glucosinolate activation plays an important role in plant defense against insect herbivory [22]–[25] , as well as the nutritional and flavor characteristics of brassicaceous crops [26] . Further , the structural outcome shows significant intraspecific diversity such that natural accessions activate a glucosinolate to either a nitrile or isothiocyanate depending upon their genotype . Thus an improved understanding of the genetic basis of variation in structural outcomes has important potential implications in evolution and ecology as well as nutrition and agriculture . Glucosinolate activation in Arabidopsis provides an excellent model for studying how development and epistasis influence quantitative traits , with a molecularly characterized biochemical pathway comprising demonstrated epistatic interactions and developmental variation . During glucosinolate activation , the myrosinase enzyme catabolically generates the unstable intermediate . The final structural outcome of subsequent rearrangement of this unstable intermediate is influenced by the presence or absence of various structural specifier proteins ( Figure 1 ) . The Epithiospecifier Protein ( ESP ) and an as yet unidentified simple nitrile structural specifier ( AtNSP ) , promote the formation of simple and epithionitriles at the expense of the default isothiocyanate rearrangement via two biochemically related yet separate rearrangements ( Figure 1 ) [22] , [23] , [27]–[31] . The Epithiospecifier Modifier ( ESM1 ) epistatically modulates ESP mediated epithionitrile and simple nitrile rearrangements ( Figure 1 ) [22] , [23] . The observation of quantitative variation influencing the developmental regulation of glucosinolate activation enabled us to explore the stability of QTLs and epistatic interactions across development [30] , [32] . We used the Bayreuth ( Bay-0 ) ×Shahdara ( Sha ) recombinant inbred lines ( RILs ) [33] to map QTLs controlling the structural outcome of glucosinolate activation in Arabidopsis . These parental accessions contain genetic variation for ESP and ESM1 and differ in developmental regulation of structural outcomes [32] . We measured structural outcomes at 30 , 35 and 42 days post germination ( DPG ) , and compared the resulting maps to assess the stability of the underlying genetic architecture across development . These DPG were chosen because day 30 represents the end of logarithmic growth in all of the lines , or Stage 1 . 10 , while day 42 is one week away from the earliest RIL flowering ( Stage 5 . 10 ) in our environmental conditions [32] , [34] , [35] . Thus , we can focus on developmental changes in what is typically considered a static rosette and is also the tissue and stages where lepidopteran insects predominate on Arabidopsis in the wild [36] . This analysis identified eleven loci and twelve pair wise epistatic interactions influencing structural outcomes , as well as eight different QTL×DPG interactions . Heterogeneous Inbred Families ( HIFs ) differing only for their genotypes at each QTL locus validated seven loci and all of the QTL×DPG interactions tested . The availability of transcript accumulation data within 211 of these RILs [37] enabled comparison of expression QTLs ( eQTLs ) for structural specifier genes and their homologs , which demonstrated collocation of eQTL clusters with eleven of the identified loci . These data reveal complex age dependent regulation of structural outcomes and suggest that transcriptional regulation is associated with a significant portion of the underlying variation , and may explain the epistatic interactions described here .
Bay-0 and Sha contain different glucosinolates due to variation at the GSL-AOP and GSL-Elong loci , such that Bay-0 has 3-hydroxypropyl glucosinolate as its main short chain aliphatic glucosinolate and Sha has but-3-enyl glucosinolate [38] . These two accessions also differ in the structures they produce following activation of these glucosinolates . Bay-0 lacks functional ESP and produces mixtures of simple nitriles and isothiocyanates , depending on the age of the plant [32] . In contrast Sha possesses a functional allele of ESP and produces mixtures of epithionitriles , simple nitriles and isothiocyanates ( Figure 1 ) . In agreement with previously published analysis , interplanted Sha parental controls had an increasing epithionitrile proportion during development for both the exogenous ( Figure 2A and C ) and endogenous glucosinolate substrates ( Table S4 ) [32] . In contrast to Sha , the Bay-0 parent showed little variation in the structural outcome of exogenous allyl glucosinolate activation between 30 , 35 and 42 DPG ( Figure 2A and C ) . The activation products for the endogenous 3-hydroxypropyl glucosinolate produced in Bay-0 could not be detected in this experiment . Thus , there is developmental variation in glucosinolate activation between the Bay-0 and Sha parental accession which allows us to investigate how the genetics of glucosinolate activation interact with plant development . We measured the structural outcome of glucosinolate activation using exogenous allyl glucosinolate in the Bay-0×Sha RILs and compared the trait distribution to the interplanted parental controls . The use of exogenous allyl glucosinolate allowed us to mask effects of variation in glucosinolate biosynthesis and accumulation . Considerable transgressive segregation was observed for simple nitrile where there were numerous RILs with values higher and lower than the Bay-0 or Sha parents ( Figure 2B ) . Further , there was transgressive segregation for epithionitrile as evidenced by the RILs with a higher value than the Sha parent ( Figure 2D ) . This suggests that alleles promoting the formation of each structure exist in both parents . This is particularly surprising in the case of epithionitrile proportions , as the Bay-0 parent can not produce epithionitriles due to a lack of functional ESP and as such might not be expected to contain genes enhancing epithionitrile formation . This population includes 212 lines producing no epithionitrile structures following activation of allyl glucosinolate ( Figure 2D ) . This is consistent with the previously observed requirement of a functional ESP allele for epithionitrile production [27] , [30] , [32] . For the sub-population of RILs with a functional allele of ESP , the distribution gradually shifted towards increased epithionitrile production from 30 DPG to 42DPG , in a manner consistent with the epithionitrile increase observed in the Sha parental controls ( Figure 2C ) . However , individual RILs showed possible differences in the both the direction of and magnitude of change in epithionitrile production from 30 to 42 DPG , suggesting that there is genetic variation in age dependent control of epithionitrile proportion from 30 to 42 DPG ( Table S4 ) . These possible differences will however require identification of the QTL and HIF validation to ensure that this is not random variation around the mean . The distribution of nitrile formation within the RILs did not show a similar ontogenic shift , but there were numerous lines that had no simple nitrile formation at 42 DPG ( Figure 2B ) . This low or undetectable amount of simple nitriles is connected to epithionitrile proportions approaching complete utilization of the glucosinolate substrate in these RILs at this DPG . This difference in ontogenic regulation supports the model of simple nitrile and epithionitrile production as being independent processes competing for the same substrate . To compare the underlying genetics controlling the biochemically distinct nitrile and epithionitrile glucosinolate activation outcomes , we estimated heritability for each glucosinolate activation structure from each measured glucosinolate ( Table 1 ) . Because we were spatially constrained to only a single measurement of each RIL at each DPG , this heritability estimate includes both RIL and RIL×DPG effects and environmental variance is not perfectly controlled . The endogenous glucosinolates are each limited to roughly one quarter of the RILs due to independent assortment of the GSL . AOP and GSL . Elong biosynthetic loci [38] . Further , the 4-methylsulfinylbutyl glucosinolate which derives from transgressive segregation at the GSL . AOP and GSL . Elong lacks a terminal alkene functional group , and can only form simple nitrile and isothiocyanate structures ( Figure 1 ) [39] , [40] . For all detectable exogenous and endogenous glucosinolates , the heritability of simple nitrile and epithionitrile proportions was approximately 50% , with the sole exception of epithionitrile from but-3-enyl glucosinolate at 69% . To identify loci controlling the diversity of structural outcomes in Bay-0 and Sha , we independently mapped QTLs for all traits at 30 , 35 , and 42 DPG . Analysis of epithionitrile proportions for exogenous allyl glucosinolate revealed nine loci ( Figure 3A , Table S3 ) , including seven novel QTLs and the previously identified ESP and ESM1 loci [22] , [23] . Three loci ( ESP , GSL . Activ . II . 13 , and ESM1 ) were detected in the RILs at all three DPG , although the additive effect of ESM1 switched direction from promoting nitrile formation at 42 DPG relative to promoting isothiocyanate formation at 30 and 35 DPG . In spite of the fact that Bay-0 lacks functional ESP , four QTLs showed a positive impact of the Bay-0 allele on epithionitrile production , which is consistent with the observed transgressive segregation in the RILs ( Figures 2D and 3A ) . We identified ten loci affecting the proportion of simple nitrile structures produced from exogenous allyl glucosinolate , including ESP , ESM1 and eight novel loci ( Figure 3B , Table S3 ) . The average allelic substitution effect of these QTLs was 27% and the median was 20% . ESP , ESM1 and six of the novel loci overlapped with epithionitrile production QTLs with all but ESP showing the same direction of allelic effect upon epithionitrile and nitrile production ( Figure 3 ) . Six of the simple nitrile proportion QTLs were significant at a single DPG , three were detected at two consecutive DPG , and one locus ( GSL . Activ . II . 13 ) was detected in all three QTL maps . Increased simple nitrile proportions were fairly evenly distributed between the Bay-0 and Sha alleles ( Figure 3B ) . One locus ( GSL . Activ . IV . 16 ) exhibited significant additive effects in opposite directions at 30 and 35 DPG ( Figure 3B ) . Isothiocyanates identified a combination of the nitrile and epithionitrile QTLs ( Figure S1 ) . This suggests that the genetic architecture underlying glucosinolate activation is much more complex than the two locus model previously assumed [22] , [41] . We proceeded to compare QTLs identified using the exogenous allyl glucosinolate to those identified with the endogenous but-3-enyl and 4-methylsulfinylbutyl glucosinolate , the two glucosinolates with the highest level of accumulation in this population . The analysis of structural outcomes of endogenous glucosinolate activation is complicated by independent assortment at the GSL . Elong and GSL . AOP biosynthetic loci , limiting each measurable endogenous glucosinolate to one quarter of these RILs [38] , [42] , [43] . QTL analysis of endogenous but-3-enyl glucosinolate activation detected three loci and nine epistatic interactions affecting simple nitrile formation , and only ESP and three epistatic interactions for epithionitrile proportions . All QTLs were consistent with those observed using the exogenous allyl glucosinolate ( Figure 3A and Table S3 ) . QTL analysis of 4-methylsulfinylbutyl glucosinolate activation identified three loci affecting simple nitrile formation , which were all identified using exogenous allyl glucosinolate ( Table S3 ) . The Sha allele of ESP increased simple nitrile formation from this glucosinolate , which lacks a terminal double bond and cannot form epithionitrile structures . In contrast , the non-functional Bay-0 allele increases simple nitrile formation from the exogenous allyl glucosinolate , possibly because this eliminates substrate competition between AtNSP and ESP . As such , the use of exogenous allyl glucosinolate provides the greatest power to independently map both structural outcomes utilizing the entire RIL population . To test how plant age in DPG altered QTL identification , we conducted an ANOVA analysis of the significant genetic loci using the data from all three assays in a single model . Each genotype is replicated within each DPG allowing for a test of marker×DPG interactions . Analysis of epithionitrile proportions using the full data set identified five marker×DPG interactions , suggesting that these loci may be involved in controlling the increase in epithionitrile formation observed from 30 to 42 DPG ( Table S3 ) . Analysis of simple nitrile proportions in the full data set also detected five significant QTL×DPG effects , suggesting age dependent regulation of simple nitrile rearrangements ( Table S3 ) . Three of the marker×DPG interactions significantly affected both simple nitrile and epithionitrile proportions , possibly as a consequence of these two rearrangements competing for the same pool of substrate or co-regulation of the two glucosinolate activation outcomes . These loci with DPG interactions provide the potential to begin understanding how ontogenic variation and genetic variation interact at the molecular level . Given the numerous QTLs controlling glucosinolate activation and the requirement of a functional allele at the ESP locus for epithionitrile production , we hypothesized that there would be significant epistasis affecting structural outcomes in this mapping population . We utilized an ANOVA to test all possible pair wise QTL interactions for significant epistasis . We identified a total of eleven different pair wise epistatic interactions for epithionitrile proportion , including the previously described ESP×ESM1 interaction [22] , [23] ( Figure 3A , Table S3 ) . Consistent with the requirement of functional ESP for epithionitrile production , all other QTLs for epithionitrile formation showed a significant epistatic interaction with ESP for at least one of these data sets ( Table S3 ) . Epistatic interactions involving ESP represent classical epistasis , where genotypes with the nonfunctional Bay-0 ESP allele produce no detectable epithionitriles , hiding the function of the interacting loci ( Figure 4A/C ) . This suggests that the QTLs epistatic to ESP may function to modulate the activity of the functional Sha allele of ESP . Interestingly , the highest levels of epithionitriles were not exclusively observed in RILs with Sha genotypes at the interacting locus . For example , the ESP×GSL . Activ . V . 18 interaction produced the highest epithionitrile proportions in lines with the Bay-0 allele at GSL . Activ . V . 18 ( Figure 4C ) . While all epithionitrile QTLs were epistatic to ESP , four epistatic interactions were detected for epithionitrile proportion that did not involve ESP ( Figure 3B and 4D and Table S3 ) . These interactions were examples of quantitative epistasis , where the effects of one locus was modified by the other locus in a quantitative manner , rather than the absolute dependence of one locus on the other locus as exhibited by interactions involving ESP ( Figure 4D versus 4A/C ) . Additionally , these epistatic interactions formed networks such that GSL . Activ . II . 13 , GSL . Activ . II . 61 and ESM1 showed all possible pair wise epistatic interactions with each other as well as with ESP ( Figure 3B and Figure 4B ) . Likewise , a similar network involves GSL . Activ . I . 16×GSL . Activ . IV . 16 and ESP ( Figure 3 ) . This suggests that complex epistasis may begin to identify underlying regulatory or protein interaction networks controlling the structural outcome of glucosinolate activation . Simple nitrile proportion from exogenous allyl glucosinolate identified fewer epistatic interactions than epithionitrile formation with most interactions also involving the ESP locus . The ESP×ESM1 epistatic interaction affected both simple nitrile and epithionitrile proportion , but with different effects on each structural proportion ( Figure 4 A and B ) . The lower number of epistatic interactions for simple nitrile formation is partly explained by the lower variation present for simple nitrile formation within these RILs ( Figure 2A/B ) . To provide additional support for the identified GSL . Activ QTLs , we obtained HIFs that contain appropriate variation for seven of the loci detected in this study ( Table S1 ) . For the GSL . Activ . II . 13 and GSL . Activ . III . 64 loci , the available HIFs only contained a non-functional ESP , thus we were unable to test the effects of these loci on epithionitrile production . We assayed glucosinolate activation in each HIF line at 24 and 38 DPG , to confirm the QTLs and any interaction between the QTLs and DPG . The HIFs confirmed seven of the GSL . Activ QTLs . This included four loci for epithionitrile formation , two for simple nitrile production and several for the total production of nitriles or isothiocyanates ( Figure 5 , Table 2 , Table S5 ) . HIF-241 and HIF-425 vary for ESP and confirm that a functional Sha allele is necessary for epithionitrile production ( Figure 5A , Table S5 ) . Interestingly , the efficiency of epithionitrile formation significantly differs between these two HIFs ( P = 0 . 047 for HIF×ESP genotype and P = 0 . 048 for HIF×ESP genotype×DPG ) , confirming the presence of background ESP modifiers . The level of validation observed in the HIF analysis is strongly supportive of the QTL mapping results , as each HIF genotype was only analyzed in six-fold replication per DPG whereas each marker genotype in the RIL study was analyzed in roughly 200-fold replication , lending more power to the RIL analysis . The original RIL QTL analysis did not replicate each line within each DPG , and as such , we designed the HIF analysis to confirm that these QTLs do interact with the plant age in DPG . All four confirmed epithionitrile proportion QTLs exhibited significant genotype×DPG effects ( Figure 5 , Table 2 and Table S5 ) . For example , HIF-157 which varies for GSL . Activ . V . 18 , showed a significant difference in epithionitrile proportions between alleles at 24 but not 38 DPG ( Figure 5B and Table S5 ) . In contrast , the locus GSL . Activ . V . 48 in HIF-350 showed no difference in epithionitrile proportions between alleles at 24 DPG , but a significant effect at 38 DPG in agreement with the QTL prediction from the RIL analysis ( Figures 3A , 5C and Table S5 ) . These results confirm that we have identified ontogenic dependent QTLs in our study . QTL mapping analysis of epithionitrile proportion consistently produced a large peak in the LOD plot at the ESP locus but there was also frequently a small shoulder ( Figure 6A ) . This could be explained by residual significance from the large-effect ESP locus or suggest the presence of a tightly linked QTL in this region . Due to tight genetic linkage with ESP , we did not include this putative locus in our statistical models . We did however identify two HIFs in this region , HIF149 with functional ESP and HIF107 with non-functional ESP . These HIFs allow us to test for the existence of this additional locus as well as its potential dependency upon ESP ( Table S1 ) . Analysis of structural outcomes in HIF149 confirmed the existence of an additional QTL teleomeric of ESP on Arabidopsis chromosome I . This QTL , GSL . Activ . I . 69 , affects both simple nitrile and epithionitrile proportions ( Table S5 and Figure 6B/C ) . GSL . Activ . I . 69 also displays an age dependent effect on simple nitrile proportion , but not epithionitrile proportion ( Table S5 and Figure 6B/C ) . Interestingly , HIF107 did not identify a significant QTL , suggesting that GSL . Activ . I . 69 is epistatic to ESP ( Table S5 ) . This HIF analysis of a QTL shoulder suggests there may be even further additional QTLs for glucosinolate activation in this RIL population . The TGG1 and TGG2 myrosinases , ESP , and ESM1 are the primary genes with a demonstrated role in controlling glucosinolate activation and associated structural outcomes within the Arabidopsis rosette [22]–[25] , [27] , [30] , [31] . However , these genes and the MBPs form gene families in Arabidopsis and some of these uncharacterized homologs may determine the genetic variation observed in the structural outcome of glucosinolate activation . To identify any potential expression level polymorphisms in these four gene families that may control the glucosinolate activation QTLs , we identified eQTLs for the full list of potential glucosinolate activation homologs ( Table S2 ) [37] . We first obtained the estimated heritability for transcripts encoding these potential glucosinolate activation genes [37] . The average transcript heritability for the 30 measurable probe sets was 63 . 0% , which is significantly higher than the genome wide average transcript heritability ( t-test , P = 0 . 045 ) . Further , the glucosinolate activation genes had more eQTLs , both cis and trans than the average Arabidopsis transcript . This excess was most dramatic with trans-acting eQTLs , with an average of 3 . 1 trans-eQTLs detected for each glucosinolate activation gene in comparison to the average Arabidopsis transcript , with 1 . 5 trans-eQTLs in this population [37] . These results agree with previous studies showing the transcripts for glucosinolate biosynthetic genes had higher heritability and variance than the average Arabidopsis transcript [38] , [44] . The previous analysis of eQTLs within the Bay-0×Sha RIL population was conducted at 35 DPG in the same growth chamber allowing their direct comparison [37] . The eQTLs controlling transcript accumulation of the glucosinolate activation genes revealed eQTL clusters collocating with ten of the eleven structural proportion QTLs ( Figure 7B ) . The eQTL clusters partitioned into both cis and trans-eQTL clusters . The two cis-eQTL clusters are associated with the genomic regions around ESP and ESM1 , which contain several ESP , ESM1 , and MBP homologues but do not overlap with known trans-eQTL hotspots ( Figure 7A and B [see the black arrows] ) [37] . These regions appear to be the result of two separate whole-genome duplications that copied an original region containing the ancestral genes for ESP , ESM1 and the MBPs ( Figure 7A ) . These consecutive duplications generated four genomic regions that contain additional non-glucosinolate genes whose paralogs also have a conserved order ( Figure 7A ) . In these two cis-eQTL clusters , we found six and eight cis-eQTLs for putative glucosinolate activation genes in the ESP and ESM1 regions respectively ( Figure 7 ) . While variation in ESP and ESM1 have been shown to control glucosinolate activation phenotypes , these local cis-eQTL clusters suggest that additional genes at each locus could contribute to the effects on the structural outcome of glucosinolate activation . In addition to cis-eQTL clusters , there were several GSL . Activ loci that co-located with clusters of trans-eQTLs suggesting that these may be regulatory loci with a measurable effect on gene expression and structural outcomes . In particular , GSL . Activ . II . 13 collocated with a large trans-eQTL cluster but not with the genomic position of any of the identified glucosinolate activation gene homologs ( Figure 7B – double black arrows highlight homologue clusters ) . Interestingly , this locus also collocates with a large trans-eQTL hotspot controls ∼1 , 200 genes suggesting that it may contain a polymorphism in a pleiotropic developmental regulator [37] . Genetic variation at the GSL . Activ . II . 13 locus alters the expression of all four ESM1 homologues , three of four ESP homologues , four MBPs , and two myrosinases . A closely linked locus , GSL . Activ . II . 42 , also collocates with a large trans-eQTL cluster controlling two myrosinases , eight of twenty potentially interacting genes and ∼2 , 500 other genes ( Figure 7B ) [37] . Thus , these two loci show that global regulatory loci can have a measurable consequence on glucosinolate activation . Interestingly , the GSL . Activ . I . 69 locus identified in the HIF analysis is associated with eQTLs controlling transcript accumulation for TGG2 , ESP , ESM1 and an MBP ( At3g16470 ) , suggesting that the GSL . Activ . I . 69 may also be a regulatory locus . However , it does not co-locate with a previously identified global trans-eQTL hotspot suggesting that it may be more specific to glucosinolate activation [37] .
The Bay-0 and Sha parents had previous been shown to have different ontogenic control of glucosinolate activation [32] . Our analysis of epithionitrile proportions revealed several differences between 30 , 35 and 42 DPG , and analysis of the full data set identified seven significant QTL×DPG interactions ( Figure 3A and Table S3 ) . Further , direct assessment of glucosinolate activation at two DPG in the HIFs confirmed these QTL×DPG interactions ( Figure 5 , Table 2 , S3 and S5 ) . This suggests that these loci are responsible for the age dependent regulation of epithionitrile production , and may act to regulate ESP expression . Similar levels of ontogenic QTL dependency were identified for simple nitrile production in spite of the absence of ontogenic variation between Bay-0 and Sha for simple nitrile production ( Figures 2A/B and 3B and Table S3 ) . Thus , transgressive segregation can also occur for interaction terms in QTL analysis . Extensive epistatic interactions controlling glucosinolate structural outcomes were detected , many involving ESP that may represent classical epistasis where the phenotypic effects of the second locus on epithionitrile production are fully masked in the absence of functional ESP ( Figures 3A and 4A/C , and Table S3 ) . These interactions formed networks wherein all possible pair wise interactions were detected ( Figure 3A ) . One such epistatic network involved the known structural genes ESP and ESM1 , with GSL . Activ . II . 13 and GSL . Activ . II . 42 . Interestingly , GSL . Activ . II . 13 and GSL . Activ . II . 42 appear to be trans regulatory loci that control the expression of ESP and ESM1 , several other putative glucosinolate activation genes and thousands of other genes suggesting ( Figure 7B ) [37] . This suggests that this epistatic network is a combination of variation in two master regulatory loci , as well as variation in the genes that they regulate , ESP and ESM1 . Further evidence for trans regulation underlying epistasis comes from the ESP×GSL . Activ . V . 18 interaction , where GSL . Activ . V . 18 collocates with an ESP trans-eQTL ( Figure 4C and 7B ) . The connection of regulators and their regulated genes in a quantitative epistatic network suggests that it may be possible to use quantitative epistasis in a manner similar to Mendelian epistasis to begin defining molecular networks and their influence upon the final phenotype . Further work will be required to validate if these regulators are directly or indirectly affecting transcript accumulation of ESP , ESM1 and the other homologs . One complication that occurs from the observed level of epistasis is a diminished statistical power to identify loci . As such , the 400 lines used may underestimate the true extent of epistasis for the structural outcome of glucosinolate activation . As a consequence of this extensive epistasis , we hesitate to eliminate QTLs not confirmed by HIF analysis as candidate loci . The genetic background of each HIF consists of a random mixture of fixed Bay-0 and Sha genotypes at all regions outside of the focal locus , and it is therefore likely that some of the available HIFs have unfavorable genotypes at interacting loci . Of the four QTLs with multiple HIFs available , three were confirmed in some backgrounds and not in others , supporting a genetic background effect ( Table 2 ) . In particular , analysis of simple nitrile proportion appears complicated by ESP genotype . Both HIFs with significant effects on simple nitrile proportions lacked ESP activity , suggesting that the ability to detect small effects on simple nitrile formation can be negatively impacted as a consequence of reduced flux in the presence of the competing epithionitrile rearrangement . Analysis of the genomic regions underlying ESP and ESM1 revealed two distinct and tightly linked clusters of structural specifiers and myrosinase interacting proteins at each locus ( Figure 7A ) . These linked clusters appear to be the product of an ancestral locus , which contained ESP , ESM1 , and an MBP and underwent a tandem duplication followed by segmental duplication to generate the ESP and ESM1 loci , with the subsequent loss of some paralogs . These four genomic regions contain the majority of the ESP , ESM1 and MBP homologs but differ in their specific composition . Further , a number of cis-eQTL were detected for these genes , and the associated QTLs have divergent effects on the structural outcome of glucosinolate activation [22] , [23] . The presence of a large number of duplicated homologs suggests that these QTL may be complex loci with numerous tightly linked polymorphisms contributing to the observed phenotypic effects . Support for this idea comes from the observation that while the ESP and ESM1 proteins were shown to explain most of the effects of their respective QTLs , complementation of both QTLs did not completely recapitulate the phenotypes associated with each QTL [23] , [42] . The association of genomic duplications with glucosinolate activation QTLs suggests that such duplications may facilitate quantitative genetic variation by creating duplicate genes . The duplicate genes can then undergo genetic sub-functionalization such that the genes have differential functions across natural genotypes [44] , [49]–[51] . This role of genome duplications and QTL association has been previously seen in polyploid plants but not characterized in diploids [52] , [53] . This relationship between genome duplications and QTLs requires the analysis of more traits and cloning of more loci to establish the generality of this connection . Nine of twelve QTLs affected both simple nitrile and epithionitrile proportions formed from allyl glucosinolate ( Figure 3 and 6 , and Table S3 ) . Additionally , four epistatic interactions were detected in both the simple nitrile and epithionitrile structural outcome data ( Figure 3 and 4 and Table S3 ) . While there was considerable overlap in the QTLs for the two distinct structural outcomes , there was not complete correspondence in the directionality of their effects . The majority of the QTLs altered the accumulation of nitrile and epithionitrile in the same direction , but the ESP locus had opposite effects on simple nitrile and epithionitrile proportions ( Figure 3 ) . In those cases where the same pair wise interactions via the ESP locus affected both simple nitrile and epithionitrile formation , they affected each proportion differently , supporting the concept of two independent rearrangements competing for the same substrate . This suggests that simple nitrile and epithionitrile formation share regulatory loci , but that there are likely separate proteins producing nitrile and epithionitrile structures from allyl glucosinolate . This is in agreement with previous observations suggesting the presence of an unidentified simple nitrile forming enzyme in Arabidopsis [23] , [25] , [32] . This study shows that there is considerable natural genetic variation controlling the age dependent regulation of structural outcomes in Arabidopsis [32] . Elucidating the basis of this regulation is necessary to obtain a better understanding of the evolution and ecological significance of developmental trajectories in this important plant defense system . Further , we describe epistatic networks that appear to link regulatory loci with the genes that they regulate . Future analyses will be required to test if quantitative epistasis can be used to generate networks in a fashion similar to Mendelian epistasis but this has potential applications in most species . Finally , the potential for whole genome duplications to be associated with multiple QTLs for the same trait may help to enhance the rate at which additional QTL can be cloned . Once one QTL is cloned for a given trait , it may immediately suggest candidate genes for QTL in genomic regions that share an ancestry through whole genome duplications . The glucosinolate system is a useful model system for quantitative genetics to begin addressing these fundamental issues in quantitative genetics , ecology and evolution .
The population of 411 Bay-0×Sha RILs [33] were chosen for QTL mapping analysis of the structural outcome of glucosinolate activation . The parents of this population differ in their glucosinolate profile and content , as well as in the structures formed upon glucosinolate activation and the developmental regulation of the structural outcome following glucosinolate activation [32] , [38] . A subset of 212 lines from this population have also been analyzed for variation in gene expression [37] , enabling comparison of gene expression to phenotypic variation . Finally , there are available HIFs , pairs of near isogenic lines fixed for alternate alleles at a single locus in an otherwise identical recombinant inbred background , which offer the opportunity to validate some of the QTLs detected in this study [54] . Seeds were imbibed and cold stratified at 4 degrees for three days to break dormancy . All seeds were sown directly onto Premier ProMix B potting soil ( Premier Brands , Inc . , Red Hill , Pennsylvania ) in 36-cell ( approximately 125 cm3 soil per cell ) flats , and grown in controlled environment chambers at 20°C with 8 h light at 100–120 µEi . Each flat contained one Bay-0 and Sha parents . All plants were free of insect pests by visual inspection . The population was grown three independent times to independently phenotype the structural outcome of glucosinolate activation at 30 , 35 , and 42 days DPG . These DPG were chosen because day 30 represents the attainment of Stage 1 . 10 or 10 mature leaves per plant for both the parents and all RILs . Further , day 42 is one week away from the earliest RIL flowering ( Stage 5 . 10 ) in our environmental conditions [32] , [34] , [35] . Thus , this range of DPG allows us to focus on developmental changes in what is typically considered a static rosette rather than query larger ontogenic shifts such as leaving logarithmic growth or the flowering transition . Within a two hour time frame centering on dawn , the rosette leaves from each and every RIL at each DPG were harvested and phenotyped for the structural outcome of glucosinolate activation . Previous work had shown that glucosinolate activation is regulated by rosette age and not the age of individual leaves within a rosette [32] . The structural outcome of glucosinolate activation was assayed using a modified version of the previously published protocol [22] , [23] . Briefly , the three fully expanded rosette leaves from a single plant were harvested and crushed in an 8 mL reaction vial containing 1 mL of 100 mM MES buffer at pH 6 . 0 and 0 . 4 µmol of allyl glucosinolate . The three leaves were consistently the first , fourth and seventh fully expanded leaf to provide a sampling of different ages . Previous work had shown that these three leaf ages had similar glucosinolate activation that was determined by the rosette age and not the leaf age [32] . This allows us to focus on rosette age rather than leaf age although the two may be intricately linked in some fashion . Exogenous allyl glucosinolate was added to enable comparisons of structural outcomes using a common substrate for all RILs despite the segregating biosynthetic variation [38] . Further , the allyl glucosinolate allows us to measure all three potential glucosinolate activation endpoints , isothiocyanate , epithionitrile or simple nitrile , whereas half the RILs do not have this capacity due to the lack of alkenyl glucosinolates [32] , [38] . Upon complete tissue homogenization the reaction vial was capped and incubated for five minutes . The reaction was stopped and glucosinolate activation products extracted with 4 mL of dichloromethane . The organic phase was removed , dried and concentrated to 200 µL for gas chromatography ( GC ) analysis using an Agilent HP 5890 with a flame ionization detector [22] . Peak identities were confirmed using a GC-mass spectral detector ( Agilent HP 6890 with an Agilent 5973N MSD ) , by comparison with published mass spectra [55] . Quantification was conducted using published response factors that were corrected using propyl isothiocyanate standards as previously described [22] . Structural outcomes are reported as the percent of simple nitrile , epithionitrile , or isothiocyanate products for a particular glucosinolate . For instance , the percent simple nitrile for allyl glucosinolate is defined as [allyl simple nitrile] / [allyl simple nitrile+allyl epithionitrile+allyl isothiocyanate] . Proper chemical names for this equation are allyl simple nitrile is 3-butenyl nitrile and allyl epithionitrile is 2 , 4-epithiobutyl nitrile . Each structural outcome for each glucosinolate was measured using a similar equation . By dividing the absolute amount of a particular structure by the sum of all possible products , the effects of myrosinase activity and differences in biosynthesis and accumulation of the endogenous substrates are cancelled , since they affect both the numerator and denominator equally . This assay is not a quantitative measure of total myrosinase activity because it reaches saturation for some samples . We obtained genotypes and genetic map information for the Bay-0×Sha RIL population from the Arabidopsis Biological Resource Center ( ABRC; www . arabidopsis . org ) [33] . To maximize our ability to detect QTLs , we utilized the data from each DPG experiment separately and as a combined data set . For each RIL , the proportion of each activation structure for each glucosinolate were independently used for QTL mapping within Windows QTL Cartographer v2 . 5 [56]–[58] . Although the proportions of glucosinolate activation structures obtained from a given substrate are not mathematically independent of one another , the simple nitrile and epithionitrile rearrangements can be separately measured for allyl glucosinolate , allowing simultaneous assessment of these partially independent processes [32] . Composite interval mapping ( CIM ) was implemented using Zmap ( Model 6 ) with a 10 cM window and an interval mapping increment of 2 cM . The declaration of statistically significant QTL is based on permutation derived empirical thresholds using 1 , 000 permutations for each trait mapped [59] , [60] . The Eqtl module of QTL Cartographer was used to automatically identify the location of each significant QTL for each trait [58] . To further test each QTL identified and query for potential epistasis , we conducted an ANOVA for the proportion of each glucosinolate activation structure . The markers most closely linked to each significant main-effect QTL were used as main effect cofactors . An automated SAS script then tested all main effects and all possible pair wise interactions between main-effect loci . Significance values were corrected for multiple testing within a model using false discovery rate adjustment within the automated script . The script returned all significance values as well as QTL main-effect estimates in terms of allelic substitution values ( Table S3 ) . In addition , the combined data were used to estimate the heritability of the different structural outcomes of glucosinolate activation . This was conducted using the general linear model procedure within SAS where broad sense heritability was defined as σg/σp ( Table 1 ) , where σg is the estimated genetic variance for the structural proportion phenotypes among different genotypes in these RILs , and σp is the estimated phenotypic variance [3] . To confirm the identified QTLs in this study we obtained sixteen HIFs , corresponding to nine of the loci detected in this experiment , from INRA ( http://dbsgap . versailles . inra . fr/portail ) ( Table S1 ) . There were no HIFs available to test GSL . Activ . IV . 16 and GSL . Activ . II . 61 . Within each HIF , only the genotypes in the region of one QTL differ while the rest of the genome is a random homozygous mixture of Bay-0 and Sha genotypes . HIFs with the functional Sha allele at ESP can be used to test QTLs controlling both epithionitrile proportion and simple nitrile proportion . For most QTLs there was a HIF available with functional ESP , except for GSL . Activ . II . 13 and GSL . Activ . III . 64 . To test simple nitrile proportion QTLs for dependence on ESP genotype , separate HIFs with the functional Sha and non-functional Bay-0 alleles of ESP were chosen when possible . Each HIF was planted with twelve independent biological replicates per allele per HIF . These were planted and grown under identical conditions as described above for the RIL population . For each HIF , six replicates of each genotype were assayed for structural outcomes at 24 DPG and six were assayed at 38 DPG to validate each QTL and survey for age dependence . These time points were chosen such that there was a sufficient developmental time difference to detect genotype×DPG effects but before epithionitrile formation reached saturation at the later time points . Due to poor germination , HIF191 , 244 , 338 , and 364 were only assayed at 38 DPG . The data for each class of glucosinolate activation product were analyzed for the effects of genotype , DPG , and DPG×genotype within each HIF using the general linear model procedure in SAS . Given that each HIF is a separate and independent test , we did not correct for multiple testing within these models . We also directly compared HIF241 to HIF425 , two independent HIFs differing at the ESP QTL , to assess background dependent effects upon ESP . We used previously published sequence data to identify the known myrosinases and structural specifier genes [22]–[24] , [61] . We utilized protein sequence data to identify Arabidopsis homologs of each major glucosinolate activation gene , myrosinase , ESP , ESM1 , and Myrosinase Binding Protein 1 ( MBP1 ) and MBP2 . For all four gene families we had two criteria to define a gene as potentially associated with glucosinolate activation . First , each included gene had to be similar to known genes based on a BLASTP score of at least e−45 . Secondly , we further restricted this list to genes that were phylogenetically limited to Arabidopsis when protein sequences from the poplar , grape and rice plant genomes were included [62] , [63] . This assumes that genes with more similar , non-cruciferous homologues are unlikely to be involved in glucosinolate activation as this system is not present in poplar , grape or rice ( Table S2 ) . Heritability , eQTL position , eQTL effect and transcript accumulation values were obtained from a previously published analysis of the Bay-0×Sha population [37] . Because these global transcription studies were conducted in the same mapping population grown under the same conditions and in the same growth chambers , it was possible to directly compare the gene expression and structural outcome data . There are no new accession numbers associated with this dataset . The microarray data set used in this study has been deposited at EBI ArrayExpress ( http://www . ebi . ac . uk/arrayexpress/ ) under numbers E-TABM-126 and E-TABM-224 . | A principal interest in biology is to understand how natural genetic variation translates into phenotypic variation . A key component of this connection is how the genetic variation interacts with other sources of variation , such as environment ( G×E ) , development ( G×D ) , or other genetic loci ( G×G or epistasis ) . To analyze the molecular underpinnings of these quantitative genetics interaction terms , we investigated the genetic architecture of an adaptive trait , glucosinolate activation , in Arabidopsis thaliana during the development of what is considered a static mature rosette . Variation in glucosinolate activation was principally controlled by epistatic and G×D interactions . Epistatic interactions identified both Mendelian epistasis , where regulatory loci controlled enzymatic loci , and quantitative interactions between regulatory loci . G×D appeared to involve master regulatory loci as determined by trans-eQTL hotspot analysis . Finally , two common glucosinolate activation QTLs appear to have evolved via gene loss and sub-functionalization following quadruplication of an ancestral genomic fragment , potentially by two whole-genome duplications . Thus , genomic duplication events may facilitate the formation of quantitative genetic variation . This study provides insights into the molecular basis of the link between genetic and phenotypic variation in a potentially adaptive trait . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"biology/plant-biotic",
"interactions",
"ecology/plant-environment",
"interactions",
"biochemistry",
"plant",
"biology/plant",
"genomes",
"and",
"evolution",
"evolutionary",
"biology/plant",
"genomes",
"and",
"evolution",
"developmental",
"biology/plant",
"growth",
"an... | 2008 | Genetic Networks Controlling Structural Outcome of Glucosinolate Activation across Development |
Natural protein sequences contain a record of their history . A common constraint in a given protein family is the ability to fold to specific structures , and it has been shown possible to infer the main native ensemble by analyzing covariations in extant sequences . Still , many natural proteins that fold into the same structural topology show different stabilization energies , and these are often related to their physiological behavior . We propose a description for the energetic variation given by sequence modifications in repeat proteins , systems for which the overall problem is simplified by their inherent symmetry . We explicitly account for single amino acid and pair-wise interactions and treat higher order correlations with a single term . We show that the resulting evolutionary field can be interpreted with structural detail . We trace the variations in the energetic scores of natural proteins and relate them to their experimental characterization . The resulting energetic evolutionary field allows the prediction of the folding free energy change for several mutants , and can be used to generate synthetic sequences that are statistically indistinguishable from the natural counterparts .
To study the co-occurrence of mutations in a sequence alignment of a particular protein family , [39] proposed a Hamiltonian or energy expression which resembles a Potts model: E ( s → ) = - ∑ i = 1 L h i ( a i ) + ∑ i = 1 L ∑ j = i L J i j ( a i , b j ) ( 1 ) where the set of {hi ( ai ) } parameters , one for each amino acid in each position , accounts for a local propensity of having a specific residue on a particular site of the protein , and the set of {Jij ( ai , bj ) } indicates the strength of the ‘evolutionary’ interaction between each possible amino acid in every pair of positions along the protein . There are q = 21 possible values of ai and bj , one for each amino acid and one for the gaps included on the multiple sequence alignments . This expression is evaluated on a particular sequence on an alignment , and the summations go over the L columns of the alignment . A sequence is more favorable or more energetic if it gets lower values of E ( s → ) . It can be expected that the population of sequences follows a Boltzmann distribution P ( s → ) = 1 Z e - E ( s → ) [40] . The parameters are thus fitted to reproduce the frequencies of occurrence of each amino acid in each position ( fi ( ai ) ) and the joint frequencies of amino acids ( fij ( ai , bj ) ) in an alignment of natural sequences used as input: f i ( a i ) = ∑ a k , k ≠ i P ( s → ) ( 2 ) f i j ( a i , b j ) = ∑ a k , k ≠ i , j P ( s → ) ( 3 ) Nevertheless , for repeat proteins there is another feature we want to capture with an evolutionary energy: the high identity of amino acids constituting consecutive repeats , arisen by the repetitiveness of these families and probably a signature of their evolutionary mechanisms ( Fig 2 ) . Therefore , we propose the following model for repeat proteins: E ( s → ) = - ∑ i = 1 L h i ( a i ) + ∑ i = 1 L ∑ j = i L J i j ( a i , b j ) - λ I d ( s → ) ( 4 ) This expression is designed to be applied in sequences constituted by two repeats . λId is a parameter that aims at reproducing the probabilities of the percentage of identity ( %Id ) between consecutive repeats in natural proteins ( pid ) . Basically , it accounts for higher order correlations not captured by the pairwise terms . For a given sequence we calculate the %Id of the adjacent repeats and sum the parameter λId corresponding to that %Id value . When the correct parameters are obtained , this equation can be used to produce an ensemble of sequences consistent with the constraints ( fi ( ai ) , fij ( ai , bj ) and pid ) . We work with pairs of repeats as it is the minimum unit that includes the interaction between repeats and the possibility of measure sequence identity between consecutive repeats . In the following section we will show the convergence of the method and the relevant information that can be obtained from it . For further details about the procedure to assign values to the parameters , please refer to Methods section .
We construct an alignment of pairs of repeats for each family: ANK ( PFAM id PF00023 , and final alignment of 20513 sequences of L = 66 residues each ) , TPR ( PFAM id PF00515 , and final alignment of 10020 sequences of L = 68 residues each ) and LRR ( PFAM id PF13516 , and final alignment of 18839 sequences of L = 48 residues each ) . See Methods for further details of construction . We measure fi ( ai ) , fij ( ai , bj ) and pid . Using a gradient descent procedure we obtain a set of parameters in eq 4 which are able to reproduce fi ( ai ) , fij ( ai , bj ) and pid . In principle , the number of parameters is large: Lq hi parameters , ( L q ) 2 - L q 2 Jij parameters and L 2 + 1 λId . For example , for pairs of ANK repeats this means 1386 hi , 959805 Jij and 34 λId . To reduce the number of free parameters to fit we use a L1-regularization which fixes to zero those parameters which do not contribute significantly to fit the frequencies . This regularization allows us to set to exactly zero between 85 and 91% of the Jij parameters which , when they are free to vary , only reach small values ( S3 Fig ) . We bound the maximum error permitted in the frequency estimations to 0 . 02 . Refer to Methods for more details . In the three families studied , the parameters obtained allow us to generate ensembles of sequences which reproduce natural fi ( ai ) , fij ( ai , bj ) and pid ( Fig 2A ) . Notice that most frequencies are fitted with an error of an order of magnitude lower than the maximum bound imposed ( S2 Fig ) . The pid distributions are also very well reproduced ( Fig 2B ) . Not only the general shape , but also the populated long tail for highly similar repeats . It is not possible to obtain the same distribution only by fitting amino acid frequencies fi ( ai ) and fij ( ai , bj ) , it is mandatory to explicitly include the pid by including the parameters λId ( S1 Fig ) , suggesting that higher order correlations must be accounted for describing these systems . Once the set of parameters {hi ( ai ) , Jij ( ai , bj ) , λId} is obtained , it can be used to score any sequence of L amino acids via eq 4 . In this section we test if this measure is capable of distinguishing polypeptides that fold in a three dimensional structure similar to members of the repeat protein family from those that do not . We calculate the distribution of energies of different sets of sequences ( Fig 3 ) . The ensembles of natural sequences of each protein family used to learn the parameters have a unimodal distribution of energies centered around -100 ( Fig 3 , red lines ) . These distributions are clearly differentiated from the energies of random chains of residues ( Fig 3 , yellow lines ) , which constitute a basic negative control for our model . For a positive control we evaluate designed proteins which have been experimentally synthesized . For the ANK family , we consider the library of repeat sequences built by Plückthun’s laboratory [13] ( green lines , Fig 3A ) . This library was constructed by fixing on each repeat 26 positions out of 33 to the most frequent residue in the multiple sequence alignment . This resulted in a set of sequences that have small variations with respect to the ANK consensus ( the sequence with the most frequent amino acid in each position ) . In our expression , they score a very low energy distribution , overlapping with the most negative tail of the distribution of natural sequences . It is notable that consensus designed ANK have been shown experimentally to be extremely stable . For the TPR family , consensus designed was done by Regan’s laboratory [11 , 12] . All pairs of repeats synthesized have the same amino acid sequence , and it’s energy score is indicated by a green full square in Fig 3B . Again , the designed sequence matches values at the most left side of the energy distribution of natural sequences , and coincidentally reports high folding stability . From it , other variants with few point mutations to improve binding to a specific ligand have been synthesized . As shown in empty green squares [41] and diamonds [42] in Fig 3B , they have higher energy , but still in the left most side of natural sequences distribution . Recently , a different design strategy was done [43] . Based on a non-repetitive protein , but similar to TPR fold , they put togheter various repetitions of the fold , using TPR loops to link them . They obtained a three-repeats protein whose pair of repeats energy are represented on triangles on Fig 3B . This time , they match natural sequences distribution in higher values . Finally , for the LRR family we contrast with the library of proteins designed by Plückthun’s group based on the consensus sequence [14] . The repetition they considered has 57 amino acids , which includes two types of repeats , one of 28 residues and the other one of 29 . As the repeat we are using for LRR is 24 residues long , we aligned both definitions and evaluated the library removing the amino acids not matching our definition . Again , their scores form a narrow distribution , but this time it is not placed on the most favorable side of the natural sequences distribution ( Fig 3C ) . Coincidentally , selected species studied do not show such a high folding stability as the ANK library did . With these parameters , we are able to generate an ensemble of sequences which are in agreement with the constraints used , via a Monte Carlo simulation ( see Methods ) . The distribution of energies of these simulated sequences matches the natural sequences energies distribution with remarkable accuracy . Moreover , we randomly choose 100 sequences from the natural ensemble and 100 sequences from the simulated one , perform a Smith-Waterman pairwise alignment all against all , calculate the pair similarity using BLOSUM62 matrix and used it as a distance method to plot a dendogram of the sequences ( S4 Fig ) . Both species appear interspersed , showing that it is not possible to distinguish a natural sequence from a constructed one . Also , we tested familiarity to the ANK family as defined in [44] and found overlapping distributions for both species ( S5 Fig ) . Therefore , simulated sequences represent possible variants to natural repeats . The wide distribution of natural proteins suggests that it should be possible to engineer sequences with more variable repeats , more dissimilar among neighbors and to the consensus than the ones published up to date . Are there any invariant properties shared by low energy sequences ? Given that repeat-proteins may evolve by other mechanisms besides point substitutions , we analyze if low energy sequences are constituted by highly similar repeats and if they are close to consensus sequences . On Fig 4A we show the relation between the %Id between the repeats and the energy of the sequence . It is evident that low energy sequences are constructed by pairs of highly similar repeats . This could be a transitive effect: if low energy sequences are very similar to the consensus sequence , and the consensus sequence is formed by two identical repeats , we would be seeing that more similarity between repeats causes lower energies . We can see that it is not the case ( Fig 4B ) . We plot the %Id to the consensus against the energy of each sequence . The consensus was calculated with the most frequent amino acid in each position on sequences used as input . We can see that there is no evident correlation between the energy and the similarity to the consensus . Thus , low energy sequences that differ from the consensus one may be constructed . Also , there are no sequences which get a high %Id to the consensus . We conclude that there are different repeats which have low energies within a protein family , and not only the consensus sequence . Consensus designed ANK proteins are very stable upon chemical and thermal denaturation [13] , and , as shown in Fig 3 also score a very low evolutionary energy according to eq 4 . Can we quantify the relationship between the stability and the evolutionary energy ? A potential test can be performed by comparing to experiments in which the effect of point mutations was evaluated . These incorporate one , two or three point mutations in natural proteins , and characterize the unfolding free energy ΔG of the wildtype and the mutated variant . A higher ΔG reports a more stable protein . We compare the change in the ΔG between the mutated and the wildtype protein ( ΔΔG ) , and the difference of energy for their sequences according to eq 4 . Although the energy expression is learned for pairs of repeats , we can easily extend it to an array of repeats making use of the elongated structure of repeat proteins in which only adjacent repeats interact . From our expression we have parameters assigned to intra-repeat positions ( hi with i = 1 … L 2 and Jij with i , j= 1 … L 2 ) , and inter-repeat interactions ( Jij with i = 1 … L 2 and j = L 2 + 1 … L , and λId ) . Then for each repeat we can assign an internal energy ∑ i = 1 L / 2 h i ( a i ) + ∑ i = 1 L / 2 ∑ j > i L / 2 J i j ( a i , b j ) and a interaction energy ∑ i = 1 L / 2 ∑ j = L / 2 + 1 L J i j ( a i , b j ) + λ I d , which of course depends on the amino acids constituting each repeat . On Fig 5A , we show the comparison between ΔΔG and the evolutionary energy calculated using Eq 4 , done for three different ANK proteins: IκBα [45 , 46] , Notch [47] and p16 [48] . It should be noted that different experimental techniques return different values for ΔG for the same protein , non overlapping within experimental error , pointing that other factors contribute to the experimental quantification of ΔΔG . A linear fit returns R2 ≈ 0 . 61 . From 152 mutations we analyzed , 124 ( 82% ) are predicted favorable when the mutation stabilized the folding of the structure , and unfavorable when they have also been measured to destabilize . The predictions that deviated the most are mutations in Notch from Serine to Proline , which is a structural disruptor , and were not considered in the linear fit . A comparison against FoldX [49] predictions can be found on S6 Fig . On Fig 5B , we show reported mutations on pp32 [50] , a protein belonging to LRR family . Again , measurements with different methods report different values of ΔΔG . The linear fit returns a poor R2 ≈ 0 . 21 , but 30 ( 75% ) mutations are both predicted and reported unstabilizing . A similar comparison was performed by [31] for small globular proteins with an expression related to Eq 1 . To reduce the number of interaction parameters Jij ( ai , bj ) they explicitly used structural information and set to zero all interactions between positions which are not in contact in the native structure . In contrast , we use a L1-regularization to fix to zero those parameters which do not contribute significantly to the fitting process and obtain Jij ( ai , bj ) = 0 and Jij ( ai , bj ) ≠ 0 in all pairs of positions , regardless they are supposed to be in contact or not in the 3D structure . Are the obtained parameters related to structural properties of these proteins ? Local fields , hi ( ai ) , should account for the local propensity of each amino acid in each position , and therefore are expected to be related to fi ( ai ) . Fig 6A shows that the inferred hi ( ai ) parameters are different from the initial condition ln ( fi ( ai ) ) for the ANK family; that is , the values obtained for the parameters that account for higher order correlations are relevant . In red we highlight the points related to the consensus amino acid in each position . All of these residues have a strong local field associated to them , justifying why the construction of sequences with these amino acids results in foldable proteins . We also show a contact map of two ANK repeats ( PDB id: 1N0R ) on Fig 6B: gray background indicates that the two positions given by x and y axis are in contact in the native structure , and white that they are not . On the upper triangle of the figure and in blue crosses , we mark the positions involved in the highest Jij parameters , i . e . those which imply higher coupling . A darker blue indicates that there are more Jij ( more combinations of amino acids ) between those positions . Most of the highest Jij match a pair of positions in contact in the 3D structure , or two which correspond to the same residue in the adjacent repeat patterns , i . e . i-th position in the first repeat and position j = i+33 in the second repeat . In red crosses we show the lowest Jij , that mark a negative constraint . Again , a darker red means that there are more Jij with low values between those positions . It is apparent that these also involve mostly residues in contact , but shows that other regions are responsible for negative design .
We propose a statistical model to account for fine details of the energy distribution in families of repeat proteins using only the sequences of amino acids . The model consists of a specialization of a Potts model to account for the local and pair-wise interactions and an extra term that includes higher order correlations , accounting for the similarity between consecutive repeats . The model is constrained by evolutionary characteristics of the families of proteins: we measure the frequencies of amino acids , co-occurrence of amino acids and the identity between repeats in extant natural proteins . To statistically define these quantities it is necessary to have a large set of sequences , which we showed are currently available for several repeat-protein families [37] . No information about the native folded conformation is required . The computation of the evolutionary energy field is computationally demanding , mostly due to long times spent in rigorous Monte Carlo simulations , but once the fitting is done the parameters can be used to score individual sequences fast and easily . We studied three popular repeat protein families: ANK , TPR and LRR . After pre-processing of the alignments , we had enough sequences ( ≈ 20500 , 10000 and 18800 respectively ) to fit the model to pairs of repeats of each family . We scored the evolutionary energy of all natural sequences in PFAM , and it allowed us to clearly distinguish between natural proteins and random sequences of amino acids: the first have energy values < -50 and show a large spread while all random sequences have energy values ≈ 0 . We evaluated designed repeat proteins which have been shown to fold and found that they score within the natural sequences distribution of energies . For the ANK and TPR families , these designed proteins have been shown to be highly stable upon thermal and chemical denaturation and , coincidentally , they are located at the most favorable side of the energy distribution of natural proteins , suggesting that the evolutionary energy score can be related to folding stability . The energetic model can be used in Monte Carlo simulations to generate sequences that agree with the natural constraints of a given protein family . This ensemble of simulated sequences matches the amino acid frequencies , the identity between repeats and also the energy distribution of natural proteins . We found this set of simulated sequences is statistically indistinguishable from natural counterparts . Thus , the proposed model can be used as a tool to design repeat-protein sequences that have all the natural characteristics evaluated to date . Repeat proteins bind to other polypeptides and are candidates for specific binder scaffolds . Designed repeat proteins have been successfully synthesized and adapted to biomedical applications . Nevertheless , consensus design limits the possible variants as only a small proportion of residues are free to vary . Furthermore , they are extremely stable . Including coupling information can wide the possible sequences that can be studied , and could lead to more malleability of the designed molecules . Moreover , the stability change upon single point mutation can be well predicted by the model using just sequence information . For the ANK family , evolutionary energy variations correlate with the experimental values with an R2 ≈0 . 6 . This improves FoldX [49] performance , which additionally requires a reference structure . Moreover , from the 152 experiments analyzed , the 82% predicts the direction of the stability change upon a point mutation . For the LRR family , the correlation is considerably lower , but 75% of the mutations are both predicted and reported in the available bibliography as destabilizing . For both the simulated sequences and for natural counterparts , we found that the similarity between consecutive repeats correlates with lower energy values , and that these are not necessarily similar to the consensus sequence of the family , pointing out that duplication of stretches of sequences may well be an important factor in the evolution of these systems [51] . The existence of a simple and reliable energy function to score the ‘evolutionary energy’ of repeat-proteins can be used to trace the biological forces that acted upon their history , and to explore to which extent these conflict with the physical necessities of the systems [52] . Mapping the energy inhomogeneities along the repeat-arrays may allow us to infer the population of excited states in these proteins , many of which have been related to their physiological mechanisms .
Multiple sequence alignments of repeats were obtained from PFAM 27 . 0 [53] . The aligned sequences usually have misdetected initial and final residues . The amino acids at the ends of the repeat-detection do occur in the polypeptide chains ( they are not actual deletions ) and incorporating them improves the statistics of the real sequences . We completed these positions with the amino acids present on the actual proteins using the provided headers on the alignment and crossing information with UniProt database [54] . This leads to a reduction on the number of gaps in our alignments , which usually derives into noisy predictions in correlation analyses [31] . After , we created the alignment of pairs of repeats , joining sequences of repeats which are consecutive in a natural protein . Finally , we removed insertions from the alignments by deleting positions which have gaps in more than 80% of the sequences in the alignment . Our model fits the occurrence of amino acids in every position , which we call the marginal frequency of residue ai at position i of the alignment and denote fi ( ai ) , and the joint occurrence of two amino acids ai and bj simultaneously at two different positions of the alignment , fij ( ai , bj ) . To avoid biases by the overrepresentation of some proteins in the database , we used CD-HIT [55] to cluster sequences at 90% of identity and chose a representative sequence from each cluster . Finally , we computed by counting the fi ( ai ) and fij ( ai , bj ) , and divided by the total number of sequences . From the same alignment explained in Frequencies calculations , for a sequence which has L residues constituting two consecutive repeats , the %Id between the repeats is the number of amino acids in positions i and i + L 2 , for i = 1 … L 2 which are exactly the same . Gaps are treated as an amino acid . Once we have the values for all sequences in an alignment , we define pid as the proportion of sequences within the alignment with the same %Id between repeats . Given a set of parameters hi , Jij , λId and Eq 4 , we use a Monte Carlo procedure and the Metropolis criterion to generate an ensemble of N sequences of length L each . We initiate with a random string of L residues . At each step , we produce a point mutation in any position . If this mutation is favorable , i . e . the energy is lower than that of the original sequence , we accept the mutation . If not , we accept the mutation with a probability of e−ΔE , where ΔE is the difference of energy between the original and the mutated sequence . When accepted , the mutated sequence is used as the original one for next step . We add one sequence to our final ensemble every t steps ( we used t = 1000 ) . Our model is proposed to reproduce fi ( ai ) , fij ( ai , bj ) and pid from the alignment of natural sequences . To learn the set of parameters hi , Jij , λId which reproduce them , we used a gradient descent procedure . In each step , an ensemble of N = 80000 sample sequences was produced via Monte Carlo using as energy the expression 4 and the trial parameters . We measured its marginal , joint frequencies and pid and updated the local parameters according to: h i t + 1 ← h i t - ϵ s f i ( a i ) - f i m o d e l ( a i ) ( 5 ) As the number of parameters for coupling is large ( = 212L2 ) , we used a regularization L1 to force to 0 those parameters which are not contributing significantly to the modeled frequencies . Then , we update these parameters by: Jijt+1←0ifJijt=0and|fi ( ai , bj ) -fimodel ( ai , bj ) |<γϵjfij ( ai , bj ) -fijmodel ( ai , bj ) -γsign ( fij ( ai , bj ) -fijmodel ( ai , bj ) ) ifJijt=0and|fij ( ai , bj ) -fijmodel ( ai , bj ) |>γJijt+ϵjfij ( ai , bj ) -fijmodel ( ai , bj ) -γsign ( Jijt ) ifJijt+ϵj ( fij ( ai , bj ) -fijmodel ( ai , bj ) -γsign ( Jijt ) ·Jijt>00ifJijt+ϵj ( fij ( ai , bj ) -fijmodel ( ai , bj ) -γsign ( Jijt ) ·Jijt<0 ( 6 ) Finally , the parameters λId are updated according to: λ I d t + 1 ← λ I d t + ϵ I D p i d ( % I d ) - p i d m o d e l ( % I d ) ( 7 ) We iterated until the maximum difference between the predicted frequencies and the natural sequences was below 0 . 02 . This value was chosen according to the robustness of the frequencies estimations on the available data . We calculated the frequencies on half of the available sequences and compared the results to the frequencies counts on all the available sequences . The largest differences were slightly below 0 . 02 . We believe that this maximum error thus reflects the actual error in the data and it is not reasonable to ask the model for more accuracy than that of the data itself . The code was written in C++ and is available at GitHub: https://github . com/proteinphysiologylab/2017_Espadaetal . | Unlike most natural proteins that are made with apparently random strings of amino acids , repeat-proteins are formed with tandem stretches of similar elements . The statistical description for these occurrences can be captured with a simple energetic model that accounts for evolutionary mechanism that gave rise to these proteins . The resulting energetic model can be used to infer folding stability and can generate sequences that are indistinguishable from the natural ones . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"sequencing",
"techniques",
"split-decomposition",
"method",
"globular",
"proteins",
"multiple",
"alignment",
"calculation",
"mutation",
"protein",
"structure",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"sequence... | 2017 | Inferring repeat-protein energetics from evolutionary information |
Yersinia pestis , the etiologic agent of the disease plague , has been implicated in three historical pandemics . These include the third pandemic of the 19th and 20th centuries , during which plague was spread around the world , and the second pandemic of the 14th–17th centuries , which included the infamous epidemic known as the Black Death . Previous studies have confirmed that Y . pestis caused these two more recent pandemics . However , a highly spirited debate still continues as to whether Y . pestis caused the so-called Justinianic Plague of the 6th–8th centuries AD . By analyzing ancient DNA in two independent ancient DNA laboratories , we confirmed unambiguously the presence of Y . pestis DNA in human skeletal remains from an Early Medieval cemetery . In addition , we narrowed the phylogenetic position of the responsible strain down to major branch 0 on the Y . pestis phylogeny , specifically between nodes N03 and N05 . Our findings confirm that Y . pestis was responsible for the Justinianic Plague , which should end the controversy regarding the etiology of this pandemic . The first genotype of a Y . pestis strain that caused the Late Antique plague provides important information about the history of the plague bacillus and suggests that the first pandemic also originated in Asia , similar to the other two plague pandemics .
In 541 AD , eight centuries before the Black Death , a deadly infectious disease hit the Byzantine Empire , reaching Constantinople in 542 and North Africa , Italy , Spain , and the French-German border by winter 543 [1] . The so called “Plague of Justinian” , named after the contemporaneous emperor , led to mass mortality in Europe similar to that of the Black Death . It persisted in the territory of the Roman Empire until the middle of the 8th century and likely contributed to its decline , shaping the end of antiquity [1] . Based on historical records , this disease has been diagnosed as bubonic plague although discrepancies between historical sources and the progression of Y . pestis infections have led some authors to suppose that the Plague of Justinian was caused by a different pathogen ( as discussed in [2] ) . This vivacious discussion was recently reinforced by an ancient DNA study of the second pandemic that also questioned whether Y . pestis was truly the causative agent of the first pandemic [3] , [4] . Western scientists have traditionally subdivided Y . pestis strains into three biovars: Antiqua , Medievalis , and Orientalis; depending on their abilities to ferment glycerol and reduce nitrate [5] . However , this system ignores many other Y . pestis biovars that have been designated and described by other scientists [see 6 , 7 , 8] . Biovars , which are based upon phenotypic properties , do not always correspond directly to specific molecular groups because the same phenotype can result from different mutations [9] . As a result , it has been suggested that groupings within Y . pestis , or assignment of unknown strains to specific populations should be based upon molecular signatures and not phenotypes [9] . Fortunately , the recent construction of highly-accurate rooted global phylogenetic trees for Y . pestis [10] , [11] ( reproduced in Figure 1 ) have facilitated the assignment of isolates to distinct populations . The most recent global phylogeny is based upon single nucleotide polymorphisms ( SNPs ) identified from the genomes of 133 global strains [11] . All clones that caused the third pandemic belong to populations assigned to the molecular group 1 . ORI [10] , [11]; the basal node for this group is N14 ( Figure 1 ) . Two recent studies [3] , [12] have queried key SNPs in DNA samples obtained from victims of the second pandemic ( 14th century AD ) , facilitating the phylogenetic placement of these samples in the most recent global phylogeny [11] . These samples are along the branch between nodes N07 and N10 ( Figure 1 ) close to the “big bang” polytomy at node N07 , where major branches 1–4 split from major branch 0 [11] . Specifically , ancient Y . pestis DNA samples from two of these studies [3] , [12] , which were obtained in England and France , are along branch N07-N10 – just one SNP away from the polytomy at N07 [11] . An additional sample from one of these studies [12] , which was obtained in the Netherlands , occurs farther along this same branch – three SNPs away from the polytomy at N07 [11] . Only a few previous studies [13]–[15] have described the isolation of Y . pestis DNA from victims of the Late Antique pandemic and only one work group [13] , [14] attempted to genotype the samples , assigning them to biovar Orientalis , which is now also designated molecular group 1 . ORI [9] . However , the authenticity of these results has been questioned repeatedly because current stringent ancient DNA anticontamination protocols ( e . g . independent replication ) were not utilized [16] , [17] . In addition , the robustness of the genotyping approach utilized in one of these studies [13] has been questioned [18] . Finally , it has been suggested [12] , [19] that the resulting phylogenetic assignment ( i . e . membership in the 1 . ORI group ) could not have existed at the time of the Justinianic Plague . Indeed , it seems impossible that isolates from the 1 . ORI group caused the first pandemic as this group likely evolved only over the last ∼200–210 years [10] , [11] . Against this background , we analyzed and genotyped new material from putative Justinian plague victims dated to the 6th century A . D . from an Early Medieval graveyard in Bavaria , Germany . This cemetery , called Aschheim , contained 438 individuals in total and is characterized by a striking number of double and multiple burials clustering in the second half of the sixth century [20] . In an earlier study [15] , we reported isolation of Y . pestis DNA from two individuals from Aschheim . However , this previous study failed to utilize all of the contamination controls and authentication of results that has been recommended for studies that describe the detection of pathogen DNA in human remains from archeological sites [12] , [21] . In this current study we utilized these more stringent approaches and our results confirm that Y . pestis was indeed responsible for the Justinianic Plague . More importantly , we were able to genotype the Y . pestis DNA present in samples from one individual using five key SNPs from the recent global Y . pestis phylogenies [10] , [11] . The genotyping results confirm that the Y . pestis strain from the Ascheim victim is more basal on the global phylogeny than the Y . pestis populations that caused the Black Death and the third pandemic ( Figure 1 ) .
Assuming that plague victims might have been buried together , we collected teeth from 19 individuals originating from 12 multiple burials from the 6th century at Aschheim ( Table 1 ) . All samples were tested for Y . pestis specific DNA in a newly built specialized aDNA laboratory in Munich using both quantitative Real-Time PCR ( qPCR ) and a conventional PCR approach; these approaches targeted a 70 nt portion and a 133 nt portion of the Y . pestis-specific plasminogen activator gene ( pla ) , respectively . This gene , which is located on the multi-copy plasmid pPst that is specific to Y . pestis , has been used in several previous studies to test samples from plague skeletons dating to the time of the Black Death [e . g . 12] , [22] . Using qPCR , we repeatedly obtained a specific pla amplification fragment from samples obtained from eight individuals although , with the exception of sample A120 , the target copy number was extremely low in most of the analyzed DNA extracts ( Table 1 ) . In addition , via conventional PCR we repeatedly obtained a longer pla amplification fragment from samples from two of these individuals ( A82 and A120; Table 1 ) . These amplicons contained pla sequences ( GenBank accession number KC170159 ) that were 100% identical to the type strain CO92 . Concurrently , four samples obtained from intact teeth from four different individuals were independently analyzed in a second DNA laboratory ( Mainz; Table 1 ) . This analysis involved amplification of a 148 nt pla fragment by conventional PCR [12] . Only one of the four samples ( from individual A120 ) produced an amplicon ( Table 1 ) . The observable differences in pla amplification success across the three PCR approaches utilized in this study ( Table 1 ) are likely a function of the target PCR amplicon sizes . In agreement with typical ancient DNA behavior [23] , our amplification success decreased with increasing target length ( Table 1 ) . We attempted to genotype all of the positive samples . However , likely due to differences in DNA preservation among the samples we were only able to gain reproducible results from samples from one individual , A120 ( Table 2 ) . Note that this was the only individual that was found to be Y . pestis-positive with all three PCR approaches ( Table 1 ) . We queried multiple samples from individual A120 with assays targeting five key SNPs from the most recent global phylogenies for Y . pestis [10] , [11] and determined whether these samples possessed the ancestral or derived states for these five SNPs ( Table 2 ) . These five SNPs occur along specific branches in the Y . pestis phylogeny: s545 occurs along the branch between nodes N06 and N07; s87 and s89 occur along the branch between N04 and N05 , s82 occurs along the branch between the phylogenetic branching point of Mongolian strain MNG 2972 ( see below ) and N04 , and s463 occurs along the branch between the phylogenetic branching point of strain MNG 2972 and N03 ( Figure 1 ) . In the Munich aDNA laboratory we determined that Y . pestis DNA samples obtained from individual A120 possess ancestral states for SNPs s545 , s87 , and s89; and derived states for SNPs s82 and s463 ( Table 2 ) . In the second aDNA laboratory ( Mainz ) we confirmed these results for s82 and s87 ( Table 2 ) ; assays for the other SNPs were not utilized in this laboratory . Partial alignments of selected SNP regions of sample A120 in comparison to the reference sequences of Y . pestis type strain CO92 and strain 91001 ( var microtus ) are shown in Table 3 . In all cases , extraction and PCR negative controls never produced an amplicon when tested with Y . pestis specific primers . These results indicate that the phylogenetic position of sample A120 in the global Y . pestis phylogeny is along the branch between the phylogenetic branching point of strain MNG 2972 and N04 , along branch N04-N05 , along the branch from N04 to 0 . ANT1 , or along one of the sub-branches within 0 . ANT . 1 ( Figure 1 ) .
Our analyses conducted in two separate aDNA laboratories independently confirmed our previous results [15] that some humans buried in the 6th century Ascheim cemetery were infected with Y . pestis . These findings confirm that Y . pestis was the causative agent of the Justinianic Plague and should end the controversy over the etiological agent of the first plague pandemic . This outcome is contrary to a recent study [3] that questioned whether Y . pestis was indeed the causative agent of the first pandemic based upon the assumption that only strains from major branches one and two are pathogenic to humans , which they estimated to have emerged only in the 13th century AD . However , Cui et al . [11] recently determined that most Y . pestis lineages are capable of causing human plague and suggested that this capability has been present since Y . pestis evolved from its Y . pseudotuberculosis ancestor approximately 1 , 500–6 , 400 years ago . Thus , they concluded that Y . pestis strains pathogenic to humans already existed long before the beginning of the first pandemic . Another important issue resolved by our study concerns the geographic origin of the Plague of Justinian . The phylogenetic position of our Y . pestis samples from the first pandemic ( Figure 1 ) suggests all three plague pandemics were caused by Y . pestis strains that originated in Asia . Two recent studies placed the origin of the 1 . ORI strains that caused the first pandemic in China [10] , [11] , and recent phylogenetic placement of samples from the second pandemic [3] , [12] near extant strains from China [11] ( Figure 1 ) suggests that strains that caused the second pandemic also originated in this region . The only extant Y . pestis strains assigned to the same portion of the global phylogeny ( Figure 1 ) as our Justinian samples from individual A120 are members of group 0 . ANT1 , which has only been reported from western China [10] , [11] , and strains from Mongolia [8] , such as MNG 2972 ( Figure 1 ) . Although multiple historical sources have pointed to an African origin for the Justinian Plague [1] , [5] , [24] , including speculations based on genealogies of Y . pestis [11] , they have not discussed the original sources of where the bacteria arose . Our results document that those original sources were in Asia . Cui et al . [11] recently raised the possibility that the Angola strain ( sole representative of group 0 . PE3; Figure 1 ) might have spread from Africa to all of Europe and been involved in the first pandemic . They based this hypothesis on several points . First , the Angola strain contains more SNPs than any other known strain of Y . pestis , which is consistent with a history of involvement in epidemic waves . Second , their 95% confidence intervals for the age estimates of the nodes that flank Angola ( 0 . PE3 ) in the global phylogeny , nodes N01 and N03 ( Figure 1 ) , are 2 , 775 BC – 590 AD and 932 BC – 806 AD , respectively , which overlap with the 541 AD date given for the beginning of the first pandemic . Third , they assume that the strain named Angola was actually isolated in Africa in the country of Angola . We do not dispute their first two points . However , we know of no published studies that describe the original isolation of strain Angola making its origins apocryphal . Additional contemporary Angola-like isolates would add insights into this single unique strain type . Although it remains possible that Angola-like strains ( ancestors ) , regardless of its geographic origin , may have been involved in the first pandemic , this remains just a hypothesis until additional samples from the first pandemic are genotyped and found to be closely-related to the Angola strain . Multiple independent age estimates for our samples are consistent with the timing of the first pandemic . The duration of occupancy of the row burial cemetery at Aschheim-Bajuwarenring has been determined by strong archaeological evidence to range from approximately 500–700 AD [20] . Radiocarbon dating , which has been carried out on three individuals analyzed in this study , including A120 ( Table 1 ) , is consistent with this range . Finally , the phylogenetic position of our samples on the global Y . pestis phylogeny is on main branch 0 between nodes N03 and N05 , with node N04 occurring in between ( Figure 1 ) . In their Figure S8 , Cui et al [11] provide the 95% confidence intervals for the age estimates for these three nodes . The date given for the beginning of the first pandemic , 541 AD , overlaps with the confidence intervals for nodes N03 and N04 , although not with the confidence intervals for N05 . Collectively , these various age estimates for our samples provide convincing evidence that they are of the correct age to have been involved in the first plague pandemic . Our results also provide new stimulus to the discussion about simultaneous multiple inhumations in Europe during the Early Medieval period [25] , [26] . It is often presumed that only mass graves are suggestive of a highly infectious disease [27] , whereas our results indicate that epidemics can also be indicated by a clustering of simultaneous inhumations involving only two or three individuals ( Table 1 ) . This observation may help to identify additional potential victims of the Justinianic Plague . Genetic studies of additional skeletal remains from other plague sites in different geographic regions would not only enhance our knowledge regarding the evolution of the pathogen , but also improve our understanding of the epidemics and spread of the Justinianic Plague . In addition , as there is no known historical source indicating that the Justinianic Plague reached current day Bavaria , our results provide the only evidence that the disease crossed the Alps and affected local populations there [1] .
The burial date of the individuals tested for Y . pestis in this study were previously estimated by archaeological methods [20] to fall in a range from 525 to 680 AD ( Table 1 ) . To confirm this , we carried out radiocarbon dating on three samples . For individual A58 , calibration indicated cal . 431–544 AD ( 95 . 4% probability ) as the most likely range . Individual A76 from a second burial pit was dated to cal . 443–566 AD ( 95 . 4% probability ) , and individual A120 from a third burial pit was dated to cal . 435–631 AD . ( 95 . 4% probability ) . From all 19 individuals ( Table 1 ) two or more teeth were taken and analyzed at the aDNA laboratory in Munich . For four individuals ( A58 , A76 , A105 , and A120 ) , another intact tooth was sent directly to a second aDNA laboratory in Mainz where they were analyzed independently and blindly . In Munich the pre-PCR DNA analyses , including the decontamination procedure , DNA extraction , and assembly of the reactions for PCR amplification; were carried out in the new aDNA laboratories at the ArchaeoBioCenter ( Ludwig-Maximillians-University , Munich ) . This aDNA laboratory is located several kilometers from the laboratory used for the post PCR analyses , which included the actual amplification process and sequencing; the post PCR laboratory is situated at the Bundeswehr Institute of Microbiology in Munich . Movement of samples between the laboratories was always unidirectional: from the aDNA laboratories to the post PCR laboratory . The pre-PCR laboratories are dedicated solely to aDNA analysis and have specialized equipment , such as airlocks , HEPA filtered air , positive air pressure , and UV air flow cleaner . In addition , extensive cleaning protocols using bleach and UV irradiation are in place . All possible further methodological precautions were also taken , such as mock extractions , PCR blanks , and independent replications of extractions and amplifications . In the first step , samples were subjected to decontamination procedures consisting of cleaning the outer surface with a 1% NaOCl solution and exposure to 15 min of UV irradiation on each side , with subsequent powdering using a ZrO2-coated mill . DNA extraction in Munich was performed as described previously [15] on powder aliquots of 0 . 4 g . In Mainz precautions for preventing contamination , pre-treatment of the samples and extraction protocols were as published previously [12] . Every sample analyzed in the Munich laboratory for Y . pestis specific DNA ( pla ) was tested at least for three times using the qPCR and conventional PCR approach before considering it negative . Samples that yielded amplification products in any of these PCR reactions were submitted to genotyping assays targeting five key SNPs from the most recent global Y . pestis phylogenies [10] , [11]: s545 ( qPCR approach ) ; s87 ( both qPCR and conventional PCR approach ) ; and s82 , s89 , and s463 ( conventional PCR approaches ) . For qPCR assays ( pla ) , or qPCR SNP endpoint genotyping assays ( s87 and s545 ) , we used 1× Platinum Quantitative SuperMix-UDG ( Invitrogen ) , 6 mM MgCl2 , ( Applied Biosystems ) , 0 . 4 mg/ml BSA ( Ambion/Life Technologies ) , assay specific primer and probe concentrations ( Table 4 ) ( TibMolbiol ) , and 2 . 0 to 4 . 0 µl of template DNA in a final reaction volume of 12 to 24 µl . Primer sequences are listed in Table 4 . Cycling conditions comprised an initial step at 50°C for 2 min , an activation step at 95°C for 10 min , 50 cycles at 95°C for 10 sec , and an assay specific annealing temperature for 1 min ( Table 4 ) . Final cooling was carried out at 4°C for 30 sec . QPCR assays were carried out on a LightCycler 480 II platform ( Roche , Mannheim , Germany ) . Quantification of pla-qPCR assays was possible by determination of the copy numbers per reaction by generating a standard curve using synthetic oligonucleotide constructs . Data analysis was performed using the LightCycler 480 II software version 1 . 5 ( Roche , Mannheim , Germany ) . For conventional PCR assays ( pla , s82 , s87 , s89 , s463 ) , we used 1× Qiagen Multiplex PCR Master Mix , 0 . 4 mg/ml BSA , and 2 or 4 µl of DNA in a final volume of 50 µl . Primer sequences are listed in Table 4 . The experiments were run on an Eppendorf Mastercycler Pro instrument . Cycling conditions started with an initial activation step at 95°C for 15 min . This was followed by 50 cycles at 94°C for 30 sec , an assay specific annealing temperature ( Table2 ) for 30 sec , and 72°C for 1 min , ending with a final elongation step at 72°C for 10 min . Final cooling was carried out at 8°C until analysis . Results ( pla or SNPs ) were only considered valid if they could be repeated at least three times from different extracts . Protocols for pla , s82 , and s87 analysis in the second aDNA lab ( Mainz ) were carried out as previously published [12] . All amplified products were verified by DNA sequencing and BLASTN-analysis . For the sequencing reactions in Munich we used 1× BigDye terminator v . 3 . 1 Cycle Sequencing Ready reaction Mix ( Applied Biosystems ) , 1 pmol/µl of the respective primers , and 3–5 µl of purified DNA template in a final volume of 10 µl . The reaction was run on a GeneAmp 9700 ( Applied Biosystems ) instrument , starting with an initial denaturation step for 1 min at 96°C , followed by 25 cycles at 96°C for 10 sec , 50°C for 5 sec and 60°C for 2 mins , and ending with cooling at 4°C until further processing . After purification using the Dye Ex 2 . 0 Spin Kit ( Qiagen ) sequences were generated on a Genetic Analyzer 3130 ( Applied Biosystems ) instrument . Sequences were further analyzed using the program CodonCodeAligner version 4 . 0 . Analyses of the results of the SNPs assays were carried by aligning the amplicons to Y . pestis type strain CO92 ( AL590842 . 1 ) , which possessed the derived state for all of the queried SNPs , and Y . pestis microtus strain 91001 ( AE017042 . 1 ) , which possessed the ancestral state for all of the queried SNPs . Sequencing in Mainz was carried out as previously described [12] . If long enough , sequences were deposited in GenBank ( Accession numbers KC170160-KC170162 ) and the alignments are shown in Table 3 ( only partial sequences are shown for longer sequences ) . The global Y . pestis phylogeny in Figure 1 is reconstructed from Figures 1A and S3B in Cui et al . [11] . Their phylogeny was constructed using SNPs discovered from the genomes of 133 modern isolates . We have indicated the main branches and molecular groups identified by Cui et al . [11] but not all of their sub-branches and sub-groups . The phylogenetic branching point for Mongolian Y . pestis strain MNG 2972 was determined using SNP information provided for this strain in Riehm et al . [8] . Note that , based upon the five SNPs queried in this study , this contemporary Mongolian strain possesses a distinct genotype when compared to the ancient Y . pestis DNA samples utilized in this study; the Mongolian strain possesses the ancestral state for s82 . The GenBank ( http://www . ncbi . nlm . nih . gov ) accession numbers for DNA sequences longer than 50 nt determined in this paper are KC170159-KC170163 . | Plague is a notorious and fatal human disease caused by the bacterium Yersinia pestis that is endemic in many countries around the world . Three of the most devastating pandemics in human history have been associated with plague . The second pandemic originated in central Asia and peaked in Europe between 1348 and 1350 ( a period of time known as the Black Death ) . The third pandemic began in the Yunnan province of China in the mid-1850s and subsequently spread to Africa , the Americas , Australia , Europe , and other parts of Asia . The second and third pandemics are well documented and scientifically proven . However , the first pandemic , which began in the 6th century and is also known as Justinianic Plague , is still a matter of controversy . Recently it has been suggested that Justinian's plague was not caused by Y . pestis . We detected Y . pestis DNA in samples obtained from multiple 6th century skeletons from Germany . This confirms that Justinianic Plague crossed the Alps and affected local populations there , including current day Bavaria . Furthermore , we used DNA fingerprinting approaches to determine Asia as the likely geographic origin for these strains . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"social",
"and",
"behavioral",
"sciences",
"biological",
"anthropology",
"genetics",
"biology",
"microbiology",
"anthropology"
] | 2013 | Yersinia pestis DNA from Skeletal Remains from the 6th Century AD Reveals Insights into Justinianic Plague |
Mutations in human Gli-similar ( GLIS ) 3 protein cause neonatal diabetes . The GLIS3 gene region has also been identified as a susceptibility risk locus for both type 1 and type 2 diabetes . GLIS3 plays a role in the generation of pancreatic beta cells and in insulin gene expression , but there is no information on the role of this gene on beta cell viability and/or susceptibility to immune- and metabolic-induced stress . GLIS3 knockdown ( KD ) in INS-1E cells , primary FACS-purified rat beta cells , and human islet cells decreased expression of MafA , Ins2 , and Glut2 and inhibited glucose oxidation and insulin secretion , confirming the role of this transcription factor for the beta cell differentiated phenotype . GLIS3 KD increased beta cell apoptosis basally and sensitized the cells to death induced by pro-inflammatory cytokines ( interleukin 1β + interferon-γ ) or palmitate , agents that may contribute to beta cell loss in respectively type 1 and 2 diabetes . The increased cell death was due to activation of the intrinsic ( mitochondrial ) pathway of apoptosis , as indicated by cytochrome c release to the cytosol , Bax translocation to the mitochondria and activation of caspases 9 and 3 . Analysis of the pathways implicated in beta cell apoptosis following GLIS3 KD indicated modulation of alternative splicing of the pro-apoptotic BH3-only protein Bim , favouring expression of the pro-death variant BimS via inhibition of the splicing factor SRp55 . KD of Bim abrogated the pro-apoptotic effect of GLIS3 loss of function alone or in combination with cytokines or palmitate . The present data suggest that altered expression of the candidate gene GLIS3 may contribute to both type 1 and 2 type diabetes by favouring beta cell apoptosis . This is mediated by alternative splicing of the pro-apoptotic protein Bim and exacerbated formation of the most pro-apoptotic variant BimS .
The Kruppel-like zinc finger protein Gli-similar ( GLIS ) 3 plays a critical role in pancreatic development , and loss-of-function mutations in this transcription factor lead to a syndrome characterized by neonatal diabetes , hypothyroidism and other congenital dysfunctions [1] , [2] . Genome-wide association studies in large numbers of individuals with type 1 ( T1D ) or type 2 ( T2D ) diabetes indicated that common variants near GLIS3 gene are associated with both forms of diabetes [3]–[7] , making GLIS3 one of the few candidate genes for both T1D and T2D . It remains to be proven , however , that susceptibility alleles for T1D and T2D actually decrease expression of GLIS3 in pancreatic beta cells . GLIS3 is also implicated in the regulation of human fasting glucose and insulin [4] , [8] and glucose-stimulated insulin release [5] , suggesting a key role for the transcription factor in human beta cell development/function . GLIS3 deficient mice have a major decrease in beta cell mass and develop neonatal diabetes [9] , [10] . These mice also have decreased expression of several key transcription factors required for the endocrine development of the pancreas , i . e . Neurogenin3 , NeuroD1 , MafA and Pdx1 [9] , [10] . Moreover , conditional knockout of GLIS3 in adult mice causes defective insulin secretion and increase susceptibility to high fat diet-induced diabetes [11] . In vitro knockdown ( KD ) or overexpression of GLIS3 in rat insulinoma 832/13 cells showed that the transcription factor binds to a cis-acting element in the rat insulin 2 ( Ins2 ) , modulating its transcriptional activity [12] . GLIS3 also synergizes with the beta cell transcription factors Pdx1 , MafA and NeuroD1 , increasing insulin promoter activity , besides directly regulating the expression of MafA ( another important inducer of the insulin promoter ) [12] . These observations suggest that GLIS3 plays an important role for the development of mature pancreatic beta cells and for the transcription of its key hormone insulin . There is , however , little information on the role of GLIS3 in beta cell susceptibility to immune- or metabolic-induced apoptosis and little data on its impact on adult beta cells . Beta cell apoptosis contributes to the two main forms of diabetes [13] , [14] . Diabetes candidate genes expressed in beta cells may have a major impact on cell survival/function in T2D [15] , [16] , [17] and T1D [18]–[22] and in the local inflammatory responses leading to insulitis and chronic autoimmunity in T1D [19] , [21] , [23] . We have presently developed an in vitro model of GLIS3 deficiency in beta cells by using siRNAs targeting different regions of the GLIS3 mRNA . GLIS3 KD increased beta cell apoptosis under basal condition and sensitized cells to death induced by interleukin 1β ( IL-1β ) + interferon-γ ( IFN-γ ) or palmitate , agents that may contribute to beta cell loss in respectively T1D and T2D . This increase in apoptosis was secondary to the activation of the intrinsic pathway of apoptosis through alternative splicing of the pro-apoptotic BH3-only protein Bim at least in part via inhibition of the splicing factor SRp55 . The present data provide the first indication that a candidate gene for diabetes may modify alternative splicing and thus hamper beta cell survival .
GLIS3 KD in INS-1E cells ( Figure 1A–1E ) significantly decreased key transcription factors for the maintenance of the beta cell phenotype , namely MafA and Pdx1 , the glucose transporter Glut2 and INS2 . These observations were reproduced using a second siRNA targeting GLIS3 ( Figure S1A and S1B ) , and were confirmed in primary rat beta cells , where a 50% KD of GLIS3 led to a decrease in INS2 expression and a trend for decreased Glut2 expression ( Figure 1F–1H ) . These changes in gene expression by GLIS3 KD had a functional impact , with decreased basal and glucose-stimulated glucose metabolism and of glucose +/− forskolin-induced insulin release in INS-1E cells ( Figure 1K–1L ) and a 25% decrease in insulin accumulation in the medium of human islets transfected with GLIS3 siRNA , as compared to controls ( Figure 1J ) . We next evaluated whether GLIS3 KD affects beta cell viability under basal condition or following exposure to stress signals that may be relevant for type 1 diabetes , namely the pro-inflammatory cytokines IL-1β + IFN-γ or the viral by-product double stranded RNA ( dsRNA ) [18] , [19] , [21] , tested here as the synthetic analog PIC , or for type 2 diabetes , namely the free fatty acids oleate and palmitate [13] . GLIS3 KD by two independent siRNAs increased basal and cytokine-induced apoptosis in INS-1E cells ( Figure 2B , Figure S1C , Figure S2A and S2B ) . Importantly , GLIS3 KD by two independent siRNAs also augmented apoptosis in human islet cells , under both basal condition and following exposure to IL-1β + IFN-γ ( Figure 2C and 2D , Figure S1D and S1E ) . The KD of GLIS3 ( Figure 2E and 2G ) also sensitized INS-1E cells to apoptosis induced by PIC ( Figure 2F ) , oleate and palmitate ( Figure 2H ) . Thus , even a partial decrease in GLIS3 expression , as may be the case in some of the diabetes-predisposing gene polymorphisms , enhances beta cell sensitivity to basal , immune- or metabolic stress-induced apoptosis . In a mirror image of these experiments , GLIS3 overexpression using an adenoviral vector ( Figure S3A ) lead to increase MafA expression ( Figure S3B ) and decreased by >50% cytokine-induced apoptosis in INS-1E cells ( Figure S3C ) . Apoptosis secondary to GLIS3 KD and exposure to pro-inflammatory cytokines was mediated by the intrinsic ( mitochondrial ) pathway of apoptosis , as suggested by increased cleavage of caspases 9 and 3 ( Figure 3A; densitometry in Figure S2A and S2B ) , cytochrome c release to the cytosol ( Figure 3B; densitometry in Figure S2C ) and Bax translocation to the mitochondria ( Figure 3C ) . A possible mechanism for cytokine-induced apoptosis in beta cells is increased nitric oxide production and consequent endoplasmic reticulum ( ER ) stress and Chop activation [24] , [25] . GLIS3 KD , however , did not increase nitric oxide production ( Figure S4A ) or Chop expression ( Figure S4B ) , making it unlikely that these are relevant mechanisms for beta cell apoptosis following GLIS3 inhibition . Interestingly , GLIS3 KD led to a decrease in Chop expression under basal condition or at some time points following cytokine exposure . Beta cells express markers of ER stress even under basal condition , probably due to the high load on the ER caused by physiological and fluctuating insulin production [26] . It is conceivable that the decrease in Ins2 mRNA expression observed in GLIS3 KD cells ( Figure 1D ) contributes to the observed decrease in Chop expression . Beta cell survival is critically dependent on the balance between anti- and pro-apoptotic Bcl-2 proteins [27] . To examine whether GLIS3 modulates these proteins we measured expression of two key anti-apoptotic proteins , namely Bcl-2 and Bcl-xL . GLIS3 inhibition did not affect Bcl-2 and Bcl-xL expression under basal condition or following exposure to cytokines ( Figure 4 ) , and neither was there a change in a third anti-apoptotic protein , namely Mcl-1 ( data not shown ) . We next examined the pro-apoptotic BH3-only proteins DP5 and PUMA . These proteins have previously been shown to contribute to IL-1β + IFN-γ-mediated beta cell apoptosis [28] , [29] , but their expression was not increased by GLIS3 KD ( Figure S4C and S4D ) . If anything , there was a decrease in PUMA expression at some time points . Another important mediator of cytokine-induced beta cell apoptosis is the BH3-only protein Bim . Previous studies from our group have shown that STAT-1-induced Bim expression [30] , [31] and JNK-induced Bim phosphorylation on serine 65 [20] contribute to beta cell apoptosis . GLIS3 KD ( Figure 5A ) increased basal Bim mRNA expression and led to a mild increase in its expression following cytokine treatment at 2 and 8 h , with a decrease after 16 and 24 h ( Figure 5B ) . This was independent of STAT-1 activation , since GLIS3 KD did not modify total or phospho-STAT1 expression following exposure to IL-1β + IFN-γ for 0 . 25–24 h ( data not shown ) . Bim has three main isoforms generated by alternative splicing , namely BimEL , BimL , and BimS [32] . Western blot showed a preferential and nearly 2-fold increase in the expression of BimS in GLIS3 KD cells both before and after exposure to cytokines ( Figure 5C; the blots are quantified in Figure 5D ) . There was a less marked increase in BimEL and BimL at some of the time points following cytokine exposure ( Figure 5C; densitometry in Figure S5A and S5B ) . The BimS up-regulation seems to be secondary to GLIS3-modulated alternative splicing , since GLIS3 KD induced a nearly 2-fold increase in BimS mRNA expression basally and following cytokine exposure ( Figure 5E ) , with only a minor and transient increase in BimEL and BimL mRNA ( at 2–8 h of cytokine treatment ) which was followed by a significant decrease after 16 and 24 h ( Figure S5C and S5D ) . Importantly , these findings were reproduced in human islets , where KD of GLIS3 led to nearly 50% increase of BimS ( Figure 5F ) with no significant increase in the other two splice variants ( Figure S5E and S5F ) . In INS-1E cells exposed to palmitate , there was also a significant increase in BimS expression ( Figure 5G ) and less marked changes in BimEL and BimL ( Figure S5G and S5H ) . The mirror image was seen in gain-of-function experiments: adenoviral GLIS3 overexpression ( Figure S7A ) decreased BimS expression and caspase 3 cleavage ( Figure S7B ) . The decreased caspase 3 activation corroborates the finding that GLIS3 overexpression protects against cytokine-induced apoptosis ( Figure S3C ) , probably via inhibition of BimS ( Figure S7B ) . We have previously shown that this Bim siRNA markedly decreases expression of the three splice variants of Bim in cytokine-treated INS-1E cells [30] . In both INS-1E cells , primary beta cells and human islet cells Bim depletion by >50% ( P<0 . 05 ) ( Figure S6A , S6B and data not shown ) abrogated the basal increase in apoptosis observed following GLIS3 KD ( Figure 6A , 6B and 6C ) . Interestingly , while Bim depletion protected human islet cells against apoptosis ( Figure 6C ) , it failed to prevent the decrease in insulin secretion secondary to GLIS3 KD ( data not shown ) , indicating dissociation between the functional and pro-apoptotic effects of GLIS3 KD . Bim KD also partially prevented the increase in cell death induced by GLIS3 KD + cytokines ( Figure 6A , 6B and 6C ) . These observations were confirmed with a second siRNA ( Figure 6D ) that induced a preferential inhibition of BimS ( 71±4% inhibition of BimS , P<0 . 001 ) . To examine whether this beneficial effect of Bim KD was restricted to cytokines , we performed double KD for GLIS3 and Bim and then exposed the cells to palmitate ( Figure 6E ) . Palmitate treatment also preferentially increased expression of BimS in INS-1E cells ( Figure 5G , Figure S5G and S5H ) . Bim KD had only a minor protective effect against palmitate alone , in agreement with recent data suggesting that DP5 and PUMA are the main mediators of palmitate-induced beta cell apoptosis [33] , but it abrogated the additive effect of GLIS3 KD upon palmitate exposure , decreasing cell death to the levels observed with palmitate alone ( Figure 6E ) . To address the mechanisms by which GLIS3 affect Bim splicing , we examined the potential role of Pnn and SRp55 , two splicing factors described in other tissues as potential regulators of Bim splicing [34] , [35] and detected as present and modified by cytokines in human islets exposed to cytokines [21] . Pnn expression was not modified by GLIS3 KD ( data not shown ) . On the other hand , GLIS3 KD decreased protein expression of SRp55 in INS-1E cells ( Figure 7A ) ( 43%±8% inhibition of SRp55 protein expression , p<0 . 05 , n = 7 ) , while GLIS3 overexpression augmented SRp55 expression basally and following cytokine exposure ( Figure S7B ) . To assess the functional impact of decreased expression of SRp55 , we inhibited it with two specific siRNAs ( Figure 7B ) . After KD of SRp55 , there was a significant increase of BimS expression under both basal condition and following cytokine treatment ( Figure 7C ) . We next evaluated whether SRp55 KD affects beta cell viability and observed an increase in apoptosis under basal condition and following cytokine exposure ( Figure 7D ) indicating a relevant role of SRp55 in viability . Double KD of SRp55 and BimS ( 71%±4% inhibition of BimS , p<0 . 001 ) counteracted the increase in apoptosis caused by SRp55 KD ( Figure 7D ) , suggesting a role for this splicing regulator in the downstream effects of GLIS3 ( Figure 7E ) . cAMP generators have been previously shown to protect beta cells against both cytokine- and palmitate-induced apoptosis [36]–[39] , and we evaluated whether forskolin could prevent beta cell apoptosis following GLIS3 KD . Interestingly , forskolin nearly completely prevented the basal increase in apoptosis following GLIS3 KD ( Figure 8A ) , which was accompanied by a significant decrease in the expression of BimS but not BimEL or BimL ( Figure 8B and 8C ) . In cytokine-treated GLIS3 KD deficient cells forskolin induced only a mild and partial protection , which was paralleled by a progressive restoration of BimS expression ( Figure 8B and 8C ) .
Genome-wide association studies have allowed the identification of a large number of associations between specific loci and T1D or T2D . The mechanisms by which most of these candidate genes predispose to diabetes remain to be clarified . This emphasizes the need for detailed studies on the function of candidate genes in the key tissues involved in the development of diabetes . Taking into account the central role for beta cell failure in both T1D and T2D [13] , it is of particular relevance to clarify the potential impact of these “diabetes genes” on pancreatic beta cell dysfunction and death . There is little convincing genetic link between T1D and T2D to date [40]–[42] , with the possible exception of Latent Autoimmune Diabetes in Adult ( LADA ) , a particular form of diabetes that has been reported to share some susceptibility risk factors from both T1D and T2D [43] . To our knowledge the GLIS3 locus is the only one showing association with genome-wide significance for both T1D , T2D or glucose metabolism traits in non-diabetic subjects , adults or children and adolescents , and in population-based cohorts [3]–[8] . GLIS3 is the single gene located within the confidence interval of the region of association with T1D [3] , and the SNPs that have been reported to be associated with T1D , T2D and T2D-related traits are all in very strong linkage disequilibrium ( LD ) to each other ( pairwise correlation coefficient r2 of 0 . 95 to 1 . 0 between the strongest associated SNPs for the key studies [3] , [4] , [6] , [7] , [44] ) , supporting the hypothesis that a unique variant near GLIS3 may be responsible for all the reported associations with these common diabetes and related traits . Furthermore , a review of all the published genetic studies and available data on T1D , T2D and T2D-related traits indicated that the orientation of association is concordant between all these traits ( C . Julier , unpublished observations ) , with the same allele associated with increased risk of T1D , increased risk of T2D , increased fasting glucose , decreased fasting insulin level , decreased HOMA-B and glucose stimulated insulin release ( nominal P-values for association with these traits <10−3 [4] , [5] ) , suggesting the role of a shared mechanism between both forms of diabetes . Pancreatic islets from T2D patients have a nearly 50% decrease in GLIS3 mRNA expression as compared to islets obtained from non-diabetic subjects ( P<0 . 001; data re-calculated from [45] and confirmed by RT-PCR analysis of whole islets and FACS-purified human beta cells; Bugliani M , Marselli L and Marchetti P , unpublished data ) , but it remains to be determined whether this is a direct effect of the risk alleles on GLIS3 expression or secondary to chronic exposure to high glucose levels . Similarly , GLIS3 was found to be one of the most differentially expressed genes between beta cells from T2D and non-diabetic subjects [46] . The fact that recessive loss-of-function mutations in GLIS3 cause severe neonatal diabetes in humans [1] and in transgenic mouse models [9] , [10] , secondary to a major decrease in beta cell mass , suggests that this transcription factor is necessary for beta cell development and differentiation . Together , these genetic and functional observations indicate that GLIS3 itself is the susceptibility gene responsible for the observed associations with T1D , T2D and T2D-related traits . The region strongly associated with T1D as defined by Barrett et al . [3] maps to the 5′ region of the GLIS3 long transcript , which is pancreas and thyroid specific [1] and includes the first exons and corresponding promoter region . Of note , all the SNPs in LD with diabetes and associated SNPs are non-coding . This suggests that the responsible variant affects the regulation of GLIS3 expression in pancreatic beta cells , most likely through a reduction of GLIS3 expression predisposing to T1D and T2D . It is thus important to understand whether these milder phenotypes affect the resistance of adult beta cells to challenges provided by immune- , viral- or metabolic-mediated stress . These stresses may cross talk with candidate genes for T1D and T2D . Our present observations suggest that a relatively mild reduction of GLIS3 gene expression in beta cells by two independent siRNAs decreases expression of Pdx1 , MafA , Ins2 and Glut2 and inhibit glucose oxidation and glucose-induced insulin secretion . These findings are in line with evidence obtained in foetal , neonatal or adult mouse beta cells [9] , [11] , and suggest a key role for GLIS3 in maintaining the beta cell differentiated phenotype . Of particular interest in the context of diabetes is the observation that GLIS3 KD increases rat beta cell apoptosis under basal condition and sensitizes the cells to death induced by pro-inflammatory cytokines ( IL-1β + IFN-γ ) , the viral by-product dsRNA , and the free fatty acids oleate and palmitate , while GLIS3 up-regulation protects against cytokine-induced apoptosis ( present data ) . GLIS3 KD also increases apoptosis of human islet cells under both basal condition and following exposure to IL-1β + IFN-γ . This broad range of sensitization to pro-apoptotic stimuli by GLIS3 KD suggests that GLIS3 , besides contributing to maintain beta cell function , provides signals required for preservation of cell viability . In line with these observations , suppression of Pdx1 , a key transcription factor for the maintenance of the differentiated phenotype of beta cells , triggers beta cell death via dissipation of the mitochondrial inner membrane electrochemical gradient Deltapsi ( m ) [47] . GLIS3 KD also contributes to beta cell apoptosis via a mitochondrial phenomenon , namely triggering of the intrinsic pathway of apoptosis as a result of the activation of the BH3-only protein Bim ( see below ) . Decreased Pdx1 expression sensitizes pancreatic beta cells to ER stress [48] , but this is not the case for GLIS3 KD , as indicated by normal expression of Chop ( present findings ) and other ER stress markers ( data not shown ) . The increase in cell death in GLIS3 deficient cells is secondary to activation of the intrinsic pathway of apoptosis , as indicated by Cytochrome c release to the cytosol , Bax translocation to the mitochondria and activation of caspases 9 and 3 . A detailed analysis of the upstream pathways implicated in GLIS3 KD-induced beta cell apoptosis indicated modulation of alternative splicing of the pro-apoptotic BH3-only protein Bim , favouring expression of the most pro-apoptotic splice variant of Bim , namely BimS [49] , [50] . In agreement with these observations , Bim depletion abrogated the pro-apoptotic effects of GLIS3 KD alone or in combination with pro-inflammatory cytokines or palmitate . Bim can bind to and inhibit most anti-apoptotic Bcl-2 proteins , besides directly activating the pro-apoptotic protein Bax [51] . Importantly , Bim contributes to cytokine- [20] , [30] , virus- [23] and high glucose-induced [52] pancreatic beta cell apoptosis . Previous observations in pancreatic beta cells indicated that Bim can be regulated by cytokines at the transcriptional [30] , [53] or phosphorylation [20] level . The present study is the first to show regulation of Bim function in beta cells by changes in splicing . There are three main isoforms of Bim , namely BimEL , BimL , and BimS that are generated by alternative splicing [32] . BimEL and BimL have a binding site for the dynein light chain 1 which decreases their pro-apoptotic activity via sequestration to the cytoskeleton [32] , [54] , while BimS is free to exert its potent pro-apoptotic activity [49] , [50] . Alternative splicing affects more than 90% of human genes [55] . It generates enormous proteome diversity , and may have a major impact on cell survival , exposure of novel antigenic epitopes , alteration of surface location of antigens and posttranslational modifications . There is a growing interest in the role of alternative splicing in several autoimmune diseases [56] , but nearly nothing is known on its role in pancreatic beta cell dysfunction and death in diabetes . We have recently shown that beta cell exposure to pro-inflammatory cytokines modifies alternative splicing of hundreds of expressed genes and affects expression of more than 50 splicing-regulating proteins [21] , [57] . Palmitate also modifies alternative splicing of a different group of genes in human islets ( Cnop M , Sammeth M , Bottu G and Eizirik DL , unpublished data ) . The present observations provide the first indication that a candidate gene for diabetes may act by regulating alternative splicing . This effect of GLIS3 KD is mediated , at least in part , via down regulation of the splicing factor SRp55 ( Figure 7 ) . This was confirmed by the reverse experiment , i . e . GLIS3 overexpression induced SRp55 and prevented BimS production ( Figure S7B ) . In line with this , the inhibition of SRp55 led to an increase in BimS expression and beta cell apoptosis ( Figure 7 ) . These results suggest that GLIS3 regulates the expression of splicing factors and consequently the splicing of their target genes . It remains to be clarified whether this is a direct effect or a secondary phenomenon via downstream regulation of other genes . In conclusion , the present observations suggest that modifications in expression of the candidate gene GLIS3 may contribute to both T1D and T2D by favouring beta cell apoptosis . This takes place to a large extent via modified alternative splicing of the pro-apoptotic protein Bim . Additional studies are now required to characterize this new avenue for functional studies on candidate genes for diabetes , namely their cross-talk with alternative splicing and other processes regulating generation of gene/protein diversity .
Human islet collection and handling were approved by the local Ethical Committee in Pisa , Italy . Wistar rats were used according to the rules of the Belgian Regulations for Animal Care with approval of the Ethical Committee for Animal Experiments of the ULB . INS-1E cells ( kindly provided by C . Wollheim , Centre Medical Universitaire , Geneva , Switzerland ) at passages 60–72 were cultured in RPMI 1640 GlutaMAX-I medium , supplemented with 5% heat-inactivated foetal bovine serum ( FBS ) , 50 units/ml penicillin , 50 µg/ml streptomycin , 10 mM HEPES , 1 mM Na-pyruvate , and 50 µM 2-mercaptoethanol in a humidified atmosphere at 37°C and 5% CO2 . Isolated pancreatic islets of male Wistar rats ( Charles River Laboratories , Brussels , Belgium ) , housed following the guidelines of Belgian Regulations for Animal Care , were dispersed and beta cells purified by autofluorescence-activated cell sorting ( FACSAria , BD Bioscience , San Jose , CA , USA ) [58] , [59] . Beta cells ( 93±2% purity; n = 6 ) were cultured in Ham's F-10 medium containing 10 mM glucose , 2 mM glutamine , 50 µM 3-isobutyl-L-methylxanthine , 0 . 5% fatty acid-free bovine serum albumin ( BSA ) ( Roche , Indianapolis , IN , USA ) , 5% FBS , 50 units/ml penicillin , and 50 µg/ml streptomycin [59] . The same medium but without FBS was used during cytokine exposure . Human islet cells from 8 non-diabetic donors ( age 66±5 years , five men/three women , body mass index 25 . 7±0 . 9 Kg/m2 ) were isolated in Pisa , with the approval of the Ethics Committee of the University of Pisa . Islets were isolated by enzymatic digestion , and density-gradient purification [60] . They were then cultured in M199 medium containing 5 . 5 mM glucose and shipped to Brussels , Belgium within 1–5 days of isolation . After overnight recovery in Ham's F-10 containing 6 . 1 mM glucose , 10% FBS , 2 mM GlutaMAX , 50 µM 3-isobutyl-1-methylxanthine , 1% BSA , 50 U/ml penicillin and 50 µg/ml streptomycin , islets were dispersed , transfected with siCTL , siGLIS3 , siBim or siGLIS3/siBim and exposed or not to cytokines for 24 h . The same medium but without FBS was used during cytokine exposure . The percentage of beta cells in the dispersed islet preparations , as determined by immunohistochemistry for insulin [37] , was 48±6% . The siRNAs used in the study are described in Table S1 . The optimal concentration of siRNA used for cell transfection ( 30 nM ) was established previously [61] . Cells were transfected using the Lipofectamine RNAiMAX lipid reagent ( Invitrogen , Carlsbad , CA , USA ) as previously described [31] . Allstars Negative Control siRNA ( Qiagen , Venlo , the Netherlands ) was used as negative control ( siCTL ) . siCTL does not affect beta cell gene expression or insulin release , as compared with nontransfected cells [31] , [61] , [62] . Beta cells transfected with siRNAs were used for experiments 24–48 h after transfection . To express GLIS3 in insulin-secreting cells , we obtained from SIRION Biotech ( Munich , Germany ) a recombinant adenovirus comprising fragments of the mouse GLIS3 mRNA ( GenBank: NM_175459 ) . The murine GLIS3 coding region was amplified by PCR from cDNA clone BC167165 purchased from Source Bioscience ( Berlin , Germany ) and was cloned via Nhe1 and EcoRV into the shuttle vector pO6-A5-CMV to give pO6-A5-CMV-GLIS3 . The CMV-GLIS3-SV40-pA region of pO6-A5-CMV-GLIS3 was then transferred via recombination in a BAC vector containing the genome of a replication deficient Ad5-based vector deleted in E1/E3 genes . Presence and correctness of the GLIS3-ORF in the resulting BAC-vector BA5-CMV-GLIS3 was confirmed by DNA-sequencing . An adenovirus expressing the luciferase protein ( Ad-LUC ) was used as control [63] . INS-1E cells were infected as previously described [63] . The cytokine concentrations used were based on previous dose-response experiments performed by our group [64] , [65] and were 10 units/ml or 50 units/ml of recombinant human IL-1β for INS-1E cells or primary rat beta cells/human islet cells , respectively ( a kind gift from Dr . C . W . Reinolds , National Cancer Institute , Bethesda , MD-USA ) and 100 units/ml or 500 units/ml of recombinant rat IFN-γ for INS-1E cells and primary rat beta cells or 1000 units/ml of recombinant human IFN-γ for human islet cells ( R&D Systems , Abingdon , UK ) . Culture supernatants from cytokine-treated cells were collected for nitrite determination ( nitrite is a stable product of NO oxidation ) at OD540 nm using the Griess method . The synthetic dsRNA polyinosinic-polycytidylic acid ( PIC; Sigma , St Louis , LO , USA ) was used at the final concentration of 1 µg/ml [19] . Cellular transfection with PIC was made as described for siRNA , with the difference that Lipofectamine 2000 was used instead of Lipofectamine RNAiMAX [19] . Oleate and palmitate ( sodium salt , Sigma , Bornem , Belgium ) were dissolved in 90% ( vol . /vol . ) ethanol and diluted 1∶100 to a final concentration of 0 . 5 mM in the presence of 1% charcoal-absorbed BSA , corresponding to a free fatty acid/BSA ratio of 3 . 4 [66] , [67] . Forskolin was diluted in DMSO and used at final concentration of 20 µM ( Sigma ) . mRNA was extracted and reverse transcribed as described [59] . Expression of target genes was determined by real-time PCR using SYBR Green [59] , [68] and comparison with a standard curve [69] . Expression values were corrected by the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) for INS-1E and primary rat beta cells and β-actin for human islet cells . GAPDH or β-actin expression is not modified under the present experimental conditions [18] , [66] , [70] . Primer sequences are described in Table S2 . Primers for MafA , Pdx1 , INS2 , Glut2 , Chop , Dp5 and Puma were described previously [53] , [71] . D-[U-14C] glucose ( specific activity: 300 mCi/mM , concentration: 1 mCi/ml , Perkin Elmer , Waltham , MA , USA ) was used to evaluate glucose oxidation in control and GLIS3 KD cells exposed to different glucose concentrations as described [72] . The rate of glucose oxidation was expressed as pmol/120 min . 105 cells . For determination of insulin secretion , INS-1E cells were incubated for 1 h in glucose-free RPMI GlutaMAX-I medium and then incubated for 30 min in Krebs-Ringer solution . Cells were then exposed to 1 mM , 10 mM or 10 mM glucose with forskolin ( 20 µM ) for 30 min . Insulin was measured in the supernatant by the rat insulin ELISA kit ( Mercodia , Uppsala , Sweden ) . Results were normalized by the insulin content measured after cell lyses . Insulin accumulation in the medium of cultured human islets was measured by the human insulin ELISA kit ( Mercodia , Uppsala , Sweden ) . The percentage of viable , apoptotic and necrotic cells was determined following 15 min of incubation with 5 mg/ml of the DNA-binding dyes propidium iodide ( PI , Sigma ) and Hoechst 33342 ( HO , Sigma ) . This method is quantitative and has been validated for use in pancreatic beta cells and INS-1E cells by comparison with electron microscopy , caspase-3 activation and DNA laddering [18] , [59] , [66] , [73] , [74] . A minimum of 600 cells was counted in each experimental condition . Viability was evaluated by two independent observers , one of them unaware of sample identity . The agreement between findings obtained by the two observers was >90% . In some experiments apoptosis was confirmed by Western blot analysis of cleaved caspase-9 and -3 , cytoplasmic cytochrome c release and BAX translocation to the mitochondria . INS-1E cells were lysed in Laemmli buffer and equal amounts of total protein were heated at 100°C for 5 min , resolved by electrophoresis in 10–14% SDS-polyacrylamide gel and electro-blotted onto nitrocellulose membranes . Immunodetection was performed after overnight incubation with antibodies for cleaved caspase 9 and 3 ( Cell Signaling , Danvers , USA ) , Bcl-2 ( Cell Signaling , Danvers , USA ) , Bcl-xL antibody ( Cell Signaling , Danvers , USA ) , Bim and p-Bim antibodies ( Cell Signaling , Danvers , USA ) , SRp55 antibody ( LifeSpan Biosciences ) , STAT1 and p-STAT1 antibodies ( Cell Signaling ) . α-tubulin ( Cell Signaling ) was used as the loading control . Membranes were then exposed to 150 ng/ml secondary peroxidase-conjugated antibody ( anti IgG ( H+L ) -HRP , Invitrogen ) for 2 h at room temperature and visualized by chemiluminescence ( SuperSignal , Pierce Biotechnology , Rockford , IL , USA ) . Bands were detected by a LAS-3000 CCD camera ( Fujifilm , Tokyo , Japan ) . The densitometry of the bands was evaluated using the Aida Analysis software ( Raytest , Straubenhardt , Germany ) . For the assessment of cytochrome c release , INS-1E cells harvested in cold PBS were centrifuged ( 500 g for 2 min ) and resuspended with 50 µl lysis buffer ( 75 mM NaCl , 1 mM NaH2PO4 , 8 mM Na2PO4 , 250 mM sucrose , 21 µg/µl aprotinin , 1 mM PMSF and 0 . 8 µg/µl digitonin ) and vortexed for 30 s . After centrifugation ( 20 , 000 g for 1 min ) the supernatant was collected as the cytoplasmic fraction . The pellet was resuspended in 50 µl lysis buffer containing 8 µg/µl digitonin , centrifuged ( 1 min at 20 , 000 g ) and the supernatant collected as the mitochondrial fraction [23] , [37] . Equal amounts of proteins were used for Western-blotting with antibodies for cytochrome c ( BD Biosciences ) ( cytoplasmic protein ) , apoptosis-inducing factor ( AIF ) and cytochrome c oxidase ( COX IV ) ( mitochondrial proteins ) ( Cell Signaling ) . β-actin was used as the loading control . INS-1E cells were plated on polylysine-coated glass culture slides ( BD Biosciences ) . After transfection and treatment , cells were fixed for 15 min in 4% paraformaldehyde , washed with PBS and permeabilized in Triton X-100 0 . 1% for 5 min . Slides were then blocked using 5% goat serum and incubated overnight at 4°C with a Bax antibody ( Santa Cruz Biotechnology ) plus ATP synthase β antibody ( mitochondrial marker ) ( BD Biosciences ) . Cells were washed with PBS and incubated for 1 h with the appropriate Alexa fluor 488 or 555-conjugated antibodies ( Invitrogen ) . Cells were stained with Hoechst 33342 , mounted and photographed using fluorescence microscopy ( Axio Imager , Carl Zeiss , Zaventem , Belgium ) [62] . Data are presented as mean ± SEM . Comparisons were performed by two-tailed paired t-test or by ANOVA followed by paired t-test with Bonferroni correction , as adequate . A P value<0 . 05 was considered as statistically significant . | Pancreatic beta cell dysfunction and death is a central event in the pathogenesis of diabetes . Genome-wide association studies have identified a large number of associations between specific loci and the two main forms of diabetes , namely type 1 and type 2 diabetes , but the mechanisms by which these candidate genes predispose to diabetes remain to be clarified . The GLIS3 gene region has been identified as a susceptibility risk locus for both type 1 and type 2 diabetes—it is actually the only locus showing association with both forms of diabetes and the regulation of blood glucose . We show that decreased expression of GLIS3 may contribute to diabetes by favouring beta cell apoptosis . This is mediated by the mitochondrial pathway of apoptosis , activated via alternative splicing ( a process by which exons are joined in multiple ways , leading to the generation of several proteins by a single gene ) of the pro-apoptotic protein Bim , which favours formation of the most pro-apoptotic variant . The present data provides the first evidence that a susceptibility gene for diabetes may contribute to disease via regulation of alternative splicing of a pro-apoptotic gene in pancreatic beta cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"insulin",
"gene",
"regulation",
"endocrine",
"system",
"genetics",
"molecular",
"genetics",
"gene",
"expression",
"biology",
"anatomy",
"and",
"physiology",
"gene",
"splicing",
"endocrine",
"physiology"
] | 2013 | GLIS3, a Susceptibility Gene for Type 1 and Type 2 Diabetes, Modulates Pancreatic Beta Cell Apoptosis via Regulation of a Splice Variant of the BH3-Only Protein Bim |
Long non-coding RNAs ( lncRNAs ) ( > 200 bp ) play crucial roles in transcriptional regulation during numerous biological processes . However , it is challenging to comprehensively identify lncRNAs , because they are often expressed at low levels and with more cell-type specificity than are protein-coding genes . In the present study , we performed ab initio transcriptome reconstruction using eight purified cell populations from mouse cortex and detected more than 5000 lncRNAs . Predicting the functions of lncRNAs using cell-type specific data revealed their potential functional roles in Central Nervous System ( CNS ) development . We performed motif searches in ENCODE DNase I digital footprint data and Mouse ENCODE promoters to infer transcription factor ( TF ) occupancy . By integrating TF binding and cell-type specific transcriptomic data , we constructed a novel framework that is useful for systematically identifying lncRNAs that are potentially essential for brain cell fate determination . Based on this integrative analysis , we identified lncRNAs that are regulated during Oligodendrocyte Precursor Cell ( OPC ) differentiation from Neural Stem Cells ( NSCs ) and that are likely to be involved in oligodendrogenesis . The top candidate , lnc-OPC , shows highly specific expression in OPCs and remarkable sequence conservation among placental mammals . Interestingly , lnc-OPC is significantly up-regulated in glial progenitors from experimental autoimmune encephalomyelitis ( EAE ) mouse models compared to wild-type mice . OLIG2-binding sites in the upstream regulatory region of lnc-OPC were identified by ChIP ( chromatin immunoprecipitation ) -Sequencing and validated by luciferase assays . Loss-of-function experiments confirmed that lnc-OPC plays a functional role in OPC genesis . Overall , our results substantiated the role of lncRNA in OPC fate determination and provided an unprecedented data source for future functional investigations in CNS cell types . We present our datasets and analysis results via the interactive genome browser at our laboratory website that is freely accessible to the research community . This is the first lncRNA expression database of collective populations of glia , vascular cells , and neurons . We anticipate that these studies will advance the knowledge of this major class of non-coding genes and their potential roles in neurological development and diseases .
More than 98% of the human genome does not encode proteins . A large number of transcribed sequences are non-coding transcripts [1–5] . Thousands of long non-coding RNAs ( lncRNAs: usually > 200 bp in length , often spliced and polyadenylated , but lacking protein-coding potential ) were recently discovered and many of them have been shown to play crucial roles in diverse biological processes [6 , 7] . Emerging evidence indicates that lncRNAs may have important roles in Central Nervous System ( CNS ) development , homeostasis , stress responses , and plasticity [6] . For example , many lncRNAs are expressed in the mouse brain and show region-specific expression patterns [8] . Many lncRNAs exhibit dynamic expression patterns during neuronal-glial fate specification and oligodendrocyte lineage maturation [6] . In addition , lncRNAs have been shown to be involved in some neuropsychiatric diseases [9] . An increasing effort is being devoted to lncRNA identification [2 , 3 , 10]; however , it is not trivial to build a comprehensive lncRNA catalog . Compared to their protein-coding counterparts , lncRNAs are generally expressed at lower levels , which make it difficult to detect and assemble these transcripts , especially if the lncRNAs are expressed in the minor cell types within a tissue [2 , 3 , 8] . In addition , lncRNA genes may be regulated in opposing directions in different cell types , so their expression can appear to be static in composite tissue data . Traditionally , microarrays were used to capture lncRNA , but microarray data is limited in its sensitivity and by the probes that can match lncRNAs . Targeted capture ( using tiling arrays to target selected portions of the transcriptome ) followed by RNA-Sequencing ( RNA-Seq ) can be used to validate transcripts that are expressed at low levels [10 , 11]; however , this approach requires a priori knowledge of the target region . Another challenge is to investigate the potential functions of lncRNAs . The functional roles of most characterized lncRNAs were first inferred by transcriptional profiling of different samples , an approach that presumes a cause-and-effect relationship between gene expression and cellular context . This ‘guilt-by-association’ strategy has proven to be a powerful tool for discovering the biological functions of lncRNAs [2 , 12] . Nevertheless , it is still critically important to validate the predicted functions of lncRNAs by classical genetic approaches such as loss-of-function experiments [13] . Additional genomic information , such as specific transcription factor binding that provides evidence of active regulation , can increase the precision of candidate selection for functional validation experiments [14] . We have recently employed RNA-Seq to characterize the transcriptome of various purified cell types isolated from mouse brain [15] . The expression levels of classic cell-type specific markers were high in their corresponding cell types , but very low in the other cell populations , demonstrating the high purity of the isolated brain cell types ( S1 Table; pericytes were excluded from the analyses because of relatively lower purity ) . In the current study , we also sequenced neural stem cells . We identified lncRNAs de novo from these purified cell types and generated a more comprehensive lncRNA annotation database by combining the lncRNAs that we identified with those from multiple other sources including GENCODE , RefSeq , Ensembl , lncRNAdb , and lncRNAs recently identified by several other groups [3 , 4 , 10 , 16 , 17] . Predicting lncRNA functions using purified cell types revealed potential functions for lncRNAs in CNS development . Moreover , to further dissect the functional roles of these lncRNAs , we performed TF motif searches in ENCODE DNase I digital footprint ( DNase-DGF ) experimental data and Mouse ENCODE promoters to infer TF binding proximal to these lncRNAs at various CNS developmental stages [18] . This comprehensive database of lncRNAs from purified brain cell types that we have integrated with TF binding and predicted functional information provides a powerful framework for systematically identifying lncRNAs that are essential for brain cell fate determination . Because Oligodendrocyte Precursor Cells ( OPCs ) play a crucial role in myelination/remyelination , and understanding the determination of OPC fate is critical for harnessing their potential for cell-based therapies [19] , we chose to investigate lncRNAs that may have essential functions in OPC fate determination . Based on our integrative analysis , the top candidate , which we named lnc-OPC , showed highly specific expression in OPCs , remarkable sequence conservation among placental mammals , and OLIG2 binding in its upstream regulatory region , as shown by ChIP-Seq and luciferase assays . Furthermore , the depletion of lnc-OPC significantly reduces OPC formation and affects global expression of genes associated with oligodendrogenesis upon the differentiation of OPCs from NSCs . Interestingly , we found transposable elements ( TE ) inserted in the intron of lnc-OPC in the mouse lineage , implying that TEs might have been involved in the evolution and regulation of the expression of lnc-OPC . These results substantiated the role of lncRNA in OPC determination and established a valuable framework that can be applied to future large-scale functional lncRNA screens in other cell types . We have presented our datasets and analytical results as online resources freely available to the research community ( http://jiaqianwulab . org/braincell/lncRNA . html ) ( Username: lncRNA; Password: rnaseq ) . We anticipate that our study will advance the knowledge of this major class of non-coding genes and their potential roles in neurological development and diseases .
Previous mouse brain lncRNA catalogs created using tissue or organ samples might have missed lncRNAs that can only be found in minor cell types and are hence likely not comprehensive [10] . To further broaden current lncRNA catalogs , we set out to identify lncRNAs that are expressed in various purified brain cell types by employing an ab initio transcriptome reconstruction approach . We previously reported RNA sequencing of poly ( A ) + mRNA from mouse cerebral cortex tissue samples , as well as from highly purified astrocytes , neurons , oligodendrocyte precursor cells ( OPCs ) , newly formed oligodendrocytes ( NFOs ) , myelinating oligodendrocytes ( MOs ) , microglia ( MGL ) , endothelial cells ( Endo ) , and pericytes ( Peri ) [15] . In this previous work , we reported 811 lncRNAs with the criterion of FPKM > 1 from the eight brain cell types based on GENCODE annotation . In order to better study transcriptome dynamics during cell lineage commitment , we now include the published RNA-Seq data for mouse embryonic stem cells ( ESCs ) [3] in our analyses . In addition , we have also sequenced transcripts from primary NSC isolated from mouse cortex using the same library construction procedures as described previously [15 , 20] . In total , ~1 . 2 billion 101-bp paired-end reads were collected from nine cell types and from samples of the whole cortex ( average ~63 million reads/sample ) ( Fig 1A ) . We performed ab initio transcript assembly and detected expression of 5040 ( 5107 , including the whole cortex samples ) multi-exonic lncRNAs from 4059 loci in these brain cell types . There are 1717 novel lncRNA loci compared to the IncRNA genes annotated by GENCODE ( M3 version ) and RefSeq . Ab initio transcriptome reconstruction using cortex samples alone ( at the same sequencing depth ) recovered only 3032 lncRNAs ( from 2519 loci ) , among which 1080 loci are novel compared to GENCODE and RefSeq . Most lncRNAs not detected in cortex samples were expressed at lower levels in respective cell types ( Fig 1B and 1C ) , indicating the limitation of the data obtained from tissues and highlighting the importance of cell-type specific transcriptome profiling for lncRNA identification . Using a previously proposed index [21] , we evaluated the cell specificity of the expression patterns of lncRNAs , as well as that of protein-coding genes . Consistent with previous observations , the lncRNA genes showed more cell-type specific expression patterns than protein-coding genes ( Fig 1D ) . In an effort to obtain a comprehensive lncRNA catalog for downstream analysis , we surveyed available lncRNA annotations in the public domain . Numerous lncRNA annotations were retrieved from databases including lncRNAdb , GENCODE , Ensembl , and UCSC known gene and RefSeq genes . Additionally , we also incorporated lncRNA annotations from several recent transcriptome studies performed using RNA-Seq technology ( see Materials and Methods ) [3 , 4 , 10] . Because the definition of a full-length transcript boundary is not always accurate without further experimental evidence , we merged lncRNA transcripts identified from the same loci that were not annotated by GENCODE , or by any other lncRNA databases , into single transcripts . This merging procedure reduced the apparent total number of lncRNA transcripts , but increased the accuracy of expression level estimation , which is more important for prediction of function [22] . Thus , our final lncRNA annotation includes a total of 11 , 534 lncRNA transcripts from 8714 loci . Finally , we combined the comprehensive lncRNA annotation with all known UCSC genes ( from the iGenome package ) and built a non-redundant annotation . We quantified the expression levels of all annotated transcripts across cell types using this combined annotation ( S2 Table ) . Recent efforts in lncRNA functional prediction using publicly available microarray or RNA-Seq data from tissue samples are limited due to the data source [4 , 23–25] . In order to infer the potential functions of lncRNAs involved in CNS development , we adopted a previously proposed ‘guilt-by-association’ approach [4] . We used RNA-Seq profiles from fifteen types of samples ( eight brain cell types and Mouse ENCODE RNA-Seq data of seven non-brain tissues including thymus , testis , kidney , liver , lung , spleen , and heart ( GEO accession GSE36025 ) ) . The protein-coding genes were ranked by the Pearson correlation of their expression profile with that of a particular lncRNA . The ranked list of these protein-coding genes was then used for Gene Set Enrichment Analysis ( GSEA ) to identify significantly enriched gene sets . This procedure was performed for all lncRNAs against gene sets generated from Gene Ontology ( GO ) functional terms , canonical pathways , and expert-curated gene sets [26] . We then created an association matrix between functional terms and lncRNAs ( S3 Table ) . Only enriched gene sets , with false discovery rates ( FDR ) < 0 . 25 ( as recommended in the GSEA manual ) , were used for matrix creation . To validate the reliability of predicted functions , we examined lnc-OPC and lncRNAs that are known to be expressed in brain and have known functions in the literature [16] ( Fig 2A ) . The predicted functions of these lncRNAs matched those in the literature well , and revealed additional associated functions . For example , lncRNA Malat1 ( metastasis-associated lung adenocarcinoma transcript 1 ) is a highly abundant nucleus-restricted RNA that localizes to nuclear speckles and was suggested to coordinate the RNA polymerase II transcription , pre-mRNA splicing , and mRNA export [27] . Recent studies have also shown that Malat1 is involved in cell-cycle progression and that it enhances cellular proliferation [28] . Consistent with that result , our functional prediction for Malat1 is closely associated with terms such as ‘Pediatric cancer markers’ , ‘Establishment of RNA localization’ , ‘RNA transport’ , ‘Cell division’ , ‘Mitotic cell cycle’ , and ‘Cell cycle phase’ , among others . Interestingly , other functional terms , such as ‘Forebrain development’ , ‘Neural tube development’ , and ‘Brain development’ were also significantly associated with Malat1 . Furthermore , we selected a recently characterized functional lncRNA , Tuna ( also named Tunar ) , as an example to illustrate the importance of using cell-type specific data for function prediction . The lncRNA Tuna is evolutionarily conserved and displays CNS-specific expression patterns [13] . Loss-of-function experiments have revealed that Tuna is required for neuronal differentiation , and depletion of Tuna in zebrafish greatly impaired locomotor functions . The authors thus concluded that depletion of Tuna impairs CNS function , probably due to neuronal defects . Consistent with this observation , in our analysis , gene sets such as ‘neuron markers’ , ‘synapse’ , ‘transmission of nerve impulse’ , and ‘neuron projection’ are among the most highly enriched terms for Tuna . Intriguingly , the gene set ‘oligodendrocyte markers’ is also highly enriched . We then checked the expression level of Tuna across all brain cell types and found that the expression of Tuna in newly formed oligodendrocytes is comparable to that in neurons ( NFO: FPKM = 5 . 78; Neuron: FPKM = 3 . 77 ) ( Fig 2B ) . Thus , Tuna might also have a functional role in oligodendrocytes , which could be one reason that inhibiting Tuna causes impaired locomotor function . Such observations could not be made if cell-type specific transcriptomic data were not available . We also performed weighted gene co-expression network analysis ( WGCNA ) and constructed co-expression networks comprised of both protein-coding and lncRNA genes using cell-type specific RNA-Seq data ( S1 Fig ) . We identified 32 co-expression modules . For 29 co-expression modules that have >100 members , we performed functional term enrichment analysis using the DAVID bioinformatics tools ( Figs 2C and S2 Fig; S4 and S5 Tables ) . Interestingly , three modules were associated with Alzheimer’s disease , Huntington’s disease and Parkinson’s disease . A close look at these modules revealed that they are all enriched for mitochondrion-related functional terms . Consistently , a number of studies have suggested that mitochondria could play a critical role in neurodegenerative diseases [29 , 30] . This raises the possibility that the 24 lncRNAs found in these modules may have potential roles in regulating mitochondrial gene expression and may be related to neurodegenerative diseases . The lower expression levels of lncRNAs relative to their protein-coding counterparts raised the question of whether lncRNAs are actively regulated , or whether their expression is merely transcriptional noise [31] . To address questions as to the involvement of lncRNAs in CNS development and to facilitate functional tests of these possible roles , we analyzed the binding status of transcription factor binding sites across the whole genome . In eukaryotes , transcription is regulated in a cell-type and condition-specific manner through the association of transcription factors with the chromatin . The dynamics of chromatin accessibility in the regulatory regions of lncRNAs during developmental processes can be used as a further indicator of active regulation . We used genome-wide maps of in vivo DNase I footprints data retrieved from ENCODE to assess the dynamics of transcription factor binding in the proximal regulatory regions of lncRNAs and protein-coding genes . DNase I hypersensitivity mapping and genomic footprinting have been used extensively to delineate cis-regulatory DNA and TF binding at nucleotide resolution in various model organisms [32 , 33] . Global mapping of TF footprints provides a powerful tool for assessing the interactions of hundreds of TFs with chromatin in a single experiment . To assess the dynamics of TFs during CNS development , we analyzed genomic DNase-DGF data from ESC and three other available ENCODE datasets related to CNS development ( Whole Brain E14 . 5 [WBE14] , Whole Brain adult 8 weeks [WB8wks] , and Retina Newborn 1 Day [Retina1D] ) . In brief , we scanned the whole mouse genome for TF-binding sites using well-annotated TF-binding motifs collected from databases and the literature , then inferred their binding status from their specific DNase I cleavage profiles using a Bayesian method named CENTIPEDE [34] . Only TF-binding sites with a posterior probability of >0 . 99 were considered to be actively bound by TFs ( S3 Fig ) . Genes under complex transcriptional regulation have been suggested to be subject to a larger ‘control set’ of cis-regulatory modules ( defined as a stretch of DNA in which a number of TFs can bind and regulate expression of nearby genes ) than are genes responsive to only a few regulatory signals [35] . Using the size of cis-regulatory modules as an indicator , we set out to test whether lncRNAs are under the control of TFs during CNS development . To this end , we calculated the number of base pairs that are bound in the promoter regions ( defined as 2 kb upstream and 1 kb downstream of the TSS ) of TFs , non-TF protein-coding genes , lncRNAs , and randomly selected intergenic regions across the DNase-DGF datasets . The results of this analysis showed that , in general , TFs have larger cis-regulatory modules than do other categories of genes , an observation that is consistent with their role as master regulators . Importantly , we also found that lncRNA promoter regions were actively bound by transcription factors more frequently than random intergenic regions ( Fig 3A ) . Most lncRNA genes have a smaller cis-regulatory module than non-TF protein-coding genes and TF genes; however , there are lncRNA genes that are part of a comparable , or even larger , cis-regulatory module than are protein-coding genes , suggesting a finer degree of control of their expression by TFs . To analyze the regulatory DNA dynamics of lncRNAs , we analyzed the DNase-DGF experiment data to identify dynamic DNase I hypersensitive sites ( △DHSs ) at which TFs can potentially bind across the four datasets ( Fig 3B , left ) . A large number ( 13 , 369 ) of △DHSs ( defined as enriched more than three fold in one sample ) were found to reside in proximity to lncRNA genes ( within 10 kb of the TSS ) . There are 6157 lncRNAs that contain at least one △DHS proximal to their TSS . To examine whether the △DHSs can reflect specific TF binding in a developmental context , we first analyzed △DHSs that are specifically activated in ESC compared to three other available ENCODE datasets related to CNS development . We performed enrichment analysis of TF recognition sequences within ESC-activated △DHSs . This analysis yielded several well-known key transcription factors associated with embryonic development and pluripotency , such as KLF4/5 , OCT4 , SOX2 , MYC , and ESRRB , among others ( Fig 3B , right ) . This observation is consistent with previous reports showing that many lncRNAs are targets of key pluripotent transcription factors [4 , 14] . We then performed enrichment analysis for TF recognition sequences in WBE14- , WB8wks- , and Retina1D-activated △DHSs in proximity to lncRNAs . Similarly , many TFs associated with CNS development were enriched , as we expected . For example , NF1 , LHX2/3 , NKX6 . 1 , ISL1 , and a number of basic-helix-loop-helix ( bHLH ) TFs , including ATOH1 , OLIG2 , and NEUROD1 , were among the top enriched TFs ( Fig 3B , right ) . Taken together , our results indicate that the expression of a substantial portion of lncRNAs is under elaborate and dynamic control by TFs during CNS development , which suggests that lncRNAs do not represent mere transcriptional noise , but may indeed play various functional roles . The information regarding TF binding can also be used as a criterion for screening lncRNA candidates for experimental validation . OPCs are distributed throughout the CNS and play the crucial role of differentiating into oligodendrocytes that ensheath axons with myelin during CNS development and remyelinate axons after damage [36] . Abnormal development or maintenance of myelin sheaths can impair efficient propagation of action potentials along axons , and lead to disorders such as multiple sclerosis ( MS ) and leukodystrophies [37 , 38] . Previous studies in animal models showed that transplanted OPCs , but not mature oligodendrocytes , can myelinate; hence , OPCs can serve as a promising cell source for transplantation therapies in demyelinating diseases [36] . Thus , understanding the sophisticated molecular mechanisms of OPC fate determination is critical for harnessing OPCs for cell-based therapies . We thus set out to identify lncRNAs that could be essential for OPC fate determination . To this end , we analyzed RNA-Seq data from NSCs and OPCs . Through analysis of differential expression , we identified 4703 ( Up-regulated: 2626 , Down-regulated: 2077 ) coding genes and 355 ( Up-regulated: 254 , Down-regulated: 101 ) lncRNAs with greater than two fold ( FDR < 0 . 05 ) changes in expression ( Fig 4A ) . RNA-seq tracks for three example lncRNAs that are enriched in an OPC lineage are shown in Fig 4B . As expected , up-regulated protein-coding genes were significantly enriched for GO terms associated with oligodendrogenesis , such as ‘axon ensheathment’ , ‘ensheathment of neurons’ , and ‘myelination’ , whereas down-regulated genes were enriched for terms associated with cell proliferation , such as ‘cell cycle’ and ‘cell division’ . Recent studies have identified non-coding RNAs that are directly regulated by key transcription factors with determinative roles in cellular differentiation [4 , 14] . Based on these observations and our TF binding analysis described above , we investigated whether the differentially expressed lncRNAs are controlled by transcription factors known to play a critical role in oligodendrogenesis . During development , the bHLH TF OLIG2 is both necessary , and in some contexts , sufficient for OPC generation [39–41] . A number of studies have suggested that OLIG2 can interact with multiple TFs to provide positional cues and thus locally regulate OPC specification [42 , 43] . ASCL1 ( also known as Mash1 ) is another bHLH TF that plays an important role in OPC specification and promotes oligodendrogenesis in brain [44] . Other TFs belonging to the Homeobox and Sox family , including NKX2 . 2 and SOX10 , also participate in oligodendrogenesis [41 , 45] . We examined whether the lncRNAs that are up-regulated during oligodendrogenesis could possibly be controlled by known oligodendrogenic TFs , particularly OLIG2 and ASCL1 . To this end , we analyzed previously identified genome-wide binding site locations of OLIG2 and ASCL1 in DNAse-DGF data ( S6 Table ) . Binding sites for OLIG2 were found in proximity ( within 10 kb of the TSS ) to 88 ( 34% ) of 254 lncRNAs that are up-regulated during oligodendrogenesis . Also , binding sites for ASCL1 were found close to the TSS of 111 ( 43% ) up-regulated lncRNAs . Similarly , up-regulated protein-coding genes were enriched for OLIG2 ( 917 ( 35% ) ) and ASCL1 binding motifs ( 1246 ( 47% ) ) . The enrichment of OLIG2- and ASCL1-binding sites is statistically significant , as determined by permutation analysis ( p-value < 0 . 01 , using randomly selected genes as control ) . In addition to DNase-DGF analysis , we also performed a separate analysis to search for TF motifs in promoter regions annotated by the Mouse ENCODE project [46 , 47] . OLIG2-binding motifs within ENCODE promoters were found in the upstream regulatory regions of 125 ( 49% ) out of 254 lncRNAs that are up-regulated during the formation OPCs from NSCs . ASCL1-binding motifs inside ENCODE promoters were found in the upstream regulatory regions of 139 ( 63% ) out of 254 lncRNAs ( S6 Table ) . A correlation analysis was carried out to identify any correlations between the expression of TFs with binding motifs within ENCODE promoter regions upstream of the lncRNAs and the target lncRNAs . A heatmap representing the correlations between the expression of lncRNAs and TFs is included in S4 Fig . Several criteria were applied to select prospective candidate lncRNAs for testing potential roles in OPC fate determination . First , the candidates had to be up-regulated in OPCs compared to NSCs , and substantially expressed in OPCs ( e . g . , FPKM > 3 in OPC ) . Second , lncRNAs had to be specifically enriched in OPCs among the cell types studied . Third , there had to be evidence of OLIG2 binding in the proximal regions of the lncRNA genes . These criteria narrowed our focus to specific candidates for further analysis . Subsequently , we performed qPCR validation and loss-of-function shRNA knockdown experiments for these candidates to investigate their possible roles in OPC fate determination . Finally , we focused on lnc-OPC , because it was among the most enriched lncRNAs in OPCs , according to RNA-Seq data . Interestingly , two predicted non-coding transcripts ( XR_873836 . 1 and XR_873835 . 1 , part of RIKEN cDNA 5330416C01Rik gene ) were documented independently in the NCBI Reference Sequence after we had identified lnc-OPC de novo , which supported our analysis result ( Fig 5A ) . As predicted by our motif search , there are OLIG2 binding sites in the upstream regulatory region of lnc-OPC and one is located within an ENCODE promoter . In order to test the in silico prediction , we performed OLIG2 ChIP-Seq and found one binding peak overlapping with these predicted motifs ( Fig 5A and S7 Table ) . Moreover , we performed OLIG2 ChIP-qPCR using NSC cell lysates . Significant enrichment over genomic input DNA and IgG control were detected by both pairs of primers in OLIG2 ChIP samples , which indicated OLIG2 binding in the upstream regulatory region of lnc-OPC ( Fig 5B and 5C ) . During CNS development , OLIG2 plays an important role in maintaining NSCs and the lineage specification of NSCs into OPCs . The observation that OLIG2 binds to the upstream regulatory region of lnc-OPC suggests that OLIG2 might regulate the expression of lnc-OPC . We created two constructs ( A1 and A2 ) containing the upstream regulatory region of lnc-OPC fused to a luciferase reporter gene , as indicated in Fig 5A . To test whether OLIG2 could regulate the transcription of lnc-OPC , 293FT cells were cotransfected with fusion luciferase reporter constructs along with the construct expressing either OLIG2 or GFP ( as a control ) . The expression of OLIG2 significantly inhibited luciferase expression from the fusion reporter constructs by more than 64% compared to the GFP control ( by 76 . 55% for A1 and 64 . 83% for A2 ) ( Fig 5D ) . However , no effect on luciferase expression was observed when cotransfecting OLIG2 and empty pGL4 . 11 vector control . These observations indicated that OLIG2 binds to the upstream regulatory region of lnc-OPC and represses the transcription of lnc-OPC in NSCs . qPCR was used to independently monitor changes in the expression of lnc-OPC during OPC differentiation from NSCs ( Fig 6A ) . The expression of lnc-OPC was elevated gradually and greater than ten fold increase was observed 2 days after NSC differentiation . We designed three lentivirus-based short hairpin RNAs ( shRNAs ) to target the last exon of lnc-OPC and two of them succeeded in knocking down more than 50% of lnc-OPC expression in NSC culture ( Fig 6B ) . Subsequently , puromycin-selected NSCs were differentiated to OPCs in culture . Knockdown of lnc-OPC resulted in a significant decrease in the expression of OPC markers ( MBP , PLP1 , CNP ) and O4+ ( oligodendrocyte surface marker ) in cells , as assessed by qPCR and immunostaining experiments ( Fig 6C , 6D and 6E ) . qPCR results indicated that depletion of lnc-OPC inhibited MBP , PLP1 , and CNP expression by >60% when compared to control cells ( Fig 6C ) . Additionally , O4-positive OPCs were significantly reduced in lnc-OPC-depleted cells compared to the control cells ( Fig 6D and 6E ) . To rule out the possibility that the reduction in OPC cell number was due merely to an impaired ability of NSCs to proliferate , we performed a BrdU cell proliferation assay and found that there was no significant influence on NSC proliferation after shRNA knockdown ( S5 Fig ) . To further confirm and examine the effect of lnc-OPC depletion at a genome-wide scale , we performed RNA-Seq to evaluate the transcriptome changes caused by lnc-OPC knockdown during OPC differentiation from NSC ( S6 Fig ) . DAVID GO functional term enrichment analysis , using genes that are differentially expressed in the lnc-OPC knockdown compared to the control , revealed significant enrichment of ‘oligodendrocyte development’ , ‘oligodendrocyte differentiation’ , ‘glia cell development’ , and ‘axon ensheathment’ terms that are associated with oligodendrogenesis ( S8 Table ) . Collectively , our experimental and computational analysis suggests an essential role of lnc-OPC in oligodendrogenesis .
The central nervous system is comprised of an intensively diverse array of cell types , which complicates the task of identifying cell-type specific transcripts and limits the utility of data obtained from tissue samples . A large number of lncRNAs are reported as expressed in the mouse brain [8 , 10]; however , the lncRNA catalogs created using tissue or organ samples have not been comprehensive [10] . Thus , gene expression dynamics occurring in rare cell types may go undetected during analysis of these kinds of samples . Hence , direct reconstruction and measurement of the transcriptomes of specific cell types is crucial to understanding the gene dynamics that underlie cellular phenotypes . This is particularly true for measuring the expression profiles of lncRNAs that are considerably more cell-type specific than are protein-coding genes [2 , 3 , 15] . Our RNA-Seq data for highly purified cell types from mouse brain represent the first lncRNA expression database for collective populations of glia and neurons . Because we have been able to identify many multi-exonic lncRNAs that were not annotated previously , our study highlights the importance of using cell-type specific data for lncRNA identification and functional characterization . The field of lncRNA functional studies is still in its early stage . Because lncRNAs are less evolutionarily conserved than protein-coding genes , analysis of their sequence conservation alone lacks the power to assess the biological significance of lncRNAs [48 , 49] . To date , the functional roles of most of the characterized lncRNAs were first inferred by transcriptional profiling and co-expression analyses of lncRNAs and protein-coding genes . We havepredicted potential functions of lncRNAs using our brain cell-type specific RNA-Seq data and non-brain RNA-Seq datasets in order to prioritize candidates for functional tests . The predicted functions of those documented lncRNAs from the CNS matched well with , but were not restricted to , their functions previously described in the literature , which suggests that lncRNAs may be involved in different biological functions in different cell types . Nevertheless , association alone cannot distinguish whether the change in expression is the cause or the consequence of differences in cellular state [13] . For example , aside from known functions in the literature , the functional terms ‘Neural tube development’ and ‘Brain development’ were also predicted to be significantly associated with Malat1; however , recent studies have shown that Malat1 is dispensable for CNS development [50 , 51] , indicating that functional predictions arising from by ‘guilt-by-association’ must be received with caution , and experimental validation is necessary . To select potential candidates for functional validation , other genomic information such as specific TFs binding is useful . We analyzed the DNase I digital footprint datasets related to CNS development from the ENCODE project to test whether lncRNAs are under active regulation by transcription factors during CNS development . Using the size of cis-regulatory modules that are bound by TFs within the promoter regions as an indicator , we showed that a portion of lncRNAs have larger cis-regulatory modules than do random intergenic regions or even some TFs . Furthermore , a large number of lncRNAs were associated with △DHSs across the four ENCODE datasets studied . Neighboring △DHSs activated in different developmental contexts are enriched for specific TF sequence motifs that are known to be associated with the respective cellular functions . Taken together , the results of these analyses indicate that lncRNAs are dynamically controlled by TFs that specify CNS development . Investigating the molecular mechanisms underlying OPC formation is critical in understanding oligodendrocyte cell fate determination and for harnessing OPCs for cell-based therapies in regenerative medicine . Although many protein-coding and microRNA genes have been shown to play a critical role in oligodendrogenesis , functional characterization of lncRNAs during OPC fate determination has not been carried out systematically . A previous study employing a custom-designed microarray also showed that a number of known lncRNAs exhibit dynamic expression patterns during oligodendrocyte lineage specification , suggesting that they may play a role in neural stem cell fate decisions and oligodendrocyte lineage maturation [6] . However , incomplete annotation of lncRNAs hindered interrogation of novel lncRNAs and thus conclusions regarding functional significance were limited . Our study identified hundreds of lncRNAs that are regulated during NSC-to-OPC differentiation , and a substantial fraction of them are under the control of key TFs associated with oligodendrogenesis . Combining the TF binding site information , differential expression analysis , and enrichment analysis , lncRNAs that are highly enriched in OPC were identified as potential candidates involved in OPC fate determination . Depletion of the most enriched lncRNA in OPC , lnc-OPC , resulted in reduced OPC formation , suggesting that lnc-OPC plays an essential role in the circuitry that controls oligodendrogenesis . Intriguingly , ChIP-Seq and luciferase assay results suggested that OLIG2 binds to the upstream regulatory region of lnc-OPC and represses the transcription of lnc-OPC in NSC . Further studies will be necessary to elucidate the molecular mechanism through which the repression of lnc-OPC transcription is released during the formation of OPC from NSC and how lnc-OPC acts to modulate oligodendrogenesis . Previous studies have indicated that lncRNAs can serve as modular scaffolds for chromatin modifying complexes that modulate the epigenetic landscape during cell fate determination [14 , 52] . Such a mechanism could operate in lnc-OPC function . Although the focus of our study is on the lncRNAs that are up-regulated in OPC compared to NSC , the possibility of functional roles for down-regulated lncRNAs is not ruled out . Some of these lncRNAs may function as inhibitors of OPC formation and may help maintain the self-renewal ability of NSCs [12] . Inspecting the multiple alignments of lnc-OPC for 60 vertebrate species in the UCSC genome browser revealed remarkable sequence conservation of the lnc-OPC loci in placental mammals . These multiple alignments , as well as pairwise alignments , indicate two inserted segments located in the last intron of lnc-OPC in the mouse ( Fig 7 ) . These two insertions were not found in rat and other placental mammals , suggesting they are mouse-lineage specific and were introduced after the divergence of the lineages leading to mouse and rat . Interestingly , the inserted segments were highly expressed in ESCs . The symmetrical expression patterns of these two segments indicate that they may be repetitive elements . Indeed , we found that two MERVL ( mouse endogenous retroviral element ) transposable elements ( TEs ) perfectly matched the RepeatMasker annotation retrieved from the UCSC genome browser . About two out of three lncRNA transcripts cataloged in zebrafish , mouse , and human are estimated to contain at least one TE-derived sequence , whereas these sequences are rarely found in protein-coding genes [53] . TEs have been shown to contribute signals essential for the biogenesis of many lncRNAs by influencing their transcription initiation , splicing , or polyadenylation [53 , 54] . In particular , TEs embedded in introns can influence transcription of the host gene , causing upstream transcript polyadenylation , thus promoting the possibility of alternative polyadenylation or alternative splicing [55 , 56] . In this case , the two TEs are located in the last intron of lnc-OPC . There are several annotated shorter transcripts upstream of the TEs ( e . g . , XR_873837 . 1 , XR_873834 . 1 , and XR_873838 . 1 ) that overlap with the 5’ part of lnc-OPC . Therefore , the TE insertions may have caused alternative splicing or alternative polyadenylation and contributed to the evolution of transcript diversity at the 5330416C01Rik gene loci . Additionally , TEs , particularly LTR/ERVs ( LTR: long terminal repeat; ERV: endogenous retroviral ) , in the vicinity of lncRNA genes can be involved in the regulation of their transcription and contribute to tissue-specific expression profiles [53 , 57 , 58] . The two MERVL TEs have higher expression signals in various other cell types than in OPC . Thus , it is possible that these TEs may be involved in lnc-OPC evolution and transcriptional control . This hypothesis and the detailed mechanisms require further investigation . Multiple sclerosis is a chronic inflammatory demyelinating disease that involves the loss of oligodendrocytes . Subventricular zone ( SVZ ) -derived progenitors can be activated and repopulate the neural lesions after demyelination insults in multiple sclerosis . We analyzed the data from a previous study using experimental allergic encephalomyelitis ( EAE ) mouse models ( GSE47486 ) [59] . Interestingly , 5330416C01Rik gene ( lnc-OPC ) was found to be up-regulated more than seven fold in glial progenitor cells ( NG2+ cells ) from the SVZ of the EAE models compared to those from wild-type mice , suggesting that lnc-OPC may be involved in this disease ( S7 Fig ) . It should be noted that the lncRNA annotation used in our study was retrieved from multiple sources including GENCODE , RefSeq , Ensembl , lncRNAdb , and collections of lncRNAs identified by several other groups . We identified lncRNAs de novo from purified brain cell types and the coding potential of identified non-coding RNAs was evaluated using the Coding-Potential Assessment Tool ( CPAT ) . However , a recent study identified products of non-canonical translation from RNAs from mouse neurons that had been classified as non-coding [60] . Thus such a possibility should be considered . Another aspect that we should point out is that only polyadenylated RNAs were selected for RNA-seq library construction in the present study . However , polyA-selected RNAs account for only a portion of non-ribosomal , non-mtRNA [61] . Future investigation using non-polyadenylated RNAs for RNA-seq library construction could help to identify and reveal the functions of non-coding RNAs that are not polyadenylated . Altogether , our study produced a rich and unprecedented database of lncRNA expression by various purified cell types from mouse brain . The integrative analysis framework that we established in the present study can serve as a model for investigating functional lncRNAs in other cell types .
Mouse neural stem cells from the cortex of embryonic ( E14 . 5 ) CD-1 mice were purchased from R&D Systems . Briefly , cells were cultured and passaged every 2 days as monolayers in MEM/F12 medium containing Glutamax , non-essential amino acids , B27 , N2 supplement , 20 ng/ml EGF , and 20 ng/ml FGF . Cells were dissociated using Accutase ( Invitrogen ) and seeded onto poly-d-lysine-coated plates or dishes . NSCs were differentiated into OPCs in DMEM/F12 medium containing Glutamax , non-essential amino acids , BSA , B27 , N2 supplement , 20 ng/ml CNTF , and 40 ng/ml T3 . For immunostaining , NSCs were allowed to differentiate for 5 days . For mRNA extraction and qPCR , NSCs were allowed to differentiate for 3 days . RNA extraction was performed using TRIzol reagent , then DNase-treated RNA was reverse transcribed with the iScript cDNA Synthesis Kit according to the instructions from the manufacturer ( Bio-Rad ) . Quantitative Real-Time PCR ( qRT-PCR ) was performed using reactions prepared with the SYBR Green Master Mix ( Bio-Rad ) performed on the ABI PRISM 7900HT Sequence Detection System . Relative gene expression was calculated by the 2-ΔΔCT method using GAPDH as the reference gene . ChIP was performed using 107 mouse NSCs per reaction . Cells were dissociated by treatment with Accutase and cross-linked in 1% ( vol/vol ) formaldehyde for 10 min at RT with rotation . Then , 0 . 125 M glycine was used to quench the cross-linking reaction . Cells were pelleted and washed with ice-cold PBS . Next , nuclei were isolated and a Bioruptor sonicator ( Diagenode ) was used to shear chromatin DNA . Either 5 μl of OLIG2 antibody ( AB9610 , Chemicon ) or 5 μl of normal rabbit IgG ( #2729 , Cell Signaling ) was added to Dynabeads Protein A ( Invitrogen ) beads and incubated for 3 h at 4°C with rotation . Then , the Dynabeads-antibody complexes were incubated with sheared chromatin DNA overnight at 4°C . After immunoprecipitation , the precipitated complex was treated with RNase A and Proteinase K , and incubated at 65°C overnight to reverse crosslinks . Primers were designed and qRT-PCR experiments were performed . ChIP-qPCR data was normalized either by the Percent Input method or relative to the IgG control . The ChIP-Seq library was constructed by using DNA SMART ChIP-Seq Kit according to the manufacturer's instructions ( Clontech ) and was sequenced on the Illumina HiSeq 2000 Sequencer . Input chromatin sample was prepared in parallel with OLIG2 ChIP sample . The generated reads were mapped to the mouse genome mm10 using Bowtie version 0 . 12 . 7 and MACS version 1 . 4 . 2 was used to call peaks for ChIP-Seq data [62] . The output from MACS was filtered using the following more stringent criteria: ( 1 ) p value cutoff <10−9; ( 2 ) fold enrichment > five fold ( 3 ) tag number >20 . RNA-seq was performed using the same procedures as described previously [15 , 20] . At least three shRNAs against lnc-OPC were separately cloned into a pLKO . 1 lentiviral vector and recombinant lentivirus was produced in 293FT cells . NSCs were transduced with recombinant lentivirus after 24 h of culture on poly-d-lysine pre-coated plates . Non-infected cells were eliminated using fresh culture medium containing 0 . 5 μg/ml of puromycin 1 day after infection . Infected cells were proliferated for 3–4 days and subsequently used for the assessment of knockdown efficiency and oligodendrocyte lineage commitment . For the oligodendrocyte surface membrane antigen O4 , cells were cultured in 24-well plates and were stained with the primary antibody O4 produced by hybridoma ( from Dr . Qilin Cao , personal communication ) at RT for 1 h . Cells were then fixed in 4% paraformaldehyde ( vol/vol ) and subsequently stained using Alexa Fluor 594 goat anti-mouse IgG ( H+L ) antibody . DAPI was used to stain the nuclei . Three independent experiments were carried out and cell number counts were obtained from 10 randomly selected fields in each experiment . Puromycin-selected NSC cells were incubated in medium with BrdU labeling reagent ( Invitrogen ) for 60 min and subsequently fixed in 70% ethanol for 20 min at RT . After three washes with PBS , cells were treated with 1 . 5 M HCl for 30 min . Cells were then immunostained with BrdU mouse antibody ( Cell Signaling ) , Alexa Fluor 488 goat anti-mouse IgG ( H+L ) antibody ( Invitrogen ) , and DAPI . Three independent experiments were carried out and the results from 10 random fields for each experiment were obtained . The lnc-OPC -luciferase reporter constructs were created by cloning 1 . 3- to 2 . 5-kb fragments upstream of the lnc-OPC transcription start site into the KpnI and XhoI multiple cloning sites of the pGL4 . 11 vector . The constructs were cotransfected with pmaxGFP or the mouse OLIG2 expression plasmid pCMV-SPORT6 . 1-OLIG2 using Lipofectamine 3000 ( Invitrogen ) in 293FT cells . Cells were harvested and examined for luciferase activity 2 days after transfection using a Luciferase Assay System kit ( Promega ) . Luciferase activity was measured using a Tecan Microplate Reader infinite M1000 and was normalized by the protein concentrations of each sample . RNA-Seq analysis was performed as described previously [15] . The Tuxedo suite was used for read mapping and transcript assembly [22] . RNA-Seq data from eight brain cell types and seven non-brain tissues were used for analysis ( pericytes were excluded because of their relatively lower purity ) . RNA-Seq data from seven non-brain mouse tissues ( thymus , testis , kidney , liver , lung , spleen , and heart ) were obtained from the Mouse ENCODE project . Data were downloaded from GeneExpression Omnibus ( GEO accession number GSE36025 ) and processed using the same pipeline as for brain cells . A previously described ‘guilt-by-association’ method was adopted for function prediction [4] . Weighted Gene Co-expression Network Analysis ( WGCNA ) R package was used to identify gene co-expression modules [63] . GSEA analysis was performed using the GSEA command line executable file downloaded from http://www . broadinstitute . org/gsea/index . jsp [26] . Potential TF-binding sites were determined using Find Individual Motif Occurrences ( FIMO ) version 4 . 6 . 1 , setting the p-value threshold to 10−5 and using defaults for other parameters [64] . The Bayesian method , CENTIPEDE , was adopted to infer the binding status of TF binding sites in the DNase-DGF data [34] . In addition to DNase-DGF , we performed motif searches in the promoter regions annotated by the Mouse ENCODE project upstream of the lncRNAs of interest . We defined dynamic DHSs as the DHSs that are enriched > three fold under one condition compared to average signals across all conditions , with signal > 10 under the enriched condition . For TFs that have public ChIP-Seq data , but that are not present in motif databases , we conducted de novo motif discovery using the MEME-ChIP included in the MEME suite [65] . We used the Homer software suite to test whether TF motifs are enriched in dynamic DHSs that are activated under different conditions . A binominal distribution was used to calculate p-values for the significance of enrichment [66] . DEGs were called using the DESeq R package [67] . Only genes with > two fold change in expression and FDR < 0 . 05 were considered as DEGs . The UCSC genome browser was used for inspecting multiple alignments and RMSK annotations . The Epigenetics browser was used to depict the track plots [68] . | Between 70 and 90% of the mammalian genome is transcribed at some point during development; however , only < 2% of the genome is associated with protein-coding genes . Emerging evidence suggests that long non-coding RNAs ( lncRNAs; > 200 bp ) play important roles in cell fate determination . In the present study , we broadened the lncRNA catalog by ab initio reconstruction of the transcriptomes of purified mouse cortex cell populations . More than 5000 lncRNAs were detected in the brain cell types studied . Predicting lncRNA functions using a ‘guilt-by-association’ approach revealed potential functions of lncRNAs in Central Nervous System development . Additionally , we analyzed transcription factor occupancy in the upstream regulatory regions of the lncRNAs . By integrating differential gene expression and transcription factor occupancy information , lncRNAs that are likely involved in oligodendrocyte precursor cell formation were identified . Loss-of-function experiments confirmed that the top candidate , lnc-OPC ( long non-coding RNA in OPC ) , significantly reduces OPC differentiation from NSCs . Interestingly , lnc-OPC is up-regulated in glial progenitors of mouse models for multiple sclerosis . Our results demonstrated the role of lncRNA in the context of oligodendrocyte cell fate determination , and provided an extensive resource and a powerful analysis framework for future functional investigations of lncRNAs in CNS cell types . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Comprehensive Identification of Long Non-coding RNAs in Purified Cell Types from the Brain Reveals Functional LncRNA in OPC Fate Determination |
Chagas disease also known as American trypanosomiasis is caused by the protozoan Trypanosoma cruzi . Over the last 30 years , Chagas disease has expanded from a neglected parasitic infection of the rural population to an urbanized chronic disease , becoming a potentially emergent global health problem . T . cruzi strains were assigned to seven genetic groups ( TcI-TcVI and TcBat ) , named discrete typing units ( DTUs ) , which represent a set of isolates that differ in virulence , pathogenicity and immunological features . Indeed , diverse clinical manifestations ( from asymptomatic to highly severe disease ) have been attempted to be related to T . cruzi genetic variability . Due to that , several DTU typing methods have been introduced . Each method has its own advantages and drawbacks such as high complexity and analysis time and all of them are based on genetic signatures . Recently , a novel method discriminated bacterial strains using a peptide identification-free , genome sequence-independent shotgun proteomics workflow . Here , we aimed to develop a Trypanosoma cruzi Strain Typing Assay using MS/MS peptide spectral libraries , named Tc-STAMS2 . The Tc-STAMS2 method uses shotgun proteomics combined with spectral library search to assign and discriminate T . cruzi strains independently on the genome knowledge . The method is based on the construction of a library of MS/MS peptide spectra built using genotyped T . cruzi reference strains . For identification , the MS/MS peptide spectra of unknown T . cruzi cells are identified using the spectral matching algorithm SpectraST . The Tc-STAMS2 method allowed correct identification of all DTUs with high confidence . The method was robust towards different sample preparations , length of chromatographic gradients and fragmentation techniques . Moreover , a pilot inter-laboratory study showed the applicability to different MS platforms . This is the first study that develops a MS-based platform for T . cruzi strain typing . Indeed , the Tc-STAMS2 method allows T . cruzi strain typing using MS/MS spectra as discriminatory features and allows the differentiation of TcI-TcVI DTUs . Similar to genomic-based strategies , the Tc-STAMS2 method allows identification of strains within DTUs . Its robustness towards different experimental and biological variables makes it a valuable complementary strategy to the current T . cruzi genotyping assays . Moreover , this method can be used to identify DTU-specific features correlated with the strain phenotype .
Chagas disease also known as American trypanosomiasis affects around 6–7 million people especially in Latin America [1] . The etiologic agent of Chagas diseases is the protozoan Trypanosoma cruzi [2] that infects several mammalian hosts and is primarily transmitted through the contamination with feces of triatomine bugs . Besides , congenital , blood transfusions , transplants and ingestion of contaminated foods represent other ways of infection [3] . Chagas disease is characterized by an acute and chronic phase . The acute phase lasts a few weeks and may present mild symptoms such as fever and swelling around the site of infection[4] . The chronic phase is in general lifelong and asymptomatic . However , 20–30% of patients develop cardiac or gastrointestinal complications[3] . T . cruzi is highly genetically diverse . In order to standardize the nomenclature facilitating the communication among scientists , T . cruzi strains were divided into six ( Tc I-VI ) discrete typing units ( DTUs ) plus Tcbat , a novel strain associated with bats . Each group represents a set of isolates that are genetically similar and can be identified by common immunological , biochemical , pathological and molecular markers [5] . T . cruzi strains characterization is extremely important to understand different epidemiological and pathological characteristics such as geographical distribution and clinical outcomes [6] . Several techniques have been introduced to improve the genotyping . In particular , the genetic diversity of T . cruzi was first recognized by multilocus enzyme electrophoresis ( MLEE ) and restriction analysis of kinetoplastid DNA minicircles [7 , 8] . Currently , the methods used for typing of T . cruzi strains are based on polymorphism of the mini-exon gene ( Spliced Leader ) and the 24Sα and 18S ribosomal RNA [9] . Other assays involve the analysis of complex electrophoretic patterns generated by restriction polymorphisms of PCR amplified genomic DNA [10–12] . These methods are able to discriminate T . cruzi strains , but they are work and time consuming and the interpretations of the results may be misleading . Multiplex Real-Time PCR Assay Using TaqMan Probes proved useful to determine DTUs of cultured T . cruzi , vector and blood samples from patients in acute infection [10 , 12 , 13] . Proteomics methods using mass spectrometry have emerged as a powerful strategy to discriminate bacteria and have been established as a valuable alternative to DNA-based bacterial identification . Indeed , MALDI-TOF MS generates an intact protein profile that is compared to a MS database of known species . This method is easy to perform , rapid and provides accurate results and has been introduced as routine in most hospitals [14] . However , this method is currently not able to distinguish at the strain level . Shotgun sequencing of peptides , derived from enzymatically digested proteins , using LC-MS/MS in combination with database search of in silico digested proteins derived from publicly available protein sequence has been proposed as a method to discriminate bacterial strains such as Helicobacter pylori and Yersinia persis [15] . This approach has great discrimination power compared to MALDI-TOF MS since it allows the identification of thousands of peptides with higher dynamic range . However , this method suffers from lack of suitable databases since many species have not been sequenced . Recently , a novel method based on MS/MS spectral matching has been shown to identify the blood meal of ticks ( Ixodes scapularis ) [16] and has also been applied to discriminate E . coli strains from different isolates [15] . Few studies have explored the possibility of using mass spectrometry as a diagnostic tool for Chagas disease . Using mass spectrometry , Ndao M . et al . [17] were able to identify full-length Apo1 as a serum biomarker of chronic Chagas disease patients using a SELDI-TOF approach . Despite several proteomic studies to understand the molecular features of T . cruzi in different biological conditions [18–22] , to date , there is no report on the use of mass spectrometry for assaying various T . cruzi strains . A pioneer work developed a peptide identification-free shotgun proteomics workflow to trace the vertebrate host that a tick ( Ixodes scapularis ) was feeding on [16] . This workflow used MS/MS spectral matching . The same strategy , named UNID , was used to profile bacterial strains [15] . In this study , we developed a platform to discriminate T . cruzi strains using MS/MS peptide spectral matching ( Tc-STAMS2 ) . The method was able to discriminate T . cruzi strains from different origins with high sensitivity and accuracy . The method was robust towards sample preparation and instrumental parameters , such as peptide purification , chromatographic gradient time , peptide fragmentation techniques and MS instruments . The MS/MS spectra were subjected to database search . More than 4000 proteins were identified in the combined six DTUs strains analyzed . A total of 1096 proteins were differentially expressed between the six DTUs and multivariate analysis allowed the discrimination of T . cruzi strains using the quantitative MS signal . In conclusion , this study describes a mass spectrometry-based method to discriminate T . cruzi strains . The Tc-STAMS2 method represents a proof-of-concept of an alternative strategy to DNA-based T . cruzi genotyping . Further studies are needed to show its applicability to biofluids in clinical isolates .
Epimastigote forms of T . cruzi cultivated in LIT ( Liver Infusion Tryptose ) supplemented with 10% fetal bovine serum [23] at 28°C of exponential culture phase were employed in the present study . DTU classification of all T . cruzi strains were confirmed by sequencing [24] . Only validated DTUs strains were used in this study ( Table 1 ) . T . cruzi cells from the Sylvio X10 cl1 ( DTU-I ) strain were collected in the exponential and stationary phase and processed as described below . Two biological replicates for the stationary ( St_1 and St_2 ) and exponential ( Exp_1 and Exp_2 ) growth phase were analyzed . Protein extraction and digestion . Epimastigote cells ( 5x108 ) were washed three times in phosphate buffered saline ( PBS ) , pH 7 . 2 ( 8 , 000g for 10 minutes at room temperature ) , re-suspended in 400 μL of lysis buffer ( 7M urea , 2M thiourea , 1 mM DTT and protease inhibitors ( Amersham ) and incubated under stirring for 30 minutes to solubilize the proteins . Proteins were reduced with 10 mM DTT ( DL-Dithiothreitol–Sigma-Aldrich ) , alkylated with 40 mM iodoacetamide ( Sigma-Aldrich ) , digested with trypsin ( Promega ) in the ratio 1:50 ( μg trypsin/μg protein ) in 50 mM ammonium bicarbonate solution at 37°C overnight . The reaction was stopped with 1% formic acid ( less than pH 3 ) and then the sample was desalted with C18 columns ( StageTips ) . Four biological replicates were prepared for each DTU . Blind test samples A ( DTU-III ) and B ( DTU-I ) were prepared using a minimum of three biological replicates according to the protocol described above . Peptide desalting was performed in acid ( acid 1 and acid 2 ) and basic ( basic 1 and basic 2 ) conditions . In particular , for the acid purification tryptic peptides were acidified with 0 . 1% TFA ( pH 3 ) and loaded onto an acid-activated StageTip microcolumn before being eluted with 50% acetonitrile: 0 . 1% TFA . For the basic purification , tryptic peptides were dissolved in 0 . 1% ammonia water ( pH 10 ) and loaded onto a base-activated StageTip before being eluted with 50% acetonitrile: 0 . 1% ammonia water . The eluted peptides were lyophilized and analyzed by liquid chromatography tandem mass spectrometry ( LC-MS/MS ) . Peptides were separated by Reprosil-Pur C18-AQ column ( 3μm; Dr . Maisch GmbH , Germany ) using Easy nano-LC HPLC ( Proxeon , Odense , Denmark ) . The HPLC gradient was 0–34% B solvent ( A = 0 . 1% formic acid; B = 90% ACN , 0 . 1% formic acid ) in 70 min at a flow of 250 nL/min . The MS analysis was performed using the LTQ-Orbitrap Velos ( Thermo Scientific , Bremen , Germany ) . The mass range was 400–1500 m/z at a resolution of 30 , 000 at 400 m/z for a target value of 1e6 ions . For each MS scan , collision induced dissociation ( CID ) fragmentation was performed on the 20 most intense ions in the linear iontrap . The parameters for data acquisition were: activation time = 15 ms , normalized energy = 35 , Q-activation = 0:25 , exclusion = available with repeat count 1 , exclusion duration = 30s and intensity threshold = 30 . 000 , target ions = 2e4 [25] . All raw data have been submitted to PRIDE archive ( https://www . ebi . ac . uk/pride/archive/ ) , project accession: PXD008088 . The robustness of the Tc-STAMS2 method was tested using different parameters: 1 ) sample amounts , 2 ) chromatographic gradients and 3 ) MS fragmentation techniques . For the different sample amounts , 0 . 5μg and 1μg , were loaded onto the analytical column before MS analysis and named low and high , respectively . The chromatographic elution time was set to 20 , 70 and 130 min from 0–34% B solvent at 250nL/min . CID fragmentation was used to develop the Tc-STAMS2 method . Higher energy collision induced dissociation ( HCD ) was evaluated as alternative peptide fragmentation type on tryptic peptides separated on a 70min chromatographic gradient . For the HCD fragmentation , each MS scan was acquired at resolution of 30 , 000 FWHM followed by 7 MS/MS scan of the most intense ions with an activation time of 0 . 1 ms and normalized collision energy of 35 . The spectral library for each DTU was developed on a LTQ-Orbitrap Velos Pro instrument ( Thermo Fisher Scientific ) located in the PR group , Department of Biochemistry and Molecular biology , University of Southern Denmark . All the other tests for testing and validating the method were performed on a LTQ-Orbitrap Velos Normal at the Biomass mass spectrometry facility ( São Paulo , Brazil ) . MS/MS spectral library generation and spectral matching . MS/MS spectral library generation and spectral matching were performed using the SpectraST software ( version 4 . 8 ) as previously described [15 , 16 , 26] . In particular , the LC-MS/MS acquired spectra were converted to an open format ( mzXML ) by MSconvert [27] , forming part of the software suite offered by TPP ( Trans-Proteomic Pipeline ) [28] . SpectraST ( version 4 . 8 ) was used to build the reference spectral library and perform MS/MS spectral matching [26] . The reference spectral library was generated with three raw files for each one of the six DTUs and one raw file for each DTU was compared against constructed reference library . The first step for the reference spectral library generation involves the application of a threshold at which MS/MS spectra originated from the same peptide precursor ion are combined to create the consensus spectrum . Moreover , low quality spectra are excluded from the library [26] . To determine the spectrum of similarity between the query spectrum and the reference library , SpectraST uses the Spectral Dataset Similarity ( SDSS ) function [16 , 29] . In particular , the unique dot product SDSS , abbreviated as “score” along the text , was chosen as T . cruzi DTU strain discrimination function and reported [16] . The statistical confidence in the identification of the correct DTU is made by data bootstrap [16] . All the spectral matching experiments reported below had a bootstrap of 1 unless reported . DiagnoProt software was used also to match the MS/MS spectra of unknown samples against a database of genotyped T . cruzi strains [21] . Default parameters for creating the spectral database were used: similarity threshold 0 . 70 , precursor tolerance 4 . 50; activation type CID; minimum number peaks: 50; minimum relativity intensity: 0 . 01; minimum retention time 10 . 00; bin offset 0 . 40; bin size: 1 . 0005; minimum bin m/z 200 . 00; maximum bin m/z 1700 . 00 . The spectral database was used to match the identity of each unknown sample . The raw LC-MS/MS files were analyzed using: Proteome Discoverer , MaxQuant and the Trans-Proteomic Pipeline . Proteome Discoverer v2 . 1 ( Thermo Scientific ) was used with the T . cruzi database using Mascot and Sequest . The searches in the database were conducted with the following parameters: precursor mass tolerance of 20 ppm; MS/MS mass tolerance 0 . 5 Da ( CID data ) . Trypsin was selected as enzyme and carbamidomethyl cysteine as fixed modification . The variables modifications were oxidation of methionine and deamidation ( NQ ) . Shared peptide sequences were grouped as grouped accessions proteins . The False Discovery Rates ( FDR ) was calculated using the algorithm Percolator with q equal or less than 0 . 01 . Protein FDR was calculated in the Proteome Discoverer software and kept below 1% . The raw files were also processed using the MaxQuant [30] version 1 . 2 . 7 . 429 and the MS/MS spectra were searched using the Andromeda search engine [31] against the Uniprot T . cruzi Protein Database ( release July 11 , 2017; 51 , 738 entries ) . The initial maximal allowed mass tolerance was set to 20 ppm for precursor and then set to 4 . 5 ppm in the main search and to 0 . 5 Da for fragment ions . Enzyme specificity was set to trypsin with a maximum of two missed cleavages . Carbamidomethylation of cysteine ( 57 . 02 Da ) was set as a fixed modification , and oxidation of methionine ( 15 . 99 Da ) , deamidation ( NQ ) and protein N-terminal acetylation ( 42 . 01 Da ) were selected as variable modifications . Bioinformatics analysis was performed using the software Perseus v . 1 . 5 . 2 . 6 [30] available in the MaxQuant environment and reverse and contaminant entries were excluded from further analysis . Protein FDR was calculated in the MaxQuant software and kept below 1% . Label Free Quantification ( LFQ ) intensity values were considered to relatively compare the abundance of proteins present in the different DTUs . Trans-Proteomic Pipeline software suite was also used to search raw files converted to mzXML [13] . The mzXML files were searched by the Comet search algorithm embedded into the TPP platform [14] . Peptide and protein FDR was estimated using the PeptideProphet and ProteinProphet algorithm embedded in the Trans-Proteomic Pipeline [32] . Identifications with less than 1% FDR were kept . Raw data from human placental tissue [33] and T . vivax ( Meta , BSF1 and EP1 ) [22] were obtained from the public MS spectra databank PRoteomics IDEntifications ( PRIDE ) and used for the negative control test . Label-free quantified peptides/proteins were analysed by the Perseus software [18] . Significantly regulated features with a p value less than 0 . 05 corrected with the Bonferroni post-hoc test were used to cluster the different DTUs . Hierarchical clustering of significantly regulated proteins/peptides was performed using the Z-score calculation on the log2 intensity values and it was represented as a heat map . The Principal Component Analysis ( PCA ) was performed using the same procedure described above in Perseus software . In addition , for generation of Venn diagram we used Venn Diagrams ( http://bioinformatics . psb . ugent . be/webtools/Venn/ Bioinformatics & Evolutionary Genomics ) .
In this study , the combination of mass spectrometry and computational approaches was used to develop a method for T . cruzi DTU discrimination , named Tc-STAMS2 . A schematic overview of the T . cruzi DTUs identification using MS/MS spectra from tryptic peptides is summarized in Fig 1 . A reference spectral library was built using a total of 586513 unique MS/MS spectra of tryptic peptides derived from three raw MS files of each one of the six T . cruzi strains ( Fig 1A ) . Each T . cruzi strain was processed and acquired in four biological replicates . MS/MS spectra acquired from three replicates of each DTU were used to build the reference spectral library using SpectraST software [26] . Following the construction of the reference MS/MS library , proteins from unknown T . cruzi strain samples were extracted and digested with trypsin before being analyzed by nLC-MS/MS ( Fig 1B ) . A LC-MS profile of four replicates of DTU-I is reported in S1 Fig and the Pearson correlation score indicates high similarity between the different runs . The LC-MS chromatographic profile of the tryptic peptides belonging to the six T . cruzi strains shows high similarity between the different DTUs , S2 Fig . MS/MS spectra from different DTUs were subjected to spectral matching comparison with the library using SpectraST software [19] ) . The identification was made by finding the reference library with the highest similarity to the sample spectral dataset , in this case , the different DTUs , to be tested . Unique dot product SDSS score was used to provide a quantitative similarity measure between two spectral datasets ( Fig 1C ) [16] . MS/MS spectra were searched against the T . cruzi proteome database as described below ( Fig 1D ) . Based on the similarity scores between the MS/MS spectra from an unknown sample with the mass spectral library , the Tc-STAMS2 was able to differentiate and accurately identify the different DTUs . We observed that the score values were between 0 . 75 to 0 . 86 for true matches and close to 0 in unmatched cases , as shown in Table 2 . The LC-MS chromatograms obtained for each replicate and for each DTU ( S1 Fig and S2 Fig ) showed high similarity; however , the developed method was capable of differentiating and identifying each of them . In order to rule out the possibility that different growth phases could influence the assignment of the algorithm , Sylvio X10/1 ( DTU-I ) epimastigotes in the exponential and stationary phase were collected . Independently of the growth phase , the algorithm was able to assign it to the correct DTU , Table 3 . This demonstrates that the identification method is not affected by the growth phase of the parasite . Moreover , the performance of the Tc-STAMS2 spectral matching approach were tested to correctly identify MS/MS data sets from: 1 ) a T . cruzi strain that is known to belong to DTU-VI ( CL14 ) [5] but was not included in the spectral library , 2 ) from a species phylogenetically related , such as Trypanosoma vivax and 3 ) from species with completely distant genome ( ex: human , E . coli , mouse ) . For this analysis , a new MS/MS library was constructed , using the MS/MS spectra from the six DTUs , including T . cruzi CL14 and T . vivax in metacyclic stage ( meta1 e meta2 ) . Firstly , MS/MS spectra from T . cruzi strains CL14 were compared with the library and the similarity score matched to the CL14 ( score = 0 . 417 ) . Interestingly , although the similarity scores of CL14 with DTUs I to V were close to zero , the similarity between CL14 and DTU-VI was comparatively high ( score = 0 . 133 ) , indicating that many spectra MS/MS of CL14 are shared with DTU-VI ( CL Brener ) , as shown in Fig 2 . It should be noted that the LC-MS chromatographic profiles of the CL14 and CL Brener strains have very high similarity , S3 Fig . However , the Tc-STAMS2 was able to clearly differentiate between the two strains within the same DTU . Subsequently , MS/MS spectra from three different life stages of T . vivax , Meta3 –metacyclic phase , BSF1 –bloodstream phase and EP1 –epimastigote were compared using the mass spectral library . Based on the similarity scores , the Tc-STAMS2 was able to correctly identify these samples to T . vivax , Fig 2 . In order to test the method with negative control , MS/MS spectra from unrelated T . cruzi species such as human , mouse and E . coli were compared to the library . For these samples , the similarity scores were close to zero , indicating that the similarity scores found between two MS/MS datasets is library-specific and not random , Fig 2 . Moreover , we also evaluated the ability of this strategy to provide correct identification from an independent sample ( blind test ) , which was collected and processed at different days or under different conditions . In particular the different datasets were obtained in: 1 ) inter-laboratory studies , 2 ) different sample preparation strategies , 3 ) different LC gradients and 4 ) different MS fragmentation methods . To assess the robustness of the Tc-STAMS2 platform , another batch of T . cruzi strains were processed and acquired using similar chromatographic and MS conditions as described in Fig 1 in an inter-laboratory study perspective . Indeed , the mass spectra library was built with data acquired in the PR group in Odense , Denmark and the blind samples were acquired in the CEFAP mass spectrometry facility in São Paulo , Brazil . Although the instrument type and conditions were similar , a different chromatographic profile was obtained , S4 Fig . However , the biological duplicate unknown samples A1 and A2 from DTU-III matched correctly to DTU-III , Fig 3 . The unknown sample B from the DTU-I also showed higher similarity scores with the MS/MS spectra library from DTU-I . The sample B was evaluated on different parameters: 1 ) sample preparation conditions such as acid and basic peptide desalting , 2 ) different chromatographic gradient such as 20 min , 70 min , 130 min , 3 ) different fragmentation methods , such as CID or HCD with a chromatographic gradient of 70 min and 4 ) different sample amount injected into the LC column . Even considering all these technical sources of variation , the similarity scores continued to match correctly to DTU-I showing the robustness of the Tc-STAMS2 towards different experimental conditions . In addition , another MS/MS spectral library search software platform , DiagnoProt , was used instead of SpectraST [21] . DiagnoProt was able to differentiate the different DTUs and to associate the CL14 strain with the DTU -VI group as shown for the SpectraST software , Table 4 . Due to that , two different spectral library search software could be implemented in the Tc-STAMS2 pipeline and used to identify T . cruzi strains , Tables 2 and 4 . Furthermore , we tested the Tc-STAMS2 method using 13 strains belonging to six DTUs such as Sylvio X10 cl1 , Sylvio X10/4 and G strains for DTU-I , Y and Esmeraldo strains for DTU-II , M6241 cl6 and 3869 strains for DTU-III , CanIII cl1 and José Júlio strains for DTU-IV , MN cl2 and NR cl3 strains for DTU-V and CL Brener and CL14 for DTU-VI . The Tc-STAMS2 method was able to correctly discriminate intra-DTU strains with high accuracy ( Fig 4 ) . Indeed , each strain matched correctly the corresponding one included into the database . Moreover , we assembled a MS/MS spectral library using two biological replicates of the 13 T . cruzi strains associated to six DTUs ( Fig 5 ) . One independent replicate of each strain was used to match the spectral library using the Tc-STAMS2 method . Each strain matched correctly to the corresponding DTU with high accuracy ( Fig 5 ) . Database search was also performed to evaluate the similarity among the DTUs using peptide and protein identification results . Three database search platforms were used ( MaxQuant , TPP and Proteome Discoverer ) . From four replicates of each DTU more than 7000 peptides and 4000 proteins were identified ( S5 Fig , S1 Table , S2 Table and S3 Table ) . DTU-I and DTU-VI had the highest number of identifications due to the protein database used for the search . Indeed , Sylvio ( DTU-I ) and CL Brener ( DTU-VI ) are the two T . cruzi strains whose genome has been sequenced and their proteome annotated and deposited in the Uniprot database . Interestingly , only 30% of the MS/MS spectra were assigned , leaving behind a wealth of information for T . cruzi strain discrimination ( S4 Table ) . Analysis of variance ( ANOVA p<0 . 05 followed by Benjamin-Hochberg FDR correction ) was applied for the log2-transformed protein or peptide intensities previously identified using MaxQuant . A total of 1096 proteins and 6130 peptides showed significant difference in abundance among the six DTUs ( S5 Table ) . The differentially expressed peptides and proteins were subjected to clustering analysis and visualized as heat-maps , S6A and S6B Fig . CL14 clustered together with CLBrener ( DTU-VI ) , indicating high similarity in the protein and peptide expression profile . In addition , principal component analysis ( PCA ) , which was applied in the differential expressed proteins , were able to discriminate the six DTUs showing a DTU-specific quantitative proteome repertoire ( Fig 6 ) . Interestingly , CL14 and CLBrener , strains belonging to the DTU-VI , were also found close to each other , confirming the Euclidean clustering result obtained previously . Additionally , to check the proteomic data reproducibility , a correlation analysis ( R squared ) between biological replicates from each DTU or between the different DTUs was performed using the log2-normalized intensities . As shown in S7 Fig , high correlation values were observed among replicates ( R squared>0 . 9 ) . Interestingly , when we compared different DTUs ( S8 Fig ) , the R squared dropped to 0 . 6–0 . 7 , but high correlation was observed between CL14 and DTU-VI ( R squared = 0 . 83 ) , confirming once more the similarity between these two DTUs ( S8 Fig ) .
Many studies have employed MS-based techniques to identify organisms , such as bacteria [34] . Önder et , al . pioneered the use of MS/MS spectral libraries to precisely identify which animal the tick Ixodes scapularis was fed even if the feeding occurred months earlier [16] . Other strategies combined genetic information in conjunction with MS-based peptide identification for the correct assignment of microorganisms [35] [36 , 37] . More recently , Shao et al . demonstrated that it is possible to identify E . coli strains with only the MS/MS fragmentation spectra , with no need for peptide identification ( Shao et al . 2015 ) . In the present study , we describe a genome-free , MS/MS spectral-matching methodology designed to identify different T . cruzi DTUs , named Tc-STAMS2 . Firstly , a peptide MS/MS spectral library using three replicates from each of the six DTUs was built . The fourth replicate was used to test the ability of the method to differentiate each DTUs using the reference library . As shown in Tables 2 and 4 and Figs 5 and 6 , this approach was able to differentiate strains belonging to similar and different DTUs . A unique assignment to the correct DTU was achieved . The method was tested using a dataset of peptide MS/MS spectra obtained from different growth conditions of T . cruzi and no bias was detected . The next step was to test the MS/MS spectral library against phylogenetically related species such as T . vivax and distant organisms such as E . coli , mus musculus and homo sapiens ( Fig 2 ) . For such a test , another spectral library database was built with the same DTUs and MS/MS spectra from the CL14 strain and T . vivax . Interestingly , it is clear that even with thousands of fragmentation spectra , the scores obtained when comparing samples of human organism or E . coli with the library were close to zero , demonstrating the specificity of the method in identifying only samples of the species/strain , whose MS/MS spectrum is present in the library . We also showed that the spectral matching method is robust even with inter and intra-laboratory source of variations using similar MS instruments , but from different laboratory and even performing changes in the sample preparation , chromatography and fragmentation method , it was still possible to correctly identify samples from DTU-III and DTU-I , as shown in Fig 3 . As expected , we also observed that the similarity score is dependent on the number of MS/MS spectra acquired . The longer the gradient time used , the greater the separation capacity of peptides prior to MS/MS analysis and the larger the number of acquired spectra , resulting in higher scores for the same sample when compared to the library . Using different fragmentation methods we also observed that CID provides scores higher than HCD when using 70 min gradient time . Although HCD provides high resolution MS/MS spectra CID provides faster MS/MS sequencing , thus generating more spectra that can be match with the spectral library . In addition , fragmentation spectra were also subjected to database search analysis for peptide and protein identification using the MaxQuant software , TPP and Proteome Discoverer . In general , the number of identifications of proteins and peptides were reproducible and consistent among different database search software . The larger number of proteins and peptides were observed for DTU-I and VI . Although the number of annotated proteins in UniProt database is greater for DTUs I and VI , the number of proteins identified to DTUs V , III and VI , was not significantly smaller . Due to that , a MS-based proteomic approach can be used to quantitatively compare the protein expression of different DTUs , even with differences in genome annotation among them and use these information’s to identify DTU-specific pathways correlated with the strain phenotype . Moreover , it was possible to cluster together DTU-VI and CL14 using the differential expressed proteins or peptides given by ANOVA test , S6 Fig . This result validates which is already known in the literature , where the CL14 belong to DTU-VI [5] and also validates the results obtained with the spectral matching , where the similarity score between CL14 and DTU-VI was higher compared to the other DTUs . Moreover , PCA analysis was also able to discriminate the six DTUs and determine the similarity between the DTUVI and CL14 , as seen by the close proximity between these two groups . Clustering analysis using protein identification in different DTUs was also performed by Telleria et , al . , [38] however only 261 ( experiment 1 ) and 172 ( experiment 2 ) 2DE protein spots were considered for this analysis . In our study , the clustering analysis was performed using 1096 proteins differentially expressed among DTUs , increasing the number of features to build a robust T . cruzi clustering ( S6 Fig ) . The Tc-STAMS2 strategy presented here is robust , accurate , easy to perform and completely automated . Dworzanski , J . P . [37] and Shao , W . et al . [15] used a similar methodology for identification of bacteria , however this is the first time that spectral matching is applied to discriminate different T . cruzi DTUs . The transmission of Chagas disease through blood transfusion and organ transplant , besides the triatomine vector , poses several public health challenges . Due to that , novel methods identify and characterize DTUs can offer opportunities to understand the DTUs diversity , link with their phenotype and provide another tool for molecular epidemiology . Recently , MALDI-based strategy was developed for the direct identification of trypanosomatids based on the MS profile , named DIT-MALDI TOF [39] . However , this method does not have the resolution to discriminate DTUs using the MS profile . In this study we present a T . cruzi strain typing based on MS/MS spectral matching . The Tc-STAMS2 method is robust , sensitive and powerful and it is based on the identification of peptides from their MS/MS spectra . The method could be used to complement other already established methods . The proposed method can help in the research of epidemiology to identify T . cruzi strains only with the use of fragmentation spectra without the need for genomic data . In this study , we used the epimastigote form of T . cruzi to implement the Tc-STAMS2 method . However , quantitative proteomic analysis of T . cruzi epimastigote and trypomastigote life stages have shown distinct protein expression profile [40 , 41] . Due to that , more studies are needed to confirm the specificity of the Tc-STAMS2 method using other T . cruzi stages . Since MS/MS spectra are available to the research community further data mining is possible in order to improve our understanding on the biology of the different T . cruzi strains . The methodology shown in this study will provide a complementary tool to the current nucleic-based testing and have the possibility to be extended to other parasitic diseases . | Chagas disease is one of the most important neglected diseases with an estimated number of 6–7 million infected individuals , the majority living in Central and South America . The Trypanosoma cruzi ( T . cruzi ) protozoan parasite is the etiological agent of Chagas disease . T . cruzi is highly genetically diverse and a new nomenclature assigned each strain to seven genetic groups ( TcI-TcVI and Tcbat ) , named Discrete Typing Units ( DTUs ) , based on their biochemical , immunological and phenotypical characteristics . T . cruzi DTUs have been correlated to diverse clinical outcomes highlighting the importance of molecular epidemiological screens . Despite the development of T . cruzi typing methods based on genetic signatures , each method presents its own advantages and challenges . The work presented here shows the application of mass spectrometry for Trypanosoma cruzi Strain Typing Assay using MS2 peptide spectral libraries ( Tc-STAMS2 ) . The novelty of the method is based on the use of peptide fragmentation spectra as strain-specific fingerprints to classify and identify DTUs . Initially , a spectra library is generated from characterized T . cruzi strains . The library is subsequently inspected using MS/MS spectra from unknown strains and confidently assigned to a specific strain in an automated and computationally-driven approach . The Tc-STAMS2 method was challenged to test several variables such as sample type and preparation , instrument setup and identification platform . Tc-STAMS2 provided high confidence and robustness in T . cruzi strain typing . The Tc-STAMS2 method represents a proof-of-concept of a complementary strategy to the current DNA-based T . cruzi genotyping methods . Moreover , the method allows the identification of strain-specific features that could be related to the biology of T . cruzi strains and their clinical outcomes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"libraries",
"tropical",
"diseases",
"microbiology",
"database",
"searching",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoan",
"life",
"cycles",
"developmental",
"biology",
"protein",
"expression",
"protozoans",
"ne... | 2018 | Development of a Trypanosoma cruzi strain typing assay using MS2 peptide spectral libraries (Tc-STAMS2) |
Rotavirus is the leading agent causing acute gastroenteritis in young children , with the P[8] genotype accounting for more than 80% of infections in humans . The molecular bases for binding of the VP8* domain from P[8] VP4 spike protein to its cellular receptor , the secretory H type-1 antigen ( Fuc-α1 , 2-Gal-β1 , 3-GlcNAc; H1 ) , and to its precursor lacto-N-biose ( Gal-β1 , 3-GlcNAc; LNB ) have been determined . The resolution of P[8] VP8* crystal structures in complex with H1 antigen and LNB and site-directed mutagenesis experiments revealed that both glycans bind to the P[8] VP8* protein through a binding pocket shared with other members of the P[II] genogroup ( i . e . : P[4] , P[6] and P[19] ) . Our results show that the L-fucose moiety from H1 only displays indirect contacts with P[8] VP8* . However , the induced conformational changes in the LNB moiety increase the ligand affinity by two-fold , as measured by surface plasmon resonance ( SPR ) , providing a molecular explanation for the different susceptibility to rotavirus infection between secretor and non-secretor individuals . The unexpected interaction of P[8] VP8* with LNB , a building block of type-1 human milk oligosaccharides , resulted in inhibition of rotavirus infection , highlighting the role and possible application of this disaccharide as an antiviral . While key amino acids in the H1/LNB binding pocket were highly conserved in members of the P[II] genogroup , differences were found in ligand affinities among distinct P[8] genetic lineages . The variation in affinities were explained by subtle structural differences induced by amino acid changes in the vicinity of the binding pocket , providing a fine-tuning mechanism for glycan binding in P[8] rotavirus .
Rotaviruses are the leading etiologic agent of viral gastroenteritis in infants and young children worldwide and are responsible for an estimated 140 , 000 deaths each year in developing countries [1] . The typical classification of rotaviruses was derived from their genome composition and the immunological reactivity of three of their structural proteins: VP6 , VP7 and VP4 . Rotaviruses are classified into at least 7 groups ( A to G ) according to the immunological reactivity of the VP6 middle layer protein , with group A rotavirus being the most commonly associated with infections in human . The two outer capsid proteins VP7 and VP4 , elicit neutralizing antibodies that can induce viral protection . Using these two proteins , a traditional dual classification system of group A rotaviruses into G ( depending on the VP7 glycoprotein ) and P ( depending on the protease-sensitive VP4 ) types was established [2] . At least 36 different G-serotypes and 51 P-types have been identified among human and animal rotaviruses [3] . Viruses carrying G1[P8] , G2[P4] , G3[P8] and G4[P8] represent over 90% of human rotaviruses strains co-circulating in most countries [2] , with the P[8] genotype , which comprises four different genetic lineages ( I to IV [4] ) , being particularly relevant [5] . Diverse interactions between histo-blood group antigens ( HBGAs ) and rotavirus have been described and it is believed that HBGAs expressed on the surface of target cells serve as viral receptors . The distal VP8* portion ( ~27 kDa , N-terminal ) of the rotavirus spike protein VP4 from P[8] , P[4] , P[6] and P[19] genotypes recognize the secretor HBGAs . P[8] and P[4] are closely related genetically and both genotypes were reported to bind the Lewisb ( Fuc-α1 , 2-Gal-β1 , 3-[Fuc-α1 , 4-]GlcNAc; Leb ) and H type-1 ( H1 ) antigens ( Fuc-α1 , 2-Gal-β1 , 3-GlcNAc ) by some authors [6] , while there are controversial reports that show no Lewisb binding for these genotypes [7] . P[6] , a slightly further related genotype , binds the H1 antigen but not Lewisb [6] , whereas P[19] binds mucin core glycans with GlcNAc-β1 , 6-GalNAc motif and the type-1 HBGA precursor [8] . In addition , P[9] , P[14] and P[25] genotypes bound specifically to the type A antigens ( GalNAc-α1 , 3-[Fuc-α1 , 2-]Gal ) [9 , 10] , whereas P[11] interacted with single and repeated N-acetyllactosamine ( Gal-β1 , 4-GlcNAc; LacNAc ) , the type-2 precursor glycan [11] . Detailed evidences of VP8*-HBGAs interactions has been obtained by X-ray crystallography of P[14] VP8* and P[11] VP8* in complex with the type A oligosaccharide [9] and LacNAc [11] , respectively . Recently , the structure of porcine P[19] VP8* complexed with lacto-N-fucopentaose I ( Fuc-α1 , 2-Gal-β1 , 3-GlcNAc-β1 , 3-Gal-β1 , 4-Glc; LNFPI ) , and the mucin core-2 oligosaccharide ( GlcNAc-β1 , 6-[Gal-β1 , 3] GalNAc ) has been solved , showing a carbohydrate binding pocket alternative to the one used by P[11] and P[14] [7] . This binding pocket was first suggested by protein sequence analyses in other members belonging to the P[II] genogroup of rotaviruses ( i . e . P[4] , P[6] , P[8] genotypes ) [7] and recently confirmed for P[4] and P[6] VP8*s [12] . However , why non-secretors individuals ( lacking α1 , 2 fucosylation in secretory HBGAs ) have reduced rotavirus susceptibility [13–15] and what is the role of the secretory L-fucose in H1 ligand recognition for the most relevant human rotavirus remains unknown , as no structure was still available for the P[8] genotype in complex with ligand HBGAs . By using VP8* from a clinical isolate belonging to the lineage III of P[8] , in the present work we show that P[8] VP8* binds H1 antigen at a similar site as their ligand HBGAs bind to P[19] , P[4] and P[6] genotypes . Our structural and functional results also show that the H1 precursor lacto-N-biose ( Gal-β1 , 3-GlcNAc; LNB ) , devoid of L-fucose , also interacts with VP8* and we discard Lewisa or Lewisb antigens as ligands of P[8] genotypes . We provide the molecular bases for the role of secretor antigen in rotavirus binding to its receptor . Our results show two-fold increase in the affinity for the H1 antigen compared to LNB . This increase is explained by reduced contacts of the L-fucose with solvent molecules and the structural stabilization of LNB moiety in the competent conformation for binding . Furthermore , we show how subtle differences at the H1/LNB binding pocket in different P[8] lineages influence antigen affinities , giving clues for the relevance of the host glycobiology in P[8] rotavirus impact in humans . The partial anti-adhesin effect of LNB against rotavirus reported in the present article and the acquired knowledge on rotavirus-host cells interaction during virus attachment might open new avenues for the treatment and prevention of rotavirus infections .
VP8* domains from P[4] , P[6] , P[9] , P[11] , P[14] , P[25] genotypes and from different genetic lineages ( I , III and IV ) from P[8] genotype were produced ( S1 Fig ) . In order to confirm their functionality , the different proteins were challenged by an ELISA-like binding assay against a panel of biotinylated histo-blood group antigens ( HBGAs ) ( Fig 1 and S1 Table ) , corroborating the previously described interactions ( S2 Fig ) . Genotypes P[4] , P[6] and P[8] recognized the H type-1 antigen ( Fuc-α1 , 2-Gal-β1 , 3-GlcNAc , H1 ) , P[11] recognized the H type-2 antigen ( Fuc-α1 , 2-Gal-β1 , 4-GlcNAc , H2 ) and P[9] , P[14] and P[25] the blood group A antigen trisaccharide ( GlcNAc-α1 , 3- ( Fucα1 , 2 ) -Gal , Atri ) . The P[8] genotype additionally displayed low binding to this trisaccharide . However , and contrarily to previous reports [7] , in our assays genotypes P[4] and P[8] exhibited very low or absence of binding to Lewisb ( Fuc-α1 , 2-Gal-β1 , 3-[Fuc-α1 , 4-]GlcNAc ) . Remarkably , VP8* from P[8] genotype recognize the H1 antigen precursor lacto-N-biose ( Gal-β1 , 3-GlcNAc , Lewisc , LNB; S2 Fig ) but differences in the binding abilities were found among different genetic lineages and strains . Thus , VP8* from the cultivable human rotavirus Wa ( P[8]Wa ) and the Rotarix vaccine ( P[8]Rotarix ) strains ( both lineage I ) gave lower signals in the ELISA tests with H1 and LNB than the VP8* from the lineage IV strain ( P[8]LIV ) and from a clinical isolate belonging to lineage III ( P[8]c ) ( S2 Fig ) . The newly discovered interaction with LNB was further characterized by testing different concentrations of VP8* from this isolate ( P[8]c ) and the cultivable P[8]Wa ( S3 Fig ) . The results showed that both proteins were able to bind H1 and LNB , although binding to the first antigen was higher . We performed a more detailed characterization of the interaction of VP8* to H1 and LNB by determining the apparent affinity constants ( Kda ) for each interacting pair by SPR ( Table 1 ) . The Kda of the VP8* P[8]c recognition of the H1 antigen ( Kda = 27 . 9 ± 0 . 71 μM; Fig 2A ) was two-fold lower compared to the LNB precursor ( Kda = 52 . 1 ± 4 . 26 μM; Fig 2B ) , this difference was significant ( p = 0 . 0045 ) suggesting that the H1 L-fucose moiety contributes actively to the binding . Surprisingly , the affinity constant for the interaction of VP8* from P[8]Wa with H1 was three times higher ( Kda = 80 . 2 ± 2 . 21 μM ) than that of P[8]c ( Fig 2C ) . Furthermore , the VP8* from P[8]Wa showed a similar apparent affinity for LNB than for H1 ( Kda = 66 . 5 ± 6 . 47 μM; p > 0 . 05 Fig 2D ) . Interestingly , interactions for VP8* from Rotarix strain ( lineage I ) and the lineage IV strain with H1 and LNB were too low to be determined under our SPR conditions . To further characterize the role of VP8* interaction with the H1 precursor LNB , complete virions of the Wa strain ( triple-layered particles; TLP ) and double-layered particles ( DLP; obtained after VP4 and VP7 removal by EDTA treatment ) were assayed by an ELISA-like binding assay . The Wa TLP , but not the Wa DLP , were able to bind H1 antigen ( Fig 3A ) and its precursor LNB ( Fig 3B ) , in a concentration-dependent manner . This indicated that the observed interaction between P[8] VP8* and LNB was also relevant in a complete rotavirus context . These results also point to the fact that despite the low affinity of P[8] VP8* from lineage I to H1 and LNB , the high avidity of a multi-binder particle ( virions contain 120 molecules of VP4 ) results in a measurable interaction . To confirm the binding of VP8* from the P[8]c genotype to LNB we settled up a binding blocking assay where LNB ( Gal-β1 , 3-GlcNAc ) and its structurally-related disaccharide galacto-N-biose ( Gal-β1 , 3-GalNAc; GNB ) , were tested as potential inhibitors of binding . The results showed a moderate but significant ( p < 0 . 05 ) reduction in the binding to the H1 antigen by both disaccharides ( 24 . 2% reduction for LNB and 30 . 1% for GNB; Fig 4A ) . The monosaccharide constituents of LNB and GNB ( D-galactose , N-acetyl-glucosamine and N-acetyl-galactosamine ) and L-fucose were also tested ( Fig 4 ) . Among these sugars only D-galactose possessed a discrete but significant blocking capacity of VP8* binding to the H1 antigen ( 14 . 7% reduction; p = 0 . 032 ) . Interestingly , soluble L-fucose significantly increased the binding of VP8* to the H1 antigen ( Fig 4A ) . As expected , when the precursor LNB was used as the ligand , soluble LNB and GNB were also able to reduce the binding of VP8* P[8]c , by 52 . 2% and 44 . 1% , respectively ( Fig 4B ) . We next investigated the role of the H1 precursor antigen in rotavirus infection by incubating MA104 cells with rotavirus Wa strain that was preincubated with LNB , GNB , their monosaccharide constituents and L-fucose . Only LNB significantly blocked viral infection ( 33% reduction; Fig 4C ) , suggesting that this HBGA precursor interferes with the binding of the rotavirus Wa strain to its receptor in MA104 cells . To understand the molecular basis of LNB and H1 recognition and binding to P[8] VP8* we determined the crystal structure of the clinical isolate P[8]c VP8* in its apo form and bound to LNB and H1 glycans . Two different crystalline forms of P[8]c VP8* in its apo form were obtained . The first form , VP8*-Apo1 , diffracts at 1 . 35 Å resolution and presents a single copy of P[8]c VP8* in the crystal asymmetric unit ( ASU ) while the second form , VP8*-Apo2 , diffracts to 1 . 5 Å and presents two copies in the ASU ( Table 2 ) . The three copies of P[8]c VP8* in these two crystalline forms present the galectin fold with two twisted β-sheets separated by a superficial cleft that conforms the glycan binding site in P[11] and P[14] genotypes [9 , 16] ( Fig 5 ) . Superimposition of the individual VP8* protomers from these crystals showed that the three molecules of P[8]c VP8* are almost identical ( RMSD 0 . 28 Å for the superposition of all the Cα; S4A Fig and S2 Table ) . Crystals of P[8]c VP8* bound to H1 or its precursor LNB were obtained in a third different crystalline form and the structures of P[8]c VP8*-H1 and P[8]c VP8*-LNB complexes were solved to 1 . 8 and 1 . 3 Å , respectively ( Fig 6A , 6B and Table 2 ) . Two protomers of VP8* were present in the ASU of each of these crystals and , remarkably , only one out of the two protomers showed a bound glycan molecule . Sugar binding induces negligible conformational changes in the VP8* ( S4A Fig ) since the structural comparison of the glycan-bounded and glycan-free P[8]c VP8* protomers showed minimal differences ( RMSD 0 . 22–0 . 35 Å; S2 Table ) . Furthermore , P[8]c VP8* are also structurally identical ( RMSDs 0 . 37–0 . 57 Å ) to the VP8* apo forms of the linage I P[8]Wa and P[8]Rotarix ( S4B Fig and S2 Table ) , supporting that the glycan binding site is preformed in the P[8]c VP8* protein . This characteristic also seems to be shared by other P genotypes belonging to the P[II] genogroup , since structural comparison with recently reported glycan-bound structures of VP8* from human P[4] and P[6] , and porcine P[19] genotypes showed modest differences ( RMSD 0 . 51–0 . 84 Å ) ( S4C Table and S2 Table ) . P[8]c VP8* binds H1 and LNB in a pocket formed by one of the β-sheets and the C-terminal α helix ( Fig 6A and 6B ) . The structures showed that the LNB moiety is embedded in the pocked interacting with the protein while the L-fucose is projected out with minimal , mainly mediated by solvent , contacts with the protein ( Fig 6A and 6B ) . Only seven residues recognize the LNB , contacting the N-acetyl-glucosamine moiety via L167 , W174 , T185 , R209 , and E212 and the galactose moiety via T184 , T185 , E212 and N216 ( Fig 6C , 6D and S3 Table ) . This pocket shared identical interacting residues to the novel VP8* binding pocket recently discovered in P[19] [7] and P[4]/P[6] [12] genotypes for their interaction with LNFPI and it differs from the previously defined carbohydrate binding site in VP8* from P[11] ( LacNAc binding [11] ) and P[14] ( A-antigen binding [9] ) genotypes , that is located in the cleft between the two β sheets ( S4D Fig ) . The N-acetyl-glucosamine has a major contribution to the binding since it is inserted in the pocket while the galactose acquires a more superficial position ( Fig 6C and 6D ) . In this way , the N-acetyl-glucosamine ring is packet between W174 and R209 that define two faces of the binding site and the O4 and O6 oxygens mediated hydrogen-bonds with E212 that it placed at the bottom of the pocket ( Fig 6C , 6D and S2 Table ) . To confirm the role of these amino acids in H1 and LNB interaction , the P[8]c VP8* mutants W174A ( M1 , VP8W174A ) , R209A ( M2 , VP8R209A ) and E212A ( M3 , VP8E212A ) were obtained ( S1 Fig ) . The three mutant VP8* lost their ability to interact with their receptor when they were assayed for binding to H1 and LNB by SPR ( S5 Fig ) , supporting the structural data and suggesting that the identified pocket is the only binding site for H1 antigen and its precursor in P[8] VP8* . Fucosylation of the H1 precursor is genetically determined by the FUT2 gene ( α1 , 2 fucosylation of the terminal galactose ) , resulting in different secretor status and by the FUT3 gene ( α1 , 4 fucosylation at the precursor N-acetyl-glucosamine ) , defining the Lewis status ( Fig 1 ) . This genetically determined glycan profile is a susceptibility factor in human rotavirus [17 , 18] . Our P[8]c VP8* structure in complex with H1 antigen confirms that the secretory L-fucose portion has reduced contacts with VP8* that are mediated by solvent molecules . Similarly , minimal or null contacts of the L-fucose moiety were observed in the complexes of LNFPI with other VP8* from the P[II] genogroup [7 , 12] . Therefore , the difference in the affinity for H1 and LNB in diverse VP8* from this genotype were difficult to explain . A close view of the relative disposition of the precursor LNB moiety in both P[8]c VP8* complexes after superimposition of VP8* proteins showed some differences but not so for the interacting residues ( Fig 7A ) . If , alternatively , the N-acetyl-glucosamine moieties from both structures are superimposed it becomes visible that the galactose moiety occupies a more solvent-exposed position in the LNB structure compared to H1 ( Fig 7B ) , explaining its weaker binding in comparison with this antigen . The atomic resolution of the diffraction data and the high quality of the density maps derived from these data ( Fig 7C and 7D ) allows modeling accurately the ligand structures to observe these differences . This observation indicates that the L-fucose moiety induces an LNB conformation more suitable for the union to VP8* . The P[8]c VP8* structure in complex with H1 shows that the L-fucose ring stacks over the acetamido group of N-acetyl-glucosamine stabilizing the glycan conformation ( Figs 6C and 7B ) . Therefore , the structures suggest that the L-fucose moiety could favor the VP8* binding by a double mechanism: inducing a competent conformation that facilities the LNB module recognition and binding and mediating indirect interactions that stabilize the glycan-VP8* complex . Oppositely , the fucosylation in the N-acetyl-glucosamine in the LNB precursor to produce the Lewisb antigen seems to be incompatible with the binding to P[8] VP8* . Docking of the Lewisb antigen ( Fuc-α1 , 2-Gal-β1 , 3-[Fuc-α1 , 4-]GlcNAc ) in the P[8]c VP8* structure taking H1 antigen as a reference showed that the α1 , 4-linked L-fucose points towards the inside of the glycan binding pocket , clashing with different residues ( mainly T185 and E212 ) that generate this cavity ( S6A Fig ) . Since glycan binding pocket is highly conserved among P[8] VP8* linages and P[II] genotypes [7 , 12] , it seems that this group of rotaviruses are non-competent to bind Lewis HBGAs . This observation is confirmed by our ELISA assays where any of the VP8* proteins from P[II] genotype showed binding capacity to these HBGAs , and it questions previous results were P[8] VP8* was shown to interact with Lewisb [6] . Following an identical approach the A-type I antigen can be docked in the P[8]c VP8* sugar-binding pocket . The docked sugar showed that the N-acetylgalactosamine ( GalNAc ) added to the non-reducing end of the H1 galactose is pointing towards the solvent without showing any interaction with the protein ( S6B Fig ) , supporting that P[8] VP8* should be able to bind this sugar . The structures of VP8* in complex with LNB and H1 obtained here confirm that P[4] , P[6] , P[19] and the prevalent human genotype P[8] of rotavirus share a common glycan binding site which is highly conserved in structure and sequence . Therefore , the differences in affinity and specificity for HBGAs observed in our ELISA and SPR assays as well as those reported by others are striking . Particularly , the differences in the capacity to interact with the H1 antigen observed among P[8] lineages are difficult to understand . Inspection of the amino acids defining the glycan binding site in the four P[8] lineages oriented by the comparison of the P[8]c , P[8]Wa and P[8]Rotarix structures revealed interesting differences . First , the VP8* protein from Rotarix strain ( lineage I ) differed from other members of the same lineage , including the Wa strain , by the replacement of leucine 167 by phenylalanine . Since L167 is placed at the bottom of the sugar-binding pocket ( Fig 6C and 6D ) , the introduction of a bulky Phe residue could explain the lack of interaction of this vaccine strain to H1 [19] that was corroborated in our study . Second , a close inspection of the P[8] structures revealed small differences in the disposition of the β-hairpin connecting the strands β9 and β10 ( S7 Fig ) . These two β-strands define the bottom of the glycan binding pocket where the LNB moiety is settled ( Fig 6C and 6D ) . These subtle movements are allowed by the presence of two Gly residues ( G170 and G171 ) conferring flexibility to the loop . Y169 and R172 at both sides of the Glys delimit the loop and play a pivotal role in the correct organization of the binding pocket since Y169 stacks over R209 and R172 interacts with W174 , the main residues recognizing N-acetyl-glucosamine ( Fig 6C , 6D and S7 Fig ) . Therefore , the disposition of this loop could modulate de glycan affinity and specificity . The sequence analysis of this region among P[8] lineages reveals that at position 173 lineage I presents a Val whereas an Ile is found in the rest of lineages ( S4 Table ) . Position 173 is placed in the base of the β-hairpin facing to the hydrophobic core of VP8* protein ( Fig 6C , 6D and S7 Fig ) . Therefore , we wondered if this subtle change could influence the glycan-binding site architecture accounting for the difference in the H1 affinities observed between lineages . To test this possibility we designed a new mutant in which I173 in P[8]c VP8* was replaced by Val ( M4 , VP8I173V ) , emulating P[8]Wa . This replacement resulted in a VP8* with a diminished interaction with the H1 antigen , with an affinity constant ( Kda = 39 . 0 ± 1 . 25 μM , S5 Fig , Table 1 ) 1 . 4-fold higher compared to the wild-type protein ( Kda = 27 . 9 ± 0 . 71 μM , Table 1; p = 0 . 0002 ) . The VP8I173V variant retained the binding ability to LNB with an apparent affinity constant ( Kda = 49 . 4 ± 1 . 74 μM , Table 1 ) that was still higher than that of H1 ( p = 0 . 0033 ) . These results support that subtle amino acid changes at the loop close to the binding pocket may have contributed to modulate the glycan affinity between P[8] VP8* from lineages I and III . Differences at this site are also evident in the structures of VP8* from the P[II] genogroup . While conservation between P[8] and P[4] is high , the sequence divergence between P[4] and P[6] ( S4 Table ) might explain the different affinities for LNFPI between these two genotypes [12] .
Virus-HBGAs interaction has emerged as an important factor in viral infectivity . Contrarily to other enteric viruses ( i . e . : norovirus ) , the relevance of HBGA interaction in rotaviruses was first neglected , and virus-host cell attachment studies were mainly focused on binding to sialic acid , until interactions with HBGA were suggested by VP8* structural analyses [20] and experimentally determined in sialidase-insensitive strains [6] . In norovirus many studies point to the human FUT2 polymorphism as a key feature affecting viral infectivity [17 , 18] . Individuals carrying two null FUT2 alleles lack fucosyl transferase-2 activity , do not express H antigen structures at the intestinal mucosa and in secretions ( non-secretors ) and are less susceptible to norovirus . While previous studies showed no correlation between the secretor status and rotavirus infection [17] , the most recent studies show that antibody titers to rotavirus [13] , rotavirus gastroenteritis incidence [14] and vaccine take [15] correlate with the FUT2 phenotype . However , the molecular mechanisms of these correlations were unknown until now . The previously reported interaction of H1 antigen ( Fuc-α1 , 2-Gal-β1 , 3-GlcNAc ) with the most common human rotavirus P genotype P[8] that has been further characterized here at the structural level , highlights the importance of the secretor phenotype on the incidence of rotavirus diarrhea . We have determined the characteristics of this interaction , acknowledging a new binding site for H1 in VP8* common for all the members of P[II] genogroup . Our results show that physical interaction between the H1 antigen and P[8] rotavirus occurs through the precursor side of the molecule ( LNB ) , reinforcing the idea that the main carbohydrate-protein contacts are made via the N-acetyl-glucosamine moiety [7] . NMR studies on A-antigen binding of P[9] and P[14] VP8* , demonstrated that the L-fucose moiety does not make contacts with VP8* and rather it remains exposed to the solvent with a high degree of flexibility [21] . However , in the same study VP8* from genotypes P[4] and P[6] , that did not recognize A-antigen in our assays , were shown to bind this antigen and L-fucose-protein contacts were evidenced [21] . Structural data from P[4] and P[6] VP8* in complex with LNFPI also showed a limited but direct interaction of the α1 , 2-linked L-fucose with the protein , namely via the R209 residue , which is conserved in all proteins from genogroup P[II] [12] . Due to the minimal interaction of the secretory L-fucose to VP8* , the authors of this study hypothesize that this glycan moiety has a low contribution to binding affinity and that a strong interaction would be expected for the unfucosylated H1 precursor , explaining the epidemiological studies that do not correlate the FUT2 status to infection by P[4] and P[6] genotypes [22] . Contrarily to this , we show that although the L-fucose moiety of H1 makes indirect contacts with P[8] VP8* , it stabilizes the competent conformation of the LNB moiety to interact with the sugar binding residues , resulting in two-fold lower Kda for H1 compared to LNB . This small but significative difference may be of relevance in the viral susceptibility context between secretors and non-secretor ( FUT2-/- ) individuals . Furthermore , a weaker interaction to LNB may also explain why infection of P[8] rotaviruses can occur , at a lower level , in non-secretor individuals [23] and it also accounts for the inhibitory effect of LNB in in vitro rotavirus infection reported here . The previously reported interaction of the P[8] genotype with Lewisb identified by ELISA assays [6] , could not be reproduced in our experiments . The structural evidence obtained here and in the analyses of the P[4] and P[6] structures in complex with LNFPI [12] argues against interaction with Lewisb , which differs with H1 in the presence of an extra L-fucose α1 , 4-linked to N-acetyl-glucosamine that generates steric hindrances to the interaction . These discrepancies , together with the differences found between ELISA and SPR for H1 binding in the different P[8] lineages , suggest that simple qualitative ELISA tests do not always provide reliable results and that other techniques need to be implemented in order to assess VP8* affinities for HBGAs . However , structural data from P[4] and P[6] genotypes predict interaction with A- and B-types HBGAs , as the N-acetyl-galactosamine ( A-type ) and galactose ( B-type ) located at the non-reducing ends in these glycans do not make any steric hindrance [12] . This coincide with our observation that P[8] VP8* from different lineages interact with blood group A trisaccharide and it is also supported by modelling additional N-acetyl-galactosamine in H1 bound to P[8]c VP8* ( S6 Fig ) . We showed that variations in the binding domain in the P[8] lineages exist that justify the differences in the affinity for H1 and LNB as measured by SPR . Additionally , even if the architecture of the binding site is similar for most P[8] lineages and other genotypes belonging to the P[II] group , other protein residues outside this site may possess epistatic effects over the capacity of the binding pocket to accommodate H1 and LNB , explaining the diverse affinities among P[II] genotypes and in the different P[8] lineages . This is exemplified by the fact that we were able to associate a residue that does not participate in direct protein-ligand contacts ( Valine 173 ) in P[8]wa VP8* to its lower affinity for H1 and LNB and that VP8* from the vaccine strain Rotateq ( lineage II ) , although sharing identical key binding residues to lineage III , do not bind H1 [19] . It is postulated that subtle changes in residues within and outside the defined pocket leads to a fine tuning in HBGA affinities that may ultimately impact host infection capacity . Epidemiologic studies revealed the occurrence of P[8] lineages I , II and III as the major circulating rotavirus with a prevalence of the lineage III [4 , 24–26] , while lineage IV is rarely found [25] . It is worth mentioning that this lineage , which is phylogenetically distant [4] , carries some differences in the H1 binding pocket and showed low affinity to H1 , has been isolated from few countries but it seems to be rapidly expanding [27] . Many studies have focused on the rotavirus genotypes circulating before and after the introduction of rotavirus vaccination programs [28 , 29] , but no work addressed the question of whether a link exists between the secretor status and the incidence of different P[8] lineages . In an study on the effect of the FUT2 status on rotavirus gastroenteritis it was shown that 100% of the patients ( n = 51 ) were secretor positive compared to a healthy control or a group of non-rotavirus gastroenteritis patients ( 14–19% of non-secretors ) and that all rotavirus involved were P[8] from lineage III [30] . Studies on the dynamics of G1P[8] rotavirus in a western population showed that ancient strains were Wa-like ( lineage-I ) and that new lineages emerged since late nineties [4] , although the three main lineages are co-circulating nowadays in most geographical locations . Some authors have argued that the lack of ( or low ) interaction of Wa-like strains to H1 may help these viruses to prevail , because they do not discriminate by the secretor status [21] , but our results suggest that only affinities for the receptor may be varying within the different P[8] lineages and it is not known how this may impact viral fitness . Notably , the amino acids comprising the H1/LNB binding pocket fall outside the defined epitopes in VP8* that elicit protection ( S8 Fig ) . Mutations in VP8* which result in antigenic variants that could escape neutralizing antibodies are frequently isolated [24–26] , but they are not affecting the H1/LNB interacting residues defined here . The impact of the different rotavirus P[8] lineages in the population depending on the secretor status deserves further studies in order to ascertain if the prevalence or co-circulation of different P[8] lineages responds to an adaptation to the HBGA profiles of the different hosts . Despite the need for more exhaustive research on the relevance of HBGA and host specificity/infectivity in P[8] rotavirus , surface glycans possess a clear application in the development of antiviral strategies . It is established that human milk , in addition to other antiviral components , carry a set of oligosaccharides ( human milk oligosaccharides; HMO ) that share structural similarities to HBGA [31] and could act as anti-adhesins by competition with pathogen ligands at the mucosa . This blocking ability by soluble carbohydrates resembling rotavirus ligands has been evidenced . HMO were shown to inhibit binding of VP8* from P[6] and P[11] genotypes [32]; P[8] and P[4] genotypes infection is inhibited by the HMO 2'-fucosyllactose , 3'-sialyllactose and 6'-sialyllactose [33]; LNFPI inhibited infection of P[19] , P[4] , P[6] and P[8] genotypes [7] , while we showed that LNB inhibited infection of Wa strain in vitro . The anti-adhesin potential of this simple disaccharide ( LNB ) is susceptible for being exploited in antiviral strategies . LNB is present in human milk in its free form [34] but mainly as a building block of type I HMO , which are predominant in human milk over type II HMO ( based on LacNAc ) , which are characteristic of other mammals and primates [31] . Thus , LNB has been considered the human milk ‘bifidus factor’ , and many bifidobacterial species from the infant gastrointestinal tract have the enzymatic machinery for its metabolism [35] . LNB would not only act as a bifidobacteria-stimulating prebiotic but also as a viral anti-adhesin to counteract rotavirus infection . Furthermore , its relatively simple synthesis , which can be undertaken enzymatically and by metabolic engineering approaches [36] , makes this disaccharide a candidate for the development of new functional foods ( e . g . improved infant formula ) . In this respect , it is important to consider that high affinity constants ( in the mM range ) have been determined for free oligosaccharides binding to VP8* [12 , 21] and , in order to obtain good competitors , conjugated multivalent oligosaccharides seem to be a better option . Detailed determination of the interactions between viruses and their host is crucial to develop appropriate antiviral strategies . We have defined the molecular interactions of P[8] VP8* from human rotavirus with its ligand HBGA giving a physical explanation as to why the secretor status influences rotaviral infectivity . Notwithstanding , extra structural elements beyond the identified binding site in VP8* are probably responsible for modulating HBGA interactions within P[8] lineages . Dissection of additional VP8* structural features affecting ligand binding is under way .
The VP8* ( amino acids 64–224 from the VP4 protein of rotavirus ) belonging to the P[4] , P[6] , P[8] , P[9] , P[11] , P[14] and P[25] genotypes were cloned into the , pGEX-2T , expression vector ( GE Healthcare ) in order to express N-terminal GST fusions . To amplify P[4] , P[6] , P[8] , P[9] and P[14] VP8*s coding region , RNA was extracted from human stool samples ( collected at Hospital Clínico Universitario de Valencia ) containing rotavirus of known P genotype using the Trizol reagent following the standard procedure ( Invitrogen ) . Viral RNA was retro-transcribed using the SuperScript Reverse Transcriptase ( Invitrogen ) and random-primers , and the cDNA was amplified by PCR using Pfu polymerase ( Stratagene ) with primers detailed in S5 Table . The cDNAs were finally cloned into pGEX-2T ( GE healthcare ) vector after digestion with BamHI ( ThermoFisher ) . The VP8* genes from genotypes P[4] , P[11] and P[25] were purchased as synthetic genes from Gene-ART technologies ( ThermoFischer ) . The expression level of the VP8* protein from the clinical sample P[4] VP8* genotype was very low in E . coli and its codon usage was optimized . P[11] VP8* and P[25] VP8* were not available as clinical samples . The recombinants GST::VP8* proteins were expressed in E . coli BL21 ( DE3 ) ( Novagen ) and purified by affinity chromatography using GSTrap columns coupled to an ÄKTA prime FPLC system ( GE Healthcare ) . All sequences are included as fasta files in the supplementary data . Selected residues in the GST::VP8* P[8] were replaced for alanine or valine according to the structural data of the LNB binding site . Four mutants ( M1–M4 ) were constructed using a Quick-Change site-directed mutagenesis kit ( Stratagene ) and appropriate oligonucleotides ( S5 Table ) , and the DNA changes were confirmed by DNA sequencing . Mutant M1 , M2 and M3 contained changes in the codons for tryptophan 174 , arginine 209 and glutamic 212 residues , respectively , that introduced an alanine at each position . Mutant M4 substituted isoleucine 173 by valine . A panel of biotinylated sugar antigens including Lea , Leb , Lec ( lacto-N-biose; LNB ) , H type-1 , H type-2 and blood group A and B trisaccharides were purchased from Glyconz ( Fig 1 and S1 Table ) . These glycans are biotinylated neoglyconjugates of a poly[N- ( 2-hydroxyethyl ) acrylamide] ( PAA ) with a size from 30 to 50 KDa . This forms a flexible polymer ideal for a multivalent presentation of glycans . Immobilized streptavidin F96 black plates ( Nunc ) were coated with the biotinylated oligosaccharides ( 2 μg/ml ) in milli-Q water and incubated during 1 hour at 37°C . After functionalization the plates were washed once with PBS containing 0 . 05% of Tween 20 ( PBS-T ) and the VP8* proteins were added ( 10 μg/ml ) and incubated at 4°C overnight . After three washes with PBS-T , a rabbit polyclonal antibody anti GST ( 1:1 , 000 ) ( Abcam ) was added and the plates were incubated one hour at 37°C . Then , the plates were washed three times with PBS-T and incubated for 1 h at 37°C with 1:10 , 000 dilution of horseradish peroxidase ( HRP ) -conjugated goat anti-rabbit ( Abcam ) . After three washes with PBS-T , the binding was detected using QuantaBlue reagent ( ThermoFisher ) kit , as recommended by the manufacturer . Fluorescence units were registered by a MultiScan microplate reader . All the binding assays were performed in triplicate . The EC50 binding of the VP8* from the clinical ( P[8]c ) and Wa ( P[8]Wa ) genotypes to the H type-1 and to its precursor ( LNB , Lec ) was determined incubating two fold serial dilutions of the VP8* proteins , ranging from 100 μg/ml to 1 . 5 μg/ml . The binding assays were performed in triplicate using the protocol described above . To confirm the binding of the VP8* from the P[8] genotype to the H type-1 precursor LNB , a blocking assay was performed using soluble LNB and its related disaccharide galacto-N-biose ( Gal-β-1 , 3-GalNAc; GNB ) produced and purified in our laboratory as previously described [36] . Streptavidin microtiter plates were coated with biotinylated H type-1 antigen or LNB at 2 μg/ml with water and incubated for 1 hour at 37°C , followed by an overnight incubation at 4°C . Blocking assays were performed in parallel , using glass tubes containing the P[8]c and P[8]Wa VP8* protein at their EC50 for each ligand and 20 mM of each of the soluble disaccharides ( LNB and GNB ) and monosaccharides ( D-galactose , GlcNAc and GalNAc ) . A positive binding control without sugar was also included . The tubes containing the mixes of VP8* with sugars were maintained 1 hour at 37°C , followed by an overnight incubation at 4°C . The next day the coated streptavidin plates were washed with PBS-T , the VP8*-sugar solutions were added to the plates and incubated during 4 hours at 4°C . The plates were washed three times with PBS-T and detection of the interactions was performed as described above . The results are presented as the percentage ( % ) of binding of each condition compared to the binding of the positive binding control ( without blocking sugar ) . All experiments were performed in triplicate . African green monkey kidney epithelial cells ( MA104 cell line; ATTC #CRL-2378 . 1 ) were used for the propagation of rotavirus Wa strain that belongs to the globally dominant human genotype G1P[8] . Briefly , ten MA104 cells confluent 150-cm2 flasks ( approximately 1 . 5 ×107 cells/flask ) were infected with Wa strain at a multiplicity of infection ( MOI ) of ≤ 0 . 1 and processed as previously described [37] . One hundred ml of medium with 1 . 5x108 virus/ml were obtained and the viral particles were concentrated by pelleting at 160 , 000 × g for 1 h at 4°C in a SW 41 rotor ( Beckman ) . The viral pellet was resuspended in TNC buffer ( 20 mM Tris-HCl pH 8 . 0 , 100 mM NaCl , 1 mM CaCl2 ) for triple-layered particles ( TLP ) or in TNE ( 20 mM Tris-HCl , pH 8 . 0 , 100 mM NaCl , 1 mM EDTA ) for double-layered particles ( DLP ) . An ELISA-like binding assay was employed to determine the binding ability of rotavirus TLP and DLP to the H type-1 antigen and to its precursor LNB . Streptavidin plates were coated with the biotinylated oligosaccharides as described above . After washing with PBS-T , two fold serial dilutions of TLPs and DLP were added to the plate ( ranging from 10 μg/ml to 0 . 078 μg/ml ) . The TLP were always maintained in TNC-T ( 20 mM Tris , 100 mM NaCl , 1 mM CaCl2 , 0 . 05% Tween 20 , pH 7 , 4 ) buffer and the binding and washing steps were always carried out in this solution . DLPs assays were carried out in Tris-buffered saline buffer with Tween 20 ( TBS-T , 20 mM Tris , 100 mM NaCl , 0 . 05% Tween 20 , pH 7 . 4 ) . TLP and DLP were incubated in the plate overnight at 4°C . After binding , the plates were washed three times in TNC-T ( for TLP ) or TBS-T ( for DLP ) with 0 . 05% Tween 20 ( TNC-T and TBS-T ) , and a mouse anti-VP6 antibody was added at 1:100 in TNC-T or TBS-T and incubated 1 h at 37°C . The plates were then washed three times with TNC-T or TBS-T , and a HRP-conjugated anti-mouse IgG was added at 1:10 . 000 and incubated at 37°C for 1 h . After three final washes the binding was revealed by QuantaBlue reagent ( ThermoFisher ) following the manufacturer recommendations . Fluorescence units were recorded by a MultiScan microplate reader . The G1P[8] Wa strain was tested on MA104 cells . The sugars LNB , GNB , GlcNAc , GalNAc , D-galactose and L-fucose were tested for their effect on rotavirus infectivity . The oligosaccharides were previously heat sterilized at 99°C for 10 min and then dissolved in serum-free DMEM containing 1 μg/ml trypsin . Serum-free DMEM containing 1 μg/ml without oligosaccharide was used as a control in each experiment . The effect of different mono- and disaccharides on rotavirus infectivity was assessed through standard fluorescent focus assays on MA104 cells [37] . The dilution of Wa virus stocks that yielded ~150 focus-forming units/well was first established . Then , sugars were added during virus inoculation at a final concentration of 5 mg/ml , incubated for 1 hour , and unbound virus was removed by washing with FBS-DMEM . The cells were allowed to be infected for 16h , washed once with PBS and fixed with 100% methanol . A mouse anti-VP6 primary antibody ( 1:50 dilution in PBS containing 3% BSA ) was added and incubation proceeded for 30 min at room temperature with gentle rotation . A secondary antibody anti-mouse IgG-FITC ( Sigma F4143 ) diluted 1:128 in PBS containing 3% BSA was added and incubated for 30 min at room temperature with gentle rotation . Individual fluorescence foci were counted on an inverted fluorescence microscope with a FITC-compatible filter . Infectivity in the absence of oligosaccharides served as the control . Each experimental condition was tested a minimum of 2 times , with technical triplicates for each oligosaccharide . The means and SD from a minimum of 6 determinations are represented for each condition . Virus titer measured in the absence of oligosaccharides was considered to be 100% infectivity , and changes in virus titer in the presence of sugars were expressed as percentage of infectivity compared with no sugar treatment . The affinity assays were based on SPR and performed in a Biacore T100 instrument ( GE Healthcare ) . H1 PAA-biotin and LNB PAA-biotin were diluted to a concentration of 1 mg/ml in water and captured with streptavidin present in a SA sensor chip ( GE Healthcare ) . H1 was immobilized in channel 2 ( 630 RU ) of the sensor chip and LNB was immobilized in channel 4 ( 624 RU ) . The channels 1 and 3 were used as the reference surfaces for channels 2 and 4 , respectively . The immobilization process was performed by conditioning the sensor chip surface with three consecutive 1-minute injections of 1 M NaCl 50 mM NaOH before biotinylated ligands were immobilized at a flow rate of 15 μl/min . The affinity assays of VP8* polypeptides to biotinylated sugars were performed at 10°C using 1X HBS-EP+ buffer ( 0 . 01 M HEPES pH 7 . 4 , 0 . 15 M NaCl , 3 mM EDTA , 0 . 005% Surfactant P20 ) , a flow rate of 5 μl/min with 2700 seconds of contact time and a dissociation time of 1800 seconds . The regeneration step consisted in a wash step with 10 mM Glycine-HCl pH 2 for 20 seconds at the same flow rate . The assays were performed with purified VP8* at different concentrations ( 45; 137; 411; 1 , 234; 3 , 703; 11 , 111; 33 , 333; 100 , 000 and 200 , 000 nM ) . Each run included three blanks without sample . The affinity data were obtained after analysis of sensorgrams performed with the BIAevaluation 2 . 0 software ( GE Healthcare ) . Since multivalent oligosaccharides are immobilized on a sensor chip surface , avidity and rebinding effects can take place and apparent affinity constants ( Kda ) are calculated with this experimental setup . Kda values were obtained from the steady-state kinetics experiment as the ligand concentration needed to achieve a half-maximum binding at equilibrium . The experiments were made in triplicate . Graphical representation of signal/concentration curves were plotted using GraphPad Prism 6 for MacOsX . The crystals were grown as hanging drops at 21°C with a vapour-diffusion approach . Initial crystallization trials were set up in the crystallogenesis service of the IBV-CSIC using commercial screens JBS I , II ( JENA Biosciences ) and JCSG+ ( Molecular Dimensions ) in 96-well plates . Crystallization drops were generated by mixing equal volumes ( 0 . 3 μl ) of P[8]c VP8* protein solution and the corresponding reservoir solution , and were equilibrated against 100 lμl reservoir solution . Both P[8]c VP8* Apo structures were crystallized at 10 mg/ml . VP8* Apo1 was crystallized in a reservoir solution of 1 . 2 M ( NH4 ) 2SO4 , 3% iso-propanol and 0 . 1 sodium citrate pH 4 . 6 , whereas VP8* Apo2 was crystallized in 1 . 5 M Li2SO4 and 0 . 1 M Tris-HCl pH 6 . 5 . In both cases 2 M Li2SO4 was used to cryoprotect the crystal when freezing in liquid nitrogen . For the crystallization in presence of glycans , the ligands were mixed with the protein at 10 mM final concentration of ligand and 10 mg/ml of protein final concentration . P[8]c VP8* LNB was crystallized in a reservoir solution consisting in 25% PEG 3 , 350 0 . 1 M Bis-Tris pH 5 . 5 . The cryosolution used for crystal freezing was its reservoir solution increased up to 35% PEG 3 , 350 . P[8]c VP8* H1 was crystallized against a a reservoir solution consisting in 25% PEG 6 , 000 , 0 . 1 M Na-HEPES pH 7 . 5 , 0 . 1 M LiCl , and PEG 6 , 000 was increased up to 35% for cryoprotection . X-ray diffraction was carried out at 100K at Alba ( Cerdanyola , Barcelona , Spain ) and DLS ( Didcot , UK ) synchrotrons and the best data sets used to solve the structures were collected at the indicated beamlines and wavelengths ( Table 2 ) . Diffraction data was processed and reduced with Mosflm[38] and Aimless[39] programs from the CCP4 suite [40] . The data-collection statistics for the best data sets used in structure determination are shown in Table 2 . P[8]C VP8* Apo1 structure was solved by molecular replacement carried out with the program Phaser [41] and using the structure of VP8* from CRW-8 porcine rotavirus ( PDB 2I2S[20] ) as a model . Initial phases from the molecular replacement were used to manually build the P[8]c VP8* structure with Coot [42] . P[8]c VP8* Apo1 structure was then used as a model for molecular replacement to solve the P[8]c VP8* Apo2 , P[8]c VP8*LNB and P[8]c VP8*H1 structures . All the final models were generated by iterative cycles of refinement using the Refmac [43] and manually optimization with Coot . Data refinement statistics are given in Table 2 . The crystals exhibited good quality control parameters and excellent stereochemistry . Atomic coordinates and structure factors have been deposited in the Protein Data Bank ( PDB ) with ID numbers 6H9W , 6H9Z , 6H9Y and 6HA0 for P[8]c VP8* Apo1 , P[8]c VP8* Apo2 , P[8]c VP8*LNB and P[8]c VP8*H1 , respectively . Structure Superposition and RMSD calculations were carried out with Superpose [44] from CCP4 suite . To assess statistical differences in the ELISA-like binding experiments where many groups were compared an ANOVA test was performed . To analyze significative differences in the Kda values obtained by SPR an unpaired t-test was applied . All statistical analyses were performed with GraphPad Prism version 6 . 0 for MacOsx ( GraphPad Software ) . p values <0 . 05 were considered to be statistically significant . This study was conducted with the approval of the Ethics Committee of the University of Valencia ( code H1544010468380 ) . The human stool samples from Hospital Clínico Universitario de Valencia were anonymized previously to their inclusion in the present study . | The interaction of viruses with host glycans has become an important topic in the study of enteric virus infectivity . This interaction modulates several aspects of the viral cycle , including viral attachment , which in most cases depends on the host glycobiology dictated by the secretor and Lewis genotypes . The capacity to synthesize secretory type-I antigen H ( fucose-α1 , 2-galactose-β1 , 3-N-acetylglucosamine; H1 ) at the mucosae , dictated by the presence of one or two functional copies of the fucosyl-transferase FUT2 gene ( secretor status ) , has been clearly linked to infectivity in important enteric viruses such as the noroviruses . However , a big controversy existed about the contribution of H1 antigen to infection in the leading cause of viral gastroenteritis in young children ( rotavirus ) . It has not been until recently that epidemiological data evidenced a diminished incidence of rotavirus in non-secretor individuals unable to produce H1 . In the present manuscript we offer the evidence that P[8] RV bind H1 via a binding site common for the P[II] RV genogroup and that the H1 precursor lacto-N-biose ( galactose-β1 , 3-N-acetylglucosamine; LNB ) is also bound to this pocket with diminished affinity . The P[8] VP8* structures show a marginal role for the L-fucose moiety from H1 in protein interaction . However , its presence provides conformational changes in the LNB moiety that increase the affinity of VP8* for the H1 ligand and would account for a stronger RV binding to mucosa in individuals expressing H1 ( secretors ) . We thus offer a mechanistic explanation for the different incidence of P[8] RV infection in different secretor phenotypes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"chemical",
"characterization",
"immune",
"physiology",
"crystal",
"structure",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"condensed",
"matter",
"physics",
"microbiology",
"reoviruses",
"viruses",
"rn... | 2019 | Unraveling the role of the secretor antigen in human rotavirus attachment to histo-blood group antigens |
Interactions between HLA class I molecules and killer-cell immunoglobulin-like receptors ( KIR ) control natural killer cell ( NK ) functions in immunity and reproduction . Encoded by genes on different chromosomes , these polymorphic ligands and receptors correlate highly with disease resistance and susceptibility . Although studied at low-resolution in many populations , high-resolution analysis of combinatorial diversity of HLA class I and KIR is limited to Asian and Amerindian populations with low genetic diversity . At the other end of the spectrum is the West African population investigated here: we studied 235 individuals , including 104 mother-child pairs , from the Ga-Adangbe of Ghana . This population has a rich diversity of 175 KIR variants forming 208 KIR haplotypes , and 81 HLA-A , -B and -C variants forming 190 HLA class I haplotypes . Each individual we studied has a unique compound genotype of HLA class I and KIR , forming 1–14 functional ligand-receptor interactions . Maintaining this exceptionally high polymorphism is balancing selection . The centromeric region of the KIR locus , encoding HLA-C receptors , is highly diverse whereas the telomeric region encoding Bw4-specific KIR3DL1 , lacks diversity in Africans . Present in the Ga-Adangbe are high frequencies of Bw4-bearing HLA-B*53:01 and Bw4-lacking HLA-B*35:01 , which otherwise are identical . Balancing selection at key residues maintains numerous HLA-B allotypes having and lacking Bw4 , and also those of stronger and weaker interaction with LILRB1 , a KIR-related receptor . Correspondingly , there is a balance at key residues of KIR3DL1 that modulate its level of cell-surface expression . Thus , capacity to interact with NK cells synergizes with peptide binding diversity to drive HLA-B allele frequency distribution . These features of KIR and HLA are consistent with ongoing co-evolution and selection imposed by a pathogen endemic to West Africa . Because of the prevalence of malaria in the Ga-Adangbe and previous associations of cerebral malaria with HLA-B*53:01 and KIR , Plasmodium falciparum is a candidate pathogen .
Major Histocompatibility Complex ( MHC ) class I molecules are present on the surface of most mammalian cells . There they function as ligands for various receptor families on two types of lymphocyte: the cytotoxic T lymphocyte ( CTL ) of adaptive immunity and the natural killer ( NK ) cell of innate immunity [1] , [2] . NK cells also contribute to reproduction , during formation of the placenta [3] . A key component of all MHC class I molecules is a short peptide , a product of intracellular protein degradation , that is bound during assembly of the MHC class I molecule in the endoplasmic reticulum . After transport to the cell surface , the complexes of peptide and MHC class I molecule are presented for surveillance by NK cell and CTL receptors [4] . In healthy tissue the presented peptides all derive from normal proteins and do not usually stimulate an immune response . In unhealthy tissue , that is infected , cancerous or in other ways damaged , changes occur in the spectrum of peptides presented , which lead to activation of NK cell and CTL mediated immunity [5] , [6] . In mammals , the selection pressures imposed by diverse and rapidly evolving pathogens have driven the evolution of gene families encoding a variety of MHC class I molecules [7]–[9] . These include conserved and highly polymorphic MHC class I molecules with species-specific character [10] . The human MHC , the HLA complex on chromosome 6p21 , has three highly polymorphic MHC class I genes , ( HLA-A , -B and -C ) each of which has thousands of alleles [11] , [12] . Some of the alleles have a worldwide or continent-wide distribution , others are more localized geographically and the majority constitutes rare variants that have been discovered through sequence-based HLA typing of huge cohorts of potential bone-marrow donors for clinical transplantation [12] , [13] . Evolving through mechanisms of point mutation and recombination , pairs of allotypes are distinguished by between one and 51 amino-acid substitutions [14] , [15] . Consistent with natural selection having driven this diversification , the common substitutions are predominantly at ‘functional’ positions of the HLA class I molecule that influence the peptide-binding specificity or the site of interaction with one of the lymphocyte receptors that engage HLA class I molecules [16]–[19] . The antigen receptors of CTL bind to the upper face of the HLA class I molecule , which is formed by the α helices of the α1 and α2 domains and the peptide bound between them [20] , [21] . The genes encoding these αβ T-cell receptors ( TCR ) are diversified during T-cell development by mechanisms of somatic recombination and somatic mutation . These processes produce acquired changes that are not passed on from one generation to the next . In addition , the conserved CD8 co-receptor of CTL , binds predominantly to the conserved α3 domain of the HLA class I molecule [22] . Largely conserved is the leukocyte immunoglobulin-like receptor ( LILR ) B1 , which also binds to the α3 domain [23] and is expressed by some NK cells [24] . NK cells and some T cells express killer-cell immunoglobulin-like receptors ( KIR ) [25] . They bind to the same upward face of the HLA class I molecule as the TCR , with an overlapping but different orientation [18] , [26] . KIR recognition of HLA class I is primarily influenced by polymorphisms in the carboxy-terminal half of the α helix of the α1 domain [18] , [27] . To a first approximation , KIR recognize four mutually exclusive epitopes of HLA-A , -B and -C molecules [28] , [29]: the A3/11 epitope carried by a small subset of HLA-A allotypes , the Bw4 epitope carried by larger subsets of HLA-A and -B allotypes , the C1 epitope carried by many HLA-C and the HLA-B*46 and -B*73 allotypes , and the C2 epitope carried by all the HLA-C allotypes that lack the C1 epitope . Each of these four ligand-receptor interactions is heterogeneous , being further diversified by allelic polymorphism of both the HLA class I and the KIR , as well as by the sequence of the bound peptide [5] , [30]–[32] . By providing resistance to specific diseases , this combinatorial diversity is believed to give individuals and populations the means to fight wide ranging pathogen diversity [9] , [28] , [29] , [33] . The KIR locus on chromosome 19q13 . 4 exhibits an extensive variability in human populations , one comparable to that of the HLA class I genes [28] , [34] . KIR haplotypes differ in the content and copy number of KIR genes and are further differentiated by allelic polymorphism of the constituent genes [28] , [35]–[37] . On the basis of gene content , human KIR haplotypes , but not their counterparts in other hominoid species [38] , divide into two groups [36] . These ‘A’ and ‘B’ haplotype groups are maintained by all human populations and are differentially associated , either alone or in combination with HLA class I , with susceptibility to diverse diseases , reproductive success , and the outcomes of therapeutic transplantation [28] , [39]–[46] . The nature of these correlations has suggested a scenario in which the A and B haplotypes are maintained by competing selection on the functions that NK cells serve in resisting infectious disease and in establishing the placenta during the early stages of pregnancy [29] . Although KIR diversity has been studied in numerous ( N = 105 ) human populations at the low-resolution of KIR gene content [47] , [48] , high-resolution analyses of allelic and haplotypic diversity have been few ( N = 4 ) and involved populations such as the Japanese and Yucpa Amerindians that have restricted genetic diversity as a consequence of historical population bottlenecks [49]–[52] . By contrast , little is known of the KIR system and its interactions with HLA class I in sub-Saharan Africans , the human populations with highest genetic diversity [53] , [54] . By using a novel combination of molecular , genetic and computational methods we have defined at high-resolution the rich diversity of KIR and HLA class I in the Ga-Adangbe population of one village in southern Ghana , West Africa .
Initial low-resolution analysis of the Ga-Adangbe KIR locus identified 19 KIR gene-content haplotypes ( Figure 1A ) and 16 different KIR genotypes ( Figure S3 ) . The 53% frequency of the KIR A haplotype ( h1 ) is comparable to the 47% combined frequency of the 18 KIR B haplotypes ( h2–h19 ) , consistent with balancing selection having been active on the two haplotype groups [49] . The number of KIR genes per B haplotype varies from four ( h9 ) to twelve ( h11 ) , with only two genes , KIR3DL3 and KIR2DL2/3 , being detected on every haplotype . By frequency , over 10% of the Ga-Adangbe KIR haplotypes ( h5 , 7–10 , 12 , 13 ) lack one of the three framework genes ( KIR3DL3 , KIR2DL4 and KIR3DL2 ) that define the structure of the KIR locus [37] and its organization into centromeric and telomeric regions; haplotypes h5 and h13 lack KIR2DL4 , whereas haplotypes h7–10 and h15 lack KIR3DL2 . In previous studies of non-African populations such haplotypes were either absent [49] , [51] or rare [50] , [52] . Haplotype h12 has a duplication of the KIR2DL4 and KIR3DL1/S1 genes , of the sort that has been described previously in Europeans [28] , [55] , [56] and South and East Asians [57] . Seven centromeric region motifs combine with six telomeric region motifs to form the 19 Ga-Adangbe KIR gene-content haplotypes ( Figure 1A ) . By far the most common motif is tA01 , which is fixed on A haplotypes and present at a frequency of 86% in this population . Consequently , the Ga-Adangbe , as well as other sub-Saharan African populations , has significantly reduced gene-content diversity in the telomeric region of the KIR locus compared to non-African populations ( p<0 . 001 , Figure 1B ) . In contrast , centromeric region KIR diversity is much higher and comparable to that of other population groups . To give a complete comparison of KIR variation in the centromeric and telomeric regions of the Ga-Adangbe KIR locus , we performed high-resolution typing to determine the allelic diversity of the component KIR genes . A total of 175 KIR variants were found , of which 126 involve allotypic differences: 32 of these being previously undiscovered ( Figure S4A–C ) . The individual KIR genes exhibit high heterozygosity ( H ) , particularly the KIR3DL3 framework gene , which is present on every haplotype and has H of 0 . 93 ( Figure S4C ) . This heterozygosity exceeds that of the highly polymorphic HLA class I genes and is clearly an outlier amongst genome-wide multi-allelic markers from West African populations ( Figure S4D–E ) . With addition of the high-resolution analysis , the 19 gene-content KIR haplotypes become subdivided into 208 allele-level haplotypes ( Figure S5 ) ; of these a large majority ( 195/208; 95% ) encode unique combinations of KIR proteins and have the potential to be functionally distinct ( Figure 2A and S4F ) . Most diverse is h1 , the canonical KIR A gene-content haplotype ( Figure 1A ) , for which there are 108 different allele combinations and 100 allotype combinations ( Figure S4F ) . Individually , none of the 18 KIR B gene-content haplotypes approaches h1 in diversity , but when both gene-content and allotype-content diversity are taken into account the 100 A and 95 B KIR haplotypes have comparable diversity as well as frequency . None of the allele-level KIR haplotypes dominate the Ga-Adangbe population; the frequency of the most common haplotype is only 6% and only 18 of the 195 functionally different haplotypes exceed a frequency of 1% ( Figure 2B ) . Thus the Ga-Adangbe population is seen to have a rich diversity of KIR haplotypes upon which natural selection can operate . The centromeric region of the Ga-Adangbe KIR locus exhibits a bimodal mismatch distribution , a network indicating successive formation and expansion of haplotypes , and a significantly elevated value for Tajima's D ( Figure 3A–C ) . All these features reflect the presence of a variety of divergent haplotypes that are at comparable frequencies and maintained by balancing selection . In contrast , the telomeric region of the KIR locus displays a unimodal mismatch distribution , a star-like haplotype network pattern ( Figure 3C ) and a Tajima's D value significantly below that expected for neutrality ( Figure 3B ) , features reflecting the presence of numerous closely-related variants under directional selection . Such difference in the evolution of the centromeric and telomeric KIR regions is not a general feature of human populations , as exemplified by comparison of the Ga-Adangbe with Yucpa Amerindians and US Europeans ( Figure S6 ) . Sliding-window analysis showed that the boundary between the high diversity and low diversity parts of the Ga-Adangbe KIR haplotype does not correspond precisely with the conventional division of the locus into centromeric and telomeric regions ( Figure 4 ) . High diversity extends into the KIR3DL1/S1 gene of the telomeric region , but sharply declines at the end of exon 3 that encodes the D0 domain , resulting in low diversity that is maintained throughout the rest of the telomeric KIR region . This result is consistent with our previous analysis of KIR3DL1/S1 polymorphism worldwide , which showed that balancing selection was restricted to the D0 domain in sub-Saharan Africans [35] . The functional consequences are first that mutations in the D0 domain can abrogate cell surface expression [58] or decrease binding to HLA-B [18] and second that reduced diversity in the D1 and D2 domains favors one particular type of ligand specificity [32] , [35] , [59] . Segments of low diversity that are of comparable length to the one in the telomeric KIR region are infrequent in the genomes of sub-Saharan Africans , as shown from analysis of Yoruba West Africans ( p<0 . 01: Figure 4 ) , a population related closely to the Ga-Adangbe [60] , [61] . In summary , intron 3 of the KIR3DL1/S1 gene marks the boundary between a diversified centromeric part and a conserved telomeric part of the KIR locus in sub-Saharan Africans . The centromeric region of the KIR locus encodes inhibitory receptors KIR2DL1 and KIR2DL2/3 that recognize the C1 ( KIR2DL2/3 ) and C2 ( KIR2DL1 ) epitopes of HLA-C , whereas the telomeric region encodes inhibitory KIR that recognize the A3/11 epitope ( KIR3DL2 ) of HLA-A and the Bw4 epitope ( KIR3DL1/S1 ) of HLA-A and -B [37] . In the Ga-Adangbe we identified 26 HLA-A , 32 HLA-B and 23 HLA-C allotypes ( Figure 5 ) . The numbers of alleles and their composition are typical of West African populations , which are readily distinguished from other African population groups by clustering analyses based on these genes alone ( Figure S7 ) . In contrast to KIR , none of the HLA class I alleles were novel or private to the Ga-Adangbe population . Significantly high values of Tajima's D provide good evidence for balancing selection having acted on all three polymorphic HLA class I genes ( Figure 6A ) , as observed previously for other populations [34] . HLA class I has various roles in immunity and reproduction that involve binding to peptide fragments , and serving as ligands for KIR and other lymphocyte receptors . To assess if any of these functions influenced HLA class I allele-frequency distributions in the Ga-Adangbe we analyzed the amino acid sequence of each of the binding motifs separately . In this analysis , we considered both the binding site for peptide antigens , and the sites of interaction with four types of lymphocyte receptor: the TCR and CD8 of CTL , and the KIR and LILR of NK cells . Of the three extracellular domains of the HLA class I molecule , the α1 and α2 domains mediate interactions with peptide , TCR and KIR , whereas the α3 domain mediates interaction with LILR and CD8 ( Figure 6B and Figure S8A–C ) . Because some of the motifs overlap , we analyzed only those residues located exclusively in each type of binding site . We analyzed the allele-frequency spectrum of each motif using the Ewens-Watterson test [62] and compared the deviation from neutral expectations for each motif using a normalized statistic ( Fnd [63] ) . The results showed strong evidence for balancing selection acting on the peptide-binding residues of all three HLA class I molecules , but no evidence for natural selection acting on the TCR binding motifs ( Figure 6C ) . The CD8 binding site is largely invariant for HLA-B and -C , whereas for HLA-A variation is introduced at residue 245 by A*68:01 . Mutation of residue 245 can influence CD8 binding [64] but there was no evidence of this being selected in Ga-Adangbe ( Figure 6C ) . These distinctions among motifs demonstrate that our analysis differentiates the effects of natural selection acting independently on each of the functional motifs of HLA class I molecules . The analysis is also consistent with codon-by-codon tests for selection , which show that HLA class I evolution in hominids has been driven by diversification of the peptide-binding motifs and not TCR or KIR binding motifs ( Figure S9 ) . Such independent evolution has likely been facilitated by the extensive intra-locus recombination and gene conversion that shaped HLA class I diversity by shuffling functional motifs among allotypes [14] , [15] . Analysis in Ga-Adangbe of the frequency distributions of HLA-A and -B motifs that interact exclusively with KIR3DL1 , gave evidence for balancing selection that was statistically significant and of magnitude greater than either the peptide-binding motifs and/or the locus as a whole ( Figure 6C ) . Although the magnitude of the Fnd values approached those of their respective peptide-binding domains , there was less evidence for balancing selection at the KIR-exclusive motif of HLA-B in populations not from West Africa , and the observation only reached statistical significance in the Ga-Adangbe ( Figure 6D and Figure S10 ) . Included in the KIR-exclusive motif is arginine at position 83 , a component of the Bw4 epitope and the only Bw4 residue necessary for HLA-B binding to KIR3DL1 [65] . In the Ga-Adangbe arginine 83 is present in 16 HLA-B allotypes having a combined frequency of 46% ( Figure 5 ) . HLA-B*35:01/*53:01 and HLA-B*49:01/B*50:01 comprise pairs of HLA-B allotypes that differ only by presence/absence of the Bw4 motif [66] . This difference determines whether these HLA-B alloypes bind to KIR3DL1 ( B*53:01 and B*49:01 ) or do not ( B*35:01 and B*50:01 ) [67] . HLA-B*35:01 and HLA-B*53:01 are both common in the Ga-Adangbe ( Figure 5 ) suggesting that distinction between binding or not binding to KIR3DL1 has been a major influence on the balancing selection acting on HLA-B , and that this variation substantially augments the diversity of peptide-binding function . Further , it implies that the presence/absence polymorphism of Bw4 is driven by the benefits of diversifying the interaction of HLA-B with KIR3DL1 , and not its interaction with peptides . For HLA-A , polymorphic residues within the KIR-exclusive motif include positions 17 and 142 and are provided primarily by the HLA-A*02 , -A*30 and -A*68 allotypes . None of these allotypes is known to interact with KIR and all are common in sub-Saharan African populations [13] , [68] . In contrast to HLA-A and -B , the HLA-C residues that interact exclusively with KIR are monomorphic ( Figure S8 ) and all expressed HLA-C allotypes are presumed to interact with KIR [31] . To examine the impact of natural selection on the LILRB1-contacting residues of HLA class I we first performed likelihood ratio tests for selection on hominid α3 domains . This analysis revealed evidence for diversifying selection on HLA-C , and codon-by-codon analysis identified the LILRB1-contacting residues for all three HLA class I molecules ( Figure S9B ) . Although statistical confidence from this phylogenetic-based analysis was low ( Figure S9B ) , frequency-based Fnd analysis suggested that balancing selection has acted on the LILRB1-interacting motifs of Ga-Adangbe HLA-A and -B ( Figure 6C ) . Their Fnd values were greater in magnitude than those of the respective peptide-binding motifs and reached statistical significance for HLA-A . In the Ga-Adangbe , HLA-A molecules that bind LILRB1 with low affinity ( 193A/194V , 47% ) are at similar frequency as the high-binding allotypes ( 193P/194I , 53% ) [69] . Together , these results demonstrate that balancing selection has acted on HLA class I in the Ga-Adangbe population , resulting in the evolution of a diversity of ligands for interaction with NK cell receptors . We next measured the scale of KIR and HLA combinatorial diversity and assessed if the interacting receptors and ligands continue to co-evolve . Each individual in the Ga-Adangbe panel has a unique compound genotype of KIR and HLA-A , -B and -C ( Figure S11 ) . Based on the known interactions between KIR and the C1 , C2 , Bw4 and A3/11 epitopes of HLA class I , we determined the number of functional ligand-receptor pairs for all members of the Ga-Adangbe panel . The frequencies of these values within the panel gave a normal distribution ( Figure 7A ) with a mean number of ligand-receptor interactions of eight ( 95% CI of 3–12 ) . To assess for co-evolution of KIR with HLA in the Ga-Adangbe , we used the Mantel test of congruence between distance matrices to look for population-wide correlation between KIR and HLA class I genotypes [70] . These analyses revealed significant correlations of matrices for KIR3DL1/S1 and HLA-B genotypes ( p<0 . 001 ) , for KIR2DL2/3 with HLA-C ( p<0 . 01 ) , and for KIR2DL1 with HLA-C ( p<0 . 01 ) ( Figure 7B–D ) . However , no correlations were observed between either KIR3DL1/S1 or KIR3DL2 and HLA-A . Residues 31 , 44 and 86 , in the D0 of KIR3DL1 , are in complete LD and were correlated in synergistic action with three groups of HLA-B residues ( Figure 7B and S12A–B ) . That the correlation also involves residues of the Bw4 epitope , is consistent with interaction between KIR3DL1 and HLA-B being the underlying mechanism driving their population frequencies . Further contributions from HLA-B are made by residue 114 , and three residues in complete LD , 24 , 45 and 194; the latter contacting LILRB1 ( Figure 7B ) and having enhanced diversity in the Ga-Adangbe ( Figure 6C ) . Residues 24 and 114 are located in the peptide binding B and F pockets , respectively , which define the anchor residues of the peptide that is presented by HLA-B ( Figure S8 and [71] ) . This result suggests that sequences of the peptides presented by HLA-B contributed to its co-evolution with KIR3DL1 in the Ga-Adangbe . A previous analysis showed replacement of isoleucine 194 in HLA-B with valine reduced the interaction with KIR3DL1 as measured by NK inhibition [65] . The study also demonstrated that polymorphism at positions in the B and F pockets of the peptide-binding site can impact 3DL1-mediated inhibition , either alone or in concert with residue 194 . Moreover , the correlations observed here between HLA-B and KIR3DL1 are all supported by the results of functional studies , which assessed the influence of the sequence of the peptide bound to HLA-B on the binding to KIR3DL1 [18] , [27] , [32] , [72] . Differences between the KIR2DL2 and KIR2DL3 subsets of KIR2DL2/3 allotypes have had major impact in the co-evolution of KIR2DL2/3 with HLA-C . For HLA-C , the major factor in this co-evolution is a group of seven residues in LD ( positions 194 , 261 , 273 , 311 , 313 , 332 , 345 ) , which includes residue 194 that contacts LILRB1 ( Figure 7C ) . This group of residues distinguishes HLA-C*07 , a common allotype in many populations , from all other C1-bearing HLA-C allotypes ( Figure 5 ) . Because of the strong LD between KIR2DL2/3 and KIR2DL1 ( D′ = 0 . 87 ) , this group of residues also correlates with C2-specific KIR2DL1 ( not shown ) although this receptor does not recognize C1-bearing HLA-C*07 [31] . The analysis revealed an independent influence from residue 49 ( Figure 7D ) which distinguishes HLA-C*04 , the most frequent HLA-C allotype in the Ga-Adangbe ( Figure 5 ) , from all other C2-bearing allotypes ( Figure 5 ) . Five residues of KIR2DL1 ( positions 154 , 163 , 182 , 216 and 245 ) contribute to its co-evolution with HLA-C . These five residues , which are in complete LD , include residue 182 that contacts HLA , and residue 245 that modulates both ligand-binding and signaling functions [26] , [73] . These are all residues that distinguish KIR2DL1*003 from KIR2DL1*004 , encoded by the common KIR2DL1 alleles of the centromeric A and B motifs , respectively ( Figure 2 and Figure S5 ) . For the cenA-containing KIR haplotypes , which carry KIR2DL3 and KIR2DL1 , 80% of the KIR2DL1 allotypes have histidine 182 and arginine 245 and are strong high-expressing C2 receptors , whereas the other 20% of allotypes have cysteine 245 and are weak , low-expressing C2 receptors . In contrast , 80% of the cenB haplotypes carry KIR2DL2 and either lack KIR2DL1 ( 49% ) or encode weak , low-expressing allotypes having arginine 182 and cysteine 245 ( 31% ) .
Variable interactions between KIR and HLA class I influence the immunological and reproductive functions of NK cells . Because of the complexity of the KIR gene family , population genetic studies have been limited in large part to low-resolution analyses of KIR gene-content variation [47] , [48] , [74] . In developing methods for high-resolution KIR genotyping , we previously focused on Asian and Amerindian populations having inherently low genetic diversity because of their demographic histories [49] , [51] . At the other end of the human spectrum are sub-Saharan African populations , who have , genome-wide , greatest genetic diversity . Reflecting this general characteristic , are the results presented here from our high-resolution analysis of KIR and HLA-A , -B and -C variation in the Ga-Adangbe population of Prampram , a coastal village in Ghana , West Africa . Segregating in this population are 81 HLA and 175 KIR variants , numbers that are four- to five-fold higher than the 19 HLA and 30 KIR variants we previously described for the Yucpa population of South American Indians [49] . Thus , we find the Ga-Adangbe population to be highly heterozygous , with every individual having a unique compound genotype for KIR and HLA class I . As they have similar levels of KIR gene-content ( Figure 1 ) and HLA class I [13] , [75] heterozygosity to other West African populations , the Ga-Adangbe provide an archetypal population for investigating immune diversity . The consequence of genetic individuality is predicted to be functional individuality in the immune responses to viruses and other pathogens against which NK cells and CTL are important elements of the defences of human immune systems . The unprecedented diversity of HLA and KIR haplotypes and alleles , and their relatively even distributions , argue that strong balancing selection on these loci has been a persistent force in the history of the Ga-Adangbe population . Probable causes of this selection include reproductive success [29] and the fluctuating pressures imposed by the variety of human pathogens in West Africa and their continual evolution to evade the immune systems of their human hosts [33] . Consistent with these roles , we identified strong balancing selection of the centromeric KIR region and co-evolution between KIR2DL1 , KIR2DL2/3 and HLA-C . Upon this background of strong balancing selection we have also identified signatures of directional selection on the telomeric region genes of the KIR locus . The telomeric region has a much lower diversity than occurs in non-African populations , due to the low frequency of the telomeric B motifs ( 14% ) and a corresponding increase in the frequency of the telomeric A motif . This bias is consistent with pressure from infectious disease [28] being stronger than that from reproductive disorders [29] . For example , KIR2DS1 , a component of telomeric B and thus infrequent in the Ga-Adangbe ( Figure 1A ) is the major KIR factor that protects against pre-eclampsia in European populations [46] . Although the two gene families are on different chromosomes , low-resolution analysis showed that KIR and their HLA ligands have evolved in concert across populations worldwide [48] , [76] . Here , using high-resolution analysis of a well-defined population having substantial genetic diversity , we identified an on-going molecular co-evolution . That the analysis only identified functionally interacting components of known ligand-receptor pairs demonstrates the correlations are due to natural selection and not chance [77] . We also identified the differential action of natural selection on the motifs of HLA class I molecules that interact with lymphocyte receptors . Diversification of peptide binding has been the major outcome of balancing selection on all three HLA class I molecules and has continued throughout hominid evolution to the present day . Through the same time period the TCR-interacting motifs have been evolving under selective neutrality , consistent with T-cell diversity being generated by somatic , not heritable , mutation [20] , [21] . Contrasting both of these patterns we detected on-going balancing selection of the KIR-contacting motif of HLA-B , and this selection was strongest in the Ga-Adangbe . Whereas varying selection pressures have resulted in a high number of different peptide binding motifs , selection on the KIR-interacting motif ( Figure 4 ) and its co-evolution with KIR3DL1 ( Figure 7 ) are likely driven by the two extreme phenotypes of receptor ligation or no ligation . This suggests these phenotypes each provide both an advantage and a potential cost to the host . This mode of balancing selection is strikingly similar to the deleterious mutants of haemoglobin that provide resistance to Plasmodium falciparum malaria but also impair erythrocyte function [78] . Illustrating the binary nature of balancing selection at the KIR-interacting motif of HLA-B are two common Ga-Adangbe allotypes that differ only at residues 77–83 . HLA-B*53:01 has the Bw4 motif and is therefore a ligand for KIR3DL1 and HLA-B*35:01 does not have the motif . HLA-B*53:01 originated in West Africa as the product of a gene conversion between HLA-B*35 and a second , unknown allele [66] . That it remains localized to West Africa [13] and combines high prevalence with low haplotype diversity is consistent with HLA-B*53:01 having risen rapidly in frequency due to natural selection likely in response to pressure exerted by P . falciparum [79] . Both B*35 and B*53 can elicit CTL responses to this pathogen through distinct but overlapping peptide repertoires [79] . Thus , the capacity of HLA-B*53:01 to also interact with NK cells may contribute to its observed protective effects , whilst parasite strain-specific differences could contribute to its detrimental effects . Supporting this interpretation are the high incidence of malaria caused by P . falciparum in the Ga-Adangbe population [80] , its impact on human health and genomes [81]–[83] and associations with combined KIR and HLA genotypes [39] . Moreover , there is no other single pathogen in West Africa that carries such a high pre-reproductive mortality as malaria [33] . In examining the sites on HLA class I that interact with different types of lymphocyte receptor we found that diversity in the LILRB1 binding site on the α3 domain of HLA-A , -B and -C is enhanced through balancing selection . We also identified co-evolution of KIR with HLA class I and also of the LILRB1 interaction with HLA class I . Supporting these results are functional data showing that the LILRB1-contacting residues and the peptide binding motif influence KIR3DL1 binding to HLA-B [18] , [27] , [32] , [65] , [72] . Thus , mutations within the LILRB1-binding motif could affect KIR ligation indirectly through their influence on HLA class I structure [65] or aggregation of receptor/ligand complexes [26] . In parallel , diversity in the LILRB1 contact site on HLA class I could serve to thwart viruses , such as cytomegalovirus ( CMV ) , that evolve mimics of HLA class I to protect virus-infected cells from NK cell attack [84] . Any collateral loss of HLA recognition by LILRB1 will be limited through presence of multiple functionally-related receptors , such as other LILR , KIR or CD94/NKG2 molecules [2] , [9] , [24] , [85] . Pointing to the selection pressure exerted by CMV are its impact on individual NK repertoires , prevalence in African populations , and the risk of mortality associated with perinatal transmission of the virus [41] , [86] , [87] .
The research we report here was conducted with approval from the Stanford University School of Medicine Institutional Review Board and the Ghanaian Ministry of Health . The population we studied were residents of Prampram , a coastal fishing village of 7 , 000 inhabitants situated 50 km east of Accra and south of the Volta Basin in the Greater Accra region of Ghana . Malaria ( 98% Plasmodium falciparum ) is endemic in Prampram , with a mean of 8 . 5 infectious bites/person/year [80] . In the course of a study to determine the patterns of malaria infection in children , samples of genomic DNA were obtained from 131 newborn infants and from 104 of their mothers [80] . The subjects are from the Ga-Adangbe ethnic group , which currently comprises 2 million individuals in total . Archaeological data and accompanying historical accounts , combined with linguistic and genetic evidence indicate that Ga-Adangbe ancestors first lived in the region of present-day Nigeria or Burkina Faso before the Bantu expansion ( ∼3000 years ago ) and then migrated to the Volta Basin 750–1000 years ago [88] , [89] . The Ga-Adangbe speak a Kwa language of the non-Bantu Niger Kordofanian family . Analysis of autosomal genetic markers indicates that the Ga-Adangbe are closely related to the Akan , also from Ghana [60] . The Akan and other closely-related Ghanaian populations , such as the Ashanti , have similar composition of both mitochondrial and Y-chromosome haplogroups , supporting the demographic model that the Ga-Adangbe derive from a population that lived in West Africa prior to the Bantu migration [90] , [91] . Nucleotide sequences were determined for the exons of KIR genes from 16 Ga-Adangbe children who were chosen at random to represent the study population . The sequences of newly discovered alleles were confirmed by re-amplification , cloning and sequencing; or by direct sequencing of the PCR products obtained from homozygous and/or hemizygous individuals . When possible , new alleles were also confirmed by amplification and sequencing of the same gene from the mother . From this dataset of Ga-Adangbe KIR sequences , we developed a pyrosequencing-based method for KIR genotyping that distinguishes all known variants , including those detected in the 16 randomly selected children ( Figure S1 and Figure S2 ) . Pyrosequencing provides a semi-quantitative measure of SNP genotypes ( the peak-height ratio ) that determines both allele identity and copy-number genotype [57] . We further exploited this feature to genotype combinations of KIR genes having exons that are difficult to distinguish using standard genotyping technology . In this manner KIR2DL1 and KIR2DS1 , which are different genes with high sequence similarity , were genotyped together , as were KIR2DL2/3 and KIR2DS2 . Similar criteria were used to distinguish exons 1 and 2 of KIR2DL5 from those of the related KIR3DP1 pseudogene . KIR2DS3 and KIR2DS5 , which are relatively uncommon in the Ga-Adangbe population , were subjected to standard Sanger sequencing in addition to pyrosequencing . The combined method targets 304 coding-region SNPs , of which 190 are non-synonymous , to discriminate 350 KIR alleles ( 247 KIR allotypes ) . Following allele-specific genotyping , 20 individuals were chosen either at random , or because of their unusual pyrosequencing patterns , and the nucleotide sequences of their KIR exons determined by standard sequencing . Pyrosequencing reactions were performed using PyroGold reagents and a PSQ HS 96A machine ( Qiagen , Valencia , CA ) . KIR gene content was confirmed by results from bead-based sequence-specific oligonucleotide probe hybridization ( SSOP ) , which tests for the presence of 13 KIR genes ( KIR2DL1-5 , KIR2DS1-5 and KIR3DL1-3 ) . The assay was performed using LABType reagents ( One Lambda , Canoga Park , CA with KIR lot #4 ) and detected using a Luminex-100 instrument ( Luminex corp . Austin , TX ) . The cohort of 235 Ga-Adangbe individuals was genotyped for HLA-A -B and -C at allele-level resolution using bead-based SSOP hybridization that was detected with a Luminex-100 instrument ( Luminex corp . Austin , TX ) . The assays were performed using lots #11 ( HLA-A ) , #14 ( HLA-B ) , and #9 ( HLA-C ) of LABType SSO reagents ( One Lambda , Canoga Park , CA ) . To identify variants that are common in the Ga-Adangbe but not detected by the probes , we further investigated all individuals who typed homozygous for HLA-A , -B , or -C by sequencing their putative homozygous genes . PCRs were performed using a Perkin-Elmer 9600 thermal cycler ( or a Veriti 96-Well instrument using 9600 emulation mode ) with a three minute denaturing step at 94°C , 10 cycles of ( 94°C 10 s; 65°C 60 s ) and 20 cycles of ( 94°C 10 s , 61°C 50 s , 72°C 30 s ) . Standard DNA sequencing reactions were performed in forward and reverse directions using BigDye Terminator v3 . 1 and analyzed using an ABI-3730 sequencer ( ABI , Foster City CA ) . When required , PCR products were cloned using Topo-pcr2 . 1 vector ( Invitrogen , Carlsbad CA ) and sequenced using M13 and internal primers . All of the newly-discovered alleles described herein were validated according to the guidelines recommended by the curators of the Immuno Polymorphism Database ( IPD ) [12] . At least five clones of the desired allele were sequenced from each individual examined . Newly identified allele sequences were submitted to Genbank and the IPD database with accession numbers indicated below and in Figure S2 . KIR genes and alleles were named by the KIR nomenclature committee [92] formed from the WHO Nomenclature Committee for factors of the HLA system , and the HUGO Genome Nomenclature Committee . A curated database is available at http://www . ebi . ac . uk/ipd/kir/ [12] . <D> denotes the number of Ig-like Domains , <L> a Long , inhibitory , cytoplasmic tail <S> a Short , activating , tail and <P> a Pseudogene . A unique DNA sequence that spans a KIR coding region is considered an allele and those that yield unique proteins are considered to define an allotype . The first three digits distinguish the allotypes , the fourth and fifth digits distinguish synonymous variation . To give an example: KIR3DL1*01501 and KIR3DL1*01502 are synonymous variants of the KIR3DL1*015 allele , and encode the KIR3DL1*015 allotype – an inhibitory receptor having three Ig-like domains . KIR haplotypes are named according to the criteria described by Pyo et al . [93] . KIR haplotypes are divided into centromeric ( c ) and telomeric ( t ) regions , or segments , that are of two forms: A and B . The two letters in the haplotype nomenclature define the four types of segment: cA , cB , tA and tB . Following these letters are two digits that uniquely define the different gene-content motifs for each type of segment: for example cA01 and cA02 . Following these designations of gene-content motif are two sets of three digits that are separated by colons and distinguish motifs having identical gene content but differing by one or more allelic polymorphisms . The first set of three digits denotes differences that include non-synonymous variation , whereas the second three digits denote differences that are only synonymous or non-coding . The high heterozygosity observed for each KIR and HLA class I gene in the Ga-Adangbe , coupled with analysis of mother-child pairs , allowed unambiguous deduction of HLA class I and KIR allele-level haplotypes . Core sets of 208 HLA and 208 KIR haplotypes were deduced by segregation analysis in 104 mother-child pairs . These sets of haplotypes were used as priors in PHASE 2 . 1 [94] analyses which deduced 54 HLA class I and KIR haplotypes from the remaining 27 unrelated individuals . The final data set consisted of 262 independent HLA class I and KIR haplotypes . Population statistics were calculated from the set of 131 children ( 2N = 262 ) . For some analyses , in which we estimated the total KIR and HLA diversity in the Ga-Adangbe population , total numbers of 366 independently segregating HLA class I and KIR haplotypes were used ( 262 haplotypes from the set of unrelated children , plus 104 non-segregating maternal haplotypes ( 2N = 366 ) ) . The distributions of HLA-A , -B and -C alleles were compared in 108 populations , including the Ga-Adangbe , for which high-resolution genotyping data were available . These comprised 103 of the 497 populations studied by Solberg et al . [13] , of which 11 are sub-Saharan Africans , and four additional sub-Saharan populations: Ugandans from Kampala [95] , Yorubas from Ibadan in Nigeria [96] , KhoeSan from Southern Africa and Hadza from Tanzania [75] . Data from a total of 31 , 298 individuals were used in the analyses described here . Statistica 10 ( StatSoft Inc . Tulsa OK ) was used to perform principal component analysis on the frequencies of every HLA-A , -B and -C allele present in four or more of the 108 populations ( 242 alleles: 70 A , 129 B , 43 C ) . Population clustering analysis , performed using STRUCTURE 2 . 3 . 3 [97] , was restricted to populations where information for each individual was available . The analysis was performed assuming the model of correlated allele frequencies among ancestral clusters , with a 1 , 000 step burn-in stage , 10 , 000 step run stage and 5 replicates . The influence of linkage disequilibrium ( LD ) between markers was reduced by including only HLA-A and -B , which are separated by ∼1 . 4 Mb . For comparison of gene-content diversity of centromeric ( cen ) and telomeric ( tel ) region KIR haplotypes across worldwide populations , haplotype frequencies were obtained from population studies that discriminated 2DL5cen ( KIR2DL5B ) from 2DL5tel ( KIR2DL5A ) and for which the data are available from allelefrequencies . net [74] . There were 72 populations satisfying these criteria with a mean N of 105 individuals per population . Tajima's D measures the impact on allele-frequency spectra of directional selection favoring a single allele ( D<0 ) , or balancing selection favoring multiple alleles ( D>0 ) [98] . Tajima's D was calculated using DnaSP 4 . 1 [99] . Statistical significance was assessed by comparing the observed values with those expected under neutral-drift equilibrium , in a range of demographic models generated using the program ms [100] . When evidence remains significant under all reasonable demographic models , the allele distributions are unlikely to have arisen through neutral genetic drift . The demographic models were as described previously [35] . Watterson's homozygosity F test provided the first evidence that balancing selection was acting on HLA molecules [101] . The statistic , which is the proportion of homozygotes expected under Hardy-Weinberg equilibrium , was calculated from the frequencies of allotypes for given HLA class I motifs using the exact test described by Slatkin [102] and implemented in the Pypop software package [103] . The reported p-value is the probability of obtaining an F statistic less than the observed value if the motif was evolving under neutrality . It is based on the null distribution of F values simulated under neutrality/equilibrium conditions and on the observed number of alleles ( k ) of any given motif and sample size ( 2N ) . In order to directly compare the magnitude of deviation from neutral expectations for motifs with differing numbers of alleles , we computed the normalized deviate of the homozygosity statistic ( Fnd ) . Fnd is the difference between the observed homozygosity , divided by the square root of the variance of the expected homozygosity . This calculation is implemented in Pypop , with variance values obtained through simulations [63] . Significant negative values of Fnd indicate balancing selection , while significant positive values of Fnd indicate directional selection . PAML 4 . 5 [104] was used to identify codons subject to positive diversifying selection . Neighbour-joining ( NJ ) and Bayesian phylogenetic analyses to provide input for PAML were performed as described previously [35] using Mega 5 [105] and MrBayes 3 . 2 . 1 [106] . The MHC-C data set used corresponded to release 2 . 21 of the IPD database [12] which included 340 alleles unique through exons 2 and 3 ( α1 and α2 domains ) of HLA-C , plus all unique chimpanzee and orangutan MHC-C alleles having sequences complete through these exons . Similarly for the α3 domains of MHC-A , -B and -C , all unique human , chimpanzee and orangutan exon 4 sequences were used . Haplotypes of coding sequence were constructed by concatenating the sequences of the KIR alleles identified by pyrosequencing . A gapped alignment was used to account for gene absence and the duplicated copies of 2DL4 and 3DL1 observed in a single individual were not included . Haplotype networks were created with the Hamming distance model using the haploNet function of Pegas 0 . 4-3 [107] . The node probability was calculated according to Templeton et al . [108] using Pegas 0 . 4-3 . Mismatch distributions were calculated with p-dist and pairwise deletion using Mega 5 [105] . For all the populations described as West African by Tishkoff et al . [83] and having N>20 , heterozygosity was calculated for each non-GATA microsatellite . The percentile range was then calculated from these 6659 data points . Heterozygosity was calculated using Nei's unbiased estimator [109] . Distance matrices ( p-distance; number of SNPs which differ , divided by number of SNPs ) between individuals in the study cohort ( N = 131 ) were calculated from SNP genotypes using the ‘dist . gene’ function in the ‘ape’ ( Analyses of Phylogenetics and Evolution: ver . 3 . 0-6 [110] ) , package for the R language for statistical computing [111] . Mantel's permutation test for similarity of matrices [70] was implemented for pairwise combinations of distance matrices using the ‘mantel . test’ function of ‘ape’ . The function compares the observed value of the z statistic for correlation to a distribution obtained by permuting the rows and columns of data . 10 , 000 permutations were performed . The SNPs were phased and haplotypes concatenated prior to analysis . In the first round single polymorphic HLA residues were compared with complete KIR genotypes; those showing significant correlation were then tested against single KIR residues . From the LD ( r2 ) values , groups of residues in linkage disequilibrium that contribute to the correlation between genotypes were then identified . Further iterations allowed the identification of single residues and groups of residues having the highest correlation between HLA class I and KIR . EU272647 ( KIR3DL2*029 ) , EU272648 ( KIR3DL2*00302 ) , EU272652 ( KIR3DL2*049 ) , EU272654 ( KIR3DL2*032 ) , EU272657 ( KIR3DL2*023 ) , EU272660 ( KIR3DL2*024 ) , FJ666320 ( KIR3DL2*035 ) , FJ666322 ( KIR3DL2*037 ) , FJ666323 ( KIR3DL2*038 ) , FJ666325 ( KIR3DL2*040 ) , FJ883770 ( KIR3DL3*032 ) , FJ883771 ( KIR3DL3*033 ) , FJ883772 ( KIR3DL3*01406 ) , FJ883773 ( KIR3DL3*00903 ) , FJ883774 ( KIR3DL3*00208 ) , FJ883775 ( KIR3DL3*01502 ) , FJ883776 ( KIR3DL3*02502 ) , FJ883777 ( KIR3DL3*01602 ) , FJ883778 ( KIR3DL3*034 ) , FJ883780 ( KIR3DL3*035 ) , GQ478175 ( KIR3DL3*02702 ) , GQ906701 ( KIR2DL4*013 ) , GU301909 ( KIR2DS5*011 ) , GU323350 ( KIR2DL1*01201 ) , GU323352 ( KIR2DL1*01102 ) , GU323351 ( KIR2DL1*01202 ) , GU323353 ( KIR2DL1*020 ) , HM211183 ( KIR2DL3*018 ) , HM211184 ( KIR2DL3*01202 ) , HM211185 ( KIR2DL2*011 ) , HM211186 ( KIR2DL2*00602 ) , HM235772 ( KIR3DL3*056 ) , HM358895 ( KIR2DS3*006 ) , JX523641/HM358896 ( KIR2DS5*00502 ) , HM602023 ( KIR2DL5B*017 ) , HM602024 ( KIR2DS3*00106 ) , HQ026776 ( KIR2DS5*009 ) , HQ191481 ( KIR3DL3*02703 ) , HQ191482 ( KIR3DL3*049 ) , HQ609602 ( KIR2DP1var1 ) , HQ609603 ( KIR2DP1var2 ) , HQ609604 ( KIR2DP1var3 ) , HQ609605 ( KIR2DP1var4 ) , HQ609606 ( KIR2DP1var5 ) , HQ609607 ( KIR2DP1var6 ) , JX523632 ( KIR2DL4*023 ) , JX523633 ( KIR2DL4_19b ) . Seven KIR3DL1/S1 alleles from this population were reported previously [35] . | Natural killer cells are white blood cells with critical roles in human health that deliver front-line immunity against pathogens and nurture placentation in early pregnancy . Controlling these functions are cell-surface receptors called KIR that interact with HLA class I ligands expressed on most cells of the body . KIR and HLA are both products of complex families of variable genes , but present on separate chromosomes . Many HLA and KIR variants and their combinations associate with resistance to specific infections and pregnancy syndromes . Previously we identified basic components of the system necessary for individual and population survival . Here , we explore the system at its most genetically diverse by studying the Ga-Adangbe population from Ghana in West Africa . Co-evolution of KIR receptors with their HLA targets is ongoing in the Ga-Adangbe , with every one of 235 individuals studied having a unique set of KIR receptors and HLA class I ligands . In addition , one critical combination of receptor and ligand maintains alternative forms that either can or cannot interact with their ‘partner . ’ This balance resembles that induced by malfunctioning variants of hemoglobin that confer resistance to malaria , a candidate disease for driving diversity and co-evolution of KIR and HLA class I in the Ga-Adangbe . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Co-evolution of Human Leukocyte Antigen (HLA) Class I Ligands with Killer-Cell Immunoglobulin-Like Receptors (KIR) in a Genetically Diverse Population of Sub-Saharan Africans |
Genetically encoded biosensors based on fluorescence resonance energy transfer ( FRET ) have been widely applied to visualize the molecular activity in live cells with high spatiotemporal resolution . However , the rapid diffusion of biosensor proteins hinders a precise reconstruction of the actual molecular activation map . Based on fluorescence recovery after photobleaching ( FRAP ) experiments , we have developed a finite element ( FE ) method to analyze , simulate , and subtract the diffusion effect of mobile biosensors . This method has been applied to analyze the mobility of Src FRET biosensors engineered to reside at different subcompartments in live cells . The results indicate that the Src biosensor located in the cytoplasm moves 4–8 folds faster ( 0 . 93±0 . 06 µm2/sec ) than those anchored on different compartments in plasma membrane ( at lipid raft: 0 . 11±0 . 01 µm2/sec and outside: 0 . 18±0 . 02 µm2/sec ) . The mobility of biosensor at lipid rafts is slower than that outside of lipid rafts and is dominated by two-dimensional diffusion . When this diffusion effect was subtracted from the FRET ratio images , high Src activity at lipid rafts was observed at clustered regions proximal to the cell periphery , which remained relatively stationary upon epidermal growth factor ( EGF ) stimulation . This result suggests that EGF induced a Src activation at lipid rafts with well-coordinated spatiotemporal patterns . Our FE-based method also provides an integrated platform of image analysis for studying molecular mobility and reconstructing the spatiotemporal activation maps of signaling molecules in live cells .
Src is a protein tyrosine kinase which plays crucial roles in cell adhesion , migration and cancer invasion [1] . In fact , epidermal growth factor ( EGF ) and its receptor EGFR has been well documented to couple with Src kinase to regulate cancer progression [2] . Before stimulation , Src is localized at microtubule-associated endosomes around the nucleus [3]–[7] . The SH3 and SH2 domains of Src kinase are coupled together by intramolecular interaction , and the catalytic kinase domain of Src is masked by the interaction with C-terminal tail , thus preventing its action on substrate molecules [8] . Upon EGF stimulation , Src can translocate to focal adhesion sites and associate with actin filaments at cell periphery [4] , [5] , [9]–[12] , possibly through the Src N-terminal tail and SH3 domain , but not the catalytic domain [3] , [10] , [13] . Recent evidence indicates that EGF can enhance the Src localization and activation at lipid rafts to regulate cancer development [14]–[16] . However , the existence of the extremely small and dynamic lipid rafts , and the mechanism on how these lipid rafts function as docking sites to coordinate signaling molecules , remain controversial [17] , [18] . It is also not clear how EGF activates Src spatially and temporally at lipid rafts to impact on cellular functions . Genetically encoded biosensors based on fluorescence resonance energy transfer ( FRET ) are powerful tools for live cell imaging [19] , [20] . A variety of such biosensors utilizing cyan fluorescence protein ( CFP ) and yellow fluorescence protein ( YFP ) have been developed to visualize the activities of important kinases in live cells , including epithelial growth factor receptor ( EGFR ) , Abl [21] , protein kinase A [22] , protein kinase B [23] , protein kinase C [24] , and insulin receptor [25] . We have also developed a genetically-encoded FRET biosensor for monitoring Src activity in live cells [21] , [26] . The investigations based on these biosensors have provided invaluable information about the spatiotemporal activation pattern of the molecules studied [27] , [28] . However , the observed FRET signal reported by these biosensors at any given spot represents the combined effect of two main factors: ( 1 ) the local kinase activity acting on biosensors and ( 2 ) the signal of activated biosensors moving in the cell among locations . The movement of these biosensors is not dependent on the motion of the targeting enzymes or their endogenous substrate molecules . Hence , the rapid motion of the biosensors can artificially dissipate the cumulative signals engendered by the in situ enzymatic activity . Therefore , it is essential to identify and subtract the effect of biosensor motility from the apparent FRET signals to allow an accurate reconstruction of the spatiotemporal activation map of the targeting kinase . The fluorescence recovery after photobleaching ( FRAP ) analysis has been widely used to estimate the apparent diffusion coefficients and characterize the motion of fluorescent molecules in live cells [29]–[32] . In classical FRAP analysis , the fluorescence recovery curve is obtained by monitoring the average fluorescence intensity in a small region after photobleaching . Based on the recovery curve , the apparent diffusion coefficient of fluorescent molecules can be estimated by parameter fitting [29] . However , this approach has specific requirements on the cell geometry , photobleached spot , and the actual photobleaching process [29]–[31] , [33] . Most recently , FRAP analysis using numerical methods , such as the computational particle method , the finite difference method , and the Monte Carlo simulation , have been developed to address these limitations [34]–[40] . Results from FRAP analysis have revealed the characteristics of transport kinetics for many important molecules [41]–[43] . Nonetheless , there is a need to apply these methods to quantify and analyze live-cell FRET images . The finite element ( FE ) method is well known for its flexibility in resolving the complex geometry of tissue and cellular structures [44] , [45] . It has been used to estimate the apparent diffusion constant in inhomogeneous tissues [46] and for modeling protein transport in single cells [47] . In this study , we have developed a new imaging analysis method based on FE and FRAP to evaluate the motility of different Src biosensors . The results revealed that the motility of biosensors tethered to lipid rafts is governed by 2D diffusion . After the effect of biosensor diffusion on FRET signals was subtracted from the apparent FRET images , the diffusion-corrected FRET signals revealed that , at lipid rafts , high Src activities upon EGF stimulation are concentrated at relatively stationary clusters around cell periphery . Our FE-based imaging analysis method , integrated with FRAP and FRET technologies , can also serve as a general method to study the spatiotemporal kinetics of other enzymatic activity in living cells .
To assess the effect of biosensor diffusion on the apparent FRET images recorded in experiments ( Figure 1 ) , we developed a FE-based method to analyze protein diffusion in FRAP experiments . Based on Fick's second law of diffusion , the change of molecular concentration in time is proportional to the second derivative of the concentration in space , i . e . , the Laplacian of concentration . This can be expressed mathematically as following: ( 1 ) where represents the time derivative of the concentration u ( x , y , t ) at a given time and location in 2D space , Δu ( x , y , t ) denotes the Laplacian of u ( x , y , t ) and D represents the diffusion coefficient of the target molecule [33] . After Eq . ( 1 ) was discretized using the FE method , the apparent diffusion coefficient can be estimated by applying a linear regression procedure on the weighted discrete Laplacian of concentration ( WDLC ) and the weighted change of concentration in time ( WCCT ) ( Figure 2 , and see Materials and Methods , “Computational Simulation and Validation of the Diffusion Model” ) . The FE-based image analysis method was validated by computational modeling of the diffusion process ( Figure 3 ) . A designated cell geometry , an initial distribution of molecular concentration to mimic the fluorescence image after photobleaching , a diffusion coefficient of 29 µm2/sec ( the diffusion coefficient of XPA-GFP which has the same size as our cytosolic Src biosensor [39] ) were first assigned . A sequence of concentration maps ( Figure 4A–4B ) was numerically generated and saved to mimic the real procedures in FRAP experiments and used for the computation of the fluorescence recovery curve ( Figure 4C ) . Based on these simulated FRAP images , FE analysis was used to triangulate the cell geometry and discretize the diffusion equation ( Figures 3 and S1 ) . Linear regression was then used to calculate the apparent diffusion coefficient ( Figure 2 ) to be 30 . 3 µm2/sec , close to the assigned diffusion coefficient . Because the simulated diffusion process is governed by Fick's law , the WDLC should be linearly correlated to WCCT . The plot of WDLC vs . WCCT on each FE mesh-node verified a linear relationship between these two quantities ( Figure 4D ) . All these results suggest that our method is accurate for modeling diffusion process . A large portion of data points in Figure 4D clustered near the zero of WCCT , suggesting that there was no significant change of the concentration at many mesh nodes distant from the photobleached spot over one time-step . Meanwhile , the noise in Figure 4D is likely due to image processing in the simulation to mimic the procedures of data processing in FRAP and FRET experiments ( saving and loading image files ) , since the same discretization method was used for simulating concentration maps and estimating diffusion coefficient . These noises can indeed be eliminated by running the simulation without saving/loading images ( data not shown ) . A Src FRET biosensor was previously modified and tethered at lipid rafts in plasma membrane through a myristoylation and palmitoylation tag at the N-terminal ( Lyn-Src ) ( see Figures S2 and 5 ) [26] , [48] . We have further developed , analyzed , and compared two other versions of compartment-localized Src FRET biosensors as shown in Figure 5 [21] , [26] . One biosensor is targeted to membrane regions outside of lipid rafts through a geranylgeranylated tag at the C-terminal ( KRas-Src ) [48] and the other is located in the cytoplasm and the nucleus ( Cytosolic-Src ) . To assess their mobility , the biosensors in a small region of a live cell were photobleached . The post-bleaching images were monitored and then normalized by the pre-bleaching images to obtain concentration maps . Subsequently , the FE analysis and linear regression method were applied on the concentration maps to estimate the apparent diffusion coefficient ( Figure 6 ) . As shown in Figure 7A–7B and Movie S1 , the fluorescence intensity of the Lyn-Src biosensor localized at lipid rafts recovered in ∼15 min after photobleaching , with an estimated apparent diffusion coefficient of 0 . 11±0 . 01 µm2/sec . To evaluate the accuracy of the diffusion model , the mobility of this Lyn-Src biosensor was simulated and compared with experimental results . The simulation-predicted concentration map of the Lyn-Src biosensor at 1 min after photobleaching precisely matches the experimental result ( Figure 7C ) . The linear relationship between WDLC and WCCT further confirmed that the motion of the Lyn-Src biosensor is dominated by diffusion and governed by Fick's law ( Figure 7D ) . These results suggest that our diffusion model can accurately predict the motility of biosensor tethered at lipid rafts . Similar approaches were employed to analyze the mobility of the KRas-Src and the Cytosolic-Src biosensors ( Figures 8–9 ) . The fluorescence intensity of the Cytosolic-Src biosensor recovered in ∼4 min after photobleaching ( Figure 9A–9B ) . The estimated apparent diffusion coefficient of the Cytosolic-Src biosensor was 0 . 93±0 . 06 µm2/sec , which is 4–8 folds higher than that of the membrane-bound Lyn-Src ( 0 . 11±0 . 01 µm2/sec ) and KRas-Src biosensors ( 0 . 18±0 . 02 µm2/sec ) . These observations are consistent with previous reports that the diffusion rate of the molecules near the plasma membrane is 2- to 3-fold slower than that in cytoplasm [30] , [49] , [50] , possibly reflecting the different nature of diffusions in 2D ( membrane ) and 3D ( cytosolic/nucleus ) . Error analysis procedures were designed to further evaluate the accuracy of the FE-based diffusion analysis for the three versions of Src biosensors ( Figure 10 , see Materials and Methods , “Error Analysis” ) . For the Lyn-Src biosensor , the model prediction matches experimental result precisely ( Figure 10A_i ) , and the scattered linear plot of data shows high confidence with the model ( Figure 10B_i ) . The KRas-Src biosensor also had relatively uniform distribution on plasma membrane ( Figure 5B ) , with reasonable agreement between experimental and simulation results ( Figure 10A_ii–10B_ii ) . It is of note that the mobility of the KRas-Src biosensor appears slightly less well predicted than for the Lyn-Src biosensor ( Figure 10 ) . On the other hand , the results for the Cytosolic-Src biosensor demonstrated an obvious disagreement between simulation and experiments ( Figure 10A_iii–10B_iii ) , which is attributable , at least in part , to the accumulation of a large fraction of the Cytosolic-Src biosensor in the nucleus in which molecules may have significantly different mobility from that in the cytoplasm ( Figure 5B ) . To gain more insights about the molecular dynamics and kinetics in lipid rafts , we investigated and compared the kinetics of the Lyn-Src and the KRas-Src biosensors in cells with MβCD treatment , which extracts cholesterol and disrupts lipid rafts . Without MβCD treatment , the Lyn-Src biosensor was found by FRAP analysis to move at a slower rate on the plasma membrane than the KRas-Src biosensor . Since the Lyn-Src biosensor is tethered on the lipid rafts , which are subdomains of plasma membrane rich in cholesterol [48] , [50]–[53] , this finding corroborates previous observations that molecules move more slowly in the cholesterol-rich than cholesterol-poor model membranes [54] . In fact , we found that the treatment with MβCD to disrupt cholesterol-associated rafts significantly increased the apparent diffusion coefficient of the Lyn-Src biosensor ( from 0 . 11±0 . 01 to 0 . 17±0 . 01 µm2/sec ) , but not the KRas-Src biosensor ( from 0 . 18±0 . 02 to 0 . 20±0 . 01 µm2/sec ) ( Figure 11A ) . This result is also consistent with earlier findings that MβCD enhances the molecular motility of HRas-tagged green fluorescence protein ( GFP ) tethering on lipid rafts , but not KRas-tagged GFP [50] . The large coefficient of determination ( R2 = 0 . 79±0 . 033 ) ( Figure 11B ) , which represents a high correlation between the experimental results and the simulated predictions by our diffusion model ( see Method , “Error Analysis” ) , suggests that the mobility of the Lyn-Src biosensor is dominated by diffusion and hence can be accurately predicted by the diffusion model . This result is also consistent with the error analysis approach ( Figure 10 ) . The mobility of the KRas-Src biosensor ( R2 = 0 . 56±0 . 06 ) is less well predicted by simulation , suggesting that transportation mechanisms other than 2D diffusion may also contribute to the mobility of biosensors tethered outside of lipid rafts . The apparent FRET images of the Src biosensors represent the combinatory effects of spatiotemporal Src kinase activity and the re-distribution of mobile activated biosensors ( Figure 1 ) . Hence the apparent FRET signals may be different from the actual distribution of Src activity or its actions on endogenous substrate molecules . In fact , many prominent substrate molecules of Src kinase , e . g . , p130cas and paxillin , are localized at subcellular regions with limited mobility in adherent cells [55] , [56] . Recent evidence indicates that lipid rafts serve as an integrated platform for Src activation [16] , [57] and the recruitment of P130cas and paxillin [58]–[60] . However , there is a lack of knowledge on the spatiotemporal pattern of Src activation at lipid rafts or its accumulative effects on the relatively immobile substrate molecules . To reconstruct the Src activation map at lipid rafts , the contribution of biosensor diffusion was simulated and subtracted from apparent FRET signals . Error analysis has shown that the FE model of diffusion can precisely predict the movement of the Lyn-Src biosensor . Control experiments suggest that the diffusion rate of the Lyn-Src biosensor does not differ significantly with or without EGF stimulation ( data not shown ) . Hence a diffusion coefficient of the Lyn-Src biosensor calculated before EGF stimulation can be applied to simulate the diffusion process through the entire time course of FRET experiment in the same cell . The subtraction of this simulated diffusion effect revealed discrete clusters of high Src activities at lipid rafts close to the cell edge , in contrast to the FRET images without diffusion subtraction which are relatively uniform ( see Figure 11C and Movie S2 ) [26] . Immunostaining of the distribution of Src activity in fixed cells upon growth factor stimulation [5] , [55] also showed high Src activities concentrated at cell periphery , consistent with our observations . It is of note that the locations with high Src activity at lipid rafts are relatively stationary upon EGF stimulation ( Figure 11C ) , suggesting that active Src remains localized without significant motion upon arrival at lipid rafts .
The timing and localization of molecular activities are crucial for their proper functions . In this paper , we have integrated FE-based imaging analysis modeling , FRAP and FRET technologies , to reconstruct and visualize the spatiotemporal Src activity in lipid rafts upon EGF stimulation . The mobility of the Src biosensor tethered in the lipid rafts of plasma membrane was shown to be dominated by diffusion . The subtraction of this diffusion effect from FRET images has helped to reconstruct the Src activation map at lipid rafts , with high Src activity localized at stationary clusters proximal to cell edge . Given the important roles of Src and lipid rafts in mediating EGF/EGFR-regulated cancer development [2] , [14] , our results should shed new lights on how cells coordinate molecular activities in space and time to orchestrate pathophysiological responses upon external stimulation . The advantage of our live-cell imaging approach is further underscored by the controversial effect of non-ionic detergents used for isolating lipid rafts in traditional assays [17] , [61] . Although the roles of Src in regulating downstream signaling pathways are well studied , the detailed mechanism of Src activation in response to EGF is not clearly elucidated [62] . It has been shown that growth factors can induce the translocation of Src from perinuclear regions to cell periphery through RhoB and actin cytoskeleton [5] , [63] . Our results suggest that Src can be transported and activated at lipid rafts . The active Src molecules upon arrival at lipid rafts appear relatively stationary with sub-compartment localization since the activation pattern of Src biosensor showed clusters with increasing size , but little motion ( Figure 11C ) . It has been shown that EGF can form complex with its receptor EGFR , which further binds to integrins [64] . Since integrins are anchored to immobile extracellular matrix and well documented to coordinate the localization of lipid rafts and its associated signaling molecules [65] , [66] , it is possible that EGF and its ligation with EGFR induce localized Src activation at lipid rafts via integrins . In fact , evidence has shown that integrin β3 can directly bind to Src through the interaction of β3 C-terminal tail and Src SH3 domain [67] . Some evidence has shown that EGFR did not colocalize with caveolae at rest state [68] . Hence it is also possible that either EGF receptor or Src is activated outside of lipid rafts and then sequestered inside lipid rafts . Further studies are warranted to elucidate the underlying mechanism for this localized and stationary Src activity at lipid rafts in response to EGF stimulation . The motility of the Lyn-Src biosensor is dominated by diffusion , as evidenced by the close match between experimental and simulated results , and by the strong linear correlation between WDLC and WCCT ( Figures 7C and 10B_i ) . The mobility of the KRas-Src biosensor , however , displays some nonlinear features between WDLC and WCCT ( Figure 10B_ii ) , suggesting that it is not completely governed by 2D diffusion . Intracellular molecule mobility is influenced by molecular interaction , diffusion , and active transportation [31] , [33] . Hence , molecular interaction or active transportation may contribute to the motion of KRas-Src biosensor besides diffusion . The mobility difference between KRas- and Lyn-Src biosensors may be attributable to the tight membrane-binding of the Lyn tag through deep insertion of side chains into the bilayer interior and the fluctuating membrane-binding of the KRas tag through electrostatic switches [69] . Because the membrane-tethered biosensors extend appreciably into the cytoplasm , it is also possible that some of the restricted motion at the proximity of the plasma membrane may be due to the interaction of the biosensor with the cortical actin cytoskeletal network [70] . These interactions may have particularly contributed to the motion of KRas-Src biosensor , which is not dominated by random diffusion . Our estimated diffusion coefficient of the Cytosolic-Src biosensor is several-fold higher than those of the membrane-targeted versions . One of the possible reasons for the difference between the diffusion coefficients of the Cytosolic-Src and membrane-targeted biosensors may be the difference in the physicochemical properties of local environment , e . g . the diffusion of the Cytosolic-Src biosensors is 3D in nature whereas that of the membrane-targeted biosensors is 2D . While our diffusion model can be used to estimate the apparent diffusion coefficient and simulate diffusion process in principle , it cannot be directly applied to study the Cytosolic-Src biosensor . The low coefficient of determination ( R2 = 0 . 33±0 . 1 , n = 5 ) suggests that the mobility of a large portion of the Cytosolic-Src biosensors cannot be described by diffusion . This is possibly because the Cytosolic-Src biosensors reside in different sub-compartments of the cell , e . g . , the nucleus vs . the cytoplasm , as shown in Figure 5B and evidenced by the results from our fluorescence loss in photobleaching ( FLIP ) experiments ( data not shown ) . The movement of the Cytosolic-Src biosensor will likely be better described by a 3D and multi-compartment diffusion model . The approach of evaluating and subtracting diffusion based on FRAP and FRET video images can also be implemented by employing other numerical methods including finite difference method , computational particle method , and Monte Carlo simulation . We decided to choose the FE-based method because it has been well-established for modeling the diffusion processes with complex geometry in 2D and 3D [44] , [71] . Since the FE methods have great flexibility in resolving the complex geometry of tissue and cellular structures [44]–[47] , no specific requirement on the cell geometry , the bleaching light beam , or the photobleaching process is needed in our new FRAP analysis method . Further , efficient solvers [72] and parallel implementation on distributed computers have been extensively developed for FE methods [73] . Thus , with the integration of 3D imaging techniques , e . g . confocal microscopy , our system can be conveniently extended to 3D analysis and parallel computing environment . In summary , our FE-based method can successfully separate the effect of biosensor diffusion from the apparent FRET signals to reconstruct the diffusion-corrected spatiotemporal activation map of membrane-tethered Src kinase . The results suggest that the EGF-induced Src activation at lipid rafts has localized and stationary patterns clustered at cell periphery . This methodology can be conveniently utilized to reconstruct other molecular activation maps from those reported by indirect and diffusion-driven biosensors .
HeLa cells ( ATCC , Manassas , Virginia ) were cultured in a humidified 95% air , 5% CO2 incubator at 37°C . The culture medium was Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 10% fetal bovine serum , 2 mM L-glutamine , 1 unit/ml penicillin , 100 µg/ml streptomycin , and 1 mM sodium pyruvate . The cell culture reagents were obtained from Invitrogen ( San Diego , CA ) . The gene for the Cytosolic-Src biosensor was constructed as described previously [26] . In brief , this Cytosolic-Src FRET biosensor consists of a peptide derived from Src substrate molecule p130cas and a phosphotyrosine-binding domain ( SH2 domain derived from c-Src ) , bracketed by monomeric ECFP and Citrine ( an improved version of EYFP ) at the N- and C-termini . The substrate peptide phosphorylated by a Src kinase can interact with the intramolecular SH2 domain , which results in a change of distance or relative orientation between ECFP and Citrine , as shown in Figure S2 . The subsequent changes of FRET between ECFP and Citrine can be represented by the ECFP/Citrine emission ratio to monitor the Src activities . The membrane-targeted ECFP was constructed by PCR amplification of the monomeric ECFP with a sense primer containing the codes for N-terminal amino acids from Lyn kinase to produce a Lyn-Src biosensor [48] . For the KRas-Src biosensor , the monomeric YFP was amplified by PCR with an anti-sense primer containing the codes for C-terminal amino acids from KRas ( KKKKKSKTKCVIM ) . For simplicity , we refer to the monomeric ECFP and Citrine by CFP and YFP respectively in text and figures . The various plasmids were transfected into HeLa cells at 80% confluence using the lipofectamine method as described by the vendor ( Invitrogen , San Diego , CA ) . For FRAP experiments , the YFP images were collected using MetaFluor 6 . 2 software ( Molecular Devices , Sunnyvale , California ) on epi-fluorescence microscopy ( Zeiss , Oberkochen , Germany ) with emission at 535DF25 and excitation at 495DF20 using 1% of the light source power . During imaging , the cells were kept in CO2-independent medium without serum ( Invitrogen ) at 25°C; and the objective focus was aimed near the basal side of the cell . The cells were monitored before photobleaching to confirm there was no detectable photobleaching during imaging . Photobleaching was conducted by exciting YFP at 495DF20 in a region of interest with full power of the light source for 15 sec , after which the recovery process was imaged at 1-sec and 10-sec intervals for the cytosolic and membrane-targeted Src biosensors , respectively . For FRET experiments , the HeLa cells expressing the desired Src biosensors were starved with 0 . 5% FBS for 36–48 hr before being subjected to EGF ( 50 ng/ml ) stimulation . The images were collected with a 420DF20 excitation filter , a 450DRLP dichroic mirror , and two emission filters controlled by a filter changer ( 480DF30 for CFP and 535DF25 for FRET ) . The pixel-wise images of CFP/YFP emission ratio were computed to assess the FRET signals , which represent the concentration of phosphorylated Src biosensor and hence Src activity in space and time . The Src biosensors were assumed to diffuse freely inside the cytoplasm or in the membrane . According to Fick's Law , the diffusion equation is given by Eq . ( 1 ) ( Results , Computer Simulation and Validation ) . Enclosed in the cell boundary , a triangular mesh was generated for the finite element discretization ( Figure S1 ) . A two-dimensional model was used because the thickness of a spread cell is relatively small compared to its length and width , and the photobleached region is sufficiently big ( ∼2 µm ) such that the 3D profile of the light beam is negligible . Using the FE method for discretizing the Laplacian operator and the Crank-Nicholson Scheme for approximating time derivative [74] , Eq . ( 1 ) can be approximated by a discrete linear system ( for details see Text S1 , “The Formulation of the Finite Element Method” ) ( 2 ) where M represents the mass matrix , K the stiffness matrix , dt the discrete interval between each time step , un and un+1 the concentration of fluorescent molecules at the nth and ( n+1 ) th time step , respectively . Here the matrices were assembled using the finite element method to incorporate the geometry of the cell . Zero flux was assumed at cell boundary . For a given initial fluorescent concentration un and an assigned diffusion coefficient , the fluorescent concentration at the next time step , un+1 , can be computed based on a simple transformation of Eq . ( 2 ) : With the interval between each time step dt set to be 0 . 0313 sec , numerical convergence of the FE method was confirmed by comparing the estimated diffusion coefficients and simulated diffusion results with those on a higher resolution mesh and a smaller time step . According to Eq . ( 2 ) , there is a linear relationship between the weighted change of concentration in time ( WCCT ) , M ( un+1−un ) , and the weighted discrete Laplacian of concentration ( WDLC ) , −0 . 5dt·K· ( un+un+1 ) . Therefore , based on the fluorescence concentration at two consecutive time steps , the diffusion coefficient can be estimated by linear fitting between these two quantities using the least square method ( Figure 2 ) . The calculated diffusion coefficient is then compared with the originally assigned diffusion coefficient to assess the accuracy of our method . The whole process of computational simulation to assess and verify the accuracy of our FE and diffusion model is illustrated in Figure 3 . All the computer-simulated concentration images were processed using a median filter with a window sized at 10×10 pixels ( Figure 3 ) . Similarly , the apparent diffusion coefficients of the Src biosensors in FRAP experiments were obtained by computing the least-square linear fitting between the WDLC and the WCCT of the concentration images . The diffusion coefficients were then used to simulate and predict the fluorescence recovery maps for comparison with the experimental concentration images ( Figure 6 ) . Different from the computer simulation which covers the entire cell , most of the FRAP images were captured with the 100× objective , so only part of the cell was captured in the image in some occasions . Therefore there may be fluxes across the image boundary , which is not part of the cell boundary . In these cases , instead of zero flux boundary conditions ( BCs ) , the BCs were computed with the apparent diffusion coefficient during linear fitting , by estimating both parameter D and r0 in Eq . S7 [46] . Using this linear regression procedure , one estimated apparent diffusion coefficient can be computed with every pair of concentration maps ( FRET ratio ) un and un+1 . The apparent diffusion coefficient was obtained by averaging the estimated diffusion coefficients of several time intervals . This strategy bears some similarity with the classic FRAP analysis where one apparent diffusion coefficient is obtained by fitting the complete recovery curve . In addition , it is required that we convert the experimental fluorescent intensity images to concentration maps , and reduce noise by smoothing the images at several stages , as described in details in Text S1 , “Pre-processing of FRAP Experimental Images” . Two kinds of error analysis were used to evaluate the accuracy of our diffusion model at each time step . First , the absolute value of the error , abs ( un−est_un ) , was used to show the difference between the simulated concentration map with experimental images . Here est_un and un denote the simulated and experimental concentration maps at the nth time step , respectively . The accuracy of our diffusion model was further evaluated by computing the coefficient of determination , which measures the percentile of total variation in the data that can be explained by the mathematical model [75] . In our diffusion model , the coefficient of determination , R2 , is equivalent to the square of the linear correlation coefficient between WCCT {xi} and WDLC {yi} . The linear correlation coefficient between these two data sets {xi} and {yi} is defined aswhere x̅ and y̅ are the mean values of {xi} and {yi} respectively . To smooth the data and reduce the computational noise , the data set of WDLC {yi} and WCCT {xi} was divided into ten equal intervals along the x-axis and averaged at each interval before computing the coefficient of determination . For statistical analysis of the estimated apparent diffusion coefficients and the coefficients of determination , we used the Bonferroni multiple comparison test of means at 95% confidence interval , which is provided by the multcompare function in the MATLAB statistics toolbox ( The MathWorks , Natick , MA ) . The estimated apparent diffusion coefficients were selected based on the criteria described in Text S1 , “Including Estimated Coefficients in Statistical Analysis” . The FRET ratio images ( CFP intensity/ YFP intensity ) were used to quantify the Src activity , or the concentration of phosphorylated Src biosensor . As shown in Figure 1 , the FRET signals originated from the diffusion of the biosensor at any given time ( Figure 1C ) was simulated by using the FRET image of the previous time step ( Figure 1A ) and the apparent diffusion coefficient estimated by previous FRAP experiments of the biosensor . This simulated FRET image ( Figure 1C ) was then subtracted from the recorded apparent FRET image at the given time ( Figure 1B ) to obtain the transient FRET changes , which represents the actions of Src kinase activity on the biosensor between these two time steps ( Figure 1D ) . These transient FRET changes were then iteratively added to the initial FRET image obtained right after EGF application to reconstruct the diffusion-corrected FRET images , which represents the cumulative Src kinase activity on its relatively immobile substrate molecules , such as those in the focal adhesion complex . | Fluorescence biosensors have been widely used to report the spatial and temporal activity of target molecules in live cells . However , biosensors can move independently of the target molecule and carry its signal to other subcellular locations . Therefore , the observed images appear to be the combination of the target molecular activity and the artifacts introduced by the movement of the biosensors ( mainly due to diffusion ) . The intriguing question is how to estimate and exclude the movement effect of biosensors from the observed fluorescent images and to reconstruct the real activity map of the target molecules . The Src molecule plays important roles in cell adhesion , migration , and cancer invasion . In this paper , we developed a novel computational method to analyze and simulate the movement of the Src biosensor , which was then subtracted from the original fluorescent images . With this computational method , we observed discrete clusters of high Src activity at relatively stationary locations on the plasma membrane . Therefore , our results highlight the coordination of molecular activities in space and time . In addition to Src , our computational method can be used to reconstruct the activity map of other signaling molecules . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/cell",
"signaling",
"biochemistry/bioinformatics",
"biotechnology/bioengineering",
"biophysics/cell",
"signaling",
"and",
"trafficking",
"structures",
"computational",
"biology/systems",
"biology"
] | 2008 | The Spatiotemporal Pattern of Src Activation at Lipid Rafts Revealed by Diffusion-Corrected FRET Imaging |
The CST ( Cdc13/CTC1-STN1-TEN1 ) complex was proposed to have evolved kingdom specific roles in telomere capping and replication . To shed light on its evolutionary conserved function , we examined the effect of STN1 dysfunction on telomere structure in plants . STN1 inactivation in Arabidopsis leads to a progressive loss of telomeric DNA and the onset of telomeric defects depends on the initial telomere size . While EXO1 aggravates defects associated with STN1 dysfunction , it does not contribute to the formation of long G-overhangs . Instead , these G-overhangs arise , at least partially , from telomerase-mediated telomere extension indicating a deficiency in C-strand fill-in synthesis . Analysis of hypomorphic DNA polymerase α mutants revealed that the impaired function of a general replication factor mimics the telomeric defects associated with CST dysfunction . Furthermore , we show that STN1-deficiency hinders re-replication of heterochromatic regions to a similar extent as polymerase α mutations . This comparative analysis of stn1 and pol α mutants suggests that STN1 plays a genome-wide role in DNA replication and that chromosome-end deprotection in stn1 mutants may represent a manifestation of aberrant replication through telomeres .
Telomeres form a specialized type of chromatin that protects the native chromosome ends from being perceived as DNA double strand breaks ( DSB ) and from eliciting a DNA damage response ( DDR ) . Telomeres also play an important role in genome duplication; their de novo synthesis by telomerase counteracts the end-replication problem caused by the inability of the conventional DNA replication machinery to fully duplicate the ends of linear chromosomes . In the vast majority of eukaryotes , the telomeric DNA consists of TG/CA-rich repeats and terminates with a 3′ TG-rich protrusion , the so called G-overhang . Telomeric DNA serves as a binding platform for a set of evolutionary conserved proteins whose function is to support telomere protection and replication . However , studies across multiple model organisms have revealed , in some cases , remarkable differences concerning the utilization and necessity of these conserved factors in telomere-related processes [1] , [2] . The protection of mammalian telomeres largely depends on the shelterin complex [3] . One of the functions of shelterin is to promote the formation and stabilization of t-loops , which are lariat structures produced by intrastrand invasion of the G-overhang into the duplex telomeric region [4] , [5] , [6] . T-loops are proposed to mediate chromosome capping by physically sequestering the G-overhang thereby limiting the access of DNA repair and processing factors to the chromosome ends . Although t-loops have been found in a variety of organisms , they do not seem to form in budding yeast . Instead , G-overhangs in Saccharomyces cerevisiae are protected by the CST complex , an RPA-like particle consisting of the subunits Cdc13 , Stn1 and Ten1 [7] . CST specifically binds to the telomeric G-overhang via OB-folds , and deletion of any subunit is lethal due to DNA damage checkpoint activation and massive nucleolytic resection , which is mainly mediated by exonuclease 1 ( Exo1 ) [8] , [9] , [10] , [11] . In addition , CST is crucial for telomere replication by facilitating the recruitment of telomerase and DNA polymerase α to the chromosome termini , and thus coordinating G-overhang extension by telomerase with the fill-in synthesis of the complementary C-strand [12] , [13] , [14] . A complex analogous to yeast CST was recently found in vertebrates [15] , [16] . It consists of the proteins STN1 and TEN1 that are orthologous to the yeast counterparts and of the CTC1 subunit , which shows little sequence homology to Cdc13 , but seems to mediate similar functions . As in yeast , protein interaction and structural studies demonstrated a similarity between human CST and RPA [16] , [17] , [18] . Furthermore , human CST binds ssDNA with a preference for the telomeric G-strand sequence [16] , [18] . However , unlike in yeast , the consequences of CST inactivation in mammals are not so detrimental . CTC1-null mice are viable , but have a greatly reduced lifespan due to bone marrow failure and G2/M checkpoint arrest in haematopoietic stem cells [19] . Down-regulation of CST components in human cell lines by siRNA led to a range of relatively mild telomere-related defects whose extent was dependent on the cell line and experimental conditions [15] , [16] , [18] , [20] , [21] , [22] . Importantly , these studies revealed that CST facilitates replication through duplex telomeric region and functions in the C-strand fill-in reaction and G-overhang maturation [20] , [21] , [23] , [24] . This led to the suggestion that mammalian CST is primarily involved in telomere replication , but not directly in protection . Thus , data from yeast and mammals imply that CST has evolved kingdom specific roles at telomeres [25] . The CST complex is also present in plants and all three subunits were identified and functionally characterized in Arabidopsis thaliana [15] , [26] , [27] . Null mutations in either CTC1 , STN1 or TEN1 have an immediate impact on the telomere structure , resulting in shorter , heterogeneous telomeres , elongated G-overhangs , aberrant telomere recombination and chromosome end-to-end fusions . While mutant plants are viable , they exhibit retarded growth and reduced fertility . These defects are apparently caused by increased cell death of meristematic stem cells triggered by an ATR-mediated DNA damage response [28] , [29] . Telomere dysfunction observed in Arabidopsis CST-deficient mutants led to the suggestion that the plant CST complex functions in both chromosome end protection and telomere replication , and may therefore represent an evolutionary bridge between budding yeast and mammals [15] , [30] . This interpretation would further imply that telomere protection and replication represent ancient functions of CST and that the telomere-protective role was lost in animals . Therefore , understanding the mechanism by which CST contributes to telomere stability in plants is important to fully decipher the role of this complex in eukaryotic genome maintenance . In this study we asked the question whether and how CST contributes to the protection of Arabidopsis telomeres by elucidating the processes that cause telomere dysfunction and genome instability in stn1 mutants . The data presented here and published in our previous paper [31] suggest that the onset of telomere deprotection phenotypes in STN1-deficient plants is gradual , and correlates with the progressive loss of telomeric sequence that is only partially counteracted by telomerase . We show that the bulk of STN1-depleted telomeres are not exposed to an excessive resection by EXO1 , although the nuclease promotes genome instability in stn1 mutants , possibly by processing critically short telomeres generated through defects during telomere replication . Interestingly , malfunction of a general replication factor ( DNA polymerase α ) produces telomeric defects similar to STN1 dysfunction . Furthermore , we present evidence demonstrating a role of STN1 in replication of non-telomeric loci . This suggests that CST plays a broader role in DNA replication and that the seemingly specific function ( s ) of this complex in telomere protection may reflect the sensitivity of this genomic region to replication stress .
In our previous work we noticed that telomere dysfunction in Arabidopsis stn1 mutants is less pronounced in plants with longer telomeres [31] . These data were obtained by analyzing stn1 mutants derived from a cross between ecotypes with short ( Col-0 , 2–4 kb ) and long ( Ws , 4–8 kb ) telomeres . The resulting first generation ( G1 ) stn1 ( W/C ) mutants displayed only mild growth defects and a low frequency of anaphase bridges . However , the severity of growth defects and genome instability increased in the subsequent generations [31] ( Figure S1 ) . This coincided with progressive telomere shortening . Although the stn1 ( C ) mutants in a pure Col-0 background exhibit profound developmental defects already in G1 , as in the case of stn1 ( W/C ) plants , these defects worsened in G2 and G3 generations ( Figures 1A and S1 ) . To quantitatively describe the occurrence of these defects , we divided plants into five phenotypic categories from wt-looking plants ( wt ) to stunted plants unable to produce seeds ( terminal ) . The intermediate phenotypes were characterized by altered flower phylotaxy ( I ) , stem fasciation and more pronounced defects in phylotaxy ( II ) , and massive stem fascination , aberrant leaf development and reduced fertility ( III ) ( Figure 1A ) . Identical phenotypes were described in Arabidopsis tert mutants and are hallmarks of telomere dysfunction [32] . While the majority of G1 stn1 ( C ) mutants exhibited phenotypes I and II , the bulk of G2 and G3 stn1 ( C ) plants were terminal ( Figure S1 ) . The earlier onset of developmental phenotypes in stn1 ( C ) plants inversely correlated with telomere length as G1 stn1 ( C ) and stn1 ( W/C ) telomeres were on average 2 , 2 and 2 , 8 kb , respectively . We also detected a further decline in the amount of telomeric DNA by dot blot hybridization in G2 and G3 stn1 ( C ) ( Figure S1D ) . These data indicate that the growth defects in Arabidopsis stn1 mutants are caused by loss of telomeric DNA and that STN1 disruption does not lead to immediate telomere dysfunction . One hallmark of telomere deprotection is the exposure of the chromosome termini to nuclease degradation . CST-deficient Arabidopsis exhibit very heterogeneous telomeres containing long G-overhangs , which may result from resection of unprotected telomeres by nuclease ( s ) [15] , [26] , [27] . S . cerevisiae telomeres depleted of Cdc13 undergo massive resection by Exo1 [11] and EXO1 orthologues degrade the telomeric C-strand and contribute to the telomere length heterogeneity in Arabidopsis ku80 mutants [31] . To assess the effect of EXO1 on STN1-null telomeres , we analyzed exo1a exo1b stn1 mutants where both Arabidopsis EXO1 paralogues have been disrupted ( hereafter referred to as exo1 stn1 mutants; Figure 1B ) . Inactivation of EXO1 largely rescued the gross developmental defects in G1 exo1 stn1 and most plants in the mutant population were fully fertile . In contrast , G1 stn1 plants derived from the same cross exhibited severe growth defects ( Figure 1 ) . This trend was also apparent in the second generation , where most G2 stn1 mutants reached a terminal phenotype , while ∼1/3 of G2 exo1 stn1 plants were still fertile and developed milder growth defects . In accordance with this result , we also observed a substantial decrease in anaphase bridges after EXO1 depletion ( Figure 1D ) . It has been previously reported that CST dysfunction elicits a chronic DNA damage response that is characterized by transcriptional upregulation of a number of DNA repair genes including PARP1 and BRCA1 [28] . Expression of these genes is reduced in exo1 stn1 plants , which further demonstrates that EXO1 exacerbates genome instability and DNA damage response after STN1 loss ( Figure 1E ) . We next wanted to know whether STN1-depleted telomeres are subject to EXO1-resection . Surprisingly , terminal restriction fragment ( TRF ) analysis and quantification of telomeric DNA by dot-blot hybridization did not reveal any discernable difference between stn1 and exo1 stn1 ( Figure 2A , B ) . These data indicate that telomere shortening and length heterogeneity are not caused by EXO1 activity . To directly assay for telomere resection , we compared the relative G-overhang size in stn1 and exo1 stn1 plants by the in gel hybridization technique ( Figure 2C , D ) . We observed a ∼6 fold increase in G-overhang signal in both mutants , showing that the excessive single-strandedness of telomeres in STN1-deficient plants is not due to EXO1 . This data is supported by the observation that inactivation of EXO1 did not reduce t-circle excision in stn1 mutants ( Figure S2 ) , a phenomenon caused by increased telomeric resection and recombination [31] . We have previously reported that only ∼50% of telomeres in Arabidopsis contain a long G-overhang , while the remaining telomeres are blunt- or nearly blunt-ended [31] . Furthermore , we showed that the blunt-ended telomeres preferentially terminate with the TTTAGGG-3′ permutation . Although the telomeric blunt-ends remain intact in stn1 mutants , the preference of the terminal sequence shifts to GTTTAGG-3′ . We proposed that the blunt-ended telomeres represent chromosome termini replicated by the leading strand machinery , whereas the G-overhang containing telomeres are the result of the lagging strand replication . Due to semiconservative DNA replication , the telomeric strand synthesized by the lagging strand-mechanism serves in the next round of replication as a template for the synthesis of a blunt-ended telomere [31] . Hence , the sequence preference of the telomeric blunt-ends likely reflects the processes involved in maturation of the lagging strand ( G-overhang containing ) telomeres . To examine whether the sequence alteration in STN1-depleted plants is caused by EXO1 , we quantified the frequency of telomeric permutations forming the chromosome termini in stn1 and exo1 stn1 mutants by adaptor ligation-mediated PCR coupled with Illumina sequencing [31] . We confirmed our previous data showing that STN1 deficiency alters the frequency of terminal permutations ( Figure 2E ) . Concomitant inactivation of EXO1 shifted the preference from GTTTAGG-3′ back to TTTAGGG-3′ and led to a permutation frequency profile similar to the one observed in wild-type plants ( Figure 2E ) . This indicates that STN1-depleted telomeres , or at least a fraction of them , are subject to a limited degradation by EXO1 . This processing does not grossly contribute to the single-strandedness of telomeres , but is detectable as a sequence alteration of the chromosome termini in stn1 plants . The extended G-overhangs in stn1 plants may result from an inefficient fill-in synthesis of the telomeric C-strand after G-overhang elongation by telomerase . To test this hypothesis , we generated telomerase-null stn1 and exo1 stn1 mutants that carry a disruption of the telomerase reverse transcriptase ( TERT ) gene ( Figure S3 ) . Loss of telomerase exacerbated the growth defects seen in stn1 mutants and the majority of the stn1 tert plants exhibited a terminal phenotype ( compare charts in Figure 3A and Figure 1C ) . Inactivation of EXO1 significantly alleviated this growth retardation and exo1 stn1 tert plants developed much milder defects ( Figures 3A and S3 ) . This observation corroborates the analysis of exo1 stn1 mutants and demonstrates that EXO1 activity is detrimental in the absence of STN1 . Comparison of the telomere length among siblings outsegrgating from STN1+/− TERT+/− and exo1 STN1+/− TERT+/− parents revealed that telomeres in stn1 tert and exo1 stn1 tert are much shorter than in the corresponding telomerase proficient controls ( Figure 3B ) . These data argue that telomerase partially compensates for the loss of telomeric DNA in stn1 mutants and mitigates the developmental defects ensuing from telomere dysfunction and genome instabilities . We next tested whether telomerase activity contributes to the elongated G-overhangs in stn1 mutants by performing an in gel hybridization assay . Quantification of the autoradiograms showed that telomerase-null plants have reduced G-overhangs in both stn1 and exo1 stn1 plants ( Figures 3C , D and S3 ) . This suggests that the G-strand extension by telomerase is partially uncoupled from synthesis of the complementary C-strand , implying that STN1 facilitates the C-strand fill-in synthesis . Inactivation of telomerase revealed that ∼1 kb of telomeric DNA is lost within one generation in stn1 tert mutants ( compare tert and stn1 tert in Figure 3B ) , which is substantially more than the amount of telomeric DNA lost due to the end-replication problem in tert mutants ( 0 . 25 kb per generation ) [33] . Interestingly , inactivation of EXO1 has no effect on the rate of shortening , indicating that mechanisms other than nuclease resection may contribute to this process . Recent studies in human cell lines suggested that CST associates with DNA polymerase α and facilitates the replication through human telomeres [20] , [21] , [23] , [34] , [35] . Hence , the telomere shortening in Arabidopsis stn1 mutants may be due to inefficient replication through the telomeres , resulting in terminal deletions of telomeric DNA . If STN1 functions with DNA polymerase α in telomere replication , impairment of DNA polymerase α may result in telomere-related phenotypes similar to the ones described in Arabidopsis cst mutants . To test this prediction , we analyzed plants that carry hypomorphic mutations in the catalytic subunit of DNA polymerase α . These alleles , named pol α and icu2-1 were recovered in genetic screens for deregulated gene silencing and contain single amino acid substitutions in the evolutionarily conserved C-terminal domain ( Figure S4A ) [36] , [37] . While both mutants are fertile , the stronger pol α allele causes severe growth retardation and developmental defects that are reminiscent of stn1 mutants of categories II and III ( Figure S4B ) . Plants carrying the icu2-1 mutation develop only very mild abnormalities and are similar to stn1 plants displaying the category I phenotype . Cytogenetic analysis revealed anaphase bridges in pol α as well as icu2-1 mutants , indicating that both mutants suffer from ongoing genome instability ( Figure 4A ) . The frequency of anaphase bridges was higher in pol α mutants , which correlates with the more severe developmental abnormalities . Since polymerase α deficiency is expected to cause genome-wide chromosome instabilities , we next determined whether some of the anaphase bridges detected in pol α and icu2-1 mutants arise from chromosome-end-to-end fusions by using a fusion PCR strategy that utilizes primers specific to subtelomeric regions [38] . We readily detected strong signals derived from telomeric fusions in stn1 samples , and weaker , but reproducible signals in pol α plants . Cloning and sequencing the PCR products verified that these represent chromosome end-to-end fusions and contain telomeric DNA . The icu2-1 samples also yielded specific products , albeit not in all analyzed samples ( Figures 4B , C ) . This correlates with the lower frequency of anaphase bridges and milder phenotype of the icu2-1 plants . Another hallmark of telomere dysfunction exhibited by Arabidopsis stn1 mutants is aberrant telomeric recombination , manifested through the excision of extrachromosomal t-circles . To examine whether this is also the case in DNA polymerase α mutants , we analyzed t-circle abundance with the t-circle amplification assay [39] . We detected elevated levels of t-circles in both pol α and icu2-1 plants when compared to their respective wild-type controls ( Figure 4D ) . Together , the presence of chromosome end-to-end fusions and increased t-circle excision demonstrate that an impaired function of DNA polymerase α results in telomere-dysfunction similar to inactivation of STN1 . To assess the impact of DNA polymerase α mutations on telomere maintenance , we performed TRF analysis . The telomeres in pol α mutants were overall shorter and more heterogeneous than telomeres in the corresponding wild-type plants ( C24 ) ( Figure 5A ) . A very similar TRF profile is characteristic for stn1 mutants . A slightly distinct telomere length profile was observed in icu2-1 mutants . While the icu2-1 telomeres were substantially shorter ( ∼3 kb ) than in wild-type plants ( En-2 , ∼6 kb ) , the heterogeneity was not as pronounced as in pol αor stn1 mutants . Nevertheless , both icu2-1 and pol αalleles have a strong impact on telomere maintenance , although the effect of icu2-1 seems to be less detrimental . Because DNA polymerase α is required for the fill-in synthesis of the telomeric C-strand , we carried out in gel hybridizations to determine the relative size of the telomeric G-overhangs . Interestingly , neither icu2-1 , nor pol α mutants yielded a discernible increase in the G-overhang signal ( Figure 5B , C and S5 ) . This suggests that both mutants are largely proficient in C-strand synthesis and that the observed telomere dysfunction is not primarily related to G-overhang maturation . This result was unexpected considering the strong parallels between pol αand stn1 phenotypes . Therefore , we next examined whether DNA polymerase α deficiency affects the structure of blunt-ended telomeres . As anticipated , we readily detected blunt-ended telomeres in pol αand icu2-1 mutants by the hairpin ligation assay , in which the complementary strands of the blunt-ended telomere are joined by a hairpin and then separated by alkaline gel electrophoresis ( Figure 5D ) }[31] . Sequence analysis of the telomeric blunt-ends by Illumina sequencing revealed that the TTTAGGG-3′ permutation is even more prevalent in C24 than in Col-0 wild-type plants ( 58% vs . 42% , respectively ) . Importantly , the pol α allele , which is in the C24 genetic background , showed a ∼14% reduction of TTTAGGG-3′ in favor of the GTTTAGG-3′ and GGTTTAG-3′ permutations ( Figure 5E; P<0 . 0001; two-tailed Student's t-test ) . This is reminiscent of the situation in stn1 mutants where the abundance of the TTTAGGG-3′ termini was reduced by ∼16% ( Figure 2E ) . The remarkable similarity in telomere-related phenotypes between stn1 and pol α mutants supports the notion that STN1 and DNA polymerase αmay act in the same telomere maintenance processes . Interestingly , stn1 pol α double mutants exhibit more severe developmental defect and genome instability than each mutation individually , suggesting partially complementary functions of STN1 and DNA polymerase α ( Figure S6 ) . Analysis of Arabidopsis DNA polymerase α mutants showed that an impaired function of a general replication factor may result in very specific telomere-related phenotypes . This prompted us to ask whether STN1 function is limited only to telomeres or whether it plays a broader role in DNA replication . To address this question , we took advantage of Arabidopsis mutants lacking the histone methyltransferases ATXR5 and ATXR6 that are responsible for the histone H3K27 monomethylation [40] . The atxr5 atxr6 double mutants exhibit a higher content of nuclear DNA due to re-replication of heterochromatic regions , which is particularly apparent in cells undergoing endoreplication [41] . We hypothesized that if STN1 plays a role in genome-wide replication , its absence could reduce the re-replication . We generated stn1 atxr5 atxr6 plants and measured the impact of STN1 deficiency on the nuclear DNA content by flow cytometry . In Arabidopsis , leaf development is accompanied by massive endoreplication giving rise to polyploid nuclei with DNA content ranging from 2C to 32C ( Figure 6A ) . A fraction of the endoreplicated nuclei in atxr5 atxr6 plants have a higher DNA content than nuclei of wild-type leaves , which is detectable on the flow cytometry profiles as broader peaks with a shoulder towards higher DNA content ( Figure 6A ) . Although the shoulder was also apparent in the stn1 atxr5 atxr6 mutants , it was less pronounced than in atxr5 atxr6 plants . STN1 deficiency causes telomere dysfunction resulting in permanent genotoxic stress . Therefore , we examined whether the reduced re-replication in stn1 atxr5 atxr6 mutants is specifically associated with STN1 dysfunction or whether is it a consequence of DNA damage checkpoint activation . To induce chronic genotoxic stress , we grew Arabidopsis seedlings for 14 days on agar plates supplemented with a moderate dose of bleomycin , a radiomimetic drug that induces DNA breaks . Elevated levels of PARP1 confirmed that this treatment induces constitutive genotoxic stress in wild type seedlings ( Figure 6B ) . Interestingly , PARP1 expression was highly elevated in non-treated atxr5 atxr6 seedlings and the expression was further increased by bleomycin . These data indicate that loss of atxr5 and atxr6 already results in a permanent DNA damage stress . To quantitatively measure DNA re-replication , we tested how the shape of 8C peaks detected by flow cytometry deviates from the Gaussian distribution . While 8C peaks in wild type seedlings perfectly matched a Gaussian distribution , 8C peaks in atxr5 atxr6 plants strongly deviate from the normal distribution due to the shoulder containing nuclei with higher DNA content ( Figure 6C ) . Quantitative analysis of 8C peaks in bleomycin treated atxr5 atxr6 seedlings did not show any reduction of re-replication , indicating that further genotoxic stress does not limit re-replication . In contrast , 8C nuclei in stn1 atxr5 atxr6 seedlings exhibited an almost normal distribution as the average Χ2 for peak normality was significantly reduced in comparison to atxr5 atxr6 mutants , albeit it was still higher than in wild-type and stn1 plants ( Figure 6C ) . These data argue that reduced re-replication by STN1 inactivation is not a result of a general DNA checkpoint activation . In addition , we found that pol αmutation also impedes re-replication in atxr5 atxr6 plants , further corroborating the multiple phenotypic parallels between stn1 and pol α . To determine whether STN1 is required for re-replication of particular genomic regions , we used Illumina sequencing to analyze genomic DNA from sorted 8C nuclei and quantified the density of reads mapped along the individual Arabidopsis chromosomes . We confirmed the previously published data on enrichment of pericentric DNA in the nuclei from atxr5 atxr6 plants over wild-type ( Figure 6B ) . The re-replication of heterochromatic DNA occurred to a much lesser extent in the stn1 atxr5 atxr6 mutants . More precisely , we observed an overall decline in the enrichment of reads across the re-replicated regions and did not detect any obvious effect on specific chromosomal loci .
CST was identified in S . cerevisiae where it protects the 5′ chromosome termini from resection . Consequently , inactivation of any CST subunit is lethal due to the excessive degradation of telomeric and subtelomeric regions along with the activation of a strong DDR [11] , [42] , [43] . These observations led to the conclusion that CST is an essential component of the telomere protective cap in S . cerevisiae . This seems to be the case also in fission yeast where Stn1 and Ten1 are crucial for cell survival and their deletion leads to the loss of telomeric DNA and chromosome end-to-end fusions [44] . The recent identification of the CST complex in higher eukaryotes raised the question of how conserved the telomere-capping function of CST is . While human CST knock-down cell lines produced a range of telomere-associated phenotypes , this was not accompanied by widespread telomere deprotection as judged from the frequency of telomere dysfunction-induced foci and the mild effects on cell viability and proliferation [18] , [22] , [24] . In addition , the observed ATM/ATR-dependent DDR was not immediately activated after CST impairment but appeared gradually after several population doublings [21] . This argues against a direct role of mammalian CST in telomere capping . The structural features of telomeric DNA pose a substantial challenge to the progression of the replication machinery , rendering this region vulnerable to fork stalling [45] , [46] . CST-depleted human chromosomes were reported to accumulate multi-telomeric signals that are hallmarks of fragile sites within telomeres and which form under conditions of replication stress [20] , [21] . In addition , the kinetics of BrdU incorporation demonstrated that CTC1/STN1 depletion slows the replication through telomeric tracks [19] , [20] , [21] . Thus , it has been proposed that , in contrast to yeast , the mammalian CST complex contributes to the maintenance of telomere integrity via promoting replication of the duplex telomeric DNA and not by providing an end-capping function . Disruption of the CST components in Arabidopsis results in more pronounced telomere-related phenotypes than downregulation of the complex in human cell lines [15] , [26] , [27] . Some of the phenotypes , such as elevated telomeric recombination , increase in G-overhang size and appearance of chromosome end-to-end fusions , are considered to be hallmarks of telomere deprotection . This led to the suggestion that , in plants , CST contributes to chromosome end-capping . Consistent with this idea , we demonstrate that the growth defects and genome instability are aggravated by EXO1 , a nuclease responsible for resecting exposed chromosome termini in a variety of settings and organisms including Arabidopsis [11] , [31] , [47] . This is similar to the situation in budding yeast where mutations reducing DSB resection rescue the lethality arising from Cdc13 inactivation [11] , [42] , [48] and further substantiates parallels between CST functions in yeast and plants . Nevertheless , three lines of observations argue against the interpretation that STN1 mediates merely telomere capping in plants . First , we show that telomeric defects in Arabidopsis stn1 mutants are mitigated by long telomeres . G1 stn1 ( W/C ) plants that have longer telomeres exhibit only mild telomere dysfunction compared to G1 stn1 ( C ) plants that have shorter telomeres . Second , although EXO1 exacerbates telomere dysfunction upon STN1 depletion , it does not cause apparent DNA degradation at the bulk of telomeres . Finally , STN1 inactivation leads to aggravated telomere shortening , indicating that chromosome end deprotection is primarily caused by insufficient maintenance of telomeric DNA . Mechanistically , loss of telomeric DNA may be promoted by de-repressing homologous recombination leading to telomere rapid deletion [49] . Indeed , the higher level of t-circles detected in Arabidopsis CST mutants supports this scenario [15] , [26] , [27] . However , comparable levels of t-circle excision have also been detected in Arabidopsis ku mutants where it does not result in telomere loss and deprotection [39] . Alternatively , as in mammals , Arabidopsis CST may facilitate replication through duplex telomeric DNA and the telomere shortening associated with CST dysfunction may be due to frequent collapse of , or failure to restart , stalled replication forks . This notion is supported by a similar set of phenotypes observed in Arabidopsis stn1 and pol α mutants with respect to telomere structure and protection . Furthermore , DNA damage signaling in Arabidopsis CST mutants largely depends on ATR , but not on ATM , an observation consistent with a role for CST in replication [28] , [29] . A surprising result of this study is that EXO1 accelerates telomere dysfunction in stn1 mutants without causing apparent resection at the majority of telomeres . Although we detected signatures of EXO1 activity in the form of altered frequencies of sequence permutations at chromosome termini , EXO1 does not seem to contribute to the G-overhang extension in STN1-null plants . We propose that while EXO1 may gain access to STN1-depleted telomeres , for example at the sites of collapsed replication forks , its activity may be greatly limited in the context of long telomeres by telomere binding proteins [50] . EXO1 activity may be unleashed at short telomeres that lack sufficient amounts of telomere binding proteins , causing excessive degradation of the remaining telomeric and subtelomeric DNA , and triggering a strong DNA damage response and chromosome end-to-end fusions . Consistent with this idea , inactivation of EXO1 in Arabidopsis limits DNA damage signaling as well as the frequency of telomere fusions in the context of short telomeres . The molecular mechanism by which CST facilitates telomere maintenance is still unknown . CST associates with DNA polymerase α and stimulates its processivity and affinity to ssDNA in vitro [34] , [51] . It has been proposed that CST might recruit DNA polymerase α to help restart collapsed replication forks and to promote the C-strand fill-in synthesis during G-overhang maturation [13] , [14] , [20] , [21] . Our data showing shorter G-overhangs in Arabidopsis stn1 tert and exo1 stn1 tert mutants indicate that the long G-overhangs in STN1-deficient plants are partially derived from an impaired C-strand fill-in or aberrant processing after telomerase elongation . Strikingly , the hypomorphic DNA polymerase α mutations described in this study appear to be proficient in the C-strand fill-in reaction , while still causing aberrant telomere length maintenance . Considering the essential role of polymerase α in DNA replication , pol α and icu2-1 alleles are expected to produce largely functional proteins capable of initiating Okazaki fragment replication . Thus , pol α and icu2-1 may represent separation-of-function mutations that affect only dispensable activities of polymerase α , such as the restart of collapsed replication forks [52] , which may be relevant to the telomere maintenance defects in these mutants . This functional separation draws parallels with the recent observation that CST promotes telomere replication and G-overhang maturation through distinct mechanisms [23] . Our discovery that hypomorphic alleles of DNA polymerase α partially recapitulate the telomere-phenotypes of STN1-deficient plants shows that the malfunction of a general DNA replication factor may lead to very specific defects at telomeres . This raises the question of whether the function of the CST complex is specifically tailored to telomeres , or whether it has a more general role . Although CST does not localize to replication loci [16] , it was shown to contribute to genome-wide replication restart after hydroxyurea treatment in human cells [20] . This observation led to the suggestion that CST may be a specialized replication factor that is needed under conditions of replication stress . Non-telomeric function of STN1 is further inferred from a study in budding yeast , where overproduction of Stn1 led to its localization to nontelomeric loci and sensitized the cells to replicative stress in polymerase αdependent manner [53] . In this study , we took advantage of unscheduled re-replication of heterochromatic regions in Arabidopsis atxr5 atxr6 mutants [41] to functionally assay the role of STN1 in DNA replication . Elevated PARP1 level in atxr5 atxr6 plants indicates that these mutants experience chronic DNA damage stress , which can be further increased by bleomycin treatment . Thus , our finding that STN1 promotes re-replication in these settings supports data in human cells suggesting a function for the CST complex in facilitating DNA synthesis under replication stress [20] . A further inference is that CST function in genome maintenance is not limited to telomeres; they may just represent a genomic region where the consequence of CST malfunction is phenotypically most apparent .
The following Arabidopsis thaliana mutant lines were used in this study: stn1-1 [26] , tert-1 [33] , exo1a-1 and exo1b-1 [31] , pol α [37] , icu2-1 [36] and atxr5 atxr6 [40] . The primers used for PCR genotyping are listed in Table S1 . The plants were grown at 22°C with 16/8 h light/dark period . The DNA was extracted from four to six weeks old plants as previously described [31] . Terminal restriction fragment ( TRF ) analysis with Tru9I restriction enzyme [54] , t-circle amplification assay [39] , hairpin-ligation assay [31] and fusion PCR [38] were performed according to published protocols . Telomere length was determined from TRF blots by using TeloTool , software for quantitative analysis of TRF data [55] . Mitotic spreads for quantification of anaphase bridges were prepared from Arabidopsis floral buds as previously described [32] . The frequency of anaphase bridges was calculated by scoring at least 200 anaphases per plant; at least three plants were analyzed for each category to calculate average frequency and SD . The in gel hybridization was performed as earlier described [56] . We used HindIII to generate long TRFs in order to minimize effect of telomere length and heterogeneity on signal detection . For the control reactions , the DNA was pre-treated with 30 units T4 DNA polymerase ( New England Biolabs ) for 30 min at 37°C to remove the 3′ G-overhangs . The hybridization signals were scanned using the Molecular Imager PharosFX Plus ( BioRad ) and quantified with the Image Lab software ( BioRad ) . To correct for loading differences , the hybridization signals were normalized to the ethidium bromid staining of the agarose gels . The G-overhang signal was obtained after subtracting the signal from the T4 DNA Pol pretreated samples . Finally , the G-overhang signal of the wild type samples was set to one and all other samples were normalized to this value . We usually perform quantitative comparisons of samples run in the same gels to minimize experimental variation . The analysis was performed as previously described with small modifications [31] . Genomic DNA ( 1 µg ) was ligated with the hairpin blunt HP 3 ( Table S1 ) at 16°C , followed by heat inactivation . The samples were digested with 20 U Tru1I ( Fermentas ) overnight at 65°C , ethanol precipitated and dissolved in 25 µl TE . 5 µl of the reaction were used as a template for 20 cycles of PCR amplification in 1xGoTaq buffer ( Promega ) supplemented with 0 . 2 mM dNTPs , 0 . 5 µM of ( TTTAGGG ) 4 primer , 0 . 5 µM HP 3 primer and 2 . 5 U GoTaq ( Promega ) in a volume of 50 µl . The PCR products were ethanol precipitated , digested with AluI and purified with the Nucleospin column purification kit ( Macherey-Nagel ) . The library for Illumina sequencing was prepared with the NEB Next DNA library Prep Reagent Set ( New England Biolabs ) according to the manufacturer's instruction using 100 ng of digested DNA . The samples were sequenced with single-end 50 bp reads using the Illumina HiSeq 2000 . Per sequencing lane , eight independent samples were pooled each carrying a unique barcode ( Table S1 ) . Per barcode , 1 . 1×106–3 . 1×106 reads were obtained and analyzed with the custom-made program “TELOMERATOR” [31] . The nuclei were prepared from 1 . 5 g mature rosette leaves ( 3rd–8th ) collected from 3–4 week old plants . The leaves were chopped on ice in 5 ml isolation buffer ( 15 mM Tris-HCl pH = 7 . 5 , 2 mM EDTA , 0 . 5 mM Spermidine , 20 mM NaCl , 80 mM KCl , 15 mM β-mercaptoethanol , 0 . 1% Triton X-100 , pH 7 . 5 ) . The released nuclei were purified by filtration through a 50 µm CellTrics disposable filter ( Partec ) and collected by centrifugation at 2800×g at 4°C . The nuclei were resuspended in 600 µl isolation buffer , filtered through a clean 50 µm filter and stained with 400 µl CyStain UV buffer ( Partec ) . After a final filtration through a 40 µm BD Falcon cell strainer ( BD Biosciences ) , the stained nuclei were sorted with the BD FACS Aria III ( BD biosciences ) using an 85 µM nozzle with sheath pressure at 45 psi and a threshold rate of ∼3000 events/s . The DNA of the 8C fraction was extracted by incubating 500 . 000 sorted nuclei in 200 µl lysis buffer ( 150 mM NaCl , 10 mM Tris-HCl pH = 8 . 0 , 100 mM EDTA , 1% SDS ) for 10 min at 65°C while shaking . The RNA was removed by RNase A treatment followed by incubation with Proteinase K ( Thermo Scientific ) . The DNA was purified by phenol:chlorophorm:isoamyl alcohol ( 25∶24∶1 , v/v ) extraction as well as ethanol precipitation . Finally , the DNA was sonicated to an average fragment size of ∼200 bp using the Covaris S2 . 20 ng DNA were used to prepare the library for Illumina sequencing with the NEXTflex ChIP-Seq Kit ( Bioo Scientific ) . NEXTflex-96 DNA barcodes ( Bioo Scientific ) were used to sequence four independent samples per lane with single-end 100 bp reads using the Illumina HiSeq2000 . For flow cytometry , nuclei from several 2 week old seedlings were prepared according to a scaled-down version of the protocol described above . The fluorescence intensity of nuclei was measured with a CyFlow counter ( Partec ) and their distribution as well as statistical analysis of 8C peaks was done using FlowMax software ( Partec ) . The sequencing reads were aligned to the A . thaliana genome ( TIAR 10 ) using Bowtie2 [57] . For reads mapping to multiple positions , the best ( or randomly selected in case of ties ) alignment was used . Multiple reads mapping to the same position were collapsed to one read . The method used for analyzing re-replicated regions derived from the strategy described by Stroud et al . [58] . The genome was tiled into 10 kb bins and the reads in the bins were summed up . Bins with a mappability of 0 and bins with more reads than the top 1% of all bins plus the median were removed . The mappability was calculated using the GEM toolbox for 100 bp reads [59] . For the atxr5 atxr6 and atxr5 atxr6 stn1 mutants where two independent samples were sequenced , the bins were averaged between replicates . For all bins , the log2 ratio to the wild type was calculated and then scaled to account for differences in read numbers and base replication ratio by finding the minimum mean ratio in a 5 Mb sliding window for each chromosome . The mean of the base window was subtracted from each bin and the result divided by the standard deviation to obtain a scaled value that is 0 for a base replication rate and varies according to the variability of each sample ( higher overall variability is penalized ) . In the figure , all bins were additionally smoothed by calculating a running mean over 10 bins . The smoothing lines were produced by cubic regression splines . | Telomeres form an elaborate nucleoprotein structure that may represent an obstacle for replication machinery and renders this region prone to fork stalling . CST is an evolutionary conserved complex that was originally discovered to specifically act at telomeres . Interestingly , the function of CST seems to have diverged in the course of evolution; in yeast it is required for telomere protection , while in mammals it was proposed to facilitate replication through telomeres . In plants , inactivation of CST leads to telomere deprotection and genome instability . Here we show that the telomere deprotection in Arabidopsis deficient in STN1 , one of the CST components , is consistent with defects in telomere replication and that STN1 phenotypes can be partially phenocopied by an impairment of a general replication factor , DNA polymerase α . In addition , we provide evidence that STN1 facilitates re-replication at non-telomeric loci . This suggests a more general role of CST in genome maintenance and further infers that its seemingly specific function ( s ) in telomere protection may rather represent unique requirements for efficient replication of telomeric DNA . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"Methods"
] | [
"biotechnology",
"chromosome",
"structure",
"and",
"function",
"telomeres",
"plant",
"science",
"dna",
"replication",
"forms",
"of",
"dna",
"plant",
"genomics",
"dna",
"chromosome",
"biology",
"plant",
"genetics",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",... | 2014 | Role of STN1 and DNA Polymerase α in Telomere Stability and Genome-Wide Replication in Arabidopsis |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.