Search is not available for this dataset
article stringlengths 4.36k 149k | summary stringlengths 32 3.35k | section_headings listlengths 1 91 | keywords listlengths 0 141 | year stringclasses 13
values | title stringlengths 20 281 |
|---|---|---|---|---|---|
The cysteine protease caspase-7 has an established role in the execution of apoptotic cell death , but recent findings also suggest involvement of caspase-7 during the host response to microbial infection . Caspase-7 can be cleaved by the inflammatory caspase , caspase-1 , and has been implicated in processing and activation of microbial virulence factors . Thus , caspase-7 function during microbial infection may be complex , and its role in infection and immunity has yet to be fully elucidated . Here we demonstrate that caspase-7 is cleaved during cytosolic infection with the intracellular bacterial pathogen , Listeria monocytogenes . Cleavage of caspase-7 during L . monocytogenes infection did not require caspase-1 or key adaptors of the primary pathways of innate immune signaling in this infection , ASC , RIP2 and MyD88 . Caspase-7 protected infected macrophages against plasma membrane damage attributable to the bacterial pore-forming toxin Listeriolysin O ( LLO ) . LLO-mediated membrane damage could itself trigger caspase-7 cleavage , independently of infection or overt cell death . We also detected caspase-7 cleavage upon treatment with other bacterial pore-forming toxins , but not in response to detergents . Taken together , our results support a model where cleavage of caspase-7 is a consequence of toxin-mediated membrane damage , a common occurrence during infection . We propose that host activation of caspase-7 in response to pore formation represents an adaptive mechanism by which host cells can protect membrane integrity during infection .
Pore-forming toxins are integral to the virulence of many microbial pathogens , including the Gram-positive bacterium , Listeria monocytogenes . This facultative intracellular pathogen can cause life-threatening disease in humans , particularly in the very old and very young , the immunocompromised , and pregnant women [1] . In macrophages , L . monocytogenes gains access to its replicative niche via the action of a pore-forming cholesterol-dependent cytolysin , Listeriolysin O ( LLO ) [2] . LLO-dependent perforation of the primary phagosomal membrane allows the pathogen to escape into the cytosol , where it grows to high titers in the apparent absence of overt cell damage until late in infection [3] , [4] . Virulence of L . monocytogenes therefore requires a delicate balance between expressing virulence factors , such as LLO , to survive host cell defenses while maintaining an intact host cell niche . Infection with L . monocytogenes expressing an overactive allele of LLO [5] , [6] or with a strain that overproduces LLO [7] results in host cell damage and attenuation in vivo , primarily due to killing of extracellular L . monocytogenes by neutrophils [5] . It can therefore be inferred that the integrity and survival of infected host cells affects virulence of L . monocytogenes . The infected macrophage plays a dichotomous role during L . monocytogenes infection , acting both as a reservoir for bacterial replication and as a source for inflammatory signals that result from recognition of microbial ligands or cellular stress . L . monocytogenes activates many inflammatory pathways in the cell that promote eventual bacterial clearance and immunity . Infection stimulates Toll-like receptors TLR2 and possibly TLR5 , and the Nod-like receptors ( NLRs ) Nod1 and Nod2 , resulting in NF-κB-dependent pro-inflammatory gene transcription [8]–[12] . Cytosolic L . monocytogenes triggers assembly of the caspase-1 associated inflammasome , a multiprotein complex whose formation can lead to an inflammatory cell death termed pyroptosis [13] . Active caspase-1 processes pro-IL-1β and pro-IL-18 , inflammatory cytokines that promote antimicrobial properties of phagocytes and stimulate adaptive immunity [14]–[16] . Several NLRs activate caspase-1 as a result of L . monocytogenes infection , including NLRC4 , NLRP3 and AIM2 , all of which require the adaptor protein ASC [17]–[20] . Studies in knockout mice have demonstrated that caspase-1 is important for primary clearance of L . monocytogenes , but the role of other caspases in the innate immune response to infection is less well defined [21] . Caspase-7 is a member of a family of cytosolic cysteine proteases that promulgate diverse biological responses , including programmed cell death and inflammation . Caspases can also promote cell survival , as caspase-1 positively regulates cholesterol biosynthesis in response to a bacterial pore-forming toxin , aerolysin [22] . A defining characteristic of caspases is specific cleavage of substrates at aspartic acid residues using a cysteine side chain as a nucleophile for peptide hydrolysis [23] . Caspases reside in the cytosol as zymogens that require dimerization and/or proteolytic cleavage before becoming catalytically active . Cleavage of caspase-7 results in a large and a small fragment; proteolysis between the large and small subunit is considered the fundamental activating event [24] . Caspase-7 was initially characterized as an “executioner” caspase whose activity directs the highly regulated cascade of events leading to cellular disassembly during apoptotic cell death ( [25] , [26] and recently reviewed in [27] ) . Recent studies have implicated maturation of caspase-7 as a consequence of infection or inflammatory stimulation . Caspase-7 cleavage occurs during infection by the Gram-negative intracellular bacterial pathogens Salmonella enterica serovar Typhimurium and Legionella pneumophila [28] , [29] . In these contexts , caspase-7 was cleaved by inflammasome-associated caspase-1 . Caspase-7-deficient macrophages allowed increased L . pneumophila intracellular growth , possibly due to delayed macrophage cell death [28] . These studies provide evidence that caspase-7 is involved in host-pathogen interactions . Here we show that caspase-7 cleavage is triggered by membrane damage during L . monocytogenes infection , and is dissociated from canonical markers of apoptosis . Caspase-7 cleavage occurred in the absence of caspase-1 , distinct from the activation cascade observed during infection by S . Typhimurium and L . pneumophila [28] , [29] . Infected macrophages lacking caspase-7 exhibited increased plasma membrane permeability , which required ongoing production of LLO . Treatment of host cells with sublytic concentrations of recombinant LLO , as well as a pore-forming toxin from Staphylococus aureus , α-hemolysin , triggered caspase-7 activation in the absence of infection . Together , these data lead us to propose that caspase-7 activation is a protective host response to plasma membrane damage that limits subsequent cytotoxicity during bacterial infection .
We first investigated whether hallmarks of proteolysis associated with caspase-7 activity could be detected in L . monocytogenes infected cells . To this end , we infected bone marrow derived macrophages ( BMDM ) from C57BL/6 ( BL/6 ) mice with either a WT strain of L . monocytogenes , or with a strain lacking the hly gene , which encodes the pore-forming toxin Listeriolysin O ( LLO− strain ) . The LLO− strain cannot escape the primary phagosome and does not replicate within macrophages . At 8 h pi , we assessed DEVDase activity , an indicator of caspase-3/7 proteolytic activity , by measuring cleavage of a luminescent DEVD containing substrate in cell lysates ( Figure 1A ) . Both caspase-3 and -7 are able to cleave exogenous DEVD substrate , but recent studies suggest that proteases have overlapping but distinct physiological substrates within the host cell [30] . We detected an increase in DEVD-specific enzymatic activity in response to infection with WT , but not the LLO− mutant , indicating that bacterial uptake per se was insufficient to stimulate DEVDase activity . The difference in DEVDase activity between cells infected at MOI1 vs . MOI5 was not attributable to differences in the number of infected cells ( Figure 1B ) nor the number of bacteria , as we isolated nearly equivalent CFU from cultures infected under these conditions at 8 h pi ( Figure S1 ) . At MOI1 and MOI5 , 80–90% of cells were infected , although at MOI5 , macrophages contained higher numbers of bacteria at early time points . To specifically determine if caspase-7 was activated during L . monocytogenes infection , we investigated whether L . monocytogenes infection induced caspase-7 cleavage in BMDM by lysing infected cells at 8 h pi and analyzing caspase-7 by immunoblot using an antibody that recognizes both the full length protein and the larger cleavage product ( Figure 1C ) . We observed cleaved caspase-7 protein in BL/6 cells infected with WT at MOI1 and MOI5 , but not in uninfected ( mock ) BMDM or BMDM infected with the LLO− strain ( MOI 5 ) , suggesting bacterial occupancy of the phagosome was insufficient to stimulate cleavage of this protease . Full-length caspase-7 protein was detected in BL/6 BMDM in all samples and was used as a protein loading control since the overall abundance of the cleaved caspase-7 product was low compared to full length . We also assessed caspase-3 cleavage during L . monocytogenes infection by immunoblot ( Figure S2 ) . Although we detected cleavage of caspase-3 in response to infection , the cleavage of caspase-7 in response to infection was more robust , and we therefore decided to further investigate caspase-7 . To determine the kinetics of caspase-7 activation during infection , we assayed cell lysates for cleavage at 2 , 4 , 6 and 8 hpi ( Figure 1D ) . Cleavage of caspase-7 was detected between 4 and 8 hpi , by which time the majority of L . monocytogenes are replicating in the cytosol . Taken together , these data indicate that caspase-7 is activated during L . monocytogenes infection , and requires LLO . Caspase-7 was originally identified as an executioner caspase whose activity was associated with activation of the apoptotic cascade [31] . To determine whether caspase-7 cleavage was also associated with molecular markers of programmed cell death during L . monocytogenes infection , we first quantified the number of infected cells at 8 hpi by staining samples with anti-Listeria antibody , followed by immunofluorescence microscopy ( Figure 1B ) . We then prepared infected cell cultures for TUNEL staining to visualize DNA fragmentation . BMDM were exposed to L . monocytogenes at MOI1 or MOI5 , and at 8 h pi cells were fixed and analyzed for DNA fragmentation by immunofluorescence microscopy . In agreement with previous findings [3] , we found no significant increase in DNA fragmentation in cells infected at MOI1 compared to uninfected control cells ( Figure 1E ) . At MOI5 , there was a statistically significant increase in the number of TUNEL positive cells compared to uninfected cells . However , comparing the overall number of TUNEL positive cells even in the MOI5 infection ( ∼10–15% ) to the number of cells infected ( 80–90% ) , we conclude that the majority of cells infected with L . monocytogenes do not display this characteristic hallmark of apoptosis by 8 h pi . We obtained similar results using Annexin V to measure phosphatidylserine exposure , an early event in apoptotic cell death ( Figure S3 ) . We also measured lactate dehydrogenase ( LDH ) release at 8 h pi as an indicator of overall cell viability ( Figure 1F ) . L . monocytogenes infection induced a modest but notable increase in LDH release at 8 h pi compared to uninfected cells . This range is consistent with previously published reports [3] , and indicates that at low MOI the majority of cells infected with L . monocytogenes retain LDH . We conclude from these results that L . monocytogenes infection stimulates caspase-7 cleavage , but is not predominantly associated with molecular markers of apoptotic cell death . L . monocytogenes infection triggers activation of caspase-1 , and during some bacterial infections , caspase-7 can be a substrate of caspase-1 [28] , [29] . We therefore tested whether the activation of caspase-7 during L . monocytogenes infection was dependent on the presence of caspase-1 by infecting csp1−/− BMDM and measuring DEVDase activity . We detected increased DEVDase activity upon infection of caspase-1 deficient BMDM ( Figure 2A ) . Analysis of caspase-7 protein by immunoblot of infected cell lysates definitively revealed cleavage of caspase-7 in csp1−/− BMDM upon infection , and this cleavage was dependent on LLO ( Figure 2B ) . ASC is a key adaptor protein necessary for caspase-1 activation during inflammasome formation [32] . To determine if ASC was necessary for caspase-7 cleavage , we assayed caspase-7 cleavage in infected Asc−/− BMDM and found that caspase-7 activation also did not require ASC during L . monocytogenes infection ( data not shown ) . These data demonstrate that the mechanism of caspase-7 cleavage during L . monocytogenes infection does not require caspase-1 , and thus appears to be distinct from mechanisms reported for other intracellular bacterial pathogens . Extracellular ( TLR2 and TLR5 ) and intracellular ( Nod1 and Nod2 ) pattern recognition receptors are able to sense L . monocytogenes and direct transcription of cytokines and chemokines to promote inflammation and clearance [8]–[12] . To determine whether these bacterial recognition pathways contributed to caspase-7 activation upon infection , we evaluated L . monocytogenes-induced DEVDase activity and caspase-7 cleavage in BMDM from myd88−/− and rip2k−/− mice . MyD88 is a critical adaptor that mediates signaling for 9 of the 10 TLRs with known ligands . RIP2 is a protein kinase that mediates inflammatory signaling through Nod1 and Nod2 . We observed infection-induced DEVDase activity and caspase-7 cleavage in myd88−/− and rip2k−/− BMDM at levels comparable to WT BMDM ( Figure 2C and 2D ) . We also detected caspase-7 cleavage in nod1−/− and nod2−/− macrophages comparable to WT ( data not shown ) . From these results , we conclude that the innate immune signaling pathways regulated by MyD88-dependent TLRs , Nod1 , Nod2 , ASC , and caspase-1 are individually dispensable for caspase-7 activation during L . monocytogenes infection . Since caspase-7 was cleaved during infection with L . monocytogenes , we hypothesized that the activity of this enzyme could impact intracellular infection . We therefore measured L . monocytogenes replication in BL/6 and csp7−/− BMDM in an antibiotic protection assay where 50 µg/ml of the cell impermeant antibiotic gentamicin was added 30 min pi to eliminate extracellular bacteria ( Figure 3A ) . We observed significant decline in the number of intracellular bacteria over time within the csp7−/− BMDM compared to BL/6 cells . We considered the possibility that csp7−/− cells were becoming permeable to gentamicin , allowing antibiotic influx into the intracellular space , killing intracellular bacteria . To test this idea , we performed a gentamicin washout , removing gentamicin from the medium 2 h pi , after which antibiotic-free medium was added and replaced every hour to limit extracellular bacterial growth ( Figure 3B ) . When gentamicin was removed from the extracellular space , there was no significant difference in bacterial replication between BL/6 and csp7−/− macrophages . These data suggest increased plasma membrane permeability of L . monocytogenes-infected macrophages in the absence of caspase-7 . To more directly assess membrane integrity of L . monocytogenes infected BMDM , we evaluated their permeability to small fluorescent molecules during infection . At 8 h pi we exposed live infected cells to rhodamine-phalloidin ( Rh-P ) , a small ( 1200 Da ) cell impermeant compound that binds F-actin in the host cell . After a 15-minute exposure , the cells were fixed and counter-stained with DAPI . Permeability to Rh-P was compared between csp7−/− and BL/6 BMDM using epifluorescence microscopy ( Figure 3C and 3D ) . Untreated BMDM excluded Rh-P , while cells treated with Triton X-100 were fully permeable to Rh-P ( Figure 3C insets ) . Although Rh-P permeability was similar between genotypes in the absence of infection , during infection the overall number of cells stained with phalloidin was significantly greater in the csp7−/− macrophages compared to wildtype BMDM . Notably , despite permeability to small molecules , infected csp7−/− cells displayed no significant increase in LDH release at 8 h pi compared to control cells ( Figure S4 ) . These findings suggest that membrane damage occurring during L . monocytogenes infection may be transient , and is insufficient to allow efflux of the large LDH tetramer ( ≈137 , 000 Da ) [33] . Taken together , these data indicate that caspase-7 promotes plasma membrane integrity during infection with L . monocytogenes . Plasma membrane damage during L . monocytogenes infection could be the result of toxin-mediated pore formation . To determine whether bacterial protein synthesis was needed to induce host membrane instability , we treated infected cells with a bacteriostatic concentration of erythyromycin ( erm ) , a macrolide antibiotic that targets the bacterial 50S ribosomal subunit . Erythromycin was added to infected cells at 2 h pi , and at 8 h pi we found a significant reduction in the number of caspase-7-deficient phalloidin positive cells upon inhibition of bacterial protein synthesis ( Figure 4A ) . We then tested the requirement for LLO in driving the plasma membrane damage we observed in infected caspase-7 deficient macrophages . LLO is an oligomeric cytolysin that forms pores in cholesterol-containing membranes [34] . The oligomerization of LLO occurs optimally at acidic pH [6] , [35] but still maintains some pore-forming activity at neutral pH [36]–[38] . LLO is actively produced by cytosolic L . monocytogenes [39] . We hypothesized that LLO production by cytosolic bacteria could result in continual but transient membrane damage during infection , which was exacerbated in csp7−/− cells . To test this hypothesis , we used a strain of L . monocytogenes with LLO expression controlled by an IPTG inducible promoter ( iLLO ) [40] . We transiently induced production of LLO during infection by the iLLO-expressing strain to permit escape from the phagosome , and then removed IPTG from the cell culture medium . Determination of endpoint CFU at 8 h pi in BL/6 macrophages infected with the WT and iLLO strains revealed no significant differences in the number of intracellular bacteria between strains ( Figure 4B ) . However , IPTG removal at 2 h pi resulted in significantly less phalloidin permeable BL/6 BMDM infected with the iLLO strain compared to cells infected with WT bacteria at 8 h pi ( Figure 4C ) . The reduction of permeability in csp7−/− macrophages infected with the iLLO-expressing strain compared to the WT strain was even more pronounced , supporting a protective role for this protease against infection associated membrane damage ( Figure 4D ) . Increasing induction time from 2 to 4 hours increased the number of phalloidin positive cells in both genotypes . Therefore , we conclude that LLO is required to drive the membrane permeability defect in macrophages lacking caspase-7 . Bacterial production of LLO was required for permeability of caspase-7 deficient macrophages during infection ( Figure 4 ) , leading us to question whether LLO was also the trigger that stimulated cleavage of caspase-7 . To test this possibility , we infected BMDM with L monocytogenes , and then added 10 µg/ml erm at 2 h pi to inhibit bacterial protein synthesis and assayed for caspase-7 cleavage by immunoblot . We were unable to detect cleavage of caspase-7 when erm was added to infected cell cultures , demonstrating that bacterial protein synthesis was necessary to stimulate this response ( Figure 5A ) . We also infected BMDM with the IPTG-inducible LLO expressing strain , incubating with IPTG for 2 h pi to promote vacuolar escape , followed by removal of IPTG from the medium to limit LLO production . Although the iLLO and WT strains grew to equivalent intracellular CFU under these conditions ( Figure S5 ) , we only observed caspase-7 cleavage in cells infected with the WT strain , not the iLLO strain ( Figure 5A ) . LLO is one of a large family of cholesterol dependent cytolysins ( CDC ) produced by Gram-positive bacterial pathogens . To determine whether activation of caspase-7 during bacterial infection was specific to LLO , or whether other CDC toxins behaved similarly , we infected wildtype and caspase-7−/− BMDM with L . monocytogenes expressing the related CDC toxin perfringolysin O ( PFO ) instead of LLO [41] . L . monocytogenes expressing PFO also stimulated caspase-7 activation ( 5B ) . Thus , we find that cytosolic bacterial replication per se is insufficient to stimulate caspase-7 cleavage , and that ongoing production of toxin is necessary for activation of the protease during L . monocytogenes infection . We next asked if exogenous LLO could activate caspase-7 independent of infection . To this end , we first determined the sublytic concentration range of purified recombinant LLO ( rLLO ) , using LDH release as a marker for loss of viability at 1 h post treatment ( Figure 5B ) . We then probed cell lysates incubated for 1 h with rLLO for caspase-7 cleavage by immunoblot . Concentrations of exogenous rLLO , i . e . , 0 . 25 µg/ml , that only induced a low level of LDH release ( ∼5% ) from BMDM up to 5 h pi ( Figure 5C and Figure S6 ) were sufficient to stimulate caspase-7 cleavage ( Figure 5D ) . As we increased the concentration of exogenous LLO , we saw a concomitant increase in LDH release that correlated with more robust caspase-7 cleavage ( Figure 5C and 5D , and data not shown ) . These data demonstrate that sublytic concentrations of LLO can activate caspase-7 in the absence of infection . To address whether LLO in its native conformation was necessary to stimulate caspase-7 cleavage , we heat-treated the protein for 10 minutes at 65°C ( LLO* ) , and compared caspase-7 cleavage against cells intoxicated with the same concentration of native protein . When LLO was heat-inactivated , we observed significantly less caspase-7 cleavage compared to cells treated with active toxin ( Figure 5D ) . These data show that native LLO is necessary for induction of caspase-7 cleavage . One interpretation of these data could be that a native epitope of LLO stimulates caspase-7 cleavage through binding of a host receptor at the plasma membrane . However , since we observed LLO-dependent caspase-7 cleavage whether LLO was extracellular or in the cytosol , we propose instead that LLO triggers caspase-7 cleavage through pore formation . To address whether caspase-7 cleavage occurs as a general response to membrane damage , we assessed the ability of exogenous detergent treatment to activate the protease . Digitonin is a non-ionic glycoside detergent that can reversibly permeabilize the plasma membrane by forming complexes with cholesterol [42] . To determine if caspase-7 could be activated in response to detergent permeabilization , we treated BMDM with lytic and sublytic concentrations of digitonin and probed for caspase-7 cleavage 1 h post treatment by immunoblot . Although digitonin treatment induced plasma membrane damage , as assessed by measuring LDH release ( Figure 6A ) , detergent treatment did not stimulate caspase-7 cleavage at any concentration tested ( Figure 6B ) . We obtained similar results using a sublytic to lytic range of the detergents Nonidet P-40 , Triton-X100 , Tween-20 and saponin ( data not shown ) . These data indicate that plasma membrane damage alone is insufficient to stimulate caspase-7 activation . We then asked if caspase-7 cleavage could be stimulated by other pore-forming toxins ( PFT ) . Staphylococcus aureus α-hemolysin is a well studied PFT distinct from the CDC family of toxins . Treatment of BMDM with α-hemolysin resulted in little to no LDH release up to 8 h post treatment ( Figure 6C and S7 ) , consistent with previous reports that nucleated cells can repair damage caused by this toxin [43]–[45] . To test the pore-forming ability of the toxin we confirmed it was lytic on red blood cells ( data not shown ) . α-hemolysin did trigger caspase-7 activation as early as 5 min post treatment ( Figure 6D ) . Some cells types activate the apoptotic cascade in response to treatment with low doses , but not high doses , of α-hemolysin [46] , [47] . However we observed no morphological hallmarks of apoptosis up to 8 h post toxin treatment ( data not shown ) . These results suggest that caspase-7 activation is generally responsive to membrane damage by bacterial pore-forming toxins .
Here we report that caspase-7 protects plasma membrane integrity during infection with L . monocytogenes . We found that the transient membrane damage observed during infection was dependent on the pore-forming toxin , LLO . We also showed that caspase-7 cleavage by L . monocytogenes infection did not individually require caspase-1 , ASC , MyD88-dependent TLR , or Nod1 and Nod2 signaling . Instead , we found that sublytic membrane damage by recombinant LLO and α-hemolysin , even in the absence of infection , could stimulate caspase-7 activation . L . monocytogenes expressing PFO also stimulated caspase-7 activity . However , treatment of BMDM with sublytic concentrations of detergent did not trigger caspase-7 cleavage . These data suggest that general membrane damage or changes in ion gradients across the membrane are insufficient to stimulate caspase-7 and the subsequent cytoprotective response . Although these observations do not rule out a role for pattern recognition of specific microbial ligands in activation of caspase-7 , our results suggest the possibility that caspase-7 participates in sensing and/or repair of PFT-induced membrane damage in the infected cell by a caspase-1 independent mechanism ( Figure 7 ) . Given the abundance and importance of PFTs for many pathogens , we propose a model whereby caspase-7 is induced by plasma membrane damage that occurs during microbial infection and initiates mechanisms of phagocyte membrane repair to protect the infected cell . Caspases are well studied for their pivotal roles in apoptotic and inflammatory cell death cascades . However , there is a growing body of evidence that links caspase activation to additional cellular functions , independent from those leading to cell death . For instance , in vitro and in vivo studies indicate that caspase-7 cleaves and is regulated by sterol regulatory element binding proteins ( SREBPs ) , transcription factors bound to endoplasmic reticulum ( ER ) and nuclear membranes [48] , [49] . When cells are depleted of cholesterol , proteolytic cleavage of SREBPs results in their nuclear translocation and subsequent transcription of genes responsible for cholesterol biosynthesis and lipoprotein uptake [50] . Gurcel et al demonstrated that activation of SREBPs in response to the pore-forming toxin , aerolysin , promoted intoxicated cell survival in a caspase-1 dependent manner [22] . However , in our studies , activation of caspase-7 did not require caspase-1 . Moreover , we observed no analogous upregulation of cholesterol biosynthesis genes ( data not shown ) . Caspases can also regulate virulence determinants during intracellular infection . Caspase-3 , a canonical “executioner” caspase highly related to caspase-7 , was recently shown to cleave the Salmonella type III secreted effector protein SipA [51] . Processing of SipA was necessary for the generation of inflammation in mouse infection models , and inflammatory signaling is critical for Salmonella pathogenesis [52] , suggesting co-evolution between host and pathogen to promote bacterial virulence in this context . Caspase-mediated cleavage of pathogen-derived proteins can also function to attenuate virulence . For instance , caspase-7 cleavage of ORF57 , an early protein critical for viral replication during KSHV infection , inhibits the expression of viral lytic proteins , thereby limiting viral spread [53] . Taken together , these studies demonstrate that caspases play a complex role during microbial infection , and individual caspases may act in a manner that benefits the host and/or the pathogen . The mechanisms by which the host cell defends itself from LLO-mediated toxicity or damage by other microbial virulence factors are not fully understood , although the concept of plasma membrane or phagosomal membrane damage as a signature of pathogenesis has been recognized [4] , [54] . Infection-associated signals , such as flagellin or membrane damage , may be sensed such that weak signals trigger a protective response , whereas strong signals result in initiation of programmed cell death through pyroptosis or apoptosis . Sublytic levels of membrane damage by pore formation may send a “weak” signal , inducing adaptive mechanisms of membrane repair . Indeed , several groups have demonstrated that mammalian cells can tolerate sublytic concentrations of the related cytolysin , Streptolysin O , which causes transient membrane perforations that allows delivery of large macromolecules into live cells [55]–[57] . Repair of such membrane lesions can occur by the exocytosis of lysosomes to the site of injury , followed by endocytosis of the damaged section of membrane [55] , [58] . Previous studies indicate that LLO is actively produced by cytosolic L . monocytogenes [39] and therefore LLO could cause ongoing transient membrane damage throughout the infection that is kept in check by membrane repair . Investigating the pore forming toxin , BT toxin , from the invertebrate pathogen Bacillus thuringiensis , Bischof et al reported that intoxication of C . elegans and mammalian cells stimulated the unfolded protein response ( UPR ) , a stress-related mechanism aimed at maintaining protein homeostasis in the ER , and that activation of this stress response was critical for nematode defense against the toxin [59] . Whether the UPR is activated in response to LLO toxicity or infection with L . monocytogenes is not known , and whether caspase-7 acts analogously or together with the UPR as a defense mechanism to protect host membrane integrity remains to be elucidated . Given that membrane-damaging virulence factors are virtually ubiquitous among bacterial pathogens , we propose that caspase-7 may represent a broad mechanism by which the host cell can sense this common insult and protect itself from ongoing damage .
Humane animal care at the University of Michigan is provided by the Unit for Lab Animal Medicine , which is accredited by the American Association for Accreditation of Laboratory Animal Care and the Department of Health and Human Services . This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Committee on the Care and Use of Animals ( UCUCA ) of the University of Michigan ( #A3114-01 ) . BMDM were infected with Listeria monocytogenes strain 10403S ( WT ) , DP-L2161 ( LLO− ) [60] which has an in-frame deletion of the hly gene , DH-616 ( iLLO ) [40] whose allele of the hly gene is controlled by an IPTG inducible promoter , and DPL-1875 [60] , a hly deletion strain expressing pfo ( perfringolysin O ) using the hly promoter . The anti-caspase-7 antibody , which recognizes full length and cleaved caspase-7 ( cat #9492 ) was purchased from Cell Signaling Technology . Rhodamine-phalloidin ( cat #R415 ) was purchased from Molecular Probes . Digitonin and α-hemolysin were purchased from Sigma-Aldrich . Bone marrow derived macrophages ( BMDM ) were isolated from WT C57BL/6 mice , except where targeted mutations were indicated . All genetically deficient strains of mice used were constructed in this background . Briefly , isolated BMDM were differentiated in DMEM with 20% heat-inactivated FBS ( Hyclone Laboratories ) , 1% L-glutamine , 1% sodium pyruvate and 30% L929 fibroblast conditioned medium . Cells were cultured in non-TC treated plates , fed with fresh media on day 3 and harvested on day 6 for infection on day 7 . Cultures were maintained in a humidified incubator at 37°C with 5% CO2 . Overnight cultures of L . monocytogenes in BHI broth were incubated statically at 30°C . Prior to infection , bacteria were pelleted and resuspended in PBS . For experiments with the iLLO strain , cultures were grown overnight statically at 30°C in BHI+1 mM IPTG . Before infection , cultures were back diluted 1∶10 in fresh BHI with 1 mM IPTG and grown with shaking at 37°C until an OD600 of 1 . 2 was achieved , at which point bacteria were pelleted and resuspended in PBS to be used for inoculation of cell cultures . All other strains was treated equivalently , but without IPTG . When iLLO was used to infect cells , 10 mM IPTG was added to the cell culture media for the indicated times post infection . Cells were infected at MOI1 or MOI5 for most strains and MOI10 for iLLO to compensate for decreased phagosomal escape of the iLLO strain . When infected in this manner , the WT and iLLO strains grew with similar kinetics and cells supported equivalent intracellular growth at 4 ( data not shown ) and 8 h pi ( Figure 4B ) . Two million BMDM were plated in 60 mm dishes 18 h prior to infection . Cells were infected at MOI5 for 30 minutes , after which the inoculum was removed , cells washed with PBS and replaced with media containing 10 µg/ml gentamicin . For samples infected at MOI1 , cells were spun at 1200 rpm for 3 min to maximize the population of infected cells , after which the inoculum was removed , cells washed , and antibiotic-free media replaced . At 30 min pi , cells were washed again and 10 µg/ml gentamicin was added to inhibit extracellular bacterial replication . At 8 h pi cells were lysed in buffer containing 1% NP-40 on ice for 15 minutes and then spun at 13 , 000 rpm for 10 min to pellet the insoluble fraction . Soluble fractions were separated by SDS-PAGE , transferred to nitrocellulose membrane , and probed with anti-caspase-7 antibody . Membranes were incubated and treated according to the antibody manufacturer instructions . Membranes from individual gels were cut to optimize exposure times for the cleaved and full-length forms of the protein . Bands were visualized using West Femto chemiluminescent substrate ( Thermo Scientific ) . For DEVDase activity assays , 4 million BMDM were plated in a 96 well format and infected according to the protocol outlined for immunoblotting . At 8 h pi , cells were exposed to the DEVDase substrate per the manufacturer's instructions ( Caspase-Glo 3/7 Assay; Promega ) and enzyme activity was quantified by luminometer . For LDH assays , BMDM were seeded into 96-well plates at a density of 4 million cells per plate . Prior to infection , medium was replaced with DMEM lacking phenol red . All subsequent steps were performed in this medium . Cells were infected as described above and supernatants harvested at indicated times post infection . Lactate dehydrogenase release was quantified using the Cytox96 Assay Kit ( Promega ) according to the manufacturer's instruction and quantified by spectrophotometer . L . monocytogenes growth curves were performed according to the following protocol . Sterile glass coverslips were placed in 24 wells , onto which 4×106 BMDM were seeded and allowed to adhere overnight . Bacteria were added to the BMDM at MOI1 and allowed to invade for 30 minutes . The inoculum was then removed , the cells were washed three times in PBS , and fresh BMDM media was added with 50 µg/ml gentamicin to inhibit extracellular bacterial growth . For gentamicin washout experiments , the antibiotic was removed from the cultures 2 h pi , and the media was replaced every hour to limit the contribution of extracellular bacteria to intracellular CFU counts . At the indicated time points , coverslips were removed and BMDM were osmotically lysed and serially diluted to enumerate CFU . CFU were counted using the aCOLyte SuperCount ( Microbiology International , Fredrick , MD ) plate reader and software . For TUNEL staining , BMDM were plated in 6-well format at a density of 2×106 cells per dish and incubated overnight . The cells were then infected according to the protocol outlined for immunoblotting and stained for DNA fragmentation per the manufacturer's instructions ( Roche ) . For host cell permeability assays , BMDM were seeded onto square coverslips in 6 well plates at a density of 5×105 per well the night before infection . The day of infection , host cells were infected with L . monocytogenes for 30 mins , after which the cells were washed 3 times with PBS and cell medium with 10 µg/ml gentamicin was added . Gentamicin was removed from the medium at 2 hpi and cells were washed once per hour with fresh media . At 8 h pi , live cells were washed twice with HBSS+30 mM HEPES , and then stained using 1∶50 rhodamine-phalloidin ( Invitrogen ) in HBSS+30 mM HEPES for 15 minutes . Coverslips were rinsed using HBSS+30 mM HEPES and fixed in 4% paraformaldehyde . After fixation , coverslips were rinsed three times in TBS+0 . 1% Triton-X 100 and counterstained with DAPI . Coverslips were mounted onto slides using Prolong Anti-Fade ( Invitrogen ) , and imaged at the Center for Live Cell Imaging ( CLCI ) at the University of Michigan Medical School using an Olympus BX60 upright fluorescence microscope ( Olympus; Center Valley PA ) . Images were collected using a DP70 CCD color camera ( RGB , 12-bits/channel; Olympus America Inc . , Center Valley PA ) using DP70 controller/manager software v3 . 02 . Automated image analysis was performed using MetaMorph software ( Molecular Devices Sunnyvale , CA ) . For purification of recombinant LLO , 10 mL LB cultures containing 50 µg/ml kanamycin ( LB Kan50 ) sulfate were inoculated with single colonies of E . coli BL21 ( DE3 ) containing plasmid pET29 that encodes for a 6xHis-tagged copy of the hly gene encoding Listeriolysin O ( LLO ) from L . monocytogenes strain 10403S from freshly streaked LB agar plates and incubated overnight at 37°C with constant agitation . Cultures were then used to inoculate 100 ml of LB Kan50 and incubated at 30°C with constant agitation for 30 minutes . IPTG was added to cultures at a final concentration of 1 mM and incubation was resumed for an additional 18 h to induce LLO expression . Protein expression cultures were pelleted at 4000×g for 15 min at 4°C and supernatants were discarded . Pellets were resuspended in 1 ml lysis buffer ( 50 mM sodium phosphate dibasic , 300 mM sodium chloride , 10 mM imidazole; pH 8 . 0 ) containing 1 mg/ml lysozyme and incubated on ice for 30 minutes . Suspensions were then subjected to 4×30 sec sonication treatments separated by 15 sec incubation on ice with a Misonix Microson Ultrasonic Cell Disruptor XL set to intensity 4 . Lysates were centrifuged for 30 minutes at 16 , 000×g at 4°C . To purify 6xHis-tagged LLO , a NiNTA spin column ( Qiagen , Cat . No . 31314 ) was loaded with 600 µl wash buffer ( 50 mM sodium phosphate dibasic , 300 mM sodium chloride , 20 mM imidazole; pH 8 . 0 ) and centrifuged at 700×g for 2 minutes; flow-through was discarded . All subsequent NiNTA spin column centrifugation steps were carried out at 700×g for 2 minutes . Lysate supernatants were passed through the column in 600 µL increments . Columns were then washed sequentially with 600 µl volumes of the following: wash buffer , 84% wash buffer/16% glycerol v/v , and wash buffer containing 700 mM NaCl . 6XHis-tagged LLO was then eluted from the column with two 200 µl volumes of wash buffer containing 400 mM imidazole . Combined elution volumes were then passed through an Amicon Ultra 3000 MWCO filter unit by centrifugation for 30 minutes at 16 , 000×g at 4°C , after which 450 µl HBSE ( 10 mM HEPES , 140 mM NaCl , 1 mM EDTA; pH 8 . 4 ) buffer was added to the column , briefly agitated , and collected by flipping the column and centrifuging the solution into a microcentrifuge tube at 16 , 000×g for 10 seconds . Protein solutions were then separated into 50 µl aliquots and stored at −80°C . Final protein concentration was measured by Bradford assay ( Thermo Scientific ) using a BSA standard curve . All p values were generated between identified samples using unpaired two-tailed t-tests and represent analysis of ≥3 replicates per condition . * P<0 . 05 , ** P<0 . 01 and *** P<0 . 001 . | Macrophages are critical early responders recruited to sites of bacterial infection . Many intracellular bacterial pathogens subvert or bypass macrophage anti-microbial defenses by expression of virulence factors and toxins . The Gram-positive intracellular pathogen , Listeria monocytogenes , secretes a pore-forming toxin that can damage host membranes . We found that toxin-mediated damage during bacterial infection triggered activation of a cysteine protease , caspase-7 . Activation of this protease was triggered by low levels of toxin alone , suggesting that sublethal toxin-mediated damage is sensed by the host cell . Infected viable macrophages lacking caspase-7 exhibited increased membrane permeability , suggesting that caspase-7 limits membrane damage in infected host cells . This study reveals an unanticipated cytoprotective role for caspase-7 during intracellular infection and provides evidence for a mechanism by which host cells can initiate an adaptive response to survive membrane-damaging toxins , which are commonly associated with pathogenic microbes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"biology",
"microbiology"
] | 2012 | Membrane Damage during Listeria monocytogenes Infection Triggers a Caspase-7 Dependent Cytoprotective Response |
Nested effects models have been used successfully for learning subcellular networks from high-dimensional perturbation effects that result from RNA interference ( RNAi ) experiments . Here , we further develop the basic nested effects model using high-content single-cell imaging data from RNAi screens of cultured cells infected with human rhinovirus . RNAi screens with single-cell readouts are becoming increasingly common , and they often reveal high cell-to-cell variation . As a consequence of this cellular heterogeneity , knock-downs result in variable effects among cells and lead to weak average phenotypes on the cell population level . To address this confounding factor in network inference , we explicitly model the stimulation status of a signaling pathway in individual cells . We extend the framework of nested effects models to probabilistic combinatorial knock-downs and propose NEMix , a nested effects mixture model that accounts for unobserved pathway activation . We analyzed the identifiability of NEMix and developed a parameter inference scheme based on the Expectation Maximization algorithm . In an extensive simulation study , we show that NEMix improves learning of pathway structures over classical NEMs significantly in the presence of hidden pathway stimulation . We applied our model to single-cell imaging data from RNAi screens monitoring human rhinovirus infection , where limited infection efficiency of the assay results in uncertain pathway stimulation . Using a subset of genes with known interactions , we show that the inferred NEMix network has high accuracy and outperforms the classical nested effects model without hidden pathway activity . NEMix is implemented as part of the R/Bioconductor package ‘nem’ and available at www . cbg . ethz . ch/software/NEMix .
Network inference benefits substantially from perturbation experiments , such as RNA interference ( RNAi ) screens . Monitoring high-dimensional effects of gene silencing enables inference of non-transcriptional network structures that cannot be learned on observational data alone [1] . Nested effects models ( NEMs ) are a class of probabilistic graphical models that aim at learning hierarchical dependencies from such intervention experiments . Upon perturbing nodes in a signaling graph , their connectivity is inferred from the nested structure of observed downstream effects . The concept was first introduced in [2] . Since then , many further additions concerning , for example , parameter inference , structure learning , and data integration , were developed [3 , 4] . In addition , dynamic models for time series data have been developed [5–7] . In [5] , a first application of dynamic nested effects models to time laps microscopy data has been described , but the model can not handle single-cell data . A Bayesian network representation of NEMs in [8] introduces a probabilistic notation for signal propagation , but in practice the signaling is kept deterministic . In all previous NEM models and applications , the signaling pathway under observation is assumed to be active and the signal flow disrupted by silencing the signaling genes one by one . In principle , RNAi experiments are a highly informative for learning NEMs . Perturbations are introduced by gene silencing in cells through RNA interference using siRNAs [9 , 10] . Effects of the knock-downs are then captured by high-dimensional down-stream observations . The screening data analyzed here , comprises imaging data of thousands of individual cells for genome-wide gene silencing . However , the experiments come at the cost of high noise levels , as well as biological and technical biases , including off-target effects [11 , 12] . These confounding factors complicate the analysis and interpretation of the screening results . On the other hand , RNAi screens currently reach very high resolution . Per knock-down , the present data sets comprise about 300 image features for several hundred individual cells , which allows for a very detailed analysis of a knock-down event . However , it has been shown that measurements from individual cells of the same experiment can differ widely , for example , due to local environmental differences [13 , 14] . Such variation on the single cell level needs to be accounted for . Otherwise , an ambiguous signal is obtained , when averaging over the cell population of a knock-down . Here , we specifically investigate single-cell observations of pathogen infection screens [15–17] . The experiments monitor cells with an siRNA knock-down during infection with human rhinovirus ( HRV ) . After siRNA knock-down , the pathogen is added to the cells , and the success of infection as well as many other cellular features are extracted from microscopy images taken of the cells from each experiment [18–20] . The aim is to infer a signaling cascade involved in pathogen entry in to the host cell . However , a challenge in the analysis of data from this experimental setup is that by experimental design even in mock controls ( i . e . , infection without knock-down ) the infection rate is far from complete . In fact , the multiplicity of infection ( MOI ) of the assay was optimized to reach 30 to 50% infected cells , such that both infection-decreasing and infection-increasing hits can be detected . Which cells in the population finally get infected is , at least to some extent , the result of stochastic effects , since cellular processes can be differently manifested in different cells . The multi-functional nature of proteins , for instance , enables a single host factor to enhance a signaling cascade , and at the same time may antagonize other processes that support or inhibit infection . Obviously , infected cells were reached by a pathogen triggering some signal to get internalized . However , for uninfected cells , it is unknown whether a pathogen actually attempted to infect them , which is crucial for determining the effect that the gene knock-down had on these cells . Wrongly assuming that the pathway is active , even though it is not , can result in conflicting knock-down schemes . In the original NEM setting , individual cell observations are summarized for each signaling gene . To address the problem of network learning when the activation state of the signaling pathway is unknown we introduce a new model , called NEMix , extending the existing NEM framework in several ways . First , we do not summarize the data across cells , but rather perform network inference using the single-cell observations directly . Furthermore , we model the unknown pathway activation with an additional hidden random variable in the graph of signaling genes . The activation state is then estimated for each individual cell . The pathway activity can be regarded as an additional hidden silencing event in the signaling graph . We introduce a general theoretical framework for probabilistic combinatorial knock-downs in NEMs . We develop our model for the most general case , not making any assumptions about the signal propagation . We have implemented the special case of one hidden variable with probabilistic knock-down , where the remaining network is kept deterministic . For inference of the hidden pathway state , we developed an EM algorithm [21] . This step is repeated for each proposal structure during the network search .
We developed NEMix , a new model based on NEMs , which allows to estimate activity of a pathway in individual cells . A NEM is a graphical model , consisting of two graphs . The transitively closed graph Φ encodes dependencies among signaling gene nodes Ss ∊ 𝓢 , which are silenced one by one . The bipartite graph Θ connects a set of observable feature nodes Ee ∊ 𝓔 uniquely to the signaling genes ( Fig . 1A ) . We seek the structure of Φ , i . e . , the topology of the signaling pathway , by inferring it from the nested structure of observed effects . For a data set 𝒟 = ( dek ) of a set of knock-down experiments k ∊ {1 , … , K} and observed features e ∊ {1 , … , m} , the likelihood function given Φ and θ is P ( 𝒟 ∣ Φ , θ ) = ∏ e = 1 m ∏ k = 1 K P ( d e k ∣ Φ , θ e = s ) , ( 1 ) where θe = s indicates that feature e is connected to signaling gene s ∊ 𝓢 . The NEMix model consists of the same two graphs Φ and Θ , but has an additional binary hidden variable Z added to the signaling graph Φ . Its connections to the signaling genes , as well as its overall knock-down probability p0 = P ( Zkc = 0 ) , are unknown and inferred for each individual cell during the network reconstruction process . Given single cell data 𝒟 = ( dekc ) with c = 1 , … , ck cells in knock-down experiment k , the likelihood function of the NEMix model , given Φ and θ , is P ( D ∣ Φ , θ ) = ∏ e = 1 m ∏ k = 1 K ∏ c = 1 c k ∑ j ∊ { 0 , 1 } p j P ( d e k c ∣ Φ , θ e = s , Z k c = j ) . ( 2 ) A detailed derivation of the model and its implementation are given in the Models section . If a signal is activating a pathway , or parts of it , the signal flow is the same as in the NEM . Also the observed knock-down effects for the features Ee are the same . However , when the pathways input signal is inactivated , the knock-down pattern of the features changes ( Fig . 1A and B , cells 7 to 15 ) . Not accounting for the pathway disruption can mislead inference of the structure Φ ( Fig . 1A , left model ) . The connectivity of Z is learned in a greedy fashion during structure inference . For the knock-down probability of the hidden variable , p0 , we implemented an EM algorithm , which estimates jointly p0 from each cell’s observation and the connections of observations to signaling genes , θ . In the following , we show improved network inference with NEMix in simulations and then infer networks of high accuracy , from single cell gene silencing experiments . To test our model , we performed a large simulation study . We generated 30 network structures with 5 signaling genes , randomly sampled from KEGG pathway maps [22] as previously described in [6] . To each network the hidden input signal was attached randomly . The resulting 30 sample networks are shown in supplementary S1 Fig . From each network , we sampled 50 data sets on 300 observed features in the following way . For each gene , we simulated single knock-downs in 200 cells . To the observed features we added another 30 noise features , not attached to any signaling gene . The data sets were generated in the following way . We sampled effects from a normal distribution with mean me = 1 and non-effects from a normal distribution with mean mn = 0 . The standard deviation for each experiment was sampled uniformly between 2 and 2 . 5 . We furthermore sampled 200 cells for control experiments . The negative control cells do not show any effects and are therefore drawn from the non-effect distribution . The positive control cells always show effects and hence are drawn from the effect distribution . The whole simulation process was repeated for five different fractions of pathway disruption , p0 ∊ {0 , 0 . 3 , 0 . 5 , 0 . 8 , 1} . NEMix inference was restarted for 16 initial networks . Each of them consists of the empty graph Φ plus a unique attachment of Z to the signaling genes . Setting the maximal out-degree of Z to two , there are 16 possible such attachments of Z . This regularization on the edges of Z reduces the search space significantly . During structure search we also imposed this restriction , but additionally allowed transitive edges that had to be added as a consequence of the insertion of any edge connecting Z to a gene ( see Models section ) . We compared NEMix to two other NEM models and , for a baseline comparison , to a random approach , where network edges are sampled uniformly with probability 1/n , where n∣Φ∣ is the number of signaling nodes . This probability was chosen as it creates networks with approximately the same number of edges as in the original graphs . To assess the impact that pathway disruption has on the cell population level , we ran the simulations on a standard NEM using the log-likelihood model introduced in [23] . For the NEM approach we had to summarize the single cell observations to the gene level . For these gene-level data sets we used p-values of a Wilcoxon test comparing the cell population of a knock-down to the control distribution . From the p-value distributions a Beta-Uniform-Mixture model was estimated . For each feature a density value is calculated from this model , indicating the effect strength of the knock-down . These density values are used as the input data , as previously introduced in [23] . The third approach , called single-cell NEM ( sc-NEM ) . is a NEMix model on individual cell observations , but with fixed p0 = 0 , i . e . , a single-cell observation-based NEM without considering uncertain pathway activity . For all three models , we applied a uniform prior on the feature attachments θ , and no prior knowledge was added for the network structures Φ . The NEMix parameter p0 was initialized by drawing from a uniform distribution in each EM restart . As NEMix and sc-NEMs infer networks on single-cell observations , we calculated log odds ratios from each observation based on the positive and negative control distributions ( see ‘Modeling the effect likelihoods’ in S1 Text ) . For NEMs and sc-NEMs , we used maximum likelihood estimation to infer θ and in the NEMix it is estimated by in an EM algorithm . Structure learning is performed using a greedy hill climbing algorithm , initialized with an empty network . Fig . 2A summarizes the overall performance for all methods and the different fractions of pathway signal perturbation p0 . We display accuracy of the edge recovery , for varying p0 . We also calculated the area under the ROC curve ( AUC ) based on the edge frequencies of the 50 replicate data sets , which yielded similar results in terms of accuracy ( see supplementary S2 Fig ) . As expected , all methods performed equally well when there is no signal disruption ( p0 = 0 ) . However , when p0 is moderate to high , NEMix performs significantly better than the other methods . If the triggering signal is always turned off , performance of all methods drops drastically . Intuitively , this is because in such a special case , all features downstream of Z always show an effect and hence they cannot be used for structure learning . For example , if , in Fig . 1B , Z is inactive for each cell , we could not infer the structure among S2 and S3 . In reality though , permanent shut down of the pathway is very unlikely . For the infection screens p0 = 1 would mean that no cell is ever infected . Pathway activity estimates are also of overall high accuracy ( Fig . 2B ) . Although simulation results demonstrate that the performance of learning Z and θ varies , depending on the network structure , the average performance is very good ( S3 Fig , S4 Fig , S5 Fig ) . Currently , one of the main obstacles for learning larger NEMix models is the fast growing run-time for n > 5 network nodes . Run-time is further increased by a factor of n , when initiating the algorithm with each possible connection of Z to one of the knock-down genes . To assess its performance on larger networks , we ran a reduced simulation study on n = 5 , 10 , and 15 genes . The setup and results of the study are described in detail in S6 Fig . Larger networks of 15 nodes can still be estimated very well ( S6 Fig . A ) and estimation of the parameter p0 even improves ( S1 Fig . D ) . However , the average time to estimate a 15-node network was 9 . 5 hours . This is substantially more than the average 1 . 9 hours needed for 10-node networks . Thus , in a highly parallelized computing environment , even larger networks can be estimated . We also assessed the connection of features to the signaling genes in the inferred graph Θ . There can be situations , where attachment of features is equally likely for several signaling genes . In these cases , where no single gene is preferred , we counted a feature as correctly attached if it was connected to any of the signaling genes with equal likelihood . Accuracy of the θ estimates is high ( > 80% ) for small p0 values and decreases with increasing p0 . For small p0 , also performance of the sc-NEMs is good , which shows the advantage of learning on the single-cell data level . However , NEMix stands out from the other methods for higher p0 . Recovery of noise features , i . e . , correct filtering of the additionally added uninformative features , is not strongly affected by the hidden signal ( see supplementary S7 Fig ) . Analyzing individual networks , one again observes high variation in performance ( see supplementary S8 Fig ) . We applied NEMix in the context of infection signaling , using the RNAi screening data monitoring HRV infection , mentioned in the introduction . Briefly , viruses were added to the siRNA transfected cells and after an incubation time , cells were fixated , stained , and then imaged . Subsequently , 360 cell features were extracted from the 9 images per knock-down experiment using the software CellProfiler [24] . For the whole experimental procedure the protocols of [17] were followed . The HRV assay is rather short with an infection time of only seven hours , resulting in measurements proximal to the infection event . The short time range is advantageous , because it leaves less room for confounding developments in the cells . Furthermore , the used antibody resulted in clean readouts , well to extract from the images . Before using the data for network inference , we performed two additional filtering steps . For each knock-down , the well is split into 9 images . They are arranged in three rows and three columns . We used only the middle image , because it is of the highest quality . In this way we avoided too many out-of-focus cells , which bias especially the cell texture features . After this filtering step , we had around 200 to 300 cells per knock-down . A second filtering step concerns siRNA off-targets [25] . We sought to avoid confounding by this effect and therefore selected only genes with low predicted off-target effects as described in ‘siRNA filtering for off-targets’ of S1 Text . We applied NEMix to a small subset of the screened genes , in order to recover a known pathway . We decided on the well-known MAP-Kinase signaling cascade as a proof of principle , for several reasons . First , it has been studied and validated in great detail [26–28] , such that the available signaling network from the KEGG database [22] can be used as a reliable source to compare to . Second , the pathway is known to be involved in HVR infection signaling , where it is associated with asthmatic and COPD exacerbation [29–31] . Finally , we observed an enrichment for low off-target siRNAs in this pathway when performing a gene set enrichment analysis [32] ( see supplementary S9 Fig ) . We then selected a small subset of 8 MAP-Kinase pathway genes for analysis based on the derived score for predicted off-target effects . Nodes of KEGG pathways can contain several genes . We selected genes such that they are all assigned to different KEGG nodes using a weighted maximum bipartite matching of low off-target siRNAs and unique KEGG nodes . After gene selection , we inferred networks for the 5 and 8 genes with lowest off-target score . Like in the simulation study above , we compared the NEMix model to the NEM and the sc-NEM approach . As input data sets , the local effect likelihoods from the selected knock-down gene experiments were computed as follows . As the experiments lack reliable controls , we instead used a random sample of cells from the plate on which the gene was located , assuming that the majority of knock-downs will not have an effect . Like for the simulation study , we derived the cell population effects for the NEM from Wilcoxon tests , comparing the knock-down experiment to the control . From the resulting p-value distributions , effect strengths for the features were estimated using the Beta-Uniform-Mixture model . Log odds ratios for sc-NEMs and NEMix in this case are calculated only based on one control distribution ( see Models section ) . NEMix inference again is repeated for the 16 initial networks of all possible connections of Z with maximal out-degree 2 to the empty graph Φ . Like in the simulation study , p0 was initialized by drawing randomly from a uniform distribution . Again we used uniform priors for θ and imposed no priors for the signaling networks other than the maximal out-degree of Z ( plus the transitive edges that need to be added ) . The known KEGG network and the inferred results for the top 5 signaling genes are displayed in Fig . 3A-E . Results for the top-8 gene network are given in S10 Fig . To assess robustness of the learned networks , we repeated the inference on 50 bootstrap samples of the original data set . Both networks show high AUC values and even better accuracy ( see Table 1 ) . As can be seen from Fig . 3F , network inference was very robust for the top-5 gene network . For the top-8 gene network , performance had a slightly higher variation . Individual plots for sensitivity and specificity are given in supplementary S11 Fig . A , B . Also the estimate of p0 shows only little variation ( S11 Fig . C ) . In all cases , the likelihood score of the known KEGG network is much lower than for the best inferred networks , indicating that under the assumptions of our model , the data and the KEGG database do not perfectly agree . Possible reasons for this observation include our model missing to explain part of the data correctly , the KEGG database being incomplete , and inaccuracies in the data generating process . Nevertheless , the accuracy value of 0 . 85 for the learned NEMix outperforms all other methods . All edges contained in the learned NEMix models are of high robustness ( > 80% for 5 genes , and > 70% for 8 genes ) . Consensus networks of the bootstrap results are shown in supplementary S12 Fig . Furthermore , the hidden root Z is attached to the same nodes in both the known KEGG graph and the estimated network for 5 genes . Also the inferred 8 node network connects Z to the same three genes . As genes were selected based on small off-target effects of their targeting siRNAs , they are not necessarily hits for HRV infection . However , of the selected genes EGFR [33] , TAB2 [34] and CACNA2D3 [35] have been shown to be involved in this process . All models have a built-in filter for uninformative features , which has been previously introduced in [36] . A comparison shows that averaged over the bootstrap samples , for all three methods , the set of used features largely agrees ( supplementary S13 Fig and S14 Fig ) . The maximum likelihood attachments of features to the knock-down genes and the null node are shown in supplementary S15 Fig and S16 Fig , together with a detailed description of the different feature types . The inferred signaling disruption of p0 = 0 . 42 seems rather high . We compared this to the average infection rate in mock experiments , i . e . , cells without siRNA knock-down . These resemble cases , where Z can be perturbed but none of the other signaling genes in the network . Mock wells from plates of the 8 genes used here , actually have a much higher percentage of uninfected cells , roughly in the range of 75 to 81% . However , this comparison should be taken with caution since control wells of these screens might have suffered from strong plate location bias , as they were located on the margins of the plate . As a general observation , NEMix-inferred networks were sparser than those obtained from NEMs , because spurious edges introduced in the latter are correctly explained by hidden pathway activity Z in NEMix . Therefore , NEMix networks have increased specificity , which might come at the cost of some missing true edges . Especially the 8-gene networks inferred by NEM and sc-NEM are much denser than the known KEGG network . A sparse network is beneficial in the sense that it allows to focus on a small set of highly specific edges . For validation experiments , it is desirable to have a low false positive rate in the predicted interactions as usually only very few of these dependencies can be experimentally tested .
RNAi screens are known to be prone to many sources of noise and bias such that their analysis is highly challenging . Here , we have identified one confounding factor , namely heterogeneous signaling pathway activation within a cell population , and incorporated it directly into a novel probabilistic model for pathway reconstruction . To address the problem of unknown activation of signaling pathways during network inference , we have introduced a general framework , building on NEMs , to handle hidden combinatorial knock-downs in a probabilistic manner . With NEMix we provide an implementation for inference under unknown pathway stimulation . For the first time , image features are explicitly used on the single-cell level for NEM inference , acknowledging large cell-to-cell variation . We have demonstrated the advantages of NEMix over current NEMs in simulations and inferred highly accurate networks in a case study on HRV infection . Especially , when the underlying true signaling networks are expected to be sparse , NEMix is beneficial . It removes spurious edges introduced due to confounding factors and therefore reduces the false positive rate , a desired property when it comes to validation of edges . A limitation of the current model formulation is the assumption of independent single cell observations . In reality , this assumption might not be met as cells can be biased due to their location and neighbors . Removing this bias either by normalization or explicit modeling , as for example in [14] , could further improve the model . Furthermore , in the current data sets cells can be in different cell cycle states . Grouping them according their states may remove further biases , but this clustering task is itself very challenging . Another general limitation of NEMs and NEMix models is that they cannot learn certain pathway features . From static data , NEMs cannot resolve any loop structures by construction . This is a general problem for network inference without time resolved data . Therefore , only performance statements based on comparing transitively closed pathways can be made . The sampled graphs in the simulation are already transitively closed and since the transitive closure is a feature inherent to all the models we compare , it should not influence the ranking based on their performance . Before comparing a network to the corresponding KEGG pathway , we also built its transitive closure . This fact should be considered when interpreting the inferred models . For example , the model does not allow for distinguishing a feed forward loop from a sequential cascade; however , the hierarchical order of genes in the network would remain the same , and this piece of information does already provide considerable insight into the biological processes . The way we have assessed performance here puts particular emphasis on this hierarchical structure of the network nodes . Further improvements could be achieved during data preparation . Image segmentation is not always perfect and might introduce technical biases into data sets , adding more confounding factors . If data is not curated carefully , we risk to capture technical biases with the additional hidden variable in NEMix models . Another interesting aspect of the data sets deserving a more thorough analysis , is the nature of the image features themselves . Here , readouts have been used to infer the graph of signaling genes . However , one could investigate in more detail how features are grouped when attaching them to the signaling genes . Some features might not contribute useful information and could be filtered in advance , others might be redundant . Future projects could use the output of NEMix models and seek for biological interpretation of feature correlations . In case of cell infection screens , infection efficacy was an obvious factor that needed to be addressed . However , the same idea could be applied to other sources of noise . For example , transfection efficacy of the knock-downs could be considered . Quality and efficacy of a knock-down can be quantified by mRNA levels ( qPCR ) or protein level ( western blot analysis ) of a gene . However , for high-throughput assays , such confirmation is not available for most gene knock-downs . In order to account for different siRNA transfection efficacies further hidden variables could be introduced . In contrast to the global Z variable introduced here , hidden knock-down rates would then be estimated for each gene individually . As a consequence , the complexity of the problem would increase substantially . Instead of one parameter , n ( number of genes ) parameters would have to be estimated . Furthermore , knock-down probabilities could only be estimated from a fraction of the observations ( e . g . , cells under the specific knock-down ) . Another drawback is that the increased number of hidden variables gives rise to identifiability problems when estimating infection efficacy in combination with the knock-down rates . For example , if the hidden variable Z was only attached to one signaling gene , effects of Z and a failed transfection could not be distinguished . Although extending the NEMix model to this situation would be an interesting future project , we believe that problems in the transfection process play an overall minor role . For the current experiments , KIF11 siRNAs ( cell killers ) were used to control transfection quality on the plate level . For the plates containing the cells used in our analysis , these controls show very high penetrance , i . e . , out of an average of 2000 cells per well , on average only 7% of cells survive in these wells . Although this test does not make a statement about the efficacy of individual siRNAs , it ensures the general functioning of the transfection process . Additionally , the library vendor claims the knock-down efficacies achieved with their smart-pool siRNAs to be in the range of 70–95% . This proportion is a result of many possible sources of imperfect gene silencing , including non-transfected cells and off-target effects . Given the above facts in combination with our off-target filtering strategy , we are convinced that the analyzed data are of high quality . We tried to minimize the general problem of confounding siRNA off-targets by considering only genes targeted by siRNAs with low predicted off-target effects . This selection step helps to achieve reasonably unbiased results with our model , but it also limits the gene sets we can analyze . Ideally , we want to be able to select any gene of interest . This scenario calls for models that can correct the off-target effects on the single-cell level . A potential solution to this issue could be delivered by NEMs directly . We could still learn the networks based on siRNA knock-downs directly , but handle the signal propagation differently . With NEMix it is already possible to use each siRNA as a combinatorial knock-down . In reality however , individual genes are knocked-down to different degrees by an siRNA . In a NEM , this would mean to split up the silencing signal of an siRNA into partial knock-downs of several genes . Then , signal propagation would have to be formulated in a fully probabilistic fashion and NEMs would have to be reformulated such that their nodes do not have binary states anymore . Further developing NEMix , by integrating the above mentioned shortcomings , will make the models more powerful for future network reconstruction tasks . Especially in the light of single cell data sets , which show large heterogeneity among individual observations , our approach is beneficial . Such data sets are becoming more and more available , and they reveal that the high cell-to-cell variation has severe consequences when summarizing such heterogeneous observations . On the population level , the signal is potentially confounded as it is only contained in part of the observations . NEMix uses the full power of single-cell experiments , as it is applied on the single-cell level directly , avoiding any data averaging . Only at this data resolution , the heterogeneity within a cell population can be accounted for and it becomes possible to investigate potentially confounding factors , such as , for example , pathway activity . NEMix is the first NEM-based method with additional unknown components in the signaling graph Φ . It is capable of inferring these missing data and provides an estimate for the fraction of signal disruption . We find such ambiguous signaling in RNAi infection screens and we have demonstrated that NEMix can improve network inference substantially by accounting for the confounding factor .
A NEM , as introduced in [2] , aims to infer the hidden dependency structure among a set of n binary signaling variables 𝓢 from the nested structure of m observed effect variables 𝓔 ( features ) . It therefore consists of two directed graphs , one describing the dependencies among the signaling genes and one connecting the features to the genes . The binary adjacency matrix of signaling genes is denoted Φ = ( ϕks ) , with ϕks = 1 if gene k propagates its effects to gene s and using the convention Φk , k = 1 , for all k . The signaling graph Φ is thus always transitively closed . If a gene is silenced , the effect is propagated deterministically along the edges of Φ . The connection of features 𝓔 to the genes 𝓢 is given by parameters θe , where θe = s indicates that feature e is linked to gene s . For a gene k and a feature e , a NEM predicts an effect of k on e if there is a gene s such that ϕks = 1 ( i . e . , k and s are connected ) , and θe = s ( i . e . , s has an effect on e ) . The observed data are denoted D = ( dek ) , where each dek is the measurement of feature e under perturbation of k ( Fig . 1A ) . Given an external signal which affects one or more of the signaling genes , each of them will have a binary signaling state . The state value is 0 if the signaling is interrupted , i . e . , does not reach the node , and 1 if the signal reaches the node , i . e . , the natural state of a stimulated pathway . For inferring the structure Φ among the signaling genes , we consider its posterior P ( Φ ∣ D ) = P ( D ∣ Φ ) P ( Φ ) P ( D ) , ( 3 ) where the marginal likelihood P ( D∣Φ ) can be obtained by integrating out the connections of features to the genes , P ( D ∣ Φ ) = ∫ θ P ( D ∣ Φ , θ ) P ( θ ∣ Φ ) d θ , ( 4 ) with prior distribution P ( θ∣Φ ) . In the absence of further knowledge , the prior is usually set to the uniform distribution . Given the network structure and assuming conditional independence of the parameters θe and of the silencing experiments k , the marginal likelihood becomes P ( D ∣ Φ ) = ∏ e = 1 m ∑ s = 1 n ∏ k = 1 K P ( d e k ∣ Φ , θ e = s ) P ( θ e = s ) . ( 5 ) The local effect likelihoods P ( dek∣Φ , θ ) denote the probability of observing an effect in feature e under knock-down of gene k . They can usually be pre-computed from the data and different approaches have been proposed [2 , 23 , 36] . For the results presented below , log-odds ratios as introduced in [36] were used ( see ‘Modeling the effect likelihoods’ in S1 Text for details ) . We first define the NEMix model and then derive it in detail . A NEMix consists of a nested effects model with effects graph Θ and an extended signaling graph Φ . The signaling graph Φ describes the dependency structure among the signaling genes and has an additional binary hidden variable Z indicating pathway activity . Z is a root of Φ , i . e . , it can be connected to any of its nodes and does not have any direct connections to features in θ . The silencing probability of Z is denoted by p0 and is a priory not known . For a set knock-down experiments k ∊ {1 , … , K} , with single cell observations c ∊ {1 , … , ck} of signaling gens s ∊ {1 , … , n} and features e ∊ {1 , … , m} , the marginal likelihood of a NEMix is P ( D ∣ Φ ) = ∏ e = 1 m ∑ s = 1 n P ( θ e = s ) ∏ k = 1 K ∏ c = 1 c k ∑ j ∊ { 0 , 1 } p j P ( d e k c ∣ Φ , θ e = s , Z k c = j ) , ( 6 ) where pj = P ( Zk = j ) . Structure learning is performed using a greedy heuristic to find an optimal network . Similar to the NEM procedure described in [3] , edges are incrementally added if the likelihood is increased ( see ‘Structure learning’ in S1 Text ) . In addition , our approach is restricted to structures without incoming edges into the hidden root Z . We initialize the algorithm with a set of initial networks . These consist of the empty graph and one edge connecting Z to one of the knock-down genes . Additionally , we limit the out-degree of Z to two . Here , by out-degree we mean only the non-transitive edges . We still allow the insertion of transitive edges from Z to any signaling gene , which has to be added in order to fulfill the transitivity requirement . This regularization reduces the search space and prevents that too many dependencies between genes are explained by Z alone . As for classic NEMs , network structure scoring involves the marginal likelihood . For the NEMix model , P ( 𝒟∣Φ ) cannot be optimized analytically . Marginalization over the feature attachments is omitted in our extended model . Instead , we estimate θ jointly with p during model inference . To do so , we approximate the marginal likelihood ( 10 ) by the expectation of the complete data log-likelihood P ( D , Z ∣ Φ , θ , p ) = ∏ k = 1 K ∏ c = 1 c k ∏ j ∊ { 0 , 1 } p j ∏ e = 1 m P ( d e k c ∣ Φ , θ e = s , Z k c = j ) Z k c ( j ) , ( 20 ) with respect to Z , where θ and p0 need to be efficiently estimated . For this task we have developed an EM algorithm . A derivation of the expected hidden log-likelihood and the maximum likelihood estimates is given in ‘Estimating the hidden signal’ of S1 Text . When starting the EM algorithm , p0 is initialized with a random draw from the uniform distribution and for θ we use a uniform initial configuration . The NEMix model is included as part of the R/Bioconductor package NEM as an additional inference type . It is invoked by calling the package’s main function NEM ( data , inference = ‘NEM . greedy’ , control ) and choosing the inference type control$type = ‘NEMix’ . ( See ‘NEMix implementation in NEM package’ in S1 Text for more detailed instructions on the implementation and usage of NEMix in R ) . To record run-times of NEMix model estimation , simulations were run without any parallelization on a 1 . 7GHz Intel i7 machine . Only one starting configuration was used , and EM iterations were performed using three restarts to avoid local optima that are globally suboptimal . For realistic data sets of 300 features and 200 cells per knock-down , NEMix estimation took on average nine minutes for 5-gene networks , with an average of 13 iteration steps until convergence of the EM algorithm . For the 8-gene network , the average run-time was 66 minutes , while the average number of iterations per EM round remained 13 also for these larger networks . The longer run-times of NEMix models as compared to NEMs are primarily due to the hidden data estimation . Each structure scored once in a NEM inference , needs to be scored 40 times on average during NEMix estimation . In addition , the input data sets are roughly 200 times larger . | Experiments monitoring individual cells show that cells can behave differently even under same experimental conditions . Summarizing measurements over a population of cells can lead to weak and widely deviating signals , and subsequently applied modeling approaches , like network inference , will suffer from this information loss . Nested effects models , a method tailored to reconstruct signaling networks from high-dimensional read-outs of gene silencing experiments , have so far been only applied on the cell population level . These models assume the pathway under consideration to be activated in all cells . The signal flow is only disrupted , when genes are silenced . However , if this assumption is not met , inference results can be incorrect , because observed effects are interpreted wrongly . We extended nested effects models , to use the power of single-cell resolution data sets . We introduce a new unobserved factor , which describes the pathway activity of single cells . The pathway activity is learned for each cell during network inference . We apply our model to gene silencing screens , investigating human rhino virus infection of single cells from microscopy imaging features . Comparing the learned network to the known KEGG pathway of the genes shows that our method recovers networks significantly better than classical nested effects models without capturing of hidden signaling . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Models"
] | [] | 2015 | NEMix: Single-cell Nested Effects Models for Probabilistic Pathway Stimulation |
Stochastic chemical reaction networks constitute a model class to quantitatively describe dynamics and cell-to-cell variability in biological systems . The topology of these networks typically is only partially characterized due to experimental limitations . Current approaches for refining network topology are based on the explicit enumeration of alternative topologies and are therefore restricted to small problem instances with almost complete knowledge . We propose the reactionet lasso , a computational procedure that derives a stepwise sparse regression approach on the basis of the Chemical Master Equation , enabling large-scale structure learning for reaction networks by implicitly accounting for billions of topology variants . We have assessed the structure learning capabilities of the reactionet lasso on synthetic data for the complete TRAIL induced apoptosis signaling cascade comprising 70 reactions . We find that the reactionet lasso is able to efficiently recover the structure of these reaction systems , ab initio , with high sensitivity and specificity . With only < 1% false discoveries , the reactionet lasso is able to recover 45% of all true reactions ab initio among > 6000 possible reactions and over 102000 network topologies . In conjunction with information rich single cell technologies such as single cell RNA sequencing or mass cytometry , the reactionet lasso will enable large-scale structure learning , particularly in areas with partial network structure knowledge , such as cancer biology , and thereby enable the detection of pathological alterations of reaction networks . We provide software to allow for wide applicability of the reactionet lasso .
Cellular processes are essentially implemented by networks of biochemical reactions . The topology of such networks is typically only partially known , rendering the identification of the correct network from experimental data a key challenge . Despite the importance of this task , only little progress has been made in devising methods to systematically and comprehensively infer topologies of non-trivial chemical reaction networks . In this work , we propose a sparse regression approach tailored to the task of large-scale model selection for chemical reaction networks . Different model classes have been developed to describe biochemical reaction systems . In order of increasing level of detail these comprise statistical time series models , such as autoregressive models and dynamic Bayesian networks , deterministic ordinary differential equation or stochastic differential equation based kinetic models [1] . The choice of model class depends on prior information for the system of interest and type of experimental data . Single cell technologies furnish further statistical information about component distributions , e . g . variances and covariances , aiding in systems identification [2] and are expected to become increasingly prevalent in routine biological research [3] . Two main computational tasks arise when learning any of these models from data: parameter inference , and structure learning . Parameter inference aims at finding model parameters ( e . g . kinetic rate constants ) . Parameter inference has been performed by sampling from posterior parameter distributions , or global non-convex or convex optimization methods [4] . Structure learning aims at additionally identifying the reaction network topology governing the dynamics of the system components . Parameter inference becomes increasingly computationally intensive for larger systems with numerous parameters [1] . Structure learning for these systems is an even more daunting task since parameter inference has to be performed for each of the possibly very many different system topologies . Therefore , structure learning is typically confined to comparison of a small , carefully selected set of candidate topologies by means of model selection criteria , such as information criteria ( e . g . AIC , BIC ) or Bayes Factors [5–7] . However , this approach requires substantial prior knowledge about the studied system in order to identify reasonable candidate models . Systematic approaches to enumerate a subset of sensible topologies have not been reported until recently . These approaches implement greedy strategies that either iteratively reduce the number of reactions of an overcomplete system of reactions or add reactions one at a time to a system with a minimal set of reactions [8] . However , such greedy approaches do not guarantee finding globally optimal topologies for non-convex fitting objectives . Furthermore , exploration of the multitude of local optima in the context of combinatorially many possible topologies becomes computationally prohibitive due to the requirement to explicitly evaluate every considered candidate topology . No global approaches have been reported to perform structure learning by comprehensively evaluating model candidates for stochastic chemical reaction networks . We propose the reactionet lasso , a convex relaxation of the structure learning task . This approach yields a single best sparse reaction set from all possible reactions by translating a recent sparse identification approach for nonlinear dynamic systems [9] to operate on and deal with non-trivial application specific parameter and noise structure for time series snapshot data acquired for stochastic chemical reaction networks .
Structure learning by the reactionet lasso takes advantage of the formal link between the chemical reaction model and the observed data that is defined by the Chemical Master Equation . This differential equation system describes the temporal evolution of the abundance distributions of species governed by a stochastic chemical reaction network [10] . The moment generating functions of the Chemical Master Equation give rise to the moment equations , a system of ordinary differential equations for the temporal evolution of the central moments Mr of the abundance distributions ( see Methods ) . M ˙ r = Σ l k l F r l ( t ; M ) , ( 1 ) with rate constants kl , time t and set of all central moments of individual species M . For mass action kinetics the terms Frl ( t; M ) are polynomials over these moments such as abundance means and variances of individual species . Frl ( t; M ) will be referred to as stoichiometric moment functions herein ( see also S1 Text ) . The moment equations constitute the formal link between the time series snapshot data and the rate constants of the underlying chemical reactions . Rate constant estimation for stochastic mass action kinetics reaction networks in this context therefore reduces to parameter estimation for the ordinary differential equation system [Eq 1] with stoichiometric moment functions determined from the time series data . Parameter estimation for a mass action kinetics network typically requires the costly integration of the moment equations for every considered parameter configuration . Imputation of the moment gradients by gradient matching procedures ( see Methods ) circumvents these type of evaluations and , in conjunction with the empirical moments , allows for parameter inference by means of a non-negative linear regression task with the least squares estimate k ^ for rate constants k given by: k ^ = arg min k ≥ 0 ∥ b ^ - A ^ k ∥ 2 2 , ( 2 ) where the response vector elements bj corresponds to the vector of empirical gradient estimates for M ˙ j ( t ) from the gradient matching procedure ( see Methods ) and the design matrix entries A ^ j l correspond to the estimates of the stoichiometric moment functions Fjl ( t; M ) : b = M ˙ ^ r 1 ( t 1 ) ⋮ M ˙ ^ r 1 ( t T ) M ˙ ^ r 2 ( t 1 ) ⋮ M ˙ ^ r 2 ( t T ) ⋮ M ˙ ^ r N ( t 1 ) ⋮ M ˙ ^ r N ( t T ) , A = F ^ r 1 1 ( t 1 ) F ^ r 1 2 ( t 1 ) ⋯ F ^ r 1 L ( t 1 ) ⋮ ⋮ ⋮ ⋮ F ^ r 1 1 ( t T ) F ^ r 1 2 ( t T ) ⋯ F ^ r 1 L ( t T ) F ^ r 2 1 ( t 1 ) F ^ r 2 2 ( t 1 ) ⋯ F ^ r 2 L ( t 1 ) ⋮ ⋮ ⋮ ⋮ F ^ r 2 1 ( t T ) F ^ r 2 2 ( t T ) ⋯ F ^ r 2 L ( t T ) ⋮ ⋮ ⋮ ⋮ F ^ r N 1 ( t 1 ) F ^ r N 2 ( t 1 ) ⋯ F ^ r N L ( t 1 ) ⋮ ⋮ ⋮ ⋮ F ^ r N 1 ( t T ) F ^ r N 2 ( t T ) ⋯ F ^ r N L ( t T ) . This linear regression formulation has been applied for parameter inference of deterministic chemical reaction models [6 , 11] . Model selection across small sets of model variants has previously been performed with information criteria [11] or model averaging [6] . The Lasso constitutes another approach for efficient and comprehensive model selection in linear regression models [12] . It introduces an L1 norm ( ‖ ⋅ ‖1 ) regularization on the parameters k to promote the identification of sparse solutions , i . e . solutions with many zero-valued parameter estimates . k ^ = arg min k ≥ 0 ∥ b ^ - A ^ k ∥ 2 2 + λ ∥ k ∥ 1 , ( 3 ) Various extensions of the Lasso method were introduced in literature to improve its shrinkage properties in the presence or absence of heteroscedasticity [13] . While the Lasso has been used in recent reports to identify general nonlinear dynamical systems [9] or to select the mechanism types ( mass action or Hill kinetics ) of a fixed reaction set defined by the deterministic Repressilator comprising six components [14] , it still remains to adapt the regression model and regularization concepts to enable more comprehensive model selection for realistic reaction systems that exhibit stochasticity and larger amount of components/reactions . The next sections will delineate in detail the challenges and solutions implemented in the reactionet lasso to achieve this goal . This section introduces the reactionet lasso ( Fig 1 ) , a computational method for learning the structure of chemical reaction networks . The overarching strategy of this procedure consists of ( 1 ) enumerating all ( or at least a significant fraction of reasonable ) conceivable unary/binary reactions between the components of a reaction system of interest and ( 2 ) applying an appropriate stepwise sparse regression approach to select the sparse subset of reactions underlying the observed dynamics in the snapshot time series data . The following properties of such structure learning instances preclude the application of conventional least squares based approaches for parameter estimation and selection: ( 1 ) noise and heteroscedasticity of the observed response b ^ ( empirical moment gradient estimates ) as well as in the observed design matrix A ^ ( stoichiometric moment function evaluations ) and ( 2 ) different scales of individual parameters ki ( rate constants ) resulting from the occurrence of large a spectrum of fast and slow reactions . The reactionet lasso addresses each of these challenges in as delineated in the following . The intrinsic variability of stochastic chemical kinetics result induces variability of the empirical estimates of moments and their gradients . Therefore the observed response vector as well as the stoichiometric moment functions in the design matrix are expected to deviate from the true latent correspondents . We capture this by defining b = b ^ + ϵ b and A = A ^ + ϵ A to be the true latent moment gradients and stoichiometric moment functions , and ϵA and ϵb to be their respective intrinsic variability induced deviations from the estimated/observed quantities . If we knew the true values of the latent variables , finding the rate constants k would translate to solving the following equation: b = A k . ( 4 ) By substituting the variables in eq 4 with the definitions for our empirical estimates of the latent variables we obtain: b ^ = A ^ k+ϵ . ( 5 ) with ϵ: = ϵA k − ϵb . Eq 5 seems to motivate a straightforward optimization strategy to compute a maximum likelihood parameter estimate given the observations for moment gradients and stoichiometric moment functions ( e . g . least squares for independent and normally distributed residuals ϵ ) . However , it becomes apparent that this strategy is not valid due to the residual ϵ being a function of the parameters k ( by virtue of the noise in the observed design matrix ) . The reactionet lasso implements a stepwise strategy to address this dependency . The first step ( Step 1 ) is a Feasible Generalized Least Squares ( FG ) estimate . It comprises the estimation of the variances of the residuals ϵb and ϵA via bootstrapping of the gradient estimates and stoichiometric moment functions on the basis the single-cell data . A preliminary least squares fit is then performed to achieve an estimate kLS for eq 5 . This estimate is expected to approximate the order of magnitude of the individual rate constants . In conjunction with the estimates of the variances of the residuals ϵb and ϵA , we use kLS to achieve an estimate of the component-wise variance Σ ϵ = diag { σ ϵ 1 2 , … , σ ϵ R 2 } of the residuals ϵ . To achieve this estimate we use only first order moments ( means ) , as they are less subjected to noise in the design matrix and provide a more robust estimate of the covariance matrix Σϵ . This estimate will allow us to operate with the rescaled observed response vector b ^ S = Σ ϵ - 1 / 2 b ^ and design matrix A ^ S = Σ ϵ - 1 / 2 A ^ to adjust for heteroscedasticity and enable effective linear regression [15] . k ^ F G = arg min k ≥ 0 ∥ b ^ S - A ^ S k ∥ 2 2 , ( 6 ) The subsequent steps aim at addressing the second challenge introduced above , i . e . the different scales of individual parameters ki , which render conventional sparse regression approaches ( such as the Lasso ) suboptimal due to the uniform penalization strength of the L1 norm ‖ . ‖1 across all components ki of the parameter vector k . The adaptive Lasso [16] constitutes an alternative to the conventional Lasso . It defines a regularization penalty that is scaled component-wise by the expected order of magnitude k i ^ of the respective component i . In Step 2 of the reactionet lasso , we apply a combination of the adaptive and relaxed Lasso , stability selection based prioritization of reactions and an additional stepwise backward regression to achieve the final set of reported reactions . We use the parameter estimates from Step 1 ( obtained with Moore–Penrose pseudoinverse matrix ) , i . e . k ˜ = k^ F G , in order to adapt the regularization penalty . To improve shrinkage , the adaptive Lasso is followed by a relaxed Lasso [17] that recomputes optimal parameter estimates with respect to the objective specified in eq 6 , while only considering the set Φ = { l : k l F G ≠ 0 } of parameters that were not set to zero in Step 1 , for which the optimal solution is k ^ A R L = arg min k ≥ 0 ∥ b ^ S - A ^ S , Φ k ∥ 2 2 + λ Σ i | k i / k ˜ i | , ( 7 ) where A ^ S , Φ contains only that columns of A ^ S , which are in a set Φ . The adaptive relaxed Lasso solution has been computed by optimizing the respective Alternating Direction Method of Multipliers ( ADMM ) formulations [18] . The adaptive relaxed Lasso is performed with five fold cross validation . We used stability selection to prioritize reactions according to their frequency of being selected across all cross validation folds [19] . Bayesian information criterion ( BIC ) was used as selection criterion ( S2 Text ) . In summary , the reactionet lasso procedure constitutes a stepwise sparse regression approach that addresses the parameter-dependent noise and heteroscedasticity in the response and design matrix for structure learning of stochastic chemical reaction systems . See also Fig 1 for a schematic overview of its steps . Software implementing the reactionet lasso can be found at http://www . imsb . ethz . ch/research/claassen/Software/reactionet_lasso . html . We first consider an extreme and yet conceptually simple scenario where we aim at learning the structure of a reaction network without any prior knowledge about the underlying reactions . While this scenario rarely occurs in a real world application because typically some prior knowledge of relevant reactions is available , we first investigate this scenario to demonstrate the structure learning capabilities of the reactionet lasso . We study two systems varying in number of components and reactions: ( 1 ) the enzymatic reaction system with four components and three reactions , ( 2 ) the receptor subunit of a recently reported kinetic model of TRAIL induced apoptosis with fourteen components and thirteen reactions , which can be combined in a total of 2275 possible unary or binary reactions , giving a total of more than 10600 possible reaction network candidates . For these systems we simulated 5 replicates each with either 103 , 104 or 105 single cell trajectories with the stochastic simulation algorithm [20] . We then generated snapshot time series datasets from the single cell trajectories by defining pools of cells at selected sets of 7 , 13 , or 28 time points . Moment gradients were estimated either with the smoothing procedure , cubic splines or the finite difference scheme ( see Methods ) . The reactionet lasso achieves structure learning of chemical reaction networks via a two step sparse regression formulation that ( 1 ) specifically accounts for heteroscedasticity in the response vector and the design matrix of the regression instances and ( 2 ) assumes a regularizer that encourages sparse reaction sets by suppressing compensatory reaction sets with small rate constants ( Fig 1 ) . The first step aims at accounting for heteroscedasticity and , most importantly at reducing the number of reaction candidates for the second step that both capture the empirical moment gradients and select for correct reactions ( Fig 2 ) . The following results are based on moment equations for all moments up to order two , i . e . means , variances and covariances . Following Step 1 of the reactionet lasso , we achieve a substantial reduction to less than 100 candidate reactions that , regardless of the moment gradient estimation technique , retains at least ten of the thirteen true reactions ( Fig 1A ) . The vast majority of the empirical moment gradients are well fit by the set of candidate reactions . The few moment gradients that are suboptimally captured correspond to higher order moments such as variances or covariances whose highly dynamic behavior precluded accurate gradient estimation by either the finite difference or spline fit . ( Fig 2B ) . Step 2 of the reactionet lasso procedure uses a relaxed adaptive Lasso estimator to estimate the rate constants of a sparse set of candidate reactions following from Step 1 . The method recovers ten out of thirteen reactions correctly with one false positive reaction when assuming no prior knowledge and selecting a suitable model with BIC ( S1A Fig ) . Similar performance is achieved for the enzymatic reaction network ( S1B Fig ) . These results demonstrate that the stepwise sparse regression strategy of Step 2 completes the structure learning task from the candidate reactions supplied by Step 1 with great sensitivity and specificity . In summary , the reactionet lasso is able to ab initio reconstruct the reaction network structure of typically-sized signaling cascades such as the fourteen component receptor subunit of TRAIL induced apoptosis [21] . We further evaluated the impact of different gradient estimation approaches on structure learning performance ( S2–S4 Figs ) . For benchmarking purposes we used the smoothed empirical moment gradient estimate as a ground truth which is not available in a real time series snapshot setting . According to these considerations , the cubic spline estimator achieves almost optimal performance for thirteen or more time points , whereas FDS is consistently inferior . These results indicate that the cubic spline estimator provides the most favorable structure learning performance for empirical moment gradients . We evaluated how measurement noise affects the ability of reactionet lasso to learn the reaction network structure . We assume a binomial measurement noise model that reflects the incomplete capture efficiency inherent to all single cell technologies ( see Methods , S3 Text ) . While structure learning performance is reduced with increasing levels of measurement noise , the reactionet lasso still recovers more than 50% of the reactions for the apoptotic receptor subunit at levels reported for single cell sequencing and mass cytometry approaches ( Fig 3 , S5 Fig ) . To assess the relative importance of the amount of available data , we varied the amount of time points and single cell recordings used at each time point . Interestingly , we found that the inclusion of additional measurement time points did not improve structure learning performance . However , the tradeoff between true and false positive reaction discoveries worsened considerably with fewer time points ( Fig 4A ) . While we found that decreasing the amount of single cell measurements per time point did result in noticeable performance losses , this situation does not constitute a limitation for flow cytometry techniques , that are easily able to generate millions of single cell snapshots ( Fig 4B ) . Cell count related performance losses can be associated with higher absolute variability and therefore reduced accuracy of empirical moment estimates ( S6 Fig ) . We conclude that careful selection of amount of single cell measurements and number as well as position of time points ( S7 Fig ) translates to accurate interpolation and subsequent gradient fitting , thereby leading to good structure learning performance of the reactionet lasso . We further investigated the impact of including different moment orders for structure learning . As expected , precisely estimated higher-order moments contain a substantial amount of information and therefore enhance the structure learning capability accordingly ( Fig 5A ) . However , although this relationship still holds for medium levels of measurement noise ( capture efficiency p = 0 . 1 ) , ( Fig 5B ) , the inclusion of second order moments becomes misleading for high levels of measurement noise ( capture efficiency p = 0 . 05 , Fig 5C ) . This observation is likely caused by the limited ability to accurately estimate higher order moments for high levels of measurement noise . However , the performance of the reactionet lasso assuming stochastic kinetics modeled with moment equations ( higher order ME ) is consistently better than assuming a deterministic kinetics modeled with mean based ordinary equations ( 1st order ODE ) . This observation demonstrates that the incorporation of higher order moment information induced by the chemical kinetics and accessible by means of single cell measurements allows for significantly improved structure learning capacity . In summary , the benchmarks above strongly advocate for the use of an experimental setup that allows for sufficiently dense sampling across time to ensure accurate empirical moment gradient estimates , as well as single cell technology , such as flow/mass cytometry , which provide 104 or more single cell measurements , for the accurate estimation of higher order moments . In these situations the reactionet lasso is capable of ab initio recovery of almost the complete reaction network structure with more than a dozen components . We now consider a scenario where we aim at learning the structure of a large reaction network with partial knowledge about the underlying reactions . For this situation we demonstrate how reactionet lasso is capable of recovering a sizable amount of the unknown reactions , for a reaction network as large as the 70 reaction TRAIL induced apoptosis cascade [21] . Structure learning tasks for chemical reaction networks typically aim at complementing already available partial knowledge on reaction sets . We assessed the ability of the reactionet lasso to complement a set of known reactions for the 70 reaction TRAIL induced apoptosis cascade . Specifically , we defined six modules for this cascade following [21] , and assumed a limited set of 22 reaction candidates connecting these modules ( S8 Fig ) and 33 uniformly distributed time points ( S1 Dataset ) . For step 1 of the reactionet lasso all possible unary and binary reactions between components within modules and the module connecting reactions serve as candidate reactions for structure learning , totaling 6828 reactions . In the absence of ground truth it is difficult to identify a regularization strength that achieves a desirable tradeoff between true and false positive reaction discoveries . We evaluated the BIC and report solutions that map to large initial improvements of BIC [22] . Structure learning without prior knowledge on the considered set of reactions achieves 32 true positive at the cost of 2 false positive reactions ( 105 single cell trajectories , 33 time points , capture efficiency 0 . 05 , Fig 6A ) . Prior knowledge on a specific reaction was encoded by a positivity constraint on the corresponding reaction rate during all regression steps of the reactionet lasso . We considered different prior knowledge settings: ( 1 ) 10% or ( 2 ) 50% randomly chosen reactions considered to be known . Settings ( 1 ) and ( 2 ) were each evaluated using ten different subsets . For 10% known reactions almost 40 ( including 7 known ) out of 70 reactions are correctly recovered with five or less false positive discoveries ( Fig 6B ) . For 50% known reactions the total number of true positive reactions is beyond 50 ( including 35 known ) . The performance doesn’t depend significantly on the choice of prior reactions . The reactionet lasso enables discovery up to dozens of novel reactions at the cost of few false positive reactions for a large signaling cascade comprising 70 reactions . While published structure learning approaches are only available for problem instances of sizes hundreds of orders of magnitude smaller , we compared these results to more simple variants of the reactionet lasso procedure , either exhibiting inferior accuracy or exceedingly high computational complexity ( S9 Fig ) . The above results demonstrate how single cell snapshot time series data and the reactionet lasso can be used to complement prior mechanistic knowledge by a sizable set of candidate reactions that is highly enriched for true positive discoveries , and do so for systems and structure learning tasks of unprecedented size [23 , 24] .
In this work we introduce the reactionet lasso for comprehensive structure learning of stochastic chemical reaction networks . Chemical reaction networks constitute a highly detailed and mechanistic description for biological processes and are qualitatively different from other popular network models in biology . These comprise probabilistic graphical models seeking to discover statistical dependencies between measured system components . These approaches range from simple correlation [25] or regression analysis [26] , to Bayesian networks [27] or more structured and robust module networks [28] and extensions thereof [29] . In contrast to chemical reaction networks , each of these model classes allows for detection of statistical dependencies without further elucidation of causality relationships and the possibly intricate dependency inducing biochemical mechanism . Physical interaction networks get closer to this goal and complement the information of reaction networks by summarizing measurements of static protein interactions [30] . By virtue of formulating the task of structure learning of chemical reaction networks as a sequence of convex optimization problems , this procedure is able to assess an unprecedented number of potential network topologies without need for explicit enumeration [23 , 24] . We demonstrate the utility of the method for ab initio structure learning of whole signaling cascades such as the apoptotic receptor subunit . The reactionet lasso originally integrates a moment based description of stochastic reaction networks with sparse regression approaches via a gradient matching to achieve an efficient and scalable structure learning procedure , overcoming the limitations of available methods for structure learning which either explicitly enumerate a small set of models or greedily search for locally optimal topologies [8 , 31] . Recent generic sparse regression approaches for identification of general nonlinear dynamical systems are in principle applicable for structure learning of biological reaction networks [9] . However , these approaches , in contrast to the reactionet lasso , do not take into account their ( 1 ) foundation in the Chemical Master Equation , ( 2 ) heteroscedastic and parameter dependent noise structure , as well as ( 3 ) parameter ranges varying across many scales , therefore failing to achieve competitive structure learning performance ( S9 Fig ) . The challenging structure learning task crucially depends on sensible experimental design yielding informative data . A central design choice concerns the selection of time points recording the relevant dynamical changes of the process of interest . These are typically chosen from prior knowledge or preliminary dense snapshot time series experiments with a cheap readout , such as population based instead of single cell measurements . Another important experimental parameter concerns the number of single cell snapshots . Our benchmarks advocate for having at least thousands of snapshots per time point . Flow cytometry experiments easily achieve snapshot counts in the order of 105 . For single cell transcriptomics experiments it seems advisable to resort to novel droplet based techniques achieving > 104 single cell snapshots per experiment [32 , 33] . Structure learning performance of the reactionet lasso depends on the accuracy of the gradient estimates from the time series snapshots . We find that estimates based on gradients obtained by rather simple approaches such as finite difference approximations or spline curve fitting achieve competitive performance . Improvements are conceivable by resorting to other techniques specifically designed for gradient estimation in differential equation systems [34 , 35] . These approaches jointly fit parameters of the curve fitting procedure and the differential equation system . The success of this strategy relies on considering a problem instance where the differentiation equation systems strongly constrains the state space . However , the problem instances we consider for systematic structure learning with the reactionet lasso assume a differential equation system defined by the moment equation for all possible unary and binary reactions . Such a system will by definition impose little constraints on the state space . Gradient matching approaches would therefore have to be adapted to avoid expected parameter overfitting resulting from their application to problem instances with such an expressive differential equation system . For this proof of principle study , we consider single time series experiments . Reactionet lasso analysis easily accommodates multiple replicates or perturbation experiments such as dose responses . Specifically , condition specific response vectors bk and design matrices Ak for each condition k are utilized to construct a problem instance by concatenation . For this problem instance reactionet lasso can be applied as described ( see also S4 Text ) . Additional experiments are expected to enhance structure learning performance . Indeed , we observe that this is the case for incorporating additional replicate time series experiments ( S10 Fig ) . The reactionet lasso is able to recover a significant proportion of missing reactions in various settings . However , integration of the moment equations for the component means assuming this set is not always able to recover the observed temporal dynamics of the system . This situation arises for instance when a single pivotal true reaction is missed and therefore precludes the correct reconstruction of downstream component dynamics . We frequently encounter this situation in the ab initio structure learning scenario ( S11 Fig ) . This scenario though constitutes an artificial setting that we only report for a proof of concept of the reactionet lasso . Real world applications comprise prior knowledge about true reactions , typically comprising specifically those pivotal reactions . It turns out that we achieve good reconstruction of integrated trajectories for the structure learning settings assuming prior knowledge ( S11 Fig ) . Until now we consider reaction systems which obey mass action kinetics . Systems of this kind can be easily translated into a series of moment equations which depend linearly upon the reaction rates However , systems with non-mass action kinetics , such as Michaelis-Menten , can still be addressed with the reactionet lasso . While appropriate moment closure approximations for certain rational rate law kinetics preserve convexity of the reactionet lasso objective [36] , generally such kinetics might yield non-convex optimization problems that would have to be dealt with using appropriate optimization techniques . The reactionet lasso generates a single point estimate for the optimal , sparse reaction network that might neglect other reasonable candidate network structures . Thus , it will be interesting to perform further in-depth analysis of the resulting network structures , for instance with Markov Chain Monte Carlo sampling techniques . For our study we assume that all relevant molecular components can be measured . Many biological applications , however , do not allow monitoring all relevant components , as for instance antibodies might only be available for a subset of components of a signaling cascade . While the aim of our study was to demonstrate proof of concept for large scale structure learning of chemical reaction networks , it will be possible to account for missing measured components by either augmenting the model by introducing latent variables or ‘lumping’ them into more complex non-mass action reaction mechanisms [37] . The reactionet lasso can be applied in its current form to systems where a significant proportion of relevant components can be measured . Considering the steady advance of single-cell technologies , we expect an increasing number of cellular signaling and metabolic processes to be assayed at single-cell resolution . While mass cytometry approaches allow for measurement of more than 30 protein components , sufficient e . g . to substantially map out the T cell receptor , epidermis growth factor and apoptosis signaling cascades , single-cell RNA sequencing opens the prospect of achieving genome-wide transcriptomic snapshots of single cells . Thus we anticipate a surge of relevant data in the near future for which the reactionet lasso can straightforwardly be applied for systematic and comprehensive structure learning of the underlying reaction networks , with direct implications for systems biology and health by providing quantitative and predictive models for scientific insight and rational intervention design .
We assume time series data with single cell resolved population snapshots obtained at discrete time points . We denote by C the number of cells measured per experiment , T the number of time points at which measurements were performed and N the number of components ( e . g . proteins ) measured in each cell . For each measurement time point t = 1 , … , T for a cell c = 1 , … , C we denote a vector of measured N protein abundances x t c = { x t , 1 c , … , x t , N c } . Therefore at each time point t vectors x t , 1 c , … , x t , N c represent a sample from a high-dimensional distribution , which evolves according to the Chemical Master Equation . We assume a biochemical reaction network of N different chemical species with abundances X1 , … , XN involved in L reactions . Each reaction l is characterized by stoichiometry vector sl and propensity function al ( x; kl ) with x representing the collection of species abundances ( system state ) and kl the reaction rate . In our work we consider systems described by mass action kinetics , resulting propensities al ( x; kl ) = kl*gl ( x ) , where gl ( x ) is a known function of the system’s state . The state of the system evolves probabilistically according to the possible reactions , with probability P ( x , t ) of occupying state x at time t . The probabilistic evolution of the system’s state is described by the Chemical Master Equation: P ( x , t ) d t = Σ l = 1 L P ( x - s l , t ) a l ( x - s l ) - P ( x , t ) a l ( x ) . ( 8 ) We denote by M r = M r 1 , … , r N = E ( X 1 - E X 1 ) r 1 … ( X N - E X N ) r N the central moment of order r = ( r1 , … , rN ) . The moment generating function of the probability distribution P ( x , t ) can be used for the derivation of moment equations [38] . Assuming mass action kinetics , we obtain Eq 1 for the time evolution of a central moment ( see also S1 Text ) . Gradient matching approaches avoid costly integration by instead interpolating the discrete snapshot time series data and estimating the empirical moment gradients Mr ( t ) , rendering the initial ODE system an algebraic equation system with the parameters as unknowns . This formulation further eliminates the need for moment closure , in contrast to integration based techniques . Previously , gradients have been estimated with spline interpolators [34 , 39 , 40] , Gaussian processes [35 , 41] or finite difference approximations [6] . Parameter estimation has been performed by least squares minimization [34 , 39 , 40] or by approximation of the parameter posterior [6 , 35 , 41] . While deterministic chemical reaction networks frequently served as application settings for gradient matching schemes , only little attention has been paid to networks with stochastic dynamics [6 , 11] . We used and compared cubic spline interpolators ( spline ) and finite difference approximations ( FDS ) to estimate empirical moment gradients for the Mr ( t ) of the moment equations . As a ground truth estimate for simulated data , we use a smoothed finite difference approximation of the single cell trajectories at the evaluation point of interest ( smooth ) . Gradient estimates are obtained via a smoothing procedure that relies on a sliding window estimate of finite differences on the simulated trajectories using the smoothing function “smooth” in Matlab . Single cell data such as obtained from flow/mass cytometry and single cell sequencing exhibit measurement noise . These technologies each detect a random fraction of the total molecular content of every individual cell . This relationship between the measurement signal and cellular analyte abundance has been frequently modeled by a binomial distribution Bi ( X , p ) whose success probability p corresponds to the capture efficiency for the analyte present at amount X [42 , 43] . We have devised an estimator to subtract the misleading measurement noise component to provide the reactionet lasso with the appropriate noise-correct empirical moment estimates for structure learning . We assume that measurement noise can be represented by the following binomial model . Let X represent the true abundance of one species at a given time point . Let Xobs be the corresponding measured signal , such that Xobs ∼ Bi ( X , p ) , where p is the capture efficiency . The binomial noise model allows for specifying the following analytical relationships between the first and second order moments of X and Xobs: E [ X o b s ] = p E [ X ] , ( 9 ) E [ ( X o b s ) 2 ] = p ( 1 - p ) E [ X ] + p 2 E [ X 2 ] , ( 10 ) V a r [ X o b s ] = p ( 1 - p ) E [ X ] + p 2 V a r [ X ] , ( 11 ) cov ( X 1 o b s , X 2 o b s ) = p 2 cov ( X 1 , X 2 ) , ( 12 ) For a derivation see S3 Text . We assume that the capture efficiency p of the single cell instrument is known [42 , 43] and estimates the empirical moments of X on the basis of the empirical moments of Xobs by solving the above equations for the respective moment of X . The resulting moment estimates are then used in the regression procedure described above to perform structure learning . | Virtually all biological processes are driven by biochemical reactions . However , their quantitative description in terms of stochastic chemical reaction networks is often precluded by the computational difficulty of structure learning , i . e . the identification of biologically active reaction networks among the combinatorially many possible topologies . This work describes the reactionet lasso , a structure learning approach that takes advantage of novel , information-rich single cell data and a tractable problem formulation to achieve structure learning for problem instances hundreds of orders of magnitude larger than previously reported . This approach opens the prospect of obtaining quantitative and predictive reaction models in many areas of biology and medicine , and in particular areas such as cancer biology , which are characterized by significant system alterations and many unknown reactions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"learning",
"cell",
"death",
"protein",
"interaction",
"networks",
"cell",
"processes",
"social",
"sciences",
"neuroscience",
"learning",
"and",
"memory",
"cognitive",
"psychology",
"mathematics",
"systems",
"science",
"network",
"analysis",
"reaction",
"dynamics",
"phy... | 2016 | Sparse Regression Based Structure Learning of Stochastic Reaction Networks from Single Cell Snapshot Time Series |
Cytomegalovirus ( CMV ) infection induces an atypical CD8 T cell response , termed inflationary , that is characterised by accumulation and maintenance of high numbers of effector memory like cells in circulation and peripheral tissues—a feature being successfully harnessed for vaccine purposes . Although stability of this population depends on recurrent antigen encounter , the requirements for prolonged survival in peripheral tissues remain unknown . Here , we reveal that murine CMV-specific inflationary CD8 T cells are maintained in an antigen-independent manner and have a half-life of 12 weeks in the lung tissue . This half-life is drastically longer than the one of phenotypically comparable inflationary effector cells . IL-15 alone , and none of other common γ-cytokines , was crucial for survival of inflationary cells in peripheral organs . IL-15 , mainly produced by non-hematopoietic cells in lung tissue and being trans-presented , promoted inflationary T cell survival by increasing expression of Bcl-2 . These results indicate that inflationary CD8 T cells are not just simply effector-like cells , rather they share properties of both effector and memory CD8 T cells and they appear to be long-lived cells compared to the effector cells from acute virus infections .
T cells are a major part of the adaptive immune system and have the ability to differentiate into memory T cells after previous stimulation with cognate antigen . Memory T cells can react much faster upon antigen restimulation by re-expansion and immediate effector functions . The induction of memory T cells , in particular of memory T cells that reside long-term in peripheral tissues , is of considerable interest for the development of T cell-based vaccines , as they afford protection in situ in case of local pathogen infection . Acute infections induce memory T cells , which mainly reside in lymphoid tissues , even long after pathogen clearance [1 , 2] . These are defined as central-memory T cells ( TCM ) expressing CD62L , CD127 ( IL-7Rα ) and CCR7 [1 , 2] . Aside TCM , acute infections also induce effector-memory T cells ( TEM , CD62L- , IL-7Rα+ and CCR7- ) , mainly at early time points after resolution of the infection , and the more recently discovered tissue-resident memory T cells ( TRM ) that can permanently lodge into tissues , are disconnected from the circulation [3] and exhibit protection during local reinfection events [4–6] . Despite the fact that acute viral infections and / or vaccinations are able to induce TRM cells , there are specific vectors that excel in the ability to induce large populations of effector-like memory cells in peripheral tissues . Cytomegalovirus , a member of the herpes virus family , is a potent inducer of a very large memory T cell pool , consisting of distinct memory subsets with different localisations , phenotypes and functions . In various species , CMV infections induce an exquisitely large and sustained population of functional memory CD8 T cells residing in peripheral tissues [7–11] . Thus , CMV-based vectors have gained broad interest in the development of T cell-based vaccines . Indeed , recombinant CMV-based vaccine vectors expressing SIV- or ebolavirus glycoprotein-derived epitopes showed protection of vaccinated rhesus macaques against SIV and ebolavirus challenge . Also , vaccinations with recombinant murine CMVs ( MCMVs ) expressing influenza A NP epitope provided protection of mice against VacV-NP challenge [12–16] . The sustained expansion and maintenance of CMV-specific CD8 T cells in peripheral tissues is a hallmark of CMV infection and accounts only for CMV-reactive CD8 T cells with certain specificities . While some CD8 T cells with defined specificities are differentiating along the "classical" expansion / contraction / memory formation pathway , specific CD8 T cell subsets follow a substantially different kinetics and differentiation trajectory . These CMV-specific CD8 T cells continue to expand also at time points when their "classical" counterparts decline in numbers , and establish themselves as a stable population of memory CD8 T cells with an activated phenotype in peripheral tissues [8 , 17–19] . Importantly , these activated memory CD8 T cells residing in peripheral tissues exhibit an exquisite capacity to control peripheral infections [20 , 21] . The mechanisms responsible for this "inflationary" behaviour of specific CMV-reactive CD8 T cell populations are currently still ill-defined . However , a number of studies have elaborated that CMV antigen presentation , resulting from CMV reactivation events , drives memory CD8 T cell inflation and that the ability of a given epitope to induce memory inflation depends on the context of gene expression and its processing by the conventional immunoproteasome [22–26] . Inflationary CD8 T cells exhibit high expression levels of the terminal differentiation marker KLRG-1 and are highly functional with respect to pro-inflammatory cytokine production and cytotoxicity , unlike functionally exhausted cells emerging in the context of highly active chronic viral infections [18 , 21 , 27–29] . Inflationary CD8 T cells were described as short-lived effector cells whose pool needs to be continuously replenished to maintain stable numbers in peripheral tissues [11] . This replenishment relies on reactivation and re-expansion of TCM cells by cognate antigen presented on non-hematopoietic cells—presumably derived from viral reactivation events—taking place in lymph nodes or in the vasculature [25 , 30 , 31] . However , it remains unclear how long these reactivated cells persist in peripheral tissues and how peripheral maintenance of inflationary T cells is regulated . In this study , we assessed these questions using a monoclonal MCMV-specific CD8 T cell population specific for an inflationary epitope in peripheral organs . Our data demonstrate that inflationary CD8 T cells in the lung exhibit a half-life between 10–12 weeks , independent of local antigen . We reveal a decisive role of IL-15 in maintaining inflationary T cells in the lungs , whereas all other common γ-chain cytokines are dispensable . We identified a dominant role of IL-15 production in non-hematopoietic cells in the lung for the maintenance of lung-resident inflationary CD8 T cells , which also bears relevance for the use of CMV-based vectors for vaccine purposes .
To assess the long-term maintenance of inflationary MCMV-specific CD8 T cells in peripheral organs , we adoptively transferred M38-specific TCR transgenic CD8 T cells ( Maxi ) into naïve C57BL/6 hosts , followed by MCMV infection . We analysed frequencies and total numbers and phenotype of transgenic Maxi cells at various time points post infection in several organs after MCMV had established latency . We found that the total number of Maxi CD8 T cells was kept constant in all organs up to nine months post infection ( Fig 1A ) . Performing intravascular labelling of Maxi cells revealed that high frequencies of Maxi cells resides either within or in close contact to the circulation in lungs , liver and spleen , but not in lymph nodes ( Fig 1B and 1C ) . In accordance with previous studies , a vast majority of inflationary cells in lung , liver , spleen and lymph nodes does not express the typical TRM markers CD69 or CD103 ( S1 Fig ) , but expresses markers associated with effector-like T cells ( CD127low , KLRG-1+ , CD62L- , CD44+ ) , particularly in the lung , liver and spleen at all time points analysed ( Fig 1D and 1E ) [32 , 33] . In contrast , the majority of Maxi cells present in the lymph nodes express markers associated with central-memory T cells ( CD127+ , KLRG-1- , CD62L+ , CD44+ ) , confirming previous data on endogenous M38-specific CD8 T cells ( Fig 1F and 1G ) . Therefore , we conclude that transgenic Maxi cells display similar properties with respect to number and phenotype as described for endogenous inflationary CD8 T cells . Furthermore , to determine more precisely the location of M38-specific Maxi cells in lung tissue , we performed fluorescence microscopy on lung sections . Maxi cells , identified by CD45 . 1 staining , were located close to CD31+ endothelial cells and were not found in association with EpCAM+ epithelial cells ( Fig 1H ) . Furthermore , the majority of the CD45 . 1 Maxi cells were labelled by i . v . administered αCD8 antibody , marked by white arrows . Since sporadic reactivation events and antigen presentation on non-hematopoietic cells are essential for driving memory inflation during MCMV latency , it is likely that the peripheral pool of inflationary CD8 T cells is continuously fuelled by recently re-activated TCM Maxi cells . Hence , the inflationary pool would decline when the TCM pool would be ablated . We tested this hypothesis in an adoptive transfer system where we transplanted TEM Maxi T cells isolated from lungs of latently infected mice into infection-matched or naïve C57BL/6J recipient mice . Specifically , at 60 days post infection , we isolated lymphocytes from the lung tissues and sorted effector-memory Maxi T cells based on CD44+ and CD62L- expression ( Fig 2A and 2B ) . In the recipients , total numbers of Maxi cells recovered from the lung tissue were quantified at three different time points to determine the half-life of the inflationary T cell pool . We found that transferred Maxi cells exhibited comparable half-lives of 10–12 weeks in lungs of naïve or infection-matched recipients , indicating that survival of inflationary cells in lungs is independent of antigen ( Fig 2C ) . This half-life is slightly longer compared to previous publications where phenotypically mixed inflationary cells from the spleen were described to have a half-life of 45–60 days [11] . Of note , comparable half-lives to lung tissue were observed in spleen ( S2 Fig ) . Being surprised by the relatively long half-life of effector-memory Maxi cells , we asked the question how phenotypically similar effector cells from an acute time point of infection would compare with respect to their half-lives to inflationary cells from the latent phase of infection . We therefore sorted CD44+ CD62L- , KLRG1+ Maxi CD8 T cells from lungs of acutely MCMV infected mice ( day 7 , Fig 2D and 2E ) , transferred them into naïve recipients and quantified Maxi cells <1 week , 6 and 12 weeks post transfer in the lungs of the recipients . These Maxi cells isolated from day 7 post infection had a strikingly shorter half-life , less than 6 weeks , despite expressing similar surface markers as the cells sorted from latent MCMV infection ( Fig 2F ) . As inflationary M38-specific CD8 T cells steadily increase their levels of the anti-apoptotic protein Bcl-2 over the course of the MCMV infection [25] , we compared Bcl-2 levels in day 7 and day 60 Maxi cells . Indeed the levels of Bcl-2 were significantly higher in effector-memory Maxi cells from latently infected mice compared to effector cells from acute infection ( Fig 2G ) , suggesting that inflationary TEM and inflationary TEFF CD8 T cells differ with respect to their longevity . As effector Maxi cells had a different half-life compared to inflationary effector-memory Maxi cells , we wondered whether this observation also holds for CD8 T cells with a different specificity . We made use of the TCR transgenic P14 mouse , whose CD8 T cells are all specific for the gp33-41 epitope of lymphocytic choriomeningitis virus ( LCMV ) in combination with recombinant MCMVs that express gp33 under the ie2 promotor ( inflationary kinetics ) or the M45 promotor ( non-inflationary kinetics ) [34 , 35] . Naïve P14 cells were transferred into naïve recipients prior to MCMV-gp33 infection . P14 effector memory cells were sorted from lungs on day 7 post MCMV-M45-gp33 infection , and at least 60 days post MCMV-ie2-gp33 infection , and further transferred them into naïve recipients ( Fig 3A ) . Notably , we corroborated the results obtained with Maxi cells , such that effector P14 cells had a diminished half-life when compared to the TEM P14 counterparts ( Fig 3B ) . The shorter half-life was once more correlated with reduced Bcl-2 levels in comparison to TEM P14 cells , suggesting an important role of Bcl-2 in the maintenance and survival of inflationary T cells during latent MCMV infection . T cells and particularly memory CD8 T cell were shown to be maintained by homeostatic cytokines . Moreover , the continuous upregulation of the anti-apoptotic protein Bcl-2 over the course of MCMV infection indicates that potential exposure to homeostatic cytokines in the organs might promote T cell survival [25 , 36 , 37 , 38] . To address the question whether and which homeostatic cytokines might promote the survival of peripheral inflationary CD8 T cells , inflationary Maxi cells were generated as described before , sorted from lung tissues , and adoptively transferred into infection matched C57BL/6 mice . The recipients were continuously treated with cytokine / cytokine receptor blocking antibodies for a period of 30 days before quantification of Maxi cells in lung tissue . Alternatively , inflationary Maxi TEM cells from lungs were transferred into naïve cytokine-deficient recipients and remaining Maxi cells were also quantified 30 days post transfer ( Fig 4A ) . Interestingly , we observed no difference in recipients that were treated with neutralising α-IL-7 antibodies ( Fig 4B ) . Also , we found no evidence for a role of IL-2 as mice treated with neutralising α-IL-2 antibodies showed no significant reduction in Maxi numbers in the lung . Strikingly , the only cytokine that had an influence on the maintenance of inflationary TEM Maxi cells was IL-15 . Namely , when recipient mice were either administered with receptor blocking α-CD122 ( IL-2/15Rβ ) antibody or Maxi cells were adoptively transferred into IL-15-deficient mice , we observed a significantly reduced recovery of Maxi cells from the lung tissue ( Fig 4B ) . We hypothesized that the reduced numbers of Maxi cells were associated with reduced Bcl-2 expression in IL-15-deficient recipients . Indeed , significantly reduced levels of Bcl-2 were apparent when the mice were treated with α-CD122 antibodies or when the recipients were IL-15 deficient ( Fig 4C and 4D ) . No reduced levels of Bcl-2 were observed in any other conditions ( Fig 4D ) . We exposed inflationary Maxi cells in vitro to recombinant IL-15 and found upregulation of pSTAT5 ( Fig 4E ) and Bcl-2 ( Fig 4F ) highlighting that inflationary cells are responsive to IL-15 . These findings suggest a pivotal role for IL-15 in the maintenance of inflationary CD8 T cells , most likely by enhancing survival via upregulation of the anti-apoptotic protein Bcl-2 . As IL-15 neutralization or use of IL-15 deficient recipient mice might also target other immune cells than the transferred TEM Maxi cells—in particular NK cells that are highly dependent on IL-15 signalling , we wanted to rule out a potential involvement of NK cells in the maintenance of inflationary CD8 T cells . Therefore , we adapted our transfer model including administration of α-NK1 . 1 antibodies over four weeks and assessed the total numbers of surviving Maxi cells in the lung tissues ( S3A Fig ) . We did not detect significant differences in total numbers of Maxi cells in any of the organs analysed ( S3B Fig ) , suggesting that IL-15 acts directly on inflationary T cells . To control for NK cell depletion , we enumerated the total numbers of NK cells in the lungs and observed a strong reduction in the NK cell numbers in all organs analysed ( S3C Fig ) . We therefore concluded that IL-15 has a direct role in providing survival signals to inflationary CD8 T cells . To confirm a pivotal role of IL-15 in sustaining memory CD8 T cell inflation during MCMV infection , we infected naïve C57BL/6 and IL-15-/- mice with MCMV and followed endogenous M38-specific CD8 T cells over the course of infection in the blood . Interestingly , we did not observe differences in percentages of M38-specific CD8 T cells during acute MCMV infection , suggesting that priming of naïve CD8 T cells and clonal expansion are not affected in the absence of IL-15 . Yet , when MCMV latency was established and lytic virus replication was absent from most peripheral organs except the salivary glands , IL-15-deficient animals had significantly reduced frequencies of M38-specific CD8 T cells compared to wild type mice ( Fig 5A ) despite comparable control of lytic MCMV infection ( Fig 5D ) . Detailed analysis of the abundance of virus-specific cells during latency in the organs revealed organ-specific reductions in total numbers and percentages of M38-specific CD8 T cells in the lungs and spleen , whereas in the LNs there were no significant differences observed ( Fig 5B and 5C ) . With most TCM cells specific for M38 being located in lymphoid tissues , particularly in LNs , this observation suggests that maintenance of TCM M38-specific CD8 T cells does not critically depend on IL-15 signals in contrast to their peripheral TEM counterparts . Consistent with previous experiments , Bcl-2 levels in M38-specific CD8 T cells were reduced in IL-15-deficient hosts in all organs ( Fig 5E ) . Interestingly , we found a higher percentage of Ki-67+ M38-specific CD8 T cells in the lungs and spleen of IL-15 deficient hosts , but not in the LNs ( Fig 5F ) . It is conceivable that increased peripheral antigen encounter due to inferior MCMV reactivation surveillance by reduced numbers of TEM cells might be responsible for this increased exit from G0 stage in M38-specific CD8 T cells . So far , we could unravel a crucial role for IL-15 in maintaining the peripheral inflationary T cell pool in lung tissue . However , it remained an open question , which cellular source is responsible for providing IL-15 . Several cellular sources have been described to be able to express IL-15 , including macrophages , dendritic cells as well as stromal cells [39–41] . Hence , we quantified IL-15 expression levels in various cellular compartments of lung tissue . We isolated lung tissue of naïve C57BL/6 mice and sorted cells into hematopoietic and the non-hematopoietic origin by staining for CD45 , isolated RNA and determined the levels of Il15 mRNA by RT-qPCR . We found that the relative Il15 expression level was higher in the non-hematopoietic compartment compared to the hematopoietic compartment ( Fig 6A ) , indicating a potential role of IL-15 production by non-hematopoietic cells in the maintenance of inflationary T cells . To confirm a potential involvement of non-hematopoietic cells in vivo , we generated bone-marrow chimeric mice with selective IL-15 deficiency in either hematopoietic or non-hematopoietic compartments ( Fig 6B ) . Chimerism of reconstituted mice was confirmed ( S4 Fig ) . We then sorted inflationary Maxi TEM cells from latently infected donor mice and adoptively transferred them into the chimeras . Four weeks post transfer we assessed the total numbers of recovered Maxi cells in the lung and in the spleens . In the lungs ( Fig 6C ) , we observed the lowest recovery of Maxi cells in mice where both non-hematopoietic and hematopoietic cells were deficient of IL-15 , consistent with previous experiments performed in IL-15-/- mice . Interestingly , we observed a 50% reduced recovery in mice lacking IL-15 on non-hematopoietic cells , whereas in IL-15-deficient hematopoietic cells no difference was observed ( Fig 6C ) . In the spleen , none of the chimeras which selectively lacked IL-15 in one or the other compartment showed a reduced recovery of Maxi cells , only the complete deficiency resulted in a reduced Maxi cell number ( Fig 6D ) . These results suggested an organ-specific role of non-hematopoietic cells in expression of IL-15 , promoting maintenance of inflationary T cells . Next , we aimed to identify the specific cellular source of IL-15 for the maintenance of inflationary M38-specific CD8 T cells in lung tissue . Thus , we analysed Il15 mRNA levels in different stromal cells from the lungs of naïve C57BL/6 mice . However , we were unable to identify a cell type exhibiting dominant IL-15 mRNA production , including epithelial cells , blood endothelial cells nor lymphoid endothelial cells ( S5 Fig ) . Next , we quantified the maintenance of inflationary T cells in mice that lacked IL-15 expression in specific cell types . We used mice with cell-type specific depletion of IL-15 , including CCL19-producing fibroblastic reticular cells ( FRCs ) ( Ccl19-Cre Il15fl/fl ) , CD11c-expressing dendritic cells and alveolar macrophages ( CD11c-Cre Il15fl/fl ) , and endothelial cells ( VE-Cadherin-Cre Il15fl/fl ) . The reasons for these choices are that 1 ) FRCs were shown to be involved in control ILC1 cells through IL-15 in the intestine [41] . 2 ) Dendritic cells and macrophages have been shown to be important to support effector-memory and central-memory T cells through IL-15Rα , trans-presenting IL-15 to the responding T cells [40] . Also , pulmonary dendritic cells were shown to maintain influenza-specific CD8 T cell responses via IL-15 signals [42] . 3 ) Endothelial cells are the major cell type of the vasculature and blood vessels and the lung tissue is heavily vascularised , and we observed that inflationary T cells are largely stained by intravascular antibody staining , suggesting localisation in or within close proximity to the vasculature ( Fig 1B ) . TEM Maxi cells from lungs of latently infected mice were adoptively transferred into cell-type specific IL-15 knockout mice . We assessed the recovery of Maxi cells from the lung tissue four weeks post transfer ( Fig 7A ) . Neither CD11c-specific ablation of IL-15 ( Fig 7C ) , nor VE-Cadherin-specific deletion of IL-15 ( Fig 7D ) had an impact on the survival of adoptively transferred TEM Maxi cells . Interestingly , we did observe a small but not significant reduction of recovered Maxi cells in Ccl19-Cre x IL-15fl/fl mice compared to the Cre-negative control mice ( Fig 7B ) , suggesting a potential role of CCL19-expressing FRCs enhancing inflationary T cell maintenance . However , our data suggest that the source of IL-15 comprises several subtypes of non-hematopoietic as well as some contribution of the hematopoietic cell compartment . IL-15 is usually 'trans-presented' to responding cells by the IL-15-expressing cells , meaning that the expressed IL-15 is bound by the IL-15Rα of the producing cell on its surface , by which it delivers signaling to the responding cell via the CD122 ( IL-2/15Rβ ) /γc complex [43 , 44] , mainly being described for dendritic cells and macrophages . We therefore sought to investigate whether trans-presentation of IL-15 by host cells is critical for the survival of transplanted inflationary MCMV-specific CD8 T cells . To this end , we crossed Il15rafl/fl mice to a constitutively expressing Cre background ( CMV-Cre ) , thereby generating recipient mice that lack IL-15Rα expression . We generated inflationary T cells by transferring naïve Maxi CD8 T cells into naïve C57BL/6 hosts and infected them a day later with MCMV . During viral latency , we isolated lungs , sorted inflationary TEM cells and adoptively transferred them into CMV-Cre Il15rafl/fl mice and Cre-negative control hosts . We assessed the total numbers of Maxi cells recovered from the lung tissue four weeks post transfer . We did observe a significant reduction of recovery of Maxi cells from the lung tissue in CMV-Cre Il15rafl/fl hosts ( Fig 8A ) , suggesting a role of both IL-15 and IL-15Rα by hosts cells to maintain inflationary CMV-specific CD8 T cells . Additionally , we also observed a reduction in the expression of Bcl-2 , which is in line with previous findings in IL-15-deficient or CD122-blocked animals ( Figs 4C and 5E ) . Overall , these data suggest an important role of IL-15 during MCMV latency to maintain the inflationary CD8 T cell pool in peripheral organs , in particular by IL-15 / IL-15Rα expressing non-hematopoietic cells .
CMV infection results in a gradual accumulation of inflationary virus-specific CD8 T cells against epitopes that are thought to be expressed in the latent phase during sporadic viral reactivation events [8 , 10] . In contrast , non-inflationary CD8 T cells do not inflate , presumably because their cognate antigens are only expressed during acute infection . Also in human CMV ( HCMV ) infection , studies have demonstrated that particularly in elderly individuals , high frequencies of CMV-specific CD8 T cells accumulate despite absence of lytic virus replication [8 , 16 , 45] . In this study , we demonstrated that inflationary T cells are kept at comparably high numbers in peripheral organs over nine months post infection exhibiting a remarkably constant phenotype . Whether such high numbers of CMV-specific CD8 T cells are important to assure control of CMV infection is rather unlikely , as CD8-deficient or transiently depleted mice are able to control lytic MCMV infection in peripheral organs , even though there is a moderate increase in virus titres , yet it is likely that there is a higher degree of latent viral loads [21 , 46–50] . Also , genetic deletion of an immunodominant inflationary epitope in MCMV resulted in increased and more progressive viral reactivation events in BALB/c mice , although not proceeding to full reactivation and production of infectious virions [51] . Nevertheless , CMV's ability to induce such pronounced numbers of peripherally located CD8 T cells is of great interest for vaccine approaches relying on T cell mediated immunity [12 , 14 , 52 , 53] . Despite a seemingly stable number of inflationary T cells in peripheral organs , there is an underlying dynamic process responsible for this stability . Sensing of CMV-derived antigens on non-hematopoietic cells during viral latency ( presumably during early viral reactivation events ) and processing of these epitopes by the conventional proteasome were shown to be prerequisites for memory CD8 T cell inflation during MCMV infection [22 , 23 , 25 , 51] . Sensing of CMV-derived antigens by TCM CD8 T cells in secondary lymphoid organs leads to their secondary expansion , effector ( memory ) differentiation and migration to peripheral tissues , thereby likely contributing to the overall maintenance of the peripheral pool of inflationary CD8 T cells . Consistent with a pivotal role of TCM reactivation in secondary lymphoid organs is the observation that the percentage of Ki-67-expressing M38-specific CD8 T cells is increased in secondary lymphoid organs compared to peripheral organs such as lung or liver during viral latency [11 , 25] . Indeed , prior to this study it was shown in adoptive transfer experiments that bulk splenic inflationary T cells have a half-life of 45–60 days [11] . However , in the spleen , there is a considerable variety of phenotypes of inflationary T cells that might influence the measured population half-lives . Here , we quantified the half-life of inflationary M38-specific monoclonal Maxi CD8 T cells in lung tissue—an organ exhibiting pronounced numbers of inflationary CD8 T cells . We demonstrated an antigen-independent maintenance of phenotypically stratified Maxi cells in lung tissue and estimated a half-life of 10–12 weeks that is slightly longer compared to previous data on bulk splenic inflationary T cells [11] . Our data suggest not only that 50% of the inflationary population has to be renewed every 10–12 weeks to guarantee stable numbers , but also that inflationary T cells themselves are rather long-lived , possibly due to relatively high expression levels of the anti-apoptotic molecule Bcl-2 . M38-specific effector CD8 T cells from acute MCMV infection exhibit significantly lower levels of Bcl-2 and consistently show a much shorter half-life and poorer maintenance when transferred into naïve recipients compared to their matured inflationary counterparts . Comparable observations were made with effector and inflationary CD8 T cells exhibiting other specificities , where differences in the expression of many other genes such as anti-apoptotic Bcl-XL have been described [54] . Bcl-XL has been associated with the anti-apoptotic effects of IL-15 on the survival of NK cells [55] . Also IL-15 has been demonstrated to induce mRNA expression of Bcl-XL in human PBMCs [56] . These results emphasise that inflationary CD8 T cells , despite sharing many phenotypical markers with effector cells , are not "just" effector cells but a population of activated T cells that have the ability to seed peripheral organs where they are maintained antigen-independently with a relatively long half-life . This notion is also corroborated by transcriptional profiling of effector , memory and inflationary CD8 T cells [54] . Searching for tissue-intrinsic cues that promote relatively long-term survival of inflationary CD8 T cells , we focussed on a potential role of IL-2 , IL-7 and IL-15 , members of the family of common γ-chain ( γc ) cytokines that are known to promote differentiation and survival of memory CD8 T cells [57] . Neutralization of these cytokines or their receptors during the maintenance phase of inflationary CD8 T cells revealed a unique role of IL-15 in promoting tissue survival of inflationary CD8 T cells . Sensing of IL-15 by inflationary CD8 T cells in vivo was associated with increased Bcl-2 expression , which might directly translate into their increased survival [38 , 58–60] . Also , in vitro , inflationary T cells exhibited STAT5 phosphorylation and increased Bcl-2 expression upon IL-15 stimulation , highlighting their intact responsiveness to IL-15 . Surprisingly , despite IL-7Rα expression on inflationary CD8 T cells , we did not observe any role of IL-7 in the maintenance of these cells . Using bone marrow chimeric mice , we found that IL-15 production by non-hematopoietic cells plays a key role for the maintenance of inflationary T cells in the lung . Our attempts to identify a specific non-hematopoietic cells type being responsible for IL-15 provision , using cell-type-specific IL-15-deficient hosts , remained largely unsuccessful , suggesting that we have either not targeted the right cell types or that there is considerably redundancy of IL-15 production by non-hematopoietic cells in the lung—a notion also supported by our observation of comparable IL-15 production in epithelial cells , blood and lymphatic endothelial cells . One of the stromal cells possibly involved in IL-15 provision are CCL19-producing FRCs , as we observed a trend for reduced survival of inflationary CD8 T cells in hosts deficient of IL-15 production in CCL19-expressing stromal cells . In contrast to a study that suggested pulmonary dendritic cells being an important IL-15 source for maintenance of effector CD8 T cells in the context of an influenza A infection , we did not identify DCs as an important cellular provider of IL-15 [42] . We propose that homeostatic IL-15 expression in non-hematopoietic cells represents a mechanism of inflationary CD8 T cell survival as we did not observe differences in their survival after transfer into naïve or latently MCMV infected hosts . Furthermore , our data suggest that IL-15 signals by non-hematopoietic cells are delivered via trans-presentation of the cytokine to the inflationary CD8 T cells , as adoptive transfer to IL-15Rα-deficient host pheno-copied the results of transfer into IL-15-deficient hosts . Overall , we demonstrate that inflationary CMV-specific CD8 T cells residing in the lung tissue have a half-life of about 10–12 weeks , when external supply is abrogated , indicating that there is substantial need of continuous replenishment to maintain constant numbers as observed in latently CMV infected hosts . Their relatively long half-life is promoted solely by IL-15 , expressed predominantly by non-hematopoietic cells in lung tissue , independent of CMV infection . The survival signals provided by IL-15 are co-dependent on IL-15Rα expression , suggesting that IL-15 acts via trans-presentation . Thus , CMV-driven inflationary CD8 T cell responses constitute a unique unit of "tissue-resident" T cells that is regulated and fuelled differently compared to TRM CD8 T cells elicited upon vaccination or acute infection , offering interesting opportunities for CMV-based vaccine vectors .
This study was conducted in accordance to the guidelines of the animal experimentation law ( SR 455 . 163; TVV ) of the Swiss Federal Government . The protocol was approved by Cantonal Veterinary Office of the canton Zurich , Switzerland ( Permit number 127/2011 , 146/2014 ) . Wild-type C57BL/6J were purchased from Janvier Elevage ( Le Genest Saint Isle , France ) . C57BL/6J , C57BL/6N-Tg ( TcraM38 , TcrbM38 ) 329Biat ( Maxi ) [25] , C57BL/6-Il15ratm2 . 1Ama/J ( IL-15Rαfl/fl ) [40] , B6 . Tg ( CMV-cre ) 1Cgn1/J ( CMV-Cre ) [61] , B6- ( Tg ( Cdh5-cre ) 7Mlia/J ) ( VE-Cadherin-Cre ) [62] , B6 . Cg-Tg ( Itgax-cre ) 1-1Reiz/J ( CD11c-Cre ) [63] , C57BL/6N-Tg ( Ccl19-Cre ) 489Biat ( Ccl19-Cre ) [64] , C57BL/6NTac-IL15tm1lmx ( IL-15-/- ) [65] , B6 . Cg-Gpi1a . Thy1a . Igha/J ( Thy1 . 1 ) [66] mice were housed and bred in specific pathogen-free facilities at the Eidgenössische Technische Hochschule ( ETH ) Hönggerberg . IL-15fl/fl mice containing LoxP sites in the Il15 gene in exon 5 were provided by Dr . K . Ikuta . Maxi transgenic ( Ly5 . 1+ ) express a TCR specific for the MCMV peptide M38316-323 [25] . P14 transgenic ( Ly5 . 1+ ) mice express a TCR specific for the LCMV peptide gp33-41 [34] . All mice were used 6–12 weeks of age and sex-matched within all experiments . Chimeric mice were generated by transferring 2–5 x 106 bone marrow cells derived from Thy1 . 1 C57BL/6J or IL-15-/- mice after lethal irradiation ( 950 rad ) of recipient mice ( IL-15-/- , Thy1 . 1 or C57BL/6J mice ) . Mice were treated in the drinking water with antibacterial Borgal 24% ( Intervet , Boxmeer , Netherlands ) during the first two weeks after reconstitution . Chimeric mice were reconstituted for 8–12 weeks before experiments . Recombinant MCMV lacking m157 ( MCMVΔm157 ) was previously described and is referred to as MCMV in this study [67] . Recombinant MCMV expressing gp33-41 within the M45 gene ( referred to as MCMV-M45-gp33 ) or within the ie2 gene ( referred to as MCMV-ie2-gp33 ) was provided by Dr . L . Cicin-Sain . MCMV strains were propagated on MEFs [68] or M2-10B4 cells [69] as previously described . Virus titres in organs were determined by standard plaque-forming assays on M2-10B4 cells as previously described [69] . Infections were performed intravenously with 106 pfu MCMV or 105 pfu MCMV-ie2-gp33 or MCMV-M45-gp33 . Mice were treated intraperitoneally with depleting / neutralizing antibodies every second day for four weeks . For cytokine neutralization / receptor blockade 500 μg anti-IL-7 ( M25 ) , 200 μg anti-IL-2 ( JESG-1A12 ) and 200 μg anti-CD122 ( TM-β1 ) were injected intraperitoneally . NK cell depletion was performed by i . p . injection of 300 μg anti-NK1 . 1 ( PK136 ) . CD8 T cells from naïve Maxi or P14 mice were purified from splenocytes using anti-CD8α MACS beads ( Miltenyi Biotech , Bergisch Gladbach , Germany ) according to the manufacturer's instructions . 104 Maxi or P14 CD8 T cells were adoptively transferred into recipient mice one day prior to infection . Memory subsets were generated in C57BL/6J mice and isolated at least 60 days post infection . Maxi or P14 TEM ( CD62L- ) from lungs were sorted from donor mice using a BD FACSAria Sorter . Sorted Maxi or P14 TEM ( number varying inter-experimentally between 0 . 5–1 . 0 x 105 cells ) were transferred into new hosts . Lymphocytes were isolated from spleen , lung , liver and lungs as described before [70] . Blood samples were obtained from the tail vein . For intravascular staining , 5 μg fluorophore-coupled anti-CD8α ( 53–6 . 7 ) antibodies were injected intravenously 3 minutes prior to euthanasia as described previously [33 , 71] . Surface staining of cells was performed for 20 min at room temperature in PBS before 5 min treatment with ACK lysis buffer for 5 min at room temperature . In vitro rIL-15 stimulation ( 50 ng/ml ) was performed for 6 hours for pSTAT5 levels or 36 hours for Bcl-2 levels at 37°C ( adapted from [72 , 73] ) . Intracellular staining of Ki-67 and Bcl-2 was performed by fixing and permeabilizing cells using anti-mouse/rat FoxP3 staining set ( eBioscience ) according to the manufacturer's instructions . After fixation with 4% paraformaldehyde and permeabilisation with 90% methanol , intracellular staining of pSTAT5 was performed for 45 min in 1 x permeabilisation/wash buffer from FoxP3 staining kit ( eBioscience ) . Multiparametric flow cytometric analysis was performed using LSRII flow cytometer ( BD Biosciences , Allschwil , Switzerland ) and FACSDiva software . Data was analysed using FlowJo software ( Tree Star , San Carlos , CA , USA ) . For the preparations of the lungs for microscopy , lungs were fixed by infusion of 1% PFA , followed by 20% sucrose ( both 20 minutes ) and finally , optimum cutting temperature ( O . C . T . ) was infused in the lung lumen . The organs were imbedded in O . C . T . and snap frozen in liquid nitrogen . Cryosections of 10 μM were made and air dried . For the staining , slides were re-hydrated in PBS and blocked with 10% rat serum in PBS . Slides were stained with antibodies diluted in PBS containing 1% rat serum and mounted with Mowiol . APC- or PE-conjugated MHC class I tetramers were generated as described before [74] . Recombinant murine IL-15 protein was purchased from eBioscience . Fluorophore-conjugated antibodies were purchased from BioLegend ( Lucerna Chem AG , Luzern , Switzerland ) , eBiosciences ( Thermo Fisher Scientific , MA , USA ) or BD Biosciences ( Allschwil , Switzerland ) . The following antibodies ( clone ) were used for Flow cytometry or fluorescence microscopy: anti-CD8α ( 53–6 . 7 ) , anti-CD8β . 2 ( 53–6 . 8 ) , anti-CD45 . 1 ( A20 ) , anti-CD45 . 2 ( 104 ) , anti-CD90 . 1 ( Ox-7 ) , anti-CD90 . 2 ( 30-H12 ) , anti-CD62L ( MEL-14 ) , anti-CD44 ( IM7 ) , anti-KLRG-1 ( 2F1 ) , anti-Ki-67 ( SolA15 ) , anti-Bcl-2 ( BCL/10C4 ) , anti-CD127 ( A7R34 ) , anti-CD4 ( RM4-5 ) , anti-CD3ε ( 145-2-C11 ) , anti-CD69 ( H1 . 2F3 ) , anti-CD103 ( 2E7 ) , anti-NKp46 ( 29A1 . 4 ) , anti-CD49b ( DX5 ) , anti-NK1 . 1 ( PK136 ) , anti-CD31 ( MEC13 . 3 ) , anti-Podoplanin ( 8 . 1 . 1 ) , anti-EpCAM ( G8 . 8 ) , anti-pSTAT5 ( pY694 ) . Live/Dead Fixable near-IR ( Life Technologies ) dead cell stain was used to exclude dead cells . Lungs and spleens from naïve mice were isolated and single cell suspensions were prepared . Hematopoietic and non-hematopoietic cell populations were sorted based on CD45 expression . RNA isolation was performed using Trizol LS reagnet ( Life Technologies ) as described before . 100 ng RNA was reverse-transcribed using the RT2 HT First Strand cDNA kit ( QiaGEN , Hombrechtikon , Switzerland ) according to the manufacturer's instructions . Real-time PCR was performed using a Rotorgene 3000 instrument ( Corbett Research , Eight Miles Plains , Australia ) to assess SYBR green incorporation ( FastStart Universal SYBR Green Master , Roche Diagnostics , Switzerland ) . The following primer pairs were used for qPCR: IL-15: 5′-CATCCATCTCGTGCTACTTGTG-3′ and 5′-GCCTCTGTTTTAGGGAGACCT-3′ [75] , β-actin: 5'-CCCTGAAGTACCCCATTGAAC-3' and 5'-CTTTTCACGGTTGGCCTTAG-3' . Data were analysed with Rotor-Gene 6000 Series Software ( Qiagen , Hombrechtikon , Switzerland ) . Expression of the housekeeping gene β-acting was used as an internal control for normalisation . Relative expression levels were calculated according to the 2-ΔΔCT method [76 , 77] . Statistical significance was determined as indicated by either non-parametric Mann-Whitney U test or unpaired two-tailed Student's t test using GraphPad Prism ( La Jolla , CA , USA ) . | A majority of the human population is infected with cytomegalovirus ( CMV ) , which results in lifelong persistence due to viral latency . CMV induces remarkably strong and sustained effector memory-like CD8 T cell responses in circulation and peripheral tissues , also referred to as memory CD8 T cell "inflation" . In tissues , these effector memory-like cells contribute to immunosurveillance and early control of CMV reactivation events . Due to the high numbers and effector-like functional properties of inflationary CD8 T cells in peripheral tissues , CMV-based vectors are gaining substantial interest in the context of T cell based vaccines that protect peripheral tissues against infections or tumors . Here , we investigated how the stable peripheral pool of inflationary CD8 T cells is maintained and show that inflationary CD8 T cells are long-lived T cells and have a markedly prolonged half-life compared to effector CD8 T cells . In peripheral organs such as lung , IL-15 cytokine is pivotal in promoting maintenance of inflationary cells by inducing expression of the anti-apoptotic molecule Bcl-2 . We show that IL-15 is mainly expressed by non-hematopoietic cells in lung tissue and that IL-15 is trans-presented to the inflationary CD8 T cells in vivo . Thus , CMV-driven inflationary CD8 T cell responses represent a unique T cell subset in peripheral tissues that is regulated differently compared to TRM CD8 T cells emerging after vaccination or acute infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"methods"
] | [
"blood",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"cytokines",
"spleen",
"immunology",
"microbiology",
"lungs",
"developmental",
"biology",
"respiratory",
"system",
"molecular",
"developme... | 2018 | Tissue maintenance of CMV-specific inflationary memory T cells by IL-15 |
Safely burying Ebola infected individuals is acknowledged to be important for controlling Ebola epidemics and was a major component of the 2013–2016 West Africa Ebola response . Yet , in order to understand the impact of safe burial programs it is necessary to elucidate the role of unsafe burials in sustaining chains of Ebola transmission and how the risk posed by activities surrounding unsafe burials , including care provided at home prior to death , vary with human behavior and geography . Interviews with next of kin and community members were carried out for unsafe burials in Sierra Leone , Liberia and Guinea , in six districts where the Red Cross was responsible for safe and dignified burials ( SDB ) . Districts were randomly selected from a district-specific sampling frame comprised of villages and neighborhoods that had experienced cases of Ebola . An average of 2 . 58 secondary cases were potentially generated per unsafe burial and varied by district ( range: 0–20 ) . Contact before and after death was reported for 142 ( 46% ) contacts . Caregivers of a primary case were 2 . 63 to 5 . 92 times more likely to become EVD infected compared to those with post-mortem contact only . Using these estimates , the Red Cross SDB program potentially averted between 1 , 411 and 10 , 452 secondary EVD cases , reducing the epidemic by 4 . 9% to 36 . 5% . SDB is a fundamental control measure that limits community transmission of Ebola; however , for those individuals having contact before and after death , it was impossible to ascertain the exposure that caused their infection . The number of infections prevented through SDB is significant , yet greater impact would be achieved by early hospitalization of the primary case during acute illness .
The 2013–2016 West Africa Ebola virus disease epidemic reached a scale never seen before . Over 28 , 600 people were infected with Ebola virus disease ( EVD ) and over 11 , 000 died [1 , 2] . Any activity involving direct , unprotected contact between living individuals and an EVD infected individual , living or deceased , are potential mechanisms for onward transmission of the virus . Transmission of EVD occurs when an uninfected individual has direct contact with the blood or body fluids of an EVD infected , symptomatic individual or objects contaminated with their blood or body fluids . Risk of EVD transmission is particularly high during caregiving practices and when bodies are being prepared for funeral rites [3 , 4] . Consequently , relatives , community members and healthcare workers without appropriate personal protective equipment have an increased risk of infection . Despite common modes of transmission , the dynamics of the 2013–2016 EVD epidemic varied widely between and within the most affected countries of Guinea , Liberia and Sierra Leone [5 , 6] . Nevertheless key pillars of the response were systematically implemented across all affected countries . These pillars , common to EVD outbreaks , focused on preventing disease transmission through case management in an Ebola Treatment Center ( ETC ) , contact tracing , social mobilization and safe burials of EVD infected bodies [7 , 8] . Deaths of EVD positive individuals that occur in the community , outside an ETC , pose a serious risk for continued EVD transmission . The infection status of the deceased is frequently unknown at the time of death and there are many opportunities for contact with the deceased . These can result , for example , from local customs including washing , redressing , and burial of the body [9 , 10] . When an EVD positive individual dies in the community , both familial abstention from contact with the deceased , and a safe and dignified burial ( SDB ) carried out by a trained team , are necessary to prevent burial- and funeral-related transmission . When an EVD infected body is buried safely , onward transmission of the virus as a consequence of burial or funeral rituals should cease . A SDB , in the strictest terms , should result in no direct contact with the infected body after the time of death . Consequently , no secondary cases should result from a safe burial , and onward transmission should be limited . The importance of SDB as an integral part of reducing EVD transmission and stopping an outbreak is accepted , but not well understood . While work has been conducted to ascertain the impact of individual EVD interventions on disease transmission , and the importance of community deaths in secondary case generation , most of the data come from limited , focal case time series or relies on data on cases and deaths as published by the World Health Organization ( WHO ) [11–14] . Until this study , neither research focusing specifically on safe burials nor quantification of the direct impact of such activities had been conducted . Using data collected during epidemiological investigations , we estimate the number of secondary cases that were potentially averted by safe burials , and describe risk factors for EVD transmission during funerals and burial rituals ( unsafe burials ) . The potential impact of the SDB program on the 2013–2016 EVD epidemic as a result of activities carried out by the National Societies of the Red Cross in Sierra Leone , Guinea and Liberia , supported by the International Federation of the Red Cross and Red Crescent Societies ( IFRC ) , is also estimated .
Authorization to conduct this research was obtained from the Sierra Leone Ministry of Health and Sanitation , the Guinea Ministry of Health and Public Hygiene and the Liberia Ministry of Health and Social Welfare . Informed consent forms were read and explained to each research participant . Due to restrictive infection prevention procedures , consent to participate was given orally in order to avoid contact between the research team and the key informants . Consent to participate was noted by the research team on the consent form . Participation was voluntary and it was made clear that consent could be withdrawn by the participant at any time and for any reason without repercussion . All epidemiological data were anonymized before being entered into data spreadsheets . All research documents were stored in a locked cabinet or electronically on a password-protected computer with access available only to the research team . Data back-ups were made on external support . The Red Cross conducted approximately 50% of all official SDBs in Sierra Leone , 100% in Guinea and 100% in Montserrado County , Liberia during the 2013–2016 epidemic . Data collection was carried out in rural and urban districts in Sierra Leone , Guinea and Liberia where the National Society in each country , supported by the IFRC ( hereafter referred to as the Red Cross ) were responsible for SDB , as depicted in Table 1 [15–17] . Data collection was carried out from June 18 , 2015 to August 4 , 2015 . Interviews were conducted for four days in each district , except for Montserrado County where they were conducted for eight days . To the furthest extent possible , attempts were made to inform the leaders of the selected villages or urban neighborhoods ( hereafter referred to as community ) of the arrival of the research team at least one day prior to their visit in order to ensure the availability of key informants and receive the necessary approvals from the village chief and community leaders . Prior to data collection in each district , a member of the local Red Cross who was familiar with the area and able to translate between English and the local language ( s ) was trained on the research methodology and data collection tool . Together with the study’s epidemiologist , they visited each community selected and met with community leaders and key informants . Key informants were identified by the Red Cross volunteer in collaboration with community members . One primary key informant was purposively selected in each community to provide information on the burial of interest . This person was generally a family member of the deceased or if a relative was no longer present , another individual identified by the community . Data collection occurred during periods of active EVD transmission in some districts . All activities were therefore conducted in accordance with strict infection prevention procedures that focused on avoiding contact between the data collection team , key informants and other community members . All members of the data collection team were provided with personal protection equipment while in the field including gum ( rubber ) boots , chlorine spray bottles and hand sanitizer containing at least 70% alcohol . In addition , team members were given daily reminders on safety procedures to adopt while in the field to ensure their safety . The sampling frame was comprised of communities ( rural villages and urban neighborhoods ) in priority districts identified by the Red Cross that had experienced at least one of the following events: a ) cases of EVD; b ) community deaths during the period of the epidemic; or c ) having a community member hospitalized in an ETC . If this information was unavailable , a list of all communities in the selected district was used in its place . During selection , one group of 10 communities ( G1 ) was randomly selected from the sampling frame . After communities were selected , key informants were identified in each community and used to ascertain: 1 ) if any cases of EVD had occurred in the community , and 2 ) if an unsafe burial of an EVD case had taken place . If the answer to both questions was yes , the community was included in the final list of communities to be visited by the study team . If at least five communities with unsafe burials were not found in G1 , another group of 10 communities ( G2 ) was selected from the sampling frame until at least five communities that reported an unsafe burial of an EVD infected individual were identified . If the history of an unsafe burial was unable to be reconstructed due to loss of key informants in the community or another cause , another unsafe burial was chosen from the list . If more than one unsafe burial related to the same source case took place in the same community , data were collected from one of the burials , generally the first one that took place . An unsafe burial was considered to be a burial of an individual with suspect EVD infection , buried by their community , family , or manipulated after death but prior to the arrival of the SDB team . The first identified suspect EVD case that was buried unsafely was considered to be the primary case for the purposes of further investigation . A contact was defined as any individual who had physical contact with the body of the deceased , their body fluids or their ( potentially infected ) belongings after death as reported by the key informant . A potential secondary case was defined as a contact that met either the baseline or ceiling definition ( see below ) . These are called “potential” secondary cases because transmission was not directly observed , and it is therefore possible that some secondary cases acquired the infection from a source other than the primary case . Data were collected on a standardized questionnaire , from key informants , regarding the deceased ( primary case ) and their contacts , by the epidemiologist in collaboration with the local Red Cross member ( s ) . For each primary case , details of age , sex , source of infection , EVD swab confirmation and secondary cases in individuals listed as having had contact with the deceased ( contacts ) were collected . For each individual reported to have had contact with the deceased , age , sex , type of contact ( s ) with the body of the deceased , current status ( alive/dead ) and EVD swab confirmation was collected . In addition , details about family ties and relationship to the deceased were collected in Liberia and Guinea . Completed questionnaires were reviewed daily for completeness and data entered into a Microsoft Excel spreadsheet . Data analysis was conducted in Stata 12 ( Stata Corp , College Station , TX ) and R ( R Foundation for Statistical Computing , Vienna , Austria ) . The average number of secondary infections potentially caused by an unsafe burial , and risks of secondary infection associated with different types of contact were estimated using two datasets , both extracted from the epidemiological data collected in the field . The first dataset from which results for the baseline estimate were produced consisted of only EVD positive cases ( as reported by the key informant ) , and contacts were limited to those that only had contact with the primary case after death ( i . e . caregivers were excluded ) . The second dataset from which results for the ceiling estimate were produced consisted of all data collected and included anyone who became sick or died following contact with the primary case ( before or after death ) as reported by the key informant . The number of secondary cases potentially averted was calculated as the average number of potential secondary cases per unsafe burial multiplied by the number of EVD positive burials carried out by the Red Cross . The number of SDB carried out by Red Cross SDB teams , by country , is presented as reported by the Red Cross . The number of SDB that were laboratory positive corresponds to the number of deaths reported from October 2014 to April 2015 in Sierra Leone for which laboratory results were available and matched based on name , age and sex using a computer algorithm ( ~75% success in matching ) in addition to all positive community deaths in Guinea and Montserrado County , Liberia . Pooled and district specific estimates are presented with medians and interquartile ranges . Estimates of risk factors among contacts of a primary case are presented as odds ratios based on two-sided Fisher’s exact test and presented with 95% confidence intervals . In order to construct the “cared for” odds ratio , the baseline case definition was altered to include caregivers otherwise the case definitions for the baseline and ceiling estimates remained the same for the risk analysis , with the baseline estimate excluding caregivers .
Across the study , 203 contacts , 65% of those recorded , were reported to have become sick after having contact with the body of the primary case during an unsafe burial ( Table 3 ) . This ranged from 89% in Western Area Rural ( Sierra Leone ) to 33% in Kambia ( Sierra Leone ) . Forty-one percent ( 83/203 ) of individuals that were reported to have fallen sick had contact with the primary case only after death; however , the key informant was not always aware if EVD infection had been confirmed by laboratory testing . The post-contact health status ( sick or EVD confirmed ) was unknown by key informants for nine contacts . For those whose health status was known , 25% ( 78/301 ) were reported to have fallen ill and have had a laboratory confirmed EVD infection ( Table 2 ) . Investigations at the household level revealed that contact was not only limited to contact with the primary case after death . Contacts that occurred after death were classified into seven categories ( Table 4 ) . Key informants frequently reported that some individuals had more than one type of contact with the primary case . Individuals who had contact with the primary case after death were frequently reported ( 46% , 142/310 ) to have had contact with the same individual during the acute phase of their illness . Twenty-three contacts ( 7% , 23/310 ) were reported to have protected themselves when having contact with the primary case during acute illness or after death . Use of protection was most frequently reported in Kailahun ( Sierra Leone ) ( 16% of contacts , 12/77 ) and Montserrado County ( Liberia ) ( 11% of contacts , 8/76 ) . The exposure that was most strongly predictive of secondary transmission was having contact with fluids ( e . g . blood , vomit ) of a primary case , followed by direct physical contact with a primary case during their acute illness ( Fig 1 ) . Overall , in the baseline estimate , 9 . 5% ( +/- 1 . 6% s . e . ; n = 45 ) of contacts became secondary cases , meaning that these contacts had exposures and outcomes that met the baseline case definition ( not caregivers , and were reported to have become infected with EVD ) . In the ceiling estimate , 68% ( +/- 2 . 6% s . e . ; n = 45 ) of contacts became secondary cases , meaning that these contacts either became sick or died after their exposure . In the baseline estimate , all laboratory confirmed EVD cases reported some form of contact after death . This precludes a finite estimate of the odds ratio associated with contact after death in the baseline estimate ( the lower bound of the 95% confidence interval on the estimate was 8 . 0 ) . In the ceiling estimate , the odds ratio was 0 . 16 ( 95% CI: 0 . 089–0 . 30 ) . This reflects that many contacts in this estimate were considered EVD positive , but had contact both before and after death . People who had contact with the primary case before death ( caregivers ) were , on average , between 2 . 63 ( 95% CI: 1 . 15–4 . 63 , baseline estimate ) and 5 . 97 ( 95% CI: 3 . 31–11 . 17 , ceiling estimate ) times more likely to become EVD infected , relative to people who only had contact with the primary case after death . For data on fluid contact , a small sample size caused the upper confidence bound to exceed 300 , but this is a sampling artifact and should not be interpreted as containing information on risk patterns . In the baseline estimate , a single unsafe burial was associated with an average of 0 . 64 ( 1 . 46 s . d . , n = 45 ) secondary cases ( Fig 2 ) . This varied among districts from 0 . 125 in Guéckédou ( Guinea ) to 3 . 00 in Western Area Rural ( Sierra Leone ) ( Fig 3 ) . In the ceiling estimate , a single unsafe burial was associated with an average of 4 . 74 ( 4 . 27 s . d . , n = 45 ) secondary cases with variation among districts ranging from 1 . 2 ( 1 . 09 s . d . , n = 6 ) in Kambia ( Sierra Leone ) , and 6 . 25 ( 7 . 39 s . d . , n = 6 ) in Guéckédou ( Guinea ) ( Fig 3 ) . In both the baseline and ceiling estimates there was no statistically significant evidence of systematic variation among districts; just evidence of an abundance of variability , likely reflecting latent heterogeneities in transmission , and the stochastic nature of the underlying disease transmission process . A total of 47 , 505 SDB were reported to have been carried out by the National Societies in Sierra Leone , Guinea and Liberia from 2013–2016 . The number of burials completed varied greatly by country ( Table 5 ) . When laboratory results were received and matched with the individuals buried , 2 , 205 ( 4 . 6% ) of the bodies buried tested positive for EVD in the laboratory . The relatively small number of bodies buried in Liberia , compared to Sierra Leone and Guinea , is related to the areas the Red Cross was responsible for SDB . In Sierra Leone and Guinea , the National Societies were responsible for burials in the entire country , while in Liberia the area of intervention was limited to Montserrado County only . Fig 4 presents estimates of the total number of reported burials carried out by the Red Cross in addition to estimates of the number of these burials that were EVD positive , contrasted with the reported cases from the WHO patient database for Sierra Leone , Liberia and Guinea . Applying the analysis presented above to the number of positive EVD cases buried , the number of secondary cases potentially caused by a typical unsafe burial leads to estimates for the number of secondary cases potentially averted by the SDB program . Using this method we estimate that between 1 , 411 ( lower estimate ) and 10 , 452 ( upper estimate ) secondary EVD cases may have been prevented by the SDB program . This represents a reduction in the total size of the 2013–2016 EVD epidemic of between 4 . 9% ( lower estimate ) and 36 . 5% ( upper estimate ) . Thus , the total number of EVD cases in the 2013–2016 epidemic could have been 30 , 012 to 39 , 053 if the SDB program had not been implemented by the Red Cross in Sierra Leone , Guinea and Liberia .
The majority of the unsafe burials investigated seemed to be relatively representative of classic EVD transmission with between 0 . 64 to 4 . 74 secondary EVD cases associated with unsafe burials , an overall average of 2 . 58 . Estimates of the average number of secondary cases potentially resulting from an unsafe burial are significant although there is variation depending on which estimate , baseline or ceiling , was used . This variation and the use of two estimates highlight the uncertainty surrounding the dynamics of EVD transmission . Additionally , both the number of individuals reported to have had contact with the primary case , and the type of contact they had during an unsafe burial , ranged widely by district . This variation may be due to many factors including the incidence of EVD in the district and the point in the epidemic during which the unsafe burial occurred . Accordingly , the highest median number of contacts per burial in this research was reported in Guéckédou ( Guinea ) , the first district affected by the EVD epidemic , and the lowest in Kambia ( Sierra Leone ) , the last of the sampled districts to be affected . This variation may reflect changes in behavior after exposure to EVD related messaging by Red Cross and other partners , and familiarity of individuals and communities with the disease . The type of contact that was had with a primary case and the timing of that contact , before or after death , are important factors in EVD transmission . Such information is normally collected during field outbreak investigations for the purposes of contact tracing and to ascertain epidemiological links to known cases . In the early days of the EVD epidemic , analysis of contact tracing data showed that of 701 confirmed EVD cases 67 ( 10% ) reported contact with EVD infected individuals after death ( only ) , while 148 ( 21% ) reported contact both before and after death [18] . Furthermore , some of the potential secondary cases identified in our study may have been infected by a source other than the primary case . In these cases , preventing exposure to the primary case would still reduce the risk of onward transmission , but would not be guaranteed to avert the secondary case altogether . Risk estimates from data presented here show , similar to previous studies , that having contact with the primary case both during illness and after death significantly increased the risk of EVD transmission compared to individuals who had contact with the primary case only after death [3 , 14 , 19] . Further analyses of risk factors for EVD transmission highlight the importance of having direct contact with the blood or body fluids of the deceased . Although our sample size was too small to provide a reliable estimate of risk due to such contact , exposure to blood and body fluids of the deceased was frequently cited by key informants and cannot be discounted as an important factor in transmission , as documented in previous risk factor analyses [3 , 19] . Reports of protective measures taken while having contact with a primary case varied in frequency and method , reflecting distinct differences in EVD awareness and behavior . Knowledge , attitudes and practices ( KAP ) surveys carried out at different points in the epidemic also highlighted geographic differences regarding EVD and its transmission [20 , 21] . Significant differences were noted in areas ( defined differently between surveys ) that had a higher incidence of EVD cases when compared to areas with lower EVD case incidence . While data collected through KAP surveys are subject to limitations , the differences between areas remains illustrative . Not only are the processes that control changes in the number of individuals infected with EVD by place or over time not well understood , the number of individuals infected during the West African epidemic is also not certain due to significant underreporting [2 , 22] . Furthermore , neither burials carried out by other organizations during the epidemic , nor burials conducted early in the epidemic by the Red Cross are taken into consideration in this estimation . Systematic data collection by Red Cross burial teams only began later in the response ( July 2014 in Liberia , October 2014 in Sierra Leone , and January 2015 in Guinea ) , laboratory data were difficult to obtain in some areas and data from other organizations involved in SDB were unavailable resulting in an underestimation of both overall SDB program impact and the impact of the Red Cross SDB program in particular . The impact of an SDB program is highly dependent on disease prevalence at the time of program initiation . As seen in the data from Guinea , it appears that initiating SDB later in the epidemic when there are fewer EVD cases results in a dramatic reduction in program impact because the bodies buried will primarily be EVD negative . The Red Cross burial program began early in April 2014 in Guinea; yet , systematic collection of samples from the deceased only began in January 2015 when the prevalence of EVD was lower . As a result , the Red Cross program in Guinea is estimated to have buried mostly EVD negative bodies as reflected in the low numbers of confirmed laboratory positive bodies . This contrasts with Liberia where less than 20% of t burials were carried out by the Red Cross; however , twice the number of positive bodies were buried . This is likely a consequence of the high prevalence of EVD when SDB was initiated in Liberia , and an EVD positivity that was higher overall . The impact of SDB on preventing chains of transmission was not investigated and may have been particularly relevant in the first months of the epidemic when the reproductive number was greater than one [5 , 18] and the epidemic was spreading quickly in communities . Furthermore , the level of knowledge about EVD varied over the course of the epidemic , as did communities’ sense of fear and/or distrust of EVD responders [9 , 10 , 20 , 21 , 23–27] . In such situations , each secondary case averted could be the first of many cases in an averted chain of transmission likely composed of multiple transmission generations and secondary cases . In Liberia , for example , one to seven transmission generations were documented in a single transmission chain with 4–35 secondary cases produced [14] . The estimates presented here correspond to the first transmission generation after the primary case and do not consider the impact that SDB had on chains of transmission averted by preventing cases beyond the first transmission generation . Consequently , the number of cases averted by SDB programs may be considerably higher than that which was directly estimated . None of the investigations of unsafe burials of suspect EVD infected individuals reported here document a superspreading event , an event such as an unsafe burial that results in a disproportionately high number of secondary transmission ( infection ) to contacts . Nevertheless , SDB may have had an even greater impact on the epidemic by preventing superspreading events . Indeed , modeling suggests that such events may have sustained EVD transmission in Sierra Leone [28] . These events have been infrequently documented but have been known to occur after the death of an influential person in the community ( e . g . elder or traditional healer ) , such as in Kissidougou ( Guinea ) where an unsafe burial of a male midwife assistant responsible for circumcisions in the community , was reported to have caused upwards of 85 secondary cases [29 , 30] . While superspreading events are relatively rare , if SDB prevented even one superspreading event , the impact on averted chains of transmission would be significant . Although such events were not identified during this research , both anecdotal reports and reports in peer reviewed literature suggest that superspreading events associated with unsafe burials did occur during the 2013–2016 outbreak [29 , 31] . The data presented here do have a number of limitations , including recall bias . As unsafe burials were investigated , the primary case was deceased and all information was collected from key informants . Age of the contact was frequently unknown and unable to be estimated with any certainty , particularly by key informants external to the family unit . Data were collected retrospectively and occasionally the unsafe burial investigated had occurred more than 12 months prior to the date of the interview , as was the case in Guéckédou ( Guinea ) and Kailahun ( Sierra Leone ) . Conversely , as in Kambia and Western Area Rural ( Sierra Leone ) , some unsafe burials had recently occurred and interviews with key informants were conducted while the individuals were still in quarantine . Frequently , key informants had themselves been infected with EVD and hospitalized due to the unsafe burial that was being investigated . In these cases it was difficult to collect data on the health outcomes of individuals who had participated in the unsafe burials if the outcome occurred during the hospitalization of the key informant . In some instances information could not be collected from familial key informants as they had died prior to the visit of the study teams . In such cases , additional key informants were identified . Furthermore , the risk factor analysis is limited by the amount of data that we were able to collect from key informants . Recall of caregiving and funeral practices was sometimes limited or not available . In addition , the status of some contacts , sick or EVD confirmed , was unknown if they did not return to their communities or share the information . In the absence of laboratory confirmation of EVD infection for all contacts reported to have become cases ( fallen sick or had an EVD infection ) , we only documented the first transmission generation from the identified primary case . To the furthest extent possible , however , we attempted to triangulate the information collected during interviews in order to ensure the quality of the data and ensure it was exhaustive as possible . Finally , not all laboratory results were available or could be matched to the burials carried out , and the period for which results were available was limited . Here , post-mortem transmission of EVD is investigated exclusively and the implications evaluated systematically for the first time . These findings underscore the substantial impact that SDB had on the 2013–2016 EVD epidemic with regards to the number of cases averted and the importance of early initiation of SDB activities during an EVD response . Previously , the importance of SDB as an integral part of reducing EVD transmission and stopping epidemics was acknowledged as a fundamental EVD control measure in terms of its impact on ending transmission , however its impact was not well understood . The timing of SDB implementation and scale-up is also important and will affect the magnitude of the outbreak . The earlier in an EVD outbreak that a SDB program can be implemented at scale , the greater the number of cases averted and consequently the larger the program’s impact on the outbreak will be . Additional research is needed in order to better understand the variation of risk for EVD transmission related to distinct care practices both before and after death . The SDB program is a cornerstone of comprehensive EVD prevention and response programs and must be implemented as a holistic intervention incorporating both public health and socio-cultural perspectives in order to enable quick scale up and achieve maximum impact . | The care of an individual infected with Ebola virus disease ( EVD ) , their death , funeral , and burial in the community rather than in an Ebola Treatment Center ( ETC ) poses a serious risk for continued disease transmission . Consequently , SDB is an essential component of EVD outbreak response; however , its impact on transmission is not well understood . During the 2013–2016 EVD epidemic the Red Cross carried out over 50% of the official burials in Guinea , Liberia and Sierra Leone . We performed epidemiological investigations in EVD affected communities to better understand disease transmission linked to unsafe burials of ( suspect ) EVD infected individuals , and risk factors for transmission linked to caring and burial practices . An average of 2 . 58 secondary cases were potentially generated per unsafe burial investigated and varied by district ( range: 0–20 ) . Additionally , the Red Cross SDB program potentially averted between 1 , 411 and 10 , 452 secondary EVD cases , reducing the epidemic by 4 . 9% to 36 . 5% . Our results quantify for the first time the potential impact this essential EVD response component had on the 2013–2016 epidemic and highlight the importance of SDB as a fundamental control measure , while also underlining the well-known importance of isolating EVD infected individuals as soon as they show symptoms in order to limit transmission . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"guinea",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"infectious",
"disease",
"epidemiology",
"geographical",
"locations",
"tropical",
"diseases",
"spatial",
"epidemiology",
"ebola",
"hemorrhagic",
"fever",
"neglected",
"tropical",
"diseases",
"africa",
... | 2017 | Estimating the number of secondary Ebola cases resulting from an unsafe burial and risk factors for transmission during the West Africa Ebola epidemic |
Many research questions in visual perception involve determining whether stimulus properties are represented and processed independently . In visual neuroscience , there is great interest in determining whether important object dimensions are represented independently in the brain . For example , theories of face recognition have proposed either completely or partially independent processing of identity and emotional expression . Unfortunately , most previous research has only vaguely defined what is meant by “independence , ” which hinders its precise quantification and testing . This article develops a new quantitative framework that links signal detection theory from psychophysics and encoding models from computational neuroscience , focusing on a special form of independence defined in the psychophysics literature: perceptual separability . The new theory allowed us , for the first time , to precisely define separability of neural representations and to theoretically link behavioral and brain measures of separability . The framework formally specifies the relation between these different levels of perceptual and brain representation , providing the tools for a truly integrative research approach . In particular , the theory identifies exactly what valid inferences can be made about independent encoding of stimulus dimensions from the results of multivariate analyses of neuroimaging data and psychophysical studies . In addition , commonly used operational tests of independence are re-interpreted within this new theoretical framework , providing insights on their correct use and interpretation . Finally , we apply this new framework to the study of separability of brain representations of face identity and emotional expression ( neutral/sad ) in a human fMRI study with male and female participants .
A common goal in perceptual science is to determine whether some stimulus dimensions or components are processed and represented independently from other types of information . In visual neuroscience , much research has focused on determining whether there is independent processing of object and spatial visual information [1] , of object shape and viewpoint [2] , of different face dimensions [3 , 4] , etc . A common approach is to use operational definitions of independence , which are linked to rather vague conceptual definitions . This approach has the disadvantage that different researchers use different operational definitions for independence , often leading to contradictory conclusions . For example , in the study of whether face identity and emotional expression are processed independently , evidence for both independence and interactivity has been found across a variety of operational tests . Evidence for independence was found by most lesion studies [5] , by lack of correlation between fMRI patterns related to identity and expression [6] , by single neuron invariance [7] , by selective fMRI adaptation release in fusiform face area ( FFA ) and middle superior temporal sulcus ( STS ) [8] , and by selective fMRI decoding of identity from anterior FFA and medial temporal gyrus [9 , 10] , and of expression from STS [10] . Evidence for a lack of independence has been provided by overlapping fMRI activation during filtering tasks [11] , by non-selective fMRI adaptation release in posterior STS [8] and in FFA–when adaptation is calculated based on perception [12]– , and by non-selective fMRI decoding from right FFA [9] . Because the different operational definitions are not linked to one another through a theoretical framework , the interpretation of such contradictory results is very difficult and necessarily post-hoc . Even more difficult is to link the neurobiological results to the psychophysics literature on independence of face dimensions , which itself is plagued by similar issues ( for a review , see [13] ) . General recognition theory ( GRT ) [14 , 15] is a multidimensional extension of signal detection theory that has solved such problems in psychophysics , by providing a unified theoretical framework in which notions of independence can be defined and linked to operational tests . Hundreds of studies have applied GRT to a wide variety of phenomena , including face perception [16 , 17] , recognition and source memory [18 , 19] , source monitoring [20] , object recognition [21 , 22] , perception/action interactions [23] , speech perception [24] , haptic perception [25] , the perception of sexual interest [26] , and many others . Here we present an extension of GRT to the study of independence of brain representations , by relating it to encoding models and decoding methods from computational neuroscience [27 , 28] . Past neuroimaging studies have been limited to choosing between decoding methods , which try to determine what stimulus information is processed in a brain region while ignoring the form of the underlying representation , and encoding models , which assume a specific representation and compare its predictions against data . [29] . We propose the concept of encoding separability as a fundamental way in which brain representations of stimulus properties can be considered independent , and we identify the specific conditions in which a decoding analysis of neuroimaging data or a psychophysical study allow inferences to be made about encoding separability . In doing so , we show that decoding methods ( and under some assumptions , psychophysics ) can be useful to make valid inferences about encoding . We also re-interpret previously-proposed tests of independence within our new theoretical framework , and provide guides on their correct use . Finally , we apply this new framework to the study of separability of brain representations of face identity and expression .
GRT is a multivariate extension of signal detection theory to cases in which stimuli vary on more than one dimension [14 , 15] . As in signal detection theory , the theory assumes that different presentations of the same stimulus produce slightly different perceptual representations . For example , as shown in Fig 1 , repeated presentations of a face identity produce a variety of values on the “identity” dimension ( orange and red dots ) , which follow a probability distribution ( red and orange curves ) . According to GRT , there are many ways in which processing of a dimension of interest , or target dimension , can be influenced by variations in a second , irrelevant dimension . GRT formally defines such dimensional interactions and links them to operational tests of independence . This allows researchers to determine whether a test can dissociate between different forms of independence , and to create new tests specifically designed to target a specific form of independence . Here we will consider the special case in which stimuli vary along two stimulus dimensions ( or more generally , components or properties ) , represented by A and B . However , the theory can easily be extended to a larger number of dimensions . Specific values of dimension A used in an experiment are indexed by i = 1 , 2 , …LA , and the specific values of dimension B are indexed by j = 1 , 2 , …LB . A stimulus in the experiment is represented by a combination of these dimension levels , AiBj . This stimulus produces a random perceptual effect in a two-dimensional perceptual space [x , y] , where x represents the perceptual effect of property A and y the perceptual effect of property B . The random vector [x , y] can be described through a two-dimensional joint probability density p ( x , y|AiBj ) , with p ( x|AiBj ) and p ( y|AiBj ) representing the marginal densities of the perceptual effects associated with components A and B , respectively ( the distributions shown in Fig 1 are examples of such marginal densities ) . Here we summarize the previous theoretical results , with an emphasis on how they can be applied to the empirical study of perceptual independence by psychologists and neuroscientists . A summary of all the theoretical relations described in previous sections can be seen in Fig 5 . First , because perceptual separability can be considered a form of decoding separability , and due to the relations summarized in Fig 3 , any failure of perceptual separability should be reflected in a failure of encoding separability somewhere in the brain . This means that any psychophysical study reporting a failure of perceptual separability provides a hypothesis to be tested by a neuroscientific study: that a corresponding failure of encoding separability should be found , probably in sensory areas thought to encode the target dimension . Second , such neuroscientific studies can be performed using direct measures of neural activity , such as those provided by single-cell recordings or local field potentials , or indirect measures of neural activity contaminated by measurement error , such as those provided by EEG and fMRI . Using traditional linear decoding strategies on indirect measures of neural activity , the decoded dimensional values still offer a basis for a valid test of decoding separability , and any violation of decoding separability found within a given brain region reflects a violation of encoding separability by the neural population in that region . It must be stressed that a failure of encoding separability is a valid inference that can be made from decoding of neuroimaging data , but such data do not provide a basis to make any strong inferences about the presence of encoding separability . A weak inference can be made , based on the lack of evidence for a violation , but this is analogous to accepting the null in a traditional statistical test . A relatively stronger inference of encoding separability could be made on the basis of assumptions about the neuroimaging measurement model , but researchers should clearly identify such assumptions . Our recommendation to researchers is to be cautious about concluding that separability ( or “invariance” ) holds at the neural level from neuroimaging data , or even from decoding of direct measures of neural activity ( e . g . [61]; a related point was made in [62 , 63] ) . Finally , we have shown that operational tests of independence available in the literature can be formally defined and re-interpreted within the framework presented here . We showed that , when some strong assumptions are met , the measurement vector orthogonality test [6 , 50 , 52] is related to the concept of perceptual independence from the traditional GRT , but it is unlikely to be related to a corresponding property of stimulus encoding . On the other hand , the classification accuracy generalization test promoted by Anzellotti and Caramazza [55–57] can lead to valid inferences about encoding separability . However , the way in which the test has been applied might lead to conclusions of invariance or separability that are in general unjustified , unless one is interested in decoding separability only , and not in the separability of underlying brain representations . In addition , the classification accuracy generalization test is likely to provide less information than our decoding separability test . The MGLM approach proposed by Allefeld and Haynes [58] suggests an alternative way to indirectly test encoding separability in neuroimaging . The resulting pattern difference invariance test seems like a valid test of violations of encoding separability , but it is based on strong assumptions about the distribution of the neuroimaging data that are not necessary when the decoding separability test is applied . Here we present an application of our framework to the study of encoding separability of face identity and expression . This application serves as a way to illustrate the kind of question that this framework can help answer and the concrete steps that researchers should take to apply the framework in their research . Information about a number of properties can be extracted from a single face , including identity and emotional expression . The influential model of Bruce and Young [3] proposed that these two face dimensions are processed independently , motivating a large number of psychophysical studies aimed at testing this hypothesis [13 , 33–35 , 64–72] . Neurobiological theories of visual face processing [4 , 73] also propose relatively independent processing of face emotion and identity , through anatomically and functionally differentiated pathways . A ventral pathway projecting from the occipital face area ( OFA ) to the fusiform face area ( FFA ) would mediate the processing of invariant aspects of faces , such as identity . A dorsal pathway projecting from the OFA to the posterior superior temporal sulcus ( pSTS ) would mediate the processing of changeable aspects of faces , such as emotional expression . Recent reviews [74 , 75] conclude that the two pathways are indeed relatively separated and functionally differentiated , with the ventral pathway being involved in the representation of face form information–including invariant aspects of face shape such as identity– , and the dorsal pathway involved in the representation of face motion information–including rapidly changeable aspects of faces such as expression . According to this revised framework , both identity and expression information may be encoded in either pathway , but exactly what information about each dimension is encoded would differ between pathways . The psychophysical and neurobiological lines of research in this area have remained relatively independent across the years , with no attempt to integrate results across levels of analysis despite the similarity of the central questions guiding their research . In addition , both lines have relied largely on operational definitions of independence that , while having face validity , are usually not linked to any theoretical definition . As indicated in the introduction , this approach makes it difficult to interpret contradictory results . Thus , the study of independence of face identity and expression is a particularly good example of an area in which our extended GRT framework can provide helpful research tools . Our theory can provide a much-needed theoretical integration across levels of analysis and tests , as well as more rigorous definitions of independence and ways to measure it . We have recently performed a GRT analysis of psychophysical data to study the perceptual separability of identity and expression [13] . The results suggested that , for the stimuli used in that study and after accounting for decisional factors , emotional expression was perceptually separable from identity , but identity was not perceptually separable from emotional expression . From these results , our current framework ( see Fig 3 ) predicts that encoding separability of identity from expression must fail somewhere in the areas representing face information , and that we should be able to find evidence of failures of decoding separability in those areas . The predictions regarding encoding separability of emotional expression are less straightforward: as there are no violations of perceptual separability in the behavioral data , violations of encoding separability seem unlikely , but are still possible . Here , we acquired fMRI data from participants while they looked at the same stimuli and completed the same task as in our previous psychophysical study ( see Materials and methods ) . The stimuli were images of four faces , which resulted from two different male identities showing two different emotional expressions ( neutral and sad ) . Participants performed a simple stimulus identification task . In each trial , a single stimulus was flashed in the screen and the participant had to identify the specific combination of identity and emotional expression that had been shown . This required participants to pay attention to both identity and expression to attain good performance . The participants received feedback about the correctness of their responses . The task was given to participants in runs that lasted around 10 minutes , during which each image was repeated 25 times . Participants completed three of such runs , and in addition they completed a standard functional localizer run [76] that allowed to obtain the approximate location of face-related regions . Performance in the task during the scanning session was high , with a mean of 81 . 67% ( SE = 5 . 18% ) . Single-trial estimates of stimulus-related activity were used as input to the decoding separability test described earlier . Because we did not have specific hypotheses about the location of areas showing failures of encoding separability , we performed a whole-brain searchlight analysis [77] , to determine which small circular regions ( radius of 3 voxels ) showed violations of decoding separability , and therefore violations of encoding separability . To spatially localize violations of encoding separability relative to areas in the face network , we found such areas with the help of a standard functional localizer . The results from this analysis did not reveal any significant violations of decoding separability , either for identity or emotion . Further exploration revealed that our standardized DDS index was consistently negative in the full-brain maps , suggesting that our method of standardization might have produced an index that was too conservative . That is , the DDS was standardized to represent a percentile value ( ranging from 0 to 100 ) re-centered around the middle of the distribution ( i . e . , ranging from -50 to 50 ) . Under the null hypothesis of decoding separability , the distribution of this DDS would be driven only by noise in the data , and we would expect the standardized measure to hover around zero , with similar areas of the brain maps being positive and negative . On the other hand , we would not expect values consistently lower than zero , as this would mean that the estimated decoding distributions are consistently more similar to one another than expected under decoding separability . As the expectation under decoding separability is that the distributions are identical , this seemed like a problem . We reasoned that one solution would be to use the difference in DDS index between the identity and emotion analyses as the main test statistic , to allow one map to serve as control for the other . We must underscore that this is an exploratory analysis , and its results should be confirmed by future research . Increasing power with a larger sample ( either a larger number of participants or a larger number of trials ) would be helpful to obtain reliable results with a conservative test . Fig 6 shows the main results of this analysis , displayed over a flat cortical map . Face-selective areas found through the functional localizer are outlined in the figure . Outlined in green are face-selective areas showing higher activity during the presentation of faces than during the presentation of other objects . Outlined in red are areas showing higher activity during the presentation of emotional faces than during the presentation of neutral faces . The figure also shows clusters of significant violations of decoding separability , depicted in red-yellow for the identity > emotion contrast . A single large cluster ( 483 2mm voxels ) was found to be significant , covering parts of the left STS and superior temporal gyrus ( peak location in MNI coordinates: -60 , -14 , 2 ) . This cluster only slightly overlapped with an area of the face network in the pSTS ( green contour ) . No significant violations were found for the emotion > identity contrast . The results shown in Fig 6 are in line with the previous psychophysical results [13] and the relations depicted in Fig 3 , as they provide evidence of stronger violations of decoding separability for identity than for emotional expression , but not the other way around . This asymmetry in the separability of neural representations is analogous to the asymmetry in perceptual separability found in our previous psychophysical study , and thus makes intuitive sense . Although this asymmetry was not a strong prediction from the theory ( which simply predicts violations of decoding separability for identity , but is ambiguous about violations of decoding separability for emotional expression ) , it suggests that there is at least an empirical correspondence between asymmetries of separability in perceptual and brain representations .
We have shown that a decoding separability test operating on indirect measures of neural activity can validly detect violations of encoding separability , but one of the conditions in our treatment of this issue was using a linear decoder . When a linear decoder is used , the relation between the target difference between decoding distributions δ ( A ^ r ) and the measured difference δ ^ ( A ^ ) is straightforward ( see Eq 26 ) , which allows us to know exactly what violations of decoding separability the test can and cannot detect . We believe that the use of a linear decoder is a reasonable requirement for the test , as they are easier to use and interpret than non-linear decoders , and decoding studies in neuroimaging have almost exclusively used linear decoders . However , one open question is to what extent using a different decoder might change the test’s ability to detect specific violations of encoding separability . That is , perhaps a specific type of violation of encoding separability is hidden by a linear decoder , but shown by a non-linear decoder , or vice-versa . This question requires considerable additional theoretical and simulation work . However , we do know that , regardless of what type of decoder is used , if encoding separability holds , then decoding separability must necessarily hold . Thus , finding a violation of decoding separability with any decoder reflects either a violation of encoding separability or a statistical error . Thus , one strategy that might prove useful in the future is performing several decoding separability tests , each using a different decoder . However , such a test should fulfill two requirements: ( 1 ) the decoders should be chosen based on previous work showing that they can detect different violations of encoding separability , and ( 2 ) a correction for multiple tests should be applied , to control for the family-wise type I error . Similarly , we linked encoding separability violations to a pattern difference invariance test by assuming a linear measurement model . This was helpful to prove that in general a violation of encoding separability may or may not lead to a violation of pattern difference invariance , but it is not clear whether some specific measurement models yield a different result . Importantly , it is unlikely that the true measurement model linking neural activity and neuroimaging measurements is simply linear . Thus , more work is necessary to reach a better understanding of exactly what violations of encoding separability can and cannot be detected using the pattern difference invariance test . Faced with the problem of operationalism in the study of neural independence , here our approach has been to propose a very general theoretical framework in which most operational tests can be interpreted and related to properties of encoding such as separability . A different approach would involve building , fitting and selecting among competing encoding models [27] . More specifically , this approach requires building several different encoding models , choosing in each case features such as the number of channels , the shape of the tuning functions , the distribution of neural noise , etc . Encoding separability would hold for some of these models and it would fail for others . The output of each model in terms of neural activities must be linked to neuroimaging data ( e . g . , estimates of activity from fMRI or EEG ) through a formal measurement model . This formal link would allow to derive or numerically approximate a probability distribution of the data given a particular model and parameter set . Once data are obtained , this probability distribution can be used to estimate the parameters of the encoding model and the measurement model that provide the best fit to the data , through maximum likelihood estimation ( or Bayesian inference , if priors are added ) . After estimation , standard model selection procedures ( e . g . , by AIC , BIC or predictive accuracy in a cross-validation scheme ) would allow to determine what model provides the best explanation for the data . The properties of the chosen encoding model , including its status regarding encoding separability , provide the best explanation for the data . This approach allows to explicitly test specific features of encoding , and some researchers argue that encoding modeling is the best way to reach valid conclusions about representation in a given brain region from neuroimaging studies [80 , 81] . Here we have shown that valid inferences about representation can be made from decoding studies , but we do believe that answering specific questions about representation may be easier through encoding modeling . However , there are three important challenges faced by anybody wanting to apply encoding models in this way . The first challenge is that this approach will pick the best model among those tested . Thus , a poor selection of competing models will lead to the wrong inferences . Building and fitting encoding models in this way requires an important level of knowledge about what stimulus properties are encoded and how they are encoded [80] . Relatedly , fitting specific models may allow to draw more precise inferences regarding encoding separability , but such inferences cannot be generalized to situations outside those included in the tested models . In contrast , a failure of decoding separability signals a failure of encoding separability regardless of the unknown specific details of the encoding and measurement models . The second challenge is that the process of fitting the models itself may require considerable technical expertise and computational resources . Likelihood functions must be derived or numerically approximated for each model , problems of model mimicry and identifiability must be assessed and solved , simplifying assumptions and constraints must sometimes be placed on the models . The consequences of decisions regarding each of these issues–and the way in which they affect inferences–might not be clear for researchers that are not experienced with modeling . The third challenge has to do with inference and interpretation . It is not always very clear what can and cannot be concluded from the fit of encoding models to data , and recent work suggests that common interpretations of the results of encoding modeling are incorrect [82 , 83] . This is complicated further by the fact that many researchers using encoding models are not explicit about their modeling assumptions . For example , many applications involve using multiple linear regression with least squares estimation of weight parameters , where the independent variables are stimulus features assumed to be encoded and the dependent variable is the measured activity in an fMRI voxel or EEG channel . Although never explicitly stated , the assumption behind these models is that there is no neural noise ( independent variables in linear regression are not random ) , the measurement model is linear , and measurement noise is Gaussian and additive . Any conclusion reached using these popular models must be qualified by this set of assumptions . Overall , we believe that encoding modeling is an excellent way to study the properties of neural encoding using neuroimaging . However , for the reasons outlined before , it seems unlikely to be adopted by experimentalists without a computational background . Indeed , researchers without such a background are probably wise to keep away from it . On the other hand , we have shown here that decoding approaches can lead to valid inferences about the independence of neural representations without being difficult to apply and interpret . We believe that using a decoding separability test offers an improvement over the operational tests of independence commonly used in the literature , without requiring a high level of expertise from researchers . Our application to face perception research is useful to highlight the kinds of questions that can be answered with the new framework and the type of analysis that should be performed to answer such questions . However , there are limitations of the present study that should be noted . First , results were obtained using a small set of naturalistic stimuli , so they should not be over-generalized . There is no guarantee that the same results will hold for other stimulus sets , and more research is needed before reaching any general conclusion about the separability of identity and emotional expression . Second , our experiment and analyses were performed at the group level . This was done to obtain a statistically-powerful test that is sensitive to violations of separability that are consistent across participants . However , the results may not be representative of individual subjects . We expect that the study of encoding separability at the individual level will require obtaining more data from each participant than what was acquired in the present study . Our theoretical work might also require further refinement . In particular , the decoding separability test can detect when encoding separability is violated , but it cannot detect when encoding separability holds ( see Fig 3 ) . Decoding separability itself is difficult to prove , as the decoding separability test is a null hypothesis test . Other approaches are necessary to prove the null of decoding separability , such as an arbitrarily small confidence interval around no effect . Such confidence intervals could be computed using resampling methods , such as bootstrapping . Providing evidence favoring the null in this way is usually difficult , as obtaining small confidence intervals requires a large amount of data . Furthermore , there is not much to gain from proving the null of decoding separability , because concluding that decoding separability holds does not lead to conclusions about encoding separability ( see Fig 3 ) . Thus , when using the decoding separability test ( or any of the other operational tests that we have discussed here ) , researchers should focus only on obtaining evidence of its failure . For many researchers , concluding that a dimension is encoded in a separable manner in a given brain region might be considered more interesting; still , an important contribution of our work is showing that this is in general not possible through indirect measures of neural activity or psychophysics . Perhaps specific assumptions about the measurement model producing the data will make it possible to establish a more direct link between decoding and encoding separability , but such assumptions need to be clearly spelled out by researchers , and data should be provided to back them up . One way in which it is possible to directly compare the evidence in favor of encoding separability versus the evidence against encoding separability is within the encoding modeling framework described earlier . As indicated earlier , this framework allows to compare two encoding models that are identical in all respects except their assumption of encoding separability . Unlike in null hypothesis testing , there is no problem with selecting the simpler model in which encoding separability holds , as long as it provides a better explanation for the data . Although our framework provides a link between neural and perceptual forms of separability , some researchers might consider this link rather weak , as we have only shown that a violation of perceptual separability should be reflected in a failure of encoding separability . Although simply indicating that encoding separability must fail is not very informative about exactly how and where it fails , it is important to understand that here precision has been traded-off for generality . That is , perceptual separability allows to conclude that encoding separability fails , regardless of how the dimension is encoded by the brain or how it is decoded for performance in a task . There is a long tradition in vision of linking neural encoding and psychophysics , and more precise conclusions can be reached by making stronger assumptions about encoding and decoding . For example , a common assumption in this literature is optimal decoding through maximum likelihood estimation [84–86] . The addition of an encoding model that is constrained by results from neurophysiology allows one to make inferences about how many neurons contribute to perception from psychophysical thresholds [84] , or about changes in neural tuning functions from changes in threshold-versus-noise functions [86] . Similarly , future research could strengthen the link between neural and perceptual forms of independence for specific dimensions , by including known features of the underlying neurobiology in the encoding models and stronger assumptions about decoding ( e . g . , optimal decoding ) , an approach that has not been explored yet in the study of independence . We must also note that the approach of trading-off precision for generality is entirely within the tradition of how GRT has been developed in the past . That is , most initial work in GRT had the goal of establishing general links between operational tests and different definitions of independence [14 , 87–89] . We believe that this groundwork is necessary before developing more powerful applications to specific problems in vision science . The notion of independent processing is central to many theories in perceptual and cognitive neuroscience , but its study has lacked the rigor and integration offered by a formal framework , like the one presented here . This framework allows development of theoretically-driven tests of independence of neural representations , which are more clear and rigorous than the operational tests used thus far . The availability of more rigorous definitions and tests to study separability is likely to advance knowledge in a number of areas in visual neuroscience interested in the notions of independence of processing and representation .
This study was approved by the Human Subjects Committee at the University of California , Santa Barbara , and written informed consent was obtained from all participants . Twenty-one male and female right-handed students at the University of California Santa Barbara were recruited to participate in this study . Each participant received a monetary compensation at a rate of US$20/hour . The stimuli and task were identical to those used in a previous behavioral study of separability of face identity and expression [13] . Stimuli were four grayscale images of male faces , part of the California Facial Expression database ( http://cseweb . ucsd . edu/∼gary/CAFE/ ) . Each face showed one of two identities with either a neutral or sad emotional expression . The faces were shown through an elliptical aperture in a homogeneous gray screen; this presentation revealed only inner facial features and hid non-facial information , such as hairstyle and color . Participants performed an identification task both outside and inside the MRI scanner . Each stimulus was assigned to a specific response key and the participant’s task was to identify the image presented in each trial . Each trial started with the presentation of a white crosshair in the middle of the screen for 200 ms , followed by stimulus presentation for a single frame ( i . e . , 16 . 667 ms at a 60 Hz refreshing rate ) . Stimulus presentation was short to make it identical to that used in our previous behavioral study . After stimulus presentation , participants pressed a response key; 500 ms later , feedback about the correctness of their response was displayed on the screen for 500 ms ( “Correct” in green font color or “Incorrect” in red font color ) . If the participant took longer than 5 s to respond , the words “Too Slow” were presented on the screen and the trial stopped . Feedback was followed by a variable inter-trial interval , obtained by randomly sampling a value from a geometric distribution with parameter 0 . 5 and truncated at 5 , and multiplying that value by 1 , 530 ms ( the TR value , see below ) . To obtain estimates of stimulus-related activity with other events in the trial ( crosshair and response ) unmixed , we used a partial trials design in which 25% of the trials included the presentation of the white crosshair not followed by a stimulus . Participants were instructed to randomly choose a response on these partial trials . Stimulus presentation , feedback and response recording were controlled using MATLAB augmented with the Psychophysics Toolbox ( psychtoolbox . org ) , running on Mackintosh computers . Participants practiced the identification task on personal Mackintosh computers outside the MRI scanner for about 20 mins . During this training , participants responded on a keyboard . During scanning , participants responded using the Lumina Response Pad System ( model LU400-Pair ) , with the same finger-stimulus mapping as during pre-training . Images were obtained using a 3T Siemens TIM TRIO MRI scanner with a 12-channel head coil at the University of California , Santa Barbara Brain Imaging Center . Cushions were placed around the head to minimize head motion . A T1-weighted high-resolution anatomical scan was acquired using an MPRAGE sequence ( TR: 2 , 300 ms; TE: 2 . 98 ms; FA: 9°; 160 sagittal slices; 1 × 1 × 1 mm voxel size; FOV: 256 mm ) . Additional scans included a localizer and a GRE field map , neither of which were used in the analyses presented here . Functional scans used a T2*-weighted single shot gradient echo , echo-planar sequence sensitive to BOLD contrast ( TR: 1 , 530 ms; TE: 28 ms; FA: 61°; FOV: 192 mm ) with generalized auto-calibrating partially parallel acquisitions ( GRAPPA ) . Each volume consisted of 28 slices ( interleaved acquisition , 2 . 5 mm thick with a 0 . 5 mm gap; 2 . 5 × 2 . 5 mm in-plane resolution ) acquired at a near-axial orientation , manually adjusted to cover the ventral visual stream and lateral prefrontal cortex . There were a total of four functional runs per participant ( with the exception of five participants who completed three functional runs ) . The first run was a standard functional localizer for face regions [76] . Neutral faces , emotional faces and non-face objects were each presented in different stimulus blocks , separated by fixation blocks . Sixteen images of the same type were presented within a stimulus block , each with a duration of 500 ms and a 250 ms inter-stimulus-interval . Fixation blocks consisted of the presentation of a black screen with a white fixation cross in the middle . The sequence started with a fixation block , followed by 6 blocks of each image category ( 18 total ) , each followed by a fixation block , for a total of 37 blocks . Blocks lasted for 12 seconds , and the whole scan lasted about 7 . 5 mins . The order of image types ( e . g . , neutral-emotional-object ) was counterbalanced across blocks . To ensure attention to the stimuli , participants were asked to push a button whenever an image was repeated in the sequence . Four of the 15 stimuli in a block were repetitions , randomly positioned in the stimulus sequence . In all other functional runs , which lasted about 10 mins each , participants performed the identification task described earlier , without feedback . Each of the four images was repeated 25 times , for a total of 100 trials per run . Stimuli were viewed through a mirror mounted on the head coil and a back projection screen . | A common question in vision research is whether certain stimulus properties , like face identity and expression , are represented and processed independently . We develop a theoretical framework that allowed us , for the first time , to link behavioral and brain measures of independence . Unlike previous approaches , our framework formally specifies the relation between these different levels of perceptual and brain representation , providing the tools for a truly integrative research approach in the study of independence . This allows to identify what kind of inferences can be made about brain representations from multivariate analyses of neuroimaging data or psychophysical studies . We apply this framework to the study of independent processing of face identity and expression . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"face",
"functional",
"magnetic",
"resonance",
"imaging",
"diagnostic",
"radiology",
"social",
"sciences",
"neuroscience",
"magnetic",
"resonance",
"imaging",
"probability",
"distribution",
"mathematics",
"algebra",
"brain",
"mappi... | 2018 | Linking signal detection theory and encoding models to reveal independent neural representations from neuroimaging data |
The modification of DNA by methylation is an important epigenetic mechanism that affects the spatial and temporal regulation of gene expression . Methylation patterns have been described in many contexts within and across a range of species . However , the extent to which changes in methylation might underlie inter-species differences in gene regulation , in particular between humans and other primates , has not yet been studied . To this end , we studied DNA methylation patterns in livers , hearts , and kidneys from multiple humans and chimpanzees , using tissue samples for which genome-wide gene expression data were also available . Using the multi-species gene expression and methylation data for 7 , 723 genes , we were able to study the role of promoter DNA methylation in the evolution of gene regulation across tissues and species . We found that inter-tissue methylation patterns are often conserved between humans and chimpanzees . However , we also found a large number of gene expression differences between species that might be explained , at least in part , by corresponding differences in methylation levels . In particular , we estimate that , in the tissues we studied , inter-species differences in promoter methylation might underlie as much as 12%–18% of differences in gene expression levels between humans and chimpanzees .
Changes in the regulation of gene expression levels have long been hypothesized to play an important role in primate evolution [1] , [2] . To begin to address this hypothesis , a large number of studies have characterized gene expression differences across primates , in particular between humans and chimpanzees [3]–[9] . These studies have pointed to several classes of biological processes ( such as transcriptional regulation , oxidative stress response , and a number of metabolic pathways ) , which might have evolved under natural selection in primates . In addition , in a few cases , comparative studies in primates have been able to draw strong connections between regulatory adaptations and ultimate physiological or anatomical phenotypes [10]–[15] . Despite the wealth of comparative gene expression data , there are many fewer studies of the mechanisms that underlie inter-primate differences in gene regulation ( e . g . , [12] , [13] , [16]–[18] ) . In particular , we know relatively little about the degree to which changes in epigenetic profiles might explain differences in gene expression levels between primates . One of the most extensively studied epigenetic mechanisms is DNA methylation – an epigenetic modification that facilitates fine-tuned regulation of transcription rates [19] , [20] . Spatial and temporal regulation of transcription by DNA methylation has been shown to play an important role in many contexts , including in female X-chromosome inactivation [21] , [22] , genomic imprinting [23] , [24] , and susceptibility to complex diseases in humans , especially cancers [25] , [26] . Methylation is also essential for proper differentiation and development of mammalian tissues [27] , [28] . For instance , the knockout of genes encoding for the DNA-methyl-transferase ( DNMT ) enzymes , which are responsible for de-novo methylation of DNA , results in embryonic lethality in mice [29] , [30] . The causal relationship between changes in promoter DNA methylation and differences in gene regulation has been well established [28] , [31] . It has been shown that hyper-methylation at promoter CpG islands typically results in decreased transcription of downstream genes [32] . When methylation is experimentally removed from promoter regions , transcription levels rise [33] . The specific mechanisms by which DNA methylation affects gene regulation are less clear , though DNA methylation is thought to interact with proteins ( such as methyl-DNA binding proteins ) that associate with histone modifications or the nucleosome in order to maintain a silenced chromatin state [28] , [31] , [34] , [35] . Additionally , it has been proposed that the binding of the transcriptional machinery and enhancer-related transcription factors to methylated genomic regions is less frequent , resulting in decreased transcription levels or absolute gene silencing [28] , [36] . Previous studies have typically described patterns of DNA methylation in a single or few tissues across species [26] , [37]–[41] or in multiple tissues or developmental stages within a single organism [26] , [27] , [34] , [42]–[45] . Comparative studies of DNA methylation across mammals have suggested that the role of DNA methylation in tissue-specific gene regulation is generally conserved . For example , after identifying Tissue-specific Differentially Methylated Regions ( T-DMRs [42] ) , in heart , colon , kidney , testis , spleen , and muscle tissues in mice , Kitamura and colleagues were able to use the methylation status in orthologous human regions to distinguish between the corresponding human tissues [44] . Irizarry and colleagues [26] , who studied genome-wide DNA methylation patterns in spleen , liver , and brain tissues from human and mouse , reported that 51% of T-DMRs are shared across both species . However , there also are a large number of potentially functional differences in methylation levels across species . In particular , in primates , Gama-Sosa and colleagues [39] found that relative methylation levels within tissues generally differ between species , with the exception of hyper-methylation in the brain and thymus , which were observed regardless of species . In addition , Enard and colleagues [38] , who compared methylation profiles of 36 genes in livers , brains , and lymphocytes from humans and chimpanzees , reported significant inter-species methylation level differences in 22 of the 36 genes , in at least one tissue . With few exceptions , however ( e . g . , [46] ) , comparative studies in primates have not explored the extent to which methylation differences between species might contribute to the genome-wide regulation of inter-species differences in gene expression levels . Towards this goal , we compared genome-wide gene expression levels and DNA methylation data in tissue samples from humans and chimpanzees .
We obtained methylation profiles from each sample ( using two independent DNA extraction replicates ) by using the Illumina HumanMethylation27 DNA Analysis BeadChip assay , which provides reproducible ( Figure S2 ) quantitative estimates of methylation levels at 27 , 578 CpG-loci near transcription start sites . Since the 50 bp probes on the Illumina array were designed to interrogate human samples , we limited our analysis to probes that were a perfect sequence match to the chimpanzee genome . In addition , we only used probes that were associated with genes for which we had expression measurements across the three tissues [8] . Following these exclusion criteria , we retained 10 , 575 CpG site probes in the putative promoter regions of 7 , 723 genes ( see Methods for more details ) . At each probe , DNA methylation levels were estimated using the Illumina-recommended β values , which are essentially estimates of the proportion of methylated DNA at each CpG site ( see Methods ) . We note that limiting our analysis to identical methylation probes in humans and chimpanzees resulted in a slight ( 0 . 5% ) but significant decrease of the median sequence divergence estimates within 500 bp windows around the retained probes ( Figure S3 ) . As a result , it is possible that , in what follows , we slightly underestimate the proportion of inter-species differences in methylation levels . However , we confirmed that limiting our analysis to identical methylation probes in the two species did not result in a noticeable shift in the distribution of expression levels of the associated genes , nor in the proportion of observed differences in gene expression levels between the two species . As a first step of our analysis , we examined patterns of promoter methylation across tissues and species . As expected [28] , [31] , we found a negative correlation between methylation and gene expression levels in each individual , whereby , regardless of tissue and species , the promoters of highly expressed genes tended to be lowly methylated while the promoters of lowly expressed genes were usually highly methylated ( Figure 1A; Figure S4 ) . We also confirmed that methylation patterns on the X-chromosome account for variation due to sex , regardless of species , as expected due to X-inactivation in mammalian females [21] ( the first component of variance , corresponding to sex , accounts for 67% of the overall variation in the X-chromosome data; Figure 1B ) . Finally , we found that genes known to be imprinted in humans tend to show a similar hemi-methylation pattern in chimpanzees ( permutation tests P<0 . 001; Figure 1C ) , suggesting that the imprinted status of this set of genes is conserved in the two species . For the remainder of the analyses , we considered only the methylation data from autosomal probes . We observed that methylation patterns across different tissues and species were quite distinct ( Figure 2; similar patterns for the expression data in Figure S5 ) . The first component of variance for the autosomal probes , accounting for 69 . 3% of the overall variation in methylation , distinguished samples based on tissue , while the second principal component ( accounting for 12 . 7% of the overall variation ) , separated the species . Overall , an average of 14 . 5% ( range of 8 . 2–26 . 1% , depending on the pairwise comparison ) of the assayed promoter CpG sites were differentially methylated between tissues within a species , while an average of 8 . 6% of the CpG sites ( range of 3 . 4–13 . 5% , depending on the tissue ) were differentially methylated between humans and chimpanzees ( at FDR<0 . 001 ) . Reassuringly , these patterns recapitulate previous observations in human and mouse [26] , [44] . We identified regions with tissue-specific patterns of methylation ( T-DMRs [26] , [42] ) by analyzing the data from each species separately ( Figure 3 ) . Specifically , we modeled the methylation data ( namely , the β values ) from each autosomal CpG site independently , using a linear mixed-effects model with a fixed effect for the tissue and a random effect to account for variation between individuals . We tested for differences in methylation levels between tissues by using likelihood ratio tests within the framework of the linear model ( see Methods ) . Using this approach , we identified 1 , 578 and 1 , 401 T-DMRs in humans and chimpanzees , respectively ( at an FDR<0 . 001; Figure 3A; Table S1 ) . Tissue-specific methylation profiles are of interest because they may underlie tissue-specific patterns of gene expression levels . To test this hypothesis , we calculated , separately for each species , Pearson correlation values between promoter methylation profiles and the corresponding gene expression levels , across the three tissues . If methylation was consistently used to silence tissue-specific gene expression across the genome , we would expect to observe an abundance of negative correlations between the estimates of methylation and gene expression levels . However , when we considered the data for all genes that were expressed in at least one tissue , we found no evidence for an enrichment of negative correlations between methylation and gene expression levels ( Figure 3B , Figure S6; 48% and 49% of the correlation values were negative in human and chimpanzee , respectively ) . In contrast , when we restricted the analysis to species-specific T-DMRs , we found an enrichment of negative correlations between methylation and gene expression levels ( Figure 3B; 64% and 67% of correlation values were negative in human and chimpanzee , respectively; Fisher's exact P<10−16 ) . This result suggests that T-DMRs underlie a subset of gene expression differences across tissues , a notion that is consistent with the important role played by DNA methylation in tissue differentiation in a wide range of species [42] . We then focused on the subset of T-DMRs with the same methylation pattern in both species . We found that 18–26% ( depending on the tissue ) of loci classified as T-DMRs in either human or chimpanzee are shared between the two species ( Figure 3A , Table S2 ) , a highly significant overlap compared to that expected by chance alone ( hypergeometric distribution P values across all pairwise tissue comparisons <10−16 ) . Importantly , the observation of a significant overlap in T-DMRs across species is robust with respect to the statistical cutoff used to classify T-DMRs ( 0 . 001≤FDR≤0 . 05; Table S2 ) . Interestingly , when we considered correlations of methylation and gene expression levels only at conserved T-DMRs , we found an even more pronounced enrichment of negative correlations ( Figure 3B and 3C; 72% of the correlation values were negative , regardless of species; Fisher's exact for an enrichment of negative correlations: P<10−23 ) , suggesting that conservation of T-DMRs often relates to functionally important tissue-specific patterns of gene regulation . It is perhaps interesting to note that we did not find a difference in the correlation of methylation and expression levels between T-DMR CpG sites that are located within or outside an annotated CpG island ( as defined by [47]; Figure S7 ) . When we examined the functional annotations of genes associated with species-specific T-DMRs as well as conserved T-DMRs ( using gene ontology annotations ) , we found an expected enrichment of genes annotated as important in ‘developmental’ processes , regardless of tissue ( P<5×10−3; FDR<0 . 3; Table S3 ) , congruent with the importance of epigenetic modification in tissue differentiation . We also found enrichments of tissue-specific biological processes , such as genes associated with cardiac muscle cell differentiation processes among heart T-DMRs ( P<5×10−3; FDR<0 . 3 ) , genes associated with embryonic organ morphogenesis and embryonic organ development processes among kidney T-DMRs ( P<5×10−4; FDR<0 . 05 ) , and genes associated with blood coagulation and with the regulation of body fluid levels ( putatively involved in homeostatic functions ) among liver T-DMRs ( P<10−5; FDR<6×10−3 and P<10−4; FDR<0 . 007 , respectively ) . The enrichment of genes associated with both developmental and tissue-specific processes among genes associated with T-DMRs is consistent with previous observations [27] , [42] . Furthermore , when we considered only conserved T-DMRs , we observed a significant under-representation of genes associated with nucleic-acid and primary metabolic processes in all three tissues studied ( all P<5×10−3; FDR<0 . 01; Table S4 ) . This result suggests that the epigenetically-mediated tissue-specific regulation of these core processes tends to be conserved between humans and chimpanzees . We next focused on the relationships between inter-species differences in methylation profiles and differences in gene expression levels between humans and chimpanzees . To estimate the relative contribution of changes in DNA methylation to inter-species differences in gene expression levels , we used linear regression analysis to account for promoter methylation effects ( per autosomal CpG site ) before analyzing the gene expression data from both species . We analyzed methylation and gene expression data in each tissue using a linear model framework similar to the one described in Blekhman et al . 2008 [8] . We then compared the evidence supporting an inter-species difference in gene expression levels before and after correcting for methylation profiles ( see Methods for more details ) . For the majority of genes ( 78% , 82% , and 77% in liver , kidney , and heart , respectively; Figure 4A ) , the evidence for a difference in expression level between the species was similar , regardless of whether or not methylation status was taken into account . For a small subset of genes ( 1% , 3% , and 2% in liver , kidney , and heart , respectively ) , we did not find compelling evidence for a difference in expression level between the species using the uncorrected expression level data , but after correcting for methylation levels using regression analysis , we rejected the null hypothesis of no inter-species differences in gene expression level ( at an FDR<0 . 01 ) . This observation , however , is unlikely to be biologically meaningful , since it is expected by chance alone ( by permutation analysis; P>0 . 434 for all tissues; Figure S8 ) . In contrast , in all three tissues , we found a significant enrichment of genes for which the evidence for inter-species differences in expression level was compelling ( FDR<0 . 01 ) before , but not after we corrected for the methylation levels ( 21% , 15% , and 21% in liver , kidney , and heart , respectively , permutation analysis yields P<0 . 001 for all tissues; Figure 4B and 4C ) . Based on the expectation of such a pattern by chance alone ( by permutations – see Methods for details ) , we estimated that , in the three tissues we studied , inter-species differences in promoter DNA methylation might underlie as much as 12–18% of differences in gene expression levels between humans and chimpanzees . When we analyzed the data considering only the sets of genes that have negative correlations between methylation and gene expression levels ( as expected if methylation is used to silence gene expression ) , we found that 8 . 1% , 7 . 6% , and 8 . 8% of interspecies differences in gene expression levels in liver , kidney , and heart , respectively , might be explained by corresponding methylation differences . The extent to which inter-species gene expression differences might be explained by methylation differences between the species was similar regardless of whether the methylated site was within or outside an annotated CpG islands ( Figure S9 ) .
We found a substantial degree of conservation of tissue-specific methylated regions in human and chimpanzee . This observation is not surprising given that previous studies found a marked conservation of T-DMRs between human and mouse , which are much more distantly related [26] , [41] , [43] , [44] . On the other hand , 7 . 0% , 21 . 6% , and 23 . 8% of the kidney , heart , and liver T-DMRs , respectively ( identified in either species ) , were differentially methylated ( in the relevant tissue ) between humans and chimpanzees , while only 3 . 3% , 8 . 0% , and 11 . 8% of non-TDMRs in these three tissues were differentially methylated between the two species ( P<10−10 for all pairwise comparisons ) . The conservation of T-DMR profiles yet the generally faster rate of inter-species change in promoter methylation at T-DMRs compared to non-T-DMRs are intriguing . These observations are difficult to explain by technical or uncontrolled aspects of the study design , because it is unlikely that those confounding factors would affect methylation at T-DMRs differently than at non-T-DMRs . Instead , it is likely that the different patterns truly reflect a functional difference between methylation at T-DMRs and at non-T-DMR CpG sites ( in the studied tissues ) . Though there is substantial evidence that DNA methylation levels upstream of genes are often inversely correlated with gene expression levels [24] , [28] , [31] , recent studies proposed that methylation of promoters may play only a relatively minor role in the regulation of tissue-specific gene expression [34] . In particular , Maunakea et al . [48] posited that methylation of gene body regions ( in regions that putatively serve as alternative promoters ) might have a greater influence on regulatory differences across tissues . While we cannot use our data to ask about the relative importance of different types and locations of epigenetic marks to tissue-specific gene regulation , our observations strongly imply that any such debate would benefit from further investigation into the evolution of epigenetic profiles . Indeed , in addition to a faster rate of evolutionary change of the methylation profiles in T-DMRs , we found evidence for an enrichment of inverse correlations between inter-tissue gene expression patterns and promoter methylation profiles at genes associated with T-DMRs , but not when we considered all genes ( the latter observation is consistent with the findings of Weber et al . [34] and Maunakea et al . [48] ) . Our results , therefore , imply that tissue-specific promoter methylation patterns may play especially important roles in regulating gene expression . The data also suggest that altered methylation levels , primarily at these sites , may underlie regulatory differences between species . We estimated that as much as 12–18% ( depending on the tissue ) of inter-species differences in gene expression levels might be explained , at least in part , by changes in DNA methylation patterns . It is important to note that this statement is based on the proposed mechanism by which DNA methylation affects the rate of transcription and overall levels of gene expression [28] , [31] . Though we did not perform experiments from which causality can be directly deduced , a causal relationship between changes in DNA methylation and gene regulation is strongly supported by previous studies ( e . g . , [24] , [28] , [31] ) . When we only consider negative correlations between methylation and gene expression levels to be indicative of a putative causal relationship , 8–9% of inter-species differences in gene expression levels might be explained by corresponding changes in DNA methylation . However , other mechanisms are also likely [34] , [43] . While DNA methylation is typically considered a silencing mechanism , high levels of methylation may be causally linked to increased gene expression levels . For example , the methylation of a repressor site could prevent the binding of repressor transcription factors , or enhancer transcription factors could favor binding to a methylated site rather than to the unmethylated site [49]–[51] . The observation of a small enrichment of positive correlations between methylation and expression when only T-DMRs are considered provides additional support for these types of mechanisms . Thus , perhaps as much as 12–18% of differences in gene expression levels between humans and chimpanzees might be explained by inter-species changes in DNA methylation . Either way , our results suggest that DNA methylation differences in promoter regions might account for , at most , a modest proportion of inter-primate differences in gene expression levels ( we confirmed that our estimates do not rely on arbitrary choices of specific statistical cutoffs; Tables S2 and S5 ) . Many inter-species differences in promoter methylation are not associated with gene expression differences between the species . One explanation for that observation may simply be that these methylation patterns are not regulatory or functional . An alternative , more interesting possibility to consider , is that a subset of genes whose regulation differed between species later acquired modifications in nearby DNA methylation patterns to accommodate ( or even partially counteract ) the original expression level changes . Since we assayed methylation using a pre-designed microarray , changes in DNA methylation in un-assayed genomic regions might explain additional regulatory differences between the species . In particular , while our assay focused on methylation at promoter regions , it has been recently shown that as a class , gene-body methylation profiles might explain a larger proportion of variation in gene expression levels than methylation profiles at currently annotated promoters [26] , [48] . With the advent of new sequencing technologies , it will soon be feasible to extend our comparative approach to characterize genome-wide patterns of methylation . In summary , we have taken some of the first steps towards characterizing variation in one mechanism that affects gene expression differences between closely related primate species [16] , [17] . In a broader context , DNA methylation is just one of many mechanisms that have been posited to regulate gene expression levels [28] , [31] , [52] . In that sense , our study is a step towards the ultimate goal of understanding the relative importance of changes in different regulatory mechanisms to human evolution . Our observations indicate that at least 82% of gene expression differences between humans and chimpanzees ( in the three studied tissues and specific promoter CpG sites examined ) are not likely to be explained by differences in promoter DNA methylation .
We collected methylation data from the same human and chimpanzee liver , kidney , and heart tissue samples used in Blekhman et al . 2008 [8] ( Figure S1; see Table S6 for details on the samples ) . DNA was extracted from each sample ( 6 human and 6 chimpanzee samples from each of the three tissues ) in two independent technical replicates using the QIAamp DNA Mini Kit ( Qiagen ) ( with the exception of chimpanzee sample CK2 , for which DNA was only available for one replicate – see Table S4 ) . The methylation profile of each sample was assayed using the Illumina HumanMethylation27 DNA Analysis BeadChip , which assays methylation at 27 , 578 CpG sites . Methylation array data are deposited to the NCBI GEO database under the accession number GSE26033 ( http://www . ncbi . nlm . nih . gov/projects/geo/query/acc . cgi ? acc=GSE26033 ) . To facilitate an unbiased comparison of methylation and gene expression levels in the human and chimpanzee samples , we first mapped the 27 , 578 50-bp Illumina probes to the human genome sequence ( hg18 ) using BLAT [53] and MAQ [54] . We retained only the 26 , 690 probes that unambiguously mapped to a single location in the human genome with a maximum of two mismatches . These probes were then associated with the nearest gene using Ensembl gene annotation , and we retained only the subset of probes associated with genes that were represented on the multi-species gene expression microarray used by Blekhman et al . 2008 [8] . This resulted in the retention of 19 , 849 probes , associated with 11 , 059 genes . Finally , since the Illumina array was designed based on human genomic sequence , we limited our analysis to probes that were a perfect sequence match to a single location in the chimpanzee genome , by mapping the remaining 19 , 849 probes to the chimpanzee genome ( panTro2 ) using BLAT [53] and MAQ [54] . We retained 10 , 575 probes that mapped uniquely to the chimpanzee genome with no sequence mismatches . This step ensures that our relative methylation measurements are not biased due to the effect of sequence mismatches on hybridization intensities . The resulting set of 10 , 575 probes is associated with 7 , 723 genes , which are present on every chromosome in the genome except for the Y-chromosome ( Figure S10 ) . The majority ( 97% ) of retained probes are located within 2 kb of an annotated transcription start site of the associated gene ( Figure S11 ) . We note that a similar screen for probes that were a perfect match to the genomes of human , chimpanzee , and rhesus macaque resulted in the retention of only 1 , 944 probes ( associated with 1 , 715 genes ) . For that reason , we limited our current study to a comparison between human and chimpanzee samples . All samples were hybridized to the Illumina HumanMethylation27 DNA Analysis BeadChip at the Southern California Genotyping Consortium facility following standard manufacturer's instructions . Basic quality checks were performed using Illumina's BeadStudio software . Of the 10 , 575 probes we considered as the final dataset , 299 had missing data for one or more individuals and were discarded in all subsequent analyses . This resulted in 9 , 911 autosomal probes ( corresponding to 7 , 291 genes ) and 365 probes on the X-chromosome ( corresponding to 266 genes ) . Since the probes map to distinct CpG island regions , which can affect downstream gene expression independently , we treated methylation levels from each CpG probe as distinct data points in all subsequent analyses . We further classified each probe as being located confidently within a CpG island region or outside of a strict CpG island region using the CpG Islands track information downloaded from UCSC [47] . For each sample , the methylation status at a probed location was summarized as: where M and U denote the signal emitted from the beads assaying the methylated and unmethylated versions at each site , respectively . Due to the number of samples being interrogated , it was necessary to hybridize the samples in two balanced batches . We observed a small difference in the mean β-value between batches , and corrected for this difference by standardizing the means across batches . After this correction , there was no further evidence for a batch effect . To further assess the quality of the data , we calculated pairwise correlations between the β-values for all hybridized samples ( Figure S2 ) . As expected , technical replicates ( which were independent DNA extractions ) were the most highly correlated ( 36 comparisons; median r = 0 . 99 ) , followed by samples from the same tissue and species ( 396 comparisons; median r = 0 . 98 ) , samples from the same tissue across species ( 432 comparisons; median r = 0 . 97 ) , samples from different tissues from the same species ( 864 comparisons; median r = 0 . 95 ) , and samples from different tissues and different species ( 864 comparisons; median r = 0 . 93 ) . To look for evidence of imprinting in both humans and chimpanzees , we focused on a set of 27 genes ( associated with 90 methylation probes ) known to be imprinted based on the Imprinted Gene Catalog ( IGC ) at http://igc . otago . ac . nz/ . To assess whether the patterns of DNA methylation at these imprinted genes were likely to occur by chance , we compared the observed proportion of hemi-methylated sites ( defined as 0 . 3<β<0 . 7 ) to the distribution obtained by analyzing methylation patterns in 1000 randomly chosen sets of 90 methylation probes , associated with an average of 27 genes ( range 26–28 ) . Measurements of gene expression levels for all samples in our study were previously described by Blekhman et al . ( 2008 ) [8] . These data are available at the Gene Expression Omnibus ( GEO ) database ( http://www . ncbi . nlm . nih . gov/geo/ ) under series accession number GSE11560 . In that study , a multi-species microarray was used to estimate gene expression levels in cDNA samples from humans , chimpanzees , and rhesus macaques . The multi-species array includes orthologous probes for 18 , 109 genes , thus facilitating comparisons of gene expression levels between species without the confounding effects of sequence mismatches on hybridization intensities [8] . Since our current study focused only on the human and chimpanzee gene expression data , we re-normalized the expression data using only the human and chimpanzee probes on the array , using the same modified quantile normalization approach described in Blekhman et al . ( 2008 ) [8] . All further analyses used these re-normalized gene expression estimates . When examining the relationships between gene expression and methylation levels , we limited our analyses to genes that were either expressed in at least one tissue ( for inter-tissue comparisons within a species ) or expressed in at least one species ( for the inter-species comparisons within a tissue ) , using a conservative threshold for defining expression , based on the entire distribution of expression values ( normalized expression value of 8; see Figure S14 in Blekhman et al . ( 2008 ) [8] ) . All statistical analyses were performed using the R statistical framework ( http://www . r-project . org ) . To identify T-DMRs , we modeled the methylation level of each CpG site separately within both humans and chimpanzees using a linear mixed-effects model . Specifically , for each of the 9 , 913 probes ( associated with 7 , 291 genes ) located on the autosomal chromosomes , if yijk represents the β value for technical replicate k ( k = 1 or 2 ) , for individual j ( j = 1 , … , 6 ) , from tissue i ( i = heart , liver , or kidney ) , we assume that: ( 1 ) where: Here , αi represents the mean methylation value at a given site in tissue i . To account for correlation between samples of the same tissue from different individuals , a random effect , ρij , which follows a N ( 0 , σ2rand ) distribution , is also included in the model . To determine whether a CpG site was likely to fall within a T-DMR , we assessed how well the model ( 1 ) fitted the data under various parameterizations of μijk . The three types of parameterizations considered are: In the simplest model ( H0 ) , the region's methylation value is assumed to be constant across all three tissues , while in the second alternative ( H2 ) the methylation value is allowed to differ between all three tissues . The first alternative ( H1 ) models the situation where the methylation level at the site of interest is constant in the two non-target tissues but differs in the target tissue . All models are fitted using a restricted maximum likelihood ( REML ) framework , and the maximum likelihoods were calculated . In this study , we are interested in identifying sites whose methylation levels are best modeled by H1 . To find such sites , we first used a likelihood-ratio test statistic ( with one degree of freedom ) to exclude sites where H2 provides a better fit to the data than H1 ( specifically , if the likelihood-ratio p-value was less than 0 . 05 , we removed these sites from the analysis ) . H2 provides a better fit for 1220 and 886 ( in humans and chimpanzees , respectively ) of the total 9911 autosomal CpG sites . For the remaining positions , we examined whether there was significant evidence to reject H0 in favor of H1 using a likelihood-ratio test statistic ( which we compared to a χ2 distribution with 1 degree of freedom ) . We corrected for multiple testing using the FDR approach of Storey and Tibshirani [55] . We used GeneTrail ( http://genetrail . bioinf . uni-sb . de ) [56] to test for enrichments of functional annotations among different classes of T-DMRs . In all tests , we used a background set of genes that were present in our study and classified as expressed in at least one tissue ( conditional on a normalized expression value of 8 ) . The tests were performed using all GO categories and KEGG pathways . We calculated p-values using a Hyper-geometric distribution and report false discovery rates for each p-value . To examine whether changes in gene expression levels between humans and chimpanzees ( within each tissue ) can be explained by inter-species differences in methylation levels , we extended the linear mixed-effects model framework described in Blekhman et al . ( 2008 ) [8] to include methylation as a covariate . However , since we have to correct the multi-species array data for probe-effects [8] , it is difficult to interpret the methylation coefficient when it is added directly to the model , since it is confounded with the probe effects . Consequently , we used an alternative approach in which we used regression to correct for the methylation effect . Specifically , for each gene-tissue combination , we tested for differences in expression level between human and chimpanzee after regressing out the following effects: To do this , we used a fully parameterized model where gene expression probe effects , CpG-probe methylation values , and species effects were explanatory variables . Additionally , a random effect was used to account for variability between biological replicates . Specifically , if ysroi denotes the normalized log2 intensity expression value for individual i ( i = 1 , … , 6 ) , from species s ( s = human or chimpanzee ) measure at probe r ( r = 1 , … , 7 ) , which is derived from species o , we assume that: where: Here , μs denotes the species effect , πro is a fixed-effect representing the probe effect for each individual probe within a probe-set and the composition effect of species-specific orthologous probes , and κsro is a fixed-effect representing the attenuation of hybridization intensities due to sequence mismatches between species of RNA and a species-specific derived probe , which are different for each individual probe within a probe set ( see [8] for more details ) . Additionally , γsi is a random effect ( following a N ( 0 , σ2rand ) distribution ) and βsi denotes the β value for the methylation probe of interest for individual i from species s . Upon fitting this model , using the lmer package within the R statistical framework , estimates of the parameters and the residuals were obtained . To obtain corrected measures of expression for each individual from each species , when probe and methylation effects are regressed out ( scenario 2 ) , we defined . When we only regressed out probe effects ( scenario 1 ) , the corrected values are defined as . In both of these scenarios , once the corrected data were obtained , we tested for differences in gene expression levels as follows . If , for each tissue-gene combination , xsik denotes the ( corrected ) level of expression for replicate k of individual i from species s , we modeled these data as follows:where:Here , αs is a species effect , and ρsi is a random individual effect . Subsequently , to test for inter-species differences in expression levels , we compare the following hypotheses:Here , the null model assumes equal expression level between the two species , and the alternative assumes different expression levels . Evidence against the null model was determined using a likelihood-ratio test statistic ( compared against a chi-squared distribution with one degree of freedom ) . By performing this analysis independently for each CpG-gene combination in all tissues , we obtained a p-value indicating the strength of the evidence against the null hypothesis , before ( under scenario 1 above ) and after ( under scenario 2 above ) accounting for the region's DNA methylation status . By comparing these p-values , we were able to identify genes within each tissue where the difference in expression level between human and chimpanzee was likely explained by inter-species differences in DNA methylation . To assess the statistical significance of our observations , we permuted the methylation values for a given gene across all individuals ( maintaining replicate correlations , but allowing labels to permute across species classifications ) . Subsequently , we repeated the analysis described above to obtain an expected distribution of discrepancies between the methylation-corrected and uncorrected data . We performed 1000 permutations and p-values were calculated based on the number of times we observed as many or more discrepancies in the permuted compared to the real data . In order to estimate the proportion of genes for which methylation differences might underlie gene expression differences , we treated the medians of the permutation distributions from each tissue as background levels . For each tissue , we then subtracted the background level from the observed proportion of genes with reduced evidence for inter-species differences in gene expression levels , once methylation was taken into account . | It has long been hypothesized that changes in gene regulation have played an important role in primate evolution . However , despite the wealth of comparative gene expression data , there are still only few studies that focus on the mechanisms underlying inter-primate differences in gene regulation . In particular , we know relatively little about the degree to which changes in epigenetic profiles might explain differences in gene expression levels between primates . To this end , we studied DNA methylation and gene expression levels in livers , hearts , and kidneys from multiple humans and chimpanzees . Using these comparative data , we were able to study the evolution of gene regulation in the context of conservation of or changes in DNA methylation profiles across tissues and species . We found that inter-tissue methylation patterns are often conserved between humans and chimpanzees . In addition , we also found a large number of gene expression differences between species , which might be explained , at least in part , by corresponding differences in methylation levels . We estimate that , in the tissues we studied , inter-species differences in methylation levels might underlie as much as 12%–18% of differences in gene expression levels between humans and chimpanzees . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"evolutionary",
"biology/human",
"evolution",
"genetics",
"and",
"genomics/epigenetics",
"genetics",
"and",
"genomics/gene",
"expression",
"evolutionary",
"biology/genomics"
] | 2011 | A Genome-Wide Study of DNA Methylation Patterns and Gene Expression Levels in Multiple Human and Chimpanzee Tissues |
The repression of competition by mechanisms of policing is now recognized as a major force in the maintenance of cooperation . General models on the evolution of policing have focused on the interplay between individual competitiveness and mutual policing , demonstrating a positive relationship between within-group diversity and levels of policing . We expand this perspective by investigating what is possibly the simplest example of reproductive policing: copy number control ( CNC ) among non-conjugative plasmids , a class of extra-chromosomal vertically transmitted molecular symbionts of bacteria . Through the formulation and analysis of a multi-scale dynamical model , we show that the establishment of stable reproductive restraint among plasmids requires the co-evolution of two fundamental plasmid traits: policing , through the production of plasmid-coded trans-acting replication inhibitors , and obedience , expressed as the binding affinity of plasmid-specific targets to those inhibitors . We explain the intrinsic replication instabilities that arise in the absence of policing and we show how these instabilities are resolved by the evolution of copy number control . Increasing levels of policing and obedience lead to improvements in group performance due to tighter control of local population size ( plasmid copy number ) , delivering benefits both to plasmids , by reducing the risk of segregational loss and to the plasmid-host partnership , by increasing the rate of cell reproduction , and therefore plasmid vertical transmission .
The evolution of cooperation is a fundamental problem in biology: why help another individual to reproduce , if this comes at a cost to one's own reproductive success ? This dilemma is reflected in the trade-off between an individual's immediate reproductive gains and its longer-term prospects of success as part of a collective , whose stability and overall performance is undermined by internal competitiveness . The eroding consequences of competition are exemplified by the “tragedy of the commons” [1] , in which a common resource is over-exploited and eventually destroyed by a group of self-interested individuals who act in order to maximize their immediate personal yield from that resource . The conflict between individual and group interests , however , does not prevent the emergence of cooperation: from genes on genomes and chromosomes in cells , to multicellularity , eusociality and beyond , harmonious cooperative behavior is both widespread and persistent across all levels of biological complexity . The mechanisms underlying the moderation of individual competitiveness ( self-restraint ) , as a means of promoting cooperation , rely on the social interactions between individuals within a group . Sufficiently high genetic relatedness among interacting individuals can promote intra-specific cooperation and the evolution of self-restraint through the carriage and transmission of shared “cooperative” genes [2] . However , kin selection can not account for the maintenance of costly cooperation when individuals are distantly related or even not related at all , in which case various forms of reciprocity can support cooperative behavior by assuring direct fitness returns to focal actors ( for reviews , see [3]–[5] ) . At low relatedness , the repression of internal competition for the benefit of the collective can also be achieved through individual investment in appropriate enforcement mechanisms such as mutual policing , resulting in a level-playing field within the group that motivates individuals to contribute towards the enhancement of the group's efficiency and productivity in order to increase their own reproductive success [6]–[9] . A particularly elegant example of a policing mechanism for the repression of competition among individuals within a group is the replication control system of bacterial plasmids . Plasmids are extra-chromosomal DNA elements , organized as , typically circular , collections of discrete genetic modules [10] , [11] , including those encoding functions necessary for their survival and propagation such as self-replication and its control , active partitioning during cell division , and conjugative transfer . Plasmids replicate autonomously by making use of the replication machinery of their host; they also encode a policing mechanism for controlling their replication [12] , [13] . The role of this mechanism is to ensure that each plasmid copy replicates once per cell cycle on average , so as to maintain a stable characteristic copy number under constant conditions . In plasmid R1 for example , copy number control ( CNC ) is achieved through the constitutive synthesis of trans-acting replication inhibitors , in the form of the widely used plasmid-coded antisense RNAs [14] , that decay rapidly so that their concentration is proportional to the plasmid's copy number [15] , [16] . Inhibitors act by binding to and deactivating a plasmid-specific target that is rate-limiting for the initiation of plasmid replication . The presence of inhibitors induces the establishment of a negative feedback loop between the plasmid copy number and individual plasmid replication rates: higher copy numbers result in a higher concentration of trans-acting inhibitors in the cell , thereby effecting a reduction in the frequency of plasmid replication and vice versa . The plasmid CNC mechanism encapsulates the two fundamental traits of standard generic policing models [6]–[9] , [17] , namely individual competitiveness ( selfishness ) captured by the production of rate-limiting constitutive factors ( such as Rep proteins [18] , [19] or RNA primers [20] ) that are responsible for the initiation of plasmid replication , and mutual policing captured by the synthesis of trans-acting replication inhibitors that act upon the target initiation factor , thus mediating the repression of competition in the plasmid replication pool . In addition to these two fundamental traits , the mechanistic structure of the CNC system motivates the consideration of a third one: the obedience of individuals to the policing resources that they themselves produce , as expressed by the binding affinity of designated inhibitor targets on individual plasmids to the inhibitor molecules . The distinction between contributing towards the production of policing resources ( by producing generic trans-acting inhibitors ) and being sufficiently obedient to policing ( by appropriately responding to these inhibitors ) creates the potential for subversive strategies: an individual plasmid can gain a competitive advantage in the intra-cellular replication pool by contributing towards the collective production of policing resources , provided that its own sensitivity to these resources is lower than the sensitivity of its coresident plasmids . Nevertheless , the good of the collective ( host cell ) can prevail over the short-sighted self-interest of individual plasmids , provided that , first , there is limited migration of individuals between collectives ( i . e . slow rate of horizontal transmission ) ; second , each new collective is founded by a limited number of collectives ( e . g . each daughter cell results from binary fission of a single parent cell ) ; and , third , the number of collectives exceeds the number of individuals per collective ( i . e . host population size far exceeds the per-host plasmid copy number ) [21] . The CNC mechanism for the collective restraint of plasmid selfishness via mutual policing operates in a clear inter-specific context: plasmids often code for accessory adaptive traits that can provide their hosts with a variety of competitive selective advantages under particular environmental conditions [22] . Examples include resistance to antibiotics and heavy metals , the ability to exploit new niches and to metabolize unusual environmental elements , the capacity to synthesize toxins and virulence factors etc . However , the carriage of locally beneficial allele ( s ) is not always sufficient to guarantee a mutualistic outcome in the host-plasmid relationship , since the host also bears the metabolic cost of plasmid maintenance , due to the plasmids' usage of the cellular machinery for the purpose of gene expression , replication etc . The cost of plasmid carriage is positively correlated with the copy number [23]–[25] and if the number of plasmid copies within the focal cell is such that the metabolic costs accociated with plasmid maintenance exceed the benefits of the focal trait then the mutualistic host-plasmid relationship degrades to parasitism . The policing mechanism for the control of plasmid replication is subject to a dynamic evolutionary conflict between two levels of selection [26] , [27]: at the intra-cellular level , plasmid mutations that induce an increased rate of initiation of plasmid replication ( selfishness ) or a reduced binding affinity to the trans-acting inhibitor ( obedience ) would result in plasmids that proliferate faster in comparison to more frugal plasmids with a higher degree of adherence to the CNC mechanism . Hence , intra-cellular selection will favor plasmid selfishness and oppose obedience to collective policing . The consequent escalating drive towards the ratcheting of plasmid replication rates is an example of a tragedy of the commons , as hosts inhabited by selfish and disobedient plasmids become increasingly unable to bear the metabolic burdens associated with the elevated copy numbers . However , such hosts find themselves in a disadvantageous position compared to fellow hosts inhabited by more obedient plasmids , where stricter replication control results in more moderate metabolic costs . This way , selective pressures towards plasmid recklessness at the intra-cellular level are counter-balanced by inter-cellular selection that penalizes hosts with disobedient plasmids and , therefore , disobedient plasmids themselves . The pressure for effective CNC is particularly strong for non-conjugative plasmids , i . e . purely vertically-transmitted symbionts that lack the capacity for horizontal ( infectious ) transmission to neighboring hosts . By forgoing horizontal transfer , vertically transmitted symbionts hook their reproductive fate to that of their hosts , thus forging an alliance of fitness interests that can support an elaborate mutualistic host-symbiont relationship [28] , [29] . In this paper , we explore a range of social dilemmas facing non-conjugative plasmids carrying beneficial alleles . These dilemmas are present in the mechanistic interactions between selfishness , policing and obedience that determine the efficiency of the CNC system . More specifically , we use a mathematical model of the symbiotic relationship between hosts and plasmids in order to investigate how the conflict between intra-cellular selection ( favoring plasmid recklessness with respect to replication ) and inter-cellular selection ( favoring an optimal metabolic balancing of costs and benefits in hosts ) orchestrates the evolution of the plasmid-coded CNC mechanism .
Our model for cellular growth and division ( or death ) is based on the premise that cell metabolism produces biomass and when this increasing mass reaches a certain threshold the cell divides . Specifically , let be the cell's biomass , with an initial value , which is updated at every time interval according to . When the biomass doubles , i . e . when , the cell divides . Inefficiencies in cell metabolism can lead to negative growth ( ) and , eventually , to death when the biomass of a shrinking cell falls below zero . Changes in biomass are determined by the chromosomal and plasmid contributions to cellular metabolism , as well as the metabolic costs of plasmid maintenance paid by the host . We consider a plasmid type , which has a positive contribution to host growth that saturates with copy number and converges towards a value ( a saturating gene dosage effect ) . In essence , measures the strength of the selective pressure for the plasmid trait in the homogeneous environment of the cellular population . Let be the general cost of maintaining a single plasmid copy , including the costs of gene expression , plasmid replication etc . We define the rate of cell growth , expressed as the change in biomass per unit time step , according to: ( 1 ) where is the metabolic contribution of the cell's chromosome or the cell's basal growth rate and characterizes the steepness of the curve that describes the saturating beneficial contribution of the plasmid to host growth as a function of the copy number . The form of implies that there exists a copy number value for which the positive offset between the plasmid's contribution to cell growth and the plasmid's burden to cell metabolism is maximum , resulting in an optimal cellular growth rate ( see Figure 1 ) . As such , cells that sufficiently deviate from the average copy number that is optimal for their growth , due to the increasing recklessness of their resident plasmids , will be penalized by slower growth rates , while cells that do not will be rewarded with faster growth rates . In effect , Equation 1 establishes the dependence of the cellular growth rate on the plasmid copy number , which varies over the cell cycle as plasmids replicate autonomously within the host , a process to which we now turn our attention . A wide range of mathematical models of varying specificity and complexity have been proposed for describing the autonomous replication of plasmids within a host , mostly based on the replication systems of plasmids R1 and ColE1 [15] , [27] , [30]–[34] . We adopt a generic approach , according to which an unstable plasmid-coded trans-acting replication inhibitor ( e . g . in the form of antisense RNA ) binds to and deactivates a plasmid-specific target that participates in the initiation process , thereby down-regulating the individual plasmid replication rate [35] . Each plasmid is , first , characterized by a basal replication rate , which indicates the plasmid's replication rate in the absence of any copy number control system . Each plasmid also has a binding affinity to a trans-acting , generic replication inhibitor , which is synthesized by all plasmids at a rate per plasmid , where is the ( short ) averarge lifetime of the inhibitor , resulting in a total inhibitor concentration of in the host , as the inhibitor is diluted with increasing biomass . The weighted response of plasmid to the collectively produced inhibitor modulates its basal replication rate and yields the actual plasmid replication rate according to: ( 2 ) where is the number of plasmids in the host . As the binding affinities are constrained by the limits of physico-chemical interactions , we set . Similarly , we assume that the basal replication rate and the rate of inhibitor production are limited by physiological and biochemical constraints , therefore we also set . Our model also assumes that there is no influence of the host growth rate on the capacity of plasmids to replicate . If that were the case , it would effectively function as a secondary mechanism for CNC . Parameters ( baseline reproduction rate ) and ( obedience ) of a given plasmid act in cis , since their values influence only the replication rate of the plasmid itself , while ( policing ) acts in trans , since its value influences the replication rate of every plasmid within the host through the aggregate term . Plasmid mutations inducing a higher basal replication rate or a lower responsiveness to the collectively produced inhibitor are favored by intra-cellular selection due to the consequent increase of the mutant's replication probability . This drive towards ever-increasing plasmid recklessness is counterbalanced by selection acting at the level of host cells and creates the context within which the evolution of CNC can be investigated .
We begin by focusing our attention to a single cell that contains a population of plasmids that are identical with respect to their replication profile . We initially ignore the stochastic nature of replication and cell division and consider the copy number to be a continuous variable , with both the host growth and the copy number described by deterministic differential equations ( see Equations 1 and 2 in Supplementary Text S1 ) . Our model allows us to establish a relationship between the number of plasmids at the beginning of the cell cycle ( ) and the number of plasmids at the end of the cell cycle at the time of cell division ( ) , where represents the number of plasmids in the resulting daughter cells , assuming equipartitioning during cell division . Figure 2 demonstrates a typical relationship between and , which is dependent on the plasmid replication rate which is , in turn , a function of the plasmid parameters , and ( see Equation 2 ) . Subject to the initial number of plasmids in the parent cell , the resulting daughter cells will have either fewer ( ) , more ( ) or as many plasmids ( ) as their parent cell initially contained . The latter case represents the cross-generational equilibrium of a particular copy number for a given set of plasmid parameter values . This equilibrium can be either stable or unstable , depending upon the response of the system to fluctuations in copy number . The conditions for a stable equilibrium where there is a stable characteristic copy number can be defined according to: ( 3 ) so that when a cell has fewer than plasmids at the beginning of the cell cycle , plasmids over-replicate and when it has more than plasmids , plasmids under-replicate . In both cases , which can emerge as a result of stochasticities in plasmid replication or segregation upon cell division , the tendency is towards an equilibrium future cell cycle with plasmids again . An example of a stable characteristic copy number is the point marked with a filled circle in Figure 2 . The point marked with an open circle on the same curve is an unstable equilibrium copy number , any perturbation to which will lead either to plasmid over-replication or under-replication , the latter case in this instance resulting in movement towards the stable equilibrium . For any configuration of plasmid replication parameters , and , we can determine whether a stable characteristic copy number exists , what its value is , and what is the corresponding cell fitness , which is defined as the reciprocal of the time required for a cell to divide . Figure 3 presents the results of exploring the space of plasmid parameters for two distinct CNC regimes . First , we consider the case in which plasmids have self-determined replication rates independent of the presence of other plasmids in the same host ( , NO-CNC ) . In this case , the only mechanism of controlling the copy number is plasmid self-restraint ( in the form of ) , what early theoretical approaches termed “passive” copy number control [36] . Second , we consider the case of an active negative feedback loop between copy number and plasmid replication rate that is realized through the synthesis of trans-acting replication inhibitors ( , with , CNC ) . In the former case ( NO-CNC ) , we observe that the system has a stable non-zero characteristic copy number for an extremely limited region of basal plasmid replication rates . This stable region is surrounded by regions of plasmid instability characterized by consistent under- or over- replication of plasmids for any initial copy number ( described by the low and high curves in Figure 2 ) . Within the region of plasmid stability , cell fitness increases with until a critical point which marks the transition to instability where plasmids consistently over-replicate . Hence , under no CNC , inter-cellular selection would favor increasing values of the plasmid replication rate ( ) , since this translates to higher cell fitness , until the critical point of plasmid instability is reached . This transition ( denoted by a vertical dotted line in Figure 3 ) occurs at maximum cell fitness and is characterized by the collapse of the stable and unstable equilibrium copy numbers to a singular copy number ( an event that is captured by the critical curve in Figure 2 ) . Consistent over-replication results in future cell cycles with an increasing number of plasmids which leads , eventually , to plasmid explosion and cell death . The activation of the CNC system ( , with ) effectively widens the range of plasmid stability and , crucially , decouples the point of transition to instability from the point of optimal cell growth . As a result , the configuration of plasmid replication parameters that yields optimal cell growth is now surrounded by suboptimal , yet also stable , regions . Under these conditions , inter-cellular selection for increased cell division rates would favor cells containing plasmids with stable copy numbers . Our previous simulations demonstrated that , when stochasticity is ignored , there is only a limited range of that allows for stability in plasmid replication . In the absence of CNC , optimization of the cell division rate would result in plasmid replication rates at the edge of stability , whereas the presence of CNC both broadens the range of stable replication rates , as well as creates a situation in which optimal cellular fitness is located in the interior of this region of stability . We now proceed by extending the deterministic unicellular framework we have considered so far in order to investigate plasmid stability and host performance in the context of stochastic multicellular simulations that describe the asynchronous growth and division ( or death ) of hosts and the autonomous replication of plasmids within such hosts . We introduce stochasticity in plasmid replication by considering the expected number of replication events for plasmids in each host at each discrete time point to be Poisson-distributed , as specified by Equation 2 ( for more details see Supplementary Text S1 ) . The performance of a particular strain in such a stochastic simulation , in which hosts are infected by plasmids with identical replication parameter values , can be evaluated by calculating the average net host growth rate as the difference between the average host division and death rates . Figure 4 displays the resulting fitness landscape of independent strains as a function of the corresponding values of the plasmid replication parameters and given . We note that , due to the homogeneity of the plasmid population , parameters and are interchangeable ( see also Equation 2 ) and , as such , the fitness landscape of and given is identical to the fitness landscape of and given . The region of plasmid stability in this landscape is dominated by the gradient of the net host growth rate leading to an area of optimal growth in which obedience to policing is maximally strong ( ) . Just as in the unicellular deterministic case , the stable region is surrounded by regions of plasmid instability , in which plasmids are eliminated from the host population , due to the absence of a stable characteristic copy number and the consistent under- or over- replication of plasmids . Consistent under-replication leads to the gradual dilution and eventual disappearance of plasmids from the population ( white area below the stable region in Figure 4 ) . Consistent over-replication leads to an elevated copy number that slows down cellular growth ( see also Equation 1 ) , providing more time for plasmids to replicate , thereby further compromising cellular growth . As such , plasmid-free hosts , resulting from stochastic segregational errors , are able to outgrow over-infected hosts , until the population is completely plasmid-free ( white area above the stable region in Figure 4 ) . In the case of extreme plasmid selfishness ( at high , low ) , the host population collapses under the weight of excessive plasmid over-replication , before plasmid-free segregants are given the chance to outgrow over-infected hosts and form a plasmid-free population ( black clusters in Figure 4 ) . The stochasticities in plasmid replication and segregation upon cell division give rise to a distribution of copy numbers in the population that occur across all plasmid-infected hosts over the course of a simulation . We explored the effects of obedience ( or policing , since these are interchangeable due to the homogeneity of the plasmid population ) on the features of these distributions , by considering a cross-section of the fitness landscape for a fixed value of the plasmids' basal replication rate , which corresponds to the optimal net growth rate at maximal CNC ( ) . The weak CNC regime along this cross-section of the fitness landscape ( ) is unstable and plasmids are eventually eliminated from the population ( see Figure 5 ) . The broadness of the copy number distributions in this regime reflects the extensive variation and drift of copy numbers in the population , due to the amplification of stochastic copy number fluctuations [15] . The transformation of the distributions begins at the intermediate range of obedience ( ) , with the emergence of a clear peak; nevertheless , the presence of a heavy tail indicates the persistence of plasmid replication instabilities . These instabilities are minimized in the region of strong CNC ( ) as the distributions become progressively less skewed , due to the increasing efficiency in controlling stochastic copy number fluctuations and the corresponding reduction in copy number variation . At the same time , the discrepancy between the distributions' average copy number and the copy number that is optimal for host growth ( see also Equation 1 ) becomes lower with increasing obedience , thus inducing an acceleration of the net average host growth up to the optimal rate at maximal CNC ( ) . Having explored the effects of homogeneous plasmid cooperation on host growth and plasmid stability , we now ask how these effects influence the evolution of the plasmid replication parameters in the broader context of the conflict between the levels of selection . To this end , we introduce plasmid variation in our multicellular stochastic simulations: each plasmid replication event implies the possibility of mutation with probability , in which case the value of exactly one of the plasmid's replication parameters , chosen at random with equal probabilities , is modified . Starting with an initial population of plasmids that do not respond to inhibitors ( i . e . ) , we allow and to evolve given a fixed rate of inhibitor production . The resulting evolutionary dynamics , shown in Figure 4 , demonstrate the emergence of efficient replication control as driven by the synergies between intra-cellular selection favoring immediate plasmid reproductive gains ( higher selfishness , lower obedience ) , and inter-cellular selection favoring evolutionary adjustments towards those regions of the plasmid parameter space where the net host growth rate increases . In effect , and due to the transient and epigenetic nature of stochastic copy number fluctuations , inter-cellular selection operates upon the net host growth rate accumulated over a few generations and , therefore , upon the copy number distributions associated with particular configurations of the plasmid replication parameters [27] . As such , the evolutionary adjustments favored by inter-cellular selection come in the form of cooperative plasmid parameter mutations , such as increased plasmid self-restraint ( lower selfishness ) or increased sensitivity to the inhibitor ( higher obedience ) , which alter the mode of plasmid replication so as to ensure a reduction , first , in the discrepancy between the optimal and mean copy numbers , and , second , in the magnitude of copy number fluctuations ( i . e . the variation of the copy number distribution ) . This way , the evolution of CNC unfolds with an escalating succession of selfish and cooperative plasmid parameter mutations that develops along the gradient of the host fitness landscape , leading the system towards the region of optimal host growth at maximal CNC . Further escalation is prevented due to the limits that are imposed on plasmid parameter values; the absence of such limits would yield a ratcheting effect whereby the succession of selfish and cooperative mutations would continue indefinitely , limited only by the costs of producing the corresponding factors involved in plasmid replication ( initiators and inhibitors ) . The same cooperative outcome ( evolution of an efficient CNC system ) is obtained when we allow all three plasmid replication parameters , and to evolve from an initial state where plasmids neither produce ( ) nor respond ( ) to inhibitors ( see Figure 6 ) . In this case , cooperative mutations can be either cis-specific ( higher obedience ) as before , or trans-specific ( greater policing ) , in which case a mutation that increases the rate of inhibitor production by an individual plasmid will influence not just the mutant but all plasmids in the intra-cellular replication pool ( due to the term in Equation 2 ) . The cis-specificity of implies that a cooperative mutation inducing a higher sensitivity to the inhibitor is costly at the intra-cellular level , since it decreases the mutant's chances of immediate reproductive success in the replication pool . The trans-specificity of introduces a coercive element to cooperation , because the production of inhibitors regulates the replication of all plasmids in the pool . It also creates the potential for subversive plasmid strategies according to which individual plasmids can gain an advantage in the replication pool by zealously producing the inhibitor ( high policing ) while maintaining a low sensitivity ( obedience ) to that inhibitor themselves . Nevertheless , this scope for opportunistic behavior does not prevent the emergence of the policing CNC mechanism . In fact , we find that plasmid parameter variation within cells is quite low ( see Table 1 ) , so that hosts are inhabited by a , more or less , homogeneous plasmid population ( i . e . plasmids are highly related to their intra-cellular neighbors ) , due to the lack of plasmid migration ( horizontal transmission ) between hosts . The degree of plasmid homogeneity is reduced by approximately an order of magnitude between hosts , compared to its value within hosts , thus generating the host growth differential upon which inter-cellular selection operates by favoring stricter control over plasmid replication . We also investigated the influence of policing costs to the evolution of collective restraint and the overall performance of the population , by introducing an additional cost term in Equation 1 , where is the cost of production per unit of inhibitor paid by the host . This implies that there is now selection at the inter-cellular level against the production of policing resources , due to the associated policing costs that slow down host growth . Figure 7 shows that the increase in policing costs corresponds to a decrease in the production of policing resources ( ) , but not a collapse of plasmid obedience ( ) to policing . On the contrary , obedience is not only sustained but also slightly increases with rising policing costs . At the same time , the basal plasmid replication rate ( ) decreases so as to compensate for the gradual reduction in the availability of policing resources ( ) . As a result , the CNC system remains functional throughout ( since there is still selection at the inter-cellular level for high obedience ) but becomes less efficient with increasing policing costs and the performance of the population deteriorates with lower division rates for hosts and higher rates of segregational loss for plasmids ( see Figure 7 ) . The positive effects of CNC are not limited to hosts but extend to plasmids as well . We evaluated the advantages of CNC for hosts and plasmids by comparing the results of our multi-cellular stochastic CNC simulations ( CNC; evolve ) to those of the baseline model where policing is absent and plasmids replicate independently of the presence of other plasmids in the same host ( NO-CNC; evolves , ) . Host performance was evaluated in terms of the average division and death rates , while the performance of plasmids was assessed on the basis of the fidelity of vertical transmission and the spread of plasmids in the host population . Figure 8 demonstrates that all measures were significantly improved when CNC was functional ( CNC simulations ) , compared to the case where the CNC mechanism was absent ( NO-CNC simulations ) . As such , the beneficial effects of the CNC mechanism on plasmid stability , due to the stricter control of stochastic copy number fluctuations , allow for widespread host infection and the minimization of segregational losses within the margins allowed by inter-cellular selection , thus solidifying the persistence of the plasmid lineage in the population . Finally , we also investigated the persistence and stability of an established policing mechanism among plasmids against invasion by selfish individuals , i . e . plasmids that are insensitive to replication inhibitors and replicate independently of the presence of other plasmids in the same host . More specifically , we simulated the competition between the NO-CNC ( only with ) and the CNC ( ) types ( a ) by mixing both types equally within hosts ( within-host heterogeneity ) and ( b ) by distributing the two types separately and equally among different hosts in the population ( between-host heterogeneity ) . In every case that we examined , we observed the rapid displacement of the selfish type from the population . The complete prevalence of the CNC type demonstrates the robustness and stability of the mechanism of collective restraint in the face of invasion by selfish elements that bypass the policing mechanism in order to gain a relative advantage in the intra-cellular replication pool .
The dependence of vertically transmitted plasmids upon their hosts for survival and propagation mediates the reconciliation of two opposing forces , namely the plasmids' drive towards recklessness with respect to replication , which is favored by intra-cellular selection , and the host's requirement for an optimal configuration of metabolic benefits and burdens , which is favored by inter-cellular selection . The coupling between the levels of selection in our model is defined in terms of the influence of plasmid copy number on host growth; the latter depends on the mode of plasmid replication which , in turn , is a function of the intra-cellular plasmid replication parameters , and ( see Equation 2 ) . The region of stability in the plasmid parameter space is defined by the existence of stable characteristic copy numbers ( see Equation 3 ) . In the absence of CNC ( ) , where plasmids replicate independently of the presence of other plasmids in the same host , the edge of plasmid stability coincides with optimal host growth . In an evolutionary context , this implies that plasmids will evolve to the edge of plasmid stability driven by intra-cellular selection ( which favors higher ) and this drive will be consistent with selective forces at the inter-cellular level since host fitness is also increasing . The transition to instability stimulates the conflict between the levels of selection as net host growth slows down due to the increasing metabolic costs resulting from the consistent over-replication ( lack of inter-generational stability ) of the selfish mutants . The activation of the CNC system ( , ) mitigates the detrimental effects of these transient tensions between the levels of selection by expanding the region of plasmid stability so that the plasmid parameter configuration that yields optimal growth no longer coincides with the edge of plasmid stability . CNC is realized by cooperating plasmids that participate in the policing mechanism that they themselves construct and maintain . This introduces a collectivist element to the repression of competition in the intra-cellular replication pool , due to the trans-specificity of the plasmid-coded replication inhibitors . Increasing cooperation through collective restraint has beneficial effects on hosts and plasmids alike , by strengthening the system's defenses against copy number fluctuations resulting from stochasticities in plasmid replication and segregation upon cell division . These fluctuations are reflected in the copy number distributions , whose variation decreases with increasing CNC ( see Figure 5 ) . As a result , hosts grow faster and plasmids maximize their spread in the population and minimize their loss due to segregational errors , within the margins allowed by inter-cellular selection . The landscape of population growth , expressed as a function of the plasmid replication parameters , is dominated by a gradient that leads progressively to the region of optimal growth at maximal CNC ( see Figure 4 ) . The synergies between the levels of selection are reflected in the co-evolution of the plasmid replication parameters that develops along this gradient and unfolds with a succession of selfish and cooperative plasmid mutations that drive the system to its optimal state at full cooperation . Along this gradient , we find a most accurate alignment between the average copy number ( a proxy for the characteristic copy number ) and the copy number that is optimal for host growth . We expect the accuracy of this alignment to decline when plasmids adopt migratory strategies such as conjugative transfer , the horizontal nature of which undermines the strong mutualistic nature of the host-symbiont relationship . The regulation of plasmid replication by means of the CNC mechanism is characterized by a more general trade-off between the immediate reproductive gains of an individual and the longer-term success of the collective to which that individual belongs; the latter can only be improved if the urge to satisfy the former can somehow be repressed . To this end , the repression of competition in the plasmid replication pool is achieved through the obedience ( binding affinity ) of plasmids to the policing resource ( replication inhibitor ) that they themselves produce . Hosts in which plasmid obedience to policing is strong or under development , will outgrow fellow hosts in which obedience is weaker or absent , thus motivating the reinforcement of cooperation and the eventual establishment of the CNC mechanism . | Mutual policing constitutes an important mechanism for the emergence and maintenance of cooperation through the repression of intra-group competition among a population of self-interested individuals . Existing models of mutual policing have been highly abstract and distant from the properties of real biological systems . In this paper , we construct a bottom-up , multi-scale computational model reflecting the biology of , perhaps , the simplest example of such a mechanism: replication control in non-conjugative plasmids , a class of vertically transmitted , molecular symbionts of bacteria . We simulate the emergence of plasmid copy number control through the co-evolution of two interacting plasmid traits: policing , realized as the production of trans-acting replication inhibitors , and obedience , expressed as plasmid-inhibitor binding affinities . We demonstrate and explain the intrinsic replication instabilities that arise in the absence of policing and we show how increasing levels of policing and obedience resolve these instabilities and improve both plasmid stability and host performance . | [
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
] | [
"coevolution",
"population",
"modeling",
"evolutionary",
"modeling",
"biology",
"computational",
"biology",
"evolutionary",
"biology",
"evolutionary",
"processes"
] | 2013 | The Evolution of Collective Restraint: Policing and Obedience among Non-conjugative Plasmids |
Mathematical modelling has proven an important tool in elucidating and quantifying mechanisms that govern the age structure and population dynamics of red blood cells ( RBCs ) . Here we synthesise ideas from previous experimental data and the mathematical modelling literature with new data in order to test hypotheses and generate new predictions about these mechanisms . The result is a set of competing hypotheses about three intrinsic mechanisms: the feedback from circulating RBC concentration to production rate of immature RBCs ( reticulocytes ) in bone marrow , the release of reticulocytes from bone marrow into the circulation , and their subsequent ageing and clearance . In addition we examine two mechanisms specific to our experimental system: the effect of phenylhydrazine ( PHZ ) and blood sampling on RBC dynamics . We performed a set of experiments to quantify the dynamics of reticulocyte proportion , RBC concentration , and erythropoietin concentration in PHZ-induced anaemic mice . By quantifying experimental error we are able to fit and assess each hypothesis against our data and recover parameter estimates using Markov chain Monte Carlo based Bayesian inference . We find that , under normal conditions , about 3% of reticulocytes are released early from bone marrow and upon maturation all cells are released immediately . In the circulation , RBCs undergo random clearance but have a maximum lifespan of about 50 days . Under anaemic conditions reticulocyte production rate is linearly correlated with the difference between normal and anaemic RBC concentrations , and their release rate is exponentially correlated with the same . PHZ appears to age rather than kill RBCs , and younger RBCs are affected more than older RBCs . Blood sampling caused short aperiodic spikes in the proportion of reticulocytes which appear to have a different developmental pathway than normal reticulocytes . We also provide evidence of large diurnal oscillations in serum erythropoietin levels during anaemia .
Mathematical modelling has proven an important tool in elucidating and quantifying the mechanisms that govern physiological processes . Here we are concerned with understanding the processes responsible for the age structure and population dynamics of red blood cells ( RBC ) . There are two main fields of research developing mathematical models of the dynamics of the erythropoietic system . Interestingly , both fields appear to progress in isolation from each other . The first , pioneered by Mackey [1] , has mainly focused on explaining periodic haematological diseases by exploring how the feedback from the circulating RBC population influences self-renewal , maturation and apoptosis of RBC progenitors ( erythroid progenitors ) in bone marrow ( [2]–[9] , see [10] for a detailed review of the literature ) . The approach is fairly theoretical: modifications to previous models—in the light of new evidence in the literature—are examined from a dynamical systems point of view , often with an emphasis on integrating partial differential equation models to time-delay ordinary differential equation models . Model testing against real data is rarely undertaken ( [9] is one exception ) , and fits are qualitative . The second field of research , pioneered by Veng-Pedersen and co-workers [11] , has focused on developing pharmaceutic PK/PD models to explain perturbations to RBC , immature RBC ( reticulocyte ) , erythropoietin ( Epo ) and haemoglobin dynamics induced by phlebotomy , recombinant human Epo or chemotherapeutic drugs [12]–[18] . This approach is directed at explaining reticulocyte dynamics as this is commonly used for clinical diagnoses and management of anaemia ( see [16] and references therein ) . Model fitting to data and parameter estimation is integral to this approach . One of our main areas of study is early blood stage malaria infection of mammalian hosts , particularly in mice as these are our main experimental model of malaria infection . Malaria parasites invade RBCs , killing them in the process and causing , in some cases , severe anaemia . Some species also prefer to invade reticulocytes over normocytes [19] . In our work , we use mathematical models to elucidate and quantify the processes that govern RBC and malaria parasite dynamics in infections [20] . The standard model for RBC production in malaria infection models is to assume that RBC growth rate is proportional to the instantaneous difference in normal and anaemic RBC concentrations ( see , for example , [21] , [22] ) . Extensions to this model include a time lag between RBC concentration and growth rate [20] and age structure [23] . Although relatively simple , these models may be sufficient to explain malaria infection data . Whether they are or not , however , has never been adequately tested . The aim of the work described in this paper is to gain a better understanding of the processes that govern age structure and dynamics of reticulocyte and RBC concentrations under anaemic conditions in general . The results can then be used to inform our mathematical models of malaria infections . A complicating factor with modelling the erythropoietic system in malaria infections is that the immune responses against the parasite are known to interact with the erythropoietic system [24] . It is therefore easier to elucidate erythropoietic mechanisms under anaemic conditions in the absence of immune responses . One method of inducing anaemia without an immune response is by administering phenylhydrazine ( PHZ ) . PHZ has been used for over 100 years as a means to induce haemolytic anaemias in experimental animals [25] . For our purposes it perturbs the RBC population away from equilibrium . The resulting transient dynamics give us a window into the mechanisms that govern RBC population dynamics . Quantitative fitting of mathematical models to data integrated with statistical analysis is a powerful and sensitive method of testing and comparing hypotheses and for development of evidence based models . Here , we attempt to fit a set of mathematical models constructed from combinations of sets of competing hypotheses to some new experimental data from our lab . These hypotheses reflect uncertainty in the qualitative and quantitative nature of mechanisms that govern RBC age structure and dynamics . In particular , there are five mechanisms we wish to examine in detail: the feedback from RBC concentration to production rate of reticulocytes in bone marrow , the release of reticulocytes from bone marrow and their subsequent ageing and clearance in the circulation , the effect of PHZ on RBC age structure , and the effect of blood sampling on reticulocyte proportion . The preliminary phase of the work reported here was a trial model development and fitting exercise to some existing unpublished data on PHZ treated mice . These data consisted of RBC and reticulocyte concentrations measured on day 0 and days 3–8 after PHZ treatment . Reticulocytes take about 1 to 2 days to mature into normocytes in the circulation . In normal conditions they are present at very low concentrations of about 1% of total RBCs . Anaemia causes their proportion to rise dramatically as production of erythroid progenitors in the bone marrow is up-regulated in response to increased Epo production in the kidney . The reticulocyte proportion is an excellent variable to test hypotheses about erythropoietic mechanisms because cells pass rapidly through this development stage , thus providing a much finer scale resolution on RBC dynamics than total RBC concentration . The trial model fitting exercise informed the experiments described in this paper . We used more mice ( 25 instead of 3 ) to improve statistical inference . We sampled for longer ( 114 days instead of 9 ) to examine return to equilibrium . We sampled twice per day on days 5 to 9 to characterise the fast reticulocyte dynamics that occur during this period . We measured serum Epo concentrations to relate RBC concentration during anaemia to production and release rates of bone marrow reticulocytes . We quantified RBC concentration and reticulocyte count error structures and variances to perform statistical analyses . We performed two control experiments: one to control for blood sampling , another to control for housing conditions in order to prevent aggressive interactions . In the Results section we describe these experiments , the justification and formulation of the hypotheses we tested , and the statistical techniques we used to test them .
Our best fitting model consisted of hypotheses A1 , B1 , C1 , D1 , and E1 . The fits to all mice are shown in Figure 4 ( A: first 14 days , B: all data , top panels: RBC concentration , bottom panels: reticulocyte proportion ) . The solid lines are the median fits , and the grey regions the 95% posterior predictive intervals . More instructive plots for model assessment are the overlaid standardised residuals for RBC concentration ( Figure 5 , top panel , note nonlinear ) and reticulocyte proportion ( Figure 5 , bottom panel ) of all mice in the main experiment . Inadequate model fits are suggested by extreme residuals and serial correlation between residuals . The fits to RBC concentration are generally adequate but there are two causes for concern . The RBC concentration on day 0 , i . e . , just before PHZ treatment , is underestimated ( Figure 5 , top panel , ) . It is clear in Figure 1B that RBC concentration on day 0 is significantly higher ( mean of 9×109 cells/ml ) than after about day 60 when presumably RBC concentration has returned to equilibrium ( 8 . 5×109 cells/ml ) . Moreover , mean RBC concentration on day 0 in the control experiments ( Figure 3 ) is lower than in the main experiment . These two pieces of evidence lead us to suggest that RBC concentrations on day 0 in the main experiment were higher than normal . This could be because of pre-experiment conditions or transportation , although we can only speculate on this . The second cause for concern is the serial correlation of residuals between days 13 and 61 . This corresponds to the transition from anaemia to polycythemia and return to equilibrium . The reticulocyte proportion data is not explained as well as for the RBC concentration data . The main causes for concern are an underestimation of the peak on day 9 , and an overestimation on day 26 . Between day 19 and 26 there is a significant , if albeit small , drop in reticulocyte proportion ( Figure 1B , inset ) that is not explained in our models by reduced production or release rate of reticulocytes caused by polycythemia . The model also overestimates reticulocyte proportion on day 61 when RBC concentration returns to equilibrium ( Figure 1B ) . Aside from the possibility that sampling-influences on reticulocyte proportion exist and show temporal variation , we do not have any hypotheses to explain these discrepancies . In the following analysis we take as our baseline model the model consisting of hypotheses A1 , B1 , C1 , D1 and E1 ( one of the best fitting models ) . Then , taking each mechanism in turn , we replace its corresponding baseline hypothesis with another hypothesis of that mechanism . These models are fitted to the data and compared to the baseline model using Bayes factors ( Table 2 ) . During day 7 , the reticulocyte proportion in mouse 8 dropped dramatically to almost normal levels ( 0 . 036% , Figure 4A ) , before rapidly rising again . This could be because the blood smear was mixed up with another , although we are certain this did not occur . It could be an anomalous count , although the probability of counting just 18 reticulocytes in 500 RBCs instead of around 117 reticulocytes ( which is the average of the other 9 mice at this time ) is unlikely ( ) . Or it could be a real phenomenon . None of our models capture this behaviour . Instead we have assumed in the above analysis , and for this mouse only , that the release rate of reticulocytes from time to returns to baseline . The marginal distributions for is 126 ( 121 , 131 ) hrs and for 64 ( 54 , 78 ) hrs ( Figure 9 ) . We have tried allowing a drop in production rate but with no success . We have no explanation for why a temporary return to baseline release rate may have occurred only in this mouse .
The aim of this paper was to synthesise ideas from previous experimental data and the mathematical modelling literature with new data in order to test old and new hypotheses about mechanisms that govern reticulocyte and RBC age structure and population dynamics in bone marrow and the circulation of mice . In contrast to previous model fitting studies we quantified measurement error . This allowed us to perform model fitting , model testing , model comparison and parameter estimation under a Bayesian framework . We tested multiple hypotheses that reflect uncertainty in erythropoietic mechanisms . Some of the hypotheses do not adequately explain the data and others do . So although we can rule out certain hypotheses , we are left with some that our data cannot discriminate between . This therefore suggests which mechanisms should be explored with further experiments and modelling . The picture of the erythropoietic system emerging from this synthesis of data and modelling is as follows . Under normal conditions ( see Figure 8 ) reticulocytes are produced at about 107 cells/ml serum/hour in bone marrow . Reticulocytes mature for between 2–4 days with limited release ( with rate constant of about 10−3 hr−1 ) into the circulation ( we estimate about 2–5% of reticulocytes are in the circulation with the rest in bone marrow ) . Once matured into normocytes , any cells remaining in the bone marrow are immediately released into the circulation . RBCs undergo random clearance with a rate constant of about 0 . 002 hr−1 independent of age , which causes a negative exponential age distribution . On reaching about 50 days old there is rapid clearance of any remaining RBCs ( we cannot estimate this rate with our data ) . We postulate that treatment with a single dose of PHZ may age rather than kill RBCs . Older RBCs are aged past their maximum lifespan and are immediately cleared causing a rapid loss of RBCs within 24 hrs . Younger RBCs , which are affected more than older RBCs by PHZ , are not immediately cleared , but cause a higher than normal clearance rate for several days until all are lost ( see upper three panels in Figure 7 ) . Serum Epo response is delayed for about 3 days , and may correlate with a delay in increased production of reticulocytes in some mice . We have also discovered that Epo appears to exhibit pronounced diurnal oscillations at least during anaemia , but we cannot tell if this causes oscillations in reticulocyte production rate . Production rate of reticulocytes is linearly and negatively correlated with RBC concentration—at least across the range of concentrations we observed . On treatment with PHZ , release rate of reticulocytes from bone marrow increases immediately and dramatically . We estimate that at maximum anaemia , between 35 to 95% of reticulocytes were in the circulation . After about 12 days of increased production , RBC concentration overshoots normal levels causing polycythemia . This results in a lower than normal reticulocyte production rate causing a lower than normal reticulocyte proportion . Normal equilibrium conditions are restored after the PHZ-induced cohort of RBCs reach the end of their lifespan . Blood sampling caused aperiodic and pulsed release of reticulocytes into the circulation in mice housed in groups but did not in mice housed individually . Reticulocytes comprising spikes appear to have a different developmental pathway than normal reticulocytes because their mechanism of release into the circulation is different . Erythropoietin mediates the negative feedback between haemoglobin concentration in the circulation and apoptosis of erythroid progenitor cells that develop into cell-cycle arrested reticulocytes [34] . We initially included erythroid progenitors in our model and assumed that they were irreversible committed to differentiate into reticulocytes . Under this assumption , we found that all the parameters that governed their age distribution were unidentifiable given the data we had . All we were able to quantify was the rate at which erythroid progenitors differentiate into reticulocytes in bone marrow ( ) and how this rate is modified by RBC concentration ( ) . Including erythroid progenitors in the model did not improve its fit . In contrast , [9] assume that erythroid progenitors take 4 days to mature but do not estimate its value from model fitting , and [15] estimate their maturation time in rats to be 43 hrs with a coefficient of variation of 7 . 5% by fitting to data . Although our model can adequately explain most of the data , where it does not , raises some interesting questions . For example , by what mechanism does blood sampling induce spikes in reticulocyte counts ? What causes the temporal pattern of spikes and why were they synchronous in the ten main experimental mice but not in the control mice ? Why were spikes less apparent and reticulocyte proportion lower in individually housed mice compared to grouped mice ? What caused the return to baseline release rate in mouse 8 ? Why is there a 3 day delay in increased serum Epo concentration ? Also our model does not appear to adequately explain RBC and reticulocyte dynamics between days 13 to 61 when RBC concentration overshoots its normal levels and becomes polycythemic . We hope to address theses questions with further experimental and modelling work .
All experimental procedures were regulated and carried out under the U . K . Home Office Animals ( Scientific Procedures ) Act , 1986 . Adult male MF1 mice were used for all experiments for two reasons: their large size reduces the effect of blood sampling their erythropoietic state , and their docile and non-aggressive nature reduces the chance that stressful interactions when housed in groups affects RBC and reticulocyte dynamics [35] , [36] . At each sampling time point we took 2 µl blood to measure RBC concentration , we took a blood smear with approximately 1 µl blood to measure reticulocyte proportion and we took 20 µl blood to measure serum Epo concentration . We diluted 1∶40 , 000 the 2 µl blood sample and measured RBC concentration by coulter counter . We stained blood smears with 10% Giemsa in pH 7 . 2 buffer and estimated reticulocyte proportion by counting the number of reticulocytes observed in at least 500 RBCs by light microscopy . We measured Epo concentration using an Elisa adapted from [37] to measure Epo concentration from small blood samples and thus permit repeat sampling of each mouse . We collected 20 µl of blood in 5 µl of heparin for each sample point . We centrifuged this at 13 , 000 rpm for 3 minutes , and collected and stored the serum at −80°C . We prepared the Elisa plates by adding of 4 µg/ml rat anti-mouse Epo IgG1 antibody ( BD , 554651 ) before incubation at 4°C overnight . We washed the plates with 0 . 1% Tween 20 in PBS and blocked them with of 1% Bovine Serum Albumin ( BSA ) for 1 hr at room temperature . We loaded the plates with rmEpo ( Roche , 112769640010 ) standards in a 2× dilution series ( 500-0 . 98 mU/ml ) . We defrosted the serum samples and diluted 20 µl of each in 80 µl of blocking buffer which we then used to make a 2× dilution series for each sample . We left the plates to incubate overnight at 4°C . After washing 3 times , we added , at 10 µg/ml , of polyclonal rabbit anti-hEpo antibody ( R&D , AB286NA ) and incubated them at room temperature for 1 hr before washing a further 3 times . We incubated the plates with of HRP-conjugated goat anti-rabbit IgG ( Bio-Rad , 1708241 ) at a 1∶3000 dilution in blocking buffer at room temperature for 1 hr . We detected the bound conjugate by adding 0 . 1ml/well of ABTS substrate with 0 . 036% hydrogen peroxide and developed for 20 minutes before reading at wavelength 405 nm with 595 nm as reference . We obtained the concentrations of Epo in the serum samples by comparison with values of the rmEpo standard curve . The main experiment consisted of 10 mice treated with 40 mg/kg PHZ intraperitoneally and housed together . Their blood was sampled 35 times: from just prior to PHZ administration to 113 days post treatment . We measured RBC concentration and reticulocyte proportion for each sampling point and measured Epo concentration up to day 42 . As a temporal control for blood sampling , we took the 20 µl of blood required for Epo measurement at each sample point for all mice in all experiments whether or not it was used for Epo measurement . As our trial experiment on PHZ-treated MF1 mice demonstrated fast reticulocyte dynamics from days 5 to 9 post PHZ treatment we sampled twice per day during this period to ensure improved model discrimination . We took samples daily for days 0 to 4 inclusive and days 10 to 14 inclusive and weekly for another 13 weeks . We took samples at 9am , and , for twice daily sampling , at 3pm . We carried out additional experiments to investigate any effects of blood sampling and group housing on erythropoietic dynamics . This involved monitoring reticulocyte proportion and RBC concentration in three groups of five mice: group-housed , non-PHZ-treated mice , group-housed , PHZ-treated mice , and individually-housed , PHZ-treated mice . Sampling frequency was the same as for the main experiment except that samples were only collected up to day 14 . We use a two-compartment , continuous-time , age-structured formalism for our models . The two compartments are bone marrow and the circulation . We model reticulocyte concentration in bone marrow and the circulation ( with conversion to a proportion of total RBC concentration for model fitting ) and RBC concentration in the circulation . In this section , we describe the mathematical formulation of the base model . We then consider a range of competing hypotheses that reflect uncertainty in the mechanisms of the erythropoietic system . The youngest bone marrow cells we consider are erythroid progenitor cells that have just differentiated into cell-cycle arrested reticulocytes ( Figure 8 ) . We define their age to be 0 . Under normal equilibrium conditions we assume that their production rate is with units of cells/ml blood serum/hr . We assume that reticulocyte maturation time is hours . The production rate of reticulocytes is mediated by erythropoietin ( Epo ) , a glycoprotein hormone produced in the kidney and other organs [38] , [39] . Epo controls erythroid progenitor growth by retarding their DNA breakdown and preventing apoptosis [34] . It was our intention to construct models that related reticulocyte production ( from erythroid progenitors ) to serum Epo concentration which in turn was related to the difference between normal and anaemic circulating RBC concentrations . However , on analysis of our Epo data , we concluded that we could not have absolute confidence in the precision of the Elisa data , possibly due to the small volumes of blood analysed and the accuracy of interpolation from our standard curve . Moreover , we had no quantitative estimate of the experimental error in Epo concentration so we could not assess the adequacy of a model fit to it . Instead we directly relate reticulocyte production rate to the difference between normal and anaemic circulating RBC concentrations and neglect serum Epo concentration as an intermediate regulator . We therefore assume that under anaemic conditions reticulocyte production rate , is multiplied by a function which is the fold-change in reticulocyte production rate compared to normal . Time since PHZ administration is with units of hours and is the circulating RBC concentration at time with units of cells/ml . In normal conditions in humans , reticulocytes mature for about 3 days in bone marrow , are released into the circulation , and complete maturation within a day [32] . As RBC concentration falls , maturation time remains constant at about 4 days but residence time in bone marrow decreases linearly with RBC concentration [32] . Other studies have demonstrated that reticulocytes in humans , sheep and mice are released early from bone marrow in phlebotomy-induced anaemia [13] , [16] , [32] and intravenous dosing with recombinant human Epo [31] . Immature and mature reticulocytes can be differentiated by their RNA content because they lose RNA as they mature . Data from humans [31] show that , in normal conditions , about 5% of reticulocytes in the circulation are immature ( using the definition of maturity in [31] ) and 95% are mature . Stimulus of erythropoiesis , in this case by an injection of Epo , caused an immediate release of immature reticulocytes from bone marrow into the circulation . There was no increase of mature reticulocytes for another 3 days . This suggests the inclusion of the following two processes in our model . i ) In normal conditions , most immature reticulocytes remain in bone marrow with some small amount of migration into the circulation . ii ) Mature reticulocytes are released into the circulation before maturation into normocytes . We formulate these processes as follows . In normal conditions we assume that bone marrow reticulocytes are released into the circulation at a rate . As for reticulocyte production rate , we model the direct relationship between the difference in normal and anaemic RBC concentrations and release rate , ignoring the intermediate regulation by Epo . Thus we assume that is modified under anaemic conditions by the function which is the fold-change in reticulocyte release rate compared to normal . We assume that there is a maximum residence time , of cells in the bone marrow . If all reticulocytes exit the bone marrow before maturation , and if reticulocytes mature into normocytes before entering the circulation Figure 8 . Let be the age distribution of bone marrow reticulocyte ( plus normocyte if ) concentration with units of cells/ml/hour . The partial differential equation describing its rate of change is given by ( 1 ) where is cell age . The left-hand-side ( LHS ) of Equation 1 describes cell ageing , the first term on the right-hand-side ( RHS ) describes release of cells between ages 0 and from bone marrow into the circulation and the second term ( a Dirac-delta function ) describes release of any remaining cells at age . Let be the age distribution of circulating RBC concentration ( reticulocytes plus normocytes ) with units of cells/ml/hour . RBCs are cleared by phagocytosis in the spleen , liver and bone marrow . Cohort labelling with radioiron in mice suggests that RBCs are randomly cleared at an age-independent rate , with negligible RBCs surviving more than about 50 days [33] . In humans , however , there appears to be little clearance before about 120 days with rapid clearance thereafter [40] . We combine these two pieces of evidence and assume that , up to days old ( about 50 days ) , RBCs are randomly cleared at a relatively slow rate . After that , clearance rate accelerates . With our data we can estimate , but we cannot quantify the acceleration in clearance rate . We therefore assume that RBCs have a maximum lifespan of hours; in other words , any RBCs that reach the age of are immediately cleared . The partial differential equation describing the rate of change of circulating RBC age distribution is given by ( 2 ) The LHS of Equation 2 describes RBC ageing , the first two terms on the RHS describe release of bone marrow cells into the circulation , and the last two terms describe RBC random clearance and immediate clearance at their maximum lifespan respectively . The boundary conditions for Equations 1 and 2 at are the production rate of bone marrow reticulocytes and 0 ( because RBCs enter via the bone marrow ) respectively , i . e . ( see Figure 8 ) , ( 3 ) ( 4 ) Note that , in normal conditions . The concentrations of circulating reticulocytes , and circulating RBCs ( reticulocytes plus normocytes ) , are ( 5 ) ( 6 ) All variables , functions and parameters are listed in Table 3 . In the following sections we formulate the hypotheses we will test . | Red blood cells are made in the bone marrow and released into the blood stream . Their population is kept at equilibrium by complex negative feedback mechanisms . Genetic diseases , pathogens such as malaria , and extreme environmental changes can alter this equilibrium , either directly by prematurely killing red blood cells or indirectly through disrupting these feedback mechanisms . In this paper we test competing hypotheses about the nature of some of these mechanisms by fitting mathematical models to experimental data . Our results give us a better understanding of the mechanisms that govern the red blood cell population and will improve models of diseases that affect red blood cells . | [
"Abstract",
"Introduction",
"Models",
"Discussion",
"Methods"
] | [
"mathematics/statistics",
"hematology/hematopoiesis",
"hematology/anemias",
"computational",
"biology"
] | 2009 | Quantitative Analysis of Mechanisms That Govern Red Blood Cell Age Structure and Dynamics during Anaemia |
We studied the global relationship between gene expression and neuroanatomical connectivity in the adult rodent brain . We utilized a large data set of the rat brain “connectome” from the Brain Architecture Management System ( 942 brain regions and over 5000 connections ) and used statistical approaches to relate the data to the gene expression signatures of 17 , 530 genes in 142 anatomical regions from the Allen Brain Atlas . Our analysis shows that adult gene expression signatures have a statistically significant relationship to connectivity . In particular , brain regions that have similar expression profiles tend to have similar connectivity profiles , and this effect is not entirely attributable to spatial correlations . In addition , brain regions which are connected have more similar expression patterns . Using a simple optimization approach , we identified a set of genes most correlated with neuroanatomical connectivity , and find that this set is enriched for genes involved in neuronal development and axon guidance . A number of the genes have been implicated in neurodevelopmental disorders such as autistic spectrum disorder . Our results have the potential to shed light on the role of gene expression patterns in influencing neuronal activity and connectivity , with potential applications to our understanding of brain disorders . Supplementary data are available at http://www . chibi . ubc . ca/ABAMS .
While the brain can be studied at many different scales and with many modalities , one of the most established is the study of brain regions and their connectivity . These “macroconnections” between neuroanatomically-defined brain regions are thought to number between 25 , 000–100 , 000 in the mammalian brain [1] , forming a complex network . Knowledge of the “connectome” is used to diagnose neurological disorders such as ischemic stroke , to interpret brain imaging results and to computationally model the brain . There is also growing evidence of connectivity abnormalities in disorders such as autism and schizophrenia [2] , [3] , [4] . Because of the fundamental importance of the wiring of the brain , there has been a recent push to create more comprehensive “connectome” maps [5] , [6] , paralleling efforts to understand the brain at the level of genes . The most comprehensive studies of connectivity have been done in the worm Caenorhabditis elegans ( at the level of single neurons ) and the macaque monkey [7] , [8] . Recent work has begun plumbing the properties of these networks , examining node degree distribution [9] , network motifs [10] , and modularity [11] . It has been shown that anatomical neighbours tend to be connected [12] , and there is evidence that wiring cost partially explains network structure [13] , [14] . There is also increasing interest in the integration of neuronal connectivity and information about genes . This is in part driven by the fact that many genes show spatially-restricted or varying expression in the nervous system , but in many cases the reasons for the expression patterns are not clear [15] , [16] , [17] , [18] . The idea that gene expression is related to connectivity is not new . For example , the expression of a transmitter must be coupled with expression of appropriate receptors in the postsynaptic target . To regulate neurite outgrowth and plasticity hetero- and homophilic cell adhesion molecules require appropriate expression patterns in connected neurons [19] , [20] . In a study of the mouse hippocampus , Dong et al . [21] identified seven genes which are differentially expressed between the dorsal and ventral CA1 field and have a correlated expression pattern in the corresponding projection fields in the lateral septal nucleus . The availability of detailed information on expression patterns in the mouse brain [15] , [16] , [17] suggests that a global examination of gene expression and connectivity in the mammalian brain would provide additional insights . While there is no large-scale analysis of gene expression and connectivity in the mammalian brain , three groups have examined this issue in the nematode worm Caenorhabditis elegans . The groups used cellular level expression data for a few hundred genes and a neuron level connectivity map [8] . By combining the data , Kaufman et al . [22] used classification and Mantel tests to predict genes involved in synaptogenesis and axon guidance . They concluded that expression profiles of neurons “carry significant information about their connectivity” . Varadan et al . [23] used a different methodology to discover biologically meaningful gene sets that provide connectivity information . Within the resulting gene sets they found high levels of multivariate synergy , suggesting interacting genes are more important than single genes . In a third study , Baruch et al . [24] predicted a neuron's postsynaptic partners using expression patterns of a small number of interacting genes . In this paper we examine gene expression patterns and macroconnectivity in the adult rodent brain , using data from the Allen Brain Atlas [17] and the Brain Architecture Management System [25] , [26] . Our results suggest that in the mammalian brain , as in Caenorhabditis elegans , there is a correlation between gene expression and connectivity , and the relevant genes are enriched for involvement in neuronal development and axon guidance .
We obtained data sets of macroconnectivity in the rat brain and gene expression data on mouse ( see Materials and Methods and Figure 1 ) . By carefully mapping brain regions across them , we identified 142 distinct ( non-overlapping ) brain regions in common ( the “common” regions; see Materials and Methods ) . In total these regions account for nearly half of the volume of the brain . A notable omission is many regions of the neocortex , which is not sub-parcellated in our data set . The expression data set , which is filtered to remove unexpressed genes ( see Materials and Methods ) consists of the expression levels of 17 , 530 genes in the 142 regions . Because many genes were assayed more than once in the Allen Atlas ( independent “image series” in their terminology ) , there are 22 , 771 rows in the expression data matrix . The connectivity data consists of the connectivity profiles of 942 regions with the 142 common regions ( Figure 1 ) . In this binary matrix , a value of 1 at index ( i , j ) indicates a connection exists between region i and region j . In most of our analyses , we considered the directionality of connectivity . Of the 142 common regions , 112 have efferent ( outgoing ) connections , and 141 have afferent ( incoming ) connections; there are 5216 outgoing connections and 6110 incoming connections . Our results are based on various direct and indirect comparisons of the connectivity and expression data matrices or their corresponding correlation matrices . We began our study with some relatively simple analyses designed to explore the relationship between connectivity , gene expression and other parameters such as spatial distribution and size of brain regions . We first tested the simple hypothesis that regions which are connected might have more similar expression patterns . This is in effect a more global search for patterns like the ones identified by Dong et al . [21] ( note that the CA1 subregions studied by Dong et al . were not represented in our data ) . To do this we compared the distribution of correlations in expression profiles for regions which are connected to the distribution for regions that are not connected ( Figure S1 ) . We found that on average , regions that are connected ( ignoring directionality; 912 connected pairs among the 142 regions ) have more similar expression profiles than the 8 , 187 non-connected region pairs ( 0 . 79±0 . 06 for connected; 0 . 76±0 . 06 for unconnected; p-value<2 . 2×10−16 , t-test ) . This is an initial indication that structural connectivity and gene expression are related . We then examined the degree of connectivity of a region with its expression profile . The degree of connectivity is computed by summing the columns of the connectivity matrix in Figure 1 . The correlation of this vector was then computed with each gene expression profile ( the rows of the expression matrix ) . After correcting for multiple testing , 887 and 1127 genes ( represented by 929 and 1175 Allen Brain Atlas image series , respectively ) had expression levels positively and negatively correlated with the number of connections , respectively . The highest rank correlations between expression levels and connectivity degree were ∼±0 . 64 . While the interpretation of this result is not clear ( a Gene Ontology annotation enrichment analysis did not yield any strong patterns ) , we noted that all three neurofilament cytoskeleton genes ( light [NCBI gene ID:18039] , medium and heavy neurofilament polypeptides , Nefl-3 ) are negatively correlated with connectivity; that is , they are expressed at higher levels in regions that have few connections . Neurofilament content is correlated with axonal diameter and length , with enrichment in motor and long-projecting neurons [27] , [28] , [29] and our results suggest another relationship with connectivity . We found that the size of a region is significantly correlated with its connection degree ( Spearman's rank correlation , ρ = 0 . 22 ) . We also noted that the more posterior the region , the fewer connections it has ( ρ = 0 . 55 ) . Regions containing motor neurons that project long axons to the spinal cord or muscles were found to have significantly fewer connections ( they also tend to be in posterior locations; p-value = 1 . 32×10−6 , Wilcoxon–Mann–Whitney test ) . Table S1 provides brain region statistics for degree , location and motor classification . While the above analyses suggest some interesting generic patterns relating connectivity to expression and other parameters , they are not able to expose more complex relationships . Like Kauffman et al . [22] and Varadan et al . [23] , we hypothesized that expression patterns carry information about specific neural connectivity patterns involving multiple regions . To test the global correlation between expression and connectivity profiles we used the Mantel test . Unlike the test used above to examine the relationship between pair-wise connectivity and expression patterns ( using the direct connectivity matrix ) , here we are asking if the similarity of the connectivity profiles of two regions is related to the similarity of the expression profiles of the two regions , regardless of whether those two regions are themselves connected . In this analysis we are comparing the correlation matrices for the expression data set and the connectivity data ( Figure 1 ) . A key finding is that , as in Caenorhabditis elegans ( at the level of individual neurons ) , we find that brain regions that have similar connectivity patterns tend to have similar patterns of gene expression . The Mantel correlation ( “correlation of correlations” ) between expression and incoming connectivity patterns ( 141 regions ) is 0 . 248 ( p-value<0 . 0001 ) . Using the outgoing connectivity profiles for 112 regions yielded a correlation of 0 . 226 ( p-value<0 . 0001 ) . This relationship holds separately for some of the 5 major neuroanatomical divisions in the Allen reference atlas . For outgoing profiles the Mantel test is significant at p-value<0 . 001 for the interbrain ( r = 0 . 42 ) , cerebrum ( r = 0 . 30 ) and hindbrain ( r = 0 . 21 ) divisions but not midbrain or cerebellar divisions . For incoming connectivity only the cerebrum ( r = 0 . 29 ) and interbrain ( r = 0 . 34 ) divisions have significant Mantel correlations with expression . We note that unlike our observation of similar expression profiles among connected regions , here we are comparing connectivity patterns of regions , which does not require that the regions be connected to each other . One factor in this analysis is that regions which are near each other tend to be connected [12] and also might be expected to have higher correlations in expression patterns ( because nearby regions will tend to be of the same embryonic origin , for example ) . This will tend to obscure the degree to which expression is specifically correlated with connectivity ( and in turn obscure the degree to which expression is specifically correlated with location ) . We assessed the overall degree of spatial autocorrelation by performing the Mantel test as above , but comparing expression or connectivity to a matrix representing physical distance or , alternatively , nomenclature distance ( relationships in the nested hierarchy of brain regions ) . As expected , the Mantel test results are all significant ( Figure 2 ) . The connection data ( r = 0 . 32; p-value<0 . 001 , Mantel test ) appears to be less spatially autocorrelated than expression ( r = 0 . 49; p-value<0 . 001 , Mantel test ) . We visualized the spatial correlation structure with Mantel correlograms ( Figure 3 ) . The Mantel correlogram displays the correlation between a data matrix and a matrix formed by grouping region pairs into distance classes . The correlogram will not be flat if it is possible to predict the distance class of a pair based on connectivity or expression correlations alone . As shown in Figure 3 , there is indeed an effect of distance on the correlation between connectivity and expression . We therefore attempted to correct our analysis for the effect of spatial autocorrelation , using regression . We calculated regressions between the distance and expression or connectivity correlations for all region pairs . The residuals of these regressions provide proximity-controlled correlations . As shown in Figure 3 , an improvement in the correction is obtained when using log-transformed distances . Using the log-transformed distance matrix from above , we can control for spatial autocorrelations by applying the partial Mantel test [30] , [31] . The partial Mantel test applies the same regression mentioned above to both the connectivity and expression similarity matrices . Then a standard Mantel test is calculated between the two spatially-corrected residual matrices . We found that after correction , the partial Mantel test between connectivity and expression remains significant , indicating the relationship is not entirely due to neighbourhood effects . However as expected the correlations are lower . Using the spatial correction , the correlation between incoming connectivity and expression is 0 . 109 ( p-value = 0 . 008 , Mantel test ) , for outgoing it is 0 . 126 ( p-value = 0 . 001 , Mantel test; summarized in Figure S2 ) . As a further confirmation for the effectiveness of the correction based on spatial distance , we found that the correlation between nomenclature distance and expression or connectivity correlation drops substantially , though the correlations are still significant ( Mantel correlation −0 . 089 for expression , p-value = 0 . 006; 0 . 11 for connectivity , p-value<0 . 001 ) . This incomplete correction is perhaps not surprising as the nomenclature hierarchy reflects connectivity as well as spatial location . The above tests use expression information for all expressed genes in the Allen Brain Atlas , but we expect that many genes will not contribute any information on connectivity . To find the most informative genes , we applied a greedy algorithm that identifies subsets of the data which maximize the correlation between connectivity and expression patterns ( see Materials and Methods ) . Figure 4 displays the change in the Mantel correlations as genes are iteratively removed . As shown in Table 1 , this yields much smaller sets of genes ( 357 and 433 for outgoing and incoming , respectively ) and much higher Mantel correlations ( 0 . 56 and 0 . 65 for outgoing and incoming connectivity respectively ) . Figure S3 provides a visualization of these results by intersecting region pairings with high expression and connectivity correlations . As a control , we performed the same procedure on multiple shufflings of the expression data , yielding a maximum correlation across ten runs of r = 0 . 42 and r = 0 . 51 for outgoing and incoming respectively . We also carried out the same procedure for the spatial correlations instead of connectivity , yielding a “spatial proximity” list of 401 genes and a Mantel correlation of 0 . 934 . Eighty-five image series ( 89 genes ) were found to overlap between the lists for incoming and outgoing connectivity , which is not surprising because there is a fair amount of reciprocal connectivity . Twenty-one image series ( 31 genes ) overlap across the spatial proximity list and one or both of the connectivity gene sets , suggesting that for the most part , different genes provide information about connectivity and proximity . The top twenty image series for the rankings are provided in Table 2 ( full results are available as Tables S2 , S3 , and S4 ) . If we consider just the top 20 genes , the Mantel correlations are 0 . 516 ( incoming ) , 0 . 460 ( outgoing ) and 0 . 590 ( proximity ) . As an additional control , we found that the correlations obtained for the optimized gene sets are robust to the completeness of the connectivity network ( tested by , for example , randomly removing brain regions and recomputing the Mantel correlations ) . Thus , while the connectivity map of the rodent brain is incomplete , the correlations with expression appear robust . We next examined the expression patterns of the optimized gene lists in more detail . It was of interest to determine , for example , if all the genes had similar expression patterns , which would suggest a single overwhelming signal in the data . A hierarchical clustering and visualization of the expression patterns of the optimized gene sets suggested that the patterns are in fact diverse ( Figures S4 and S5 ) . This is supported by a comparison of the distributions of gene-gene correlations within the optimized outgoing list , which are on average slightly lower than the full data set ( 0 . 10±0 . 21 for top outgoing genes; 0 . 15±0 . 21 for all genes; p-value<2 . 2×10−16 , t-test , Figure S6 ) . This suggests that many different gene expression patterns are contributing to the overall correlation between connectivity and gene expression . Figure 5 shows the expression patterns for two genes that rank high in the “outgoing” gene list , overlaid on schematics of the connectivity data . In Figure 5A , we show the pattern for Pcp2 ( Purkinje cell protein 2; Figure 5A ) . Although Pcp2's function is unknown , it is almost exclusively expressed in the projection neurons of the cerebellar cortex ( Purkinje cells ) . We did not expect this specific expression pattern to carry information about connectivity because no other regions express Pcp2 . However , the connections of the cerebellar cortex are also unique and specific: of the 112 outgoing regions , 69 place the cerebellar cortex in the bottom tenth percentile of similar regions based on proximity controlled connectivity . As a result , the optimization procedure finds that Pcp2's expression pattern marks the cerebellar cortex's unique connectivity profile . Figure 5B shows the expression pattern of Pgrmc1 ( Progesterone membrane component 1 ) , a gene that may play roles in axon guidance [32] , [33] . In contrast to Pcp2 , which is expressed in only one brain region , expression of Pgrmc1 in two regions is correlated with a connection between them ( Figure S7 ) . Thus , clusters of highly connected regions tend to show higher levels of Pgrmc1 expression ( Figure 5B ) . While the strong relationships shown in Figure 5 are not representative of the data set as whole , they serve to illustrate how expression patterns can contain information on connectivity . One concern about using high-throughput in situ hybridization data might be the potential for artifacts . While all of the image series we used had passed the Allen Brain Atlas project's ( ABA ) own quality control criteria , we did note occasional spatial artifacts such as dust or bubbles , though there was no indication such problems were more common in the genes we ranked highly . In addition , while there is good evidence that the ABA data are reliable , with a high quantitative and qualitative agreement with other data [34] , [35] , there are genes ( ∼6% in ABA ) for which ABA has disparities [35] and a few of those genes show up in our results ( at approximately the expected proportion; see Dataset S1 ) . To help address these concerns , we extracted a higher-confidence subset of results by considering genes measured more than once in the Allen Brain Atlas . These “duplicate” image series vary primarily by the RNA probe sequence used and the plane of section ( sagittal vs . coronal ) , and it seems unlikely that results which are concordant across image series would be due to expression analysis artifacts . Seventeen genes in our top outgoing connectivity list have two concordant image series . In the case of incoming connectivity , 16 of the genes on our list are represented by at least two image series ( Rprm has three , and Calb2 has four of its 20 image series across the atlas ) . We refer to these as the “high-confidence” lists . The next stage of our analysis was to consider in greater detail the types of genes which are correlated with connectivity . We accomplished this through a combination of Gene Ontology ( GO ) annotation enrichment analysis and manual review of the literature relating to the genes , particularly those on our high-confidence lists . We specifically hypothesized that genes that play roles in neural development might be found , as suggested by previous work on Caenorhabditis elegans [22] , [23] . In agreement with this hypothesis , our Gene Ontology analysis of the “outgoing” list revealed significant enrichment in categories related to neuronal development ( Table 3; note that many of the top groups have overlapping gene members . No GO terms were significant for the “incoming” or “proximity” lists . Full GO analysis results are in Table S5 ) . A manual examination of the connectivity top gene lists ( Tables S2 and S3 ) makes it clear that this is due to the presence of many different genes that play a variety of roles in neuronal development , but axon guidance was a prominent theme . Our lists contain a total of 14 members of three major axon guidance families ( Semaphorin , Ephrin , and Slit families ) [36] ( Table 4 ) . These gene families express cell-surface or secreted proteins that function to provide guidance signals to growing axons . This was most striking for the Semaphorin family , with ligands , receptors and co-receptors appearing in the incoming or outgoing top gene lists ( Table 4 ) . Six of the 17 genes from the high-confidence “outgoing” list function in neuronal development and axon guidance . Two of these six , Gpc3 and Hs6st2 encode a heparan sulfate proteoglycan and a heparan sulfate sulfotransferase respectively . Two additional heparan sulfotransferases , Hs3st1 and Hs6st1 appear with one image series on outgoing top gene list . Heparan sulfate proteoglycans are membrane proteins that have been linked to neurogenesis , axon guidance and synaptogenesis [37] . Hs6st2 has been specifically linked to retinal axon targeting in Xenopus [38] . Another gene on the high-confidence list is the L1 cell adhesion molecule ( L1cam ) , a recognition molecule involved in neuron migration and differentiation [39] . Vesicle-associated membrane-protein ( Vamp2 ) is another gene connected to connectivity through two image series; in addition Vamp1 occurs once in the outgoing list . Recently Vamp2 has been linked to attractive axon guidance but not repulsion in chick growth cones [40] . Neurturin is another high-ranking gene with two image sets linked to outgoing and one linked to incoming . Neurturin is well known to promote neuronal survival and induce neurite outgrowth [41] . Lastly , Serinc5 is enriched in white matter and Inuzuka et al . [42] suggest its major role is to provide serine molecules for myelin sheath formation . In the case of genes correlated with patterns of incoming connectivity , 4 of the 16 of the genes on our high confidence list have previously suggested roles in brain connectivity . Neurensin-1 shows up with two image series and is known to be involved in neurite extension [43] . Recently , Stat5a has been labelled a key effector molecule in the mammalian CNS , affecting axon guidance in the spinal cord and cortex [44] . Thirdly , Uchl1 is mutated in the GAD mouse strain that presents axon targeting and genesis defects [45] . Finally , ciliary neurotrophic factor receptor ( Cntfr ) appears twice on the top ranked list and is known to promote neuron survival and plays important roles in nervous system regeneration and development [46] , [47] . Another trend we notice from the GO results is that groups of genes with negative regulatory roles are much more prominent than the corresponding “positive” groups ( e . g . , “negative regulation of neurogenesis” ) though these groups are not statistically significant after multiple test correction . The high ranking of these terms ( which share members ) is due to 11 genes: Hdac5 , Notch3 , Nrp1 , Cd24a , Cit , Apc , Nr2e1 , Ptk2 , Gpc3 , and Runx2 . The “negative” aspect of the function of these genes varies but all have roles in neuronal development and/or plasticity . For example Nrp1 is a coreceptor for semaphorins and triggers inhibition of axonal growth [48] , while Hdac5 is a histone deacetylase whose activity is associated with repressed chromatin conformations that are altered after addictive stimuli [49] . We also conducted a search among our high-confidence list for genes whose homologs are implicated in human disorders of the nervous system . We found evidence for such a role for five of the 30 genes . Prominent among the five is L1Cam , defects in which cause several brain disorders including partial agenesis of the corpus callosum [50] . Two genes in the high confidence lists have been linked to heritable forms of Parkinson's disease ( alpha-synuclein ( Snca ) [51] and Uchl1 [52] ) . Finally , two genes have been linked to autistic spectrum disorder ( ASD ) . The human homolog of Cadps2 has been linked to autism and lies in the 7q autism susceptibility locus ( AUTS1 ) [53] , [54] . Another , Btg3 is in a genetic locus linked to autistic children characterized by a history of developmental regression [55] . By examining our expanded list of genes , we found several more of our connectivity linked genes are in AUTS1 and have been studied in the context of autism: Reln [56] , Mest [57] , Ptprz1 [58] , Dpp6 [59] and En2 [60] . To further explore the potential connection between our results and autism , we downloaded all autism candidate genes from the AutDB database [61] . Of those genes , 163 were available in our dataset , and 17 appear in at least one of the connectivity linked lists ( 14 for incoming connectivity and Nrp2 , Cadps2 , Ntrk1 , and Apc appear in both incoming and outgoing lists ) . The probability of this occurring by chance is 0 . 00029 ( hypergeometric test; considering the incoming list alone the p-value is 5 . 43×10−5 ) . In contrast , the proximity-ranked list contains only 5 genes in the AutDB set ( p-value = 0 . 32 ) .
Our analysis revealed a number of interesting relationships between gene expression and patterns of connectivity in the adult mammalian brain . Our key finding is that genes whose expression patterns carry information on connectivity are enriched for genes involved in neural development , and axon guidance in particular . While our results are based on analysis of the brains of rodents , it is of potential importance that many of the genes we identify have human homologs implicated in disorders of the nervous system including ASD . Because there is an increasing interest in the idea that ASD and other disorders are in part due to abnormalities in connectivity [4] , [62] , and given the heritability of many such disorders , the relationship between gene expression and connectivity is pertinent . The enrichment of homologs of autism candidate genes in our results suggests that these patterns could be relevant to the understanding of behavior in autism and potentially avenues for treatment . To our knowledge ours is the first study comparing gene expression and connectivity in mammals at a global level . Interestingly , a previous focused examination of the correlation between expression and connectivity for two brain regions identified some of the same genes we did . Dong et al . [21] examined correlations between genes that are differentially expressed between the dorsal and ventral hippocampus ( which we were not able to treat as separate regions in our analysis ) . For nine of their genes , they observed matching expression patterns in a connected brain region , the lateral septal nucleus . Three of these seven genes appear on our connectivity correlation lists ( Gpc3 , Man1a , Wfs1 ) ; this is unlikely to occur by chance ( p-value = 0 . 0045 , hypergeometric test ) . In contrast , none of the seven appear on the proximity gene list . We stress that because what we observe are correlations , it is difficult to ascribe a definite mechanism or meaning to the patterns . In addition , in absolute terms the Mantel test correlations may seem low when we considered all genes . However , we do obtain a correlation of 0 . 65 between gene expression patterns and proximity-controlled incoming connectivity after gene selection . We also point out that at the neuron to neuron level in Caenorhabditis elegans , Kaufman et al . [22] reported statistically significant correlations of 0 . 075 and 0 . 176 between expression and incoming and outgoing connectivity , respectively . Thus the patterns we observe in the adult mammalian brain are at least as strong as those observed in previous studies . An obvious question is whether the signals we observe are strong enough to predict patterns of connectivity . Unfortunately , while the signals we observe are statistically significant , they are not strong enough to allow prediction of connections based on expression patterns . Kaufman et al . [22] attempted this with their data and achieved very low accuracy . Using similar data , Baruch et al . [24] attained statistically significant results in predicting the direction of connectivity between neurons known to be connected or which share a common synaptic partner . Using advanced imaging techniques on human subjects , Honey et al . [63] attempted to predict diffusion tensor imaging ( DTI ) based cortical connectivity from fMRI functional connectivity . By setting thresholds on functional connectivity , they achieved an AUC value of 0 . 79 that could predict only ∼6% of inferred DTI connections [63] . Despite these limitations , our results suggest some underlying models that in turn provide some testable hypotheses . Many of the genes we find to be associated with connectivity patterns in the adult are thought to be primarily active in the developing brain , when large-scale connectivity is determined . The reasons for expression of these genes in the adult brain is not fully understood , though there is evidence in some cases that they continue to play roles in the maintenance or tuning of neuronal connectivity at finer scales [16] , [64] . There is even less known about why the genes show regionally restricted patterns in the adult brain . Our results are the first to link the expression signatures of some of these genes to macroscopic connectivity . Our results have at least two possible biological interpretations . One is that the expression patterns in adulthood are a “residue” of the developmental pattern that reflects processes occurring when connectivity is laid down , but that the adult expression pattern is not causally related to connectivity at the scale we studied . An alternative is that the expression patterns in adulthood are functionally relevant with respect to connectivity , perhaps in modulating activity in certain pathways . The patterns we identified could be used to design experiments to distinguish between these alternatives . While we have provided evidence for a relationship between connectivity and gene expression in the mammalian brain , our analysis is surely hindered by the incompleteness of connectivity and expression information . There are many brain regions for which we had expression data but no connectivity . While some of these regions might never have been studied , there are many reports in the literature that are not included in the current connectivity databases . Advances in the generation of connectivity information from new experiments or from more complete use of existing reports will be essential . The availability of additional expression data would also improve our ability to interpret the patterns we observe . In particular , having detailed information on gene expression patterns during development , and their relationships to the developing projection patterns in the brain , could permit stronger inference of causal relationships . A final limitation is that the structural connections we use cannot be easily linked to specific states or functions of the brain . Because of this we could only interpret our results in the context of gene function information . It would be of interest to employ functional connectivity data to link gene expression to more dynamic and task specific states of the brain , especially in the context of genetic variation .
For neuroanatomical connectivity knowledge , we used the Brain Architecture Management system ( BAMS ) . BAMS contains extensive information about neural circuitry curated from neuroanatomical atlases and tract tracing experiments [25] , [65] . The version of the BAMS database we use contains 7 , 308 structural connections between 961 rat brain regions and is accessible via bulk download ( http://brancusi . usc . edu/bkms/xml/swanson-98 . xml ) . Instead of parsing the original XML we used a converted semantic web version created by John Barkley ( http://sw . neurocommons . org/2007/kb-sources/bams-from-swanson-98-4-23-07 . owl ) . The BAMS system stores information on projection strength , number of reports , report citations and absence of connections but it is not available in the database version we obtained . However , directions of the neuroanatomical connections are known , allowing splitting of our analysis between incoming and outgoing connection profiles . The BAMS curators comprehensively studied the bed nuclei of the stria terminalis ( BNST ) and indicate that its connection matrix is considered complete [65] . We were concerned that this unusually well-studied region would bias our results , as it has more known connections than the other regions ( we considered regions that lack a documented connection to be unconnected ) . For example , it has over seven times the average number of outgoing connections . To reduce this bias in the dataset , we removed connection information for the BNST and its subparts . We do not suspect the quality of these connections but wished to prevent one well-characterized region from being overrepresented . We believe the complete connectivity matrix of the BNST will be very valuable for future focused analysis . We considered using gene expression profiles from SAGE and microarray experiments , but spatial resolution was too low . Therefore we used high-resolution colourmetric in situ hybridization ( ISH ) measurements produced by the ABA [17] . The complete expression matrix from the ABA ( kindly provided by the Allen Institute for Brain Research ) consists of 5 , 380 , 137 entries formed by 25 , 991 ISH image series and 207 brain regions . In many cases a gene was assayed more than once , using a different probe or plane of sectioning . The ABA provides values for expression “energy” , “level” and “density” across a region . Because level and density had a large fraction of data missing ( ∼40% ) we choose to use expression energy ( 3% missing ) . Expression energy is defined as the sum of expressing pixel intensities normalized by the number of pixels in a region . The natural logarithm of expression energy values formed our gene expression matrix . Genes that do not have detectable expression in the ABA were removed . The list of non-expressing genes list was provided in Lein et al . as supplementary data [17] . After removing the non-expressing genes the final gene expression profiles contain 22 , 771 image series representing 17 , 530 genes . The names of brain regions are formalized in hierarchies both in BAMS [26] , [66] and the ABA data [67] , but the schemes are not identical . In addition , the BAMS dataset contains information at a finer neuroanatomical resolution than ABA . To maximize the use of connectivity information , we created connection profiles of coarser scale by using an up-propagation procedure . Up-propagation maps the brain region to its parent region until the desired level in the neuroanatomical hierarchy is reached . This procedure was applied to all connection pairs in BAMS . For example , a connection between region A and region B will be expanded to the set of all possible connections between the neuroanatomical parents of both region A and region B . To prevent enrichment of up-propagated connections we kept regions that had zero connections to the ABA mapped regions . Although the two datasets have common objects - brain regions , the organisms differ . The rat brain with a wealth of neuroanatomical information is bigger and for some regions like the cerebellum , more complex . In contrast , genetics and molecular research is more commonly performed on the smaller mouse brain . For this work we considered neuroanatomical differences between the mouse and rat to be minor at the level of granularity we used [68]; for example , the Paxinos mouse atlas was guided by several rat brain atlases [69] , and brain regions names largely coincide between the two . These common names allowed quick lexical mapping for most of the regions . To join the two data types we mapped nomenclatures manually . We used primarily a region's name , then secondarily its parent region and spatial borders to pair brain regions . The mappings for the Allen Brain regions are provided in Table S6 . The neuroanatomical atlases from ABA [67] and BAMS [70] provide information on which brain regions are neuroanatomical children or parts of others . These relations create correlations in the gene expression profiles and the connectivity data ( due to up-propagation ) . To negate this effect we used only 149 of 207 Allen brain regions for the primary region list . These remaining regions have no neuroanatomical subparts in the ABA dataset . The Allen Atlas provides a differing grouping of regions than the BAMS hierarchy . The superior colliculus is one example . The ABA divides its regions into motor and sensory areas , while the BAMS atlas groups the regions into optic , gray and white layers . Differences were resolved by creating “virtual regions” in the BAMS atlas space that contained the corresponding subregions of the Allen Atlas . The connectivity profiles of the mapped regions were joined using a logical OR operation to provide the virtual region's BAMS connections . For example the superior colliculus sensory related virtual region has all of the BAMS connections of the zonal , optic and superficial gray layers . In addition to the superior colliculus , virtual regions were created for the pallidum medial region and nucleus ambiguus . After mapping of brain regions , the ABA data is an x ( number of regions in the ABA ) by y ( number of genes ) matrix , and the BAMS connectivity data is a square w ( number of regions in BAMS ) by w ( region ) matrix ( Figure 1 ) . The two matrices are not directly comparable because the number of regions in BAMS is greater than those in ABA ( w>x ) . Rather than discarding all information from regions which lack expression information , we use the x by w submatrix of the BAMS data . Thus each of the x regions has a y-dimensional expression vector and a w-dimensional connectivity vector . This maximizes the use of connection information , but we note that the connectivity profiles include information from regions for which we lack expression information . Correlations between gene expression values and connection degree were computed using Spearman's rank correlation coefficient ( ρ ) . Connection degree for each brain region is the sum of its propagated incoming and outgoing connections . Significance of the correlation was corrected for multiple testing using the Bonferroni method . We generate a ranked list of genes so that a gene's rank is proportional to its contribution to the connectivity correlation score . To achieve this we reduce the number of genes in the expression profiles while maximizing the Mantel test correlation score . Since it is not feasible to compute all possible subsets of the image sets , we approximate an optimal candidate list of genes . Again , we take guidance from Kaufman et al . [22] and use a greedy backward elimination algorithm with the Mantel test . Each iteration of the algorithm involves ranking each gene by its contribution to the global correlation , removing the least informative gene , and repeating the test on the remainder . For the connectivity gene rankings we optimized a partial Mantel correlation that modelled proximity in the connection matrix but not the expression correlations ( due to computational constraints ) . For functional enrichment analysis we employed the ErmineJ software to explore the roles of the candidate genes [72] . Overrepresentation analysis was used on the set of genes removed after correlation reached a maximum . To increase resolution of the genes , NCBI identifiers were used instead of gene symbols . Gene Ontology ( GO ) groups included in the analysis required 5 to 200 measured gene members and were limited to the biological process division . Benjamini-Hochberg false discovery rate was used to control for testing multiple GO groups [73] . GO groups were sorted by corrected p-value to determine rankings . For creation of Figures S4 and S5 we employed average linkage hierarchical clustering on both the image series and brain regions . The clustered data was converted to a heatmap using matrix2png with rows normalized to zero mean and variance of 1 [74] . Values were then constrained to the range of −3 to 3 . | We tested the idea that the “wiring diagram” of the adult brain has a relationship with where genes are expressed . We were inspired by similar work carried out by groups examining the nematode worm Caenorhabditis elegans . By using large-scale databases of brain connectivity and gene expression in rodents , we found that many genes involved in the development of the brain show correlations with anatomical connectivity patterns . Some of the genes we found have been implicated in disorders such as autism , which is suspected to affect brain wiring . While the biological causes of the patterns we found are not yet known , we believe they provide new insight into the patterns of gene expression in the brain and will spur further study of this problem . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience",
"genetics",
"and",
"genomics/bioinformatics",
"genetics",
"and",
"genomics/gene",
"expression"
] | 2011 | Relationships between Gene Expression and Brain Wiring in the Adult Rodent Brain |
Gene targeting of mouse Sushi-ichi-related retrotransposon homologue 11/Zinc finger CCHC domain-containing 16 ( Sirh11/Zcchc16 ) causes abnormal behaviors related to cognition , including attention , impulsivity and working memory . Sirh11/Zcchc16 encodes a CCHC type of zinc-finger protein that exhibits high homology to an LTR retrotransposon Gag protein . Upon microdialysis analysis of the prefrontal cortex region , the recovery rate of noradrenaline ( NA ) was reduced compared with dopamine ( DA ) after perfusion of high potassium-containing artificial cerebrospinal fluid in knockout ( KO ) mice . These data indicate that Sirh11/Zcchc16 is involved in cognitive function in the brain , possibly via the noradrenergic system , in the contemporary mouse developmental systems . Interestingly , it is highly conserved in three out of the four major groups of the eutherians , euarchontoglires , laurasiatheria and afrotheria , but is heavily mutated in xenarthran species such as the sloth and armadillo , suggesting that it has contributed to brain evolution in the three major eutherian lineages , including humans and mice . Sirh11/Zcchc16 is the first SIRH gene to be involved in brain function , instead of just the placenta , as seen in the case of Peg10 , Peg11/Rtl1 and Sirh7/Ldoc1 .
Mammals , including human beings , have evolved a unique viviparous reproductive system using a placenta and a highly developed central nervous system . How did these unique characteristics emerge in mammalian evolution ? Retrotransposons occupy approximately 40% of the mammalian genome . They recently have attracted attention as one of the driving forces of genomic evolution , providing novel endogenous genes [1–11] as well as rewiring the genetic network in the form of novel cis-elements , such as promoters , enhancers , insulators and transcription factor binding sites [12–16] . In a series of KO mouse experiments we have demonstrated that at least three LTR retrotransposon-derived genes are essential for mammalian development and reproduction via multiple placental functions; Peg10 is involved in the formation [8] and Peg11/Rtl1 in the maintenance of the placenta [9] , while Sirh7/Ldoc1 is involved in endocrine regulation via the differentiation/maturation of a variety of placental cells [11] , suggesting that they all have profoundly contributed to the evolution of viviparity during mammalian evolution [8–11 , 17 , 18] . Two major families of genes derived from Ty3/Gypsy LTR retrotransposons have been identified: one is the SIRH family , comprising the 11 genes mentioned above , while the other is the paraneoplastic MA antigen ( PNMA ) family derived from a gypsy_12DR-related retrotransposon comprised of at least 19 and 15 genes in humans and mice , respectively [7 , 19–22] . It should be noted that Peg10 is the only gene commonly conserved in both the eutherians and marsupials [23] , while all the others exist as eutherian- or marsupial-specific genes [11 , 22 , 24 , 25] . Among the SIRH family , Sirh11/Zcchc16 ( also called Mart4 ) is unique because it does not exhibit any placental expression during development , but rather , is expressed in the brain , testis , ovary and kidney . In this work , we set out to examine whether Sirh11/Zcchc16 plays a role in organs other than the placenta , then generated and analyzed Sirh11/Zcchc16 KO mice . Interestingly , the Sirh11/Zcchc16 KO mice exhibited a variety of behavioral abnormalities related to cognition , indicating Sirh11/Zcchc16 is involved in brain function . We also found abnormal regulation of the NA level in the prefrontal cortex of KO mice . As the noradrenergic system in the LC in the brainstem sends projections to virtually all brain structures , including the prefrontal cortex of the cerebrum , and has been proposed to be involved in cognitive function , such as impulsivity , attention , working memory and their associated behaviors in mammals [26–29] , we investigated the potential role of the Sirh11/Zcchc16 protein in the noradrenergic system , suggesting the relationship to human mental disorders and the impact on brain evolution in eutherian mammals .
Mouse Sirh11/Zcchc16 encodes a Gag-like protein comprising 304 amino acids with a typical CCHC RNA-binding motif at the C-terminus ( Fig 1A ) . It exhibits 37 . 5% homology with the entire sushi-ichi retrotransposon Gag , consisting of 371 amino acids , except for the N-terminus . Sirh11/Zcchc16 is located on the X chromosome between Trpc5 and Lhfpl1 . Its location is conserved in all of the eutherian lineages , euarchontoglires , laurasiatheria , afrotheria and xenarthra ( Fig 1B and 1C ) . However , it became a pseudogene by frameshift and nonsense mutation in xenarthran species , such as the armadillo and sloth ( S1 Fig ) . In the case of the armadillo ( Dasypus novemcinctus ) , contig including pseudoSIRH11/ZCCHC16 is short and thus does not reach LHFPL1 or TRPC5 . However , the presence of several evolutionarily conserved sequences ( ECSs ) in its surrounding 20 kb sequence confirms that it is orthologous to SIRH11/ZCCHC16 ( Fig 1B , lower column ) . The absence of SIRH11/ZCCHC16 from marsupials , monotremes and birds was also confirmed , because there is no orthologous gene between TRPC5 and LHFPL1 in the opossum and Tasmanian devil , or between TRPC5 and AMOT in the chicken and platypus , respectively ( Fig 1B and 1C ) . This indicates that the insertion of SIRH11/ZCCHC16 occurred in a common eutherian ancestor after the spilt of the eutherians and marsupials 160 million years ago ( Ma ) , before the diversification of the three major eutherian lineages , boreoeutheria ( including euarchontoglires and laurasiatheria ) , afrotheria and xenarthra , 120 Ma [30] . Importantly , the dN/dS ratio in the pairwise comparisons of SIRH11/ZCCHC16 orthologs between mouse and seven eutherian species other than xenarthran species is approximately 0 . 35~0 . 45 ( < 1 ) ( Table 1 ) [31] , suggesting that SIRH11/ZCCHC16 has been subjected to purifying selection after its domestication ( exaptation ) in the common eutherian ancestor . Furthermore , to examine whether the functional constraint is relaxed in the xenarthran lineage , we also performed an analysis using the Phylogenetic Analysis by Maximum Likelihood ( PAML ) program for comparing two models [31] . One is that all the twelve species including the four xenarthran species have the same ω value ( dN/dS ratio ) ( model 1 ) . The other is based on the assumption that the ω value in the xenarthran branch , including the two armadillo and two sloth species , is different from that of all the other eight eutherian species ( model 2 ) . The analysis demonstrated that model 2 is statistically significant ( p = 8 . 80E-04 ) : ω2 = 1 . 11 for the xenarthran branch , with ω1 = 0 . 766 for all of the others ( Fig 1D ) , that is , relaxed or neutral evolution is ongoing in xenarthra . All of these results indicate that SIRH11/ZCCHC16 is a protein-coding gene in the eutherians except in xenarthra . Sirh11/Zcchc16 is basically comprised of 7 exons , with its ORF in the last exon ( Fig 2A ) . According to the NCBI database , there are at least three variants with different first exon sequences , presumably dependent on the tissues and organs where it resides . A low level of Sirh11/Zcchc16 expression was observed in the brain , liver and heart on embryonic day 14 . 5 ( d14 . 5 ) , with a moderate level of expression in the brain , kidney , testis and ovary in adults ( 8 weeks ( 8 w ) ) ( Fig 2B ) . We generated Sirh11/Zcchc16 KO mice using TT2 ES cells by means of a complete deletion of its protein coding sequence ( S2 Fig ) . After removing the neomycin cassette , Sirh11/Zcchc16 KO mice were backcrossed to B6 more than 10 generations . We confirmed the lack of the ORF region by RT-PCR using a primer set ( F5R5 ) . However , it should be noted that the RT-PCR experiment using primer sets amplifying its 3’-UTR region ( such as F9R9 ) exhibited an approximately 1 . 5 fold increment in these organs in the KO mice ( although not significant in the whole brain or kidney ) , suggesting the existence of a feedback mechanism regulating Sirh11/Zcchc16 at the protein level ( Fig 2A and 2C ) . Sirh11/Zcchc16 KO mice did not exhibit lethality or growth retardation in the pre- and postnatal periods in either female ( homo ) or male ( null ) KO mice ( Tables 2 and 3 ) . Despite the relatively higher expression of Sirh11/Zcchc16 in the testis and oocyte , both the male and female KO mice were fertile , even in the case of mating between female homo KOs and null male mice , indicating that Sirh11/Zcchc16 has no apparent role in sperm and/or egg production ( Table 2 ) . No abnormalities were detected in the urine of the Sirh11/Zcchc16 male KO mice ( 10 w ) , such as pH or the amount of glucose , total protein , urobilinogen , ketone body , bilirubin and occult blood , suggesting that kidney function is also normal in these KO mice ( S3 Fig ) . Although no evident structural abnormalities were found in the KO mouse brain ( S4 Fig ) , it was noticed that abnormal behaviors were exhibited . For example , they displayed agitated movement in their cages when staff personnel entered the breeding room , and sometimes jumped out when their cages were exchanged . Therefore , comprehensive behavior tests were carried out using 8–10 w males . Sirh11/Zcchc16 KO mice exhibited no abnormality in the open field test ( 9 and 10 w ) , but in the Light/Dark transition test ( 8 w ) the latency before entering into the light chamber was significantly decreased ( unpaired two sample t-test , t ( 12 ) = 2 . 52 , p = 0 . 0269 ) , while the number of transitions was significantly increased ( unpaired two sample t-test , t ( 12 ) = -2 . 58 , p = 0 . 0242 ) compared to the wild type ( WT ) , suggesting a reduced attention and/or enhanced impulsivity ( Fig 3A ) . They also exhibited significantly higher activity during the dark period of the home-cage activity test ( 10 w ) , especially just after the “light to dark ( Zeitgeiber time ( ZT ) 12 and 13 ) ” ( effect of genotype , F ( 1 , 60 ) = 7 . 86 , p = 0 . 00681 and effect of genotype , F ( 1 , 60 ) = 5 . 35 , p = 0 . 0241 , respectively ) as well as just before the “dark to light ( ZT23 ) ” ( F ( 1 , 60 ) = 9 . 77 , p = 0 . 00274 ) transition periods . There was also lower activity at ZT19 ( F ( 1 , 60 ) = 7 . 56 ) , p = 0 . 00788 ) , while there was no significant difference in the light periods ( ZT0-ZT11 ) , suggesting hyperactivity , especially when the light conditions are changed or about to be changed ( Fig 3B ) . Next , the Y-maze test was conducted to assess spatial working memory by recording spontaneously alternating search behavior during a 5-min session in a Y-maze ( 16–18 w ) . It is considered that alternating search behavior reflects a primitive working memory capacity because it is based on the tendency of normal animals to enter the arm of the Y-maze which was least recently explored [32] . The Y-maze test also allows the simultaneous assessment of hyperactivity independently of spatial working memory . Hyperactivity may interfere with learning and memory , therefore , its assessment is crucial for the interpretation of memory test results [33] . Importantly , KO mice exhibited a lower level of alternation ( Mann-Whitney U test , p = 0 . 032 ) , although the total number of arm entry events was the same ( Mann-Whitney U test , p = 0 . 358 ) as the WT controls , suggesting that they have a poor working memory ( Fig 3C ) . Some of the KO mice jumped out of the Y-maze stage before the test started , while none of the control mice exhibited such behavior . Although we excluded this data from the results , this also appears to indicate that these KO mice tend to exhibit extreme behavior when transferred to a new environment . From these results , we reasoned that Sirh11/Zcchc16 KO mice have some abnormality in cognition , possibly related to monoamine function in the brain , that impacts impulsivity , attention and/or memory . We used null KO male mice for behavioral tests . Although Sirh11/Zcchc16 is also expressed in the kidney , testis and ovary , there was no evidently unusual phenotype in these organs and also no viability or growth effects observed , as mentioned . Therefore , we think that the abnormal behaviors observed in the KO mice mainly reflect an impairment of brain function . In addition , we used mice generated by in vitro fertilization ( IVF ) between KO males and hetero KO females for these behavioral tests , as described in Materials and Methods . The fertilized eggs were transferred to the pseudopregnant ICR females , after which the pups were born naturally and taken care of by the ICR mothers . Thus , harmful effects on the pups from the potentially abnormal KO mothers in terms of their nursing behavior were completely excluded . Therefore , we believe that these results reflect a difference in genotype . We carried out microdialysis analysis in the prefrontal cortex of the cerebrum to directly examine the monoamine levels in the KO brain [34–35] because it is well documented that prefrontal cortical NA as well as dopamine ( DA ) plays an important role in spatial working memory [36] . The levels of various monoamines were measured , mainly at 11~13 w of age , including DA , NA , adrenaline ( AD ) , 3-methoxytyramine ( 3-MT ) , 5-hydroxyindole acetic acid ( 5-HIAA ) , serotonin ( 5-HT ) , 3 , 4-dihydroxyphenylacetic acid ( DOPAC ) , homovanillic acid ( HVA ) , 3-methyl-4-hydroxy-phenylglycol ( MHPG ) and normetanephrine ( NM ) using isoproterenol ( ISO ) as the control . During overnight experiments for long-term recording , we found that these levels varied suddenly and unexpectedly , presumably because of abrupt changes in sound and vibration in the experiment room . The levels were also greatly dependent on the initial condition of the mice . As it was difficult to obtain reproducible data in the normal experimental setting , we adopted the perfusion method to overcome this difficulty . Perfusion by high-K solution intensively stimulates the release of neurotransmitters and hence enables an evaluation of the sum of neurotransmitter content at the synapse and in the neuron . High K solution was applied twice , with a 120-min interval . The release of neurotransmitter in response to the second perfusion of the high K solution is dependent upon the catabolic activity during a 120 min period . It was demonstrated that the recovery of the NA level was significantly delayed compared with the DA level ( Mann-Whitney U test , p = 0 . 0159 ) , while the levels of two other DA metabolites , DOPAC and 3-MT ( Mann-Whitney U test , p = 0 . 532 and 0 . 310 , respectively ) , were unchanged in the KO mice ( Fig 4A ) . DA is catabolized in three different pathways to NA , DOPAC and 3-MT by the enzymes dopamine beta monooxigenase ( Dbh ) , monoamine oxidase ( MAO ) and catechol-O-methyltransferase ( COMT ) , respectively [37] . Dbh exhibits a dominant expression pattern in the brainstem , approximately 32 fold greater than the cerebrum ( Fig 4B ) although its brainstem level was not affected in the Sirh11/Zcchc16 KO . Sirh11/Zcchc16 also exhibited a relatively higher expression in the mesencephalon , diencephalon and brainstem compared with other parts of the brain ( Fig 4C ) . Interestingly , both the Dbh and Sirh11/Zcchc16 expression levels fluctuated in the brainstem , but their expression patterns exhibited a negative correlation ( i . e . the test for non-correlation , Pearson correlation coefficient ( r ) = −0 . 591 , p = 0 . 033 , Fig 4D ) . Therefore , it is possible that they are reciprocally regulated by the same environmental signals . Although the precise biological role of Sirh11/Zcchc16 as well as the relationship to Dbh in the brain are presently unknown , all of these data suggest that behavioral defects of the Sirh11/Zcchc16 KO mice are somehow related to a dysregulation of the noradrenergic system in the brain . We used two cohorts of mice , one for the comprehensive behavioral tests , including biochemical analyses , and the other for expression analysis in a range from embryos to adult tissues , as well as microdialysis and the Y-maze test . We believe that the results of the monoamine analysis ( microdialysis ) accurately reflect the Y-maze test . They also are consistent with the Light/Dark transition and home-cage activity tests in the first cohort of mice .
It has been proposed that the LC-NA system plays an important role in cognitive function , presumably by regulating the balance between focused versus flexible responding , or selective versus scanning attention [25–28] . In the phasic mode , LC cells exhibit selective phasic activation for target stimuli , but only a moderate level of tonic discharge , leading to excellent performance on the specific task with few errors and focused , selective attention . In the tonic mode , the LC cells fail to respond phasically to any task stimuli , but rather , exhibit higher levels of ongoing tonic activity , leading to poor performance with many errors and a form of scanning , labile attention . In their overarching theory , Bouret and Sara proposed that phasic activation of the NA neurons of the LC takes place in time with the cognitive shifts that facilitate dynamic reorganization of target neural networks , permitting rapid behavioral adaptation to the demands of changing environmental imperatives [27] . In this work , Sirh11/Zcchc16 KO mice exhibited certain clearly evident behavioral abnormalities , such as increased activity during the light/dark transition test , higher daily activities in the period just after dark as well as just before light and a lower score in alternating search behavior in the Y-maze test , indicating that Sirh11/Zcchc16 plays a role in cognition . Together with certain other sudden and unexpected movement activity that was observed , it is likely that Sirh11/Zcchc16 KO mice tend to exhibit behaviors related to the classic tonic LC mode . In support of this idea , a significantly lower recovery rate of NA compared to DA in the prefrontal cortex was observed in a microdialysis analysis of the KO mice after perfusion with high potassium solution . Importantly , activation of the LC-NA system is also associated with an increased accuracy of the response to task-relevant stimuli [26] . Using microdialysis analysis , Rossetti and Carboni demonstrated that both prefrontal cortical DA and NA are involved in the modulation of working memory [36] . From an application of the T-maze test using rats that analyzes such function along with memory and spatial learning via an application of various stimuli , they demonstrated that the prefrontal cortical DA and NA dialysate levels are both phasically increased when rats perform correctly in a delayed alternation task in a T-maze . Together with the findings from other experiments , they ultimately concluded that DA is primarily associated with reward expectancy , whereas NA is involved in the active maintenance of goal-related information as well as the rules for realizing the goal [36] . This is in good accord with the results of our Y-maze test , indicating that Sirh11/Zcchc16 KO mice with a lower NA recovery rate in the neurons of the prefrontal cortex have impaired spatial working memory . It has also been proposed that the phasic as opposed to tonic LC activity participates in certain critically important normal behavioral functions as well as severe mental problems , including attention-deficit/hyperactivity disorder ( ADHD ) and a variety of emotional and affective disorders [25–28] . Therefore , the behavioral abnormalities observed suggest a possible role for SIRH11/ZCCHC16 in mental disorders . Although a relatively higher expression level was observed in the brainstem , diencephalon and mesencephalon in the RT-PCR experiment , we have no direct evidence at present as to precisely where Sirh11/Zcchc16 is expressed in the brain or whether the putative Sirh11/Zcchc16 protein in the NA neurons from the LC . This is despite numerous attempts using in situ , Western blotting and immunostaining . However , the conservation of the amino acid sequence of the Sirh11/Zcchc16 in both boreoeutheria and afrotheria clearly demonstrates that it is subjected to purifying selection ( dN/dS <1 ) , providing indirect but nonetheless supportive evidence that Sirh11/Zcchc16 is a protein-coding gene ( Table 1 ) . This conclusion is consistent with the finding that the xenarthran pseudoSirh11/Zcchc16 bearing many mutations in its coding frame has been subjected to relaxed or neutral evolution ( dN/dS ~1 ) ( Fig 1D ) . The Sirh11/Zcchc16 protein possesses a conserved CCHC zinc finger domain with potential RNA-binding capability . Therefore , it is possible that it functions as a part of messenger ribonucleoprotein particles ( mRNPs ) in neurons because mRNAs synthesized in the soma are transported to neurites and/or synapses as mRNPs by binding to RNA-binding proteins and are translated there [38–40] . However , since the possibility that Sirh11/Zcchc16 functions as a non-coding RNA has not been completely excluded , these issues need to be further addressed in more detail in the future . Human SIRH11/ZCCHC16 is located on Xq23 , where several X-linked intellectual disability ( XLID ) genes have been mapped , such as PRPS1 ( –5 Mb from SIRH11/ZCCHC16 ) , ACSL4 ( –2 . 5 Mb ) , PAK3 ( –1 Mb ) , DCX ( –1 Mb ) , AGTR2 ( +4 Mb ) , LAMP2 ( +8 Mb ) and GRIA3 ( +11 Mb ) . As some of the genes responsible for XLID remain to be identified in this chromosomal region , SIRH11/ZCCHC16 may also be a good candidate for XLID [41 , 42] . Interestingly , Cho et al . recently reported that some XLID patients have mutations in the SIZN1/ZCCHC12 ( PNMA10 ) gene that locates approximately 6 Mb downstream of SIRH11/ZCCHC16 and indicated this gene to be a good candidate for XLID [43] . PNMA10/ZCCHC12 is another eutherian-specific Ty3/Gypsy LTR retrotransposon-derived gene that is known to be involved in transcriptional regulation [44] . Mouse Pnma10/Zcchc12 is expressed in the embryonic ventral forebrain in a cholinergic-neuron-specific manner and binds to SMAD family proteins . It also acts as a transcriptional co-activator for bone morphogenic protein ( BMP ) signaling [44] . Both Sirh11/Zcchc16 and Pnma10/Zcchc12 encode Gag-like proteins with the CCHC zinc-finger motif , so it will be of interest to determine whether these two retrotransposon-derived proteins have different activities which function in a noradrenergic- and a cholinergic-neuron-specific manner , respectively , in the extant eutherian mammals . Comparative genome analysis clearly demonstrated that SIRH11/ZCCHC16 is highly conserved in the three major groups of eutherian mammals , but not in xenarthra , strongly implying that the presence of Sirh11/Zcchc16 is beneficial in these three eutherian groups , including humans and mice . Our results suggest that SIRH11/ZCCHC16 contributed to the evolution of the brain by modulating the NA neuronal network in a complex manner . The noradrenergic system is also conserved in other vertebrates , such as fish , amphibians , reptiles and birds , and the general system characteristics are strikingly preserved across a wide range of phylogenetic groups [27] . Therefore , the biological function of the Sirh11/Zcchc16 protein is of great interest in terms of elucidating the evolution of the neuromodulatory system of the brain in the eutherian mammals . What was the function that was replaced in the noradrenergic system by domestication of SIRH11/ZCCHC16 in these three eutherian lineages , thus permitting rapid behavioral adaptation to changing environmental imperatives ? Do the xenarthran species , i . e . the armadillos and sloths , have diminished and/or distinct cognitive activity related to SIRH11/ZCCHC16 ? Although a variety of issues of this type require future investigation , this is the first demonstration that one of the SIRH genes plays a role in cognitive function in the brain , presumably via the noradrenergic system . We previously demonstrated that the three SIRH genes , such as Peg10/Sirh1 , Peg11/Rtl1/Sirh2 and Sirh7/Ldoc1 , play essential roles in eutherian development and reproduction , and proposed that these three SIRH genes have profoundly contributed to the evolution of viviparity in mammalian evolution as newly acquired genes [8–11] . In addition , the present investigation of Sirh11/Zcchc16 provides further insight into the impact of LTR retrotransposon-derived genes on the neuromodulatory system in the brain as key step in the evolution of the eutherian mammals .
The present animal experiments were performed in strict accordance with the guidelines of Tokai University and Tokyo Medical and Dental University ( TMDU ) , and were approved by the Animal Investigation Committees of Tokai University and TMDU . The open-field test , Light/dark transition test , home-cage activity test and urinalysis were performed in accordance with the guidelines issued by the RIKEN Bioscience Technology Center in their “Outline for Conducting Animal Experiments” ( August 1999 , revised October 2001 ) . Sushi-ichi gag ( AAC33525 . 1 ) and mouse Sirh11/Zcchc16 ( NP_001028967 . 2 ) protein sequences were obtained from NCBI . Amino acid identity and similarity were calculated using the EMBOSS Water program ( http://www . ebi . ac . uk/Tools/psa/emboss_water/ ) in the default mode . The orthologues of SIRH11/ZCCHC16 were identified by search of NCBI Gene ( http://www . ncbi . nlm . nih . gov/gene/ ) using ZCCHC16 as the query term . Genomic homology analysis was performed using the mVISTA LAGAN program ( http://genome . lbl . gov/vista/mvista/submit . shtml ) . We obtained TRPC5-AMOT genomic sequences from the NCBI database . The sequences used for analysis were the following: Chicken ( Gallus gallus ) : gi|358485508:c13102445-12913720; Platypus ( Ornithorhynchus anatinus ) : gi|149729612:c11732082-11450308; Opossum ( Monodelphis domestica ) : gi|126362945:c69157606-68740637; Mouse ( Mus musculus ) : gi|372099090:144381671–145505458; Human ( Homo sapiens ) : gi|568815575:111774314–112840908; Dog ( Canis lupus familiaris ) : gi|357579592:84841858–85807530; African savanna elephant ( Loxodonta Africana ) gi|343530165:c19381570-18463351; Armadillo ( Dasypus novemcinctus ) gi|476561443; Sloth ( Choloepus hoffmanni ) gi|692243298|gb|KN194663 . 1| . PseudoSIRH11/ZCCHC16 protein sequences were aligned to Florida manatee SIRH11/ZCCHC16 using Clustal Omega ( http://www . ebi . ac . uk/Tools/msa/clustalo/ ) . Genomic DNA was isolated from frozen muscle using the DNeasy Blood & Tissue Kit ( QIAGEN ) . For PCR , the primers were designed at the 5'- and 3'-UTR of pseudoSIRH11/ZCCHC16 using the consensus sequence between Dasypus novemcinctus and Choloepus hoffmanni . The PCR reaction was performed using ExTaqHS ( TaKaRa ) with the following conditions: 30 cycles of 98°C , 10 sec; 60°C , 30 sec; 72°C , 1 min . The following PCR primers were used: pseudoSIRH11-F1: 5'-CTTACTGCCTGCCCATTGGT-3' and pseudoSIRH11-R1: 5'-GGATTTTAAAAGTTGGTGCAGG-3' . PCR products were direct-sequenced using the above primers after Exo-SAP-IT ( USB ) treatment . DNA Data Bank of Japan ( DDBJ ) accession numbers: LOC064756 for Tolypeutes matacus SIRH11/ZCCHC16 and LOC064757 for Choloepus didactylus SIRH11/ZCCHC16 . A phylogenic tree was constructed with ClustalW2 ( Neighbor-joining method ) ( http://www . ebi . ac . uk/Tools/msa/clustalw2/ ) using protein coding and pseudo SIRH11/ZCCHC16 sequences obtained from twelve species . The codon alignment of cDNA was created with the PAL2NAL program ( http://www . bork . embl . de/pal2nal/ ) [45] . The non-synonymous/synonymous substitution rate ratio ( ω = dN/dS ) was estimated by using CodeML in PAML [31] . To generate Sirh11/Zcchc16 MT mice ( Accession No . CDB0557K: http://www . clst . riken . jp/arg/mutant%20mice%20list . html ) , we obtained three genomic fragments , the 5’-arm ( 8 kb: 145111478–145119486; NC_000086 ) , middle arm ( 1 . 1 kb: 145119487–145120587; NC_000086 ) and 3’-arm ( 2 . 1 kb: 145120588–145122703; NC_000086 ) by recombination from the RP23-319K12 BAC clone ( BACPAC Resources ) , and then cloned them into a targeting vector . The targeting vector was introduced into TT2 ES cells ( C57BL/6 × CBA genetic background ) by electroporation [46] . To generate chimeric mice , ES cells in which homologous recombination had occurred were injected into 8-cell stage embryos . Germ line transmission of the Sirh11/Zcchc16 MT allele was confirmed by Southern blot and PCR using the genome prepared from pups in which male Sirh11/Zcchc16 chimeric mice had been crossed with female C57BL/6J . To remove the flox region , we injected a pCAG/NCre plasmid [47] into the fertilized eggs generated by in vitro fertilization ( IVF ) from Sirh11/Zcchc16 MT hetero eggs and C57BL/6J sperm , thus establishing Sirh11/Zcchc16 neo mice . To obtain Sirh11/Zcchc16 KO mice , we injected a pCAGGS-FLPe plasmid ( Gene bridge ) into the fertilized eggs generated by IVF from Sirh11/Zcchc16 neo hetero eggs and C57BL/6J sperm . Exclusion of the neo cassette was confirmed by genomic PCR of the pups’ DNA . Southern blot analysis was performed using a standard protocol . Five micro grams of genomic DNA from the tail were digested by restriction enzymes NheI ( for the 5’ probe ) and NcoI ( for the 3’ probe ) , respectively . Hybond-N+ ( GE Healthcare ) membranes blotted with digested DNA were hybridized in Church buffer with radio isotope-labelled probes . The 5’ and 3’ probes were generated by genomic PCR using the following sequences: 5’ probe: 145109101–145110193; NC_000086; 3’ probe: 145124154–145124725; NC_000086 . The Sirh11/Zcchc16 KO allele was detected by genomic PCR . Genome DNA was prepared from the tail or ear tip using a DNeasy Blood & Tissue Kit ( QIAGEN ) . PCR was performed using ExTaqHS polymerase ( TaKaRa ) with the following primers: Sirh11-F1: 5’-ATGTATCCTAAGGTGATCCG-3’ and Sirh11-R2: 5’-ATGTGATGCCACAGCAACTC-3’ . Sirh11/Zcchc16 KO mice were backcrossed to C57BL6/J for more than 10 generations . Total RNA was prepared from frozen tissues using ISOGEN ( NIPPON GENE ) and ISOGEN-LS ( NIPPON GENE ) . The cDNA was made from total RNA ( 1 μg ) using Revertra Ace qPCR RT Master Mix ( TOYOBO ) . Quantitative RT-PCR analysis was performed using Fast SYBR Green Master Mix ( Life technologies ) and a StepOnePlus System ( Life technologies ) by means of an absolute quantification method . Data was normalized by Actb expression . Student’s t-test was used for statistical analysis . The following primer sequences were used: Sirh11-F9: 5’-TGGTGCTGGTGTATTTCCCC-3’ and Sirh11-R9: 5’-TGGCACAGTGGTTAGTGAGGC-3’; Sirh11-F5: 5’-AAGAGGAGGATAGGAAATCACTTTG-3’ and Sirh11-R5: 5’-GTTGTTAGGACAAGGTTGAGG-3’; Dbh-F1: 5’-ACTGAACGGAGAAGCCCTGGAC-3’ and Dbh-R1: 5’-CACCAGAGGACCAACAGGGTCG-3’; Actb-F: 5’-AAGTGTGACGTTGACATCCG-3’ and Actb-R: 5’-GATCCACATCTGCTGGAAGG-3’ . To reproduce the hetero and wild-type progeny for the behavior-screen , in vitro fertilization ( IVF ) was performed . Wild type male mice were used as the source for sperm , while hetero female mice were used as the source for the oocytes used for IVF . ICR female mice ( CLEA Japan , Tokyo , Japan ) were used as pseudopregnant recipients for embryo transfer . Sperm were collected from the caudae epididymides of adult male mice ( 20 w ) and allowed to diffuse in human tubal fluid ( HTF ) medium . After preincubation for approximately 1 hour to allow for capacitation , the sperm were used for insemination . Meanwhile , immature hetero female mice ( 4 w ) were superovulated using intraperitoneal injections of PMSG and HCG ( Serotropin and Gonatropin; ASKA Pharmaceutical Co . , Tokyo , Japan ) with an interval of 48 hours between injections . Approximately 15–17 hours after the HCG injection , the oocyte-cumulus complexes were collected from the oviducts of the superovulated female mice . The complexes from several female mice were then placed in the HTF fertilization medium . Insemination was performed by adding the sperm suspension to the fertilization medium containing complexes and culturing at 37°C with 5% CO2 in air . Twenty-four hours after insemination , 2-cell embryos were transferred into the oviducts of pseudopregnant ICR females mated to vasectomized ICR males . All pups were delivered naturally after embryo transfer . After weaning at four weeks age , they were employed as breeding individuals in a single-breeding cage . Each mouse was placed in the corner of an open-field apparatus ( 400 mm wide x 400 mm long x 300 mm high; O’Hara & Co . , Ltd . , Tokyo , Japan ) made of white polyvinyl chloride . The distance traveled by each animal in the open field was recorded for 20 min with a video-imaging system ( Image OF9; O’Hara & Co . , Ltd . , Tokyo , Japan ) . The mice were tested on two separate occasions at 9 and 10 w . The unpaired two sample T-test was used for statistical analysis . A commercially available light/dark chamber ( O'Hara & Co . , Ltd . ) was used for the light/dark transition test . The apparatus consists of a light chamber ( 200 mm long × 200 mm wide × 250 mm high ) made of white vinyl chloride plates and a dark chamber with the same dimensions made of black vinyl chloride plates . The apparatus has an opening ( 50 mm wide × 30 mm high ) in the middle of the wall that joins the two chambers . The opening is controlled by a guillotine door . The latency for entering into the lighted chamber and number of transitions between the light and dark chambers were measured . The mice were tested at 8 weeks of age . The unpaired two sample T-test was used for statistical analysis . Each mouse was placed alone in a testing cage ( 227 mm wide x 329 mm long x 133 mm high ) under a 12-h light–dark cycle ( light on at 08:00 h ) and had free access to both food and water . After 1 day of acclimation , spontaneous activity in the cage was measured for 5 continuous days ( starting at 08:00 ) with an infrared sensor ( activity sensor , O’Hara & Co . , Ltd . ) . The mice were tested at 10 w . The two-way analysis of variance ( ANOVA ) ( effects of genotype and date ) was used for statistical analysis . The Y-maze test was performed using male mice at 16–18 weeks of age ( WT: N = 6 , KO: N = 8 ) . The apparatus was a black , plastic maze with three arms ( 400 mm long × 30 mm wide × 150 mm high , 120 degrees ) . Mice were placed at the center of the apparatus and allowed to move freely through the maze for 2 min . The sequence and total number of arm entries were recorded by video camera for 5 min . When all 4 limbs of the mouse were within a pathway , it was considered an entry . An “alternation” was counted when a mouse successively entered 3 different arms . Spontaneous alternating search behavior was calculated by the following equation: alternation behavior ( % ) = [the number of alternations/ ( total number of arm entries – 2 ) ] × 100 . The results are given as the mean and standard error of the mean ( S . E . M . ) . Statistical analysis was conducted using computer software ( Prism , version 6 . 0c , GraphPad Software , San Diego , CA , USA ) for a comparison across the experimental conditions . Statistical evaluations for the measurement of behavior or the NA level were carried out using Mann-Whitney U test . A P-value <0 . 05 was considered to be significant . We performed in vivo microdialysis measurements of extracellular monoamines in the prefrontal cortex of 11–13 week-old mice . A guide cannula ( AG-4; EICOM , Kyoto , Japan ) was implanted into the prefrontal cortex ( +1 . 9 mm anteroposterior and +0 . 5 mm mediolateral relative to the bregma and −2 . 0 mm dorsoventral relative to the dura of the skull ) under inhalation anesthesia with nitrous oxide , oxygen and isoflurane ( 2% ) . Two days after surgery , a dialysis probe ( AI-4-01 , 1-mm membrane length; EICOM ) was inserted through the guide cannula and perfused at a flow rate of 1 μl/min with artificial cerebrospinal fluid ( 147 . 0 mM NaCl , 4 . 0 mM KCl , 2 . 3 mM CaCl2 ) or high potassium-containing artificial cerebrospinal fluid ( 51 . 0 mM NaCl , 100 . 0 mM KCl , 2 . 3 mM CaCl2 ) . Samples were collected every 20 min and injected directly into an HPLC column ( EICOMPAK CA-5ODS; EICOM ) by an auto injector ( EAS-20; EICOM ) . The concentrations of the monoamines in the dialysate were determined by HPLC with an electrochemical detector ( ECD300; EICOM ) . After the experiments , 1 μl of 0 . 3% Evans blue dye was microinjected through the cannula to histologically verify the position of the probe , and only data from animals with a correct probe placement were used in the analysis . The results are given as the mean and standard error of the mean ( S . E . M . ) of the data as described in the Y-maze test . | Retrotransposon-derived DNA sequences occupy approximately 40% of the mammalian genome , compared with only 1 . 5% of protein coding genes . They have been commonly considered “junk DNA” and even potentially harmful for host organisms . However , a series of knockout ( KO ) mouse analyses demonstrated that at least some of the LTR retrotransposon- and retrovirus-derived sequences play essential roles in the current mammalian developmental system as endogenous genes , such as Peg10 , Peg11/Rtl1 , Sirh7/Ldoc1 , SYNCYTINs and FEMATRIN-1 , which are active in multiple aspects of placental function . Here we demonstrate that another LTR retrotransposon-derived gene , Sirh11/Zcchc16 , plays an important role in cognitive function in the brain . Sirh11/Zcchc16 KO mice exhibit abnormal behaviors related to cognition , including attention , impulsivity and working memory , possibly due to the locus coeruleus-noradrenaline ( LC-NA ) system , suggesting that human SIRH11/ZCCHC16 may be involved in X-linked intellectual disability and/or attention-deficit/hyperactivity disorder . Comparative genome analysis demonstrates that SIRH11/ZCCHC16 was acquired in a common eutherian ancestor , suggesting that it contributed to eutherian brain evolution because it confers a critically important advantage in the competition that occurs in daily life . This study provides further insight into the impact of LTR retrotransposon-derived genes on mammalian evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Cognitive Function Related to the Sirh11/Zcchc16 Gene Acquired from an LTR Retrotransposon in Eutherians |
Kinetic models of metabolism require detailed knowledge of kinetic parameters . However , due to measurement errors or lack of data this knowledge is often uncertain . The model of glycolysis in the parasitic protozoan Trypanosoma brucei is a particularly well analysed example of a quantitative metabolic model , but so far it has been studied with a fixed set of parameters only . Here we evaluate the effect of parameter uncertainty . In order to define probability distributions for each parameter , information about the experimental sources and confidence intervals for all parameters were collected . We created a wiki-based website dedicated to the detailed documentation of this information: the SilicoTryp wiki ( http://silicotryp . ibls . gla . ac . uk/wiki/Glycolysis ) . Using information collected in the wiki , we then assigned probability distributions to all parameters of the model . This allowed us to sample sets of alternative models , accurately representing our degree of uncertainty . Some properties of the model , such as the repartition of the glycolytic flux between the glycerol and pyruvate producing branches , are robust to these uncertainties . However , our analysis also allowed us to identify fragilities of the model leading to the accumulation of 3-phosphoglycerate and/or pyruvate . The analysis of the control coefficients revealed the importance of taking into account the uncertainties about the parameters , as the ranking of the reactions can be greatly affected . This work will now form the basis for a comprehensive Bayesian analysis and extension of the model considering alternative topologies .
Kinetic models of metabolism require quantitative knowledge of detailed kinetic parameters ( e . g . maximum reaction rates , enzyme affinities for substrates and regulators ) . However , our knowledge about these parameters is often uncertain . When the parameters are measured , various sources of error can affect the results: experimental noise at the technical and biological levels , systematic bias introduced by parameters being measured in vitro instead of in vivo or by the choice of specific experimental conditions ( pH , temperature , ionic strength , etc . ) . Moreover , a substantial number of important parameters have never been measured and the estimates included in models are based either on values measured in closely related species or on the general distribution of similar parameters [1] . Few general methods for dealing with this uncertainty have been suggested [2]–[6] . Here we present an analysis of the effect of parameter uncertainties on a particularly well defined example of a quantitative metabolic model: the model of glycolysis in bloodstream form Trypanosoma brucei [7] ( see Fig . 1 ) . This ordinary differential equation ( ODE ) model is mainly using parameters measured on purified enzymes rather than fitted , and , since its first publication in 1997 , it has been updated [8] and extended [9]–[11] several times , making it one of the most highly refined dynamic models of a metabolic pathway published to date . The model has been successfully used to predict the “turbo explosion” that would happen in the absence of the glycosome , the subcellular compartment in which the first seven enzymes of glycolysis are localized in T . brucei [8] . This important property was confirmed experimentally more than 10 years after the model was initially proposed [11] . In this paper we used the last updated version of the model published [11] with slight modifications to take into account the equilibrium constants of all reactions ( see methods ) . Explicitly considering the uncertainties of parameters in the analysis of the model allowed us to gain interesting new insights into its behaviour . Most importantly , our analysis allowed us to quantify the degree of confidence concerning diverse properties of the system , including the hierarchy of control which is relevant for prioritizing potential drug targets . The resulting quantitative profile of model uncertainties , including the identification of major fragilities and areas in need of further examination , provides a solid basis for future model extensions . These will in turn introduce new uncertainties and should be dealt with using the same general framework established here .
In order to specify the uncertainty associated with each parameter , we gathered all available information relating to the sources of the values used in the model . Information included data on how kinetics were measured , the number of replicates and the standard error of mean values when available , additional calculations used to estimate the parameter from the observed values , and any “corrections” for additional factors such as temperature or pH . For this purpose , we created the “SilicoTryp” wiki , a MediaWiki-based ( http://www . wikimedia . org ) website dedicated to the detailed documentation of the sources of parameters used in the latest version of the model of glycolysis in T . brucei ( http://silicotryp . ibls . gla . ac . uk/wiki/Glycolysis ) [10] . Each reaction is described on its own page , which contains the rate equation and the detailed references and calculations for each parameter ( see Fig . 2 for an example ) . From the information collected , probability distributions could be inferred for each parameter as described in Methods . supplementary text S1 shows the estimated distributions for all parameters . To model the effect of uncertainty , we sampled values for each parameter according to its probability distribution , generating a ensemble of alternative models . Together these alternative models accurately represent our degree of uncertainty about the correct parameters , assuming that our knowledge of each parameter value is independent of the other parameters ( see Methods for one example , the equilibrium constant , where this assumption is violated and needs to be accounted for ) . This collection of models can then be used to analyse model behavior and the associated uncertainties . The same properties that were studied with the fixed parameter version of the model can be studied with each alternative model . The distribution of the results shows the robustness and the degree of certainty we have about the inferred model properties ( e . g . the steady-state concentrations of the metabolites and the control coefficients ) considering our current knowledge about the parameters and the topology of the model .
Dynamic models of metabolism are powerful tools to infer interesting and often unexpected properties of cellular physiology . However , the data used to build models from diverse sources can lack accuracy and precision . Here we demonstrate how model output can vary when the uncertainties associated with incomplete and variable datasets are explicitly considered in studying a model . We took as an example the well characterised model of the compartmentalised glycolysis in the parasitic protozoan T . brucei . It should be noted that our assessment of the effect of parameter uncertainty on the conclusions that are possible is very conservative . Whenever possible , we have restricted our uncertainty estimates to the level of experimental uncertainty seen within a single assay . This ignores the systematic effects of differences in , e . g . , temperature , pH or ion compositions , or biases introduced in sample preparation , all of which would increase uncertainty as can also be seen when parameter values from different laboratories are compared . However , even with these relatively limited uncertainties , we were able to assess the robustness and variability of various properties of the model . The first property that we studied is the ability of the model to reach steady-state rapidly . Surprisingly , a significant proportion ( 60% ) of the models we generated by sampling the parameters did not allow the model to reach steady-state within 300 minutes , due to the accumulation of either 3-phosphoglycerate or pyruvate in the cytosol . This phenomenon could be attributed to two individual parameters , the maximal reaction rates of phosphoglycerate mutase and pyruvate transport which , when operating below their mean value ( but still very close to it ) , caused the accumulation of two metabolites ( 3-phosphoglycerate and pyruvate respectively ) . For the pyruvate transporter , the analysis suggested a mechanism that could avoid this problem: alanine aminotransferase has been shown , unexpectedly , to be essential in bloodstream form T . brucei [18] , and its activity comparable with the rate of pyruvate efflux . This would be sufficient to exert a substantial influence on the intracellular pyruvate concentration . The maximal reaction rate of phosphoglycerate mutase is difficult to measure directly [27] , therefore further experimental and theoretical studies are required to refine our knowledge about this reaction . Indeed , the model predicts that current values for PGAM are probably lower than those operative in T . brucei , and some effort should be made to determine whether the values are indeed higher . We then analysed the distribution of the steady-state fluxes between the pyruvate and glycerol producing branches of glycolysis both in aerobic and anaerobic conditions . In totally anaerobic conditions , the distribution was very well conserved . Indeed , this property is entirely constrained by the topology of the model and thus this result was expected . Our analysis shows that the distribution of the fluxes is more variable in aerobic conditions , consistent with previously unexplained variation in experimental observations ( although changes in oxygen tension within different cells in measured populations would create the same effect ) . Further analysis of the steady-state concentrations allowed us to distinguish the metabolites that are only moderately affected by the parameter uncertainties and follow an approximate log-normal distribution , such as and NADH , from the metabolites that follow a more complex distribution such as glycosomal ATP . ATP is constrained by a conserved sum , therefore its steady-state concentration always stays within reasonable limits . Technical limitations mean that the concentration of glycosomal ATP is not directly accessible for experimentation ( glycosomes cannot be purified efficiently enough ) . Therefore , only by acquiring additional data about the parameters of the model can assumptions about these concentrations at steady-state be refined . Finally , we analysed the control coefficients of each enzyme using our collection of models . These properties are especially important in the case of glycolysis in T . brucei , as they allow us to identify potential drug targets . Our analysis reveals that , although the reaction that has the most control over the glucose consumption flux is the glucose transporter in 40 . 3% of the models , two other reactions maximally control the flux in a significant proportion of the models: PGAM ( 31 . 1% ) and GAPDH ( 28 . 5% ) . Moreover , the activity of GAPDH has been reported to be inhibited by an unknown metabolite [25]; if this inhibition occurs in vivo , it might have an important role in the control of glycolytic flux . Interestingly , partial inhibition of GAPDH has been shown to affect parasite growth and glycolytic flux [26] , and selective inhibitors of the T . brucei enzyme have been shown to be trypanocidal [28] . The rest of the control coefficient hierarchy is more variable . Either this variability is a true reflexion of biological noise or the result of our lack of knowledge about some parameters of the model . The data derived from the work performed here point to several further studies , including analysis of the role of alanine amino transferase in the regulation of pyruvate concentration and more exact quantification of pyruvate transport and phosphoglycerate mutase kinetics . The detailed description of parameter uncertainty will now form the basis for a comprehensive Bayesian analysis and extension of the model using alternative topologies [29] . These analyses will allow us to quantify our posterior belief about the parameters of the model when it is confronted with new experimental data such as measured metabolite concentrations in different conditions .
The model used in this paper is the last updated version [10] , [11] of the glycolysis model of T . brucei first published by Bakker et al . in 1997 [7] ( see Fig . 1 ) . To allow a straight-forward sampling of parameters , the rate equations were rewritten to contain the equilibrium constant instead of the ratio of values ( reverse over forward ) , using the Haldane equation [30] . This does not change the rates , but simplifies the sampling of the parameters , as we do not need to check for consistency with the thermodynamic equilibrium constant . For example , the phosphoglucose isomerase ( PGI ) rate equation was: ( 1 ) where . The Haldane equation gives: ( 2 ) Therefore , the rate equation of PGI can be rewritten as: ( 3 ) The list of sources used to compute the values of the equilibrium constants is available in supplementary text S3 and on the SilicoTryp wiki ( http://silicotryp . ibls . gla . ac . uk/wiki/Glycolysis ) . The model in [11] considered the transport reactions between the cytosol and the glycosome and adenylate kinase ( see special cases ) to be very fast compared the other reactions of the model . Therefore , they were not explicitly modelled . To enable consideration of the effect of parameter uncertainty on the rate of these transport reactions , we modelled them explicitly using mass action kinetics . As we considered that these reactions have an equilibrium constant of unity ( no preferential accumulation or exclusion in one of the compartments ) , we used a single rate parameter for each transport reaction . For example , the transport of glucose between the cytosol and the glycosome is modelled as: ( 4 ) The model is available as supplementary dataset S1 ( SBML file [31] ) . The parameter values are as in [11] . The equilibrium constant are calculated from the values and the ration of over when necessary . In order to sample the model parameters , we needed to define a probability distribution for each parameter . These distributions can be defined empirically using arbitrary shapes , but for the sake of convenience it is usually appropriate to use standard shapes ( e . g . normal or log-normal distributions ) and then to estimate the parameters of these distributions ( usually the mean and standard deviation ) . All parameters were sampled using the MT19937 random number generator of Makoto Matsumoto and Takuji Nishimura [37] implemented in the GNU Scientific Library ( GSL ) [38] . The random numbers where then transformed to follow their assumed probability distribution using the random number distribution function implemented in the GSL library . The steady states were calculated using the SOSlib library [39] . Steady-state is considered if the mean+standard deviation of the rates of change of all metabolite concentrations is lower than a user-defined parameter of SOSlib . The initial conditions were set using the steady-state concentrations calculated using the mean values of all parameters . For any sampled model , it is assumed that steady state should be reached within 300 minutes of simulated time ( steady state detection threshold , parameter 1 per simulated minute ) . We checked that the steady-state calculations give similar results in COPASI [40] and PySCeS [41] using their default parameters . We also verified that the parameter sets that do not allow the model to reach steady state in these conditions show accumulation of individual metabolites beyond reasonable concentrations ( hundreds or thousands of millimol per liter , see Results and Fig . 3 ) . The control coefficients were computed using the methodology described by Bakker et al . [8] . The computation of control coefficients requires more precise steady-states calculations . Therefore , the parameters of SOSlib were set to: maximal time minutes and the threshold . | An increasing number of mathematical models are being built and analysed in order to obtain a better understanding of specific biological systems . These quantitative models contain parameters that need to be measured or estimated . Because of experimental errors or lack of data , our knowledge about these parameters is uncertain . Our work explores the effect of including these uncertainties in model analysis . Therefore , we studied a particularly well curated model of the energy metabolism of the parasite Trypanosoma brucei , responsible for African sleeping sickness . We first collected all the information we could find about how the model parameters were defined on a website , the SilicoTryp wiki ( http:///silicotryp . ibls . gla . ac . uk/wiki/ ) . From this information , we were able to quantify our uncertainty about each parameter , thus allowing us to analyse the model while explicitly taking these uncertainties into account . We found that , even though the model was well-defined and most of its parameters were experimentally measured , taking into account the remaining uncertainty allows us to gain more insight into model behavior . We were able to identify previously unrecognised fragilities of the model , leading to new hypotheses amenable to experimental testing . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biochemistry",
"biology",
"computational",
"biology",
"metabolism"
] | 2012 | Dynamic Modelling under Uncertainty: The Case of Trypanosoma brucei Energy Metabolism |
Genetic variants underlying reduced male reproductive performance have been identified in humans and model organisms , most of them compromising semen quality . Occasionally , male fertility is severely compromised although semen analysis remains without any apparent pathological findings ( i . e . , idiopathic subfertility ) . Artificial insemination ( AI ) in most cattle populations requires close examination of all ejaculates before insemination . Although anomalous ejaculates are rejected , insemination success varies considerably among AI bulls . In an attempt to identify genetic causes of such variation , we undertook a genome-wide association study ( GWAS ) . Imputed genotypes of 652 , 856 SNPs were available for 7962 AI bulls of the Fleckvieh ( FV ) population . Male reproductive ability ( MRA ) was assessed based on 15 . 3 million artificial inseminations . The GWAS uncovered a strong association signal on bovine chromosome 19 ( P = 4 . 08×10−59 ) . Subsequent autozygosity mapping revealed a common 1386 kb segment of extended homozygosity in 40 bulls with exceptionally poor reproductive performance . Only 1 . 7% of 35 , 671 inseminations with semen samples of those bulls were successful . None of the bulls with normal reproductive performance was homozygous , indicating recessive inheritance . Exploiting whole-genome re-sequencing data of 43 animals revealed a candidate causal nonsense mutation ( rs378652941 , c . 483C>A , p . Cys161X ) in the transmembrane protein 95 encoding gene TMEM95 which was subsequently validated in 1990 AI bulls . Immunohistochemical investigations evidenced that TMEM95 is located at the surface of spermatozoa of fertile animals whereas it is absent in spermatozoa of subfertile animals . These findings imply that integrity of TMEM95 is required for an undisturbed fertilisation . Our results demonstrate that deficiency of TMEM95 severely compromises male reproductive performance in cattle and reveal for the first time a phenotypic effect associated with genomic variation in TMEM95 .
Impaired reproductive performance is a prevalent condition in both sexes of many species and up to 15% of couples are affected in humans [1] , [2] . The disability to reproduce is defined as infertility ( i . e . , sterility ) , whereas subfertility refers to any form of reduced fertility [3] . Low sperm concentration ( i . e . , oligospermia ) and the absence of spermatozoa ( i . e . , azoospermia ) , respectively , are frequently diagnosed in males with impaired fertility [4] . Further aberrant semen quality traits ( e . g . , abnormal sperm morphology [5] , reduced motility [6] , [7] ) account for another substantial fraction of reduced male fertility . However , semen analysis of a considerable number of males with impaired reproductive performance remains without any apparent pathological findings ( i . e . , unexplained/idiopathic infertility ) [8] , [9] . Semen quality traits have low to medium heritability in cattle populations [10] . Numerous genetic variants underlying routinely assessed semen quality traits have been identified so far in humans [11] , [12] , model species [13] and livestock populations [14] . However , the number of known genetic mechanisms causing idiopathic male subfertility is very small [15] , [16] and identified polymorphisms explain only a small fraction of its genetic variation [17] . Artificial insemination ( AI ) is predominant over natural service in most cattle populations and all ejaculates are closely examined immediately after semen collection . Only semen samples without any apparent abnormalities , such as low sperm count , reduced progressive motility , low viability , abnormal morphology of spermatozoa , are used for insemination . However , the reproductive performance indicated by the proportion of successful inseminations varies considerably among AI sires [18] , [19] . So far , genome-wide association studies ( GWAS ) for male reproductive traits were of limited success in cattle populations [20] , [21] and only one putatively causative mutation has been identified [22] . Here we report a new recessively inherited variant of idiopathic male subfertility in the Fleckvieh ( FV ) cattle population . The mapping of the underlying genomic region was facilitated by using high-density genotypes in a large sample of artificial insemination bulls with phenotypes for reproductive performance assessed based on 15 million artificial inseminations . Exploiting whole-genome re-sequencing data revealed a causative loss-of-function mutation in the transmembrane protein 95 encoding gene TMEM95 .
Phenotypes for male reproductive ability ( MRA ) were obtained for 7962 bulls of the FV population based on 15 . 3 Mio artificial inseminations ( AI ) . The values for MRA range from −40 to +13 and reflect the bulls' reproductive performance as percentage deviation from the population mean . Male reproductive ability is highly correlated ( r = 0 . 59 ) with the 56-day non-return rate ( NRR56 ) in cows . The NRR56 is the proportion of cows that are not re-inseminated within a 56-day interval after the first insemination . After visual inspection of the distribution of MRA , forty-nine bulls with exceptionally poor reproductive performance ( MRA<−20 ) were considered as subfertile ( Figure S1 and Table 1 ) . Animals with values for MRA below −20 ( = five standard deviations below the population mean ) were used as case group in a case-control design . Using MRA as quantitative trait in a genome-wide association study ( GWAS ) yielded a strong association signal on bovine chromosome ( BTA ) 19 ( P = 4 . 38×10−20 , Figure S2 ) . However , the association signal was more pronounced using 49 subfertile animals ( MRA<−20 ) as case group and the remaining 7913 animals as controls ( Table 1 and Figure 1A ) . The most significantly associated SNP is located at 30 , 220 , 186 bp ( ARS-BFGL-NGS-11488; P = 4 . 08×10−59 ) . Autozygosity mapping revealed a common 1386 kb segment ( 26 , 580 , 096 bp–27 , 956 , 634 bp ) of extended homozygosity in 40 subfertile bulls containing 80 genes ( Figure 1B and Table S1 ) . None of 7913 bulls with normal reproductive performance was homozygous for the 1386 kb segment , indicating recessive inheritance . Semen samples of 40 homozygous bulls had been used for 35 , 671 artificial inseminations with an average of 892 inseminations per bull . This is a typical number for test inseminations performed with semen samples of young bulls in progeny testing based breeding programmes . However , only 619 ( 1 . 74% ) of those inseminations were successful ( Table S2 ) . There was no evidence for the presence of large structural variants ( i . e . , copy number variations ) within the segment of extended homozygosity ( Figure S3 ) . The proportion of missing genotypes did not significantly differ between cases and controls ( P>0 . 09 ) for all SNPs located within the associated region . The frequency of the subfertility-associated haplotype amounts to 7 . 2% . Of 7962 genotyped bulls with phenotypes for MRA , 1068 ( 13 . 41% ) carry the deleterious haplotype in heterozygous state . The carrier frequency increased considerably within the last years ( P = 0 . 0002 , Figure S4 ) . The reproductive performance of heterozygous bulls is normal , indicating recessive inheritance ( Figure 2 ) . Of 1952 primiparous cows , 291 are heterozygous and 17 are homozygous for the subfertility-associated haplotype . The haplotype neither affects reproductive performance nor milk production traits in females ( Table S3 ) . The haplotype distribution does not deviate from the Hardy-Weinberg equilibrium , neither in females ( P = 0 . 303 ) nor in males ( P = 0 . 817 ) . Both , haplotype and pedigree analysis allowed to trace the mutation back to the bull HAXL ( *1966 ) ( Figure S5 ) . HAXL appears in the pedigrees of 7779 out of 7962 bulls ( 97 . 70% ) and can be considered as the most important ancestor of the current FV population [23] . Whole genome re-sequencing of 43 animals and subsequent multi-sample variant calling yielded genotypes at 17 . 17 million sites [23] . Among them , 5965 ( 5287 SNPs and 678 INDELs ) are located within the subfertility-associated region on BTA19 ( 26 , 580 , 096 bp to 27 , 956 , 634 bp ) . Six of the 43 re-sequenced animals were identified as carriers of the associated haplotype via high-density genotypes . The sequence data were filtered for variants compatible with the supposed recessive inheritance , i . e . , heterozygous in carriers and homozygous for the reference allele in non-carriers ( see Material & Methods , Figure S6 ) . After filtering , 26 SNPs and six INDELs were retained as candidate causal mutations ( Table S4 and S5 ) . The functional effects of those variants were predicted based on the gene annotation of the UMD3 . 1 assembly of the bovine genome [24] . Four of the 32 compatible variants were located in coding regions ( Table 2 ) . Among them , we considered a nonsense mutation in TMEM95 ( rs378652941 , c . 483C>A , p . Cys161X , Chr19: 27 , 689 , 622 bp ) as the prime candidate causal mutation ( Figure 3A and 3B ) . The nonsense mutation was subsequently confirmed in the re-sequenced animals by classical Sanger sequencing ( Figure S7 and S8 ) . Genotypes for two non-synonymous substitutions in ACDVL and KIF1C and for the nonsense mutation in TMEM95 were obtained for cases and controls using TaqMan genotyping assays ( Table 3 ) . Only c . 483C>A , introducing the premature stop-codon in TMEM95 ( p . Cys161X ) , was perfectly associated . All animals , which are homozygous for the subfertility-associated haplotype , are homozygous for the non-reference allele , whereas none of 1396 FV bulls with normal reproductive performance are homozygous . The polymorphism is present in the FV breed only; 277 Holstein-Friesian and 278 Braunvieh animals are homozygous for the reference allele . The c . 483C>A-mutation is not segregating among 15 Jersey , 47 Angus and 129 Holstein-Friesian animals which were sequenced in the context of the 1000 bull genomes project [25] . TMEM95 encodes a highly conserved single-pass type I transmembrane protein consisting of 183 amino acids with a predicted extracellular N-terminal signal peptide , a 23-amino acid transmembrane domain ( amino acid position 153 to 175 ) and a 8-amino acid intracellular C-terminal domain ( Figure 3C and Figure S9 and S10 ) . The premature stop codon introduced by the c . 483C>A-mutation is located within the predicted transmembrane domain and truncates the protein by 22 amino acids . Semen quality ( morphology , vitality , total motility ) was analysed using cryopreserved semen samples of 30 bulls ( 10 wt/wt , 10 wt/mt , 10 mt/mt ) . In all ejaculates , spermatozoa showed less than 20% morphological alterations and less than 5% morphological alterations of the head . Total motility after thawing ranged from 50 to 65% . Statistical analysis showed no significant differences in the proportion of motile spermatozoa from wt/wt , wt/mt and mt/mt bulls ( Table 4 ) . As shown by eosin staining , 40 to 70% of the spermatozoa were viable after thawing . There were no significant differences in the percentages of viable spermatozoa between wt/wt , wt/mt and mt/mt bulls . Additionally , ejaculate volume , sperm concentration and progressive motility were assessed in fresh semen samples of 203 AI bulls ( 177 wt/wt , 21wt/mt , 5 mt/mt ) . Ejaculate volume was above 5 ml , sperm count was above 1 . 42 Mio/µl and the proportion of spermatozoa with progressive motility was above 70% for all animals ( Table 5 ) . A mouse-derived polyclonal antibody generated against human transmembrane protein 95 was used to locate its position in spermatozoa of 33 bulls ( 10 wt/wt , 10 wt/mt and 13 mt/mt ) . In spermatozoa of wt/wt bulls , TMEM95 was distinctly located on the plasma membrane of the acrosome ( Figure 4A ) . Staining was also visible on the equatorial segment of the head . The sperm neck was regularly labelled . Spermatozoa of wt/mt and wt/wt bulls showed an identical staining pattern ( Figure 4B ) , whereas spermatozoa of mt/mt bulls did not show any staining on the head ( Figure 4C ) . Weak fluorescence was detected in the midpiece of the tail in spermatozoa of all animals due to the autofluorescence of the mitochondria . In the negative controls , there was no signal detectable on the sperm head whereas the midpiece of the tail showed weak autofluorescence ( Figure S11 ) .
The genome-wide association study ( GWAS ) with imputed genotypes for 7962 artificial insemination bulls identified a genomic region on BTA19 for male reproductive ability ( MRA ) in the FV population . Autozygosity mapping revealed a common 1386 kb segment of extended homozygosity in 40 bulls with unexplained exceptionally poor reproductive performance . None of the bulls with normal reproductive performance was homozygous indicating recessive inheritance . Only 1 . 74% of inseminations performed with semen samples of affected bulls were successful , although semen quality parameters were within a normal range , reflecting idiopathic subfertility [26] . The newly identified congenital defect is denominated as “Bovine Male Subfertility” and accounts for 82% of FV bulls with exceptionally poor reproductive performance . However , we cannot exclude the possibility that homozygous males are infertile and that the very low proportion of successful inseminations reflects errors in parentage recording which might be as high as 10% in dairy cattle breeding programmes [27] . In progeny testing based breeding programmes , semen doses of young bulls are used for approximately 1000 test inseminations [28] . These artificial inseminations are performed within very short time , precluding the early identification of subfertile/infertile bulls . Identifying and culling bulls with poor fertility prognosis ( i . e . , homozygous bulls ) before they are used for artificial insemination is now possible . There was no evidence for any additional genomic region underlying idiopathic male subfertility in the FV population , although the reproductive ability of nine bulls which are not homozygous for the c . 483C>A-mutation , is very low . However , the number ( n = 9 ) of subfertile bulls not attributable to the BTA19 locus might not be sufficient for detecting additional loci ( Figure S12 and Table S6 ) . The potential of targeted or whole genome re-sequencing for the identification of causal trait variants has been demonstrated in several species ( e . g . , [29]–[31] ) including cattle [32]–[34] . Causal trait variants for monogenic disorders are traditionally identified by sequencing case/control-panels and by subsequently comparing allele counts in affected and unaffected individuals . However , the concept of the present study is different: the identification of the underlying mutation was based on whole genome re-sequencing data of 43 unaffected FV animals explaining a vast majority of the population's genomic variation [23] . As the frequency of the mutation was reasonably high ( 7 . 2% ) , the affected haplotype was present in heterozygous state in six of the re-sequenced animals . Filtering the re-sequencing data for variants compatible with the supposed recessive inheritance pattern revealed a plausible candidate causative loss-of-function mutation in TMEM95 encoding the transmembrane protein 95 . The nonsense mutation was perfectly associated in 1990 animals representing three different breeds . To our knowledge , this is the first report of a phenotypic effect associated with variation in TMEM95 in any organism . So far , there are no clues about the precise function of TMEM95 . However , it seems likely that TMEM95 is involved in sperm-egg interactions , which has been shown to be the main function of sperm-specific transmembrane proteins ( e . g . , [35] , [36] ) . The phenotype in the present study resembles phenotypic patterns of Caenorhabditis elegans resulting from an impaired function of sperm-specific transmembrane proteins [37] , [38] . Taken together , our findings evidence genomic variation within TMEM95 to severely compromise the reproductive performance in cattle . The causative polymorphism ( c . 483C>A , rs378652941 ) introduces a premature stop codon in TMEM95 ( p . Cys161X ) . The affected codon resides within the predicted transmembrane domain of TMEM95 most likely resulting in a disturbed anchorage of the truncated protein in the sperm plasma membrane bilayer . It is also likely that the resulting truncated protein is absent due to nonsense-mediated mRNA decay [39] . Our data show no evidence that the mutation affects any of the routinely assessed semen quality parameters in vitro . However , we cannot exclude the possibility that the mutation affects semen quality parameters , e . g . , vitality and motility , in vivo [40] , [41] . Transmembrane protein 95 is primarily located on the acrosomal membrane of the sperm head indicating that it may be involved in the acrosome reaction . Spermatozoa of mt/mt animals showed no fluorescence at the acrosomal membrane implying deficiency of TMEM95 . Thus , successful fertilization by spermatozoa of mt/mt animals might be compromised . This is supported by the fact that the equatorial segment of the acrosome , which provides the first contact of the spermatozoon with the cell membrane of the oocyte [42] , is also labelled in spermatozoa of fertile animals . The sperm neck contains the centriole and is essential for cell division and development of the early embryo [43] . Labelling of the neck indicates an additional potential role of TMEM95 after fertilization during the first cell divisions of the early embryo . Although spermatozoa of mt/mt animals showed fluorescence , neither on the acrosomal membrane and the equatorial segment nor at the centriole , weak unspecific fluorescence was observed at the midpiece of the tail . This fluorescence pattern is also present in spermatozoa of wt/wt and wt/mt animals . The midpiece is the only region of spermatozoa that contains mitochondria [44] . The weak fluorescence of the midpiece is due to unspecific autofluorescence of the mitochondria , which has been described in several organs and species [45]–[47] . Male subfertility is also present in other species besides cattle [9] , [48]–[50] , and compromised sperm surface proteins account for a substantial number of males with distinctly reduced reproductive ability in humans [15] , [40] . Our results demonstrate that TMEM95 is another sperm surface protein , which is likely to be involved in sperm-egg plasma membrane interactions . Its protein sequence is highly conserved among species ( Figure S10 ) and genetic variants disrupting TMEM95 are likely to induce male subfertility also in other species than cattle . Numerous polymorphic sites have been identified in human TMEM95 , among them several potential loss-of-function variants ( Figure S13 ) . Based on our findings it is highly recommended to systematically survey variants in TMEM95 as potentially causal for idiopathic male in- or subfertility in any species . Frequencies of variants that disrupt protein-coding genes are usually low in human populations [51] , [52] . However , in livestock populations , the frequency of deleterious alleles might increase rapidly because individual sires can generate tens of thousands of progeny by artificial insemination ( e . g . , [53] , [54] , [32] ) . The loss-of-function mutation in TMEM95 can be traced back to HAXL ( *1966 ) , the most important ancestor of the current FV population . Within eight generations , the frequency of the deleterious allele increased to 8 . 9% and 1443 ( 13 . 92% ) animals of the present study are heterozygous . This increase of the allele frequency had been possible because phenotypic effects become apparent in homozygous males only . There are no phenotypic effects detectable neither in heterozygous nor in homozygous females ( Table S3 ) . In agreement with previous findings in livestock [55] and humans [15] , our results evidence that standard assessment of spermatozoa ( i . e . , morphology , motility and vitality ) is not sufficient to reliably anticipate male reproductive performance . All routinely assessed semen parameters of bulls homozygous for the nonsense mutation in TMEM95 comply with current requirements for artificial insemination in cattle [56] . It might be advisable to develop functional assays , e . g . , for the integrity of sperm surface proteins , for an efficient prospective monitoring of male fertility .
Semen samples were collected by approved commercial artificial insemination stations as part of their regular breeding and reproduction measures in cattle industry . No ethical approval was required for this study . Male reproductive ability ( MRA ) was evaluated in 7962 AI bulls of the German FV population . Semen samples of those bulls were used for 15 , 321 , 171 artificial inseminations with an average of 1924 artificial inseminations per bull . The phenotypes for MRA were obtained from routine breeding value estimation for reproductive traits , which are jointly estimated for males and females [57] . The resulting phenotypes for MRA represent the bulls' reproductive performance adjusted for environmental and genetic effects ( i . e . , year , season , flock , female mating partner ) . The lower the value for MRA , the worse is the bull's reproductive performance ( i . e . , the smaller the proportion of successful inseminations ) . A total of 3545 animals ( 1475 AI bulls , 2070 primiparous cows ) of the FV population were genotyped with the Illumina BovineHD Bead chip comprising 777 , 962 SNPs . Another 7073 AI bulls were genotyped with the BovineSNP50 Bead chip comprising ∼54 , 000 SNPs . The chromosomal position of the SNPs was determined according to the UMD3 . 1 assembly of the bovine genome [58] . Mitochondrial , Y-chromosomal and those SNPs with unknown chromosomal position were not considered for further analyses . Stringent quality control was carried out for each dataset separately using PLINK v1 . 07 [59] . Animals and SNPs with call-rate <0 . 95 were excluded , as well as SNPs with minor allele frequency <0 . 5% and those SNPs deviating significantly from the Hardy-Weinberg equilibrium ( P<10−6 ) . The pairwise genomic relationship [60] was compared with the pedigree relationship tracing pedigree records back to 1920 [61] . Animals showing major discrepancies of the pedigree and the genomic relationship were removed from the dataset , as such patterns indicate sample swaps . After quality control , the high-density dataset contained 3332 animals and 652 , 856 SNPs with an average per-individual call-rate of 99 . 17% . The medium-density dataset contained 7031 animals and 42 , 907 SNPs with an average per-individual call-rate of 99 . 75% . Genotype imputation was performed to extrapolate medium-density genotypes to higher density using a pre-phasing based approach . Haplotypes were inferred for the two datasets separately using Beagle [62] and subsequent haplotype-based imputation was performed with Minimac [63] . This approach yields high imputation accuracy in cattle [64] . The imputed dataset comprised 10 , 363 animals ( 8411 AI bulls/1952 primiparous cows ) and genotypes for 652 , 856 SNPs . Phenotypic records for MRA were available for 7962 bulls . Genome-wide association studies were performed applying a variance component based approach to account for population stratification . We used EMMAX [65] to fit the mixed model , where Y is a vector of phenotypes , b is the SNP effect , X is a design matrix of imputed SNP genotypes , u is a vector of additive genetic effects assumed to be normally distributed with mean 0 and ( co ) variance , with being the additive genetic variance and G being the realized genomic relationship matrix ( GRM ) of the 7962 bulls with phenotype information built based on 635 , 224 autosomal SNPs ( see above ) and where e is a vector of random normal deviates . The genomes of 42 key and contemporary animals of the FV population were sequenced at low- to medium coverage ( ø 7 . 4-fold ) and one animal was sequenced at high coverage ( 25-fold ) using Illumina GA IIx and HiSeq 2000 instruments [66] , [23] . Paired-end reads were obtained and mapped to the bovine reference sequence ( UMD3 . 1 [58] ) using the Burrows-Wheeler Aligner ( BWA ) [67] . PICARD ( http://picard . sourceforge . net ) was used to mark PCR-duplicates . Subsequent multi-sample variant calling with mpileup [68] yielded genotypes at 17 . 17 million sites . The re-sequencing data were contributed to the 1000 bull genomes project [25] and all variants were submitted to dbSNP [23] . Beagle phasing and imputation within the 43 sequenced animals improved the primary genotype calls ( a detailed overview of the entire variant calling pipeline and all obtained variants is presented in Jansen et al . [23] ) . Of 17 . 17 million sites , 5287 SNPs and 678 INDELs were located within the 1386 kb segment ( 26 , 580 , 096 bp to 27 , 956 , 634 bp ) of extended homozygosity on BTA19 . Of the 43 sequenced animals , six were identified as carriers of the subfertility-associated haplotype via high-density genotypes , among them the animal sequenced at high coverage . Assuming perfect correlation between the subfertility-associated haplotype and the causal mutation , the allele frequency of the causal mutation should be 7% ( 6 of 86 affected alleles ) in the sequence-derived genotypes . To account for inaccurately genotyped variants due to the low-coverage sequence data ( e . g . , mis-calling of heterozygous genotypes for rare variants [69] ) [70] and for possible phasing errors , a conservative mutation scan was performed to identify variants compatible with recessive inheritance ( Figure S6 ) . The 5965 polymorphic sites were filtered for variants that met three conditions: ( i ) the frequency of the non-reference allele is below 10% , ( ii ) the variant is heterozygous in the animal sequenced at high coverage and ( iii ) the variant is present in heterozygous state in at least three of five carrier sequenced at low coverage . PCR primers ( 5′-CACCCTGCCTTGTCTTTCAT-3′ and 5′-AGGCTCTGTCCTCGTTTTCA-3′ ) were designed for exon 6 of TMEM95 to scrutinize the rs378652941-polymorphism by classical Sanger sequencing in the re-sequenced animals as recommended by Jansen et al . [23] . Genomic PCR products were sequenced using the BigDye Terminator v1 . 1 Cycle Sequencing Kit ( Life Technologies ) on the ABI 3130x1 Genetic Analyzer ( Life Technologies ) . Genotypes for rs378652941 ( TMEM95:c . 483C>A , p . Cys161X , Chr19:27689622 ) , rs381722524 ( KIF1C:c . 197A>G , p . Gln66Arg , Chr19:27042848 ) and rs385135118 ( ACADVL:c . 706C>A , p . Pro236Thr , Chr19:27570146 ) were obtained by TaqMan genotyping assays ( Life Technologies ) in 1990 , 1222 and 1206 animals , respectively , representing three different breeds ( BV , FV , HF ) . The primer and probe sequences are listed in Table S7 . The topology of bovine transmembrane protein 95 ( NCBI reference sequence: XP_002695846 . 1 ) was predicted with SPOCTOPUS [71] and PHOBIUS [72] . Both methods predict the transmembrane protein topology while accounting for the existence of a N-terminal signal peptide sequence . The protein topology was visualized with SOSUI [73] . ClustalW [74] was used for multi-species alignment of the protein sequence and for the prediction of conserved regions . Cryopreserved ( −196°C ) sperm specimens of 30 bulls ( 10 wt/wt , 10 wt/mt , 10 mt/mt ) were obtained from artificial insemination companies . Two different ejaculates of each bull were evaluated with two straws pooled per ejaculate . After thawing ( 37°C , 30 s ) , sperm morphology was assessed by staining with Diff-Quik ( Siemens Healthcare , Germany ) . Sperm total motility was investigated immediately after thawing by phase-contrast microscopy using the Leica DM 1000 microscope ( Leica , Germany ) . Viability of spermatozoa was investigated after thawing by mixing 5 µl of the thawed ejaculate with aqueous Eosin Y solution ( Sigma Aldrich , Germany ) in a volume ratio of 1∶1 . Intact and viable spermatozoa stay colourless whereas spermatozoa with disturbed membrane integrity stain red . Counting of viable sperm was done within 10 seconds after mixing . Two hundred spermatozoa in at least two different fields of view were investigated at a magnification of 400× to analyse morphology , viability and motility . Fresh semen traits ( ejaculate volume , sperm count and progressive motility ) of 203 AI bulls ( 177 wt/wt , 21wt/mt , 5 mt/mt ) were provided by an artificial insemination company . Semen quality was analysed based on 10 , 682 ejaculates with an average of 52 . 6 ejaculates per bull . At semen collection , the age of the bulls ranged from 1 . 1 to 2 . 5 years . Immunohistochemistry on cryopreserved sperm specimens of 33 bulls ( 10 wt/wt , 10 wt/mt , 13 mt/mt ) was repeated for 3 times . After thawing ( 37°C , 30 s ) , spermatozoa were washed in phosphate buffered saline ( PBS ) twice and were diluted in PBS to a concentration of 500 , 000 spermatozoa/ml . Thereafter , drops of 7 µl were placed on 3-aminopropyl-ethoxysilane-coated slides and dried on a heating-plate at a temperature of 38°C . Subsequently , the slides were fixed in Bouin's solution for 7 min and washed in PBS twice . Non-specific binding was blocked by incubation in blocking buffer ( 0 . 1% bovine serum Albumin in PBS , Sigma-Aldrich , Germany ) for 5 minutes and in normal goat serum ( dilution 1∶5 in PBS , Invitrogen , Germany ) . Next , the spermatozoa were incubated with the first antibody Yomics Ab989 ( mouse-derived against human TMEM95 , Primm , USA ) in a dilution of 1∶200 in blocking buffer at 4°C overnight . The secondary antibody was the Fluorescein ( FITC ) -conjugated AffiniPure Goat anti Mouse IgG ( H+L ) ( Dianova , Germany , dilution 1∶200 ) . Negative controls were done by a ) replacing the first antibody with PBS and b ) using a non-relevant anti-mouse antibody directed against Villin ( 1∶75 , Beckman Coulter ) . Specimens were evaluated by using a confocal laser scanning microscope ( Leica DM IRBE ) in magnifications from 400 to 800 . | Impaired male fertility is a prevalent condition in many species and is often explained by aberrant semen quality . In some cases , male fertility is severely compromised although semen quality is without any apparent pathological findings ( i . e . , idiopathic male subfertility ) . The genetic mechanisms underlying idiopathic male subfertility often remain unexplained . In the present paper , we report a recessively inherited variant of idiopathic male subfertility in a cattle population . We use 650 , 000 genome-wide SNP markers genotyped in >7900 artificial insemination bulls to pinpoint the underlying genomic region . We take advantage of whole-genome re-sequencing data of 43 animals to identify a causal loss-of-function mutation in TMEM95 encoding a nondescript transmembrane protein . We demonstrate that transmembrane protein 95 is located at the plasma membrane of spermatozoa of fertile animals whereas it is absent in spermatozoa of subfertile animals . Our results indicate that integrity of transmembrane protein 95 is required for an undisturbed fertilisation . This is the first report to reveal a phenotypic effect associated with genomic variation in TMEM95 in any organism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome-wide",
"association",
"studies",
"animal",
"genetics",
"population",
"genetics",
"animal",
"breeding",
"gene",
"function",
"genome",
"sequencing",
"genome",
"analysis",
"tools",
"veterinary",
"diagnostics",
"trait",
"locus",
"analysis",
"molecular",
"genetics",
... | 2014 | A Nonsense Mutation in TMEM95 Encoding a Nondescript Transmembrane Protein Causes Idiopathic Male Subfertility in Cattle |
The cellular concentration of Bcl-xL is among the most important determinants of treatment response and overall prognosis in a broad range of tumors as well as an important determinant of the cellular response to several forms of tissue injury . We and others have previously shown that human Bcl-xL undergoes deamidation at two asparaginyl residues and that DNA-damaging antineoplastic agents as well as other stimuli can increase the rate of deamidation . Deamidation results in the replacement of asparginyl residues with aspartyl or isoaspartyl residues . Thus deamidation , like phosphorylation , introduces a negative charge into proteins . Here we show that the level of human Bcl-xL is constantly modulated by deamidation because deamidation , like phosphorylation in other proteins , activates a conditional PEST sequence to target Bcl-xL for degradation . Additionally , we show that degradation of deamidated Bcl-xL is mediated at least in part by calpain . Notably , we present sequence and biochemical data that suggest that deamidation has been conserved from the simplest extant metazoans through the human form of Bcl-xL , underscoring its importance in Bcl-xL regulation . Our findings strongly suggest that deamidation-regulated Bcl-xL degradation is an important component of the cellular rheostat that determines susceptibility to DNA-damaging agents and other death stimuli .
The Bcl-2 proteins are grouped into those that promote cell survival and those that promote programmed cell death [1] . It is thought that the balance of activity of these two groups of proteins serves as a rheostat that determines whether the cell lives or dies [2] . The activity of the prosurvival Bcl-2 proteins is normally dominant in a cell . Most antineoplastic agents and other proapoptotic agents induce changes in Bcl-2 proteins that tip the balance towards the prodeath activity [3] . Importantly , this may involve a decrease in the activity of prosurvival proteins , an increase in the activity of prodeath proteins , or a combination of both . There is substantial evidence that the level of the prosurvival Bcl-2 family protein Bcl-xL is one of the most important cellular determinants of patient outcome in a broad range of tumors . For example , increased Bcl-xL expression portends a worse prognosis in pancreatic cancer [4] , thyroid cancer [5] , follicular lymphoma [6] , ovarian cancer [7] , [8] , hepatocellular carcinoma [9] , and prostate cancer [10] and it has been specifically shown that increased levels of Bcl-xL correlate with treatment failure in thyroid cancer [5] , ovarian cancer [8] , and oropharyngeal cancer [11] . In support of a functional role for Bcl-xL in determining the prognosis and treatment response of patients with these cancers are the findings that ( i ) there is a “striking” correlation between resistance to treatment with a panel of 122 chemotherapeutic agents and Bcl-xL expression levels when assessed in 60 different types of tumor cells [12]; ( ii ) overexpression of Bcl-xL confers a multidrug resistance phenotype to tumor cells [13]; ( iii ) a small molecule or antisense that selectively inhibits Bcl-xL increases sensitivity to chemotherapy in vivo [14] , [15]; ( iv ) at least in some cells , there is a bcl-x gene-dosage effect for resistance to DNA-damaging agents [16]; and ( v ) increased Bcl-xL expression increases susceptibility to carcinogen-induced tumor formation in mice [17] . When considered together , these findings suggest that tumor cell Bcl-xL levels have an important functional role in determining patient outcome . The expression level of Bcl-xL is also important in determining the extent of damage in certain forms of tissue injury; in fact , Bcl-xL levels may be upregulated to protect against certain forms of injury . For example , liver cells with decreased Bcl-xL levels demonstrate increased susceptibility to injury [17] , [18]; conversely , transgenic overexpression of Bcl-xL protects against liver injury [19] . In this context , it is intriguing that hepatic Bcl-xL expression is upregulated in response to liver injury [20] , [21] . Similarly , Bcl-xL levels are upregulated in the esophageal mucosa in response to chronic acid reflux [22] . It is likely that the increased Bcl-xL in these and other instances protects against tissue injury . The findings outlined above underscore the importance of understanding the mechanisms by which Bcl-xL levels are regulated . We and others have previously shown that two asparagines in human Bcl-xL undergo deamidation to aspartyl or isoaspartyl residues and that the rate of deamidation of these asparagines increases in susceptible tumor cells that are treated with DNA-damaging agents [16] , [23] , [24] . We now present evidence that asparagine deamidation has been conserved in Bcl-xL-like proteins from the simplest extant metazoans through the human form of Bcl-xL . This extent of conservation suggests that deamidation has a critical role as a regulatory posttranslational modification of Bcl-xL . Indeed , we demonstrate here that the rate of deamidation dynamically modulates the cellular level of Bcl-xL because deamidation is a continuous but regulated process that , like phosphorylation in other proteins , activates a conditional PEST sequence to target Bcl-xL for degradation . Importantly , we show that in susceptible tumor cells , DNA-damaging agents decrease Bcl-xL levels , which increases cellular susceptibility to death signaling , because these agents induce an increase in the rate of deamidation of Bcl-xL and , consequently , an increase in the rate of degradation of Bcl-xL . In contrast , however , we have previously shown that at least in some nontransformed cells the increased rate of Bcl-xL deamidation and consequent degradation that would otherwise occur upon treatment with DNA-damaging antineoplastic agents is suppressed by p53-retinoblastoma protein ( pRb ) signaling; hence , Bcl-xL levels remain static in these cells when they are treated with DNA-damaging antineoplastic agents [16] . Therefore , Bcl-xL deamidation is a regulatable process and certain stimuli can shift the balance of cellular prosurvival and prodeath activity by altering the rate of Bcl-xL deamidation .
Asparagine deamidation is a nonenzymatic posttranslational modification . Although asparagine deamidation occurs spontaneously , its rate can be regulated by changes in the pH , ionic composition , or temperature of the surrounding cellular microenvironment [25] . An asparagine is most susceptible to deamidation when it is immediately followed by a glycine in a conformationally flexible region of a protein because deamidation is initiated when the peptide bond nitrogen of the N+1 amino acid attacks the carbonyl carbon of the asparagine side chain—this is facilitated by the reduced steric hindrance of glycine and flexibility of the surrounding sequence [25] . Human Bcl-xL has a large conformationally flexible region between the BH4 and BH3 domains that is referred to as its flexible loop ( Figure 1A ) [26] and we have previously demonstrated that two asparagines that are immediatedly followed by glycines that lie within the flexible loop undergo deamidation [16] . We now report that Bcl-xL-like proteins from sponge through human contain asparagine-glycine sequences within a region that is predicted to be conformationally flexible between the BH4 and BH3 domains ( Figure 1B ) [27] , [28] . The widespread presence of these asparagine-glycine sequences in the flexible region is striking as there is no other obvious sequence similarity within this region across all species ( Figure 1B ) , and it suggests that the presence of an asparagine-glycine sequence per se is a conserved feature of the flexible loop of Bcl-xL . Additionally , there are a number of species that express Bcl-xL–like proteins that have a long flexible region immediately upstream of the BH4 domain ( Figure 2A ) , and we found that each of these proteins contains an asparagine-glycine sequence in this region ( Figure 2B ) , suggesting that an asparagine-glycine sequence within a flexible region is a conserved feature of Bcl-xL . To objectively assess whether asparagine-glycine sequences are indeed a conserved feature of the Bcl-xL flexible loop , we performed a de novo analysis of an independently assembled group of Bcl-xL-like proteins , the Bcl-xL homology group of the Bcl-2 family database[29] , using the MEME conserved motif discovery algorithm[30] . MEME is a widely used tool that searches for sequences that are reiterant within an input group of proteins and assigns each an E-value , a statistical estimate of the probability that the sequence would occur with an equal or greater frequency than it occurs in the input group of proteins if the amino acids of the proteins were positionally randomized[31] . Sequences with E-values of less than 1×1−2 likely represent conserved , and therefore functional , motifs[32] . In a dataset consisting of the sequences of the entire region between the BH4 and BH3 domains from all of the members of the Bcl-xL homology group in the Bcl-2 database ( Figure S1 ) [29] , the asparagine-glycine dipeptide occurs in more sequences and with a greater frequency than any other dipeptide and it is assigned an E-value of 2 . 0×10−3 by MEME . Furthermore , almost half ( 91/187 ) of the asparagines in the dataset are a component of an asparagine-glycine sequence . These findings strongly suggest that there is selective pressure to maintain the asparagine-glycine sequence in this region . This implies that deamidation is a conserved feature of the Bcl-xL flexible loop because , to our knowledge , the only function asparagine-glycine dipeptides could have in this context is to serve as deamidation sites . We examined this further by determining if deamidation occurs within three flexible loops in which the asparagine-glycine sequences are surrounded by widely disparate sequences: the flexible loops in the human , Xenopus laevis , and zebrafish forms of Bcl-xL ( Figure 3A ) . We have previously demonstrated that two asparagines in the flexible loop of human Bcl-xL undergo deamidation[16] . Deamidation is readily detected in human Bcl-xL because the deamidated forms migrate more slowly than the native form during SDS-PAGE[16] . The more slowly migrating forms do not develop—that is , deamidation does not occur , if the susceptible asparagines are mutated to alanines to block deamidation . Conversely , Bcl-xL in which these asparagines are mutated to aspartates to mimic deamidation migrates at the same rate as the more slowly migrating , deamidated forms of wild-type Bcl-xL[16] . Additionally , deamidation , and therefore , the development of the more slowly migrating forms , can be further induced by incubating Bcl-xL at an alkaline pH in vitro[16] . We assessed the asparagine-glycine sequences in the flexible loops of Bcl-xL from Xenopus laevis ( Xenopus Bcl-xL ) and zebrafish for deamidation using the approach outlined above . Wild-type Xenopus Bcl-xL , which contains a single asparagine-glycine sequence in its flexible loop ( Figure 3A ) , forms a doublet when it is expressed in mammalian cells and evaluated by SDS-PAGE ( Figure 3B ) , but only the upper band forms during SDS-PAGE when the cell extract containing the wild-type Xenopus Bcl-xL is first incubated at an alkaline pH ( Figure 3B ) ; wild-type zebrafish Bcl-xL , which contains three asparagine-glycine sequences in its flexible loop ( Figure 3A ) , forms multiple bands ( Figure 3B ) , and there is a relative decrease in the most rapidly migrating band with a concomitant increase in the more slowly migrating bands when it is incubated at an alkaline pH ( Figure 3B ) . Mutation of the asparagines of the asparagine-glycine sequences to alanines to block deamidation in Xenopus Bcl-xL and zebrafish Bcl-xL [Xenopus Bcl-xL ( N37A ) and zebrafish Bcl-xL ( 3N/3A ) , respectively] blocks the formation of the more slowly migrating bands ( Figure 3B ) ; mutation of the asparagines to aspartates to mimic deamidation yields forms of Bcl-xL [Xenopus Bcl-xL ( N37D ) and zebrafish Bcl-xL ( 3N/3D ) , respectively] that migrate with the upper bands of their respective wild-type proteins ( Figure 3B ) . Finally , the mutant forms of Xenopus and zebrafish Bcl-xL are unaffected when incubated in an alkaline buffer ( Figure 3B ) ( we note that there is a protein band that is most readily visualized in the lanes of the two mutant forms of zebrafish Bcl-xL that migrates at an intermediate rate and appears to be unaffected by incubation in an alkaline buffer—the nature of this band is unknown ) . These findings demonstrate that the human , Xenopus laevis , and zebrafish forms of Bcl-xL all have the potential to undergo deamidation . That deamidation could occur at asparagine-glycine sequences in flexible loops with such disparate sequences as those in human , Xenopus laevis , and zebrafish Bcl-xL-like proteins is consistent with the finding that asparagines that are followed by glycines in flexible regions of proteins are exquisitely labile to deamidation [25] , [33] and the finding in model peptides that the deamidation rate is determined primarily by the amino acid that immediately follows the asparagine with the amino acids in surrounding positions having little or no effect [33] . When considered in this context , our findings suggest that deamidation could occur at the asparagine-glycine sequences in the flexible loops of Bcl-xL-like proteins irrespective of the immediate surrounding sequence . Therefore , our findings suggest that deamidation is a feature of the flexible loop of Bcl-xL-like proteins across a wide range of species . We next wanted to determine if Bcl-xL deamidation occurs in nonmammalian cells . When expressed in Drosophila Schneider 2 cells and analyzed by SDS-PAGE , we found that wild-type human Bcl-xL forms a doublet ( Figure 3C ) . The lower band of the doublet migrated with a mutant form of human Bcl-xL in which the deamidation is blocked by replacement of the asparagines with alanines [Bcl-xL ( N52A/N66A ) ] [16] , while the upper band of the doublet migrated with a mutant human Bcl-xL construct in which the susceptible asparagines are replaced with aspartates to generate a constitutively deamidated form of Bcl-xL [Bcl-xL ( N52D/N66D ) ] ( Figure 3C ) [16] . Additionally , when the Drosophila Schneider 2 cell lysates were incubated at an alkaline pH prior to SDS-PAGE , the wild-type Bcl-xL migrated in the position of the upper band of the doublet , while the mutant forms were unaffected ( Figure 3C ) . These findings suggest that the lower band of the doublet found in intact Drosophila Schneider 2 cells is the native form of human Bcl-xL and the upper band of the doublet is deamidated Bcl-xL . That Bcl-xL deamidation occurs in both insect and human cells strongly suggests that deamidation of Bcl-xL–like proteins can occur in a wide range of species . The rate of Bcl-xL deamidation is increased in response to treatment with DNA-damaging agents , such as cisplatin , etoposide , and γ-radiation , in several types of tumor cells [16] . We and others have found that a form of Bcl-xL in which deamidation is blocked affords tumor cells increased resistance to these agents when compared to the effect of wild-type Bcl-xL [16] , [23] , [24] . Additionally , Zhao and coworkers found that the suppression of Bcl-xL deamidation by an oncogenic tyrosine kinase contributes to etoposide and γ-radiation resistance in a mouse tumor model [23] and in human myeloproliferative disorders [34] , and there is evidence that suppression of Bcl-xL deamidation is a component of hepatocellular carcinogenesis [35] . These findings suggested that deamidation decreases cellular Bcl-xL prosurvival activity . We originally reported that deamidation decreases the prosurvival activity of Bcl-xL by disrupting its ability to sequester prodeath Bcl-2 family members such as Bim in vivo [16]; however , we subsequently found that our conclusion was based on artifactual results ( please see erratum , reference [36] ) . Surprisingly , though , another group has since published that deamidation does indeed disrupt the ability of Bcl-xL to sequester Bim both in vivo and when in solution in vitro [23] , [37] . Their in vitro findings were particularly surprising because ( i ) the deamidation sites are positioned near the center of the large unstructured region of Bcl-xL; ( ii ) the unstructured region is not necessary for the interaction with Bim or for the antiapoptotic activity of Bcl-xL [38]; ( iii ) the unstructured region remains unstructured in the deamidated form of Bcl-xL [39]; and ( iv ) the native and deamidated forms of Bcl-xL “adopt an essentially identical backbone structure” in solution [39] . Therefore , we reexamined the effect of deamidation on the ability of Bcl-xL to bind Bim using several different approaches and controls . We found that deamidation has no effect on the ability of Bcl-xL to bind Bim or Bax ( Text S1 and Figure S2 ) . This is consistent with the finding that both the native and deamidated forms of Bcl-xL bind equally to PGAM5 , a protein that has been implicated in oxidative stress-induced apoptosis [40] . Furthermore , Bcl-xL encodes several additional presumed prosurvival activities , such as the ability to bind to p53 [41]–[43] and the ability to regulate mitochondrial membrane permeability by forming an ion channel [44]–[46] . It seemed unlikely that deamidation within the unstructured loop could directly inactivate all of these functions . Therefore , we sought the mechanism by which deamidation decreases cellular Bcl-xL prosurvival activity . We noted that the levels of endogenous Bcl-xL decreased as deamidation increased in several of our previous experiments ( e . g . , figure 2 and figure 6 in reference [16] ) and a correlation between deamidation and decreased Bcl-xL levels in maturing erythrocytes was noted by Koury and coworkers [47] . Furthermore , in cells in which apoptosis was induced by oxidative damage , a fragment of Bcl-xL , but not full-length Bcl-xL , was found to be bound to an enzyme that binds deamidated proteins [48] , which suggests that Bcl-xL is rapidly degraded upon deamidation . Therefore , we considered the possibility that deamidation decreases the cellular activities of Bcl-xL by targeting Bcl-xL for degradation . Indeed , we found a clear correlation between the DNA damage-induced increase in Bcl-xL deamidation and a decrease in Bcl-xL levels ( Figure 4A ) . To begin to determine if it is specifically the deamidated forms that are targeted for degradation , we first blocked synthesis of the native form of Bcl-xL using cycloheximide . We found that the level of the native Bcl-xL decreases first and then , once the native Bcl-xL is depleted , the level of deamidated Bcl-xL decreases ( Figure 4B ) . The simplest explanation for this finding is that the native Bcl-xL is constantly deamidated , even in cells that have not been treated with DNA damaging agents , and the deamidated forms are degraded . To confirm that the deamidated forms are specifically targeted for degradation , we compared the stability of wild-type Bcl-xL and a form of Bcl-xL in which deamidation is blocked because the susceptible asparagines are mutated to alanines , [Bcl-xL ( N52A/N66A ) ] [16] . We have previously shown that the signaling that increases the rate of Bcl-xL deamidation in cells that are treated with DNA damaging agents is suppressed in wild-type mouse embryo fibroblasts ( MEFs ) and that the suppression is dependent upon the activation of pRb by p53 signaling [16] . Therefore , to determine if deamidation targets Bcl-xL for degradation we reconstituted Bcl-xL expression in bcl-x−/−/p53−/− MEFs with either wild-type Bcl-xL or Bcl-xL ( N52A/N66A ) . Importantly , we expressed each protein using retroviral infection at a multiplicity of infection of <1 without polybrene treatment or centrifugation so that instead of overexpressing the Bcl-xL constructs at high levels , we approximated the level of Bcl-xL found in wild-type MEFs as closely as possible . After antibiotic selection for the infected cells , we treated the pooled cells with etoposide or cisplatin to induce increased deamidation of Bcl-xL . Whereas the level of wild-type Bcl-xL decreased progressively after etoposide or cisplatin treatment , the level of Bcl-xL ( N52A/N66A ) , the form of Bcl-xL in which deamidation is blocked , remained relatively constant ( Figure 4C ) . As would be expected , the cells expressing Bcl-xL ( N52A/N66A ) were more resistant to the apoptotic effects of etoposide and cisplatin than were the cells in which the wild-type Bcl-xL was expressed ( Figure 4D ) . These findings strongly suggest that deamidation mediates the inactivation of Bcl-xL prosurvival activity by mediating the degradation of Bcl-xL . Proteins that are subject to regulatory degradation often contain PEST sequences and the presence of PEST sequences is specific to such proteins—that is , PEST sequences are rarely found in long-lived cellular proteins[49] . PEST sequences are hydrophilic stretches of at least 12 amino acids that are enriched in prolines , glutamates , aspartates , serines , and threonines that are flanked by but do not contain histidines , arginines , or lysines [50] . The PESTfind algorithm identifies potential PEST sequences and assigns them a score that predicts the likelihood that they truly function as a degradation signal sequence [50] . A score above zero denotes a potential PEST sequence[50]; the higher the score , the more likely the sequence functions to target the protein for degradation . Whereas the most well characterized PEST sequence , the PEST sequence in IκBα , has a PESTfind score of 5 . 90[49] , human Bcl-xL contains a PEST sequence with a score of 10 . 79 , which suggested that we would find that the human Bcl-xL PEST sequence truly functions as a proteolytic signaling sequence . It is also notable that ( i ) the PEST sequence is conserved among all mammalian forms of Bcl-xL ( Figure 5 ) ; ( ii ) even though the sequences themselves differ considerably from the mammalian sequence , there are sequences that are identified by the PESTfind algorithm as potential PEST sequences in a similar position in the Bcl-xL-like proteins from a wide range of nonmammalian species ( Figure 5 ) —that is , suggesting that there is conservation of a specific function at this position even though the sequence is not conserved; and ( iii ) PEST sequences only occur infrequently and indeed , there are no other sequences with PESTfind scores greater than zero at any other position within any of the Bcl-xL-like proteins listed in Figure 1B . These findings argue strongly for the importance of a functional PEST sequence at this position . Importantly , a PEST sequence may either constitutively or conditionally target a protein for proteolysis [49] . Therefore , it was intriguing that the PEST sequences either encompass or are in close proximity to the deamidation sites ( Figure 5 ) . This was intriguing because phosphorylation within or in proximity to certain conditional PEST sequences increases proteolytic signaling [49] and the products of deamidation , aspartyl residues , can functionally mimic phosphorylated amino acids [51] . Similarly , because deamidation adds an aspartyl residue , it increases the hydrophilicity and , hence , the PESTfind score of the PEST sequence ( e . g . , the PESTfind score of human Bcl-xL increases from 10 . 79 to 13 . 40 upon deamidation ) , which suggested that deamidation increases the activity of the PEST sequence . Therefore , we assessed the possibility that , like phosphorylation in other proteins , deamidation activates the PEST sequence as a signal for the proteolysis of Bcl-xL . To test this , we generated a human Bcl-xL construct in which the three prolines of the PEST sequence are mutated to alanines [Bcl-xL ( 3P/3A ) ] to partially disrupt its activity . We found that the level of deamidated Bcl-xL ( 3P/3A ) relative to the native form is increased when compared with wild-type Bcl-xL in untreated cells and cells treated with etoposide ( Figure 6A ) and that this is due to increased stability of the deamidated forms ( Figure 6B ) . Furthermore , the cells expressing Bcl-xL ( 3P/3A ) were significantly more resistant to etoposide- and cisplatin-induced apoptosis than those expressing wild-type Bcl-xL ( Figure 6C ) . The simplest explanation for these findings is that the function of Bcl-xL deamidation is to increase the proteolytic targeting activity of the Bcl-xL PEST sequence . Importantly , in the experiment depicted in Figure 6A and in several of the experiments discussed below , the Bcl-xL constructs are overexpressed and they therefore prevent induction of the later phases of apoptosis . However , even the overexpressed Bcl-xL undergoes an increase in deamidation-regulated degradation upon treatment with DNA-damaging agent agents . This indicates that deamidation-regulated degradation of Bcl-xL is a function of changes that occur in the cell during the premitochondrial phase of apoptosis , the phase in which decreases in Bcl-xL would increase susceptibility to prodeath signaling . This finding and the conservation of the PEST sequence together provide strong evidence of the functional significance of the deamidation-regulated degradation of Bcl-xL as an integral component of the rheostat that regulates cell death . Bcl-xL is cleaved by calpain both in vitro and in vivo [52]–[54] , which is notable because PEST sequences can target proteins for calpain-mediated degradation [55]–[57] . Therefore , to begin to identify the protease ( s ) that mediate degradation of deamidated Bcl-xL , we treated HTB-9 and C33a cells with calpain inhibitor I and found that it causes primarily an increase of deamidated Bcl-xL in both ( Figure 7A ) . Additionally , the deamidated forms of Bcl-xL are increased by calpain inhibitor I when Bcl-xL deamidation is further induced by etoposide treatment ( Figure 7B ) . Importantly , the increase in the deamidated forms in the cells treated with calpain inhibitor I is due to an increase in stability as assessed by a pulse chase experiment ( Figure 7C ) , demonstrating that calpain inhibitor I increases Bcl-xL levels by blocking its degradation . Importantly , calpain inhibitor I inhibits several different proteases , not just calpain . In fact , calpain inibitor I also inhibits the proteasome , albeit at a higher concentration than that which is required to inhibit calpain , and PEST sequences can target proteins for proteasomal degradation . We therefore assessed a known proteasomal target , Mcl-1 , on the same blot in which we had examined the effect of calpain inhibitor I on Bcl-xL in HTB-9 cells . We also examined total cellular ubiquitinated proteins in the same cell lysates . Whereas 5 µM calpain inhibitor I had caused a near maximal increase in the level of the deamidated forms of Bcl-xL ( Figure 7A ) , Mcl-1 and total ubiquitinated proteins only reached near maximal levels when the cells were treated with 15–20 µM calpain inhibitor I ( Figure 7D ) . We also found that the specific proteasome inhibitor lactacystin had only a relatively small , if any , effect on Bcl-xL compared with its its effect on MCL-1 and total ubiquitinated proteins in HTB-9 cells ( Figure 7E ) . These findings suggest that the proteasomal activity does not have a signficant role in the degradation of the deamidated form of Bcl-xL . Bcl-xL has also been shown to be degraded by caspases [58] . However , we found that stable expression of a dominant negative form of caspase 9 had no effect on Bcl-xL degradation in response to etoposide treatment in SAOS-2 cells even though the dominant negative caspase 9 blocked activation of caspase 3 ( Figure 7F ) and apoptosis ( unpublished data ) . Expression of the retinoblastoma protein ( pRb ) , which blocks Bcl-xL deamidation [16] , was used as a control ( Figure 7F ) . That expression of the dominant negative caspase 9 fails to block degradation of deamidated Bcl-xL is consistent with the finding that overexpressed Bcl-xL is degraded even though its overexpression should block caspase activation . Together these findings demonstrate that caspase activity is not necessary for DNA damage-induced Bcl-xL degradation , at least in certain cell lines . Finally , to further examine the potential role of calpain in the degradation of deamidated Bcl-xL , we examined Bcl-xL in fibroblasts that lack calpain activity [59] . The Capn4 gene encodes the small subunit of calpain , which is necessary for all calpain activity . When Capn4−/− MEFs in which calpain activity was rescued by expression of the Capn4 gene were treated with cycloheximide , Bcl-xL decreased ( Figure 7G ) , as it does in other cells that have calpain activity when they are treated with cycloheximide ( e . g . , Figures 4B and 6B ) . However , Bcl-xL accumulated in its deamidated form in Capn4−/− MEFs when they were treated with cycloheximide . These findings are consistent with a role for calpain in the degradation of deamidated Bcl-xL . It is widely accepted that there is a rapid fall in cytosolic pH of ≈0 . 3–0 . 4 units that occurs in apoptosis upon mitochondrial outer membrane permeabilization [60] , [61]; however , several groups have reported that cytosolic alkalinization to as high as pH 8 . 0 occurs early in certain forms of apoptosis , including DNA damage-induced apoptosis [62]–[67] . This is notable because based on structural considerations , Bcl-xL is predicted to be exquisitely susceptible to nonenzymatic deamidation at pH 7 . 4 [16] , [68] and it has been demonstrated that the rate of Bcl-xL deamidation in reticulocyte lysates is increased significantly by increases in pH within the range of pH 7 . 0 to pH 8 . 0 [47] . These findings strongly suggested that DNA damage-induced Bcl-xL deamidation is regulated by changes in pH in the cell . Indeed , while this work was in progress , it was confirmed that the DNA damage-induced increase in Bcl-xL deamidation in lymphocytes is induced by the increase in cytosolic pH that occurs in response to DNA damage [37] and we have confirmed that this is also true in the cells of human solid tumors ( Text S2 and Figure S3 ) . Notably , the finding that the rate of deamidation is increased by increased pH is further evidence that the DNA damage-induced increase in deamidation of Bcl-xL occurs in the premitochondrial phase of apoptosis , because , as noted above , the onset of the postmitochondrial phase is characterized by a rapid acidification of the cytosol [60] , [61] , which would be expected to decrease the rate of deamidation of Bcl-xL . We previously reported that expression of pRb in SAOS-2 osteosarcoma cells blocks both the DNA damage-induced increase in Bcl-xL deamidation and apoptosis [16] . Indeed , we now report that expression of pRb decreases pH in these cells at baseline and after treatment with DNA-damaging agents ( Figure 8A ) . This strongly suggests that Rb blocks an increase in the rate of Bcl-xL deamidation by maintaining the cytoplasm at a relatively low pH after treatment with DNA-damaging agents . This is notable because we found that inhibition of Bcl-xL expression renders SAOS-2 cells that express pRb susceptible to DNA damage-induced apoptosis [16] . Together these findings strongly suggest that the increased rate of deamidation-regulated degradation of Bcl-xL is an important function of the increase in pH that occurs in response to treatment with DNA-damaging agents—that is , alkalinization is necessary to induce an increased rate of deamidation-regulated degradation of Bcl-xL , which in turn is necessary for apoptosis to occur , but alkalinization is not necessary if Bcl-xL is absent . We have also reported that the DNA damage-induced increase in Bcl-xL deamidation is suppressed in wild-type MEFs , but it occurs in p53−/− MEFs . This is notable because while pRb is activated by DNA damage in wild-type MEFs , it remains inactive after DNA damage occurs in p53−/− MEFs [16] . Therefore , we hypothesized that the activated pRb in the wild-type MEFs suppresses Bcl-xL deamidation . Consistent with this and our finding that pRb suppresses the alkalinization in SAOS-2 cells , we found that while p53−/− MEFs are susceptible to the DNA damage-induced alkalinization , wild-type MEFs are not ( Figure 8B ) . Finally , we found that even though bcl-x−/− MEFs , which have an intact p53-pRb signal transduction pathway , are exquisitely susceptible to apoptosis [16] , they do not exhibit a DNA damage-induced increase in cytosolic pH prior to undergoing apoptosis ( Figure 8C ) . This last finding is further evidence that the increased rate of deamidation-regulated degradation of Bcl-xL is an important target of the increase in pH that occurs in response to treatment with DNA-damaging agents in susceptible tumor cells .
Asparagine deamidation was long thought to be a purification artifact; however , in 1968 Flatmark provided the first demonstration that a protein undergoes deamidation within the cell [69] . It is now well accepted that many proteins undergo deamidation within the cell , but deamidation is still viewed nearly universally as a form of protein damage or aging that is detrimental to the organism . This is because deamidation has been thought by most to be an unregulated , spontaneous process that disrupts protein function through the nonspecific disruption of protein structure . Furthermore , whereas deamidation has been implicated in the dysfunction underlying several pathologic processes , such as Alzheimer's disease [70] and cataract formation [71] , there has been only limited evidence that it could serve a beneficial role [72] . We have now demonstrated that Bcl-xL deamidation is a process that activates a conditional PEST sequence . The degree of organization underlying both the regulation and functional consequence of Bcl-xL deamidation together with the fact that it is conserved across a wide range of species clearly suggests that deamidation can play a beneficial regulatory role . It is possible that the deamidation that occurs in Alzheimer's disease , cataract formation , and other pathologic processes represents a dysregulated state of a process that normally has an important cellular function . This would be analogous to the contribution of the dysregulation of the phosphorylation of certain proteins to tumorigenesis [73] . Indeed , there is evidence that the dysregulation of Bcl-xL deamidation contributes to the development of hepatocellular carcinoma [35] and myeloproliferative disorders [34] . Notably , in addition to pH , the rate of deamidation is affected by the buffer ion , tonicity , and temperature [74] . A change in any of these that results in a decrease in the rate of Bcl-xL deamidation would have the potential to increase tumor cell viability and inhibit the tumor cell response to treatment , worsening patient outcome . Additionally , we have shown that even modest changes in Bcl-xL levels can alter the extent of tissue damage in response to certain types of injury [17] . The finding that mutation of the PEST sequence or treatment with calpain inhibitor I in otherwise untreated cells results in a relative increase of the level of deamidated Bcl-xL demonstrates that Bcl-xL levels are continuously modulated by deamidation , even in normally growing cells . Therefore , any change in factors that affects the rate of deamidation could alter the extent of tissue damage in response to certain types of injury . Finally we note that asparagine deamidation is an extraordinarily simple posttranslational modification in that it only requires a water molecule to proceed . Its simplicity suggests that it was an early form of posttranslational modification . In this context , it is notable that asparagine is the evolutionary offspring of aspartate and it is thought that asparagine “captured” what were originally two aspartate codons to serve as its codons [75] . Thus we speculate that asparagines replaced certain aspartates as proteins evolved so that a residue with an inducible negative charge , asparagine , could replace a residue with a fixed negative charge , aspartate . This substitution would have afforded a greater degree of control of protein function . Indeed , it may have been the selective advantage of the potential to switch from a neutral residue to a charged residue that initially drove the stable incorporation of asparagine into proteins .
The protein sequences listed in the Bcl-xL ( BCL2L1 ) homology group of the Bcl-2 database ( Figure S1 ) [29] that contain both a BH4 and a BH3 domain were identified using the online Batch Search tool of the Conserved Domain Database[76] . The intervening sequences between the BH4 and BH3 domains in these proteins were compiled to form the dataset that was submitted to the MEME server for analysis . Importantly , there are species in the database that express more than one protein in which the sequence between the BH4 and BH3 domains are identical; such proteins are typically the result of alternative splicing . In such instances , the sequence was only included once in the analysis . SAOS-2 cells ( ATCC HTB-85 ) , C33a cells ( ATCC HTB-31 ) , and MEFs were maintained in DMEM with 10% FBS . HTB-9 cells ( ATCC HTB-9 ) were maintained in RPMI-1640 with 10% FBS . Drosophila Schneider 2 cells were maintained in Shields and Sang M3 ( Sigma ) with 10% FBS . bcl-x−/− , p53−/− , and bcl-x−/−/p53−/− MEFs and Rb-inducible SAOS-2 cells were described previously [16] . Capn4−/− MEFs were described previously [59] . Bcl-xL–inducible SAOS-2 cells were generated using the T-Rex system ( Invitrogen ) . Xenopus and zebrafish Bcl-xL cDNAs were amplified by RT-PCR using primers 5′-ATATATCCATGGCAGAGGGCAGCAGTAGAGATCTGGTGG-3′ and 5′-TATATACAGCTGTCGGCGCCTCATGTAGCAGACC-3′ with Xenopus mRNA and 5′-ATATATCCTGGCATCTTACTATAACCGAGAACTGGTGG-3′ and 5′-TATATACAGCTGCAGGCGTTTCTGTGCAATGAGTCCCCC-3′ with zebrafish mRNA ( the underlined sequences in the primers were used for cloning purposes ) . The products were cloned between the Nco I site and Pvu II site in the plasmid pTriEx-1 . 1 ( Novagen ) . The sequences of the inserts were confirmed as identical to the Xenopus Bcl-xL sequence ( Genbank accession no . NP001082147 ) and the zebrafish Bcl-xL sequence ( Genbank accession no . NP571882 ) listed in the NCBI databases . All mutations were made using the QuikChange Kit ( Stratagene ) . Xenopus Bcl-xL codon 37 was changed from AAT to GCT to generate Xenopus Bcl-xL ( N37A ) and codon 37 was changed from AAT to GAT to generate Xenopus Bcl-xL ( N37D ) . Zebrafish Bcl-xL codon 42 was changed from AAT to GCT , codon 54 from AAT to GCT , and codon 81 from AAT to GCT to generate zebrafish Bcl-xL ( 3N/3A ) , and codon 42 was changed from AAT to GAT , codon 54 from AAT to GAT , and codon 81 from AAT to GAT to generate zebrafish Bcl-xL ( 3N/3D ) . For expression of the wild-type and mutant forms of human Bcl-xL in Drosophila Schnieder 2 cells pCMA-Bcl-xL , pCMA-Bcl-xL ( N52A/N66A ) , and pCMA-Bcl-xL ( N52D/N66D ) were constructed by ligation of PstI Bcl-xL encoding fragments from pSFFV-Bcl-xL , pSFFV-Bcl-xL ( N52A/N66A ) , and pSFFV-Bcl-xL ( N52D/N66D ) [16] into the PstI site of pCMA [77] . Retroviral vectors for expression of wild-type and mutant forms of human Bcl-xL were generated as follows . The retroviral construct pBABE-blast and pBABE-blast-HA were generated by ClaI/HindIII digest of pBABE-puro and pBABE-puro-HA ( removes the puromycin resistance gene ) and blunt-end ligation of the blasticidin resistance gene with its promoter from pcDNA/TR ( Invitrogen ) into these sites . pBABE-blast-Bcl-xL , pBABE-blast-Bcl-xL ( N52A/N66A ) , and pBABE-blast-HA-Bcl-xL were constructed by ligation of the EcoRI Bcl-xL encoding fragments from pSFFV-Bcl-xL and pSFFV-Bcl-xL ( N52A/N66A ) [16] into the EcoRI site of pBABE-blast and pBABE-blast-HA . In pBABE-blast-Bcl-xL and pBABE-blast-HA-Bcl-xL , Bcl-xL codon 38 was changed from CCA to GCA , codon 48 from CCC to GCC , and codon 55 from CCA to GCA to generate pBABE-blast-Bcl-xL ( 3P/3A ) and pBABE-blast-HA-Bcl-xL ( 3P/3A ) . pCDNA3-Flag-dominant negative caspase 9 was described previously [16] . Retroviral particles were produced by transient transfection of Phoenix E cells with either pBABE-blast-Bcl-xL or pBABE-blast-Bcl-xL ( N52A/N66A ) . The pBABE-blast-Bcl-xL and pBABE-blast-Bcl-xL ( N52A/N66A ) supernatants were collected from the Phoenix E cell cultures and diluted 1∶5 in fresh media . The diluted retrovirus was added to the medium of the MEFs without polybrene or centrifugation . Twenty-four hours later , blasticidin ( 1 . 0 µg/ml ) was added to the media . After selection , 1×105 cells were plated on 60 mm dishes and treated 24 h later with 5 µM of etoposide . Standard retroviral techniques were used for assessment of the PEST sequence in MEFs . SAOS-2 , HTB-9 , and C33a cells were transfected using the calcium phosphate method . Drosophila Schneider 2 cells were transfected using nucleofector ( Amaxa ) . Survival was quantified by flow cytometry using the Live/Dead kit ( Molecular Probes ) or by microplate reader at 450 nm using the Cell Counting kit-8 ( Dojindo Molecular Technologies ) . The following antibodies were used: anti-Bcl-xL ( 610211 ) and anti-Bcl-2 ( 610538 ) from Transduction Laboratories; anti-Bcl-xL ( 2764 ) from Cell Signaling; anti-tubulin ( sc-9104 ) , anti-actin ( sc-1616 ) , anti-Mcl-1 ( sc-819 ) , and anti-Ubiquitin ( sc-8017 ) from Santa Cruz Biotechnology; anti-HSV-Tag ( 69171 ) from Novagen; and anti-HA ( 1867423 ) from Roche . Immunoblotting and immunoprecipitation were performed as previously described [16] . For immunoprecipitation , lysis buffer ( 50 mM HEPES ( pH 7 . 0 ) , 250 mM NaCl , 1 mM EDTA , 0 . 2% NP-40 , and Complete Protease Inhibitor ( Roche ) was used . Bcl-xL–inducible SAOS-2 cells were induced by doxycycline treatment and pulsed with 35S-methionine for 4 h . Cells treated with calpain inhibitor I as indicated and chased for the specified times . The cell lysates were prepared and immunoprecipitated for Bcl-xL as described previously [16] and then analyzed by SDS-PAGE and autoradiography . Cells were grown in a HEPES-buffered medium instead of the standard HCO3−/CO2 buffer system to avoid the rapid shifts in pH that occur when cells in the HCO3−/CO2 buffer system are removed from the 5% CO2 atmosphere of an incubator . Sixty hours after DNA-damaging agent treatment , cells were washed with PBS . The studies were purposefully biased towards the assessment of cells in the earlier stages of apoptosis by measuring the pH of only those cells that remained adherent to the tissue culture dish . These cells were loaded with 5 µM of SNARF-1 for 10 min and then washed with PBS just prior to assessment by flow cytometry . The SNARF-1 was excited at 488 nm and emissions were read at 585 nm ( pH-dependent ) and 640 nm ( pH-independent ) . The pH-independent emission allows for the normalization of SNARF-1 loading differences between cells . The ratio of the emissions was calculated and the pH was read from a calibration curve . An in situ calibration curve was generated as follows: Cells are loaded with SNARF-1 as above . HEPES was used to make buffers at 0 . 5 pH unit intervals ranging from pH 6 . 5–8 . 5 . These contained the ionophore nigericin 13 µM and K+ 140 mM to render the cells permeable to the buffers . Cells were equilibrated in the buffers for 20 min . The cells were then analyzed by flow cytometry , and a calibration curve of the pH versus the ratio of the pH-dependent and pH-independent emissions was plotted . | Cellular levels of the pro-survival protein Bcl-xL are an important determinant of cellular susceptibility to many death stimuli , including most cancer therapies . We previously showed that human Bcl-xL undergoes deamidation – the conversion of two neutral asparaginyl side-chains into negatively charged aspartyl side-chains – a process that occurs spontaneously but is accelerated by the treatment of tumor cells with DNA-damaging agents . Here , we show that deamidation activates a hitherto undetected signal sequence within Bcl-xL that targets it for degradation by a pathway involving the proteolytic enzyme calpain . This increased degradation of Bcl-xL , and the consequent enhanced cellular susceptibility to programmed cell death , may contribute to the ability of DNA-damaging agents to kill tumors . We also demonstrate that deamidation of Bcl-xL has likely been conserved from the simplest metazoans to humans , underscoring the importance of deamidation in the regulation of Bcl-xL . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biology"
] | 2013 | Control of Cellular Bcl-xL Levels by Deamidation-Regulated Degradation |
Herpes simplex virus type 1 ( HSV-1 ) and HSV-2 are highly prevalent viruses that cause a variety of diseases , from cold sores to encephalitis . Both viruses establish latency in peripheral neurons but the molecular mechanisms facilitating the infection of neurons are not fully understood . Using surface plasmon resonance and crosslinking assays , we show that glycoprotein G ( gG ) from HSV-2 , known to modulate immune mediators ( chemokines ) , also interacts with neurotrophic factors , with high affinity . In our experimental model , HSV-2 secreted gG ( SgG2 ) increases nerve growth factor ( NGF ) -dependent axonal growth of sympathetic neurons ex vivo , and modifies tropomyosin related kinase ( Trk ) A-mediated signaling . SgG2 alters TrkA recruitment to lipid rafts and decreases TrkA internalization . We could show , with microfluidic devices , that SgG2 reduced NGF-induced TrkA retrograde transport . In vivo , both HSV-2 infection and SgG2 expression in mouse hindpaw epidermis enhance axonal growth modifying the termination zone of the NGF-dependent peptidergic free nerve endings . This constitutes , to our knowledge , the discovery of the first viral protein that modulates neurotrophins , an activity that may facilitate HSV-2 infection of neurons . This dual function of the chemokine-binding protein SgG2 uncovers a novel strategy developed by HSV-2 to modulate factors from both the immune and nervous systems .
Herpes simplex virus type 1 and 2 ( HSV-1 and HSV-2 , respectively ) are highly prevalent , neurotropic human pathogens [1] . Initial infection occurs in epithelial cells , generally within the skin and the mucosa of the oral tract and genitalia [1] . Then , HSV reaches and infects free nerve endings ( FNE ) of sensory neurons and colonizes ganglia of the Peripheral Nervous System ( PNS ) . The mechanism ( s ) facilitating HSV neurotropism , which is crucial for latency and pathogenesis , are not well understood . Since herpesviruses are highly adapted pathogens that modify several aspects of both the immune and nervous systems , it is conceivable that they may modulate factors influencing neuronal functions to gain access to the nervous system . Several axonal guidance cues and neurotrophic factors involved in neural targeting have been identified [2] . Among them , neurotrophins are a family of secreted proteins that play relevant roles in neuronal survival , axonal growth and guidance in the PNS . Members of this family include nerve growth factor ( NGF ) , brain-derived neurotrophic factor ( BDNF ) , neurotrophin 3 ( NT3 ) and NT4/5 [3] . Each neurotrophin binds with high affinity and activates tyrosine kinase receptors known as Trks . NGF binds TrkA , BDNF and NT4/5 bind TrkB , and NT3 binds TrkC . Moreover , NT3 can also bind TrkA and TrkB , although with lower affinity [3] . Both mature neurotrophins and immature precursors ( proneurotrophins ) also bind p75 neurotrophin receptor ( p75NTR ) , a member of the tumor necrosis factor ( TNF ) receptor superfamily . p75NTR has multiple and diverse functions [4] . Another important family of neurotrophic factors is the glial cell line-derived neurotrophic factors ( GDNF ) family ligands ( GFLs ) formed by GDNF and artemin among others . GFLs interact with co-receptors of the GDNF Family Receptor α ( GFRα ) protein family , allowing the activation of the tyrosine kinase receptor RET ( rearranged during transfection ) [5] . Peripheral neurons innervating skin and mucosa show a strong dependency on neurotrophic factors both ex vivo and in vivo [6 , 7] . In order to colonize the PNS , HSV must reach FNE , dynamic structures capable of degeneration and regeneration [8] in response to neurotrophic factors [6 , 7] . The possible relevance of neurotrophic factors in the initial steps of HSV infection in neurons is not completely understood . We hypothesized that HSV could modify nerve ending navigational cues during the early steps of PNS colonization . HSV glycoprotein G ( gG ) is the least conserved of the glycoproteins shared by HSV-1 and HSV-2 [9] . HSV-1 gG ( gG1 ) and HSV-2 gG ( gG2 ) have a similar C-terminal domain present at the virion and at the surface of the infected cells , composed by an extracellular proline-rich domain , a transmembrane region and a short cytoplasmic tail . The N-terminal domain of HSV-2 gG , but not that of gG1 , is proteolytically cleaved and secreted ( termed here SgG2 ) during infection [10 , 11] . Whether these differences between gG1 and gG2 have any functional consequence is unknown . Recombinant soluble gG1 ( SgG1 ) and SgG2 bind chemokines with high affinity enhancing chemokine function [12] , in sharp contrast to all previously described viral chemokine-binding proteins ( vCKBPs ) that inhibit chemotaxis . Since neurotrophic factors are also secreted proteins regulating many aspects of peripheral neurons , and have been previously involved in immune related-functions [13] , we investigated whether HSV gG could bind neurotrophic factors modifying their activity . Here we show that SgG2 interacts with several neurotrophic factors . SgG2 , but not SgG1 or M3 , a viral chemokine binding protein from murine gamma herpesvirus 68 ( MHV-68 ) [14] , transiently enhances NGF-dependent axonal growth of superior cervical ganglion ( SCG ) neurons ex vivo . The molecular mechanism beneath this enhancement involves modulation of the NGF receptor TrkA . SgG2 alters TrkA localization in lipid rafts , and NGF-dependent TrkA internalization , signaling and retrograde transport . In vivo , both infection with HSV-2 and expression of SgG2 in the external layers of the epidermis modifies the termination zone of the TrkA dependent FNE . Our data indicate that SgG2 may reconfigure neurotrophin signaling during HSV-2 primary infection to attract specific terminal axons to the infection site and may facilitate neural invasion .
To test whether gG from HSV-1 or HSV-2 interacts with neurotrophic factors we performed surface plasmon resonance ( SPR ) assays . As a control we used another vCKBP from MHV-68 , M3 . Recombinant SgG2 interacted with members of the neurotrophin family such as NGF and those of the GFLs family , like artemin or GDNF ( Fig . 1A and S1 Table ) . SgG1 and M3 bound neurotrophic factors but saturable binding was not demonstrated in these SPR experiments and therefore non-specific binding cannot be excluded ( S1A Fig . and S1 Table ) . However , binding specificity was suggested by the lack of interaction of the vCKBPs tested with interferon ( IFN ) -α , TNF-α , and interleukin ( IL ) -1 , and the negative binding of M3 to artemin ( Fig . 1B , S1B Fig . and S1 Table ) . Saturation experiments using SPR further demonstrated the specific , high affinity interaction of SgG2 and NGF ( Fig . 1 D and S1 Table ) . We could also show binding of SgG1 and SgG2 to NGF coupled in a chip ( S1C Fig . and Fig . 1C , respectively ) , with a similar affinity ( KD 1 . 4 x 10−8 M ) to that calculated for NGF interacting with SgG2 coupled in a chip based on kinetics analysis ( not shown ) , supporting the specificity of the interaction . HSV-2 glycoprotein D ( gD ) was used as a negative control for NGF binding ( Fig . 1C ) . We confirmed binding of these vCKBPs to NGF by crosslinking assays using iodinated rat NGF ( [125I]-rNGF ) and soluble , recombinant M3 , SgG1 and SgG2 ( S1D Fig . and S1 Text ) . The formation of the vCKBP-NGF complex was competitively inhibited with increasing concentrations of unlabeled NGF ( S1E , F Fig . ) . Addition of increasing amounts of cold NGF resulted in the formation of higher molecular weight bands that probably correspond to NGF oligomers since they appear also in the absence of viral proteins . The formation of NGF oligomers at high concentration may affect the pattern of binding to the viral proteins in our SPR assays . Since HSV-1 and HSV-2 gG also bind chemokines , we tested whether NGF and chemokine binding takes place through different regions or whether the binding sites overlap and the interaction of NGF may be competitively inhibited with chemokines . We injected CXCL12β or NGF alone or in combination with NGF in a chip containing SgG2 . The binding detected after simultaneous injection of chemokine and NGF nearly corresponded to the sum of the binding obtained when the chemokine and NGF were injected independently ( Fig . 1D ) . Furthermore , injection of one of the analytes ( NGF or CXCL12β ) in an SgG2-coupled chip preincubated with the other analyte ( CXCL12β or NGF ) caused no displacement of the prebound analyte ( not shown ) . It is important to note that the affinity of the interaction of SgG2 with CXCL12β ( 2 . 2x10−9 M ) [12] is higher than that determined for NGF ( 2 . 2x10−8 M ) . We addressed whether the vCKBP could affect the function of neurotrophic factors . FNE of sensory neurons are the main targets of HSV-1 and HSV-2 in the skin and mucosa . We focused on NGF and artemin due to their relevance in epidermal homeostasis and innervation [6 , 7] . In order to have a feasible and robust model we used mouse sympathetic neurons from SCG that express high levels of NGF and artemin receptors , and depend on NGF and artemin both in vivo and in culture [15] . We cultured SCG as 3D explants using collagen matrix containing recombinant viral proteins , and we measured the area comprised by axons normalizing to the ganglion perimeter . In the presence of NGF , SgG2 enhanced axonal growth of SCG neurons compared to HEPES control ( Fig . 2A , middle panels ) . On the contrary , M3 ( Fig . 2A , middle panels ) did not induce such increase . No changes in axonal growth induced by SgG2 or M3 were detected in the absence of trophic factors ( Fig . 2A , upper panels ) , or when artemin was used ( Fig . 2A , lower panels ) . Next , we addressed whether SgG1 had the same effect on NGF as SgG2 . SgG1 did not significantly increase axonal growth of SCG neurons 24 h post-incubation ( Fig . 2B ) . SgG2 enhancement of NGF-mediated axonal growth was also observed when using dissociated SCG neurons ( Fig . 2C ) . To further characterize this effect and to discriminate between axonal growth and directionality , transfected HEK-293T cells were co-cultured with SCG in 3D explants with NGF , and the proximal/distal ( P/D ) ratio was analyzed . HEK-293T cells express endogenous axonal repulsive cues like semaphorin 3A [16] and several class-A ephrins [17] . Accordingly , using low concentrations of NGF ( 0 . 25 nM ) , control transfected HEK-293T cells induced repulsion of SCG axons ( Fig . 2D ) . Such repulsion was also observed when HEK-293T cells were transfected with a V5-tagged M3-expressing plasmid ( V5-M3 ) . On the contrary , the expression of V5-SgG2 significantly reduced the repulsion of SCG axons ( Fig . 2D middle panel ) . Viral protein expression was detected by Western blot ( Fig . 2D ) . Altogether , these data showed that SgG2 specifically increases axonal growth of SCG neurons in an NGF-dependent manner . We addressed whether the increase in NGF-dependent axonal growth mediated by SgG2 was related to changes in TrkA . Since SgG2 bound NGF and modulated its function we decided to address whether it could interact with TrkA . We incubated dissociated SCG neurons with NGF in the presence or absence of SgG2 . Following immunoprecipitation of TrkA , SgG2 was detected only when NGF was present , indicating that SgG2 and TrkA belong to the same complex in the presence of NGF and the interaction of SgG2 with NGF is required ( Fig . 3A ) . Then we focused on TrkA signaling . Dissociated cultures of SCG neurons were incubated with NGF , M3 or SgG2 alone or NGF together with the viral proteins , and the downstream signaling was analyzed by Western blotting . Incubation with SgG2 or M3 in the absence of NGF did not induce TrkA phosphorylation or activation of downstream pathways ( Fig . 3B–F ) . As expected , NGF promoted TrkA phosphorylation [3] . SgG2 significantly increased NGF-dependent TrkA phosphorylation at tyrosine 490 ( Tyr490 ) at 15 min , but not at 120 min post-incubation , when compared to control and M3 conditions ( Fig . 3B , C ) . Extracellular signal-regulated kinases ( ERK ) 1/2 activation in response to NGF , but not that of AKT ( also known as protein kinase B ) , was significantly increased in the presence of SgG2 both at 15 and 120 min post-incubation ( Fig . 3B , D , E ) . We also tested the activation status of cofilin , an actin-severing protein responsible of actin turnover that is inactivated by phosphorylation [18] . We detected higher levels of cofilin phosphorylation induced by SgG2-NGF 120 min post-NGF stimulation when compared to NGF alone ( Fig . 3B , F ) . At the cell surface there are different and segregated types of rafts depending on their lipid and glycosphingolipid composition , leading to diverse biological functions [19] . A fraction of TrkA accumulates in GM1 lipid rafts , and TrkA appears to signal through GM1 rafts [20 , 21] . We addressed whether SgG2 could affect TrkA recruitment to GM1-enriched lipid rafts in SCG dissociated neurons . In the absence of NGF , a small fraction of TrkA co-localized with GM1 staining ( Fig . 4A , top row ) . As predicted , NGF stimulation maintained and increased TrkA co-localization with GM1 staining 2 min post NGF stimulation ( Fig . 4A , third row ) . SgG2 disrupted NGF-dependent and-independent TrkA localization in GM1 rafts at 2 ( Fig . 4A , fourth row ) and 10 min post-stimulation ( Fig . 4B , fourth row ) . We hypothesized that SgG2 could divert TrkA to other types of rafts such as GM3 rafts [19] . Stimulation with NGF reduced the localization of TrkA in GM3-rich rafts ( Fig . 4C , D , third row ) concomitantly with NGF-dependent TrkA increase in GM1 rafts , when compared to HEPES treatment . However , addition of SgG2 retained TrkA within GM3 rafts at 2 min post-incubation ( Fig . 4C , fourth row ) and even more at 10 min post-incubation ( Fig . 4D , fourth row ) . Overall , our results indicated that SgG2 recruits TrkA to GM3-enriched lipid rafts , thereby altering the normal NGF-TrkA signaling that usually occurs within GM1 rafts . The NGF receptor p75NTR is present in GM1 rafts [20] and interacts with TrkA in an NGF-dependent manner [3 , 22] . We addressed the effect of SgG2 on the interaction between TrkA and p75NTR . As shown in S2 Fig . , SgG2 disrupts the NGF-induced TrkA-p75NTR interaction in agreement with TrkA raft relocation mediated by SgG2 , whereas SgG1 did not . As TrkA recruitment to GM3 appeared to be independent of the presence of NGF in the culture we wanted to determine whether SgG2 alone was sufficient to induce changes in TrkA signaling . To explore this , we preincubated dissociated SCG neurons with HEPES or with SgG2 for 10 min followed by NGF stimulation in the presence or absence of SgG2 . Preincubation with SgG2 did not affect NGF dependent TrkA or ERK phosphorylation ( S3A–C Fig . ) and , in agreement with Fig . 3B , changes on signaling were detected only when SgG2 was added together with NGF . These results indicated that SgG2 induces TrkA raft relocalization that may be necessary , but not sufficient , to enhance NGF-TrkA signaling . Careful examination of the results obtained following 10 min of NGF incubation ( Fig . 4B , D ) suggested a possible effect of SgG2 at the level of plasma membrane TrkA . To confirm this observation we analyzed the internalization rate of TrkA in response to NGF and SgG2 . Similar amounts of TrkA were detected in almost every experimental condition in the absence of NGF ( Fig . 5A , time 0 min ) . Addition of 1 nM NGF promoted maximal TrkA internalization at 15 min post NGF incubation in both control ( HEPES ) and M3 conditions ( Fig . 5A ) . However , when SgG2 and NGF were added simultaneously we observed higher levels of TrkA at the plasma membrane ( Fig . 5A ) . Following 120 min of NGF exposure , this difference was maintained although TrkA at the neuronal surface increased both in control and M3 conditions , probably due to receptor recycling ( Fig . 5A ) . In vivo , neurotrophins are secreted by specific tissues and only the axon terminals are directly exposed to them [23] . For a suitable signaling on neurons to occur , neurotrophins must be transported in a retrograde manner from distal axons to their cell bodies . TrkA internalization seems to be required for TrkA retrograde transport [23] . Cofilin is a critical mediator of TrkA retrograde transport [18] . Stimulation of SCG neurons with NGF and SgG2 induced significant higher levels of cofilin phosphorylation , resulting in its inactivation [24] , at late time points ( Fig . 3 ) . We addressed the effect of SgG2 on TrkA retrograde transport by monitoring the localization of phosphorylated TrkA ( p-TrkA ) on SCG neurons , grown in microfluidic devices , whose axon terminals were incubated with NGF alone or in combination with vCKBPs during 120 min . We found very low levels of p-TrkA in the distal axons or cell bodies of neurons not exposed to NGF ( Fig . 5B , upper row ) . As expected , exposure of distal axons to NGF during 120 min promoted a moderate increase on p-TrkA staining in distal axons and a high increase of p-TrkA staining in the cell bodies ( Fig . 5B , second row ) . Similar results were obtained when distal axons were exposed to NGF and M3 ( Fig . 5B , fourth row ) . Exposure of distal axons to NGF and SgG2 induced a significant accumulation of p-TrkA staining in distal axons whereas low level staining was present in the cell bodies ( Fig . 5B , third row and graph ) . Moreover , the growth cone had a normal morphology following 120 min exposure to NGF in both control and M3 conditions , with large actin-rich filopodial protrusions . The presence of SgG2 severely disturbed growth cone morphology , presenting only few filopodia ( NGF-HEPES: 88 , 8% spread growth cones , NGF-M3 94 , 7% spread growth cones and NGF-SgG2: 6 , 2% spread growth cones; P = 0 , 00303681 , P = 0 , 00052529 , respectively ) . Altogether , these data indicated that SgG2 partially blocks NGF-dependent TrkA endocytosis and retrograde transport with p-TrkA accumulating at the distal axon in the presence of NGF-SgG2 . Also , the actin cytoskeleton is perturbed . We hypothesized that SgG2-modification of NGF function could attract axons to infected sites in vivo . NGF is secreted by keratinocytes in the skin [6] . In adult mice glabrous skin there are two morphologically different intra-epidermal FNE , with distinct termination zones ( [25] , Fig . 6A ) : ( i ) FNE that reach and meander through the stratum granulosum , an external layer of epidermis ( Fig . 6A , solid arrowhead ) , most of them being non-peptidergic and expressing RET; and ( ii ) FNE with straight trajectory that remain in the inner stratum basale or stratum spinosum ( Fig . 6A , open arrowhead ) , most of them being from peptidergic TrkA+ neurons [25–27] . Since SgG2 specifically modifies NGF-dependent TrkA+ signaling , we analyzed peptidergic FNE in vivo during HSV-1 or HSV-2 infection . We topically applied PBS , HSV-1 or HSV-2 onto mice hindpaw ( glabrous skin ) following mild skin exfoliation . We did not detect any macroscopical difference in the skin between the different conditions 48 h post-infection . We detected HSV infection with an anti-gBgD antibody and the peptidergic FNE using an anti-calcitonin gene related peptide ( CGRP ) antibody . The number of peptidergic FNE was significantly reduced in skin infected with both HSV-1 and HSV-2 ( Fig . 6B ) . Most peptidergic FNE remained in the stratum basale or in the stratum spinosum in PBS-treated skin , as described [25] ( Fig . 6C left panel , 6D left panel ) . HSV-1 infection promoted a subtle , although not significant , change in the trend of the termination zone of the remaining peptidergic FNE ( Fig . 6C , right panel ) . However , following HSV-2 infection nearly half of the remaining peptidergic FNE increased in length , reaching the stratum granulosum ( Fig . 6D , right panel ) . We hypothesized that changes in straight peptidergic FNE termination zone induced by HSV-2 , but not by HSV-1 , could be mediated by SgG2 through its interaction with NGF . We transfected mice hindpaw skin with empty vector or viral cDNA coding for V5-SgG2 or V5-M3 . Transfection did not affect epidermal layers , general nerve arrangement or the number of peptidergic FNE per field in all the conditions tested . In vector-transfected skin , nearly all straight trajectory-peptidergic FNE remained in the stratum basale or in the stratum spinosum , ( Fig . 6E , left panel , 6F left panel ) . However , in the fields where V5-SgG2 expression was detected , around 20% of straight trajectory-peptidergic FNE showed increased length , reaching the stratum granulosum ( Fig . 6E , right panel ) . This phenomenon was not detected when V5-M3 was expressed ( Fig . 6F , right panel ) . Altogether these data suggested that HSV-2 , but not HSV-1 , modifies the termination zone of the straight peptidergic FNE and that SgG2 may account , at least partially , for this phenomenon .
Neurotropism is a major evolutionary advantage for HSV . The infection of neuronal FNE permits HSV to establish latency in sensory ganglia and persist for the lifetime of the individual . We have detected that infection of glabrous skin with HSV-1 and HSV-2 causes a reduction in the number of peptidergic FNE . This could be due to the secretion of neurotoxic molecules following infection , such as IL-1β and TNF-α [28 , 29] . HSV may have developed molecular mechanisms to counteract this toxicity . The presence , plasticity and topology of FNE during development and adult remodeling depend on axonal navigational cues and trophic factors like neurotrophins . Our results show that HSV-2 SgG specifically binds to several neurotrophic factors from the neurotrophin family ( NGF , BDNF and NT3 ) and the GFL family ( GDNF and artemin ) . Interactions of HSV-1 gG and MHV-68 M3 with neurotrophic factors were also observed but we cannot rule out the possibility that these interactions are non-specific . Further research may identify biological relevance for these interactions in the biology of HSV-1 and MHV-68 . Only the interaction between SgG2 and NGF is biologically relevant in our experimental model resulting in an increase in FNE growth of TrkA+ neurons . This constitutes the first description , to our knowledge , of a protein expressed by a human pathogen with the ability to modulate neurotrophic factors and we hypothesize that this interaction may contribute to HSV neurotropism . To test this hypothesis we focused our functional studies on neurotrophic factors regulating innervation of the epidermis and that are important in skin homeostasis and inflammation , like NGF and artemin [6 , 7 , 13] . Using SCG explants and dissociated cultures , only SgG2 , but not SgG1 or M3 , modified NGF-dependent axonal growth ex vivo . In vivo only SgG2 modified the growth and termination site of the TrkA+ FNE , corresponding to peptidergic neurons . Importantly , similar results were observed when infecting the skin with HSV-2 but not with HSV-1 . One of the most intriguing results of our work is the difference between HSV-1 and HSV-2 . HSV-1 seroprevalence is higher than that of HSV-2 [30–32] . HSV-1 is normally transmitted during childhood and is linked to facial herpes whereas HSV-2 is normally sexually transmitted and associated with genital herpes . However , both viruses can infect these two anatomical areas , and genital herpes due to HSV-1 is increasing [31 , 33] . Of note , HSV-1 reactivates in the genital tract less frequently than HSV-2 [34] . Therefore , differences in prevalence , transmission route and recurrent disease between HSV-1 and HSV-2 in the oro-labial or genital area exist . The lack of effect with HSV-1 in our in vivo experiments does not rule out the possibility that HSV-1 may modulate neurotrophic factors in another setting or utilizes a different mechanism to modulate axonal growth . In this regard , HSV-1 latency associated transcript induces axonal regeneration and growth in a post-entry phase following serum deprivation [35] or NGF starvation [36] , and this could facilitate release of HSV-1 into the peripheral tissue following anterograde transport from the neuronal cell body [36] . From a molecular perspective , the different activity of gG1 and gG2 could be due to their structure and location . HSV-1 gG is a transmembrane protein that remains anchored at the surface of infected cells and the virion envelope whereas gG2 is proteolytically processed secreting an N-terminal domain , SgG2 , with the potential of reaching neighboring cells . Mature peptidergic neurons express TrkA . These type of FNE represent 40% of the total FNE in glabrous skin [25] , but they are the predominant population in mucosa and in the internal organs [37 , 38] . The fact that HSV-2 displays a specific molecular mechanism involved in attracting peptidergic FNE points to possible implications on a more efficient colonization of different anatomical niches or subsets of neurons . In this regard , recent evidence indicates that sensory A5+ neurons ( corresponding to CGRP+/TrkA+ neurons ) support HSV-2 productive infection while they are non-permissive for HSV-1 productive infection in vitro [39] . This difference in susceptibility is dependent on the latency associated transcript [40] . Whether the lack of HSV-1 effect on TrkA+ FNE shown here may influence A5+ neuron susceptibility requires further investigation . The genitalia are enriched in TrkA-expressing peptidergic innervation [41 , 42] . For instance , TrkA+ projecting neurons rise up to 60% from first sacral ganglia to rat penis [43] . We propose that SgG2 may facilitate the infection of FNE innervating the genitalia and/or subsequent spread . In agreement with this hypothesis , pharmacological modulation of capsaicin-sensitive peptidergic neurons reduces the severity of cutaneous HSV-2 genital infections both in female and male pigs [44] . Alternatively , infection of peptidergic neurons may be relevant for nociception . In summary , we cannot conclude from our study that HSV-2 SgG-mediated modulation of NGF is essential for the infection of FNE and subsequent colonization of the nervous system , but our data lead us to propose that SgG2 may facilitate the infection of specific subsets of neurons and this may have consequences for transmission and disease . Deletion or disruption of us4 , the gene encoding gG , in HSV-1 results in partial attenuation in mouse models of infection [45–47] . There are no reports analyzing the role of gG2 in vivo using animal models . Transient enhancement of NGF-mediated axonal growth by SgG2 involves several related events affecting TrkA localization and trafficking . SgG2 alone modifies TrkA association with gangliosides , increasing its relative presence in GM3- versus GM1-rich rafts . This effect could be mediated by the SgG2 ability of binding glycosaminoglycans ( N . M . -M . , A . V . -B . and A . A . submitted manuscript ) ) . However SgG2 mediated TrkA recruitment to GM3 is not sufficient to induce changes in TrkA signaling by itself . SgG2 must be bound to NGF to promote the described changes as previously observed for SgG2 enhancement of chemokine function [12] . GM1 is the preferential ganglioside for canonical NGF-TrkA signaling [21] . GM1 is a marker of caveolae and caveolae-like membranes [20] . The SgG2-mediated TrkA exclusion from GM1 rafts may influence the function of TrkA and could explain the reduced TrkA-p75NTR interaction upon NGF stimulation since p75NTR is located mainly within caveolae and caveolae like membranes [20] . TrkA-p75NTR interaction appears to be required for NGF-TrkA endocytosis in some models [22] and could account at least partially for the reduced TrkA endocytosis detected in the presence of SgG2 during NGF stimulation . The fact that TrkA is a raft resident protein [48] , whereas RET remains outside lipid rafts and is transiently recruited to rafts upon binding to GFLs in cis [49 , 50] , could explain the lack of SgG2 effect on Artemin function . Since TrkA endocytosis is a pre-requisite for its retrograde transport [18 , 51] , this could explain the blockage of NGF-mediated TrkA retrograde transport caused by SgG2 . Finally , accumulation of p-TrkA in distal axons mediated by NGF-SgG2 could induce a local increase in axon length as proposed for NT-3 activation of TrkA [51 , 52] . The SgG2-NGF-TrkA complex promotes an aberrant downstream signaling . The enhanced NGF-mediated TrkA phosphorylation in the presence of SgG2 could be due to the presence of higher levels of TrkA at the plasma membrane , to differences in NGF-TrkA sensitivity , or to the modification of TrkA-ganglioside association since the raft environment determines the interaction of receptors with specific signaling components [19] . One of the most intriguing questions that arise from SgG2-NGF-TrkA signaling resides in the long-term inactivation of cofilin through phosphorylation . Cofilin is an actin-severing protein that regulates actin turnover [24] . Besides , cofilin is a key component of the TrkA retrograde transport complex [18] . Our data showed that SgG2 blocks NGF-induced TrkA retrograde transport and disturbs growth cone morphology . Both results are in agreement with the observed phosphorylation of cofilin . An increased amount of p-TrkA in distal axons may lead to a rapid and transient axonal growth . Growth cones appear to function as a probe [53] , and exposure to SgG2-NGF reduces filopodia resulting in a blunt growth cone . These blunt growth cones could be less responsive to repulsive navigational guidance cues and , together with an increased TrkA and ERK signaling , may reach distant , non-permissive sites . All these molecular events promoted by SgG2 , raft relocation of TrkA , reduced internalization of TrkA in response to NGF and activation of modified NGF signaling pathway , appear to be interconnected finally resulting in the invasion of external epidermal layers by peptidergic FNE when SgG2 is ectopically expressed in mouse footpad skin , probably facilitating HSV-2 access to this type of FNE . We have shown that HSV gG is a vCKBP with the unique property of enhancing the activity of chemokines [12] , immune proteins involved in the migration and activation of leukocytes [54] . Both gG1 and gG2 potentiate chemokine-mediated migration in vitro and in vivo , and this may facilitate the infection of cells recruited to areas of infection and thereby virus dissemination to other tissues . The migration of uninfected epithelial cells towards HSV infected sites has been shown [55] supporting the ability of HSV to modulate cell migration . Here we demonstrate that HSV SgG2 interacts with neurotrophic factors and enhances NGF-dependent growth of FNE to sites of infection probably facilitating transmission to the PNS . Interestingly , SgG2 uses a common mechanism to enhance chemokine and NGF activity , causing the relocalization of their receptors to specific ganglioside-rich sites in the plasma membrane and delaying the internalization rate of the receptors , increasing their levels at the cell surface ( this report and N . M . -M . , A . V . -B . and A . A . , manuscript submitted ) . SgG2 seems to bind both chemokines and NGF simultaneously and this property could play a relevant role during in vivo infection due to the contribution of chemokines and neurotrophins in the crosstalk between the immune and nervous system . Chemokines are essential in the antiviral response but they can also modulate the responsiveness of axons to guidance cues , acting as antagonists of axonal repulsion and inducing axonal sprouting [56–58] . Neurotrophic factors are essential elements of the nervous system and also play relevant roles in inflammation and pain modulation [59] . NGF is the major contributor of axonal growth of peptidergic neurons and a key component of the neurogenic inflammatory response in the epidermis in response to stress or under pathological conditions , permitting a coordinated response between immune cells and peptidergic neurons [60–62] . Our results in vitro , ex vivo and in vivo , lead us to propose that SgG2 modifies NGF-TrkA signaling to induce peptidergic neurons to grow to external layers of the epidermis , facilitating access of HSV-2 to peptidergic neurons . This is , to our knowledge , the first example of a viral protein , encoded by a highly prevalent neurotropic human pathogen that interacts with components of both immune and nervous systems , and both activities may help HSV-2 to reach FNE to establish latency in neurons . A similar strategy may be used by other relevant human pathogens . The interaction of SgG2 with neurotrophic factors sheds light on the complex network of virus-host interactions and uncovers a new molecular framework to investigate the colonization of the nervous system by HSV-2 , an important human pathogen .
All animal experiments were performed in compliance with national and international regulations and were approved by the Ethical Review Board of the Centro de Biología Molecular Severo Ochoa under the project number SAF2009–07857 . The procedures employed complied with the National ( “Real Decreto” 1201/2005 and 53/2013 ) and European ( Directives 86/609/CEE and 53/2013 ) regulations . The interactions between neurotrophic factors and vCKBPs were characterized by SPR technology as before [63] , with Biacore X and X100 biosensors ( GE Healthcare ) . SgG1 , SgG2 and M3 were coupled to CM4 or CM5 Biacore chips through amine coupling . In the screening experiment , to determine binding to SgG1 , SgG2 and M3 , neurotrophic factors ( Peprotech ) were injected at 100 nM in HBS-EP buffer ( 10 mM HEPES , 150 mM , NaCl , 3 mM EDTA , 0 . 005% ( vol/vol ) surfactant P20 , pH 7 . 4 ) at a flow rate of 10 μl/min , and association and dissociation were monitored . Other analytes used ( IFN-α , TNF- α and IL-1 , Peprotech ) were also injected at 100 nM . To determine the association and dissociation kinetics , different concentrations of neurotrophic factors ( with the exception of artemin ) were injected at a flow rate of 30 μl/min , and association and dissociation were monitored . The number of relative units in each chip was 831 ( for SgG1 ) , 187 ( for SgG2 ) and 375 ( for M3 ) . Artemin was coupled to a CM5 chip and different concentrations of SgG1 , SgG2 and M3 were injected . Human NGF was coupled to a CM4 chip and different concentrations of SgG1 , SgG2 and HSV-2 gD were injected . In some experiments NGF and CXCL12β were injected alone or in combination into a chip containing SgG2 to address whether SgG2 could bind both simultaneously . All Biacore sensorgrams were analyzed with the software Biaevaluation 3 . 2 and Evaluation 2 . 0 . 1 ( for sensorgrams obtained with the Biacore X and Biacore X100 , respectively ) Mouse sympathetic SCG were cultured in a 3D collagen matrix [64] . In brief , 340 μL of rat tail collagen I ( BD Biosciences , San Jose , CA ) were mixed with 40 μL of 10x MEM , 10 μL of HEPES or the vCKBPs at a final concentration of 50 nM , and mouse NGF 2 . 5S ( N-100 , Alomone labs , Jerusalem , Israel ) at a final concentration of 0 . 25 nM . This mix was neutralized using 0 . 8 M NaHCO3 , and immediately spotted in drops where ganglia were included . Mouse sympathetic neurons from SCGs were cultured as previously described [65] . Ganglia were dissected from newborn mice ( postnatal day 0–1 ) , digested in collagenase and trypsin ( Worthington , Lakewood , NJ ) , dissociated by trituration and plated on dishes previously coated with rat tail collagen I ( BD Biosciences ) in DMEM containing 50 ng/mL NGF ( Alomone labs ) , 10% fetal bovine serum and 5ng/mL of aphidicolin ( A . G . Scientific , San Diego , CA ) for 5–7 days . For retrograde transport analysis SCG neurons were cultured in polylysine ( 100 μg/mL ) —laminin ( 10 μg/mL ) using microfluidic devices AXIS Axon Isolation Devices , 450um from Millipore ( Billerica , MA ) HEK-293T were cultured in MEM , 10% FBS and antibiotics . HEK-293T were transfected with Lipofectamine 2000 ( Life Technologies , CA , USA ) according to the manufacturer´s instructions , and 6 h after transfection the cells were detached to prepare cell aggregates using the “hanging drop” method [64] . Dissociated neurons were grown during 5 to 7 days in vitro ( DIV ) and starved of NGF during 16 h when indicated . NGF and vCKBP were mixed in DMEM prior stimulation . To calculate NGF molarity we considered NGF as a dimer ( 26kDa ) . The concentrations used were 0 . 5 nM NGF with 100 nM vCKBP for signaling experiments; 1 nM NGF with 200 nM vCKBP for TrkA-p75 interaction analysis and TrkA internalization assays; and 5 nM NGF with 1 μM vCKBPs were applied in the distal axon compartment for retrograde transport analysis . Pearson´s coefficient ( PC ) is a standard statistical analysis designed to measure the strength of a linear relationship between two variables , in this case fluorescent intensities from two images . PC generates a range of values from 1 , a perfect positive correlation , to −1 , a perfect but inverse correlation , with 0 representing a random distribution . The intensity correlation analysis ( ICA ) [66] method is based on the principle that if two proteins are part of the same complex then their staining intensities should vary in synchrony , whereas if they form part of different complexes or structures they will exhibit asynchronous staining . Intensity correlation quotient ( ICQ ) was used to provide an overall index of whether the staining intensities were associated in a random , a dependent or a segregated manner . Given that the outlining of regions in which two probes may distribute is required to obtain accurate measurements of colocalization , we drew regions of interest ( ROI ) corresponding to the entire plasma membrane of the neuron , where the colocalization analysis was carried out . A minimun of three different sections were quantified in the assays shown . ICA analysis was perfomed using Image J 1 . 43 software . Mice were anesthetized with a mixture of ketamine xylazine prior to infection . A region located between proximal pads and heel of ventral hindpaw was exfoliated by rubbing the skin 30 times with an exfoliation sponge ( 3 M , Heavy Duty ) , followed by tape stripping . This procedure results generally in transient and mild erythema but not in hemorrhage and scar formation [67] . Following exfoliation 10 μL of PBS , HSV-1 or HSV-2 containing 1x104 plaque forming units were applied on the footpad . To transfect mouse hindpaw skin , a similar exfoliation procedure was used . The DNA was transfected using In-vivo Jet Pei ( Polyplus Transfection , Illkirch , France ) [67] . Mice were euthanatized 48 h after in vivo transfection or infection . Transfected or infected hindpaw skin was immediately removed by using a 3 mm biopsy punch and fixed in Zamboni´s fixative for 6 h . Then biopsies were washed , cryoprotected in 20% sucrose for 24 h , embedded in OCT and sectioned . Staining of FNE was performed by adapting previously established protocols [25 , 68] . 40 μm free floating sections were washed in PBS with 0 . 3% Triton X-100 ( PBS+TX ) , blocked for 30 min in 10% horse serum in PBS+TX . Antibodies used for FNE staining were anti-protein gene product ( PGP ) 9 . 5 rabbit antibody ( Cedarlane-Ultraclone , Ontario , Canada ) and anti-CGRP rabbit antibody . Anti V5-tag mouse antibody to detect viral proteins containing a V5 tag was from Sigma . Secondary antibodies used for immunostaining were from Sigma . Confocal analysis was performed with LSM 510 Confocal Laser Scanning Microscope from Carl Zeiss . Analysis and treatment of images was performed using LSM Image Browser , Fiji and Adobe Photoshop . | Herpes simplex virus type 1 and 2 ( HSV-1 and HSV-2 , respectively ) establish latency in peripheral sensory ganglia , where they remain for the lifetime of the infected individual . Understanding the mechanisms that allow these viruses to colonize the nervous system will permit devising antiviral strategies . We show that HSV-2 glycoprotein G ( SgG2 ) binds to and increases the function of nerve growth factor ( NGF ) , a neurotrophin expressed in the skin and mucosa essential for axonal growth and neuronal survival . This constitutes the first description , to our knowledge , of a human pathogen with the ability to augment neurotrophic factor function . The enhancement in NGF activity results in an increase in axonal growth of neurons expressing the receptor for NGF . These results were obtained in vitro , ex vivo and in the infected mouse , suggesting that this effect may permit a more efficient infection of NGF dependent free nerve endings by HSV-2 . Absence of a similar function for HSV-1 gG may indicate a preference for the infection of particular subsets of neurons by these viruses . These results shed light on the modulation of neurotrophic factors by relevant human pathogens and on the mechanisms of colonization of the nervous system by HSV . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Secreted Herpes Simplex Virus-2 Glycoprotein G Modifies NGF-TrkA Signaling to Attract Free Nerve Endings to the Site of Infection |
We report a study of genome-wide , dense SNP ( ∼900K ) and copy number polymorphism data of indigenous southern Africans . We demonstrate the genetic contribution to southern and eastern African populations , which involved admixture between indigenous San , Niger-Congo-speaking and populations of Eurasian ancestry . This finding illustrates the need to account for stratification in genome-wide association studies , and that admixture mapping would likely be a successful approach in these populations . We developed a strategy to detect the signature of selection prior to and following putative admixture events . Several genomic regions show an unusual excess of Niger-Kordofanian , and unusual deficiency of both San and Eurasian ancestry , which were considered the footprints of selection after population admixture . Several SNPs with strong allele frequency differences were observed predominantly between the admixed indigenous southern African populations , and their ancestral Eurasian populations . Interestingly , many candidate genes , which were identified within the genomic regions showing signals for selection , were associated with southern African-specific high-risk , mostly communicable diseases , such as malaria , influenza , tuberculosis , and human immunodeficiency virus/AIDs . This observation suggests a potentially important role that these genes might have played in adapting to the environment . Additionally , our analyses of haplotype structure , linkage disequilibrium , recombination , copy number variation and genome-wide admixture highlight , and support the unique position of San relative to both African and non-African populations . This study contributes to a better understanding of population ancestry and selection in south-eastern African populations; and the data and results obtained will support research into the genetic contributions to infectious as well as non-communicable diseases in the region .
The analysis of high-throughput genotype data has revealed global patterns of human haplotype variation , casting light on the pre-history of human populations [1 , 2 , 3 , 4 , 5] . The International HapMap consortium [1 , 5] ) and Human Genome Diversity Project ( HGDP ) [6] , among others , have facilitated the analysis of human genome-wide variation , and linkage disequilibrium in disease association studies [1 , 4 , 5] and also helped refine estimates of recombination rates [7] . Comparative genome-wide genotype data among humans , Neanderthals and Chimpanzees have also shown that selection has played a significant role in human adaptation to the environment [8 , 9 , 10 , 11] . These data have provided additional support for the African origin of modern humans [12 , 13] and highlight the effects of migration both within Africa and out of Africa . In general , African populations exhibit less linkage disequilibrium between adjacent markers than their non-African counterparts , consistent with a migratory bottleneck in the latter [1 , 2 , 5] . Such differences in the extent of linkage disequilibrium have a profound effect on the power of case-control association studies , since these studies depend largely on linkage disequilibrium between disease variants and genotyped single nucleotide polymorphisms ( SNPs ) . Substantially more SNPs are required to capture genomic variation in African populations than populations of European ancestry [1 , 5] . In addition , African populations are characterized by higher levels of genetic diversity [13 , 14 , 15 , 16] and considerable population substructure [17 , 18 , 19] , probably the combined result of several migration events , effective population size changes , population differentiation through genetic drift and local selective forces operating in ecologically diverse environments [18] . Hypotheses of migration within Africa based on mitochondrial DNA ( mtDNA ) suggest that at least three major migration events are plausible that could account for the patterns of mtDNA variation within Africa [17]; ( 1 ) the divergence of southern African San and east African populations who share the ancestral mtDNA haplogroup ( L0d ) and associated lineages in their maternal gene pool from an ancestral parental population circa 200 kya , ( 2 ) the establishment of west African maternal haplogroups ( L1’5 & L0abf ) from an east African source ( circa 100 kya ) , and ( 3 ) the Bantu expansion from the Niger-Congo region into central , eastern and southern Africa ( < 5 kya ) . Although a southern African versus east African origin of modern humans cannot be fully evaluated with current data , multiple lines of evidence from mtDNA [16] , Y chromosomes [20] , Alu insertions [21] , and autosomal SNPs [3] place the divergence of the San at the root of modern humans with at least 100 ky of isolation from other non-San African populations [17 , 22] , and relatively recent ( < 5 kya ) admixture with Bantu-speaking populations [16 , 23 , 24 , 25 , 26 , 27] , followed by subsequent admixture ( < 5 kya ) in the region [16 , 28 , 29 , 30] . Given this relative isolation of present-day San in southern Africa , it is expected that many SNPs ascertained in HapMap populations may not necessarily be polymorphic in San , unless the polymorphisms arose well before the divergence of these populations . Southern Africa was occupied exclusively by the San prior to the arrival of Bantu-speaking populations within the past 1 , 500 years , a consequence of the Bantu-expansion out of west Africa some 5000 years ago [16 , 23 , 24 , 25 , 26 , 27 , 31] . Migrations across equatorial central Africa to the region of the Great Lakes in east Africa , followed by southern African migrations [16 , 25] established the eastern and southeastern Bantu-speaking groups , respectively . Migrations along the west coast of Africa contributed to western and southwestern Bantu-speaking groups , the latter , currently extending to Namibia [16 , 25 , 26 , 27 , 28 , 29] . According to our findings , the label “Khoe-San” represent populations resulting from the mixture of predominately San , Eurasian and Bantu-speaking populations . Over hundreds of years , indigenous San and Khoe-San communities have undergone a sharp decline in population size , largely due to warfare and diseases such as smallpox which arrived with colonialists [29 , 32] . It is estimated that the population decline ( i . e . 90 percent ) of both San and Khoe-San populations was due to smallpox [31 , 32] . Recently , Lachance et al . [33] used the whole-genome sequences of five individuals in each of three different hunter-gatherer populations , including Pygmies from Cameroon , Khoe-San-speaking Hadza and Sandawe from Tanzania , and identified several genomic regions with evidence of archaic introgression in the hunter-gatherers . In addition , Lachance et al . [33] demonstrated that distribution of the time to the most recent common ancestors for these regions was similar to that observed for introgressed regions in Europeans [33] . Ancient and relatively recent contact between immigrants from Europe , Asia and Indonesia with sub-Saharan Africans [24 , 26 , 34] have resulted in varying degrees of admixture between these populations . Furthermore , a recent study by Gurdasani et al . [35] presented a broad survey of polymorphisms in a novel array genotyping data set of ∼1 , 481 individuals from 18 self-identified ethnic/linguistic and low coverage whole genome sequencing data set of 320 individuals from 7 self-identified ethnic/linguistic in Sub-Saharan Africa , and suggested that Eurasian back migrations to Africa and contributions to ancestry has a substantial impact on differentiation among some sub-Saharan African populations . These mixtures have also contributed to shaping the gene pool of the derived populations in south-eastern Africa [28 , 35] . Other disciplines , such as archaeology , history and anthropology , have given us clues about the prehistory of African populations . The study by Pickrell et al . [16] convincingly demonstrated waves of two-way admixture between Niger-Congo-speaking African and west Eurasian ( European or Middle Eastern ) populations to form eastern and southern African ( admixed ) populations . However , the role of native indigenous San in the south-eastern African region and the genetic contribution of this population to the southern and eastern African admixed populations has not been elucidated . The present study makes use of genetic markers to investigate which factors , and to what extent , they have contributed in shaping the gene pools of extant southern and eastern African populations . More specifically , we used the Affymetrix Genome-Wide Human SNP Array 6 . 0 , to examine ∼900K SNPs and copy number variants in five indigenous populations comprising 25 Ju\’hoansi San from Namibia ( KHS ) , southeastern Bantu-speakers [25 Sotho-Tswana ( STS ) , 36 Xhosa ( XHS ) , 25 Zulu ( ZUL ) ] as well as 25 Herero ( HER ) , a southwestern Bantu-speaking group from Namibia . These data were used in conjunction with other published data to examine the genetic origins of southern African populations . Importantly , our study demonstrates the admixture of the indigenous San , Niger-Congo-speaking populations and populations of Eurasian ancestry in southern and eastern African populations . We have also developed two complementary approaches to identify signatures of selection prior to and following putative admixture events in the southern African populations .
The sample consisted of unrelated individuals belonging to the following five self-identified ethnic/linguistic populations of southern Africa: southeastern Bantu-speaking [25 Sotho-Tswana ( STS ) , 25 Zulu ( ZUL ) and 36 Xhosa ( XHS ) ] , southwestern Bantu-speaking [25 Herero ( HER ) ] , and 25 Ju\’hoansi San ( KHS ) . The Sotho-Tswana and Zulu samples were collected in Johannesburg , the Xhosa from Khayelitsha in Cape Town , the Herero from Windhoek , and the Ju\’hoansi from Tsumkwe [36] . The Blood samples were collected with the subject’s informed consent , and the use of DNA samples for population genetics research was approved by both the University of the Witwatersrand and University of Cape Town . DNA samples were shipped to Affymetrix ( http://www . affymetrix . com ) for genotyping using the Affymetrix Genome-Wide Human SNP Array 6 . 0 , containing 906 , 600 SNPs and more than 946 , 000 probes for the detection of copy number variation . These data were used to examine patterns of migrations , genetic ancestry and effects of selection in this study . Other populations included in this study are listed in S1 Table . The separation of Africans from non-Africans is clearly evident ( Fig . 1 ( A ) ) ; this has also been previously reported with both microsatellite data [37 , 38] as well as with other SNP data [2 , 3 , 5] . From pairwise population genetic distance estimates , we find that there is little genetic difference among Bantu-speaking populations ( S2 Table ) . In addition , Fig . 1 ( A ) shows a distinct separation of San populations ( San ( SAN ) and Ju\’hoansi ( KHS ) and Khoe-San populations ( Bushmen ( BUS ) , ‡Khomani ( KHO ) ) , consistent with previous studies [16 , 26 , 33 , 39 , 40] . This result suggests Khoe-San , and both eastern and southern Bantu-speaking populations have undergone admixture . Furthermore , this result is consistent with the 3-population test [39 , 40] result displayed in S3 Table , which shows clear evidence of admixture between Yoruba ( YRI ) and KHS in the southern Bantu ( ZUL , STS , XHS ) . Furthermore , the ‡Khomani ( KHO ) , and eastern Bantu-speaking populations also reflect a three-way admixture of Caucasian ( CEU ) , Yoruba ( YRI ) and KHS . The results in Fig . 1 ( A and D ) suggest that the genetic make-up of the southeastern Bantu-speaking groups ( ZUL , STS , XHS ) includes ancestral contributions from Niger-Congo ( 26% ± 0 . 3% ) and San populations ( 74% ± 0 . 4% ) . However , consistent with previous findings [40] , the data in Fig . 1 ( B-C ) , suggests Niger-Congo ancestry ( 17% ± 1 . 2% and 57% ± 1 . 6% ) , San ancestry ( 70 ± 1 . 3% and 15% ± 0 . 4% ) , and notably Eurasian-related ancestry ( 13% ± 1% and 28% ± 2% ) in the genetic make-up of ‡Khomani ( KHO ) and Sandawe ( SAW ) , respectively . The admixture observed in the Khoe-San ( KHO ) , and in the eastern African populations , ( particularly ) Sandawe ( SAW ) reflects the gene flow from Bantu-speaking agriculturalists and/or eastern African pastoralists within the past 1 , 200 years and sea-borne immigrants from Europe , Asia and Indonesia [33 , 35 , 39 , 40 , 41] . Our observation of Eurasian ancestry in both eastern ( SAW ) and southern ( KHO ) African populations is consistent with archaeological , genetic , climatological and linguistic data [24 , 25 , 26 , 27 , 28 , 35] . Furthermore , Pickrell et al . [16] previously demonstrated multiple waves of population mixture in the history of many eastern and southern African populations , and that genetic material from Eurasians or related populations entered eastern Africa 2 , 700–3 , 300 years ago , and southern Africa 900–1 , 800 years ago [16 , 41] . In addition , our study demonstrates the genetic contribution of the San population to the waves of admixture in the ancestry of the southern and eastern African populations . Using the Mantel test with N = 10000 permutations ( Materials and Methods ) , we found a significant positive correlation between genetic and geographic distance in the southern African populations ( Pearson’s r = 0 . 64; p-value = 1 . 0 × 10−4; Fig . 2 ) . To analyse more closely the outlier points in Fig . 2 , we calculated the perpendicular distance between each point and the regression line . Analysing the concentration of points around the linear regression , we therefore defined outliers as points which are greater than 0 . 05 distance units from the regression line . When analysing the scatter plot ( Fig . 2 ) , there are 10 outlier points , which suggest possible obstacles to migration ( S4 Table ) , assuming that populations have used the shortest path during their migrations . To assess patterns of migrations and to capture the genetic drift in southern African populations , we used a maximum likelihood tree and Gaussian approximation to the genetic drift model; implemented in Treemix [40] . We observed not only a major split between the African and European continent exhibited on this population tree , but also sub-lineages within African , and particularly within the southern African populations ( S1 Fig . ) which is consistent with previous results [16 , 26 , 34 , 39 , 40] . S1 Fig . ( B ) shows the inferred graph with three migration events , explaining the model for the relationship of southern and eastern Africans and non-Africans . This provides evidence for a shared origin for San-and Eurasian- and Bantu-related populations in Sandawe ( SAW ) and ‡Khomani ( KHO ) . The latter possibility would be consistent with known south-east African admixture in the Sandawe ( SAW ) and ‡Khomani ( KHO ) . We clearly see four population branches in southern Africa: ( i ) one formed from the southern Bantu-speaking populations , which are very distinct from the Niger-Congo and eastern Bantu-speaking populations , ( ii ) the second group formed with eastern Bantu-speaking populations , and ( iii ) the third , and ( iv ) the fourth group formed with San ( KHS+SAN ) and Khoe-San ( BUS+KHO ) , both hunter-gatherers which are quite distinct , and are split into two distinct groups , including San populations ( SAN and Ju\’hoansi ( KHS ) ) and Khoe-San populations ( BUS and KHO ) . This is also consistent with the admixture results shown in Fig . 1 , reaffirming the concordance between genetic data with geographic origins of populations and their linguistic affinities . Consistent with previous observations [13] , the mean haplotype block lengths are substantially shorter in African populations than in non-Africans ( Fig . 3 ( A ) and S5 Table ) . Mean block lengths are remarkably consistent across the southern African populations in this study and easily distinguishable from the non-African block lengths . Similarly , decay of linkage disequilibrium with physical distance along the genome is rapid in southern Africans when compared with non-Africans ( Fig . 3 ( B ) ) . Ascertainment biases have been shown to result in faster decay of linkage disequilibrium compared to a sample of non-ascertained markers [42] . We performed coalescent simulations ( S1 Text and S2 Text ) in order to investigate the effects of ascertainment bias when markers are ascertained in a population divergent from that in which they are genotyped . Consistent with previous reports [42] , we found the rate of decay of linkage disequilibrium to be greater with ascertained SNPs ( S2 Fig . ( A ) ) . Similarly , haplotype block lengths are similar , irrespective of whether markers were ascertained in the genotyped population , or in a divergent population ( S2 Fig . ( A ) ) . Frequency spectra , however , differ when SNPs are ascertained in a divergent population ( S2 Fig . ( A ) ) . Indeed more monomorphic SNPs , and thus lower overall SNP diversity , are evident when markers are ascertained in a population divergent from that in which they are genotyped . This is further evident in distributions of minor allele frequencies from empirical data , in which the distribution of minor allele frequencies of San more closely resembles the theoretical expectation for a non-ascertained sample ( S2 Fig . ( B ) ) , mostly due to the abundance of monomorphic SNPs . In addition to differences in demographic processes , such as bottlenecks , differences in the extent and pattern of linkage disequilibrium may be the result of differences in the patterns of fine-scale recombination rate . We assessed the impact of fine-scale recombination events to differences in linkage disequilibrium patterns using a coalescent-based method [7] . Interestingly , we found that the southern African Bantu-speaking populations share proportionally more recombination hotspots with both Yoruba ( YRI ) and Europeans ( CEU ) than with the Ju\’hoansi ( KHS ) ( Fig . 4 , S6 Table ) , where a shared hotspot is identified as a region with greater than five times the background recombination rate within a 10kb window . The proportion of hotspots shared between southern Africans and both European ( CEU ) and Yoruba ( YRI ) samples was generally low ( Fig . 4 ) . Our empirical analyses indicate that few recombination hotspots are shared between southern Africans and the HapMap populations , with San being the most extreme . More results on recombination hotspots and the test of whether increased frequency of low frequency and monomorphic SNPs improves the power to detect recombination hotspots are detailed in S4 Text and S7 Table . To assess the accuracy with which missing SNPs in southern African populations can be imputed using Yoruba ( YRI ) or European ( CEU ) reference populations , we removed SNPs , imputed them and checked for correctness in imputation ( detail in S1 Text and S3 Text ) . Our results show that YRI appears to be useful for imputation , at least for some of the southern Bantu-speaking groups included in the study , namely Sotho/Tswana ( STS ) , Zulu ( ZUL ) , Herero ( HER ) and Xhosa ( XHS ) , but less so for the San , for whom imputation accuracy is significantly lower than for other African populations ( S3 Fig . ) . Xhosa ( XHS ) also had lower imputation accuracy , compared with other Bantu-speaking groups . We first developed an approach to select polymorphisms that exhibit large allele frequency differences between ancestral populations of Sandawe ( SAW ) , Xhosa ( XHS ) and ‡Khomani ( KHO ) ( see Materials and Methods ) . We constructed 3 different panels of AIMs [for Sandawe ( SAW ) , Xhosa ( XHS ) and ‡Khomani ( KHO ) ] , where selected SNPs have a certain level of admixture LD with each other and with at least 1MB spacing between adjacent genetic markers on a chromosome ( Materials and Methods ) . This was to avoid linkage disequilibrium ( LD ) in the ancestral population . Such background LD could contribute noise ( or bias ) to the estimation of ancestral allele frequencies and locus-specific ancestry [43] . Thinning down the SNPs to a 1Mb spacing may result in a reduction in power to detect cases of deviation in ancestry or allele frequency differences that result from selection . Consequently , our strategy to detect regions of unusual differentiation between the admixed southern African populations and their source populations , and unusual deviation in local ancestry , is conservative . We evaluated whether there is an excess of common SNPs with large allele frequency differences ( expressed as a χ2 ( 1 d . o . f . ) statistic under a model ( see Materials and Methods ) of neutral genetic drift ) between putative ancestral populations of each admixed southern African population [‡Khomani ( KHO ) , Sandawe ( SAW ) and Xhosa ( XHS ) ( Table 1 and S5 Fig . ) ] . An unusual extent of population differentiation can suggest the action of population-specific natural selection . We observed several SNPs within chromosomal regions ( Table 1 ) for which the evidence of unusual population differentiation was genome-wide significant between the Sandawe ( SAW ) and Caucasian ( CEU ) populations ( S5 Fig . ) , and a small number of SNPs ( on chromosome 17q25 . 1 and 12q24 . 21 ) showed unusual genome-wide significant differentiation between SAW and its two other putative ancestral populations , Yoruba ( YRI ) and Ju\’hoansi ( KHS ) ( S5 Fig . ) . Chromosome region 3p11 yielded ( to ) a genome-wide significance of unusual differentiation between the Xhosa ( XHS ) and Ju\’hoansi ( KHS ) ( p = 9 . 5e-10 , lowest p-value ) , and between ‡Khomani ( KHO ) and Ju\’hoansi ( KHS ) ( p = 7 . 6e-09 , lowest p-value ) . Furthermore , unusual allele frequency differences between the Yoruba ( YRI ) and Xhosa ( XHS ) were identified on chromosome 1q41 . No significant signal of unusual allele frequency differences between Yoruba ( YRI ) and ‡Khomani ( KHO ) were observed , which may be explained by the fact that the Niger-Congo contribution to admixture in the Khoe-San groups , in particular the ‡Khomani ( KHO ) ( Khoe-San population ) occurred too recently for it to have a significant impact on their allele frequencies . All these identified candidate SNPs of unusual allele frequency differences lie in or near known genes ( Table 1 ) . Their biological functions in the GeneCards database [44] , are putatively linked with diseases of high prevalence in southern Africa; their detailed annotations are presented in Table 1 . We selected the best proxy parental populations of Xhosa ( XHS ) based on a pool of Click-speaking and Bantu-speaking populations using PROXYANC [45] . Yoruba ( YRI ) and Ju\’hoansi ( KHS ) were chosen as best proxy ancestral populations for Xhosa ( XHS ) . Similarly , among the populations in the study , Yoruba ( YRI ) , European ( CEU ) and Ju\’hoansi ( KHS ) were chosen as best non-San , European and San proxy ancestral populations for both ‡Khomani ( KHO ) and Sandawe ( SAW ) ( Materials and Methods ) . Using AIMs panels , LAMP-LD [46] was employed to estimate the distribution of genetic contributions of ancestry across the genome ( Materials and Methods ) to provide additional reassurance from our data that we obtain unbiased results in the absence of possible background LD . The average locus-specific Ju\’hoansi ( KHS ) and Yoruba ( YRI ) ancestry proportions across the Xhosa ( XHS ) samples were estimated to be 27% ± 3 . 1% and 73% ± 3 . 1% ( mean ± SD ) , respectively . We obtained 12% ± 0 . 8% , 77% ± 1 . 1% and 11% ± 0 . 9% ( mean ± SD ) locus-specific Yoruba ( YRI ) , Ju\’hoansi ( KHS ) and Caucasian ( CEU ) average ancestry contributions , respectively along the genome of the ‡Khomani ( KHO ) . For the Sandawe ( SAW ) , the locus-specific ancestry proportions were 12% ± 0 . 9% , 70% ± 0 . 7% and 18% ± 1 . 0% for Yoruba ( YRI ) , Ju\’hoansi ( KHS ) and Caucasian ( CEU ) average ancestry , respectively . The above estimates of average locus-specific ancestry are all consistent with the related genome-wide average proportion estimates in the admixture analysis section , indicating that there is no evidence of systematic distortion in our local ancestry estimates . The plots of these average locus-specific ancestries of these admixed southern African populations , namely Xhosa ( XHS ) , ‡Khomani ( KHO ) and Sandawe ( SAW ) are in S6 Fig . . In the next two sections , we examined signals of selection , consisting of unusual deficiency or excess of ancestry in the admixed southern Xhosa ( XHS ) , Sandawe ( SAW ) and ‡Khomani ( KHO ) populations . Such regions in admixed populations have served in previous studies as signatures of natural selection that occurred after admixture [43 , 47 , 48 , 49 , 50 , 51] . Here , we considered not only the regions of strong deviation from ancestry , but we also implemented an approach that is now incorporated in PROXYANC [45] to test for unusual deficiency or excess ancestry using the inferred locus-specific ancestry across the genomes of admixed populations . The loci showing unusual ancestry patterns , i . e . four standard deviations above ( excess ancestry ) or below ( reduced ancestry ) the genome-wide average , were identified as candidates of post-admixture natural selection ( Materials and Methods ) . Examining the genome-wide distribution of ancestry in Xhosa ( XHS ) , we detected the natural selection events post-admixture ( Table 2 ) . We identified a region on chromosome 3p11 ( chr3: size: 17 , 184 ( bp ) , p = 1 . 4e-10 ) with strongly reduced Ju\’hoansi ( KHS ) ancestry in Xhosa ( XHS ) ( Table 2 ) . This region yielded a genome-wide significance with an unusual difference of ancestry , suggesting a signal of selection after admixture . The SNP in the 3p11 region with the lowest p-value , rs4858960 , is associated with POU1F1 , which in turn interacts with five other genes [52] , including ETS1 , NR3C1 , JUN , NR1I3 and MED1 . These genes are known to play a role in a metabolic pathway that positively affects growth traits and hormone deficiency [53] . Furthermore , the 3p11 region showed strong differences in allele frequencies between Xhosa ( XHS ) and Ju\’hoansi ( KHS ) ( p = 9 . 5e-10 ) ( Table 1 ) . Since San and Khoe-San communities have undergone a sharp population decline in their history , this differentiation suggests an environmental pressure that the San ancestors of the Xhosa ( XHS ) may have experienced before population admixture , and we speculate a possible adaptation of Xhosa ( XHS ) to the local environment . Mutations in the POU1F1/PIT1 gene , a pituitary-specific transcription factor , affect the development and function of the anterior pituitary and lead to combined pituitary hormone deficiency [53] . In spite of slight predominance of Ju\’hoansi ( KHS ) , San ancestry in ‡Khomani ( KHO ) compared to Sandawe ( SAW ) , and European ( CEU ) related ancestry in Sandawe ( SAW ) compared to ‡Khomani ( KHO ) , consistent with previous findings [16 , 26 , 34 , 40] , our results from both admixture ( Fig . 1 ) and locus-specific ancestry analyses ( S6 Fig . ) have shown a potential ancestral link between the admixed Sandawe ( SAW ) and ‡Khomani ( KHO ) . Three chromosomal regions ( 12q24 . 1 , 18p11 . 31 and 18p11 . 2 ) , each within several SNPs with moderate and significant p-values , appear with excess of Yoruba ( YRI ) ancestry in both Sandawe ( SAW ) and ‡Khomani ( KHO ) ; an additional region ( 13q14 . 3 ) was also identified as an excess of Yoruba ( YRI ) ancestry in ‡Khomani ( KHO ) , ( Tables 2 and 3 ) . These four candidate regions ( Tables 2 and 3 ) showed strong unusual difference of ancestral contributions ( p < 1 . 0 e-08 , chi2 test ) , and have been associated with various important diseases , including malaria , T-cell leukemia , congenital muscular dystrophy , Noonan syndrome [53] , and others listed in Tables 2 and 3 . That some genes in these regions are associated with ‡Khomani ( KHO ) - and Sandawe ( SAW ) -specific high-risk diseases ( such as malaria ) [53] , suggests a functional role these disease-related genes ( or other genetic elements in these regions ) might have played in their migration and particularly local adaptation due to such selective pressure resulting from shared gene-culture co-evolution and cultural practices in Bantu-speaking and Click-speaking populations . Overall , in the results of genome-wide allele frequency differences between Yoruba ( YRI ) and these two admixed populations ( Tables 1 , 2 and 3 ) , only the 12q24 . 1 region was replicated significantly between Yoruba ( YRI ) and Sandawe ( SAW ) . This may indicate different environmental pressures that the ‡Khomani ( KHO ) and Sandawe ( SAW ) experienced post-population-admixture . We observed two other regions ( 12p13 . 31 and 14q13 . 2–14q13 . 3 ) , with significant difference ( Tables 2 and 3 ) of ancestry ( p < 4 . 8e-08 ) showing a strong relative reduction of Caucasian ( CEU ) and Ju\’hoansi ( KHS ) ancestry in both ‡Khomani ( KHO ) and Sandawe ( SAW ) . These regions were also identified as candidates of the natural selection after admixture ( Tables 2 and 3 ) . Importantly , these two regions ( Tables 2 and 3 ) are also associated with some important diseases such as breast cancer , lung cancer , tumour inflammation , diabetes mellitus , Parkinson's and other diseases [44 , 53] , Although these regions have been associated with diseases , there is no indication of whether this points to any mechanistic association . However , it is tempting to speculate that factors such as food , pathogens , and life style , could also be responsible for such reduction in ancestry and may therefore play a role Our approach to analyzing copy number variation in southern African populations involved the detection of known copy number polymorphisms ( CNPs ) using a Gaussian mixture model , and the identification of potential novel copy number variants ( CNVs ) using a Hidden Markov Model ( HMM ) ( S5 Text ) . The number of CNPs ( S5 Text ) in Yoruba ( YRI ) is greater than that found in the European ( CEU ) and the southern African populations ( Table 4 ) . The former is probably the result of bottlenecks in non-Africans and subsequent loss of CNPs of low frequency [54 , 55 , 56] , whereas the latter is likely the result of ascertainment bias . Given that CNP probes were ascertained in HapMap populations ( including Yoruba ( YRI ) ) , lower levels of CNP diversity for populations that are divergent from ascertained populations is expected . However , southern African populations , which are approximately matched for sample size , show marked differences in the distribution of the number of CNPs , particularly in the San ( Ju\’hoansi ( KHS ) ) with fewer CNPs than other southern African populations ( Table 4 ) . Distributions of derived allele frequencies of CNPs suggest higher purifying selection on duplications ( S7 Fig . ) . In contrast , however , there appears to be little difference in the degree of purifying selection on duplications and deletions in novel CNVs detected with the HMM ( S7 Fig . ( A ) ) . We detected a total of 1873 CNVs ( Table 5 ) , of which 1231 were deletions . Only 137 of the CNVs were singletons , with 87 deletions and 50 duplications ( Table 6 ) . A total of 397 were novel with respect to the Database of Genomic Variants [55 , 56 , 57 , 58] . At least 157 of these were unique CNVs , which occurred in only one population . The number of CNVs per individual is generally similar between populations ( S7 Fig . ( B ) ) , except San which had significantly fewer deletions than other populations [e . g . Herero ( HER ) vs Ju\’hoansi ( KHS ) ]: Student’s T-test , t20 = 22 . 4 , P = 1 . 3e-15 ) . Furthermore , distributions of derived allele frequencies of CNPs suggest purifying selection on duplications ( S7 Fig . ( A ) ) . In contrast , however , there appears to be little difference in the degree of purifying selection on duplications and deletions in novel CNVs detected with the HMM ( S7 Fig . ( A ) ) .
In this study , we have conducted a systematic population genomics survey and investigated demographic histories of indigenous southern African populations , making it possible to address questions about the signature of selection prior to and following purported ancient admixture events . Consistent with previous studies [16 , 26 , 33 , 34 , 35 , 39 , 40] , we demonstrated stratification among indigenous southern African populations . Both the geographic distribution of genetic variations and the population structure , suggested a complex human population history generally within the African continent , and specifically in southern and eastern Africa . Incorporating the data from other Click-speaking populations from previous studies [16 , 26 , 33 , 34 , 39 , 40] together with that from our 25 Ju\’hoansi ( KHS ) subjects , it was possible to investigate the relationship between Click-speaking and southern Bantu-speaking populations thought to represent an early diverging branch of modern humans . The admixture analyses , particularly that of southern African populations , lends support of gene flow between San and Niger-Congo-speaking populations due to their contact following migrations of Bantu-speaking populations across the continent [17 , 18 , 26 , 27 , 33 , 34 , 35] . Consistent with previous studies [16 , 26 , 33 , 34 , 39 , 40] , our admixture ( Fig . 1 ) and tree-mix analyses ( S1 Fig . ) suggested a division between south-west ( San ) and south-east ( Khoe-San mostly admixed ) populations . Our findings confirm an ancient link between San and some eastern African populations , including Sandawe , consistent with previous findings [16 , 26 , 35 , 34 , 39 , 40] . The Eurasian ancestral components in south-east Khoe-San and some eastern Bantu speaking populations ( such as Sandawe , Hadza ) may be a consequence of an early Eurasian genetic contribution into Africa [16 , 28 , 35] , Furthermore , the f-3 statistic test ( S3 Table ) confirms southern Bantu speaking populations , in particular Xhosa ( XHS ) to be two-way admixed , and both ‡Khomani ( KHO ) and Sandawe ( SAW ) are at least three-way admixed . The San ( KHS ) exhibit higher levels of homozygosity ( S9 Table ) , increased relatedness ( S9 Table ) and higher proportions of monomorphic SNPs ( S8 Table ) than other African populations . However , we have shown that ascertainment of markers in a divergent population results in a reduction of diversity in the genotyped population , probably the result of polymorphisms arising after the divergence of the ascertained and genotyped populations , and the loss of polymorphisms in the genotyped population through fixation . Improved statistical models are therefore needed for the comparison of populations that have varying degrees of divergence from the population in which markers were ascertained . Our copy number analysis included identification of both known CNPs , which are copy number loci previously identified in HapMap populations [55 , 56 , 58] , and putatively novel CNVs . CNPs are highly ascertained , since they have been selected to be polymorphic and segregating at allele frequencies > 1% in HapMap populations [56] . CNVs , however , are less ascertained and should have more similar levels of polymorphisms in all of the studied populations [55] . In the case of CNVs , deletions are observed more frequently than duplications . This appears to be inconsistent with the proposal that deletions are under stronger purifying selection [58 , 59 , 60] , which has also been inferred previously based on a lower degree of overlap between deletions and both genomic regions [59] , and disease-related genes [59] . However , the disparity in the number of deletion and duplication CNVs probably reflects the relative difficulty of detecting the latter , due to a smaller relative change in copy number ( 3:2 versus 2:1 ) [59] , rather than stronger purifying selection on duplications . In the southern African data , deletions and duplications have similar distributions to that of derived allele frequencies for CNVs , suggesting little difference in the relative degree of purifying selection . The number of deletion CNVs per individual differs markedly between the San ( KHS ) and other African populations . This may be an effect of sample size; however Herero ( HER ) , with a similar sample size to San ( KHS ) for copy number calling , have no reduction in the number of deletions . In addition , copy number variants called for the Zulu ( ZUL ) panel with only 20 samples , were more than 99 . 9% concordant at normal , and 81 . 6% concordant at abnormal copy number regions , with those called in conjunction with other Bantu populations . Alternatively , some hybridization probes may have lower intensities in the San ( KHS ) due to probe-target mismatch mutations . However , such probe effects are likely to cause increased numbers of deletions in the San ( KHS ) . Finally , population demographic and selective effects may cause differences in the number of deletion CNVs . In summary , copy number results suggest San ( KHS ) to be unique , although they should ideally be validated using trios , as shown previously [55 , 56] . Haplotype blocks show very similar patterns of linkage disequilibrium between African populations , with this collective group having substantially shorter haplotype blocks , and less linkage disequilibrium , than Non-African populations . For instance , patterns of linkage disequilibrium surrounding the lactose tolerance ( LCT ) gene , known to have undergone a selective sweep in Europeans [7] , have strong levels of linkage disequilibrium in Europeans , yet not in southern African populations ( S2 Fig . and S4 Fig . ) . Khoe-San , however , appear to have increased levels of linkage disequilibrium associated with LCT than the other African populations [particularly the Sotho/Tswana ( STS ) and Zulu ( ZUL ) ; S2 Fig . ] . This may be due to a weak selective sweep or the result of gene admixture with the San ( KHS ) , a pastoral group from Namibia known to be lactose tolerant [29] . In addition , it was particularly interesting to examine the signature of selection in the indigenous and admixed southern African populations , including ‡Khomani ( KHO ) , Xhosa ( XHS ) and Sandawe ( SAW ) due to the high mortality of the San population , historically . Following the recommendation of Bhatia et al . [61] , we additionally implemented two strategies to detect possible evidence of population-specific natural selection in southern African populations . The first strategy , involved evaluating whether there is an excess of common SNPs with large allele frequency differences between admixed southern African populations , including ‡Khomani ( KHO ) , Sandawe ( SAW ) and Xhosa ( XHS ) and their purported parental populations . The power of this analysis was based on an approach we developed to select three panels of 502 SNPs with at least 1MB spacing between adjacent genetic markers on each individual chromosome . Several SNPs on chromosomal regions for which there is evidence of unusual population differentiation between Sandawe ( SAW ) and Caucasians ( CEU ) , are displayed in Table 1 . Importantly , most of the signals of selection identified through this strategy are linked with specific high-risk diseases such as malaria , influenza , tuberculosis , and AIDs/HIV , which have a high prevalence in southern African populations ( e . g . in the Sandawe , ‡Khomani and Xhosa populations ) ( Table 1 ) . The allele frequency differences between southern African populations ( including some putative parental populations ) follow the null distribution predicted by neutral drift as a consequence of the recent origin of southern African population structure . This may yield a risk of false positive associations due to population stratification in disease association studies , despite the fact that there are differences between southern African populations [62] . The second strategy to detect possible evidence of population-specific post-admixture selection involved a signal of unusual excess or deficiency of ancestry in the admixed southern African populations [‡Khomani ( KHO ) , Sandawe ( SAW ) and Xhosa ( XHS ) ] . The recent studies by Bhatia et al . [61 , 63] showed that loci with significant deviation in local ancestry ( from the genome-wide average ) may due to insufficient correction for multiple hypothesis testing and/or due to possible systematic errors in local ancestry inference . We have employed the minor allele frequencies from the correct proxy ancestral populations of the admixed population to correct for possible systematic errors on the inferred local ancestry that may lead to false positive deviations in local ancestry . Moreover our study did not only rely on the deviation ( more than 4 . 0 standard deviations ) in local ancestry from the genome-wide average; we additionally used the distribution of difference in locus-specific ancestry along the genome admixed population to evaluate the genomic regions showing unusual excessive or reduced ancestry which are likely to be signatures of natural selection after admixture [43 , 48 , 49 , 50 , 51] . Several recent studies have detected excessive or reduced ancestry contributions in admixed populations as signals of post-admixture selection , using reference ancestral parental populations [43 , 48 , 49 , 50 , 51] . Our study used selected best proxy ancestral populations and AIMs panels for our admixed southern African populations , and we extended previous approaches to test for unusually increased or decreased ancestry contribution along the genome . We identified three and four regions showing a significant excess of Yoruba ( YRI ) ancestry in Sandawe ( SAW ) and ‡Khomani ( KHO ) , respectively ( Tables 2 and 3 ) . Three other regions showed unusually reduced Caucasian ( CEU ) and San ( KHS ) ancestry in both ‡Khomani ( KHO ) and Sandawe ( SAW ) ( Tables 2 and 3 ) . Since some of the genes in these regions are linked with specific high-risk diseases such as malaria in the ‡Khomani ( KHO ) and Sandawe ( SAW ) , as has also been noted in the recent study by Gurdasani et al . [35] , it is plausible that these disease-related genes might have played a role in population adaptation historically . Among the identified genomic regions , the 12q24 . 1 region was found in both strategies for detecting signals of natural selection , supporting evidence of environmental pressures that the ‡Khomani ( KHO ) and Sandawe ( SAW ) experienced . Furthermore , two other candidate regions pointing to natural selection were identified in both ‡Khomani ( KHO ) and Sandawe ( SAW ) , showing strong deficiency of European and San ancestry components , and also an unusual population differentiation in these regions . These two regions are also linked with some important diseases such as breast cancer , lung cancer , inflammation , diabetes mellitus and Parkinson's disease [53] , which are known to occur at a relatively higher prevalence in European populations , when compared to indigenous southern African populations [59] . African , and particularly southern and eastern African populations , face a heavy burden of diseases including HIV/AIDs , tuberculosis and malaria , and a growing burden of non-communicable diseases [17] . Of note , all the reported regions with signals of selection are in admixture LD and with significant deviation in average local ancestry ( or unusual difference in allele frequency ) . In addition , our constructed AIMs panels for southern and eastern admixed populations may potentially be utilized for further admixture mapping studies in these populations . Nevertheless , further investigations are required to reveal the targets and agents of selection that have played important roles in shaping the admixed gene pool of these southern and eastern African admixed populations . With extensive admixture , both between none-San and San populations , and between African and non-African populations , southern and eastern African populations have a great potential for the identification of genes which determine susceptibility to both communicable and non-communicable diseases and to understand the African genetic variations with response to drugs/treatment variability . The southern Bantu and Khoe-San populations are 'admixed' and future genome-wide studies will need to correct for this stratification or may need to use the locus-specific ancestry to increase power in association studies . Admixture mapping in the African-American and some other three-way admixed populations ( such as Latinos , Puerto ) has been successful for some disease traits [43 , 51] . Since the admixed southern African populations have similar admixture proportions to admixed American populations , we hypothesize that admixture mapping would likely be a successful approach in many southern Bantu and Khoe-San cohorts , and particularly in the Xhosa , ‡Khomani and Sandawe . A large proportion of the currently active genomic studies being conducted as part of the recently launched H3Africa programme ( H3Africa , http://h3africa . org/ ) and the more recently described African Genome Variation Project [35] , involve genome wide association studies [64] . A significant number of these studies involve large collections of sub-Saharan African subjects , and would benefit from this knowledge .
This study , investigating the genomic structure of indigenous southern African populations , was approved by the Research Ethics Committees of the University of Cape Town , and Witwatersrand University ( REC Ref 305/2009 for the Project: Genome Wide Microarray Analysis of southern African Human Populations [65 , 66] . Consider a pair of populations k and l from a pool of K ancestral populations of an admixed population and assume that the minor allele frequencies at SNPs i and j are greater than 0 . 005 . Similar to Glaubitz et al . [67] , we defined the admixture linkage disequilibrium as Lij=mLijk+ ( 1−m ) Lijl+m ( 1−m ) δikl×δjkl ( a ) Where m is the ancestral proportion , δi and δj are differences in allele frequency at SNPs i and j in population k and l , respectively . Assuming for each pair of SNPs i and j there is no linkage disequilibrium in ancestral populations , it thus follows , Lij=m ( 1−m ) δikl×δjkl ( b ) 1=m ( 1−m ) δikl×δjklLij ( c ) At a given pair of SNPs i and j in the admixed population , Equation ( c ) establishes a relationship between the observed linkage disequilibrium Lij in a recently admixed population and ancestral population differentiation . One can expected the ratio ( part 2 ) in Equation c to be closer to 1 when the two reference ancestral populations contributed to the admixture of the related admixed population . Equation ( c ) is a total ancestry content ( AC ) at a pair of SNPs i and j . Let Iij denote the ration in Equation c , assuming a uniform ancestral proportion , and summing Equation ( c ) over all possible pairs of proxy ancestral populations , we can obtain the ancestry informativeness Iij of each pair of SNPs i and j as follows , Iij=14K∑k≠lδikl×δjklLij Let M be the total number of SNPs . For i ∊ {1 , … , M} , let Ni be the total number of pair-wise LD j with i , where j ≠ i , ∀ j ∊ {1 , … , M} within SNP i , we obtain the ancestry informativeness at SNP i as a weighted sum of Iij , Ii=∑j=1NiIijM . We applied this method to construct the AIMs panel for Xhosa , ‡Khomani and Sandawe . This approach of selecting ancestry informative markers ( AIMs ) is implemented in the PROXYANC program ( http://web . cbio . uct . ac . za/proxyanc/ ) . We estimated the pair-wise genome-wide level of relatedness using a previously described relatedness statistic [67] applied to a random selection of 2500 putatively unlinked SNP markers with minor allele frequencies between 0 . 3 and 0 . 5 . These SNPs were randomly selected across each chromosome , with a minimum spacing of 1 MB , to prevent inclusion of SNPs in strong linkage disequilibrium , which would violate the assumption of marker independence . Principal Component Analysis ( PCA ) was performed , using EIGENSOFT [68] , on the combined HapMap3 , HGDP , other African data from [26 , 34 , 39 , 40] and southern African genotypes , which included a total of 50K SNPs shared between these different panels . In addition to the PCA analysis , an FST matrix using the smartpca program was generated . Admixture analysis [68 , 69] was performed on combined panels based on 900K SNPs using the ADMIXTURE program [69] . To evaluate the genetic relationships among the above populations , we used the TreeMix software [40] to infer the structure of a graph from genome-wide allele frequency data and a Gaussian approximation to genetic drift . Furthermore , to identify some aspects of ancestry not captured by the tree , we also examined the residuals of the model’s fit and sequentially added the migration events to the tree . We also used copy number variants as a population marker in an additional population structure analysis , but only for HapMap3 and southern African samples for which the intensity data ( CEL files ) necessary for copy number calling were publicly available . Copy number variants , detected with a Hidden Markov model that identifies novel copy number variation [55] , were preferred over previously described copy number polymorphisms , since these are affected to a lesser extent by ascertainment bias . We randomly selected a total of 2869 copy number variable positions , corresponding to 1 marker every 1Mb , across all chromosomes and specified copy number alleles as either a deletion , normal or duplicated state dependent on the copy number state called in the Birdseye algorithm [55] . We only selected simple copy number variants consisting of either a deletion or duplication , but not both . Here , we used all available southern African population data , including HER , SAN , XHS , XHS , LWK , BUS , ZUL , SAW , a Niger-Congo-speaking population ( YRI ) and a non-African population , which included CEU . We made use of the Haversine formula to compute the geographic distance ( in kilometre ) between pairwise populations based on great circle distances using the way points between continents . The way-points used are Egypt ( 29 . 998392 , 30 . 999751 ) and Turkey ( 41 . 015472 , 27 . 986336 ) . Thus , we computed the correlation between FST and Geographic distance using a linear regression equation as FST= 1 . 298 × 10−5× Geographic distance + 1 . 709 × 10−2 We analysed the scatter plot of the relationship between FST and geographic distance . To address this , we computed the perpendicular distance between each point and the regression line . This enabled us to define outliers as points whose distance to the regression line is greater than or equal to 0 . 05 units . To minimize deviation from the normality assumption , SNPs with minor allele frequencies < 0 . 05 are excluded . Thus , at a given locus i , the difference ( pik−pil ) between observed variant allele frequencies of two populations , k and l , can be approximated as a normal distribution under neutral drift with mean 0 and variance [60] p ( 1−p ) ( 2FST+1Nk+1Nl ) , ( d ) Where FST is the genetic distance between the population k and l . To avoid overestimating the degree of differentiation at single SNPs due to sample size difference , we used the estimator of FST in by Bhatia et al ( 63 ) . Nk and Nl are total variant allele counts in each population , and p is the ancestral allele frequency that is commonly approximated as the average of the two observed variant allele frequencies . Similar to [60] , we test unusual difference in allele frequency Ukl from population k and l as follows t Ukl1= ( pik−pil ) 2p ( 1−p ) ( 2FST+1Nk+1Nl ) , ( e ) Ukl2= ( pik−pil ) 2p ( 1−p ) . ( f ) Equations e and f are the χ2 distributed with 1 degree of freedom ( d . o . f ) , and can be applied to unrelated ( Equation b ) and related samples ( Equation c ) , respectively . An excess of large values of the χ2 statistic indicate deviations from the null model equation ( Equation e and f ) , suggesting the action of natural selection [60] . We applied this method to the data from the Xhosa population using Ju\’hoansi and Yoruba as ancestral populations . We also applied this method to KHO and SAW using KHS , CEU and YRI populations . All gene annotations and associated diseases were obtained using both the GeneCards and MalaCards databases [44 , 53] . We used LAMP-LD to infer locus-specific ancestry in admixed populations [46] . The model in LAMP-LD leverages the structure of linkage disequilibrium in the proxy ancestral populations . LAMP-LD achieved highest accuracy in both simulation and real data in the study of Puerto Rico and Mexico populations [43] . Here , we applied LAMP-LD to infer local ancestry in three potential southern African populations , including KHO , XHS and SAW . Following the population structure result and the proxy ancestry selection approach developed in PROXYANC [45] , YRI , KHS and CEU was selected as reference ancestral populations from a pool of Bantu-speaking , Click-speaking and European populations , respectively . We obtained phased haplotype data by running Beagle software [70] on KHS , CEU and YRI data . To estimate the distribution of genetic contributions of ancestries to XHS across the genome , we used haplotypes of 80 YRI and 80 KHS . In addition , the haplotypes of 80 YRI , 80 CEU and 24 KHS were used to compute the locus-specific genetic contributions to KHO and SAW using the AIMs panel . Admixed populations provide special opportunities for investigating recent selection . Prior to admixing , the ancestral populations have been isolated geographically , and their genomes may have evolved in distinct environments . Migration of previously isolated populations may have brought individuals of the ancestral populations into an unusual environment , and may consequently introduce life-style changes or changes in pathogens they are exposed to . This type of selection may differ from that faced by stationary populations , for which the local environmental changes may occur gradually , allowing for rare advantageous alleles to increase in frequency [43] . Here , we adopted an approach to detect ancestral signatures of selection by looking in an admixed population for genomic regions that exhibit unusually large deviations in ancestry proportions compared with what is typically observed elsewhere in the genome . Given the genome-wide ancestral proportions , αk , from ancestral populations k ∊ {1 , … , K} in N samples of an admixed population , let φki , m be the estimated locus-specific ancestry of individual i at genetic marker m ∊ {1 , … , M} , from the kth ancestral population . We computed the deficiency or excess of ancestry , at each SNP using the estimated admixture proportion as a baseline . We thus define the deficiency/excess of ancestry from ancestral population k at marker m as , δkm= ( 1N∑φki , m ) −αk=φ¯km−αk where φ¯km is the average locus-specific ancestry at SNP m . δkm can be approximated as a normal distribution under neutral drift with mean 0 and empirical variance , derived from the distribution of φki , m values among the N individuals [43 , 51] . We can fit a chi-square on φki , m as follows , Zkm= ( δkm ) 2var ( φki , m ) is a χ2 with 1 degree of freedom . A large value of the chi2 statistic indicates deviations from the null model and 4 standard deviations above ( excess ancestry ) or below ( deficiency ancestry ) the genome-wide average , suggests the action of natural selection post-admixture [51] . Summing-up the equation above over all SNPs assigned to a gene , we obtain the deficiency/excess of ancestry at the gene level . This allows us to assess the statistical significance of a deficiency/excess of ancestry at the SNP and gene level . To assess unusual difference in deficiency/excess of ancestry between a pair of ancestral populations given SNP m ∊ {1 , … , M} within a gene , we compute t˜kl=∑ ( ( δkm−δlm ) 2[var ( φki , m ) +var ( φli , m ) ]/N ) Which is a two-sample t-statistic with M − 2 degrees of freedom , assuming equal sample size N . For a pair of populations , k ≠ l ∊ {1 , … , K} , we compute the overall unusual difference in a deficiency/excess of ancestry , t˜=∑∑ ( ( δkm−δlm ) 2[var ( φki , m ) +var ( φli , m ) ]/N ) In order to summarize the types of loci and explore the potential adaptive genetic architecture implicated by our genome-wide selection scans , we identified all protein coding genes within 40 kb downstream or upstream of SNPs showing signatures of selection . To achieve this , we downloaded genomic coordinates for all genes from the NCBI ftp-server ( ftp://ftp . ncbi . nih . gov/ ) , retaining only entries for the human reference sequence and protein-coding genes . We updated genomic coordinates to the latest assembly using the Lift-Over tool on GALAXY ( https://main . g2 . bx . psu . edu/ ) . We obtained the genomic predicted human genes from the GeneCard database [44] . We investigate the roles of genes and cells in disease processes using the MalaCard database [44; 53] . | Genome-wide analysis of human populations is useful in shedding light on the evolutionary history of the human genome , with a wide range of applications from reconstructing past associations between different population histories to disease mapping . In this manuscript we report on the application of genome-wide data to southern African populations and the identification of genome-wide signatures of selection pre- and post-admixture . Several signals of selection , before and after admixture , were identified , some of which involved loci associated with human diseases , including malaria , influenza , tuberculosis and HIV/AIDS . These results may reflect adaptations of southern African populations to infectious diseases . Consistent with previous studies , this study highlights the significance of the San in the genetics of human populations , as they are distinct from the other populations in many respects i . e . haplotype structure , locations of recombination hotspots , copy number and population structure . Furthermore , our study demonstrates the admixture of the San , Bantu-speaking populations and populations of Eurasian ancestry in some of the southern and eastern African populations . It illustrates the value in correcting for this stratification in future genome-wide association studies , and suggests that a future admixture mapping in these populations would likely be warranted and successful . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | A Genomic Portrait of Haplotype Diversity and Signatures of Selection in Indigenous Southern African Populations |
The aim of the present study was to determine whether there is a correlation between phylogenetic relationship and inflammatory response amongst a panel of clinical isolates representative of the global diversity of the human Mycobacterium tuberculosis Complex ( MTBC ) . Measurement of cytokines from infected human peripheral blood monocyte-derived macrophages revealed a wide variation in the response to different strains . The same pattern of high or low response to individual strains was observed for different pro-inflammatory cytokines and chemokines , and was conserved across multiple human donors . Although each major phylogenetic lineage of MTBC included strains inducing a range of cytokine responses , we found that overall inflammatory phenotypes differed significantly across lineages . In particular , comparison of evolutionarily modern lineages demonstrated a significant skewing towards lower early inflammatory response . The differential response to ancient and modern lineages observed using GM-CSF derived macrophages was also observed in autologous monocyte-derived dendritic cells and murine bone marrow-derived macrophages , but not in human unfractionated peripheral blood mononuclear cells . We hypothesize that the reduced immune responses to modern lineages contribute to more rapid disease progression and transmission , which might be a selective advantage in the context of expanding human populations . In addition to the lineage effects , the large strain-to-strain variation in innate immune responses elicited by MTBC will need to be considered in tuberculosis vaccine development .
High-throughput sequence analysis has allowed reconstruction of the evolution of human Mycobacterium tuberculosis Complex ( MTBC ) , differentiating the bacteria into six main phylogenetic lineages [1] , [2] . Three lineages , including two whose members are known as M . africanum [3] , which branched off from a common ancestor at an early stage of evolution , are referred to as evolutionarily “ancient” lineages; three separate evolutionarily “modern” lineages diverged at a later time point [1] . It is proposed that the branches reflect the history of human migration out of Africa , with the current geographic distribution of the different lineages being determined by the expansion and migration of their corresponding host populations . This phylogeny provides a rational framework to assess whether the genotypic diversity of MTBC is associated with diversity in biological phenotype [4] , [5] . Several studies suggest that this may be the case [5] , [6] . A comparison of pulmonary TB with TB meningitis in Vietnam demonstrated that strains belonging to the modern Euro-American lineage were significantly less likely to cause extra-pulmonary disease [7] . In contrast , the modern Beijing/W lineage was found significantly associated with extra-thoracic TB in comparison with non-Beijing/W lineages , though no difference was found when looking at other read-outs of virulence such as the amount of cavitation [8] . A recent study in Madagascar found that infection with an ancient lineage induced a significantly higher immune response , measured as interferon-γ production by peripheral blood T cells [9] . Finally , a series of studies , including the report from Madagascar , have described a reduced immune response to members of the “Beijing family” ( part of the modern East Asian Lineage 2 ) , and its association with rapid progression to severe disease in humans and experimental animals [9] , [10] , [11] , [12] , [13] , [14] . Two studies that analyzed individual isolates involved in TB outbreaks concluded that a low inflammatory response was linked to increased virulence [10] , [15] . The rationale is that a reduction in innate immune recognition will result in a delay in engagement of the adaptive response , providing the pathogen with a significant advantage during the early stage of infection . The low inflammatory phenotype of M . tuberculosis HN878 , a member of the Beijing family implicated in an outbreak in Texas , was reversed by disruption of the gene encoding an enzyme required for biosynthesis of a phenolic glycolipid molecule ( PGL ) [11] . However , it was shown later that the production of PGL was variable across strains from the Beijing/W lineage [16] . Moreover , the role of this particular glycolipid in the virulence of HN878 could not be reproduced by restoring its production in the genetic background of another modern strain belonging to Lineage 4 , highlighting a rather confusing inter- and intra-lineage diversity in the molecular mechanisms of M . tuberculosis pathogenicity . In contrast , the low inflammatory phenotype of M . tuberculosis CAS , responsible for an outbreak in Leicester , was linked to a chromosomal deletion and could be reversed by restoration of the functional gene [15] . Several other studies have described differences in the inflammatory response induced by different isolates of M . tuberculosis [12] , [13] , [17] , [18] , [19] . In the present study , we used a collection of strains covering the global genetic diversity of MTBC to test the hypothesis that inflammatory phenotype would be linked to genotype .
To test for a link between genotype and inflammatory phenotype , we selected 26 isolates representative of the global diversity of human MTBC from a well-characterized clinical strain collection [1] , [4] , [20] ( Figure 1 ) plus two laboratory adapted strains as references ( M . tuberculosis H37Rv and M . bovis BCG Pasteur ) and measured their ability to induce production of inflammatory cytokines by human GM-CSF monocyte-derived macrophages ( T1-MDMs ) [21] . Figure 2 shows cytokine levels from culture supernatants harvested 24 hours after infection with each of the strains for two human donors , and highlights three important observations . First , there were clear differences in the level of pro-inflammatory cytokines produced by a single donor in response to different strains; ranging from a few hundred picograms to several nanograms . Second , although the absolute amount of cytokine varies between individual donors ( results in Figure 2 are plotted against two separate y-axes ) , the relative hierarchy of low and high responses is the same in the two donors . This is further illustrated in Figure 3 , summarizing results from eight donors , with strains ranked according to their ability to induce a cytokine response after median normalization of the dataset . In addition to robust consistency across different human donors , a similar hierarchy was observed when the same strains were used to stimulate bone-marrow-derived macrophages from Balb/c mice ( Figure S1 ) . A third observation is that the different pro-inflammatory cytokines – IL-6 , IL-12p40/p70 and TNFα – all showed a similar pattern . Again , this is particularly clear when analyzed across the panel of eight donors , showing a highly significant correlation between production of IL-6 and IL-12p40/p70 for example ( Figure 3C; Spearman rank correlation coefficient = 0 . 988 , P<0 . 001 ) . We next compared the pro-inflammatory responses across the main MTBC lineages . Combining results from the eight individual donors after median normalization revealed higher heterogeneity in IL-6 production in response to Lineage 5 , 6 and 1 ( Figure 4A ) . However , there was overall a statistically significant effect of lineage on the cytokine levels observed ( Kruskal-Wallis rank test , P<0 . 001 ) . To look for higher order grouping of the lineages we performed a principal component analysis ( PCA ) using the genetic variation among the strains . PCA distinguished three major groups: strains belonging to Lineages 5+6 ( M . africanum ) , strains belonging to Lineage 1 , and strains belonging to the Lineages 2+3+4 ( these three lineages have been referred to as evolutionarily “modern” based on previous work [1] , [22] ) ( Figure 4B ) . These three major groupings were consistent with the most recent genome-based MTBC phylogeny published earlier this year [2] and shows that the “modern” lineages are more genetically homogenous than “ancient” lineages . Comparing the level of cytokine induction across these three groups showed that the modern group of strains consistently induced a lower IL-6 response when compared to Lineage 5+6 or Lineage 1 ( Figure 4C ) , though Lineage 1 had an intermediate phenotype reflecting again the higher heterogeneity of the ancient group when compared to the modern strains . Even so , combining the lineages according to the ancient/modern grouping revealed an overall difference in inflammatory phenotype , with strains from the modern lineages always eliciting significantly lower levels of IL-6 ( Figure 4C ) and other cytokines and chemokines such as IL-12p40/p70 , TNFα , IL-15 , MIP-1α , CCL5 and others ( Figure 5 ) . Therefore for the rest of the study we decided to keep the "ancient"/"modern" dichotomy to highlight the different behavior of the later group when compared with the other lineages . However , relevant figures splitting the data in three groups ( "M . africanum" , Lineage 1 and "Modern" ) are made available as supplementary material ( Figure S2 A-E ) . As a further test of the reproducibility of the differential inflammatory response , experiments were repeated using a second batch of the same strains of MTBC separately cultured and quantified . While a few individual isolates showed evidence of batch variation in inflammatory phenotype , the overall pattern of a lower response to the modern lineage was maintained as shown by a statistically significant correlation test ( Spearman correlation test , P<0 . 05 ) ( Figure S3 ) . There was a differential increase in the production of pro-inflammatory cytokines as the infection progressed , and the difference between ancient and modern lineages observed at the 24-hour time point had markedly diminished by 72 hours ( Figure 6; Mann-Whitney U test , P<0 . 05 ) . Thus , the reduced response to the modern strains is due to a delay in the kinetics , rather than a complete inhibition of the immune response . The differential response pattern was not observed in infection experiments using unfractionated peripheral blood mononuclear cell ( PBMC ) preparations in place of differentiated monocytes . In these experiments , we observed a lower amount of TNFα but with no significant difference between strains from the ancient and modern lineages , and a statistically significant higher amount of IL-6 in comparison to T1-MDMs ( Figure 7A; Mann-Whitney U test , P<0 . 05 ) . Replacement of GM-CSF by M-CSF during monocyte differentiation generated a “Type 2” macrophage population characterized by a lower level of production of pro-inflammatory cytokines , along with increased expression of IL-10 [21] . Figure 7B illustrates the differences between Type 1 ( T1-MDMs ) and Type 2 ( T2-MDMs ) polarization in response to LPS stimulation . Comparative flow cytometry analysis is shown as supplementary material ( Figure S4 ) . Infection of T2-MDMs with the different strains of MTBC resulted in lower production of IL-6 and IL-12p40/p70 as compared to T1-MDMs , with a significant difference between lineages only for IL-6 ( Figure 7C , D ) . In contrast to T1-MDMs , a consistent IL-10 response was induced during infection of the T2-MDMs . The IL-10 response was also variable between isolates , but no significant difference was observed when comparing ancient and modern lineages ( Figure 7E ) . IL-10 is an anti-inflammatory cytokine that has been shown to act as an auto-regulatory inhibitor of pro-inflammatory cytokine production by human monocytes [23] , [24] . Notably , IL-10 production has been associated with the anti-inflammatory phenotype of a recent outbreak strain belonging to the modern lineage [15] . IL-10 is produced and effective at very low concentrations . To test whether the differential inflammatory response observed in T1-MDMs might be influenced by variations in production and consumption of IL-10 that were not detected by ELISA , we repeated infection experiments in the presence of anti-IL-10 blocking antibodies . Consistent with the idea that IL-10 is produced and acting at very low levels , IL-10 blockage resulted in a systematic increase in pro-inflammatory cytokines TNFα and IL-6 , but not in IL-12p40/p70 . However , it did not affect the differential between ancient and modern lineages ( Figure 8 ) . We also matured monocytes from three different donors in the presence of GM-CSF and IL-4 in order to generate monocyte-derived dendritic cells ( MD-DCs ) [21] . Compared with autologous T1-MDMs , MD-DCs significantly down-regulated CD14 and expressed higher levels of MHC class II molecules and CD86 ( Figure S4 ) . The response of MD-DCs to infection with the different strains resembled that of autologous T1-MDMs . Although the absolute amounts of immune mediators were generally lower , the trend towards weaker responses to the modern lineage was preserved for a largely overlapping panel of cytokines and chemokines but also IL-1β , IL-1RA , CXCL8 and GM-CSF ( Figure 9 ) .
Using a panel of strains representative of human MTBC genetic variability , we have found that genetically diverse strains of MTBC vary widely in their induction of an early inflammatory response during infection of human macrophages , and that these differences are linked to MTBC lineages . Overall there was a significantly lower response to evolutionarily modern lineages as compared to ancient lineages . Previous reports have described differences amongst M . tuberculosis isolates in their inflammatory phenotype [10] , [12] , [13] , [15] , [18] , [19] but the present study is the first to link this to MTBC phylogeny . Consistent with previous reports , we observed a higher inflammatory phenotype for the laboratory strain H37Rv in comparison to the Beijing strain , HN878 [11] . We found that these differences were robust and reproducible using different batches of bacteria to infect monocyte-derived macrophages and autologous dendritic cells across different human donors . We were concerned that an experimental bias could result from inaccuracies in quantifying bacterial preparations . Our quantitation is based on measurement of colony forming units and , while we made every effort to use comparable actively growing cultures , to minimize clumping artifacts and to repeat multiple measurements , errors could arise if there is a variation between strains in the rate of accumulation of dead cells during culture . While it is difficult to rigorously exclude this possibility , we consider it an unlikely explanation for our results since ( a ) differences in accumulation of dead cells would need to span several orders of magnitude , and ( b ) in some experimental systems – infection of PBMCs or T2-MDMs – the same preparations showed opposite or no differences in cytokine response . Convergence in responses of T1-MDMs to different strains at later time points is also consistent with equivalence between preparations . Although we did observe a link between inflammatory phenotype and the various bacterial genotype clusters , each of the phylogenetic lineages included strains that induced high and low levels of inflammatory cytokines . For example , strain N0024 belonging to Lineage 3 consistently elicited a stronger inflammatory response as compared to the other strains from the same lineage . Future comprehensive genetic and biochemical analysis of these strains will be performed with the aim of deciphering the origin of this difference . The variation observed within lineages suggests that differences in inflammatory phenotype cannot be explained simply by the presence or absence of a single molecular determinant , and one model to account for the dispersal of high and low inflammatory strains across the phylogenetic tree is that a diverse range of mutations might influence innate immune recognition and arise independently in different lineages . This model is consistent with the known complexity and diversity of mycobacterial products that have been shown to stimulate or inhibit inflammatory responses [11] , [25] , [26] , [27] , [28] . Considering the multiple families of cell-surface and intracellular receptors involved in mycobacterial recognition [29] , one might expect that the consequences of molecular changes that affect different ligands would depend on the exact receptor repertoire expressed by different donors [20] . However , the consistent hierarchy in inflammatory response that was observed across independent donors suggests that there is limited human variability in the initial immune response to genetically different mycobacteria . Our demonstration of a robust hierarchy in inflammatory phenotype within the different lineages of MTBC poses a challenge to our understanding at a molecular level of the microbiology and pathogenesis of tuberculosis as multiple mechanisms might converge towards similar phenotypic effects in distinct MTBC lineages . This observation may hold important lessons for development of new vaccines . In spite of the heterogeneity within MTBC lineages , we observed a statistically significant distribution towards more pro-inflammatory strains amongst members of the ancient lineages , and lower inflammatory responses to strains from the modern lineages . The low inflammatory phenotype of modern strains is in agreement with previous studies of individual Beijing strains [10] , [12] and other strains [15] belonging to the modern lineages . Our results are also consistent with the previous suggestion that a low inflammatory response may lead to a reduction in the adaptive response [9] . A possible model to rationalize this finding is that the respective high and low inflammatory responses could reflect different virulence strategies that emerged during the evolution of the ancient and modern lineages . The characteristic latency in TB has been suggested to represent an evolutionary adaptation to low host densities , with reactivation after several decades allowing the pathogen to access new susceptible birth cohorts [1] , [30] . By contrast , the low inflammatory response induced by evolutionary modern outbreak strains has been associated with an enhanced ability to cause early progressive disease [10] , [15] . Such a strategy may be an advantage in the context of high human population densities , where the number of susceptible hosts is large , and rapid lethality does not threaten to exhaust the pool of new uninfected hosts . We previously presented an evolutionary scenario for human TB based on population genetic analyses of multilocus sequence data [1] referred to recently as “the most well defined phylogeny of the MTB complex” [31] . According to this scenario , M . tuberculosis originated in Africa and accompanied early modern humans on their Out-of-Africa migrations . In those times , human populations were small , and M . tuberculosis might have benefited from the latency strategy [30] . During the last few hundred years , the three modern lineages of M . tuberculosis experienced strong population expansions as a consequence of the recent human population increases in Europe , India and China [1] . The overall lower inflammatory responses observed in the modern lineages of M . tuberculosis might be a consequence of their access to rapidly increasing numbers of susceptible hosts resulting in selection for faster progression to active disease . In support of this hypothesis , a study in the Gambia showed that members of the modern strain lineages were three times more likely than members of the ancient lineages to cause active TB in recently exposed contacts [32] . To our knowledge , this is the first time that the immune response to a particular infectious agent has been measured in a systematic manner by selecting representative strains belonging to the major human MTBC lineages and grasping the global M . tuberculosis genetic diversity , including notably M . africanum . As we show here , the combination of genotypic , phenotypic and epidemiological studies offers the potential for novel insights into the biology of this pathogen .
Mycobacterial cultures of clinical isolates were obtained from a single colony forming unit . One volume of a stationary phase culture of mycobacteria in Middlebrook 7H9 medium with ADC supplement ( BD Biosciences ) , 0 . 05% Tween-80 ( Sigma-Aldrich ) and in some cases sodium pyruvate 40 mM ( e . g . M . africanum strains [33] ) was diluted with 100 volumes of the same medium in the absence of detergent and incubated for 10 days at 37°C . Gentle culture dispersion was performed manually every 48 h . Mycobacteria were pelleted , supernatants discarded and pellets dispersed by manual shaking for 1 min with equal volumes of 2–3 mm glass beads . Mycobacteria were resuspended in PBS and centrifuged at 260 xg for 10 min to remove clumps . Cleared supernatants mostly composed of single particles [34] were supplemented with 5% glycerol and titrated on 7H11 agar plates complemented with OADC ( BD Biosciences ) and sodium pyruvate 40 mM before and after freezing and storage . Peripheral blood mononuclear cells ( PBMCs ) from healthy anonymous donors were isolated from buffy coats processed by the National Blood Services , Colindale , UK . PBMCs were prepared on a Ficoll-Paque density gradient ( Amersham Biosciences AB , Uppsala , Sweden ) by centrifugation ( 800 xg , 30 min at room temperature ) . Recovered PBMCs were washed twice with RPMI ( Gibco , Invitrogen ) and resuspended in RPMI/FCS ( 4% ) /methyl-cellulose ( 2% ) /DMSO ( 9% ) for gradual overnight freezing in a NalgeneTM Cryo 1°C container before storage in liquid nitrogen . Monocytes were selected from fresh or frozen PBMCs by magnetic cell sorting using CD14 microbeads ( Miltenyi Biotec , Auburn , CA , USA ) according to manufacturer's recommendations . Cell purity checked by flow cytometry was always >95% . Macrophages were differentiated from monocytes after 6 days of culture in the presence of recombinant human GM-CSF or M-CSF ( Peprotech Ltd ) for T1-MDMs or T2-MDMs respectively and monocyte derived dendritic cells ( MD-DCs ) in the presence of GM-CSF and IL-4 as previously described [21] . Cells were recovered after 15 min Trypsin/EDTA ( 2 mM ) treatment , resuspended in RPMI and 5% FCS , and evenly distributed at 8×104 to 1×105/well ( according to experiment ) in tissue culture treated 96 well plates or 12 . 5×103/well in 384 well plates before mycobacterial infection at a multiplicity of infection of 1∶1 unless specified otherwise . LPS stimulation was performed at a final concentration of 10 ng/ml . IL-10 blocking experiments were performed as described elsewhere [35] , antibodies were added prior to infection at a final concentration of 0 . 1 µg/ml . Mice were bred in the animal facilities of the National Institute for Medical Research and provided after being sacrificed in line with code of practice for the humane killing of animals under schedule 1 to the animals ( scientific procedures ) Act 1986 . Authors were not involved in the handling and/or sacrificing of live mice . Femurs from dead Balb/c mice were flushed with 1 ml complete medium ( RPMI1640 , 1 mM sodium pyruvate , 2 mM glutamine , 10 mM HEPES , 0 . 05 mM β-mercapthoethanol and 10% FCS ) using a 25 G needle . Cells were pelleted , red blood cells lysed for 5 mins with 10 ml 0 . 83% ammonium chloride , filtered through a 70 µM strainer and washed twice with complete medium before incubation in a CO2 incubator at 37°C for 4 days in 90 mm Petri dishes ( 4×106 in 8 ml complete medium containing 20% L-cell medium ) . On day 4 , 10 ml conditioned medium was added and cells were harvested on day 7 by removing supernatant and adding 5 ml PBS containing 2 mM EDTA to detach the macrophages . Recovered cells were pelleted and resuspended in complete medium and plated out as described for human monocyte derived macrophages . Cell supernatants were recovered at indicated time points , sterilised twice using 96 well filter plates ( 0 . 2 µm , Corning ) and stored at −20°C until analysis . IL-6 , IL-12p40/p70 , TNF-α and IL-10 were measured using either ELISA kits ( Peprotech ) or Luminex 30-plex kit ( Invitrogen ) following manufacturer's recommendations . Data analysis , correlation study , paired t-tests , Mann-Whitney U tests and Kruskal-Wallis rank test were performed using GraphPad Prism software and STATA s . e . m . version 10 . Without assuming a pre-defined distribution of the response tested , non-parametric statistical analysis has been used all across the study . Unless otherwise stated , we used the individual measures for each combination of donor and strain for all the statistical analysis ( n = 200 ) . Principal component analysis was conducted using STATA s . e . m . version 10 with the polymorphic positions found in Hershberg et al . 2008 for the strains used in the present study . | Mycobacterium tuberculosis is a long-standing human pathogen spread by aerosol transmission between individuals interacting in close social groups . It can be anticipated that the evolution of M . tuberculosis will parallel the evolution of human societies , and the phylogeny as determined by whole genome sequencing of clinical isolates is indeed consistent with emergence of the pathogen with modern humans in Africa and its subsequent dissemination along routes of human migration and trade . The present study was designed to test the hypothesis that the genetic diversity of M . tuberculosis isolates would be reflected in a corresponding diversity in their biological properties . In particular , we explored the interaction of different isolates with the innate immune system , which plays important contrasting roles in initial resistance to infection and in disease transmission . We observed a difference in the innate immune response when we compared isolates belonging to “modern” lineages that have evolved amongst high-density populations in regions of recent massive demographic expansion , with isolates belonging to “ancient” lineages selected in older low-density human populations . Our results provide insights into host-pathogen co-evolution and into fundamental mechanisms underlying the pathogenesis of M . tuberculosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"infectious",
"diseases/bacterial",
"infections",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"immunology/innate",
"immunity"
] | 2011 | Human Macrophage Responses to Clinical Isolates from the Mycobacterium tuberculosis Complex Discriminate between Ancient and Modern Lineages |
Mutations conferring resistance to antibiotics are typically costly in the absence of the drug , but bacteria can reduce this cost by acquiring compensatory mutations . Thus , the rate of acquisition of compensatory mutations and their effects are key for the maintenance and dissemination of antibiotic resistances . While compensation for single resistances has been extensively studied , compensatory evolution of multiresistant bacteria remains unexplored . Importantly , since resistance mutations often interact epistatically , compensation of multiresistant bacteria may significantly differ from that of single-resistant strains . We used experimental evolution , next-generation sequencing , in silico simulations , and genome editing to compare the compensatory process of a streptomycin and rifampicin double-resistant Escherichia coli with those of single-resistant clones . We demonstrate that low-fitness double-resistant bacteria compensate faster than single-resistant strains due to the acquisition of compensatory mutations with larger effects . Strikingly , we identified mutations that only compensate for double resistance , being neutral or deleterious in sensitive or single-resistant backgrounds . Moreover , we show that their beneficial effects strongly decrease or disappear in conditions where the epistatic interaction between resistance alleles is absent , demonstrating that these mutations compensate for the epistasis . In summary , our data indicate that epistatic interactions between antibiotic resistances , leading to large fitness costs , possibly open alternative paths for rapid compensatory evolution , thereby potentially stabilizing costly multiple resistances in bacterial populations .
Bacteria evolve to endure antibiotics through the acquisition of genes or chromosomal mutations that confer resistance to the drugs [1–3] . Resistance mutations are widespread in clinical [2 , 4 , 5] and environmental [6 , 7] bacterial populations , providing a reservoir that can be transmitted by horizontal gene transfer [3 , 8] . Moreover , as microbes become resistant to a specific drug , the subsequent use of alternative antibiotics might select for additional resistances , thus leading to the increasing menace of multidrug-resistant strains [2 , 9–11] . This is a prevalent scenario in Staphylococcus aureus , E . coli , or Mycobacterium tuberculosis , for instance , in which multiple resistances pose a serious threat to human health [9 , 12–14] . Thus , disclosing the evolutionary factors governing the maintenance of multiple resistance is key for designing effective treatments . Chromosomal mutations conferring resistance , the focus of this study , are typically deleterious in the absence of antibiotics [15 , 16] , since often they affect proteins involved in essential cellular machinery , such as the ribosome ( i . e . , streptomycin ) [17] or the RNA polymerase—RNAP— ( i . e . , rifampicin ) [18–20] . Nevertheless , bacteria can curb this deleterious effect by acquiring compensatory mutations [15 , 21–23] . This allows antibiotic resistance to become stabilized in the population while paying little or even no cost [24] . Acquiring additional mutations to overcome the fitness cost of resistance is also more likely to occur than the genetic reversion of the resistance mutation , since the range of targets for compensation is much broader [24 , 25] . Compensation for the cost of resistance has been widely described both in clinical [11 , 19] and laboratory [17 , 18 , 20] settings . The dynamics of compensatory adaptation depends on population parameters , such as bottlenecks [17] and mutation rate [21] , and on the distribution of fitness effects of compensatory mutations , which depends on the genetic background [22] . Studies of compensation have focused on single resistances ( e . g . , [17 , 20 , 23 , 26] ) . However , resistance mutations often interact such that the fitness cost of multiple resistance tends to differ from that expected given the effects of each resistance [24 , 27] . These epistatic interactions might originate from the fact that resistance mutations affect interconnected functions in the cell [2] . Thus , cooccurring resistance mutations can lead to synergistic or antagonistic functional interactions [24–26] . This leads us to hypothesize that compensatory evolution of multiple-resistant bacteria may substantially differ from that of single-resistant strains . This may be especially important for chromosomal resistance mutations that exhibit negative epistasis , since the proteins involved in the interactions between the resistance mutations ( which potentially contribute to the epistasis ) could become targets for compensation . In that scenario , additional mutations beyond those that compensate for each individual resistance may be expected to occur . Thus , we further postulate that mutations that specifically compensate for epistasis should emerge during the evolution of multiple-resistant bacteria . To test these hypotheses , we followed the compensatory adaptation of single- and double-E . coli–resistant strains in antibiotic-free media . We chose three founder genotypes carrying chromosomal resistance mutations as a case study: RpoBH526Y ( conferring resistance to rifampicin ) , RpsLK43T ( conferring resistance to streptomycin ) , and a strain harboring the two mutations ( RpoBH526Y RpsLK43T ) , which show negative epistasis , causing a strong decrease in bacterial fitness [24 , 28] . To dissect the process of compensation for the mentioned founder genotypes , we analyzed the dynamics of the frequencies of neutral markers during compensatory evolution , performed whole-genome sequencing of evolved clones and allelic reconstruction of putative compensatory mutations in all genetic backgrounds . Our results demonstrate faster compensation in low-fitness double-resistant bacteria and unveil the identity of mutations that compensate for the negative epistasis between resistances . These mutations are advantageous only for double-resistant genotypes , i . e . , they are either neutral or deleterious for single resistant and sensitive clones , as well as in environments where the epistasis between resistances is absent .
To study the dynamics of compensation and determine its pace , we performed experimental evolution in an antibiotic-free medium of six E . coli clones representing three founder genotypes , each carrying either of two neutral markers ( Cyan fluorescent protein [CFP] or yellow fluorescent protein [YFP] , S1 Fig ) : two carrying a RifR ( RpoBH526Y ) allele , two carrying a StrR ( RpsLK43T ) allele , and two carrying both resistant alleles RifR StrR ( RpoBH526Y RpsLK43T ) . The single resistances cause different fitness costs ( 0 . 06 ± 0 . 001 for RpoBH526Y and 0 . 03 ± 0 . 01 for RpsLK43T ) and generate strong negative epistasis—0 . 27 ± 0 . 01 >> ( 0 . 06 ± 0 . 01 ) + ( 0 . 03 ± 0 . 01 ) —in the double mutant ( see S2 Fig ) . The neutral markers allow us to readily identify the emergence of compensatory mutations , as the frequency of a neutral marker is expected to rapidly increase when adaptive mutations sweep through evolving populations . Conversely , a marker should decrease in frequency when a beneficial mutation occurs in the subpopulation carrying the other marker . The number of generations , together with the relatively large population size and the mild bottlenecks during the propagations ( see materials and methods ) , ensure that adaptation will mostly be driven by positive selection [29] . However , it is possible that some alleles that increase in frequency are neutral ( or even slightly deleterious ) and arise through genetic hitchhiking with beneficial mutations [30 , 31] . We followed the changes in marker frequencies over 22 days ( ~180 generations ) in 12 independently evolving RifR populations , 12 StrR populations , and 24 RifR StrR populations ( Fig 1A–1C ) . We observed that , for the double-resistant RifR StrR population ( Fig 1C ) , markers deviate from their initial frequency faster and with steeper slopes than for either of the single-resistant bacteria , RifR ( Fig 1A , One-way ANOVA with Tukey’s Honest Significant Difference [HSD] correction , p < 0 . 0001 ) or StrR ( Fig 1B , One-way ANOVA with Tukey’s HSD correction , p = 0 . 001 ) . In only 4 days of evolution , all double-resistant populations show signature of evolutionary adaptation , indicated by the changes in the frequency of the markers , that either reach frequencies close to fixation ( >0 . 95 ) or show strong fluctuations . The latter indicate competition between clones carrying adaptive mutations ( clonal interference ) in both CFP and YFP backgrounds . Rapid and strong changes in marker frequency dynamics are expected for genetic backgrounds with low fitness—as it is the case for the double-resistant clones—which have been reported to adapt faster [32–35] . In single-resistant clones , near fixation of a marker is observed in a minority of the populations ( three out of 12 RifR and 1 out of 12 StrR populations ) and only much later ( 8–10 days ) . Interestingly , the dynamics of RifR populations show no strong signs of clonal interference , whereas this phenomenon can be observed in a few StrR populations . This suggests that more targets for adaptation may be available for StrR than for RifR , albeit with smaller effects . To understand which process , mutation or selection , is responsible for the large differences in the dynamics of compensation of double versus single resistances , we measured the competitive fitness of the evolving populations along the adaptive process ( defined by the increase in competitive ability of the evolving populations against a nonfluorescent sensitive strain , see materials and methods ) . A clear increase in competitive fitness is observed in all resistant backgrounds along the propagations ( t-test p = 1x10−6 for RifR , p = 6x10−6 for StrR , and p < 2x10−16 for RifR StrR , at day 22 of evolution , Fig 1D–1F ) , but the dynamics of fitness change differ among the different evolved resistant populations . The first 5 days show a strong and fast competitive fitness increase in RifR StrR bacteria compared to the other backgrounds ( Fig 1D–1F , ANCOVA , F ( 2 , 70 ) = 45 . 08 , p < 0 . 0001 , with 0 . 048 ± 0 . 003 fitness increase per day , between day 0 and day 5 , significantly higher than RifR [0 . 01 ± 0 . 004] or StrR [0 . 01 ± 0 . 004] ) backgrounds . After day 5 , the RifR populations have a more pronounced fitness increase along time compared to the other two backgrounds , in which competitive fitness seems to approach stabilization ( generalized linear mixed effects model [GLMEM] , Chi2{2} = 362 . 9 , p < 0 . 0001 , with 0 . 006 ± 0 . 0006 competitive fitness increase per day between day 5 and day 22 , versus StrR [0 . 0015 ± 0 . 0006] and RifR StrR [0 . 0017 ± 0 . 0004] ) . Overall , the high initial cost of the double resistance ( compare black dots with red dashed line in Fig 1F ) is , on average , largely mitigated at day 5 ( 0 . 24 mean increase in competitive fitness , calculated as the difference between the competitive fitness of the evolved populations at day 5 and that of the founder double mutant , standard deviation [SD] = 0 . 06 , across the 24 populations ) . Interestingly , the fitness increase of double-resistant bacteria appears to slow down at different fitness levels in the different populations . For instance , in population six , no significant fitness increase above 0 . 32 is observed after day 5 ( linear regression’s differential fitness [dW] = y = −0 . 0005x + 0 . 32; R2 = 0 . 024 ) , whereas for population 12 , no significant gain above 0 . 15 is detected from day 9 onwards ( dW = −0 . 001x + 0 . 149; R2 = 0 . 08 , see Fig 1F and also S3 Fig ) . This suggests that the first step of adaptation may involve mutations with different fitness effects , which might condition the following adaptive steps . The faster fitness increase in double-resistant bacteria can result from a higher rate of acquisition of beneficial mutations ( due to , for instance , a larger target size ) and/or from acquisition of mutations with stronger beneficial effects . To infer the rate and the distribution of effects of arising beneficial mutations in each of the backgrounds , we used a modified version of the One Bi-Allelic Marker Algorithm ( OBAMAv2 ) [36] that considers both the marker dynamics and the fitness trajectories along the adaptive walk ( see materials and methods ) . OBAMAv2 infers that the rate of acquisition of beneficial mutations ( U ) is not significantly different between the double resistant and the RifR populations ( U ~ 3x10−6 per cell and generation ) , being higher in the StrR background ( U ~ 10−5 per cell and generation ) ( Fig 2A ) . This suggests a larger target size for compensation of StrR . However , the mean selective effect of beneficial mutations is higher for the RifR StrR background than for either of the single-resistance backgrounds ( 0 . 18 for the double versus 0 . 1 for the RifR and 0 . 05 for the StrR ) ( Fig 2B ) . This could be due to the acquisition of similar mutations causing different effects across backgrounds with different initial fitnesses , as has been previously observed [33 , 34 , 37] . It may also be that the mutations underlying compensation could be different in single- and double-resistant bacteria . In the latter case , double-resistant bacteria would have access to mutations with higher mean selective benefit , which could potentially include compensatory alleles for the high cost resulting from the epistasis between RifR and StrR . To identify the compensatory mutations emerging in each of the studied founder genotypes , we performed whole-genome sequencing of pools of evolved clones . We pooled one clone from each evolved population carrying the fluorescence marker with the highest frequency at the end of the experiment ( Fig 1A–1C ) . This allows capturing a broad sample of the compensatory landscape ( Fig 3 ) . We found mutations affecting compensatory targets previously described ( 10 out of 16 allelic changes in RifR , 2 out of 15 in StrR , and 16 out of 57 in RifR StrR ) , such as rpoB itself , rpoA , and rpoC for rifampicin resistance and rpsE , rpsD , and tufA for streptomycin resistance ( Fig 3 ) . Indeed , several mutations found , affecting ribosomal genes , have been previously described as compensatory for the streptomycin resistance allele RpsLK43N in Salmonella Typhimurium [17] ( RpsEA110V in the StrR background and RpsDD50Y in the RifR StrR background ) or affect similar residues ( RpsDT86A , RpsET103P , and RpsEG109R in the RifR StrR background ) . In addition , mutations in rpoA and in rpoC found in the RifR background ( RpoAT196I and RpoCH450P , respectively ) were also previously identified as compensatory for the allele RpoBH526Y [22] . We identified some putative compensatory targets that are common to single- and double-resistant backgrounds . At the level of functional category , we observed overlap in genes encoding membrane proteins ( i . e . , ybjO and nanC in RifR; cdgI and yojI in StrR; ompF and ydiY in RifR StrR ) and ribosomal proteins ( rplL in RifR StrR and rplI in StrR ) . At the gene level , there is a high parallelism , mostly involving the known compensatory targets previously mentioned , such as rpoA , rpoB , rpoC , rpsE , and tufA . At the nucleotide level , we found three mutations common to the StrR and the double mutant: an insertion sequence [IS] element in a known hotspot for IS elements upstream from the flhDC operon [39] , a nonsynonymous change in tufA , and a 6 bp deletion in rplL . A single compensatory mutation was found in common between RifR and the RifR StrR backgrounds: a reversion in rpoB ( Fig 3 and S1 Table ) . Reversions are typically rare [21 , 40] , but clinically relevant , since bacteria regain sensitivity to the antibiotic . To determine the extent to which reversion to sensitivity occurred in our evolving populations , we phenotyped clones ( n > 40 ) from each evolved population at day 22 . We streaked each clone sequentially on both the medium supplemented with antibiotic ( s ) , at a concentration similar to where the founder genotypes were originally selected ( i . e . , 100 ug/ml of rifampicin , streptomycin , or both drugs ) , and in the antibiotic-free medium . We define “sensitivity” as the inability of a clone to grow in the presence of the antibiotic but able to grow in the drug-free medium . Remarkably , we found that , in 6 out of 24 populations of the RifR StrR background , sensitivity to either rifampicin or streptomycin emerged . We also found the emergence of sensitive clones in the single-resistance evolved populations: in the RifR background , in two populations , fixation of sensitive clones was detected; in the StrR background , the frequency of streptomycin-sensitive clones reached 64% ( 58%–71% , 95% confidence interval [CI] ) in one population and 69% ( 62%–75% ) in another ( see S2 Table ) . Curiously , while the restoration of rifampicin sensitivity is mediated by a genetic reversion ( RpoBY526H ) , the tested clones that regained sensitivity to streptomycin still harbor the original mutation ( RpsLK43T ) and thus experienced a phenotypic reversion . Notably , phenotypic reversion of streptomycin resistance due to mutations in tufA ( which appeared as target in our study ) has been previously reported [41 , 42] . Genetic reversions are expected to cause strong beneficial effects in the double-resistant background , given that they provide a direct solution for the negative epistasis . For instance , a reversion in rpoB in the double-resistant bacteria would cause a 24% fitness increase , from a competitive fitness of 0 . 73 ( 27% cost ) to 0 . 97 ( 3% cost ) ( see S2 Fig ) . Its rarity in the RifR StrR background is a further indication of the emergence of other compensatory mutations with strong fitness effects , which is consistent with the very large fitness gains observed in the double-resistant populations ( Fig 1C and Fig 1F ) . Interestingly , mutations affecting genes involved in DNA replication and DNA repair ( i . e . , dnaG and mutL ) were detected in double-resistant evolved populations ( Fig 3 ) . Such mutations may have caused mutation rate changes in some of the evolved clones , thus resulting in a higher number of mutations accumulated . However , since the frequency of these mutations in the pooled run was only 3 . 6% , these mutations appeared in only a few populations ( two at most , in the case of mutL ) and thus are unlikely to have affected the estimation of evolutionary parameters ( Fig 2 ) . Apart from the previously mentioned IS element detected in some populations , both in the StrR and RifR StrR populations ( Fig 3 ) , no other types of structural variation , such as duplications , could be found , as assayed by looking for increases in coverage for any given region of the sequenced evolved lines . In summary , the limited overlap among nucleotide changes across backgrounds suggests that single- and double-resistant bacteria acquire distinct compensatory mutations . This might be due , at least partially , to the existence of mutations that compensate for the epistasis between resistances in double-resistant bacteria . To test the hypothesis that epistatic interactions between the resistance alleles provide additional compensatory targets , we generated derivatives of all resistant backgrounds harboring selected compensatory mutations that we have only found in double-resistance populations and could potentially compensate for the epistasis . We used genetic reconstruction to isolate the compensatory effect of each individual mutation for double-resistant bacteria , as well as measuring its fitness effect in the sensitive and single-resistant backgrounds . To do so , we performed competitions between derivatives of the founder genotypes ( RpoBH526Y , RpsLK43T , or RpoBH526Y RpsLK43T ) harboring the compensatory mutation ( s ) and their corresponding ancestral ( not carrying the compensatory mutation ) and estimated their fitness effect from differences in growth over a 24 h competition ( see materials and methods ) . We selected a single nucleotide polymorphism ( SNP ) in the rpoC gene ( RpoCQ1126K ) because , although rpoC is a common compensatory target for rifampicin resistance [19 , 26 , 43] , this change was detected at a very high frequency ( 26 . 1% ) in the double RifR StrR populations but not detected in any of the single-resistant populations ( Fig 3 and S1 Table ) . Additionally , we chose two regulatory mutations observed exclusively in the double-mutant background: one in the promoter region of the secE-nusG bicistronic transcript [44] ( nusG hereafter ) , and another in a regulatory secondary structure in the 5′ untranslated region ( 5′-UTR ) of the operon that encodes rpsJ ( nusE ) [45] ( nusE hereafter ) . The latter two mutations were selected because they are good candidates for being compensatory to epistasis , since NusG and NusE mediate the molecular interaction between the ribosome and the RNAP [46] ( Fig 4E ) . We found the nusE mutation to be beneficial across all resistance backgrounds but neutral in the sensitive bacteria ( Fig 4A and S4A Fig ) . In contrast , the mutation in rpoC confers a strong beneficial effect only in the double RifR StrR background , where it emerged , being neutral or even slightly deleterious in either the single-resistance or the sensitive backgrounds ( Fig 4B and S4B Fig ) . Finally , the nusG mutation is exclusively beneficial in the double RifR StrR background , being highly deleterious in all other backgrounds ( Fig 4C and S4C Fig ) . Therefore , the beneficial effects of RpoCQ1126K and nusG compensatory mutations are strongly conditioned by the presence of the two resistance alleles , suggesting that , in both cases , the target of compensation is the epistatic interaction between the resistance alleles . To further corroborate this we measured the effects of RpoCQ1126K and nusG mutations in minimal medium supplemented with glucose , an environment where RpoBH526Y and RpsLK43T were previously shown not to exhibit negative epistasis [28] . We found that RpoCQ1126K loses most of its beneficial effect in this environment ( from 25 . 4% to 4% ) , indicating that it partially compensates for the epistasis . Remarkably , nusG is indeed deleterious in minimal medium supplemented with glucose , demonstrating that this mutation compensates specifically for the epistasis between antibiotic resistances ( Fig 4D , and S4B and S4C Fig ) . On the other hand , nusE does not seem to compensate specifically for the negative epistasis ( Fig 4A ) , as it is equally beneficial in minimal medium supplemented with glucose , where the epistasis is not present . In sum , we found that while nusE is a global compensatory mutation to the costs of resistance , RpoCQ1126K and nusG are highly idiosyncratic and only likely to be found linked to double-resistant backgrounds , since they compensate for the negative epistasis between these resistance alleles . The striking difference in the fitness effects of nusE and nusG regulatory mutations prompted us to analyze their effects on gene expression in the sensitive background by quantitative real-time PCR ( RT-qPCR ) . The change in the leader transcript of nusE ( Fig 5A ) caused a small but significant increase ( less than 2-fold ) in its expression level ( Fig 5C ) , consistent with the mutation affecting a region involved in its transcriptional and translational attenuation [47] . The mutation in nusG affects the −10 sequence of its promoter region ( Fig 5B ) [44] , resulting in a strong increase ( above 4-fold ) in its expression ( Fig 5C , rightmost column ) . Interestingly we found that the nusG mutation also causes upregulation of nusE expression ( Fig 5C , third column ) . This cross-regulation might be due to the fact that NusA participates in the attenuation of nusE operon [47] and that NusG can interact with NusA [48] , interfering with its activity [49] . Importantly , the effects of the nusE and nusG regulatory mutations on gene expression reflect their effects on the fitness of sensitive bacteria: the strong and pleiotropic induction caused by the nusG mutation generates a high fitness cost ( Fig 4C ) , while the milder and less pleiotropic change on expression caused by the nusE mutation entails a close to neutral fitness effect ( Fig 4A ) . This suggests that the expression of these genes constitutes an important mechanism connecting genotype , phenotype , and fitness .
As spread of multidrug-resistant bacteria increases , it is crucial to understand how multiple resistant alleles can be maintained in populations [52] . Thus , investigating the process of acquisition of compensatory mutations is essential . Although compensation for single-resistance mutations has been thoroughly studied , our findings suggest that previous observations are not readily transferrable to compensation in multiresistant strains . We demonstrate that the pace of compensation is faster in a bacterium resistant to both rifampicin and streptomycin , due to stronger effect mutations being accessible to this low-fit background . Compensation by stronger effect mutations is supported by direct competitive fitness assays , in which high-fitness increments were detected ( Fig 1 ) . In particular , we found a mutation ( RpoCH450P ) beneficial in the RifR background ( s = 0 . 05 ± 0 . 01 ) that shows a stronger beneficial effect ( s = 0 . 10 ± 0 . 01 ) in the double mutant ( S5 Fig ) . This is consistent with the general pattern recurrently observed that low-fitness genotypes tend to acquire stronger effect mutations [33 , 34 , 53 , 54] and tend to adapt at a faster pace [32 , 35] . Interestingly , when studying the evolvability of other RifR mutants , Barrick et al . found that lower-fitness genotypes were able to adapt faster due to access to stronger effect mutations [37] . Thus , our observations fit the previous found patterns . Curiously , we observed that , in some adapted populations , the competitive fitness exceeds the initial fitness cost of the ancestral genotype ( see Fig 1D ) . This indicates that either there is further adaptation to the environment ( i . e . , the ancestral genotype is not at or very near to the fitness peak ) beyond compensation and/or that the initial resistance mutation displaced the genotype in the fitness landscape , allowing the exploration of other , potentially higher , fitness peaks [55] . In order to determine the extent to which adaptation to the environment occurs in our experimental conditions , we propagated sensitive bacteria during 22 days in an antibiotic-free medium , performed competitive fitness assays , and determined the mutation targets associated with such propagation ( S6 Fig ) . We found that marker frequency changes at a much slower pace ( S6A Fig ) , but it is not static , and associated increases in competitive fitness can be detected ( S6B Fig ) . Therefore , other mutations associated with the adaptation to the environment occur . Thus , adaptation to the environment could also have occurred in the resistant backgrounds . We note , however , that we could not detect any common mutational targets between sensitive and resistant backgrounds ( S6C Fig ) . Interestingly , in the double-resistant backgrounds , the dynamics of fitness appear to stabilize at different plateaus for different populations ( Fig 1F and S3 Fig ) . However , the period of propagation done in this study is too short to assess if these populations are indeed stabilizing at different peaks . Nevertheless , it is possible to conceive the existence of a complex network of epistatic interactions among the compensatory mutations . In that scenario , the initial acquisition of a given compensatory mutation could influence the fitness effects of subsequently acquired compensatory mutations , restricting the range of evolutionary trajectories during further adaptation [56] . When analyzing the genetic basis of compensation , we detected several new putative compensatory targets ( Fig 3 ) . In principle , the general lack of overlap across the three mutant backgrounds and the sensitive strain ( Fig 3 , S6C Fig , S1 Table ) suggests that the majority of the mutations found are not due to general adaptation to the selected growth conditions but , instead , are driven by compensation to the cost of resistance . Moreover , although the target size for adaptation to growth conditions might be large enough to make overlap relatively infrequent , and the sampling method used ( one clone per population ) may limit detecting overlapping mutations , many of the targets identified are directly related to transcription and translation , suggesting that they are indeed compensatory . The observation of parallelism in targets such as the mglBAC operon in the RifR StrR populations or the cytR gene in StrR populations suggests that targets of compensation not directly related to rpsL and rpoB might be available . By genetic reconstruction , we found that a mglA mutation is beneficial in both the RifR and RifR StrR backgrounds , being deleterious in the StrR background and neutral in the sensitive ( S7 Fig ) . This points to the mglA mutation being indeed compensatory for RifR and not generally adaptive , which is also supported by its absence in the mutational targets detected from the propagation of the sensitive strain ( S6C Fig ) . The functional unrelatedness of the mglBAC operon ( sugar transport [57] ) with the cellular machinery affected by the original RifR mutation ( RNAP ) suggests that the mutations in the mglBAC operon do not correct for the damage generated by the resistance mutation . Instead , they may somehow alleviate its consequence and , therefore , its fitness cost . Several alleles reconstructed ( RpoCQ1126K , nusG , and nusE ) in the different backgrounds were found to be beneficial in the single- or double-resistant genotypes but not in sensitive bacteria , providing direct evidence of their compensatory role . The presence of these compensatory mutations has no discernible impact on the ability of sensitive , single- , or double-resistant bacteria to withstand rifampicin and streptomycin , as no significant change in their minimum inhibitory concentrations was detected ( S3 Table ) . The dynamics of marker frequency and the distribution of fitness effects in the propagation experiment ( Fig 1C and 1F ) suggest that double-resistant bacteria already acquired compensatory mutations with large fitness effect by day 5 . Indeed , population-sequencing of the evolved double-mutant bacteria at day 5 ( S1 Table ) confirmed that several mutations present at day 22 appear already at day 5 ( for instance , RpsET103P , RpoAH66Q , an IS insertion between insA and uspC , RpoBE745K , RpoCA784V , RpoCQ1126K , RpoCE1030A , and RpoCI1248T ) . The maintenance of these alleles in the populations suggests that their fitness effects are strong enough to have prevented them from being outcompeted by other mutants during the evolution experiment . Consistently , one of these mutations ( RpoCQ1126K ) was shown to cause a large compensatory effect ( Fig 4B ) . Among the compensatory mutations found , we identified specific sets of mutations with sign epistasis , i . e . , that shift from beneficial to deleterious depending on the genetic background in which they occur ( Fig 4 ) . Indeed , we show that compensation in RifR StrR bacteria can specifically target the epistatic interaction between the resistant alleles . Importantly , mechanisms specifically targeted by these compensatory mutations include the coupling between transcription and translation , modulating the interaction between the resistances themselves . For instance , NusG and NusE proteins , which constitute the molecular link between RNAP and ribosomes [46] ( Fig 4E ) , are mutational targets of compensation in the double-resistant strains . NusE is part of the 30S ribosomal complex [58] and , besides connecting with the RNAP through NusG [46] , it has been suggested to bind to it directly [59] . Therefore , a beneficial effect in the different resistance backgrounds , but not in sensitive bacteria , could be expected . Moreover , the regulatory mutation in nusE likely affects the expression of its entire operon , which encodes ten other ribosomal proteins , whose potential compensatory effects will have to be analyzed in future research . On the other hand , the mutation affecting NusG , a universally conserved transcription termination/antitermination factor , shows a beneficial effect specific to the epistasis between streptomycin and rifampicin resistance alleles . A tentative explanation for this effect relies on the coupling of transcription and translation in prokaryotes . RpoBH526Y and RpsLK43T resistance mutations generate dysfunctional RNAP [60] and ribosome [61] , respectively , affecting transcription and translation efficiencies . These altered processivities likely generate discoordination of these two machineries , therefore challenging the transcription–translation coupling . In that scenario , mutations causing increased expression of factors involved in the physical interaction between the RNAP and the ribosome ( NusG and NusE ) [46] may help to enhance their coordination , restoring the coupling . Indeed , the RpoCQ1126K mutation partially compensates for the epistasis between resistances , and the analysis of the E . coli RNAP structure available ( PDB 3LUO ) indicates that the residue Q1126 is at the surface of the RpoC protein and maps closely to the region of the protein described to interact with NusG ( Fig 4E and S8 Fig ) . This suggests that the RpoCQ1126K mutation could also be mechanistically linked to the interaction between the RNAP and the ribosome and , therefore , to transcription–translation coupling . To the best of our knowledge , this is the first time that the NusG–NusE complex has been identified as a compensatory target for antibiotic resistance . This shows the power of experimental evolution to identify novel compensatory targets in resistant bacteria and also reveals its potential at uncovering important mechanisms in cellular fitness [62] . Moreover , the identification of mutations targeting the regulation of gene expression raises the possibility of the existence of nonmutational mechanisms for phenotypic compensation of the fitness cost of antibiotic resistances , as it has been recently described in mycobacteria [63] . Importantly , although our observations are limited to the set of resistance mutations studied , given that the mechanisms involved in their interaction ( transcription and translation ) are commonly targeted by antibiotics [64 , 65] , these observations likely apply to other resistance mutations . It will be interesting to study other multiresistant backgrounds in the future in order to test this possibility . Resistance mutations typically affect essential cellular functions and more often than not those mutations tend to exhibit epistasis [28 , 66] . Indeed , observations in clinical populations of bacterial pathogens suggest the existence of undiscovered epistatic interactions [67] . Thus , there is an urgent need to uncover the spectrum of mutations that suppress the costs arising from epistatic interactions in multidrug-resistant pathogens . Our results suggest that this cannot be achieved by focusing on the study of single-resistant strains . The ability to identify and determine how specific mutations affect epistasis between resistant alleles is essential for understanding how antibiotic resistances persist in bacterial populations . Interestingly , this knowledge may provide new grounds for the development of novel antimicrobial strategies that specifically exploit potential weaknesses derived from epistasis between antibiotic resistances in multidrug-resistant bacteria [68] .
All the strains used in this study ( S4 Table ) are derivatives of an E . coli K12 MG1655 marked with constitutively expressed YFPs or CFPs inserted in the galK locus and the entire lac operon deleted ( ΔlacIZYA galK::cat::PLlacO-1-YFP/CFP ) . Resistant mutants also harbor chromosomal-resistance mutations to either streptomycin ( RpsLK43T allele ) , rifampicin ( RpoBH526Y allele ) , or both . Single- and double-resistance strains in both fluorescent backgrounds were used in the long-term propagation experiment in Lysogen Broth ( LB ) without antibiotics . Cultures were grown in a 96-well plate incubator at 37°C with shaking ( 700 rpm ) . Nonfluorescent wild-type E . coli K12 MG1655 was used as reference strain for the competition fitness assays , except when specified otherwise . In order to acclimatize bacteria to the environment , strains were grown separately from frozen stocks in LB media ( 150 ul per well ) in 96-well plates at 37°C with shaking ( 12 replicates per strain were inoculated in a checkered format to avoid cross-contaminations ) . After 24 h , 10 ul of bacteria culture diluted by a factor of 10−2 was transferred into 140 ul fresh LB medium and allowed to grow for an additional 24 h . Isogenic strains differing only in the marker were diluted again by a factor of 10−2 and then mixed based on their cell numbers given by the Flow Cytometer ( LSR Fortessa ) in order to obtain an initial ratio of 1:1 . CFP was excited with a 442 nm laser and measured with a 470/20 nm pass filter . YFP was excited using a 488 nm laser and measured using a 530/30 nm pass filter . A total of 48 competitions were initiated by inoculating 140 ul of LB medium with 10 ul of each mixed population , which were allowed to grow for 24 h , reaching a concentration of , approximately , 109 CFU/ml . The separate growths were done to minimize the occurrence of common compensatory mutations during acclimatization , a phenomenon that is difficult to avoid . After every 24 h of growth , and for 22 days , these cultures were propagated by serial passage with a constant dilution factor of 10−2 ( 10 ul of diluted culture was transferred into 140 ul of fresh medium ) . In parallel , cell numbers were counted using the Flow Cytometer in order to measure the frequency of each strain in the mixed population during the experiment , by collecting a sample ( 10 ul ) from the spent culture each day . Samples were frozen at days 5 , 9 , 12 , 15 , 18 , and 22 . The relative fitness of each evolved population at the end of the propagation experiment , at day 22 , was measured by competitive growth against the nonfluorescent ( unmarked ) isogenic reference strain E . coli K12 MG1655 . The competitor ( each evolved population , potentially composed of both YFP and CFP ) and the reference ( unmarked ) strains were first unfrozen and acclimatized separately for 48 h ( with two growth periods of 24 h ) and then mixed in a proportion of 1:1 , using a method similar to the one previously described . To assess the cost of the resistances themselves before any compensation , control competitions were performed between the ancestral of each resistant mutant and the reference strain by mixing 25% of YFP + 25% of CFP with 50% of the unmarked strain . Thus , a total of 52 competitions were initiated by inoculating 140 ul of LB media with 10 ul of each mixed population and allowed to compete for 48 h . The initial and final frequency of the strains were obtained by counting their cell numbers in the Flow Cytometer . Generation time was estimated from the doubling time of the reference strain ( approximately eight generations ) , and the fitness was determined as the average of three independent replicates for each competition . Evolved cultures were grown in 96-well plates with LB for 24 h and subsequently plated in antibiotic-free solid media . From each evolved population , clones were picked with a pipette tip , which then was used to streak sequentially both on antibiotic plates ( containing 100 ug/ml of rifampicin , streptomycin , or both drugs ) and antibiotic-free plates ( LB agar only ) . Clones were classified as sensitive if they grew on an antibiotic-free media but not on media with antibiotic ( s ) . Fifty clones were initially tested for each population . Populations in which sensitive clones were found were subsequently tested for another 150 clones . The method used to estimate the rate of acquisition of beneficial mutations and the distribution of selective effects was modified from the one published in [36] . Briefly , the method simulates 1 million marker dynamics with parameters ( mutation rate , shape , and mean of a gamma distribution for the selective effects of new mutations ) chosen from a uniform distribution . Afterwards , marker frequency and average population fitness are used to summarize both the experimental data and the simulated dynamics . These data are used as summary statistics to be compared by Approximate Bayesian Computation ( ABC ) . The modification in the method done here is that a simple ranking and rejection method was used ( instead of using the neural network function in ABC ) , in which simulations are ranked by the Euclidean distance between summary statistics from the simulated and experimental data , and the top 20 simulations are chosen for the distributions of parameters . This change in the method was required , as the neural network method failed to provide biological reasonable parameters for the observed experimental data . We note that , in our estimates , we assumed that mutations can only arise during the simulated period ( t > 0 ) . Thus , we have ignored the short period of acclimatization . Given the possibility that mutations might have occurred during acclimatization , this could cause an overestimation of the rate of appearance of beneficial mutations . The evolved populations were plated onto LB agar plates and grown for 24 h at 37°C . One colony , with the most frequent marker , was picked from each population and was grown overnight into 10 mL of liquid LB , at 37°C with shaking . The bacterial DNA was then extracted , and its concentration and purity were quantified using Qubit and NanoDrop , respectively . Following quantification , equal concentration of each DNA sample was taken and pooled together with the clones from the populations of a similar resistant background , resulting in three pools of DNA ( RifR pool with 12 clones , StrR pool with 12 clones , and RifR StrR pool with 24 clones ) . The DNA extracted was sequenced using the Miseq Illumina platform . Coverage of the different pools was as follows: 521x for 12 clones in the RifR pool , 640x for 12 clones in the StrR pool , and 631x for 24 clones in the RifR StrR pool . The reads were filtered using SeqTk version 1 . 0-r63 . The resulting sequences were analyzed in Breseq version 2 . 3 , using E . coli K12 genome NC_000913 . 2 as a reference , and with the polymorphism option selected and the following parameters: ( a ) rejection of polymorphisms in homopolymers of a length greater than three , ( b ) rejection of polymorphisms that are not present in at least three reads in each strand , and ( c ) rejection of polymorphisms that do not have a p-value for quality greater than 0 . 05 . All other Breseq parameters were used as default . Mutations in homopolymers and pseudogenes were also discarded . The genomic data in Fig 3 was plotted using the BLAST Ring Image Generator ( BRIG ) software [69] . The mutations located in nusE and nusG regulatory regions were constructed by Lambda-Red recombineering [70] using a resistance cassette inserted in a nearby gene ( gspB and tufB , respectively ) as a marker for selection , followed by transference to RifR , StrR , and RifR StrR backgrounds by P1 transduction [71] using the same marker . Derivatives of these strains harboring the marker in the nearby gene but not the mutations in nusE and nusG were constructed , in order to be used as references in the competition experiments . The putative compensatory mutations located in rpoC were constructed by pORTMAGE recombineering [72] in both the RifR and the StrR backgrounds and subsequent transfer to the sensitive , RifR , StrR , and RifR StrR backgrounds was done by P1 transduction , using as recipient auxotroph derivatives carrying a mutation in the nearby gene argE and selecting for growth in minimal medium . The primers used are listed in the S5 Table . The presence of the desired mutations in the transductant isolates was assessed by PCR-mediated amplification of the corresponding gene and sequencing . Both reconstructed strains and the corresponding reference strains were unfrozen and acclimatized during 24 h to avoid compensatory mutations appearing in the more costly backgrounds . Reconstructed mutants were competed against the respective ancestral strain to assess their competitive advantage per generation , assuming the competition lasts eight generations , or 24 h . The fitness landscapes of the main reconstructed compensatory genotypes ( S4 Fig ) were plotted using the MAGELLAN ( Maps of Genetical Landscapes ) software [73] . To determine changes in gene expression caused by the mutations in the regulatory regions of nusE and nusG , RNA extraction from 0 . 5 ml of cultures at the late exponential phase ( OD600 0 . 6 ) of isogenic strains harboring , or not , each mutation was performed using a Direct-Zol RNA miniprep kit ( Zymo Research ) according to manufacturer's specifications . Bacterial RNA was subsequently treated with RQ1 DNase ( Promega ) according to manufacturer's protocol . Reverse transcriptase reaction was performed on 880 ng of RNA , using M-MLV RT ( Promega ) and random primers ( Promega ) according to manufacturer's protocol . RT-qPCR was executed in a Bio-Rad CFX 384 Real-Time PCR Detection System , using iTaq Universal SYBR green Supermix ( Bio-Rad ) . The cDNA was diluted 200-fold before being used for RT-qPCR . The cycling conditions were as follows: one step of 10 min at 95°C and then 40 cycles of 30 s at 95°C , 30 s at 67°C , and 30 s at 72°C . The primers used are listed in the S5 Table . Melting-curve analysis was performed to verify product homogeneity . All reactions included three biological replicates for each strain , and three technical replicates for each sample . For analysis , data were normalized using the algorithm specific for multiple reference genes described by Hellemans et al . [74] , using rrsA and ssrA housekeeping genes as references . The MIC of rifampicin and streptomycin in the ancestral and the compensated backgrounds was determined by using MIC test strips ( Liofilchem S . R . L . ) in LB agar medium , ranging from 0 . 016 to 256 μg/ml for rifampicin and from 0 . 064 to 1 , 024 μg/ml for streptomycin . The strains resistant to the maximal concentrations in the MIC test strips were assayed for higher concentrations in liquid medium ( LB broth ) , up to 768 μg/ml for rifampicin and 8 , 192 μg/ml for streptomycin . The assessment of different slopes for single and double mutants ( Fig 1A–1C ) was performed using a One-way ANOVA , corrected with Tukey’s HSD to compare between the different resistance backgrounds . The significance of the change in competitive fitness ( Fig 1D–1F ) was performed using an analysis of covariance ( ANCOVA ) between the ancestral populations and the evolved populations at day 5 and GLMEM ( with restricted maximum likelihood ) for the fitness changes between day 5 and the end of the experiment ( day 22 ) . In both cases , we have used Hochberg’s adjustment to correct for multiple testing . The differences in estimated rate of acquisition of beneficial mutations ( Fig 2A ) , and also the differences in their effects ( Fig 2B ) , were analyzed using a two-sample Kolmogorov , adjusted for multiple comparisons using the Benjamini and Hochberg method [38] . The significance of the fitness effect of compensatory mutations across backgrounds ( Fig 4A–4C and Fig 5C ) was assessed by ANOVA with Tukey's correction for multiple testing . The CIs for classifying a population “sensitive” ( or partially sensitive ) ( S2 Table ) were assessed using a Binomial proportion CI on the number of colonies assayed that show a sensitive phenotype . | Antibiotics target essential cellular functions , such as translation or cell wall biogenesis , and bacteria can become resistant to antibiotics by acquiring mutations in genes encoding those functions . This causes most drug-resistance mutations to be detrimental in the absence of the drug . However , bacteria can reduce this handicap by acquiring additional mutations , known as compensatory mutations . Compensatory evolution is crucial for the maintenance and dissemination of antibiotic resistances in bacterial populations . While compensation for single resistances has been extensively studied , compensatory evolution of multidrug-resistant bacteria remains unexplored . Importantly , interactions between resistance mutations are frequent , and this may cause compensation of multidrug-resistant bacteria to differ significantly from that of single-resistant strains . By comparing compensation of single- and double-drug–resistant E . coli , we found that double-drug–resistant bacteria compensate faster than single-drug–resistant strains . This is due to the acquisition of compensatory mutations with larger effects and possibly driven by the large fitness cost of double-drug resistance . Strikingly , we identified mutations that compensate specifically for the interaction between drug resistances , since they are beneficial only for double-drug–resistant bacteria and in conditions in which the interaction between resistances occurs . In summary , our data indicate that certain interactions between antibiotic-resistance mutations can open alternative paths for rapid compensatory evolution , thereby potentially stabilizing multiple drug resistances in bacterial populations . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"bacteriology",
"antimicrobials",
"organismal",
"evolution",
"medicine",
"and",
"health",
"sciences",
"microbial",
"mutation",
"drugs",
"microbiology",
"cloning",
"epistasis",
"antibiotic",
"resistance",
"mutation",
"streptomycin",
"antibiotics",
"microbial",
"evolution",
"... | 2017 | Multidrug-resistant bacteria compensate for the epistasis between resistances |
Two-component signaling systems ( TCS ) regulate bacterial responses to environmental signals through the process of protein phosphorylation . Specifically , sensor histidine kinases ( SK ) recognize signals and propagate the response via phosphorylation of a cognate response regulator ( RR ) that functions to initiate transcription of specific genes . Signaling within a single TCS is remarkably specific and cross-talk between TCS is limited . However , regulation of the flow of information through complex signaling networks that include closely related TCS remains largely unknown . Additionally , many bacteria utilize multi-component signaling networks which provide additional genetic and biochemical interactions that must be regulated for signaling fidelity , input and output specificity , and phosphorylation kinetics . Here we describe the characterization of an NtrC-like RR that participates in regulation of Type-IV pilus-dependent motility of Myxococcus xanthus and is thus named NmpR , NtrC Modulator of Pili Regulator . A complex multi-component signaling system including NmpR was revealed by suppressor mutations that restored motility to cells lacking PilR , an evolutionarily conserved RR required for expression of pilA encoding the major Type-IV pilus monomer found in many bacterial species . The system contains at least four signaling proteins: a SK with a protoglobin sensor domain ( NmpU ) , a hybrid SK ( NmpS ) , a phospho-sink protein ( NmpT ) , and an NtrC-like RR ( NmpR ) . We demonstrate that ΔpilR bypass suppressor mutations affect regulation of the NmpRSTU multi-component system , such that NmpR activation is capable of restoring expression of pilA in the absence of PilR . Our findings indicate that pilus gene expression in M . xanthus is regulated by an extended network of TCS which interact to refine control of pilus function .
Two-component signaling systems ( TCS ) regulate numerous bacterial responses to environmental signals . These biological machines typically contain a membrane associated sensor histidine kinase ( SK ) and a cognate cytoplasmic response regulator ( RR ) that function together to propagate a response that is facilitated by the phosphorylation and dephosphorylation of the RR by the dual-function SK [1] . Following activation , RRs may function via different output domains , but the most common response is transcriptional regulation . Overall , TCS are critical for bacterial survival under varying environmental conditions and have well-established roles in metabolism , stress responses , virulence , motility , and many other physiological processes [2–9] . Despite the vast array of potential inputs and outputs , TCS have evolved to maintain remarkable specificity regarding phosphotransfer between cognate SK/RR pairs [10–12] . However , how TCS control the flow of information in the context of complex signaling networks within a given cell is less well-understood , especially in bacterial species with a high number of closely related TCS that have arisen from gene duplication . One particularly important subfamily of RRs is known as the NtrC-like RRs , named after the nitrogen regulator NtrC of Escherichia coli and Salmonella typhimurium [13 , 14] . This family of RRs is required for transcription initiation via interaction with RNA polymerase holoenzymes that specifically contain the σ54 sigma factor , which cannot initiate transcription on their own [15] . Unlike typical RRs , NtrC-like RRs contain a central ATPase domain that is required for hexamer or heptamer oligomerization [16–20] and is necessary for transcriptional activation by providing the energy for opening of the transcriptional bubble [21 , 22] . Also , NtrC-like RRs may cause significant DNA bending , may function at relatively large distances ( 100–1000 bp ) from transcriptional start points , and have been demonstrated to bind DNA elements both upstream and downstream of the start point . Thus , NtrC-like RRs are thought to function similarly to eukaryotic transcriptional machinery and are known as bacterial enhancer binding proteins [15 , 23] . NtrC-like RRs are important transcription factors in numerous bacterial species , though many bacteria encode only a limited repertoire . For example , E . coli encodes only four ( NtrC , ZraR , AtoC , and GlrR ) . However , some bacteria , especially environmental species with large genomes , encode an expanded number of NtrC-like RRs [24] . Myxococcus xanthus , a member of the δ-proteobacteria , encodes at least 27 bona fide NtrC-like RRs and a similar number of functionally related proteins that contain the σ54-interacting central ATPase domain but alternative sensing or output domains [25 , 26] . It is believed many of the M . xanthus NtrC-like RRs arose via gene duplication and they are highly related at the nucleotide and amino acid level , making M . xanthus an ideal model organism to study how closely related signaling molecules maintain signaling fidelity or interact within complex networks [27] . Indeed , several NtrC-like RRs and related proteins collectively described as NtrC-like activators ( Nla ) are known to coordinate development and motility of M . xanthus [3 , 25 , 26 , 28–36] . Consistent with the paradigm that TCS maintain signaling fidelity at the biochemical level , broad cross-phosphorylation ( "cross talk" ) between these systems does not seem to occur in M . xanthus [12] . This suggests the NtrC-like pathways of M . xanthus interact to modulate behavior at other levels , such as integration of environmental signals or overlapping transcriptional outputs . Based on this view , we hypothesized that additional regulation of the expression of motility genes should exist for M . xanthus which depends largely on motility for its survival and coordination of complex social behaviors such as multi-cellular development and predation . One method of M . xanthus motility is dependent on Type-IV pili ( T4P ) . This motility is thought to be strictly dependent on PilR , the RR of the PilSR TCS in many bacteria including M . xanthus and Pseudomonas aeruginosa [5 , 6 , 26 , 35 , 37] . PilR is a member of the NtrC-like RR family and , under standard laboratory conditions , is necessary for transcription of pilA encoding the T4P monomer . In the absence of PilR , M . xanthus is unable to generate PilA and is therefore unable to move [26 , 35] . However , in this study , we identify suppressor mutations in an M . xanthus ΔpilR strain that restore motility by activating a previously uncharacterized TCS . This system contains an NtrC-like RR ( Mxan_4240 ) with no previously described function . We identified several independent mutations in mxan_4240 that restore expression of pilA , and designate this gene nmpR ( NtrC Modulator of Pili ) . Each mutation occurred in well-conserved domains within NmpR and likely affect activity by promoting or mimicking the phosphorylated RR state . Biochemical and genetic data also demonstrate that this NtrC-like RR is part of a complex signaling system that contains at least four components: an SK ( Mxan_4246 , NmpU ) , a potential phospho-sink protein with two isolated RR receiver domains ( Mxan_4245 , NmpT ) , a hybrid RR-SK ( Mxan_4244 , NmpS ) , and the output RR ( Mxan_4240 , NmpR ) .
We have previously demonstrated that two RRs ( PilR and PilR2 ) encoded within the pil locus of M . xanthus are independently necessary for T4P-dependent social motility [26] . In particular , the RR PilR ( of the PilSR TCS ) is necessary under standard laboratory conditions for expression of pilA encoding the major T4P subunit and is therefore essential for T4P-dependent motility [26 , 35 , 38] . M . xanthus T4P-dependent motility is assayed on 0 . 5% agar and is readily distinguished as radial expansion from an initial colony inoculum . Strains that are unable to move on this medium do not expand and have a smooth , delineated colony edge [39 , 40] . During the course of our previous studies [26] , we observed “flares” extending out from the edge of the M . xanthus ΔpilR colony indicating this otherwise non-motile strain acquired mutations allowing for restored motility ( Fig 1A ) . Because of growth and colony motility rates , we typically assess M . xanthus T4P-dependent motility over the course of two days , but these flares only developed after prolonged incubation ( 7–14 days ) at 32°C . Suppressor mutant cells were isolated from individual flares by transferring a ~1 cm2 area from the leading edge to fresh 0 . 5% agar . Colonies were allowed to move away from the transferred spot to limit any potential contamination of the original ΔpilR strain and then transferred to broth medium for preparation for long-term storage and future characterization . All isolated suppressor mutants were confirmed by PCR to have maintained their original background deletion of pilR . Independent cultures of the parental ΔpilR strain were then spotted twice more ( total n = 3 ) and suppressor mutant flares were again observed with similar kinetics ( i . e . 7–14 days ) . At the same time , the non-motile ΔpilA strain was assayed , and no suppressor mutant flares were observed , as expected . Therefore , suppressor mutations must have occurred in a gene distinct from pilR , yet require pilA , and thereby represent pilR-bypass suppressor mutations . To begin to determine the causative nature of restored motility in these suppressor strains , a subset of mutants was characterized for motility related phenotypes ( Fig 1B ) . A total of nine mutant strains were assayed for three phenotypes associated with wild-type M . xanthus T4P-dependent social motility: 1 ) motility rates as measured by colony expansion over the course of five days; 2 ) the production of extracellular polysaccharides ( EPS ) that are essential for orienting cells within swarming colonies , stimulating T4P retraction , and regulating cell reversals [41 , 42]; and 3 ) the percentage of cells that display sedimentation in standing culture due to the “stickiness” of both T4P and EPS ( Fig 1B ) . All nine suppressor mutants displayed partial restoration of T4P-dependent motility , but at rates significantly lower than the wild-type cells . Similarly , all strains displayed varying levels of EPS production and sedimentation that were typically less than the wild-type but more than the ΔpilR . Collectively , these phenotypes strongly suggested that the restoration of motility to the suppressor mutant cells was due to their ability to produce T4P which requires production of the major pilin subunit , PilA . Indeed , anti-PilA immunoblot analyses of whole cell lysates of the nine suppressor mutants confirmed that they all produced near wild-type levels of PilA ( Fig 1C ) . Because pilA gene expression could have resulted from mutations in its promoter region , we sequenced the pilA promoter ( 186 bp upstream of the translation start codon ) in each of these strains . This analysis revealed no mutations had arisen in this known promoter region . Collectively , we concluded that restoration of PilA production , T4P-dependent motility , EPS production , and sedimentation of M . xanthus ΔpilR cells must have arisen at a unique locus relative to the pilA promoter . To identify the location of the suppressor mutations , the nine characterized strains along with the wild-type M . xanthus DZ2 and ΔpilR parental strain were subjected to whole genome sequencing with the Illumina MiSeq platform . Raw sequences were assembled with SeqMan NGen ( DNASTAR , Madison , WI ) using the annotated M . xanthus DK1622 genome as a template [27] . The sequence data resulted in ~60x coverage for each genome . Importantly , the ΔpilR parental strain had no mutations relative to the laboratory wild-type M . xanthus DZ2 strain , and the pilR deletion was reconfirmed in each suppressor . Seven of the nine suppressor strains had an identifiable mutation ( Table 1 and Fig 2 ) in mxan_4246 ( nmpU ) encoding a SK and one strain had a mutation in mxan_4240 ( nmpR ) encoding an NtrC-like RR . As the gene number indicates , these genes are in close proximity on the chromosome ( Fig 2A ) . Mutations in the SK nmpU spanned the length of the gene and each resulted in either a premature stop codon or frame shift which we predict would result in truncated forms of the SK ( Table 1 and Fig 2C ) . The single mutation identified in the RR nmpR was a missense mutation that changed a valine to glutamic acid at amino acid 87 ( V87E ) ( Table 1 , Fig 2B ) . Because the suppressors led to enhanced motility , we hypothesized that the amino acid change in the RR NmpR was the result of a gain-of-function mutation . To test this , we first constructed an in-frame nmpR deletion in the wild-type ( M . xanthus DZ2 ) , ΔpilR , and NmpRV87E strains . As expected , in the otherwise wild-type genetic background ΔnmpR cells remained motile ( Fig 3A ) since the native copy of pilR remained intact . In the absence of pilR , deletion of nmpR had no effect and this strain remained non-motile ( Fig 3A ) . Finally , deletion of nmpRV87E in the suppressor strain resulted in a loss of motility ( Fig 3A ) . Together these data strongly suggested that the NmpR V87E substitution confer a gain-of-function and confirmed this mutation is responsible for the restoration of motility for the ΔpilR cells . Several additional attempts to develop suppressor mutants in the ΔpilRΔnmpR strain were unsuccessful despite several passages on motility agar plates , strongly suggesting that under the conditions tested NmpR is the only additional RR that can restore pilA expression and motility of the ΔpilR strain . In order to determine if the NmpR V87E allele is sufficient to control expression of pilA and restore motility in the ΔpilRΔnmpR mutant background , we generated a series of complementation constructs . Complementation of mutations was performed using expression constructs that are capable of integration in single copy at the Mx9 phage attachment site ( see Materials and Methods ) . Expression of a wild-type allele of nmpR from its putative native promoter ( pNat; 585 bp upstream of mxan_4236; Fig 2A ) or from a high expression promoter ( pHigh; 623 bp upstream of mxan_4894 , groES [43 , 44]; S1 Fig ) was not able to restore motility to the ΔpilRΔnmpR strain . In contrast , expression of nmpR V87E from its native promoter was sufficient to restore motility to ΔpilRΔnmpR cells ( Fig 3B ) . Because RRs typically require phosphorylation for activity , we expressed the phosphomimetic mutant form of NmpR ( D54E ) and a non-phosphorylatable variant ( D54A ) . The D54E mutant form was sufficient to restore motility while the D54A mutant form was not ( Fig 3B ) . Given that only the phosphomimetic version of NmpR ( D54E ) was capable of restoring motility , we hypothesize the suppressor variant ( V87E ) mimics or promotes an active conformation of this RR to restore pilA expression and motility to the ΔpilR strain . During the complementation experiments described above , we were unable to recover transformants expressing the nmpR V87E allele expressed from the pHigh promoter in the absence of an endogenous copy of nmpR . This suggested to us that the NmpR gain-of-function is toxic to M . xanthus when over-produced . Nonetheless , we hypothesized that additional mutations in nmpR would lead to a similar gain-of-function and rationalized that over-expression of wild-type nmpR in the ΔpilR strain would reveal additional suppressor mutations as the endogenous allele of nmpR in this strain may provide protection from any detrimental consequences of constitutive activation of the heterologously expressed nmpR . As expected , over-expression of nmpR in ΔpilR did not restore motility , yet after prolonged incubation on 0 . 5% agar suppressor mutations were readily observed . These suppressor strains were isolated as before and confirmed to have restored T4P-dependent motility on fresh 0 . 5% agar . Subsequently , we selectively PCR amplified the over-expression nmpR construct ( with a forward primer in the heterologous promoter to differentiate it from the wild-type allele ) and sequenced . From this suppressor screen strategy , we were able to identify an additional nine independent , gain-of-function mutations in nmpR ( Fig 4 ) . Importantly , in a subset of these strains , we also sequenced the wild-type nmpR and nmpU loci and found no mutations , indicating the restored motility was due to the identified mutations in the over-expression nmpR construct . These mutations clustered in the receiver domain , especially near the phosphorylation pocket , and also in the variable “Q-linker” [45] domain that connects the receiver domain with the central σ54 activation domain ( Fig 4 ) . Therefore , these mutations likely influence the activation of NmpR by promoting or mimicking the phosphorylated state . Another smaller cluster of mutations was identified near the C-terminus of the σ54 activation domain and in the low-homology linker between the activation domain and the helix-turn-helix domain , suggesting that these might change the ATPase activity and/or the DNA binding affinity ( Fig 4 ) . Collectively , identification of these suppressor mutations in nmpR emphasizes a role for activated NmpR during modulation of pilA expression . Nearly all of the initial suppressor mutants identified were in the gene mxan_4246 encoding a SK that we have designated NmpU . However , within the same genomic locus is another SK that we have designated NmpS . Both of these SK were plausible candidates for cognate kinases of NmpR , so we sought to systematically and genetically determine if these signaling proteins are in a shared pathway . First , we deleted nmpU in several genetic backgrounds ( Fig 5 ) . Deletion of nmpU did not impair motility in wild-type M . xanthus DZ2 ( Fig 5A ) , consistent with the phenotype of deletion of nmpR ( Fig 3A ) . This further emphasizes that in the presence of PilR , neither the SK NmpU or the RR NmpR is necessary for motility and both likely play modulatory roles under certain environmental conditions . In the context of the NmpRV87E suppressor , deletion of nmpU did not result in a loss of motility , indicating the V87E alteration of NmpR in this suppressor strain is “blind” to the activity of NmpU ( Fig 5A ) . Finally , deletion of the mutated allele of nmpU in the suppressor strain NmpUQ283Stop ( Table 1 and Fig 2; C5 ) did not result in a loss of motility , consistent with the prediction that NmpU is non-functional in this strain ( Fig 5B ) . To demonstrate NmpU and NmpR are in a shared pathway , we next performed epistasis analysis , combining in-frame nmpU and nmpR deletions . When nmpR was deleted in the NmpUQ283Stop suppressor strain , it caused a return to a non-motile phenotype ( Fig 5B ) . Therefore , ΔnmpR is epistatic to the nmpU suppressor mutation and is a downstream output of a signaling system containing NmpU . This epistasis was also observed in two additional nmpU suppressor strains ( NmpUQ61Stop , C2 and NmpUQ217Stop , C4 ) , indicating nmpR is the output regardless of the location of the nmpU mutation . Next , complementation experiments similar to those presented in Fig 3B were conducted in the NmpUQ283Stop ΔnmpR strain ( Fig 5C ) . In this strain , complementation with wild-type nmpR expressed from its native promoter was sufficient to restore motility . However , over-expression of the wild-type allele from the pHigh promoter was not sufficient . We interpreted this to mean that when over-expressed , most of the NmpR would be unphosphorylated , even in the suppressor background . This is consistent with observations mentioned above that the presumed stoichiometry of the active to non-active state of NmpR is critical to influence motility and the wild-type , unphosphorylated NmpR is dominant to the suppressor mutations . Indeed , when an unphosphorylatable nmpR ( D54A ) was over-expressed , motility was not restored , while over-expression of a phosphomimetic nmpR ( D54E ) was sufficient for restored motility ( Fig 5C ) . Given that there are two additional TCS proteins encoded in the same genomic locus as nmpU and nmpR ( mxan_4244 , nmpS , and mxan_4245 , nmpT; Fig 2A ) , we sought to test whether the hybrid sensor kinase NmpS also plays a role in a signaling pathway that includes SK NmpU and RR NmpR . Returning to our epistasis analysis , deletion of nmpS in the NmpUQ283Stop returned it to a non-motile state ( Fig 5B and 5C ) . Therefore , ΔnmpS is also epistatic to the suppressor mutation of nmpU and is in a signaling pathway that includes the NmpU and NmpR . Remarkably , in the non-motile NmpUQ283Stop ΔnmpS strain ( Fig 5B ) , additional suppressor mutants developed and when sequenced were once again identified in nmpR , indicating that NmpR is the output of this multi-component signaling pathway . Activation of NmpR is sufficient to restore motility regardless of the loss of activity of the SKs NmpU or NmpS . Finally , epistasis analysis described here was recapitulated in the parental ΔpilR strain; deletion of nmpU was sufficient to restore motility , and deletion of nmpS or nmpR were epistatic to the nmpU deletion ( Fig 5D ) . The epistasis analysis described above clearly demonstrated that NmpR , NmpS , and NmpU are in a single signaling pathway . Yet , we could not be sure of the flow of phosphoryl groups and therefore biochemical regulation of the pathway . To conclusively demonstrate which SK is communicating with which RR , we purified each signaling component of the pathway individually as His-tagged protein constructs . To limit the complexity of this initial analysis , only the kinase domains ( i . e . no sensor domain ) or response regulator receiver domains ( i . e . no output domain ) were purified . Following well-established protocols based on enzymatic kinase activity and radio-labeled ATP [12] , we performed in vitro autophosphorylation and phosphotransfer assays ( Fig 6 ) . Indeed , both NmpS and NmpU displayed autokinase activity confirming they are active kinases ( Fig 6A ) . These SKs were then individually incubated pair-wise with each RR receiver domain . Based on the epistasis analysis , NmpU is the “top” kinase in the pathway and the phosphotransfer analysis supports this conclusion . NmpU phosphorylated the receiver domain of the hybrid SK NmpS but did not phosphorylate the receiver domain of the final signaling output NmpR ( Fig 6C ) . In addition , NmpU phosphorylated the first receiver domain of the dual receiver domain protein NmpT ( Fig 6B ) , consistent with results presented elsewhere [46] and which we suggest means that NmpT is a phospho-sink of NmpU . In direct contrast to the activity of NmpU , the hybrid SK NmpS only phosphorylated the receiver domain of NmpR ( Fig 6D and 6E ) . NmpS did not phosphorylate its own receiver domain or either of the receiver domains of NmpT . Together the phosphotransfer data suggests signaling fidelity within this branched pathway is dictated by the specificity residues of the SK/RR pairs [11] . The specificity residues of the first receiver domain of NmpT and the receiver domain of NmpS are nearly identical which likely explains the ability of NmpU to phosphorylate both ( Fig 6F ) . On the other hand , the unique specificity residues of NmpR support the observed phosphotransfer by only NmpS . Furthermore , the mutations identified in NmpR are not near the specificity residues ( Fig 4D ) and the V87E amino acid change of NmpR did not affect the specificity of the pathway . NmpRV87E is still phosphorylated by NmpS , but not by NmpU ( Fig 6C and 6D ) . Thus , our data supports the assertion that the specificity residues determine phosphotransfer specificity even in the context of complex pathways . Based on the epistasis analysis and the phosphotransfer data presented above , we hypothesized that the RR receiver domain of the hybrid kinase NmpS acts as a “sensing domain” and that phosphorylation of this domain by NmpU maintains NmpS in an “off” conformation . To test this , we again took both a biochemical and genetic approach ( Fig 7 ) . Full-length NmpS was purified and assayed in autokinase assays as described above . Full-length NmpS does indeed have baseline activity , indicating the amino-terminal RR receiver domain does not prevent autokinase activity in vitro ( Fig 7A ) . A NmpS construct containing a substitution at the conserved aspartic acid to alanine displayed greater activity than wild-type . Furthermore , complementation of the ΔpilR strain with a NmpS D59A construct was able to restore motility to this strain , while complementation with the full length wild-type NmpS did not ( Fig 7C ) . To date we have been unsuccessful in demonstrating direct regulation of NmpS kinase activity by phosphorylation of its receiver domain in vitro and more research will clarify the role of this RR receiver domain in signaling of this atypical SK . Yet , these data are consistent with the conclusion that the aspartic acid and its phosphorylation are not necessary in vitro or in vivo and that unphosphorylated NmpS ( i . e . NmpS D59A ) is the more active form of this SK . Having established the flow of phosphoryl groups through the multicomponent Nmp signaling system with in vitro kinase assays and epistasis analysis , we sought to establish the final NmpR output . That is , we tested whether NmpR directly regulates pilA or if the resulting suppressor phenotype is facilitated through an alternative indirect mechanism . To answer this question , full-length NmpR was purified and used in electromobility shift assays to determine the DNA regions that NmpR binds ( Fig 8 ) . Initially , three probes were designed: the pilA promoter ( -217 bp upstream of the pilA start site ) , the putative native promoter of nmpR upstream of mxan_4236 ( Fig 2A; used for the complementation experiments ) , and the groES promoter as a negative control ( again , the same region used for the complementation experiments , pHigh ) . As expected , increasing concentrations of NmpR did not shift the pgroES probe . Conversely , a reproducible shift was evident with increasing concentrations of NmpR and the p4236 probe . This suggests that NmpR is autoregulatory by binding to a promoter upstream of mxan_4236 . Finally , contrary to our original hypothesis , NmpR did not shift the ppilA probe , suggesting that NmpR does not bind this region to directly regulate pilA ( Fig 8A ) . This observation suggested that the regulation of pilA in the suppressor strains might be indirect , for example by regulating the expression of a different transcription factor that in turn regulates pilA . Alternatively , NmpR could regulate another portion of the pil locus that would then explain its ability to restore motility in the suppressor strains , as nearly all genes within the pil locus are co-transcribed [26] . Therefore , overlapping 800 bp DNA probes spanning the entire pil region upstream of pilA were generated and used in electromobility shift assays ( Fig 8B ) . NmpR shifted a probe containing 800 bp immediately upstream of pilR was specific; no other probe tested was bound by NmpR . A follow-up assay confirmed that NmpR does not bind within the pilR sequence but does bind a sequence upstream of a putative σ54 binding site within the pilS open reading frame . The potential σ54 binding site ( TGGCACGTGACGTACG ) begins at -351 bp relative to the pilR start codon in wild-type M . xanthus and is positioned -537 bp from the pilA start codon in the ΔpilR strain . These results suggested that pilS contains an internal NmpR-dependent promoter for regulation of pilR that is utilized to rescue pilA expression and therefore motility in the ΔpilR suppressor strains . Indeed , when pilS was deleted in two suppressor backgrounds , the strains returned to a non-motile phenotype ( Fig 8D and 8E ) . Therefore the sequence containing the NmpR binding site is necessary for the suppressor phenotypes and strongly suggests that this promoter is NmpR-dependent , highly specific , and physiologically relevant . In summary , the genetic and biochemical characterization of the NmpRSTU pathway leads us to propose a model ( Fig 9 ) whereby the “top” kinase NmpU phosphorylates NmpS to maintain it in an off state . NmpU also phosphorylates NmpT , though at this time it is unclear the role of NmpT as either a phospho-sink and/or inhibitor of the pathway . In the absence of phosphorylation of the NmpS RR receiver domain , NmpS is active and phosphorylates NmpR , making NmpRS a bona fide cognate SK-RR pair . NmpR appears to be autoregulatory , binding to the promoter of mxan_4236 ( Fig 8A ) and potentially other promoters throughout the M . xanthus genome . Critical to the role of NmpR in motility regulation , it binds a specific promoter upstream of pilR that we propose increases the relative concentration of PilR that in turn promotes pilA expression . Thus , NmpR indirectly modulates pilA expression during certain environmental conditions , likely changes in oxygen concentrations sensed by the protoglobin domain of NmpU . Perturbations of this pathway that alters the kinase activity , especially those that promote or mimic the phosphorylated state of NmpR , lead to pilA expression that rescues T4P-dependent motility in the absence of the otherwise necessary RR PilR .
The NtrC-like RR family has been extensively studied and has served as a model protein for intramolecular structure-function dynamics [18 , 19 , 23 , 47–50] . NtrC-like RRs also continue to be discovered as important transcriptional regulators controlling diverse biological processes in diverse bacterial species [51–54] . For example , NtrC-like RRs and σ54 play critical roles in processes exterior to the bacterial cell , such as lipopolysaccharide production , EPS production , and motility [24] . M . xanthus encodes at least 27 NtrC-like RR and many of these likely arose by gene duplication during the evolution of ∂-proteobacteria [27] . M . xanthus NtrC-like RRs are known to regulate such complex behaviors as multicellular fruiting body formation , sporulation , T4P-dependent social motility , and predation [3 , 25 , 26 , 29 , 30 , 32 , 33 , 35 , 36 , 55] . There is also evidence that there are hierarchal and/or overlapping transcriptional programs mediated by these complex TCS networks [3 , 33] . Therefore , M . xanthus is an ideal organism to study how these complex networks regulate the flow of information in a global cellular context . In this study , we have identified a TCS with no previously described function that includes an NtrC-like RR that we propose indirectly modulates pilA expression and therefore T4P-dependent social motility of M . xanthus . We have named the system NtrC Modulator of Pili ( Nmp ) . Interestingly , the gene encoding NmpR was previously disrupted by insertional mutagenesis and found to have no motility or developmental phenotype [25] . The lack of a phenotype in that background is consistent with our observation that NmpR function was only revealed under standard laboratory conditions via bypass suppressor mutations that arose to restore motility in the absence of PilR . Remarkably , the same selective pressure to restore motility in order to obtain nutrients has been observed elsewhere , such as in Pseudomonas fluorescens where a deletion of fleQ , an NtrC-like RR necessary for expression of flagellum biosynthesis machinery , was rescued by suppressor mutations in ntrB ( the SK of the NtrBC system ) that lead to hyperphosphorylation of NtrC [56] . The Nmp signaling system described here adds to the already complex picture of gene regulation for the control of social motility in M . xanthus [26 , 34 , 35 , 57 , 58] and is composed of at least four components: the SK NmpU ( Mxan_4246 ) , a putative phospho-sink NmpT ( Mxan_4245 ) , an atypical hybrid SK NmpS ( Mxan_4244 ) , and a final output RR NmpR ( Mxan_4240 ) ( Fig 9 ) . Notably , within other members of the ∂-proteobacteria lineage , nmpRSTU comprise one contiguous locus ( S2 Fig ) , emphasizing their function as a unit within this clade . In contrast , genomes of more distantly related ∂-proteobacteria ( e . g . Desulfovibrionales and Desulfuromonadales orders ) contain only a homologue of the SK , NmpU ( S2 Fig ) . Yet , there are examples of homologues of nmpSTU in diverse bacterial species including Planctomyces sp SH-PL14 ( S2 Fig ) . Like M . xanthus , Planctomycetes are bacteria that have a complex lifecycle , are found in soil , and have extracellular appendages [59 , 60] . Collectively , these observations suggest that the NmpRSTU signaling unit has been gained ( or lost ) throughout evolution within the ∂-proteobacteria and that horizontal gene transfer between species within this clade and between distantly related clades may have been common . As genomic sequences of previously unappreciated bacteria become more common-place and publicly available [61–65] , this signaling system may be revealed to be more widespread than currently appreciated . A homologue of NmpU ( AfGcHK ) exists within a closely related myxobacterial species , Anaeromyxobacter sp . Fw109-5 ( S2 Fig ) . Autophosphorylation of this kinase occurs when oxygen is bound to an iron molecule found within the heme co-factor of its protoglobin domain ( Fig 2C ) [66] . Anaeromyxobacter species are considered anaerobic , but they likely evolved from an aerobic ancestor within the δ-proteobacteria lineage [67] . Moreover , Anaeromyxobacter dehalogenans produced motility flares under aerobic conditions indicating this organism is aerotolerant and may utilize the protoglobin sensor AfGcHK ( NmpU ) to respond to oxygen concentrations to influence motility . Protoglobin sensors are part of the family of globin-coupled sensors found in the majority of bacterial species and in some archaea and fungi [68–72] . They are known to regulate physiological events in response to oxygen , such as aerotaxis of Bacillus subtilis and the archaeon Halobacterium salinarum [73] . Thus , we propose that this multi-component signaling system finely tunes the expression of pilR ( and by extension pilA ) in response to oxygen concentrations . During oxygen rich conditions , we propose the SK NmpU is “on” and autophosphorylates as the first step of this pathway ( Fig 9 ) . It is worth noting that M . xanthus is thought of as an obligate aerobe and so under laboratory conditions cells are grown and assayed in oxygen replete conditions . We therefore hypothesize NmpU is typically “on” in laboratory settings , keeping NmpR unphosphorylated and unable to affect downstream promoter targets ( Fig 9 ) . In natural settings , M . xanthus must encounter environments with varying oxygen concentrations , perhaps within the soil column or during multicellular development , so therefore likely encodes the ability to respond to an oxygen gradient . Following autophosphorylation , the NmpU homologue of A . dehalogenans Fw109-5 mentioned above preferentially transfers the phosphoryl group to the first receiver domain of the NmpT homologue [46 , 66] . Our analysis fits with that data , as NmpU phosphorylated RR1 of NmpT , but not RR2 ( Fig 6B ) . We expanded the known partners of NmpU by demonstrating that this kinase also phosphorylates the receiver domain of the hybrid SK NmpS which is encoded immediately downstream of NmpT . Remarkably , the specificity residues ( amino acids that confer cognate SK/RR specificity [10 , 11 , 74] ) are nearly identical between the RR1 domain of NmpT and the receiver domain of NmpS ( Fig 6F ) which explains the ability of NmpU to phosphorylate both receiver domains . We propose that phosphorylation of NmpS maintains it in an “off” state ( Fig 7 ) . This conclusion is supported in part by the fact that deletion of nmpS was epistatic to the null mutations of nmpU ( Fig 5B and 5D ) and that a NmpS D59A construct displayed increased autokinase in vitro and was able to rescue motility of the ΔpilR strain in vivo ( Fig 7 ) . Interestingly , hybrid SKs with an amino-terminal receiver domain and carboxyl-terminal histidine kinase domain , such as NmpS ( Figs 2 and 7 ) , may utilize the receiver domain as a “sensing domain” that regulates kinase function . For example , the function of a similar hybrid kinase , EsxG of Rhizobium NT-26 , is controlled via phosphorylation of its amino-terminal receiver domain [75] . Phosphorylation of EsxG is provided by another SK and promotes a closed EsxG conformation that is then unable to phosphorylate its own downstream cognate RR . Additionally , at least one other hybrid kinase with the same domain architecture as NmpS is Lvr , a SK important for virulence of Leptospira interrogans . Like NmpS , Lvr retains wild-type levels of autokinase activity when the conserved aspartic acid was substituted with alanine [76] . We expect that the receiver domain of NmpS impacts the conformation of individual monomers of this kinase , but could also affect the oligomeric state of a collection of NmpS monomers . The phosphorylation state and stoichiometry of all components of the Nmp pathway will ultimately determine the SK on/off states of NmpS , and will inform us on the regulation of atypical SKs broadly . These mechanisms of NmpS kinase regulation will be one area of further study . We propose that the “switch” of the Nmp pathway occurs during oxygen depleted conditions ( or when NmpU is mutated to be non-functional ) when NmpU no longer phosphorylates NmpS . In the absence of phosphorylation of NmpS , the second branch of the pathway becomes “on” . NmpS is then kinase active , phosphorylates NmpR , leads to modulation of pilR and modulates pilA expression ( Fig 9 ) . The conclusion that NmpR is the output of this multi-component pathway is supported by: 1 ) deletion of nmpR was epistatic to nmpU mutations ( Fig 5B and 5D ) , 2 ) unique nmpR suppressor mutations arose in a strain in which NmpU or NmpS was non-functional or deleted , respectively ( Fig 5C ) , 3 ) attempts to isolate further suppressor mutations in a ΔpilRΔnmpR strain were unsuccessful , and 4 ) NmpR specifically binds a DNA region upstream of a σ54 binding site in proximity to the pilR start codon that is necessary for the suppressor phenotype ( Fig 8 ) . In further support of our model , we demonstrate that NmpSR must be “on” to modulate pilR expression ( Fig 9 ) . First , suppressor mutations that seem to confer activation of the RR NmpR allowed for restored motility ( Fig 4 ) . One of these activating mutations that we characterized in depth , V87E , was necessary and sufficient for the restored motility ( Fig 3 ) . Furthermore , the NmpR D54A allele that cannot be phosphorylated is unable to restore motility while the phosphomimetic NmpR D54E is sufficient to restore motility ( Fig 3B and Fig 5D ) . Second , suppressor mutations that confer activation of NmpR to restore motility ( including V87E ) were identified in the well-conserved α-helix 4 within the RR receiver domain which contains several surface exposed amino acids ( Fig 4 ) . Structural and functional analysis of both NtrC and RRs in general , have indicated that this helix undergoes significant conformational changes between the inactive and active state , making close intermolecular contacts with the SK [77] and direct intramolecular contacts with the σ54 activation ATPase domain and dimer interface [18 , 19 , 47–50 , 78] . Indeed , NmpR V87E was phosphorylated by NmpS in vitro , and is sufficient to restore motility in vivo . Collectively , the data suggests that the V87E mutation promotes the activated state of NmpR . Another key mutation suggests that the constitutively active form of NmpR promotes motility . Specifically , the K104N mutation ( Fig 4 ) occurs at a position that is nearly 100% conserved in all RR , regardless of family . This lysine forms hydrogen bonds with the phosphoryl group and is critical for the ability to turn the RR “off” ( i . e . be dephosphorylated ) . Alteration at this lysine position to a related but distinct amino acid , such as asparagine or arginine , allows the RR to be phosphorylated but not dephosphorylated . For example , when this conserved lysine of E . coli CheY ( K105 ) is altered to an arginine ( K105R ) it has significantly reduced dephosphorylation rates and remains in the “on” state [79 , 80] . Similar mutational analysis at the same lysine position has also been demonstrated in NtrC [81 , 82] . Importantly , we observed several additional mutations in NmpR that activate motility but have not been identified or characterized for activation of RRs in other organisms . Future studies of how these mutations confer activation to this NtrC-like RR will be important in our understanding of bacterial gene regulation , generally , but also in the context of important RR virulence factors and our potential in targeting these proteins with novel antibiotics [83 , 84] . Finally , given several notable properties of this newly described pathway , future investigations have the potential to further our understanding of the dynamics of complex TCS signaling , atypical regulation of hybrid SKs , functions of protoglobin containing sensors , and global cellular NtrC-like RR networks .
All M . xanthus strains are listed in Table S1 . E . coli DH5α and TOP10 ( Life Technologies , Grand Island , NY ) strains were used for cloning and grown in LB ( Becton-Dickinson , Franklin Lakes , NJ ) at 37°C . E . coli BL21 DE3 ( Life Technologies , Grand Island , NY ) was used for protein production and were grown in Terrific Broth . M . xanthus strains were cultivated in Casitone Yeast Extract ( CYE ) medium [85] at 32°C with shaking at 220 rpm . M . xanthus was grown on agar plates at 1 . 5% agar , except in motility assays where 0 . 5% agar was used . For all experiments , the M . xanthus DZ2 strain was used as the wild-type strain [86] . Kanamycin was used at a final concentration of 50 μg ml-1 , tetracycline at 10 μg ml-1 , and spectinomycin at 100 μg ml-1 ( E . coli ) or 1 mg ml-1 ( M . xanthus ) , when required . PCR reactions were performed using Failsafe polymerase with Buffer K ( Epicentre Technologies , Madison , WI ) or high-fidelity Phusion polymerase with the manufactured suggested protocol ( New England Biolabs , Ipswich , MA ) . All primers ( Table S2 ) were supplied by Integrated DNA Technologies ( Coralville , IA ) and were designed based on sequences obtained from the Microbial Signal Transduction Database ( MIST2 . 2 , [87] ) . PCR products used for cloning or deletion mutant construction were initially cloned into the pCR2 . 1-TOPO vector ( Life Technologies , Grand Island , NY ) and sequence verified ( Nevada Genomics Sequencing Center , Reno , Nevada ) . DNAStar software ( Lasergene 7 . 2 ) was used for sequence analysis . Whole genome sequencing of M . xanthus DZ2 , the ΔpilR parental strain , and all suppressor mutants reported in this paper was performed by the University of Iowa Institute of Human Genetics , Genomics Division on the Illumina MiSeq Sequencer with 2x250 bp reads . DNAStar software ( Lasergene 7 . 2 ) was used for genome alignment and mutation identification based on the M . xanthus DK1622 annotated genome [27] . The mean genomic coverage for all samples was >60x and suppressor mutations identified by a 95% cut-off . Identified mutations were confirmed by traditional Sanger-sequencing methods . In-frame deletions of M . xanthus were constructed essentially as described elsewhere [26 , 88] . Briefly , to construct in-frame deletions , approximately 800 bp fragments upstream and downstream of each gene to be deleted was amplified and cloned into pBJ113 [89] . Final plasmid constructs were electroporated into appropriate strains of M . xanthus [90] . Transformants were selected on CYE agar with kanamycin and screened for proper insertion of the plasmid by PCR , followed by counterselection on galactose [91] . All final deletion mutants were verified by PCR using primers spanning the deletion junction and/or internal to the gene deletion . Complementation of each gene was achieved by cloning the entire opening reading frame down-stream of the heterologous high-expression groES promoter ( 623 bp upstream of mxan_4894 [43 , 44]; S1 Fig ) or , the putative native promoter of nmpR ( 585 bp upstream of mxan_4236; Fig 2A ) . Mutagenesis of vector constructs was done with QuikChange Site-Directed Mutagenesis Kit ( Agilent Technologies ) or Q5 Site-Directed Mutagenesis Kit ( New England Biolabs ) per the manufacturer’s instructions and constructed mutations sequenced verified . The plasmid pSUM117 was used to integrate the complementation constructs into the Mx9 phage attachment site . This plasmid is a pBCKS+ derivative that contains the attPMx9 site for integration . Final clones were again verified by PCR . T4P-dependent motility assays were performed as previously described [40] . Briefly , cultures were grown overnight to log phase . Cells were washed in MMC buffer ( 20 mM MOPS , pH 7 . 6 , 4 . 0 mM MgSO4 , 2 mM CaCl2 ) , suspended in MMC to ~2x109 CFU ml-1 , and 10 μl cell suspensions spotted onto dry CYE 0 . 5% agar plates for growth/motility at 32°C . Overall motility rates were determined by measuring the diameter of the spreading colony at 24 hour intervals for five days . The reported rates reflect the average slope of at least three colonies . An unpaired t-test was used to analyze the significance of the differences of motility rate relative to wild-type or ΔpilR motility ( p<0 . 05 considered significant ) . Motility assays were monitored by microscopy using a Nikon SMZ10000 dissecting microscope . Images were taken using a QImaging Micropublisher charge-coupled device camera and processed with QCapture software . While we attempted to replicate motility conditions between samples , the experiments presented in this study were not all done at the same time . Individual images are representative of multiple observations . To assess the sedimentation of M . xanthus cells , assays were conducted as described previously [35] by measuring the optical density at 600 nm of standing cell suspensions prepared as for the motility assays . The reported sedimentation percentage is the remaining optical density relative to the original optical density of each strain after two hours . Quantitative measurement of EPS production was conducted as previously described [92] by again preparing cell suspensions in MMC buffer to ~2x109 CFU ml-1 with a final concentration of 30 μg ml-1 trypan blue in each preparation . The trypan blue was incubated with the cells for 30 min at room temperature , at which time the cells were pelleted and the supernatant removed . Remaining trypan blue in the supernatant was measured by absorbance at 585 nm . Initial trypan absorbance was measured with control samples of trypan blue in which no cells were included . Relative amount of trypan blue binding for each strain was compared to wild-type , which was set at 100% . M . xanthus strains were grown overnight to log phase , pelleted , suspended in 10 mM Tris-HCl pH 7 . 5 , and lysed by sonication . Total protein content was determined by Bradford assay and cell lysates were loaded on a 12% SDS-PAGE gel normalized to total protein concentration . After electrophoresis , the gel was blotted onto a PVDF membrane , blocked with 5% skim milk , and incubated with anti-PilA antiserum [35] at 1:5000 followed by secondary anti-rabbit horseradish-peroxidase-coupled antibody ( Sigma ) at 1:10 , 000 . Detection was performed using Western Lightning-ECL ( PerkinElmer ) . To assess the activity of the groES promoter , the same promoter used for complementation experiments was incorporated upstream of lacZ . This plasmid is a pSUM117 derivative containing a lacZ reporter and attPMx9 attachment site for integration . Incorporation of the plasmid in single copy at the Mx9 phage attachment site was confirmed by PCR . Expression of groES in the resulting strains was assessed following lawn growth on CYE plates for four hours . Cells were scraped from the multiple plates and pooled . To quantify β-galactosidase activity , bacteria were pelleted , suspended in Z-buffer , and lyzed by sonication . Total protein concentration was determined with Bradford reagent and spectrophotometry . Activity was quantified by incubation of normalized cell lysates at 37°C with o-nitrophenol and presented as Miller units ( ( OD420/mg protein * time in minutes ) *1000 ) . All proteins were purified using a standard column purification protocol [12] . All proteins were expressed using the isopropyl-β-D-thiogalactopyranoside ( IPTG ) -inducible vector pET28a ( Novagen ) . For purification , E . coli BL21 DE3 strains containing an expression vector was grown in 100 ml Terrific Broth in a 250-ml Erlenmeyer flask at 37°C with protein expression induced with 1 mM IPTG . After overnight incubation at 16°C , cells were pelleted at 8 , 000g for 15 min . When necessary , pellets were stored at -20°C until purification . Cell pellets were thawed , suspended in 10 ml of cell lysis buffer ( 25 mM Tris pH 7 . 6 , 125 mM NaCl , 5 mM imidazole , 1% Triton X-100 , and 0 . 625 g of CelLytic Express ( Sigma-Aldrich ) . Cells were incubated for 1 h before lysates were clarified by centrifugation at 8 , 000 g . One ml of His-Select cobalt affinity resin ( Sigma-Aldrich ) was equilibrated in wash buffer ( 25 mM Tris pH 7 . 6 , 125 mM NaCl , 5 mM imidazole , 1% Triton X-100 ) . Whole cell lysates were equilibrated over the resin by gravity , washed once with 10 ml of wash buffer , twice with 10 ml wash buffer with 20 mM imidazole , and finally eluted with 5 ml of elution buffer ( 25 mM Tris [pH 7 . 6] , 125 mM NaCl , 250 mM imidazole ) . Samples were dialyzed overnight at 4°C against dialysis buffer ( 25 mM Tris pH 7 . 6 , 125 mM NaCl , 1 mM dithiothreitol , 1% Triton X-100 , 50% glycerol , 0 . 5 mM EDTA ) . Protein purity was assessed by standard denaturing gel electrophoresis , and concentration determined using the Bradford reagent . All purified proteins were stored in dialysis buffer at -20°C . Kinase assays were performed as described previously [12] . Briefly , 5 μl of 50 μM SK stock was added to 5 μl 10x kinase buffer ( 250 mM Tris [pH 7 . 6] , 500 mM KCl , 10 mM MgCl2 , 10 mM MnCl2 , 10 mM CaCl2 and 10 mM β-mercaptoethanol ) and 35 μl dH2O . Reaction mixtures were incubated at room temperature and started by the addition of 5 μl of an ATP mix ( 250 μM ATP , 3 μM [γ-32P]ATP ) . Aliquots were removed at various time points and stopped by the addition of an equal volume of 4x SDS loading buffer . Samples were resolved by electrophoresis on 12% SDS-polyacrylamide gels . Gels were exposed for 3 hours to a phosphor screen and visualized using a Typhoon Imager ( GE ) . For phosphotransfer profiling , kinases were allowed to autophosphorylate for two hours until maximal phosphorylation levels were reached . A 5 μl aliquot of the phosphorylated SK was mixed with 5 μl of 10 μM RR , and reactions were stopped at various times by addition of 4x SDS loading buffer . Incorporation of labeled phosphoryl groups was analyzed as detailed above and figure images are representative of multiple experiments . Full-length wild-type NmpR was expressed , purified , and quantified as above . DNA probes were generated by PCR with Failsafe polymerase with Buffer K ( Epicentre Technologies , Madison , WI ) , purified with Qiagen QIAquick PCR Purification Kit , and quantified by spectrophotometry . Probes were diluted in 10 mM Tris-HCl , pH8 . 5 to 100 fmol/μl with a final quantity of 200 fmol used per individual binding reaction . Probes were incubated with increasing concentrations of NmpR , as indicated , in a reaction containing a final concentration of 20 mM Hepes pH 7 . 6 , 1 mM EDTA , 10 mM ( NH4 ) 2SO4 , 1 mM DTT , Tween 20 0 . 2% , and 30 mM KCl . Reaction conditions were determined empirically in our laboratory but were based on recommendations from the DIG Gel Shift Kit ( Sigma-Aldrich ) . NmpR was incubated with the DNA probes for 15 minutes at room temperature and then loaded on 8% acrylamide non-denaturing PAGE gels ( 8% acrylamide , 10 mM Tris-HCl , 1 mM EDTA , 400 mM glycine ) . Gels were resolved at 60 v for ~2 hours , then stained with ethidium bromide ( 10 μg per 10 ml H2O ) for 15 minutes prior to visualization . Presented images are representative of reproducible independent reactions . Several online software platforms were utilized to identify homologues of NmpRSTU . Protein sequences and some synteny analysis were obtained from the Mist 2 . 2 database [87] . Homologues were also identified by blastp ( blast . ncbi . nlm . nih . gov ) against δ-proteobacteria ( taxid:28221 ) specifically , or by excluding δ-proteobacteria . Homologues and synteny were also identified at Pfam ( pfam . xfam . org; [93] ) and SyntTax ( archaea . u-psud . fr/synttax/; [94] ) . The structure of NmpR was unbiasedly modeled using SWISS-MODEL ( swissmodel . expasy . org ) [95] . | Sensing and responding to the environment are critical for bacterial survival and adaptation . Many bacteria sense environmental signals through biological machinery referred to as two-component signaling systems . These signaling systems are composed of a sensor histidine kinase and a cognate response regulator and serve to regulate gene expression in response to environmental cues . Due to their importance for survival , many bacterial genomes encode multiple two-component systems , some of which arose from gene duplication events . However , the mechanisms by which two-component systems integrate multiple inputs within larger networks remains poorly defined . In this study , we demonstrate that suppressor mutations within a Myxococcus xanthus signaling system of previously unknown function restore Type-IV pilus-dependent motility to cells lacking PilR , a response regulator typically required for production of the pilin monomer PilA . Suppressor mutations in ΔpilR cells led to activation of a related signaling system , resulting in expression of pilA and the corresponding production of PilA to restore motility . Thus , both PilSR and NmpSR coordinately regulate motility in M . xanthus . Collectively , our results expand our understanding of the activation of response regulators via phosphorylation and the mechanisms controlling multi-component signal transduction systems . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"cell",
"motility",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"oxygen",
"pathogens",
"mutation",
"regulator",
"genes",
"gene",
"types",
"genomic",
"signal",
"processing",
"proteins",
"chemistry",
"pathogen"... | 2018 | Suppressor mutations reveal an NtrC-like response regulator, NmpR, for modulation of Type-IV Pili-dependent motility in Myxococcus xanthus |
The visual world is complex and continuously changing . Yet , our brain transforms patterns of light falling on our retina into a coherent percept within a few hundred milliseconds . Possibly , low-level neural responses already carry substantial information to facilitate rapid characterization of the visual input . Here , we computationally estimated low-level contrast responses to computer-generated naturalistic images , and tested whether spatial pooling of these responses could predict image similarity at the neural and behavioral level . Using EEG , we show that statistics derived from pooled responses explain a large amount of variance between single-image evoked potentials ( ERPs ) in individual subjects . Dissimilarity analysis on multi-electrode ERPs demonstrated that large differences between images in pooled response statistics are predictive of more dissimilar patterns of evoked activity , whereas images with little difference in statistics give rise to highly similar evoked activity patterns . In a separate behavioral experiment , images with large differences in statistics were judged as different categories , whereas images with little differences were confused . These findings suggest that statistics derived from low-level contrast responses can be extracted in early visual processing and can be relevant for rapid judgment of visual similarity . We compared our results with two other , well- known contrast statistics: Fourier power spectra and higher-order properties of contrast distributions ( skewness and kurtosis ) . Interestingly , whereas these statistics allow for accurate image categorization , they do not predict ERP response patterns or behavioral categorization confusions . These converging computational , neural and behavioral results suggest that statistics of pooled contrast responses contain information that corresponds with perceived visual similarity in a rapid , low-level categorization task .
Complex natural images are categorized remarkably fast [1] , [2] , sometimes even faster than simple artificial stimuli [3] . For animal and non-animal scenes , differences in EEG responses are found within 150 ms [4] and a correct saccade is made within 120 ms [5] . This speed of processing is also found for other scene categories [6] and may require less attentional resources compared to artificial images [7] , [8] . This suggests that relevant visual information is rapidly and efficiently extracted from early visual responses to natural scenes . However , the neural computations involved in this process are not known . Importantly , natural images differ from other image types such as white noise in low-level properties ( e . g . , sparseness ) , leading to the suggestion that the visual system has adapted to these low-level properties [9] . This idea paved the way for optimal coding models for natural images [10] , [11] and successful predictions of response properties of visual neurons [12] . Recent work identified statistical properties that differ even within the class of natural images , e . g . between natural scene parts [13] , [14] or natural image categories [15] , showing that image statistics such as power spectra of spatial frequency content or distributions of local image features are informative about scene category . The fact that it is mathematically possible to distinguish categories based on image statistics , however , does not imply that they are used for categorization in the brain . Image statistics may not be sufficiently reliable , or their computation may not be suitable for neural implementation [12] , [16] . We recently showed that statistics derived from the frequency histogram of local contrast – summarized by two parameters of a Weibull fit , Fig . 1A – explain up to 50% of the variance of event-related potentials ( ERPs ) recorded from visual cortex [17] . These parameters inform about the width and shape of the histogram , respectively , and appear to describe meaningful variability between images ( Fig . 1B ) . Importantly , we found that these parameters can be reliably approximated by linear summation of the output of localized difference-of-Gaussians filters modeled after X- and Y-type LGN cells , suggesting that this global information may be available to visual cortex directly from its early low-level contrast responses [17] . Moreover , we found that output of contrast filters with a larger range of receptive field sizes captures additional image information [18] . This is not surprising since objects in natural scenes appear at many distances and hence spatial scales [19] . In the present implementation , the model first estimates at which scale relevant contrast information is present , as well as characteristics of the distribution of contrast strengths at those scales . This model , which approximates early visual population responses based on spatially pooled contrasts , was able to explain almost 80% of ERP variance to natural images [18] . These previous findings suggest that images with more similar contrast response statistics evoke more similar early visual activity . Could these responses already contain relevant information about the stimulus for rapid categorization ? The two parameters appear to index meaningful information such as degree of clutter , depth and figure-ground segmentation [17] , but how the two dimensions in Fig . 1B influence perception has not been examined . The goal of the current study was thus to explore what type of visual information is contained in the variance of the earliest visual contrast responses that is so well described by these two parameters . Specifically , we were interested in whether these parameters cannot only predict variance in visual activity , but also ‘variance in perception’ . In other words , do images with more similar contrast statistics also lead to more similar perceptual representations , and perhaps ultimately , to similar images being considered a single category ? We aimed to answer this question in a data-driven manner , by investigating 1 ) which images group by similarity early in visual processing and 2 ) whether this grouping matches with perceived similarity of those images . For the first part of this question , we obtained reliable evoked responses to individual images . The advantage of this approach relative to traditional ERP analysis ( which is based on averaging many trials across individual images within an a priori determined condition ) is that it provides much richer data [20]–[24] that can be used for model selection . We used these single-image evoked responses to compute dissimilarities in ‘neural space’ , similar to the pattern analysis approach used in fMRI [25] , [26] . This allowed us to track , over the course of the ERP , to what extent the representation of an image is ( dis ) similar to all images in the data set . For the second part of the question , we needed to obtain an image-specific behavioral judgment of perceived visual similarity . However , simply judging similarity of natural scenes is problematic , because these images obviously contain rich semantic content: there are many features of natural scenes that can be similar or dissimilar , which is likely to lead to different categorization strategies by different subjects . Also , it is uncertain to what extent specific semantic tags that are provided by the researcher ( e . g . ‘openness’ or ‘naturalness’ , [27] ) , can be uniformly interpreted as a relevant stimulus dimension that has a linear mapping to processing in early vision . Therefore , to explore the variance explained by contrast response statistics in a bottom-up way , we used stimuli that were simplified model images of natural scenes ( ‘dead leaves’ , Fig . 2A ) , which have similar low-level structure as natural scenes ( e . g . 1/f power spectra ) but are devoid of semantic content . These images are created by filling a frame with objects - much like fallen leaves can fill a forest floor – and are used in computer vision to study , for example , how the appearance and the distribution of these objects influences the low-level structure of natural scenes [28] . By manipulating properties of the objects in a controlled manner , we created distinct image categories , and then tested whether differences between these categories in contrast statistics matched with behaviorally perceived similarity by letting human observers perform a same-different categorization task on all combinations of image categories . Specifically , we used the space formed by the two Weibull parameters to compute geometric distances between images in contrast statistics , and used these distances as quantitative predictors of dissimilarity [29]–[31] . We thus tested whether these parameters can predict the extent to which image categories induced dissimilar single-image EEG responses ( experiment 1 ) and whether they match with perceptual categorization at the behavioral level ( experiment 2 ) . We predicted that images with very different Weibull statistics would appear less similar , i . e . be less often confused than images from categories with similar statistics . By using controlled images that we quantified using a model originally derived from contrast responses to natural images , we aim to build a bridge between findings obtained with systematic manipulation of artificial stimuli and those obtained with more data-driven natural scene studies . For purpose of comparison , and to better understand which statistical information is captured by the Weibull parameters , we also tested two other global contrast statistics ( Fig . 2C ) . Following [32] we calculated the intercept and slope of the average power spectrum to parameterize spatial frequency information , a commonly used measure of low-level information in scene perception . In addition , we followed [33] to derive the skewness and kurtosis of the contrast distribution for a range of spatial scales: these higher-order properties of distributions have previously been suggested ( e . g . [34] , [35] to reflect low-level differences between images that are relevant for perceptual processing . We find that Weibull statistics explain substantial variance in evoked response amplitude to the dead leaves images , predicting clustering-by-category of occipital ERP patterns within 100 ms of visual processing . In addition , they correlate with human categorization behavior: specific confusions were made between categories with similar Weibull statistics . By comparison , Fourier power spectra and skewness and kurtosis can be used for accurate classification of image category , but fail to predict neural clustering and behavioral categorization . These convergent results provide evidence for relevance of pooled contrast response statistics in rapid neural computation of perceptual similarity .
The experiments reported here were approved by the Ethical Committee of the Psychology Department at the University of Amsterdam; all participants gave written informed consent prior to participation and were rewarded with study credits or financial compensation ( 7 euro/hour ) . Gray-scale dead leaves images ( 512×512 pixels , bit depth 24 ) were generated using Matlab . Images contained randomly placed disks that were manipulated along 4 dimensions ( opacity , depth , size and distribution ) to create 16 categories . Disks were either opaque or transparent; intensity at the outer edges of the disk was either constant ( leading to a 2D appearance ) or decaying ( 3D appearance ) , and disk size was determined by drawing randomly from a range of small , medium or large diameters ( exact settings as in [28] . Twelve categories were created by systematically varying these properties of power-law distributed disks . Four more categories were created using medium-diameter , exponentially distributed disks that could be 2D or 3D and opaque or transparent . For each category , 16 images were created using these category-specific settings: the random placement and use of ranges of diameter sizes ensured that each of these 16 images was unique . This procedure thus resulted in a total of unique 256 images , divided into 16 distinct categories , which were used for experimentation ( Fig . 2A ) .
If we set out all 256 dead leaves images against the three sets of image statistics ( Weibull parameters , Fourier parameters and skewness/kurtosis ) , stimuli cluster by category in all cases , with Fourier parameters leading to the most separable clusters ( Fig . 4A–C ) . There were considerable correlations between the various parameters ( Fig . 4D; individual correlations plots in Fig . S2 ) . Skewness and kurtosis correlated highly ( ρ = 0 . 91 , p<0 . 0001 ) , but other significant correlations are observed as well , for example between Fourier slope and the Weibull beta parameter ( ρ = 0 . 57 , p<0 . 0001 ) and also between the two Weibull parameters ( ρ = 0 . 48 , p<0 . 001 ) . A correlation of similar magnitude was also observed [17] for natural scenes , supporting the notion that the dead leaves stimuli used here have similar low-level structure as natural stimuli . Interestingly , however , the ‘similarity spaces’ formed by each set of parameters are quite different between the various models . If Weibull parameters determine the axes of the similarity space ( Fig . 4A ) , highly cluttered images with many strong edges ( e . g . 2D opaque stimuli with small disks ) are located in the upper right corner ( high gamma , high beta ) ; images containing fewer edges ( e . g . with larger disks ) are found more on the left ( low gamma ) ; and most of the transparent stimuli , with weak edges , cluster together in the bottom of the space ( low beta ) . For Fourier intercept and slope ( Fig . 4B ) , transparent categories are highly separated across the space: however , most images with strong edges end up in a similar part of the space ( low slope , high intercept ) . Based on either skewness or kurtosis ( Fig . 4C ) , a few categories are distinct , but most tend to cluster together . These qualitative results suggest that all parameters are informative about clustering of image categories , but that they index different image properties . Importantly , they give rise to different predictions about which categories should lead to similar evoked responses based on overlapping parameter values . We tested these predictions using the single-image ERP data .
Whether low-level statistics are indeed actively exploited during scene or object categorization is a topic of considerable debate . Whereas some studies report that manipulation of low-level properties influences rapid categorization accuracy [54] , [55] as well as early EEG responses [56] , [57] , other studies have shown that not all early visual activity is obliterated by equation of those properties [58]–[60] and , conversely , that early sensitivity to diagnostic information is revealed in stimuli that do not differ in low-level statistics [20] , [61] . We find that , at least for our set of simplified models of natural scene images , early differences in ERPs are correlated with low-level contrast statistics that are themselves also directly predictive of perceptual similarity . It is however likely that the degree to which low-level properties are relevant for processing of natural image categories is highly dependent on stimulus type and context , even within actual natural scene stimuli: for example , low-level information may influence rapid detection of faces to a larger extent than objects [22] and the effects of low-level statistics on animal detection may interact with scene category ( man-made vs . natural ) [62] . In addition , the present work is very different from these previous reports in that our experiments did not require formation of a high-level representation but only a same-different judgment . There are also notable differences between our ERP effects and those obtained with standardized object/scene categories: our maximum explained variance was found at around 100 ms , whereas those studies report sensitivity starting at 120 ms and onwards [63]–[66] . Maximal sensitivity of evoked activity to faces and objects is found at lateral-occipital and parietal electrodes ( PO , e . g . [58] ) , whereas our correlations are clustered around occipital electrode Oz . This suggests that the dead leaves images may mostly engage mid-level areas of visual processing , such as those sensitive to textural information , e . g . V2 [24] , [67]–[69] . Our results implicate that clustering of image similarities at this level of processing can , in principle , already predict perceptual similarity – in turn , these similarities can be derived from Weibull contrast statistics . Given that for natural scenes , the Weibull statistics explain similar amounts of variance in EEG activity as reported here , we can hypothesize that image similarities as predicted by Weibull statistics are also present in evoked activity to actual natural scenes . If Weibull statistics indeed approximate meaningful global information in natural images , which image features do they convey ? By manipulating computational image categories in their perceptual appearance , we were able to get a better understanding of the information contained in the Weibull parameters . They appear to index the amount of clutter , i . e . are related to occlusion and object size . These properties may be relevant for natural scene categorization: a forest has a higher degree of clutter ( high gamma ) and lower mean edge strength ( high beta ) compared to a beach scene . An image containing a few strong edges ( low beta ) that are sparsely distributed ( low gamma ) has high probability of coinciding with a single salient object , for example a single bird against an empty sky , suggesting that these statistics may be relevant for object detection in natural scenes . Here , behavioral confusions ( and corresponding dissimilarities in ERP signals ) were found between stimuli without coherent edge information ( transparent stimuli with either large or small disks ) , or that were highly cluttered ( opaque stimuli with small disks ) which were exactly the categories that overlapped in Weibull parameter values . For comparison , we computed Fourier power spectra and higher-order properties of the contrast distribution ( skewness and kurtosis ) , two sets of statistics that each index different sources of information in natural images: spatial frequency content and central moments of the contrast distribution , respectively . Deviations in the power spectra of natural images inform about variations in contrast across spatial scales: the slope and intercept parameters describe the ‘spectral signature’ of images [32] which is diagnostic of scene category [15] . Skewness and kurtosis were proposed to be relevant for texture perception [35] , [70] which in turn can be important for feature detection [53] , [71] and the presence of featureless regions of images [34] , [72] . Our results confirm that both frequency content and central moments of the contrast distribution inform about image properties: both lead to accurate image classification . However , in the present study they did not predict neural and behavioral categorization patterns , suggesting that these statistics may not be plausible computations involved in visual processing of the dead leaves images . Even though we used controlled , computationally defined image categories , it is still possible that an image property other that the contrast statistics tested here will provide a better prediction of the ( neural and behavioral ) data , for example one of the manipulations used to create the image categories ( e . g . , opacity ) . However , neither the observed clustering-by-category of ERPs in the RDM , nor the pattern of categorization errors in behavior mapped clearly onto one of the manipulations used to create the categories ( e . g . , opaque vs . transparent; as is visible in Fig . 7B , there are also differences within opaque and transparent categories , and this complex pattern of clustering is only predicted by Weibull statistics ) . Why is the Weibull model better than widely used contrast statistics in predicting early neural and perceptual similarity ? Although higher order moments of distributions can be diagnostic of textural differences , they may in practice be difficult for the visual system to represent [35] . In addition , it has been suggested that rather than amplitude spectra , phase information derived from the Fourier transform [73] , [74] , or the interaction between these two [75] , [76] contains diagnostic scene information . The reason that higher-order statistics derived from the phase spectrum may contain perceptually relevant information [77] is that they carry edge information . In the Weibull model , contrasts , i . e . non-oriented edges , are explicitly computed ( as the response of LGN-type neurons ) and evaluated at multiple spatial scales . The model may thus be able to capture information contained both in power spectra ( scale statistics ) as well as central moments ( distribution statistics ) . The Weibull parameters appear to reflect different aspects of low-level information: the beta parameter varies with the range of contrast strengths present in the image , reflecting overall contrast energy , whereas the gamma parameter varies with the degree of correlation between local contrast values , reflecting clutter or spatial coherence . Obviously , the Weibull fit is still a mathematical construct . However , the two parameters can also be approximated in a more biologically plausible way: with our previous single-scale model [17] , we demonstrated that simple summation of X- and Y-type LGN output corresponded strikingly well with the fitted Weibull parameters . Similarly , if the outputs of the multi-scale filter banks used here ( reflecting the entire range of receptive field sizes of the LGN ) are linearly summed , we again obtain values that correlate highly with the Weibull parameters obtained from the contrast histogram at minimal reliable scale ( S . Ghebreab , H . S . Scholte , V . A . F . Lamme , A . W . M Smeulders , under review ) . This suggests that Weibull estimation can in fact be reduced to pooling of neuronal population responses by summation , which is a biologically realistic operation . Why would summation of contrast responses of low-level neurons convey the same information as the Weibull parameters ? This is likely a result of the structure of the world itself: distributions of contrast in natural images tend to range between power-law and Gaussian , which is the family of distributions that the Weibull function can capture [78] . It appears that this statistic simply provides a good characterization of the dynamic range of the low-level input to the visual cortex when viewing natural images . Since our brain developed in a natural world , early visual processing may take advantage of this regularity in estimating global properties to arrive at a first impression of scene content . The present results extend our previous findings [17] , [18] with natural images to other image types ( computational categories ) and to prediction of behavioral categorization . Interestingly , even though the subjects in experiment 1 ( EEG ) were not engaged in categorization of the dead leaves images , their results generalize to the behavioral categorization patterns that were found in experiment 2 , suggesting that similarity of bottom-up responses measured in EEG - in a different person - can be predictive of the perceived similarity during categorization of these images . This observation is now restricted to computationally defined categories . An interesting question for future work is whether in construction of high-level categorical representations of natural stimuli - considered a computationally challenging task - the brain actively exploits the pattern of variability of the population response to low-level information , estimated from early receptive field output . Contrary to the classical view of the visual hierarchy ( e . g . , [79] ) it has been proposed that a rapid , global percept of the input ( gist ) precedes a slow and detailed analysis of the scene [80]–[83] . Natural image statistics provide a pointer to information that could be relevant for such a global percept [84] , [85] . However , the mechanism by which global information can be rapidly extracted from low-level properties is not directly evident from natural image statistics alone . As explained above , in our model , the statistics are derived from a biologically realistic substrate ( the response of early visual contrast filters ) . We suggest that to build a realistic model of natural image categorization , it is essential to understand how statistics derived from very early , simple low-level responses can contribute to gist extraction . In conclusion , our findings suggest that global information based on low-level contrast can be available very early in visual processing and that this information can be relevant for judgment of perceptual similarity of controlled image categories . | Humans excel in rapid and accurate processing of visual scenes . However , it is unclear which computations allow the visual system to convert light hitting the retina into a coherent representation of visual input in a rapid and efficient way . Here we used simple , computer-generated image categories with similar low-level structure as natural scenes to test whether a model of early integration of low-level information can predict perceived category similarity . Specifically , we show that summarized ( spatially pooled ) responses of model neurons covering the entire visual field ( the population response ) to low-level properties of visual input ( contrasts ) can already be informative about differences in early visual evoked activity as well as behavioral confusions of these categories . These results suggest that low-level population responses can carry relevant information to estimate similarity of controlled images , and put forward the exciting hypothesis that the visual system may exploit these responses to rapidly process real natural scenes . We propose that the spatial pooling that allows for the extraction of this information may be a plausible first step in extracting scene gist to form a rapid impression of the visual input . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"visual",
"system",
"cognitive",
"neuroscience",
"behavioral",
"neuroscience",
"cognition",
"computational",
"neuroscience",
"psychophysics",
"biology",
"computational",
"biology",
"sensory",
"systems",
"neuroscience",
"sensory",
"perception",
"neuroimaging",
"coding",
"mecha... | 2012 | Spatially Pooled Contrast Responses Predict Neural and Perceptual Similarity of Naturalistic Image Categories |
Patients with clinical manifestations of leishmaniasis , including cutaneous leishmaniasis , have limited treatment options , and existing therapies frequently have significant untoward liabilities . Rapid expansion in the diversity of available cutaneous leishmanicidal chemotypes is the initial step in finding alternative efficacious treatments . To this end , we combined a low-stringency Leishmania major promastigote growth inhibition assay with a structural computational filtering algorithm . After a rigorous assay validation process , we interrogated ∼200 , 000 unique compounds for L . major promastigote growth inhibition . Using iterative computational filtering of the compounds exhibiting >50% inhibition , we identified 553 structural clusters and 640 compound singletons . Secondary confirmation assays yielded 93 compounds with EC50s ≤ 1 µM , with none of the identified chemotypes being structurally similar to known leishmanicidals and most having favorable in silico predicted bioavailability characteristics . The leishmanicidal activity of a representative subset of 15 chemotypes was confirmed in two independent assay formats , and L . major parasite specificity was demonstrated by assaying against a panel of human cell lines . Thirteen chemotypes inhibited the growth of a L . major axenic amastigote-like population . Murine in vivo efficacy studies using one of the new chemotypes document inhibition of footpad lesion development . These results authenticate that low stringency , large-scale compound screening combined with computational structure filtering can rapidly expand the chemotypes targeting in vitro and in vivo Leishmania growth and viability .
Leishmaniasis is endemic in >85 developing countries with >1 . 5 million estimated cases occurring each year and an additional 350 million people at risk of infection [1] . Increased travel and migration within the tropics , subtropics , Middle East and Southern Europe as well as global climate and environmental changes are making leishmaniasis a considerable risk for populations in geographic regions previously unaffected by the disease [2]–[5] . As a result , there has been a progressive expansion of leishmaniasis endemic regions as well as a concomitant increase in the total number of reported leishmaniasis cases , often in epidemic proportions ( i . e . , with 100 , 000–200 , 000 individuals infected ) [6]–[9] . Transmission of leishmaniasis most commonly occurs via an infected phebotomine sandfly . Leishmaniasis can also be transmitted , albeit rarely , through blood transfusions , especially to individuals with immature or compromised immune systems , further expanding and globalizing the number of at-risk populations [10] . With clinical manifestations ranging from cutaneous ( CL ) and mucocutaneous ( M-CL ) to visceral , leishmaniasis has profound cultural and socioeconomic repercussions due to overt disability , disfigurement or scarring , and death [4] , [11]–[15] . Despite the prevalence of leishmaniasis and its impact on human life , there are no vaccines or prophylactic drugs for any form of the disease . Current chemotherapeutic treatments rely heavily on the use of the pentavalent antimonials , sodium stibogluconate , and meglumine antimoniate , which were first introduced more than a half century ago [16]–[18] . Significantly , these compounds have been used without refinement for decades , have serious side effects and are declining in efficacy due to chemoresistance [19]–[22] . Second-line drugs , such as pentamidine and amphotericin B , are available but they too have significant untoward effects and pharmacological liabilities [4] , [18] . Moreover , these existing leishmanicidals often require continuous clinical surveillance , have invasive or painful routes of administration and , are expensive for endemic areas . Others have attempted to augment the pool of available leishmanicidals by exploiting drugs approved for other diseases . Among newer treatments are the use of rifampicin , tamoxifen , doxocycline , monomycine , trimethoprim and nifurtimox; however , these agents are generally associated with limited anti-leishmanial efficacy [18] , [23]–[29] . To maximize effectiveness and minimize toxicity , the choice of drug dosage and duration of therapy should be individualized based on the region of disease acquisition and host factors such as immune status . Also , we know that some drugs and regimens are effective only against certain Leishmania species or strains and only in certain areas of the world . The idea that one drug might treat all forms of leishmaniasis has rapidly lost popularity . Regrettably , there is a paucity of large-scale drug discovery efforts focusing on the design of new small molecules ( i . e . drugs ) that can treat individuals with leishmaniasis . This deficiency has contributed to leishmaniasis being classified as a neglected disease , with CL being the most neglected among the clinical manifestations of leishmaniasis [13] . Thus , there is a strong need to identify potential new drug treatments for specific clinical manifestations of leishmaniasis , and especially novel chemotherapeutics for CL . As with other pathogenic diseases , genetic tools and genomic sequencing information are now available for multiple Leishmania spp . enabling a molecular target- driven approach to anti-leishmanial drug discovery [30]–[32] . Nonetheless , the low success rate of those efforts may reflect an incomplete understanding of the complexities of leishmaniasis and the significance of the proposed molecular targets to parasite growth or survival [33] , [34] . Thus , whole parasite phenotypic anti-leishmanial drug discovery remains appealing . Until recently , however , most efforts to identify new leishmanicidals via whole parasite screening have concentrated on the exploitation of limited , small-scale activities using discrete , focused compound sets or compounds with known pharmacological actions [35] , [36] . Consequently , the identification of novel leishmanicidal chemotypes has been effectively limited by screening throughput as well as compound library diversity . We postulate that the identification of new anti-leishmanial chemotypes can be rapidly accelerated by using low stringency , high throughput screening ( HTS ) methodologies with large diverse compound libraries combined with computational tools . For maximum utility , the HTS assays should be well-validated , integrated with data management and capture systems , have a simple assay format , be relatively inexpensive and , be coupled with secondary assays to expedite confirmation of the activity and specificity of novel chemotypes [37]–[39] . In the work presented herein , we developed and implemented a multi-tiered compound screening paradigm to identify and confirm novel leishmanicidal chemotypes . Our screening strategy was founded on a validated L . major ( taxonomy id 5664 ) promastigote drug susceptibility HTS assay , which we used to screen a structurally diverse 196 , 146 compound library at low stringency ( i . e . , a relatively high compound screening concentration - 10 µM ) . Promastigotes are easy to use and there is evidence that they provide a good model for gauging a compound's leishmanicidal activity [40]–[42] . The selected assay detection reagent , alamar blue , is simple , inexpensive , easily adapted to automated HTS procedures and has been frequently used to identify and characterize leishmanicidal compounds [43] , [44] . Our primary aim was to maximize the potential chemical diversity of the L . major promastigote growth inhibitory chemotypes identified . Thus , we purposefully screened a large chemical library at a relatively high initial compound concentration to yield the maximum number of active compounds . To reduce the candidate compounds to a manageable size , we exploited computational methods to cluster chemotypes . We termed this integrated approach HILCES for high throughput , low-stringency , computationally enhanced small molecule screening . Representative members of each cluster and the unassigned compounds , i . e . singletons , were then sequentially characterized with respect to potency , specificity of response , and predicted in silico ADMET . Significantly , the use of an annotated public compound library enabled us to determine compound specificity by comparing its bioactivity in up to 369 additional biochemical or phenotypic assays . Moreover , specific molecular targets were suggested that might be critical to Leishmania growth , viability and survival . Selected compounds also demonstrated in vivo efficacy in a murine model system .
Black , clear bottom tissue culture treated 384-well microtiter plates were purchased from Greiner ( Monroe , NC ) and used for all experiments . Alamar blue ( Cell Titer Blue ) was purchased from Promega ( Madison , WI ) ; tamoxifen from MP Biomedicals ( Solon , OH ) ; dimethyl sulfoxide ( DMSO ) , aphidicolin from Sigma-Aldrich ( St . Louis , MO ) ; phenyltoloxamine , clotrimazole , sangivamycin and amphotericin B from VWR ( West Chester , PA ) ; disulfiram from Fisher Scientific ( Pittsburgh , PA ) ; pentamidine from Toronto Research Chemicals ( Ontario , Canada ) and; acivicin from Biomol ( Plymouth Meeting , PA ) . The PubChem CID compounds 786799 , 742546 , 760847 , 2946668 , 757789 , 2851545 , 728862 , and 16187595 were obtained from Chembridge ( San Diego , CA ) . All purchased compounds were subjected to quality control testing by their respective manufacturers . L . major promastigotes ( MHOM/SA/85/JISH118 ) ( a kind gift from Dr . Frederick Buckner ) were maintained in Medium 199 ( pH 7 . 2 ) ( Invitrogen , Carlsbad , CA ) supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) ( Hyclone , Logan , UT ) , penicillin ( 100 units/mL ) and streptomycin ( 100 µg/mL ) as previously described in Buckner and Wilson [45] . Promastigotes were grown in vented T75 tissue culture flasks and maintained at 28°C . Promastigote cultures were initiated at 105 parasites per mL and subcultured every 3–4 days . L . major promastigote counts were performed in duplicate using a hemocytometer and particle counter ( Beckman Coulter , Fullerton , CA ) . For HTS assays , L . major promastigote cultures were harvested during exponential growth phase ( ∼2 . 0–3 . 0×107 parasites/mL ) and were not maintained past passage 20 . Axenic amastigote-like parasite populations were derived from stationary growth phase L . major promastigotes and were maintained in Schneider's medium ( pH 4 . 9 ) supplemented with 10% heat-inactivated FBS , penicillin ( 100 units/mL ) , streptomycin ( 100 µg/mL ) , L-glutamine ( 2 mM ) and cultured at 32°C with 5% CO2 . This parasite population was specifically designed to test the potency of compounds under low pH conditions . At these culturing conditions ∼80–90% of the L . major parasites exhibited an aflagellated rounded morphology and displayed similar characteristics of previously described axenic amastigotes including , but not limited to doubling time ( i . e . , ∼24 h ) , clustered growth patterns , agglutination response to PNA lectin , protease activity and protein expression profiles [46]–[49] . Characterization of this parasite population also includes genotyping studies to confirm identity . All axenic amastigote-like parasite cultures were maintained in vented T25 or T75 flasks . For drug susceptibility assays , axenic amastigote-like parasites were harvested in exponential growth phase . The library of pharmacologically active compounds ( LOPAC ) ( 1 , 280 compounds ) was purchased from Sigma-Aldrich . The DP validation set ( 159 compounds ) and the University of Pittsburgh Chemical Methodology and Library Development Center ( UP-CMLD ) diversity set ( 960 compounds ) were obtained from the UP-CMLD ( http://ccc . chem . pitt . edu/UPCMLD/index . html ) . We assayed the 196 , 146 compound library from the Pittsburgh Molecular Libraries Screening Center ( PMLSC ) for L . major growth inhibitors . Cherry-picked compounds from the PMLSC library were supplied by BiofocusDPI ( San Francisco , CA ) . In primary screening , 2 µL of a 1 mM test compound solution in 100% DMSO were diluted in 22 µL complete L . major promastigote growth medium , generating an 83 . 3 µM working concentration ( in 8 . 3% DMSO ) of library compounds . The final test compound concentration was 10 µM with a constant DMSO concentration of 1% in each assay well . The L . major promastigote drug susceptibility assay was performed in a final volume of 25 µL using our previously described 384-well microtiter plate format [38] , [39] . For automated HTS procedures , L . major promastigotes ( 5 , 000 parasites/22 µL ) in complete growth medium were seeded into each well of the microtiter plates using a MAPC2 bulk dispenser ( Titertek , Huntsville , AL ) . Test and control compounds ( 3 µL ) were added to individual wells using a Velocity 11 V-prep ( Menlo Park , CA ) liquid handling system , equipped with a 384-well dispensing head , followed by centrifugation at 50 g for 1 min . Negative ( vehicle ) controls contained 1% DMSO , positive controls contained 10% DMSO and EC50 controls contained 500 nM tamoxifen ( final well concentrations ) . Assay plates were allowed to incubate for 44 h at 28°C in the presence of 5% CO2 . Five µL of alamar blue reagent were added to each assay plate well and incubated for 4 h at 28°C with 5% CO2 . Data were captured on a Molecular Devices SpectraMax M5 ( excitation560; emission590 ) . Individual assay plate Z-factors were derived from the vehicle and positive controls , and data from plates were used only if Z-factors were >0 . 5 [50] . Primary hits were defined as compounds displaying ≥50% inhibition of signal readout . The L . major axenic amastigote-like assay was performed using the alamar-blue assay format and detection methods as the promastigote except that assay plates ( 7 , 500 parasites/well ) were incubated for 144 h at 32°C in the presence of 5% CO2 . In initial 10-point EC50 determination experiments , two µL of 1 mM test compound in 100% DMSO were diluted with 46 µL complete L . major promastigote growth medium creating a 41 . 7 µM working concentration of library compounds . A two-fold serial dilution was then performed creating a concentration range ( 0 . 08–41 . 7 µM ) . The assays were performed in duplicate with a final 10-point concentration range spanning 0 . 01–5 . 00 µM . A compound was designated a confirmed inhibitor only if the EC50 values of both replicates were ≤5 µM . L . major promastigotes were harvested in exponential growth phase and adjusted to a concentration of 2 . 1×105 parasites per mL in complete growth medium . Fifteen thousand parasites ( 75 µL volume ) were then seeded into each well of a 96 well microtiter plate and were treated with a concentration range ( 0 . 1–50 µM ) of test and control compounds . Parasite assay plates were incubated for 48 h at 28°C . Samples were prepared by transferring five µL of parasite suspension to 100 µL of ViaCount reagent ( Guava Technologies , Hayward , CA ) followed by gentle and thorough mixing to ensure an even distribution of parasites . Data were captured on a Guava EasyCyte Plus flow cytometer and analyzed using CytoSoft 5 . 0 . 2 software ( Guava Technologies ) and GraphPad Prism 5 . 0 software ( San Diego , CA ) . A total of 500–1 , 500 parasites were evaluated in duplicate per compound treatment . Mammalian cells were cultured and maintained according to ATCC specifications ( ATCC , Manassas , VA ) . Cell line drug susceptibility assays were performed in final volumes of 25 µL using our previously described 384-well microtiter plate format [38] , [39] . Briefly , for automated HTS procedures , cells ( A549 , IMR-90 and , HeLa , 1 , 000 cells; PC-3 , 750 cells and; MDA-MB-231 , 3 , 000 cells ) in complete culture medium were seeded into each well of 384-well microtiter plates using a Titertek MAPC-2 bulk dispenser . Test and control compounds were added to individual wells as described above . Vehicle and positive controls were 1% DMSO and 10% DMSO , respectively ( final well concentrations ) . Assay plates were incubated for 44–46 h at 37°C in the presence of 5% CO2 and growth inhibitory effects were determined as described above . Five µL of alamar blue reagent was added to each well and incubated for 2–4 h . Data were captured as described above . Primary HTS data analysis and subsequent compound EC50 calculations were performed using ActivityBase ( IDBS , Guilford , UK ) and Cytominer ( University of Pittsburgh Drug Discovery Institute , Pittsburgh , PA ) . To maximize the diversity of leishmanicidals , we performed the primary HTS assay at low stringency with 10 µM of each compound , which ensured a high rate of positive compound identification . Jarvis-Patrick clustering methodology ( Leadscope , Columbus , OH ) was used to computationally filter the number of compounds that proceeded through secondary hit confirmation assays [51] . This deterministic and non-iterative methodology generated non-overlapping , non-hierarchical clusters based on chemical structural similarities . The algorithm selected the number of clusters , with each cluster consisting of at least one structure , and generated non-overlapping , non-hierarchical clusters . A compound with the smallest maximum pairwise distance to the other cluster members was selected as the representative for the structural cluster . In clusters with only two compounds , either compound was selected to represent its specific cluster . This methodology enabled us to reduce the number of potential inhibitors to be evaluated from ∼20 , 000 to ∼1 , 200 ( 0 . 61% hit rate ) while maximizing the chemical diversity of the primary hit pool . Additional data visualization and statistical analysis were performed using Graphpad Prism software 5 . 0 and Spotfire ( Somerville , MA ) . The PubChem database ( http://PubChem . ncbi . nim . nih . gov ) was mined to determine if the confirmed L . major growth inhibitors exhibited bioactivity in other assays . In some instances , select compounds were tested in approximately 300 additional assays , including various molecular target based , phenotypic and cytotoxicity assays . The structural similarity of the confirmed inhibitors was determined using Leadscope software ( i . e . Tanimoto score ) . Confirmed L . major growth inhibitors were filtered further for desirable drug-like properties using ADME Boxes v4 . 0 software ( Pharma Algorithms , Toronto , Canada ) [52] , [53] . In brief , this algorithm predicted human adsorption and metabolism bioavailability for new compounds using a combination of two methods: probabilistic and mechanistic . A bioavailable compound was defined as one that should satisfy the following criteria: dissolve in the stomach or intestine under variable pH , withstand acid hydrolysis at pH<2 , permeate through intestinal membrane by passive or active transport , withstand P-glycoprotein efflux in concert with metabolic enzymes in intestine , and withstand first-pass metabolism in liver . Based on predictions , oral bioavailability was classified as follows: poor<30%; moderate 30–70%; and good >70% . The ADME Boxes software also was used to predict toxicity ( i . e . AMES , hERG , skin irritation , LD50 in mice and Cyp450 inhibition ) of compounds . For genotoxicity , we calculated the probability that a compound would register as a positive in an Ames mutagenicity screening test while hERG in silico assessment was calculated as the probability of a compound being a hERG channel inhibitor at clinically relevant concentrations . Acute toxicity was estimated as the LD50 value ( mg/kg ) after intraperitoneal , oral , intravenous or subcutaneous administration to mice . Skin irritation in silico predictions reflected measurements usually performed in a rabbit Draize test , which primarily measures the toxicity of a compound intended for topical application , cosmetic use or possibly coming into contact with human skin at a standard dose ( 100 or 500 mg ) . Toxicity predictions have an associated Reliability Index ( RI ) as defined as follows: RI<0 . 3; not reliable , RI = 0 . 3–0 . 5 , borderline reliability; RI = 0 . 5–0 . 75 , moderate reliability and RI≥0 . 75 , high reliability [53] . Adult female Balb/c mice ( 6 to 10 week old ) were obtained from ( Charles River Laboratories , Wilmington , MA ) and maintained as outlined by the National Institutes of Health Guide for the Care and Use of Laboratory Animals . All in vivo studies were carried out in accordance with protocols approved by the Institutional Animal Care and Use Committee ( IACUC ) at the University of Miami ( IACUC number C01-08 ) . Food and water were supplied ad libitum . Mice were anesthetized prior to subcutaneous inoculation with 106 stationary phase L . major parasites in 50 µL of Dulbecco's modified Eagle medium in the left hind footpad . Animals were examined daily to determine lesion development . Mice were treated with experimental compounds at a concentration of 40 or 160 mg/kg in a 200-µL total volume/mouse . Control mice were injected with an equivalent amount of vehicle control or amphotericin B ( 12 . 5 mg/kg ) . Footpad lesion size was measured using a Vernier caliper at 7 , 14 , and 21 days post-compound administration . Mice were euthanized in a CO2 chamber at day 21 .
The growth characteristics of the L . major promastigotes in a 384-well plate format were first optimized . When promastigotes were seeded at 105/mL on day 0 , the parasite exhibited conventional exponential , stationary and declining phases over seven days , as anticipated from previous reports with other plate formats [54] ( Figure S1 ) . All subsequent assay development and screening studies were performed with exponentially growing L . major promastigote cultures ( ∼2–3×107 promastigotes/mL ) . Promastigotes readily tolerated up to 1% DMSO with no degradation of growth rate , and the optimal incubation time for alamar blue was 4 h . In the 384-well format , the EC50 for amphotericin B was 207±11 nM , consistent with previously published results with L . major promastigotes in a different assay plate format [41] , [42] . Similarly , EC50 values from other known leishmanicidals including paromomycin ( 19 . 7±0 . 6 µM ) , pentamidine ( 0 . 36±0 . 02 µM ) and sodium stibogluconate ( >100 µM ) compared favorably to previously published reports with other Leishmania species [46] , [55] . An automated , three-day variability assessment with the L . major promastigote drug susceptibility assay format produced Z-factors of >0 . 5 and >10-fold signal window . The L . major promastigote drug susceptibility assay was validated for automated HTS implementation by screening the 1 , 280 compound LOPAC set . Each compound was tested in duplicate at a single concentration ( 10 µM ) and the reproducibility between the duplicate screens is represented in Figure 1 ( R2 = 0 . 94 ) . Average Z-factors were 0 . 71±0 . 03 for the two LOPAC assays , demonstrating the robustness of the developed HTS assay format . Significantly , several compounds with known in vitro and/or in vivo leishmanicidal activity were identified as primary hits , including tamoxifen , pentamidine isethionate , ketoconazole , ivermectin , niclosamide , clotrimazole , and quinacrine [1] , [28] , [29] , [56]–[60] . We also found the leishmanicidal compounds berberine and mycophenolic acid as primary hits when we screened the UP-CMLD DP validation set [61] , [62] . These data confirmed that our optimized L . major promastigote drug susceptibility HTS assay format could be used to identify compounds exhibiting in vitro as well as in vivo leishmanicidal activity . The percentage of compounds in these two validation assays that were identified as growth inhibitory was relatively high , namely 10 . 5% and 22 . 6% for the diverse LOPAC and the more focused UP-CMLD DP sets , respectively , as would be expected under low stringency conditions . To test whether our screening strategy was associated with increased chemical diversity , we used the L . major promastigote drug susceptibility assay to interrogate the UP-CMLD diversity set , which comprised 960 compounds , at 1 and 10 µM . As anticipated , the total number of compounds identified as potential growth inhibitors at 10 µM was greater than at 1 µM ( 250 versus 46 ) and , importantly , 87% of the compounds identified as actives ( ≥50% inhibition of signal ) at 1 µM were also found at 10 µM . There were more structural clusters identified at 10 µM ( 19 ) than at 1 µM ( 7 ) , confirming enhanced structural diversity with the higher screening concentration . Compounds classified as singletons remained relatively consistent across the high ( 8 ) and low screening concentrations ( 6 ) , although the composition of the singleton category changed with increasing screening concentration . Specifically , only 3 ( of the 6 ) singleton compounds detected at the 1 µM screening concentration were represented in the 8 singletons identified at the 10 µM screening concentration . Thus , we adopted a high throughput , low-stringency , computationally-enhanced , small molecule screening ( HILCES ) strategy to maximize the structural diversity of the identified leishmanicidals . We next screened 196 , 146 compounds at 10 µM in 618 plates . Performing robustly , the assay had an average Z-factor of 0 . 9±0 . 1 and an average signal to background values of 26 . 1±1 . 0 without any assay plate failures ( Figure S2 ) . Primary hits , defined as compounds that caused ≥50% inhibition of the signal readout , represented 17 , 629 compounds ( an 8 . 9% hit rate ) . We next computationally filtered the number of compounds that would progress to secondary confirmation assays using a Jarvis-Patrick clustering methodology . We identified 553 structural clusters ranging from 2–360 members and 640 compounds as unique chemical structures ( i . e . , singletons ) ( Figure 2 ) . One compound with the smallest maximum pairwise distance to all other compounds within a cluster was selected to represent a particular structural cluster . In the 84 structural clusters consisting of two compounds , one compound was selected arbitrarily because the Jarvis-Patrick methodology is based on the similarity between several neighbors . In total , the 640 singletons and 553 representative compounds ( 1 , 193 compounds ) were selected for the L . major promastigote secondary assays . Initially , compounds were reassayed at 10 , 5 , and 1 µM to confirm activity and assess potency quickly . One hundred forty-six compounds exhibited ≥50% inhibition when assayed at 1 µM and , therefore , progressed to secondary confirmation assays . All of these primary screening data have been posted for public access on the PubChem database ( http://PubChem . ncbi . nlm . nih . gov/ ) . The growth inhibitory activity of the 146 compounds was confirmed using 10-point concentration ( 0 . 01–5 . 00 µM ) response assays . In total , 137 compounds had EC50 values of <5 µM for an overall confirmation rate of 93 . 8% . Of the 137 confirmed L . major promastigote growth inhibitors , remarkably , 93 compounds had EC50 values <1 µM . In initial specificity studies , 70 of the submicromolar L . major growth inhibitors failed to inhibit the growth of the sentinel mammalian A549 cell line at 1 µM , suggesting specificity towards the L . major promastigote ( Table S1 ) . Moreover , because these compounds are part of the publicly accessible PubChem database , they have to date been screened in 99 ( lowest ) to 369 ( highest ) additional phenotypic and target-based bioassays ( Table S1 ) . Sixty-six percent of the leishmanicidal compounds registered as confirmed actives in ≤2 PubChem bioassays . None of the leishmanicidal compounds were structurally similar to the clinically used anti-leishmanial compounds sodium stibogluconate and amphotericin B ( Tanimoto score ≤0 . 3 ) , supporting our objective of expanding the pool of potential leishmanicidal chemotypes ( Table S1 ) . Importantly , however , compounds with previously documented in vivo or in vitro leishmanicidal activity were also identified using the HILCES system , including pentamidine isothionate , clotrimazole , aminacrine , aphidicolin , and acivicin , thus further validating our assay system ( Table 1 and Table S1 ) [1] , [42] , [58] , [63] , [64] . Next , we selected a representative group of 15 chemotypes and verified their leishmanicidal activity using compounds from a commercial supplier , thereby controlling for growth inhibitory effects resulting from any potential compound degradation during library storage . These compounds were balanced between compounds with known pharmacological actions ( 7 ) and new chemotypes ( 8 ) ( Figure 3 , Tables 1 and 2 ) . We confirmed the leishmanicidal activity of the 15 chemotypes ( Tables 1 and 2 ) with the majority of the compounds registering as submicromolar growth inhibitors . Significantly , there was a strong correlation between the EC50 values derived using the alamar blue assay with determinations using a flow cytometer-based format providing a second , independent methodology that confirmed the leishmanicidal activity of the test compounds ( Tables 1 and 2 ) . Subsequent testing in a human cell line panel indicated that the majority of the compounds displayed a specific and selective growth inhibitory effect toward the L . major parasite ( Tables 1 and 2 ) . None of the new chemotypes and only two of the compounds with known pharmacological actions , sangivamycin ( PubChem CID 9549170 ) and acivicin ( PubChem CID 2007 ) , inhibited the growth of human cell lines tested ( Table 1 and Table 2 ) . Amphotericin B was used as a reference compound and the results were consistent with previously reported EC50 values ( Table 1 ) [40] , [65] . We next determined the leishmanicidal activity of the 15 test compounds using an L . major axenic amastigote-like alamar blue-based assay . Thirteen compounds exhibited growth inhibitory activity , indicating that these compounds were active at pH 4 . 9 . Significantly , four compounds maintained their submicromolar activity , with three compounds PubChem CID 3117 ( disulfiram ) , 457964 ( aphidicolin ) and 760847 , exhibiting EC50 values comparable to amphotericin B ( Table 1 and Table 2 ) . Several other compounds displayed EC50 values ≤10 µM . The 15 test compounds were further classified for potential in vivo studies with respect to in silico predictive ADMET characteristics ( Table S1 ) . Twelve compounds had predicted bioavailability profiles in the good to moderate range while three compounds were predicted to have poor bioavailability . Overall , the 15 test compounds were not predicted to exhibit significant toxicity; however , two compounds ( CID 786799 and 742546 ) have high probability for skin irritation while one compound ( CID 2812 ) has a moderate probability of inhibiting Cyp3A4 at 10 and 50 µM ( Table S1 ) . To determine if any of the new leishmanicidal chemotypes identified in the L . major promastigote screen had in vivo activity , we prioritized compounds according to the empirically-derived potency and specificity data , known pharmacological activity , activity in the L . major axenic amastigote-like drug susceptibility assay , in silico predicted ADMET , previous human usage , and novelty of the leishmanicidal chemotype . Thus , disulfiram was selected for initial in vivo efficacy studies . L . major-infected Balb/c mice were treated with vehicle , disulfiram ( 40 or 160 mg/kg ) , or amphotercin B ( 12 . 5 mg/kg ) for 21 days . Drug treatment was initiated 3 days post-infection allowing for the establishment of the leishmaniasis infection . Over the course of the 10 day treatment , a decrease in the average footpad thickness was observed as compared with vehicle treated animals . With disulfiram ( 160 mg/kg ) treatment , there was a ∼43% and 50% reduction in footpad thickness observed on days 14 and 21 post-infection , respectively ( Figure 4 ) . There was a similar decrease observed in footpad thickness with 40 mg/kg disulfiram on days 14 ( 25% ) and 21 ( 35% ) , illustrating a dose- and time-dependent efficacy of the disulfiram treatment . As expected , the amphotericin B ( 12 . 5 mg/kg ) control treatment effectively reduced footpad swelling and these data were consistent with additional experiments that showed an average 80–85% reduction in footpad swelling after amphotericin B treatment ( Figure 4 and data not shown ) . Although disulfiram and amphotericin B display similar levels of growth inhibitory activity in the promastigote and axenic amastigote-like assays , there was a difference in their in vivo effects ( i . e . , 50% versus ∼85–90% reduction in footpad swelling ) . This disparity in in vivo effects may be the result of differences in bioavailability or mechanism of action .
In the current study , we illustrate the power of HILCES , a low stringency , ”forward” pharmacology , antileishmanial drug discovery strategy that employs a robust phenotypic HTS assay unencumbered by concerns for specific molecular targets [66] . HTS methodologies enabled the interrogation of a large diverse compound library , and when linked with computational methodologies , permitted refinement of the primary screening data by chemical structural clustering of chemotypes and predicted pharmacological attributes . This HILCES strategy enhanced our ability to identify novel leishmanicidal chemotypes , and , as a result , enabled us to test these new chemotypes for in vivo leishmanicidal activity , thus effectively expanding the pool of chemical structures that could be refined as potential leishmanicidal therapies . By capitalizing on multiple assay formats as well as L . major promastigote and axenic amastigote-like life cycle forms , we were able to confirm and prioritize our L . major growth inhibitory chemotypes for in vivo testing . Significantly , our preliminary studies with disulfiram indicated that our HTS and hit confirmation strategy could lead to the identification of novel leishmanicidal chemotypes with in vivo efficacy . L . major promastigotes have frequently been used to characterize the growth inhibitory activity of potential leishmanicidal agents and they are well suited for the rapid screening of large chemical libraries due to ease of culturing [44] . In fact , two smaller scale screens , ∼2 , 100 compounds ( http://www . sandler . ucsf . edu/lhf ) and ∼15 , 000 compounds [67] have been performed using Leishmania promastigotes . Moreover , there is some evidence that the promastigote form of the parasite is an effective and reliable indicator of a compound's leishmanicidal activity in cell-based and axenic amastigotes except when examining immunomodulating anti-leishmanial compounds , such as sodium stibogluconate and meglumine antimoniate [40] , [55] , [68] , [69] . Nonetheless , we acknowledge that there continues to be some debate about the physiological relevance of the L . major promastigote as an indicator of leishmanicidal activity for the cell-internalized amastigote form of the parasite , primarily because it is not the parasite stage found in humans , and they have a dissimilar response to the pentavalent antimonial compounds [40] , [44] . Even so , we suggest that the promastigote-based screening assay may effectively function as the foundation for a comprehensive screening paradigm that is designed to identify and qualify novel leishmanicidal chemotypes . We recognize , however , the significance of the ability to perform HTS in a cell-based amastigote system ( http://www . dndi . org/newsletters/n18/5_1 . php ) . The use of a publicly available annotated chemical library enabled us to cross-query a range of archived bioassays and to consider potential novel molecular targets . While the majority of the leishmanicidals failed to register as confirmed actives in other assays ( Table S1 ) , suggesting specificity for leishmanicidal activity , we found several compounds that affected previously unappreciated and provocative potential L . major molecular targets . For example , we found protein targets involved with cell proliferation , differentiation , invasion and motility , such as protein kinase D ( gene id 5587 ) , protein kinase C ( gene id 5578 ) , polo-like kinase 1 ( gene id 5347 ) , steroidogenic factor 1 ( gene id 2516 ) and phosphatase regenerating liver-1 ( gene id 7803 ) [70]–[74] . Significantly , these or related proteins are not only expressed in L . major but also in other parasites , including Schistosoma mansoni and Trypanosoma brucei , so they might also be critical for schistosome and trypanosome growth , differentiation , cell cycle regulation , motility and viability [31] , [75]–[77] . Moreover , these data suggest that compound libraries used in conjunction with genome searches may be exploited to identify potential new drug targets . In summary , we identified 70 submicromolar compounds that inhibit promastigote growth by using HILCES with a publicly available annotated library . Significantly , these compounds did not inhibit mammalian cell growth in companion counter-screening assays , suggesting an L . major-specific inhibitory response . All of the primary screening data are accessible on PubChem ( http://PubChem . ncbi . nlm . nih . gov ) and can be conveniently mined worldwide to allow for further refinement of individual compounds . A novel leishmanicidal chemotype , disulfiram , exhibited up to 50% in vivo efficacy in our animal model system . Disulfiram validated our compound screening strategy , it has a number of potential molecular targets and mechanisms . Several of the identified compounds have known molecular targets that may be relevant for this and other Leishmania species . The simple platform developed for L . major may also be useful for efforts designed to identify chemotherapeutics for other Leishmania species . | Leishmaniasis is a parasitic disease with cutaneous , mucocutaneous and visceral clinical manifestations , depending on the Leishmania spp . and human host . Globally , there are 350 million people at risk of leishmaniasis , but current treatment options rely predominantly on ancient pentavalent antimonials , which have the potential to cause serious systemic toxicity . Our research focuses on the rapid expansion of potential anti-leishmanial compounds that could function as novel chemical structures for future drug development and offer additional therapeutic options to patients with leishmaniasis . We combined high throughput screening methodologies with computational algorithms and multiple confirmatory assay formats to identify and characterize new potent L . major promastigote growth inhibitors , including one that displays in vivo activity without toxicity to human cells . Our use of a large , broadly distributed compound library enabled the identification of these new chemotypes . In addition , since this chemical library is publicly available and annotated , we were able to cross-query archived bioassays and to identify new molecular targets that may be involved in L . major growth and viability as well as identify new protein targets for future leishmanicidal drug discovery . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"chemical",
"biology/small",
"molecule",
"chemistry",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"pharmacology/drug",
"development",
"pharmacology",
"infectious",
"diseases/protozoal",
"infections",
"computational",
"biology",
"chemical",
"biology",
"infectious",... | 2009 | Identification of Potent Chemotypes Targeting Leishmania major Using a High-Throughput, Low-Stringency, Computationally Enhanced, Small Molecule Screen |
Guanine ( G ) -rich DNA readily forms four-stranded quadruplexes in vitro , but evidence for their participation in genome regulation is limited . We have identified a quadruplex-binding protein , Lia3 , that controls the boundaries of germline-limited , internal eliminated sequences ( IESs ) of Tetrahymena thermophila . Differentiation of this ciliate’s somatic genome requires excision of thousands of IESs , targeted for removal by small-RNA-directed heterochromatin formation . In cells lacking LIA3 ( ΔLIA3 ) , the excision of IESs bounded by specific G-rich polypurine tracts was impaired and imprecise , whereas the removal of IESs without such controlling sequences was unaffected . We found that oligonucleotides containing these polypurine tracts formed parallel G-quadruplex structures that are specifically bound by Lia3 . The discovery that Lia3 binds G-quadruplex DNA and controls the accuracy of DNA elimination at loci with specific G-tracts uncovers an unrecognized potential of quadruplex structures to regulate chromosome organization .
Ciliates maintain distinct germline and somatic genomes that are partitioned into different nuclei , called micro- and macronuclei , respectively [1] . At each sexual round of the ciliate life cycle , the somatic genome is destroyed , and new germline and somatic genomes are created from identical copies of a zygotic genome formed after exchange of germline nuclei between conjugating partners . The subsequent differentiation of the somatic genome involves massive genome reorganization , which includes fragmentation of the chromosomes and elimination of a large fraction of the germline-derived sequence . In the ciliate Tetrahymena thermophila , more than 6 , 000 dispersed loci , comprising nearly one-third of the genome , are eliminated [2] . These internal eliminated sequences ( IESs ) consist of both unique and repetitive sequences that are most likely evolutionarily derived from the movement of transposable elements . DNA elimination serves as an effective genome surveillance mechanism that silences these potentially deleterious sequences by removing them from the transcribed nucleus [reviewed in 3] . The eliminated sequences are targeted for excision by small-RNA-directed heterochromatin formation . The targeting small RNAs ( called scan RNAs ) are produced during meiosis of the micronucleus and then assembled into effector complexes containing the argonaute/Piwi-related protein , Twi1 [4–6] . This mechanism is the evolutionary equivalent of the piRNA pathway , which employs small RNAs to silence transposons in the germline of multicellular organisms [see 7 , 8] . In Tetrahymena , the scan RNA-Twi1 complexes enter developing macronuclei during post-zygotic development and direct histone H3 lysine ( K ) 9 and K27 tri-methylation ( me3 ) to homologous regions [9 , 10] . The modified chromatin is recognized first by chromodomain proteins Pdd1 and Pdd3 [11 , 12] and then by additional proteins [13] . Finally , the domesticated piggyBac transposase , Tpb2 , excises the IESs [14] . The widespread distribution of IESs throughout germline chromosomes , together with the high gene density of the somatic genome ( the average intergenic region is 1 kbp ) [15] , necessitates accurate removal of the IESs to prevent loss of important coding or regulatory sequences . Previous work has revealed that cis-acting sequences located in the DNA flanking each IES specify excision boundaries [16–19] . Even so , the functionally equivalent controlling sequences of different characterized IESs share no obvious sequence similarity . The best studied of these cis-acting sequences is a polypurine tract ( 5’ AAAAAGGGGG 3’ or A5G5 ) located 45–50 bp outside each excision boundary of the extensively studied M IES [16] . These sequences on each side of the eliminated region reside in opposite orientation such that the G5 portion is proximal to the IES . This A5G5 tract is both necessary and sufficient to direct accurate excision [16 , 20] . However , the actual mechanism by which this critical sequence defines the M IES boundaries is unknown . Here , we show that deletion of the novel gene LIA3 abolishes accurate excision of both the M IES and other IESs flanked by A5G5 tracts . Furthermore , we show that the Lia3 protein binds the M IES A5G5 boundary determinant when it adopts a non-canonical Guanine quadruplex ( G4 DNA ) structure . G4 DNA forms when Hoogsten base pairs stabilize interactions between four strands each composed of runs of three or more Gs [21] . G4 DNA may form during DNA replication , transcription , or other circumstances that free DNA strands from the double helix; however , in vivo evidence for formation of G4 DNA and its regulatory functions is limited . Studies have indicated that cells need to effectively manage sequences that have the potential to form G4 DNA to ensure genetic and epigenetic stability [22 , 23] . Furthermore , a G4-DNA-forming sequence was found to be critical for antigenic variation in Neisseria gonorrhoeae , illustrating that DNA elements that form non-canonical structures are indeed functional [24] . Early evidence that G4 DNA can form in eukaryotic cells came from studies of the telomeres of the multicopy nanochromosomes of Stylonychia lemnae in which telomeric G4 DNA and telomere binding proteins were shown to mediate attachment to the nuclear envelope [25 , 26] . The abundance of Stylonychia telomeres permitted ready detection of G4 DNA with the aid of structure-specific antisera . By using a similar approach , G4 DNA was more recently detected in vivo in multiple eukaryotic species including in mammalian cells [27–29] . Identification of proteins that bind and/or unwind G4 DNA has provided further evidence that these structures likely serve functional roles in vivo [21 , 30 , 31] . The in vitro binding and in vivo genetic data presented here identify a new role for G quadruplexes , in the control of genome-wide DNA elimination , and demonstrate clearly that such non-canonical DNA structures function in genetic regulation .
In our search for proteins that are important for the differentiation of the somatic genome , we identified candidates , including Lia3 , that are expressed specifically during conjugation and localize to developing macronuclei [13] . Lia3 is a novel protein , which only has obvious similarity with three other Tetrahymena proteins of unknown function . To determine whether Lia3 has a critical role in macronuclear development , we created LIA3 knockout ( ΔLIA3 ) strains lacking all germline and somatic copies of LIA3 . We confirmed the replacement of the LIA3 coding region with the neo3 paromomycin-resistance cassette through genetic crosses and Southern blot analysis ( S1 Fig ) , and loss of LIA3 expression by using rtPCR ( Fig 1A ) . When we mated two LIA3 knockout lines together , we found that they completed all stages of development , reaching the wild-type ( wt ) end-point of conjugation , having resorbed one of the two micronuclei ( Fig 1B ) ; however , when mated ΔLIA3 cells were returned to growth media , only 15% of mated pairs produced viable progeny , whereas 70% of wt pairs did so ( Fig 1C ) . These results indicated that LIA3 participates in , but is not essential for , development . During macronuclear development , the germ-line derived genome is extensively reorganized and nearly one-third of the DNA is eliminated . To assess whether DNA elimination occurred efficiently in ΔLIA3 conjugants , we monitored the excision of a well-characterized locus containing two eliminated sequences , the M and R IESs . The M IES exhibits alternative excision , eliminating either 0 . 6kbp ( Δ0 . 6 ) or 0 . 9kbp ( Δ0 . 9 ) ( Fig 2A ) . By using PCR primers outside the IES , we could detect both rearranged and unrearranged loci ( Fig 2B ) . As all parent lines used in this study possessed only the Δ0 . 9 form in their macronuclei , detection of the Δ0 . 6 form during conjugation revealed if and when new excision had occurred in differentiating nuclei . Upon mating wt cells , M IES excision began by 12 hrs of conjugation , evident by a doublet of ~600 bp bands ( Fig 2B ) ; In contrast , M IES excision in ΔLIA3 mating cells was both delayed and aberrant , as newly excised forms were not observed until 16hrs after initiation of mating , and when observed , a ladder of PCR products was visible instead of the doublet ( Fig 2B ) . We did not observe similar aberrancy in R IES elimination due to loss of Lia3 . R IES excision may be delayed in ΔLIA3 matings , as the DNA fragment representing the unrearranged form was more abundant between 10 and 18 hrs than in wt , but this could not be unambiguously determined because de novo rearrangement of this IES cannot be distinguished from the DNA present in the parental macronuclei ( Fig 2D and 2E ) . Nevertheless , no aberrant excision was evident , suggesting that the loss of LIA3 affects the accuracy of excision of only one of these two IESs . We initially observed aberrant M IES excision in ΔLIA3 mating populations for which only a portion of cells survived . To determine whether the defective excision detected occurred primarily in the fraction of the population that died , we also examined M and R IES excision in individual surviving progeny cells . The nine individual progeny lines from ΔLIA3 crosses examined possessed an array of M IES excision products , which reflects the aberrancy observed within the full mating population ( Fig 2C and 2B ) . Excision of the R IES again appeared to be largely unaffected ( Fig 2F ) . Thus , aberrant excision was not limited to the ΔLIA3 progeny that died as cells with heterogeneous excision boundaries survived conjugation . To determine how IES boundaries are positioned in the absence of Lia3 , we cloned and sequenced a number of the M IES junctions of wild-type and ΔLIA3 progeny . The boundaries of eliminated DNA can be positioned hundreds of base pairs upstream or downstream of the major wild-type boundaries ( Fig 2G ) . This aberrant elimination could occur due to improper cleavage of the genome by Tpb2 or , alternatively , cleavage could be normal , but the rejoining that must occur subsequent to cleavage could be perturbed . We took advantage of the observation that excised IESs will circularize to map presumed sites of cleavage in both wt and mutant cells [32 , 33] . The junctions of these circular , excised IESs were recovered by using PCR primers complementary to the excised regions to amplify outward across the joined ends , thus allowing us to map excision boundaries ( see Fig 2H ) . In wt matings , circular products were observed starting either at 10 or 12 hrs into conjugation . In ΔLIA3 matings , R IES circular products of the predicted size were detected at the same hour into conjugation as they were found in wt matings , but M IES circular products appeared much later and were variable in size ( Fig 2I and 2J and S1 Table ) . The IES boundaries of these circular products showed similar map positions as the boundaries of the rejoined DNA in progeny ( Fig 2B and S1 Table ) . This observation is consistent with aberrancy in the cleavage of the M IES from the genome . To determine why loss of LIA3 affects M IES but not R IES excision , we tested whether the perturbation in ΔLIA3 mutants was due to impaired recognition of the M IES or specification of accurate boundaries . Recognition of IESs occurs when complementary scan RNAs match regions of the developing somatic genome and mark them for elimination [4 , 34] , whereas boundaries are determined by sequences flanking each IES , which are retained after excision [16–19] . Two lines of evidence indicate that the M and R IESs differ from one another in both their recognition requirements and boundary determinants: 1 ) Ema1 , an RNA helicase that participates in scan RNA/Twi1 recognition , is required for M but not R elimination [35]; and 2 ) M and R IESs have functionally distinct and incompatible boundary-controlling sequences [16 , 17 , 20] . To determine whether Lia3 acts to identify the eliminated region of the M IES or its flanking boundary sequences , we generated chimeric IESs and tested their excision in both wt and ΔLIA3 conjugants . These chimeras contained the eliminated region of either the M , R , or a segment of the transposon-like TLR IES [36 , 37] inserted between either the M or R boundary-controlling flanking sequences . These chimeras were introduced into conjugating cells during nuclear differentiation on rDNA-based replicating vectors , and IES excision was monitored by Southern blot analysis of the transformant DNA . When any of the eliminated sequences , including that of the M IES , was positioned between the R IES’s flanking sequences , the chimera was accurately excised using the normal R IES boundaries , even in ΔLIA3 matings ( Fig 3A–3D ) . In contrast , when any of these IESs was positioned between the M IES’s flanking DNA , each IES was correctly and efficiently excised in wt matings , but not in ΔLIA3 matings ( Fig 3E–3H ) . We observed a significant decrease in excision efficiency for the M-flanked M and R IESs ( Fig 3F and 3G ) , whereas the M-flanked TLR IES was efficiently deleted , but its excision lacked clearly defined boundaries , evident as a ladder of products ( Fig 3H ) . The decrease in excision efficiency that we observed in ΔLIA3 progeny coincides with the decreased M IES excision observed upon deleting or otherwise mutating polypurine tracts flanking the M IES [16 , 20] . These experiments demonstrate that Lia3 acts in concert with the M IES flanking DNA to specify the boundaries of excision , but does not discriminate between the different IESs placed between these controlling sequences . The boundary-controlling flanking sequences of the M IES consist of polypurine tracts , 5’-A5G5-3’ , located approximately 45bp away from the major boundaries [16] . This A5G5 sequence is not present in the flanking region of the R IES , leading to our hypothesis that Lia3 interacts with this polypurine tract to determine the position of each excision boundary . If true , Lia3 represents the first protein known to position these boundaries . We first tested this possibility by identifying other IESs with similarly positioned polypurine tracts and assessed whether their excision was aberrant in ΔLIA3 progeny . All four additional IESs with polypurine tracts located near their boundaries exhibited aberrant excision in progeny of ΔLIA3 matings ( Figs 4A , 4B and 4C and S2A ) whereas the several other IESs tested that lacked obvious polypurine tracts were not affected by loss of LIA3 ( Figs 4D , 4E and 4F and S2B–S2F ) . These findings are consistent with our analysis of chimeric IESs ( Fig 3 ) that showed that diverse IESs are affected by loss of LIA3 only when flanked by G-rich polypurine tracts . Thus Lia3 appears to specifically control the excision boundaries of a class of IESs containing flanking polypurine tracts . The obvious interpretation of our data is that the novel protein Lia3 directly binds to the M IES polypurine tracts and controls the extent of excision . To test the ability of Lia3 to bind DNA , we purified the protein after expression in E . coli ( S3 Fig ) and used it in electrophoretic mobility shift assays ( EMSA ) . Initially , we incubated Lia3 with single stranded ( ss ) or annealed ( ds ) 30 nt oligonucleotides corresponding to either the leftmost M IES boundary ( M1 ) , centered on the A5G5 tract , or the equivalent region from the leftmost R IES boundary ( R1 ) ( Table 1 ) . Lia3 bound strongly to ssM1 and weakly or not at all to the dsM1 , ssR1 , and dsR1 . To confirm specificity , we used unlabeled oligonucleotides to attempt to compete away weaker interactions . The ssM1 oligonucleotide competed effectively for the initial binding observed when using the ssR1 and dsR1 substrates , whereas the ssR1 and dsR1 oligonucleotides could not compete for the binding to ssM1 , indicating that the interaction that Lia3 had the highest affinity for the ssM1 probe ( Fig 5A ) . In these assays , the majority of the unbound ssM1 oligo exhibited an altered electrophoretic mobility , migrating significantly slower than expected , and it was this form of the probe to which Lia3 preferentially bound ( Fig 5A , black arrow ) . This purine-rich oligonucleotide contains a run of Gs , leading us to test the possibility that the probe had adopted G4 DNA structure ( Fig 5B ) . G4 DNA is known to form readily in the presence of KCl , but poorly in the presence of LiCl or without cations [38] , so we denatured the oligonucleotide by boiling in either 10 mM Tris-HCl ( ph 7 . 5 ) alone or supplemented with either 100 mM KCl or 100 mM LiCl , then slow cooled to room temperature before native gel electrophoresis . The M1 oligonucleotide migrated as expected for ssDNA in buffer without salt or with LiCl , but abnormally slow in the presence of KCl ( S4A Fig ) . Mutation of either the first three Gs within the A5G5 segment to Cs , mutations known to abolish boundary function [20] , or even simply changing the second G to C , was sufficient to prevent the M1 oligo from forming a higher-ordered structure ( S4 Fig ) . All these observations are consistent with a G4 DNA structure . We confirmed that 30 nt , A5G5-containing oligonucleotides representing either the M1 and M2 flanking region ( the sequences from the two alternative left side M IES boundaries centered around the A5G5 , [16] ) formed quadruplex structures by performing circular dichroism ( CD ) [39] ( Figs 5C and S4 ) . Parallel G4 DNA exhibits a diagnostic positive peak at 260nm and a negative peak at 240nm [39] . Both the M1 and M2 oligonucleotides displayed CD spectra diagnostic with formation of parallel G4 DNA when in the presence of KCl , but not , at least for M1 , when in the presence LiCl ( S4 Fig ) . This observation further supports our conclusion that these probes formed a quadruplex in the conditions used in our EMSA . To rule out the possibility that a co-purifying contaminant in the extract was responsible for quadruplex binding , we performed parallel purification of Lia3 and the MS2-coat protein and used each in EMSA . Our initial binding assays were performed with a histidine-tagged Lia3 protein , which required denaturing lysis to recover from E . coli . We subsequently expressed Lia3 with a maltose binding protein ( MBP ) fused to its amino terminus , which allows purification in non-denaturing conditions , and isolated MBP-Lia3 along with MBP-MS2 ( S5 Fig ) . The MBP-Lia3 specifically bound the G4 DNA M1 probe formed in the presence of KCl , but not the ssMI probe ( in LiCl ) ( Fig 6 ) . The specificity of Lia3 for G4 DNA is further revealed by the observation that none of the residual ssM1 DNA ( lower band ) remaining in KCl treated probe samples was shifted upon addition of protein ( Figs 5 and 6 ) . The MBP-MS2 protein bound neither the probe in LiCl or KCl . Excess unlabeled M1 oligonucleotide , but not the C3G2 mutant oligonucleotide , could compete away MBP-Lia3 binding , which further supports that Lia3 preferentially binds G4 DNA . To further assess the specificity of Lia3 for parallel G4 DNA , we measured binding affinity of Lia3 to M1 G4 DNA , ssM1 , dsM1 , or Tetrahymena telomere sequence , which is known to form mixed quadruplex structures ( Figs 6B and S6 ) . We determined that the Kd of Lia3 for the M1 quadruplex was 144 nM . It also bound to the telomere quadruplex , but with lower affinity than to the M1 quadruplex ( Kd = 11 . 5 μM ) . Lia3 had much higher affinity for either quadruplex probe than for the ss or ds linear forms of the M1 probe ( extrapolated Kd over 0 . 2 mM ) . In competition experiments , oligonucleotides forming parallel G4 DNA ( M1 or M2 ) were able to compete away the interaction of Lia3 with the M1 quadruplex , whereas linear oligonucleotides did not compete for binding , which further shows that Lia3 binds specifically to parallel G4 DNA in vitro ( S7 Fig ) . It is important to note that addition of LiCl to the binding reaction did not inhibit Lia3 binding to the M1 probe when it was pre-assembled into the quadruplex form ( S7 Fig–ss competitor ) . Together , our genetic and biochemical analyses indicate that Lia3 binds to a parallel G quadruplex that forms near the boundaries of the IESs flanked by A5G5 sequences to direct accurate excision . We attempted to directly detect these structures in developing macronuclei using available anti-G4 DNA antibodies without success . We could detect putative G4 DNA in the macronuclei of unmated cells and the parental macronuclei of late stage conjugants , possibly due to the very abundant telomeres ( S8 Fig ) . The failure of this approach may indicate that Lia3 binding masks the G4 DNA epitope in developing macronuclei or that the amount of these structures present is below the level needed for detection with these reagents . Although it is easy to envision how intermolecular association of four oligonucleotides can allow formation of G4 DNA in our gel shift assays , how a four-stranded structure might form at chromosomal loci given that each side of the IES contains a single run of Gs is less obvious . We investigated one possibility , that two of the four strands could be RNA . Hybrid DNA/RNA quadruplexes can form during transcription [40] , and transcription of IESs occurs before their excision [35 , 41] . Transcription would unwind the flanking G tracts , freeing them to interact with other G-rich strands . In this model , the non-coding transcripts created provide the two additional strands needed to complete this structure . To test whether RNA is available to participate in defining M IES boundaries , we used rtPCR to look for transcripts at the time that the Lia3 protein accumulates ( S9 Fig ) and detected RNAs that span the A5G5 tract ( Fig 7A ) . We also found that RNA oligonucleotides with the M1 flanking region sequence can form quadruplexes , and that these RNA quadruplexes can compete for Lia3 binding to the M1 G4 DNA probe ( S10 Fig ) . Although these findings support the possibility that non-coding transcripts participate in controlling the boundaries of eliminated sequences , they certainly do not exclude other mechanisms discussed below .
The polypurine tracts flanking the M IES were first shown to control its excision boundaries 25 years ago [16] . Despite the identification of similarly positioned controlling sequences flanking other IESs , how these diverse cis-acting sequences are recognized has remained a mystery . We show here that Lia3 is required to accurately excise the M IES and other IESs possessing flanking polypurine tracks . Lia3 is a novel protein , expressed exclusively during post-zygotic development . In our efforts to characterize its binding to DNA , we discovered that it binds specifically to parallel G4 DNA formed by the M IES A5G5 sequence . As both this sequence and Lia3 determine IES boundaries , our data strongly support our hypothesis that G4 DNA can form at internal chromosomal loci ( not just telomeres ) and define specific regulatory domains . Although we were not surprised that the A5G5-containing oligonucleotides we used as EMSA probes could form G4 DNA , we did not expect that Lia3 would preferentially bind this structure . Each side of the IES has a single G5 tract , and formation of a quadruplex would require four independent copies to come together . The simplest way we can envision this forming in vivo involves the transcription from the C5 strand , providing G5-containing RNA copies that participate in quadruplex formation such that the quadruplex includes the G5 DNA tracts on each side of the IES and the two RNA strands ( Fig 7B ) . Four strands of the flanking regulatory A5G5 DNA are also produced during a round of DNA replication that precedes DNA elimination ( Doerder and Debault 1975 ) . If RNAs are not part of the quadruplex , it is likely that the G tracts on each side of the IES from both sister chromatids form the G4 DNA structure . It is also possible that the G tracts of different A5G5-controlled IESs interact to form a single quadruplex . By showing that Lia3 is both a parallel G4 DNA binding protein and a specific regulator of the excision of IESs containing A5G5 tracts , we report a compelling case for a role for non-canonical DNA structures in regulating genome organization . The G4 DNA bound by Lia3 appears to bring together G tracts located on each side of the IESs . The formation of G4 DNA through the association of distal G tracts located on different DNA strands is not the obvious outcome , and therefore our findings elucidate an unforeseen potential of dispersed , G-rich DNA sequences to interact . The proposed involvement of transcription in this structure serves two purposes: to unwind the DNA to allow the G tracts to interact with distal partners , and to provide additional G-rich strands to promote a four-stranded structure to form . If such a predicted structure forms in vivo , Lia3’s ability to bind these structures may permit this protein to serve as a probe for such structures in genomes beyond Tetrahymena . Long non-coding RNAs and non-genic transcription appear to be prevalent in genomes . The model we present in Fig 7 suggests a novel mechanism for these RNAs to interact with DNA and affect chromosomal DNA organization . We believe the ability of Lia3 to bind novel quadruplex structures represents another case in which studies of ciliate genome rearrangements have uncovered new regulatory potential in eukaryotes . To assist the quadruplex formation between distal G tracts , we propose that the formation of heterochromatin ( i . e . , establishment of H3K9me3 and H3K27me3 ) across the IES and subsequent binding of chromodomain-containing proteins Pdd1 and Pdd3 , and other DNA excision proteins , position the IES flanking regions in proximity to one other . This organization of IES chromatin aids the formation of the quadruplex , which is stabilized by Lia3 binding . This proposal is consistent with data showing that G4 DNA is enriched in the heterochromatic regions in Drosophila polytene chromosomes [28] . The interaction of distal G tracts on different strands represents a novel mechanism to partition chromosomal loci into distinct domains . Once bound , Lia3 guides the domesticated transposase Tpb2 to preferential boundary sites either by directly interacting with the transposase or simply preventing it from cutting elsewhere . Multiple results from mutational analyses provide evidence that cis-acting sequences on each side of an IES interact with one another . For instance , deletion or other disruption of the boundary-controlling sequence on one side of an IES did not lead simply to inaccurate specification of the boundary on the mutated side of the IES , but instead severely decreased overall rearrangement efficiency [16 , 17] . Furthermore , chimeric IESs containing one M and one R IES flanking sequence did not exhibit excision at the native M and R boundaries present in the construct , but instead used the native M boundary on one side and a novel boundary that is 45–50 bp away from a cryptic A5G5 tract present by chance within the R IES sequence [16] . These data are consistent with our model in which G5 tracts on each side of the IES come together to form parts of a common structure . Coupled cleavage on both sides of an IES may have been selected for during the domestication of the Tpb2 piggyBac transposase to ensure accurate excision and prevent aberrant double-strand breaks during genome-wide DNA elimination events . Coordinated cleavage on both sides of an IES occurs in the ciliate Paramecium [42 , 43] , which also uses a domesticated piggyBac to perform its genome rearrangements [44] , indicating that communication between IES ends is a conserved mechanism . Although we favor a model in which IES heterochromatin is established , and subsequent organization of this chromatin structure helps to bring distal A5G5 sequences together to form a G quadruplex , we cannot rule out the possibility that Lia3 stabilizes this structure prior to the completion of these chromatin modifications and acts to limit the spread of small-RNA-directed heterochromatin . In the future , we will determine the enrichment of H3K9me3 and H3K27me3 across the developing genome in the presence and absence of Lia3 to assess whether the cis-acting sequences that control IES boundaries actually serve as barrier elements blocking the spreading of chromatin modifications . Only a subset of IESs have flanking A5G5 tracts and are controlled by Lia3; yet there are thousands of IESs , which vary greatly in size and sequence , that are faithfully excised during differentiation of the somatic genome . The adjacent M and R IESs are known to use functionally distinct boundary-controlling sequences [16 , 17] . The use of distinct cis-acting sequences by neighboring IESs would prevent aberrant elimination events between the distal ends of adjacent IESs . Some likely candidates to define the ends of the non-A5G5 IESs are three Tetrahymena proteins with homology to Lia3 . Like LIA3 , each is expressed exclusively during post-zygotic development [45] , and the two of which that we have examined localize to developing macronuclei ( S11 Fig ) . Although these Lia3-like ( LTL ) proteins do not have obvious homologs in other organisms , database annotation indicates that their amino termini possess similarity to DNA binding proteins [46] . In our preliminary investigations , disruption of LTL1 ( Ttherm_00499370 ) results in aberrant excision of several non-Lia3 regulated IESs . It will be interesting to determine whether these related proteins control the boundaries of other IESs by binding to other non-canonical DNA structures . Their study could provide evidence for novel mechanisms used to bring together cis-acting sequences to define specific regulatory domains within genomes .
Tetrahymena cells were grown at 30°C in either SPP or Neff’s medium under standard conditions [47 , 48] . Strains CU428 ( Mpr1-1/Mpr1-1 [VII , mp-s] ) , B2086 ( II ) , and CU427 ( Chx1-1/Chx1-1 [VI , cy-s] ) , B*VI ( VI ) , and B*VII ( VII ) , were used to construct knockout strains or were transformed with rDNA constructs . Strains B-VII-427 ( Chx1-1/Chx1-1 [VII , cy-s] ) and B2086 were used for excision assays because both contain only the Δ0 . 9 form of the M IES in their macronuclei . To promote synchronous mating , cells were starved at 30°C overnight in 10 mM Tris , pH 7 . 5 , prior to mixing at equal cell densities ( ~2 . 5x105 cells/ml ) . Individual pairs , >6 hrs after initiating mating , were transferred to individual drops of SPP and allowed to complete conjugation . Drops containing living cells after 2 days were transferred to 96 well plates containing starved CU428 or B2086 . Throughout the day wells were screened for mating pairs . Wells containing paired cells indicated that the initial drop plates had contained back-outs instead of progeny as progeny would not be sexually mature yet . Survival was scored as the percent of drops containing progeny versus the number of drops plated . To score conjugation endpoint , cells were fixed with 2% paraformaldehyde 24 hrs into mating and stained with DAPI . Total RNA ( 4μg ) , isolated by RNAsol extraction [51] , was converted to cDNA using SuperScript II reverse transcriptase as described [50] . PCR was performed using Lia3rt_FW and Lia3rt_RV primers ( S2 Table ) to monitor LIA3 expression and HhpI_FW and HhpI_RV primers ( S2 Table ) to monitor HhpI expression as a loading control . gDNA was isolated from mating cells at indicated times . Detection of IES junctions was performed by using PCR primers that amplify across the IES junction as described [2] . Detection of excised IES circles was performed by nested PCR using primers ( S2 Table ) pointing outward from the IES [52] [53] . PCR products were gel isolated and then TA cloned prior to sequencing . Electroporation of wild-type or ΔLIA3 mating cells with IES-containing rDNA vectors was performed as described [17 , 54] . Plasmids containing M or R IES sequences or chimeric IESs ( pMgtwM_m , pMgtwM_r , pRgtwR_m , pRgtwR_r ) were created by first replacing the IES sequence with a gateway recombination cassette then recombining the desired IES sequence into the desired vector . For Southern blot analysis , three individual paromomycin-resistant progeny lines , obtained after electroporation of wild-type or mutant mating cells with IES-containing vectors , were co-cultured for genomic DNA isolation . Ten μg of each DNA preparation was digested with either NotI or BamHI , fractionated on 1% agarose gels , transferred to nylon membranes and hybridized to M or R IES ( from pDLCM3 or pDLCR5 , respectively ) [55] . DNA encoding an N-terminal His tagged Lia3 ( His-Lia3 ) was codon optimized for expression in E . coli by Life Technologies . His-Lia3 was cloned into NcoI and XbaI sites of pBAD ( a gift from Dr . R . Kranz , Washington University ) to make pBAD-HisLia3 . His-Lia3 was expressed in E . coli strain BL21 ( DE3 ) . After reaching an OD600 ~0 . 8 , 0 . 2% wt/vol Arabinose was added and cells continued to grow for 4hrs before harvesting . Cell pellet was resuspended in native lysis buffer ( 50 mM NaH2PO4 , 300 mM NaCl , 10 mM Imidazole , and protease inhibitors ) and lysed using a French press ( 1200 psi ) . Cells were centrifuged at 40 , 000 rpm for 45min at 4°C and the pellet was resuspended in denaturing buffer ( 100 mM NaH2PO4 , 10 mM Tris-HCl pH 7 . 5 , 8M Urea , and protease inhibitors ) and stirred on ice for 1 hr . Cell lysate was spun at 10 , 000 x g for 30 min and the supernatant was incubated with Ni-Nta resin for 1 hr before loading onto the column . After washing with 50 mM NaH2PO4 , 300 mM NaCl , and 20 mM Imidazole , proteins were eluted in 50 mM NaH2PO4 , 300 mM NaCl , and 125 mM Imidazole and subsequently dialyzed against 25 mM Tris pH 7 . 5 , 100 mM KCl , 1 mM DTT , 1 mM EDTA , and 10% glycerol before storage at -80°C . The His-Lia3 coding sequence was PCR amplified using oligonucleotides listed in S2 Table to add an N-terminal TEV protease cleavage sites and BamHI and HindIII sites . The amplified DNA was cloned into the pMAL-C2X expression vector ( New England Biolabs , Ipswich , MA ) to create pMAL-TEV-hisLIA3 . The plasmid was transformed into BL21 ( De3 ) cells and the recombinant protein was purified by using an amylose resin as described [56] and then dialyzed against 25 mM Tris pH 7 . 5 , 100 mM KCl , 1 mM DTT , 1 mM EDTA , and 10% glycerol before storage at -80°C . Oligonucleotides ( Table 1 ) were labeled by incubation with T4 PNK and [γ -32P] ATP for 1 hr at 37°C and then purified using Roche oligo spin columns . Oligos were made double-stranded by mixing equal amounts of complementary oligonucleotides in 10 mM Tris pH 7 . 5 , 5% glycerol , and 100mM LiCl and boiling for 5 min , followed by slowly allowing the oligonucleotides to cool to RT . Prior to gel shifts , oligonucleotides were boiled for 5 min in 10 mM Tris pH7 . 5 , 5% glycerol , and either 100 mM KCl or 100 mM LiCl and slow cooled to RT to allow structures to form . For binding and competition experiments , 50–400 nM His-Lia3 , MBP-Lia3 , or MBP-MS2 was incubated with unlabeled competitor oligonucleotides ( Table 1 ) in binding buffer ( 10 mM Tris pH 7 . 5 , 1 mM EDTA , 0 . 1 mM DTT , 5% vol/vol glycerol , 0 . 010 mg/ml BSA , and either 100 mM KCl or 100 mM LiCl ) for 15 min at RT . KCl was used in all binding reactions except when LiCl was use to limit quadruplex formation . Followed by addition of 4 nM 32P-labeled oligonucleotide and incubation for another 15 min at RT before 4 μl was loaded onto a 4 . 5% polyacrylamide ( 75:1 acrylamide:bisacrylamide ) gel . After pre-running gels at 140V for 30 min , samples were fractionated by electrophoresis at 140V for 1 hr 45 min . Gels were subsequently vacuum dried for 1 hr prior to exposure to X-ray film or to a Phosphorimager screen . For binding curve experiments , 4 nM 32P-labeled oligo was incubated with 0 , 50 , 75 , 100 , 150 , 200 , 250 , 300 , 400 , 500 , 700 , 900 , 1100 , 1300 , 1500 , 1900 , 2100 , 2500 , or 3000 nM His-Lia3 for 20 min at RT prior to loading on gel . Oligonucleotides were boiled for 5 min in 10 mM Tris pH 7 . 5 , and either 100 mM KCl or 100 mM LiCl and slow cooled to 4°C . The CD spectra were recorded on a J-810 spectropolarimeter ( Jasco ) . The measurements were carried out with 500 μL 3 μM ODN samples at 4°C under nitrogen . Spectra shown are the average of 3 scans in a range from 220 to 300 nm with a band width of 1 nm , response time of 0 . 5 s , data pitch of 0 . 2 nm , and scan speed of 50 nm/min . A blank sample of 10 mM Tris pH 7 . 5 with 100 mM KCl or 100 mM LiCl was used for baseline correction . Strains CU428 and B2086 were starved in 10 mM Tris-HCl ( PH 7 . 5 ) and mixed to induce mating . Between 9 and 10 hours post-mixing , mating Tetrahymena cells were fixed in 3% PHEMS-paraformaldehyde essentially as described [57] , incubated overnight with a 1:200 dilution of anti-G4 antisera ( 1H6- EMD Millipore , Billerica , MA ) [27] , and detected with Alexa 488-conjugated , goat , anti-mouse antisera . Cells were counterstained with DAPI ( 4' , 6-diamidino-2-phenylindole ) and imaged on a Nikon E600 epiflourescent microscope equipped a Retiga EX CCD camera ( Q imaging , Burnaby . B . C . Canada ) with Openlab acquisition software v404 ( Improvision ) . | Non-canonical DNA structures , including four-stranded Guanine quadruplexes ( G4 DNA ) , have been observed readily in vitro , but their regulatory importance within cells has been particularly challenging to demonstrate conclusively . We have discovered a G4 DNA binding protein , Lia3 , that specifically regulates programmed DNA elimination events in Tetrahymena thermophila . This ciliate deletes nearly one-third of its germline genome from each developing somatic nucleus . These genomic deletion events must be accurate as the thousands of DNA regions excised are located near genes and/or their promoters , thus aberrant excision may alter gene expression . When we knocked out the gene encoding Lia3 , we found that the boundaries of the excised regions were heterogeneous for a subset of loci that are flanked by G-rich ( 5’-AAAAAGGGGG-3’ ) boundary controlling sequences . When we tested whether Lia3 bound this sequence , we discovered that the sequence itself formed G4 DNA and that Lia3 bound only when the sequence adopted this conformation . Our findings that Lia3 binds G4 DNA and that deletion of the gene encoding Lia3 perturbs the boundaries of the excised loci which are flanked by this quadruplex-forming DNA provides compelling evidence that this non-canonical DNA structure has a critical role during development of these cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"sequencing",
"techniques",
"chromosome",
"structure",
"and",
"function",
"molecular",
"probe",
"techniques",
"dna-binding",
"proteins",
"organisms",
"telomeres",
"protozoans",
"materials",
"science",
"dna",
"dna",
"structure",
"molecular",
"biology",
"techniques",
"cilia... | 2016 | A Parallel G Quadruplex-Binding Protein Regulates the Boundaries of DNA Elimination Events of Tetrahymena thermophila |
Fungal septicemia is an increasingly common complication of immunocompromised patients worldwide . Candida species are the leading cause of invasive mycoses with Candida glabrata being the second most frequently isolated Candida species from Intensive Care Unit patients . Despite its clinical importance , very little is known about the mechanisms that C . glabrata employs to survive the antimicrobial and immune response of the mammalian host . Here , to decipher the interaction of C . glabrata with the host immune cells , we have screened a library of 18 , 350 C . glabrata Tn7 insertion mutants for reduced survival in human THP-1 macrophages via signature-tagged mutagenesis approach . A total of 56 genes , belonging to diverse biological processes including chromatin organization and golgi vesicle transport , were identified which are required for survival and/or replication of C . glabrata in macrophages . We report for the first time that C . glabrata wild-type cells respond to the intracellular milieu of macrophage by modifying their chromatin structure and chromatin resistance to micrococcal nuclease digestion , altered epigenetic signature , decreased protein acetylation and increased cellular lysine deacetylase activity are the hall-marks of macrophage-internalized C . glabrata cells . Consistent with this , mutants defective in chromatin organization ( Cgrsc3-aΔ , Cgrsc3-bΔ , Cgrsc3-aΔbΔ , Cgrtt109Δ ) and DNA damage repair ( Cgrtt107Δ , Cgsgs1Δ ) showed attenuated virulence in the murine model of disseminated candidiasis . Further , genome-wide transcriptional profiling analysis on THP-1 macrophage-internalized yeasts revealed deregulation of energy metabolism in Cgrsc3-aΔ and Cgrtt109Δ mutants . Collectively , our findings establish chromatin remodeling as a central regulator of survival strategies which facilitates a reprogramming of cellular energy metabolism in macrophage-internalized C . glabrata cells and provide protection against DNA damage .
Candida glabrata , a nondimorphic , haploid budding yeast , is emerging as an important nosocomial fungal pathogen with an attributable mortality rate of ∼ 30% [1] . It normally resides as a commensal in the normal flora of human mucosal tissues but can cause infections ranging from superficial mucosal to invasive , life-threatening systemic infections in immunocompromised patients [2] . C . glabrata accounts for 12% of total Candida blood stream infections and is the second most common cause of candidiasis after C . albicans [3] . Although C . glabrata shares some virulence factors with C . albicans including phenotypic switching , bio-film formation and ability to adhere to the host tissues , it lacks key virulence traits of C . albicans such as hyphae formation , mating and secreted proteolytic activity [4] , [5] . Despite the lack of these virulence mechanisms , C . glabrata is capable of establishing successful disseminated infections suggesting that it does possess a repertoire of other virulence factors . Niacin limitation responsive , chromatin-based regulation of a large family of adhesins is a unique virulence attribute of C . glabrata [6] , [7] . Candida spp . are highly evolved for interaction with , and survival in , the human host . Both innate and adaptive immunity contribute towards host's resistance to candidiasis [8] . C . albicans is recognized primarily by two pathogen associated molecular markers , β-glucan and mannan , which constitute ∼ 90% of the cell wall dry weight [9] . The cell wall is composed of an outer layer of mannoproteins covalently linked to an inner core of β-glucan [10] . Polymorphonuclear neutrophils , macrophages and dendritic cells are natural protective barrier against systemic candidiasis and impaired phagocytic function is a major risk factor for disseminated candidiasis [11] . Dendritic cells recognize and phagocytose C . albicans to process them for antigen presentation and can differentiate between yeast and hyphal form to initiate the T-helper cellular immune response , which is required for long term resistance to candidiasis [12] , [13] . Macrophages contribute to antifungal defense via phagocytosis and clearance of the fungal pathogen [14] . Nonpathogenic yeast Saccharomyces cerevisiae is unable to replicate in murine macrophage-like cells while C . albicans undergoes morphological switching and the hyphal form penetrates out of macrophages ultimately killing them [15] , [16] . Contrary to this , C . glabrata replicates in murine macrophages and human monocyte-derived macrophages ( MDMs ) without inducing apoptosis or causing any significant damage to macrophages [17] , [18] . Transcriptional profiling of murine macrophage-internalized C . glabrata cells revealed significant remodeling of carbon metabolism [17] . Recently , C . glabrata has been shown to modulate the phagolysosome maturation and to subvert the macrophage cytokine production [18] . Signature-tagged mutagenesis ( STM ) is a widely employed approach to identify virulence factors in microbial pathogens wherein multiple mutants are simultaneously screened for increased/decreased fit via DNA signature tags [19] . We have previously generated a Tn7 insertion mutant library by random , in vitro Tn7-based insertional mutagenesis approach in 96-uniquely tagged wild-type strains [20] . Here , we present findings from a genomic analysis of C . glabrata's interaction with human macrophages . By screening a Tn7 insertion mutant library , representing ∼ 50% of C . glabrata genome , for reduced survival in a cell culture model system via STM approach , we have identified a set of 56 genes which are required for survival/replication in THP-1 macrophages . Further , we report for the first time that C . glabrata modifies its chromatin architecture in response to macrophage intracellular environment and the mutants disrupted for chromatin remodeling display survival defects in macrophages and in the murine model of systemic candidiasis . Our data suggest a link between cellular energy status and chromatin organization which is crucial for survival of C . glabrata in macrophages .
To study the interaction of C . glabrata with human macrophages , we first established infection dynamics of C . glabrata cells with human monocytic cell line THP-1 . Infection studies of PMA ( Phorbol-12 Myristate 13-acetate ) -differentiated THP-1 cells with C . glabrata cells at a MOI ( multiplicity of infection ) of 1∶10 revealed a moderate 5- to 7-fold increase in the wild-type ( wt ) colony forming units ( CFUs ) over a period of 24 h ( Figure 1A and B ) while only 1% cells remained viable for a C . glabrata mutant ( Cgyps1–11Δ ) which lacked eleven cell surface-associated aspartyl proteases ( Figure 1A ) . Importantly , both wt and protease-defective mutant grew well in RPMI medium and were phagocytosed by THP-1 cells at a similar rate ( 58–62%; data not shown ) . Notably , replication of C . glabrata wt cells remained unaffected over a wide range of MOI although a few extracellular yeast cells were detected at 10∶1 MOI after 24 h of co-culturing with human macrophages ( data not shown ) . Confocal fluorescence microscopic analysis of C . glabrata-infected THP-1 macrophages revealed that two hours after co-incubation , many macrophages were infected with 1–2 yeast cells , however , 24 h post infection , the number increased to 6–10 yeast per macrophage ( Figure 1B ) thus corroborating the CFU-based intracellular replication of wt C . glabrata cells . C . glabrata cells were also able to replicate in mouse peritoneal macrophages wherein a 6–7 fold increase in CFUs was observed over a period of 24 h ( data not shown ) . This ability of C . glabrata to replicate in activated THP-1 and mouse peritoneal macrophages is consistent with earlier findings where a 4- to 6-fold and a 3-fold increase in CFUs was observed in mouse macrophage-like cells ( J774A . 1 ) and MDMs , respectively [17]–[18] . Further , the intracellular nature of C . glabrata cells in THP-1 cells post 24 h infection was verified by inside/outside staining using anti-Epa1 ( Epithelial adhesin 1 ) antibody which revealed no measurable extracellular yeast ( Figure S1A ) . A closer examination of C . glabrata's intra-cellular behaviour via trypan blue staining and cell cycle analysis found 10% and 20% of wt cell population to be dead after 2 h and 6 h of incubation with macrophages , respectively ( S1B and data not shown ) . However , negligible yeast cell death and a 5- to 7-fold increase in the number of internalized yeast 12 h and 24 h post infection , respectively ( Figures 1A and S1B ) suggested that C . glabrata survives an initial burst of macrophage-elicited antimicrobial response . In accord with this , C . glabrata-infected THP-1 and mouse peritoneal macrophages displayed a ∼ 5- and 14-fold elevation in the reactive oxygen species ( ROS ) levels 2 h post infection ( Figure 1C and data not shown ) . Importantly , treatment of THP-1 cells with NADPH-oxidase inhibitor DPI ( diphenyleneiodonium ) reversed the ROS production to basal levels ( Figure 1 C ) indicating that elicitation of the oxidative burst is an active process . Consistent with above data , intracellular levels of ROS in 2 and 6 h macrophage-internalized C . glabrata cells were elevated by 7-fold compared to RPMI-grown cells ( Figure S1C ) indicating that C . glabrata cells encounter oxidative stress in macrophage internal milieu . Further , labelling of C . glabrata-infected THP-1 cells with lysosomotropic , fluorescent dye LysoTracker Red DND99 , which accumulates in acidic compartments , revealed that GFP-expressing , live C . glabrata cells do not co-localize with the acidic phagosomal compartment ( Figure 1D ) . In contrast , heat-killed yeast cells exhibited co-localization with ∼ 90–95% phagosomes being lysotracker positive ( Figure 1D ) . This suggests that live C . glabrata cells possess the ability to modulate the acidification of phagosome and is consistent with the findings of a recent study [18] . Surprisingly , Cgyps1–11Δ cells , which lost viability upon macrophage internalization , were found to be competent in preventing the maturation of phagolysosome as GFP-fluorescing live mutant cells did not co-localize with the LysoTracker stained lysosomes ( Figure 1D ) . This implies that inhibition of phagosome acidification is one of the mechanisms that C . glabrata employs to survive the internal milieu of macrophages . Cytokine profiling of C . glabrata-infected THP-1 cells revealed an increase in the secretion of anti-inflammatory cytokine IL-4 compared to uninfected cells while no such induction was observed for other cytokines including IL-6 and IL-10 ( Figure S1D and data not shown ) . Notably , C . albicans infection and LPS treatment of THP-1 cells led to the induction of IL-4 and IL-6 cytokines , respectively ( Figure S1D ) . Together , these data suggest that activated THP-1 macrophages elicit an oxidative stress response and a modest induction of IL-4 secretion upon C . glabrata infection , however , C . glabrata manages to survive and counteract these antimicrobial responses using multiple strategies . To identify genes involved in survival/replication of C . glabrata in THP-1 cells , we screened a C . glabrata mutant library for altered survival profiles via STM approach . This mutant library is composed of 18 , 350 mutants and was created by homologous recombination of in vitro generated Tn7 insertions in the C . glabrata genomic clones [20] . These mutants had been assembled in a total of 192 pools wherein each pool is comprised of 96 mutants . Each of these pools carries the same set of 96 tags but within a pool each mutant contains a different tag , thus , allowing a parallel analysis of 96 mutants in each infection experiment . For a pool of tagged mutants , the ratio of hybridization in the output ( macrophage-recovered yeast cells grown in YPD medium ) and the input ( yeast cells grown in YPD medium ) pools reflected any shift in the representation of the corresponding mutant in the pool . An MOI of 1∶10 was chosen for the mutant screen because no extracellular yeast were observed even after 48 h co-incubation with THP-1 cells ( data not shown ) . For input , each C . glabrata mutant pool was grown in the YPD medium for 14 h at 37°C . For output , differentiated THP-1 cells were infected with each pool of 96-tagged mutants and intracellular yeast were recovered by lysing macrophages 24 h post infection . All the 192 mutant pools were screened and analyzed using hybridization-based STM approach . As controls for experimental reproducibility , 10 mutant pools were screened in macrophages in replicates and 15 pools were hybridized to filters in duplicates . In both cases , similar results were obtained ( data not shown ) . The mutants with an output/input ratio of ≥6 and ≤0 . 1 from the STM screen were selected as ‘up’ ( increased survival ) and ‘down’ ( reduced survival ) mutants , respectively ( Tables S1 and S2 ) . Using this cut-off value , a total of 168 ( 35 up and 133 down ) mutants were identified which displayed altered survival profiles in differentiated THP-1 cells ( Table S1 ) . For the current study , we focussed on down mutants which displayed an output to input ratio of ≤0 . 1 implying at least a 10-fold underrepresentation of the unique tag in macrophage-recovered mutant cells compared to RPMI-grown mutant cells in the STM screen ( Tables S1 and S2 ) and these , henceforth , have been referred as mutants that exhibited either≤ten-fold reduced survival or were 10-fold down in macrophages . The attenuated growth of selected mutants was validated by screening composite pools ( a mixture of down mutants and tagged mutants ( wt-like survival in macrophages ) in THP-1 cells via hybridization-based STM manner . Phenotypic profiling of 133 down mutants under several stressful conditions including pH and oxidative stress revealed the diverse nature of the identified mutants and overlapping sensitivities to different stresses were observed for very few mutants ( Figure S2A and data not shown ) thereby precluding the possibility of general sick mutants ( slow-growers ) coming through the screen . Tn7 insertion in the identified mutants was mapped , by transposon rescue and sequencing analyses , to 56 C . glabrata ORFs and 20 intergenic regions ( http://www . genolevures . org ) ( Tables S1 and S2 ) . Using the Saccharomyces Genome Database ( SGD ) Gene Ontology Slim Mapper ( http://www . yeastgenome . org ) , the identified C . glabrata genes were functionally annotated to biological processes based upon the information available for their S . cerevisiae orthologs ( Table S2 ) . 12% of the identified genes were involved in chromatin organization while genes implicated in DNA repair , golgi vesicle transport , and endocytosis constituted 9% , 9% and 7% , respectively of the total identified genes ( Table S2 ) . Of the mutants identified , we selected 10 ‘down’ mutants , which were defective for either chromatin organization or DNA replication and repair , for further analysis . The putative functions of the genes disrupted in these mutants are listed in table S2 . First , we confirmed the survival defects of Tn7 insertion mutants in THP-1 cells by single infection assays . The survival ratio of 0 . 4 to 0 . 6 for mutants defective in chromatin organization and DNA repair ( Figure S2B ) validated their inability to replicate properly in THP-1 cells . Liquid growth assays revealed no significant differences in the growth profiles for any mutant compared to the wt in YPD and RPMI medium ( Figure S2C and D ) . Cgdna2 mutant was exquisitely sensitive to methyl methane sulfonate ( MMS ) , a DNA alkylating agent , while growth of Cgchz1 mutant was impaired on plates containing MMS , hydrogen peroxide ( H2O2 ) , camptothecin ( CPT ) , a DNA topoisomerase inhibitor and hydroxyurea ( HU ) , a DNA replication inhibitor ( Figure S2E ) . Surprisingly , Cgsgs1 mutant displayed sensitivity neither to oxidative nor to replication/genotoxic stress ( Figure S2E ) . Growth of Cghfi1 was attenuated in the presence of H2O2 and HU . H2O2 also mildly inhibited the growth of Cgarp7 , Cgrtt107 , Cgrtt109 and Cgrsc3 mutants ( Figure S2E ) . Cgrtt107 mutant exhibited sensitivity to MMS and CPT ( Figure S2E ) . Together , these results indicate varied levels of sensitivity of chromatin organization and DNA repair defective mutants towards genotoxic and oxidative stress causing agents . As seven genes ( CgARP7 , CgCHZ1 , CgFPR4 , CgHFI1 , CgRSC3-A , CgRSC3-B , CgRTT109 ) , identified through the STM screen , are directly involved in maintaining chromatin architecture , we examined whether chromatin dynamics of C . glabrata cells is altered upon internalization by THP-1 macrophages via micrococcal nuclease ( MNase ) digestion assay . Chromatin isolated from internalized yeast post 2 h macrophage infection displayed sensitivity to MNase digestion similar to that of the RPMI-grown cells as observed by the appearance of nucleosomal bands ( Figure 2A ) . In contrast , chromatin extracted from 6 h and 12 h macrophage-internalized yeast was highly resistant to MNase digestion ( Figure 2A ) and this resistance was reversed in yeast recovered 24 h post infection ( Figure S3 ) suggesting that C . glabrata cells adapt to the macrophage environment by remodeling their chromatin architecture . Biochemical post-translational modifications ( PTMs ) of histones H2A , H2B , H3 and H4 , the protein components of the nucleosome core , contribute strongly to the structural organization of the chromatin [21] . To corroborate the altered chromatin architecture in macrophage-internalized C . glabrata cells , we examined both the total levels of histone proteins ( H1 , H2A , H2B , H3 and H4 ) as well as the PTMs of histones H3 and H4 in RPMI-cultured and macrophage-internalized yeast . Total levels of histones H1 , H2A and H2B in macrophage-ingested yeast were elevated throughout the 24 h infection time course ( Figures 2B and S4A ) . In contrast , H4 levels were reduced compared to the RPMI-grown cells 2 , 6 , and 12 h post infection ( Figures 2B and S4A ) . Intriguingly , H3 histone levels were significantly higher 6 h and 12 h post infection ( Figures 2B and S4A ) . This non-tandem expression of H1 , H2A , H2B , and H3 with H4 was quite unexpected and the biological significance of this observation remains to be determined . Notably , elevated synthesis of histones results in chromatin condensation which may protect yeast DNA from ROS-mediated damage . Alternately , increased levels of H1 , H2A , H2B and H3 may reflect the growth phase of cells as majority of histones in proliferating cells are synthesized during S-phase [22] . To investigate if changes in the histone amounts are due to their transcriptional regulation , we performed qRT-PCR analysis on 2 h and 10 h macrophage-internalized yeast . While a three- to eight-fold increase in the transcript levels of H2A , H2B , H3 and H4 coding ORFs was observed in 10 h macrophage-internalized yeast , no significant activation of the transcripts encoding histones was observed 2 h post infection ( Figure 2C ) suggesting that increased histone levels in 2 h macrophage-ingested yeast probably reflect stabilized proteins . The basis for the reduced levels of H4 despite the transcriptional activation of CgHHF2 and CgHHF3 , upon macrophage internalization is not clear and warrants further investigation . Examination of the histone PTMs revealed that acetylation of H3 at lysine 9 and 56 was higher in macrophage internalized yeast compared to the RPMI-cultured cells , however , the acetylation was reduced by ∼ 25–50% compared to the total H3 levels 6 and 12 h post infection ( Figures 2B , S4B and C ) . Similarly , acetylation of H3 at lysine 14 and of H4 on lysine 16 was diminished in 6 and 12 h macrophage-internalized yeast when normalized to total H3 and H4 levels , respectively ( Figures 2B , S4B and C ) . Further , histones extracted from 2 , 6 and 12 h macrophage-ingested yeast were enriched for H3K9Me3 and H3K27Me2 marks ( Figures 2B , S4B , C and D ) . Increased phosphorylation of histone H3 at serine 10 , reflective of condensed chromosomes , was also observed in internalized yeast ( Figures 2B , S4B and C ) . Overall , these findings indicate that histone modifications that mark silent ( H3K9Me3 , H3K27Me2 , and H4K20Me3 ) and transcriptionally active ( H3K9Ac , H3K14Ac , H3K56Ac , H4K16Ac ) chromatin are largely elevated and diminished in 6 and 12 h THP-1-internalized C . glabrata cells , respectively . Chromatin extracted from 24 h internalized yeast cells displayed predominantly an epigenetic signature ( a particular pattern of DNA methylation , histone modifications and chromatin structure ) of the active chromatin which is consistent with its sensitivity to MNase digestion ( Figure 2A and B ) . Collectively , these data suggest a highly compact chromatin in 6 and 12 h macrophage-internalized C . glabrata cells . To investigate the role of chromatin remodeling in survival in THP-1 macrophages , we decided to disrupt CgRSC3-A , CgRSC3-B , CgRTT107 , CgRTT109 , and CgSGS1 genes belonging to the GO category of either chromatin organization or DNA damage , and generated single deletion strains for these five and a double deletion strain for CgRSC3-A and CgRSC3-B genes . Our attempts to disrupt CgARP7 , CgCHZ1 , CgCTI6 , CgMRE11 , and CgHFI1 were unsuccessful implying that these genes are probably essential in C . glabrata . All subsequent experiments were performed with clean deletion mutants for CgRSC3-A , CgRSC3-B , CgRSC3-A and B , CgRTT107 , CgRTT109 , and CgSGS1 genes . CgRsc3-A and CgRsc3-B are orthologs of S . cerevisiae Rsc3 ( Table S2 ) which possesses sequence-specific DNA binding transcription factor activity and is a component of a 17-subunit RSC ( remodel the structure of chromatin ) complex [23] . The RSC ATP-dependent chromatin remodeling complex is essential for mitotic growth in S . cerevisiae [24] . CgRsc3-A and CgRsc3-B share an amino acid identity of 33% and 28% , respectively , with Rsc3 , and 25% identity with each other . Owing to the presence of two orthologs , CgRsc3-A and CgRsc3-B in C . glabrata and their low levels of sequence identity with S . cerevisiae Rsc3 , it is plausible that CgRsc3-A and CgRsc3-B may not share all functions with S . cerevisiae Rsc3 protein . CgRTT109 encodes an acetyltransferase , whose counterpart in S . cerevisiae acetylates lysine 56 on histone H3 and functions in DNA replication and maintenance of genomic stability [25] . Orthologs of CgRTT107 and CgSGS1 in S . cerevisiae code for a BRCT ( BRCA1 C Terminus ) domain-containing protein and a RecQ-related nucleolar DNA helicase , respectively and are implicated in DNA double-strand break repair [26]–[27] . Since the STM screen was based on competitive growth assays and the mutants were selected via a DNA hybridization-based approach , the reduced survival/replication of Cgrsc3-aΔ , Cgrsc3-bΔ , Cgrtt107Δ , Cgrtt109Δ and Cgsgs1Δ mutants was examined by infecting THP-1 cells singly with either wt or mutant strain ( Figure 3A ) . These single infection assays validated the replication defects of Cgrsc3-aΔ , Cgrsc3-bΔ , Cgrtt107Δ , Cgrtt109Δ and Cgsgs1Δ mutants in THP-1 cells with mutants displaying survival ratio of 0 . 3 to 0 . 6 ( Figure 3A ) . Further , since both Cgrsc3-aΔ and Cgrsc3-bΔ mutants displayed growth defects in macrophages , the effect of simultaneous disruption of both genes may have been additive . However , contrary to the expectation , deletion of the two ORFs coding for CgRsc3 did not aggravate the replication defect in Cgrsc3-aΔbΔ mutant ( Figure 3A ) implying functional redundancy between CgRsc3 homologs . Importantly , all strains displayed similar growth profiles in YPD and RPMI medium ( Figure S5A and B ) . Cgrtt107Δ and Cgrtt109Δ exhibited sensitivity to MMS , CPT and H2O2 and this susceptibility was complemented by expressing the respective ORF from the plasmid ( Figures 3B and S5C ) . Sensitivity of Cgrtt107Δ and Cgrtt109Δ mutants to H2O2 was also confirmed by performing liquid growth assays ( Figure S5D ) . Compared to 14% growth inhibition for wt cells , H2O2 treatment led to 27% and 33% growth attenuation in Cgrtt107Δ and Cgrtt109Δ mutants , respectively ( Figure S5D ) . Cgrtt109Δ mutant also displayed sensitivity to fluconazole ( Figure 3B ) . Although growth of Cgrsc3-aΔ , Cgrsc3-bΔ , Cgrsc3-aΔbΔ was impaired in the presence of H2O2 , sensitivity to replication and DNA damage stress was not observed ( Figure 3B and data not shown ) . Surprisingly , Cgsgs1Δ , unlike its S . cerevisiae counterpart , grew like wt in the presence of DNA damaging and oxidative stress causing agents ( Figure 3B and data not shown ) suggesting that CgSgs1 is not essential for processing of DNA double-strand breaks in C . glabrata . To demonstrate that CgRSC3-A and CgRTT109 are implicated in maintaining chromatin structure in C . glabrata , we performed MNase digestion assay . Chromatin extracted from YPD-grown logarithmic-phase Cgrsc3-aΔ , Cgrsc3-aΔbΔ and Cgrtt109Δ cells exhibited enhanced resistance to MNase digestion ( Figure S5E ) suggesting an altered chromatin architecture . Next , we checked the status of histone levels and their PTMs in RPMI-grown and macrophage-internalized Cgrsc3-aΔ and Cgrtt109Δ cells . Similar to the wt , increased amounts of H2A and H2B and decreased levels of H4 were observed , upon macrophage ingestion , in mutants ( Figures 3C , S6A and B ) . Surprisingly , macrophage-internalized Cgrsc3-aΔ and Cgrtt109Δ mutant cells displayed low levels of H3 after 2 h co-incubation with macrophages ( Figures 3C , S6A and B ) which could be due to decreased H3 transcript levels and/or protein stability . Further , while H3 levels were elevated in 6 h macrophage-ingested Cgrtt109Δ cells , no appreciable increase in H3 was observed post 6 h infection in Cgrsc3-aΔ cells ( Figures 3C , S6A and B ) . Cgrtt109Δ cells showed a low level of residual acetylation of H3 on lysine 56 ( Figures 3C and S5F ) implicating CgRtt109 in the acetylation of H3 at lysine 56 residue . Acetylation of H3 on lysine 9 and lysine 56 was reduced in 6 h macrophage-internalized Cgrtt109Δ and Cgrsc3-aΔ cells , respectively , however , H3 was acetylated at lysine 9 in Cgrsc3-aΔ cells during 6–12 h period ( Figures 3C , S6A and B ) . Notably , unlike wt cells , total H3 levels in Cgrsc3-aΔ and Cgrtt109Δ cells , acetylation of H3 on lysine 9 in Cgrtt109Δ and acetylation of H3 at lysine 56 in Cgrsc3-aΔ cells were lower 24 h post infection ( Figures 3C , S6A and B ) suggesting diminished transcriptional activity . Phosphorylation of H3 on serine 10 was observed in 12 h and 24 h RPMI-grown Cgrtt109Δ cells ( Figures 3C and S6B ) . Intriguingly , tri-methylation of H4 at lysine 20 ( a repressive methyl mark ) , upon macrophage internalization , was significantly higher and lower in Cgrtt109Δ and Cgrsc3-aΔ cells , respectively ( Figures 3C , S6A and B ) . The implication of this observation is not clear and warrants further investigation . Together , these results indicate an altered epigenetic response of the chromatin remodeling mutants to the macrophage environment . To examine if C . glabrata cells encounter oxidative stress-induced DNA damage upon macrophage ingestion , we quantified the phosphorylation of H2AX ( γ-H2AX , double-strand break-induced serine-129 phosphorylated form of H2A variant ) and found it to be significantly increased in macrophage-internalized wt cells 2 h post infection ( Figures 2B and S4A and C ) . A 40% increase in the appearance of γ-H2AX foci , markers of DNA damage and repair , in 2 h macrophage-internalized yeasts compared to the RPMI-grown cells ( Figure 3D ) corroborated the activation of DNA damage signaling upon macrophage internalization . Phosphorylation of H2AX in macrophage-internalized Cgrsc3-aΔ cells , when normalized to total H2A levels , was comparatively less implying a subdued response to DNA damage ( Figures 3C and S6A ) . As a control , we also assessed DNA damage in Cgrtt107Δ cells and found constitutively elevated numbers of γ-H2AX foci ( Figure 3D ) . Consistent with this , phosphorylated form of H2AX was higher in RPMI-grown and 2 h macrophage-internalized Cgrtt107Δ cells ( Figures 3E and S6C ) reflective of a constitutively active DNA damage signaling thus validating the role for CgRtt107 in the repair of damaged DNA . Collectively , these data suggest that improper activation of DNA damage response and/or chromatin remodeling may contribute to the reduced survival of Cgrsc3-aΔ , Cgrtt107Δ and Cgrtt109Δ strains in macrophages . To investigate if altered chromatin structure of Cgrsc3-aΔ and Cgrtt109Δ leads to differential gene expression , we performed microarray analysis on 10 h RPMI-grown and macrophage-internalized wt and mutant cells . The datasets of differentially regulated genes ( ≥2-fold change with P≤0 . 05 ) were analysed for co-regulation by hierarchical clustering and annotated with GO term for biological process . We first compared the transcript profiles of RPMI-grown Cgrsc3-aΔ and Cgrtt109Δ cells with those of the RPMI-cultured wt cells and found 724 and 819 genes to be differentially regulated , respectively ( Figure 4A ) . Intriguingly , this gene set revealed a striking overlap with 300 induced and 252 repressed genes common to both the mutants . Consistent with the role of CgRSC3-A and CgRTT109 in chromatin organization , the common induced gene set included the genes involved in chromatin silencing and remodeling , RNA metabolism , ergosterol biosynthesis , DNA replication and repair ( Figure 4 B and data not shown ) . Additionally , genes implicated in protein glycosylation were significantly represented in the induced gene dataset of Cgrtt109Δ mutant ( Figure S7A ) . The repressed gene set included genes belonging to tricarboxylic acid cycle , iron-sulfur cluster assembly and mitochondrial electron transport ( Figures 4C and S7B ) . Notably , misregulation of energy metabolism in Cgrtt109Δ cells is consistent with the down-regulation of the mitochondrial function-related genes observed in rtt109−/− mutant in C . albicans [28] . Genes implicated in DNA-dependent regulation of transcription and protein phosphorylation were uniquely down-regulated in Cgrsc3-aΔ and Cgrtt109Δ mutant , respectively . Further , expression of iron-regulated transcriptional activator CgAft2 , glucose-responsive factor CgRgt1 , oleate-activated factor CgOaf1 and MAPK signaling-activated transcriptional factor CgSte12 was 2- to 3-fold lower uniquely in Cgrsc3-aΔ mutant suggesting a positive regulatory role for CgRSC3 in their gene expression . These data suggest that CgRsc3-a and CgRtt109-mediated chromatin organization impact transcription and exert a similar global regulatory effect on cellular physiology . A total of 214 , 432 and 683 genes were found to be differentially expressed in wt , Cgrsc3-aΔ and Cgrtt109Δ cells , respectively in response to macrophage internalization ( Figure 4D and E ) . Of these , 100 , 219 and 346 were induced and 114 , 213 and 337 were repressed in wt , Cgrsc3-aΔ and Cgrtt109Δ cells , respectively . These genes were functionally annotated via gene ontology analysis performed with GO Slim Mapper at CGD ( http://www . candidagenome . org/cgi-bin/GO/goTermMapper ) . The global gene expression analysis of 10 h wt internalized yeast revealed up-regulation of the genes involved in ammonium transport , glyoxylate cycle , β-oxidation of fatty acids , meiosis , signal transduction and proteolysis ( Figures S7B , S8 and S9 ) . The repressed genes were implicated in iron transport and homeostasis , ergosterol biosynthesis , cell wall metabolism and response to stress ( Figures 4F , S8 and S9 ) . The repression of reductive high-affinity iron assimilation in 10 h macrophage-internalized C . glabrata cells implies either an iron-rich internal milieu of macrophages or an anaerobic environment . Notably , iron homeostasis has previously been linked with sterol biosynthesis and oxygen availability [29] . Similar to the wt , macrophage-internalized Cgrsc3-aΔ and Cgrtt109Δ cells showed up-regulation of tricarboxylic acid cycle , β-oxidation of fatty acids and signal transduction , however , of total up-regulated genes , only 2% of induced genes belonged to signal transduction in Cgrsc3-aΔ and Cgrtt109Δ compared to 9% signal transduction-related up-regulated genes in wt ( Figure S8 ) . Further , contrary to the wt cells , Cgrsc3-aΔ and Cgrtt109Δ cells , upon macrophage internalization , displayed induction of the genes implicated in the generation of precursor of metabolites and energy , cellular respiration , respiratory electron transport chain and cellular amino acid metabolic process ( Figures 4C , S7C , S8 ) . Genes implicated in iron-metabolism , ergosterol biosynthesis and cell wall metabolism were down-regulated in macrophage-internalized Cgrsc3-aΔ and Cgrtt109Δ cells similar to their expression pattern in macrophage-ingested wt cells ( Figures 4C and S7C ) . Interestingly , while the down-regulated gene set of wt cells contained 5% of translation-related genes , this gene class constituted only 1% and 2% of total repressed genes in Cgrtt109Δ and Cgrsc3-aΔ , respectively ( Figure S8 ) . Overall , microarray data indicate that transcriptional responses of wt and mutant cells ( Cgrsc3-aΔ and Cgrtt109Δ ) to macrophage internalization are largely overlapping with genes implicated in many processes including carbohydrate metabolic process , iron and ergosterol metabolism and cell wall organization displaying differential regulation ( Figure S8 ) . However , a major difference is the striking up-regulation of genes involved in generation of precursor of metabolites and energy and cellular respiration in macrophage-ingested Cgrsc3-aΔ and Cgrtt109Δ cells ( Figure S8 ) which may lead to an imbalance between energy demand and production in mutant cells , thus , adversely affecting the intracellular survival and/or replication . Next , we compared in parallel the mRNA profiles of macrophage-internalized Cgrsc3-aΔ and Cgrtt109Δ cells with those of the macrophage-internalized wt cells and found a striking overlap with a set of 355 induced and 472 repressed genes common to both mutants ( Figure S9A and B ) . Further , up-regulation of the genes involved in DNA replication , amino acid biosynthesis , protein glycosylation and chromatin silencing and down regulation of the genes implicated in vesicle-mediated transport was observed in Cgrsc3-aΔ and Cgrtt109Δ cells in response to macrophage environment compared to the wt cells ( Figure S7A , C and D ) . Overall , ORFs either unique to C . glabrata or with no known function represented 15%–30% of the total differentially regulated genes among various analyzed gene sets . To validate the microarray data set , we performed qRT-PCR analysis on a set of twenty seven genes including highly up-and down-regulated genes in wt and metabolic and iron-responsive genes in Cgrsc3-aΔ and Cgrtt109Δ cells and observed good agreement between the microarray expression data and the transcript level measurement by qRT-PCR ( Figure S9C , D and E ) . Collectively , the specific up-regulation of genes implicated in cellular respiration , amino acid metabolism and chromatin silencing in macrophage-internalized Cgrsc3-aΔ and Cgrtt109Δ suggests that the mutant cells , probably due to global changes in their chromatin architecture , are unable to mount an appropriate response to restrain the cellular energy metabolism . Reprogrammed carbon metabolism , characterized by decreased glycolysis and increased gluconeogenesis , glyoxylate cycle and fatty acid degradation is a major hallmark of macrophage-internalized fungal pathogens [15] . Thus , it is plausible that glucose limitation and/or presence of alternative carbon sources in the internal milieu of macrophages is a cue for remodeling of chromatin . To test this , we first examined the ability of the mutants defective in chromatin organization to utilize different compounds as sole carbon sources . As shown in figures 5A and S10A , Cgrsc3-aΔ , Cgrsc3-bΔ , Cgrsc3-aΔbΔ , Cgchz1 and Cghfi1 were attenuated for growth on plates containing oleic acid , ethanol , sodium acetate , citric acid and lactic acid as sole carbon source . A mutant disrupted for CgICL1 ( encodes Isocitrate Lyase , an enzyme of glyoxylate cycle ) was used as a control . As expected , Cgicl1Δ mutant couldn't utilize any of the alternative carbon sources ( Figure 5A ) . The inability of Cgrsc3-aΔ , Cgrsc3-bΔ and Cgrsc3-aΔbΔ mutants to utilize sodium acetate and lactic acid as carbon sources was also validated by liquid time course analyses ( Figure S10B , C and D ) . Further , chromatin extracted from wt cells grown on sodium acetate medium displayed resistance to MNase digestion and mirrored the reduced acetylation levels of H3 on lysine 56 of macrophage-internalized cells ( Figure S10E ) implying a similar cellular response to macrophage environment and utilization of alternative carbon sources . Since protein lysine acetylation is important for metabolic regulation [30] , we next checked whether the overall acetylation status of yeast cellular proteins is altered in response to the macrophage environment . Significantly lower levels of lysine-acetylated proteins were observed in macrophage-internalized yeast compared to the RPMI-cultured cells ( Figure 5B ) . In accord with this , macrophage-internalized wt cells displayed a 3- to 4-fold increase in the lysine deacetylase ( KDAC ) activity ( Figure 5C ) and a 1 . 5-fold higher intracellular ratio of NAD+ to NADH ( data not shown ) implying a low energy status . KDAC activity was also elevated when cells were grown on medium containing sodium acetate as sole carbon source ( Figure S10F ) . Surprisingly , Cgrsc3-aΔ and Cgrtt109Δ cells did not show elevated KDAC activity upon macrophage internalization indicating an impaired epigenetic and metabolic regulation ( Figure 5C ) . Cellular ATP measurement , to assess the energy status of macrophage-ingested cells , revealed no appreciable change and a 4-fold decrease in the total ATP levels after 2 and 12 h of co-incubation with macrophages , respectively ( Figure 5D ) which may be a reflection of nutrient-poor intracellular environment , low metabolic activity and/or increased ATP consumption due to activated stress responses including ATP-dependent chromatin remodeling . To examine if chromatin remodeling is important for the virulence of C . glabrata , we assessed the fungal burden for wt and chromatin organization mutants in the murine model of systemic candidiasis . While 6×105 yeast could be recovered from the kidneys of mice infected with wt C . glabrata cells , 25- to 50-fold lower yeast CFUs for Cgrsc3-aΔ Cgrsc3-bΔ and Cgrsc3-aΔbΔ and 3- to 5-fold lower CFUs for Cgrtt107Δ , Cgrtt109Δ and Cgsgs1Δ infected mice were obtained . No statistically significant differences in the fungal burden were seen in the liver and spleen of wt , Cgrsc3Δ , Cgrtt107Δ and Cgsgs1Δ infected mice ( Figure 6 ) . Interestingly , Cgrtt109Δ mutant showed reduced survival in all the three target organs , kidney , liver and spleen ( Figure 6 ) . Notably , similar to the survival in macrophages , an additive effect of CgRSC3-A and CgRSC3-B disruption was not observed in mice ( Figure 6 ) suggesting a functional redundancy . Taken together , these data implicate the CgRSC3-A , CgRSC3-B , CgRTT107 , CgRTT109 , and CgSGS1 genes in the survival of C . glabrata in the mammalian host .
C . glabrata has emerged as the second most common cause of invasive candidiasis , however , our knowledge about the strategies , it employs to multiply and avoid recognition by host phagocytic cells , is very limited . To this end , we have identified a set of 56 genes , via STM approach , that are required for survival/replication of C . glabrata in human macrophages . Of these , 53 genes are novel and orthologs of CgENA1 , CgRTT109 and CgSEF1 have been implicated in the virulence in Cryptococcus neoformans [31] and C . albicans [28] , [32] , respectively . Seven of the identified genes encode proteins which were potentially involved in chromatin organization while five of the identified gene products were implicated in DNA repair ( Table S2 ) . As a prelude to identify strategies that C . glabrata might utilize to survive and replicate in macrophages , we examined , in detail , the early response of Cgyps1–11Δ which is unable to retain its viability in THP-1cells , to the macrophage environment . However , no appreciable differences were observed between the wt and Cgyps1–11Δ in the inhibition of phagolysosome maturation ( Figure 1D ) , sensitivity to oxidative and hypoxic stress and utilization of alternate carbon sources ( data not shown ) implying defects in yet to be uncovered mechanism/s which attribute to yeast survival in THP-1 macrophages . Notably , the cytokine response of THP-1 macrophages to Cgyps1–11Δ infection was also similar to that upon wt infection ( data not shown ) . Cellular metabolism is regulated by environmental cues and the mechanisms regulating carbon and energy metabolism are tuned to sense and rapidly adapt to the varied environments . As reported previously [16]–[17] , an immediate response of fungal cells to the internal milieu of macrophage is a wholesale reprogramming of carbon metabolism with elevated gluconeogenesis , β-oxidation of fatty acids and glyoxylate cycle . Our global transcript profiling analysis on 10 h macrophage-internalized C . glabrata cells reveals that fatty acids remained the main intracellular carbon and energy source in the macrophage milieu and yeast cells utilize the acetyl-CoA produced in fatty acid oxidation via glyoxylate cycle to generate energy and intermediates for synthesis of cellular building blocks ( Figures 4 and S8 ) . Further , since macrophages are postulated to be low-iron environment [33] , the concerted down-regulation of genes involved in ergosterol biosynthesis , and high-affinity iron uptake and homeostasis raise the possibility of macrophage milieu being a hypoxic environment . In yeast , glucose acts as a source of free energy and its complete break-down to carbon dioxide and water results in the synthesis of ATP , the major energy currency molecule of the cell via glycolysis and aerobic respiration [34] . Reprogramming of bioenergetic pathways towards glucose metabolism is a pre-requisite for cell growth and proliferation . One of the most striking findings of our study is a significant overlap in the genes expression profiles of the Cgrsc3-aΔ and Cgrtt109Δ mutants , which displayed reduced proliferation and altered epigenetic modifications in THP-1 macrophages . Notably , CgRSC3-A and CgRTT109 code for a DNA binding protein and a histone acetyltransferase , respectively , and are implicated in maintenance of the chromatin architecture . Chromatin remodeling , the alteration of chromatin localization and structure , has long been associated with the regulation of eukaryotic gene expression [21] , [35] A nucleosome which consists of 146 bp DNA and an octamer of histone proteins ( histone 2A , histone 2B , histone 3 , and histone 4 ) is the fundamental unit of chromatin [21] . Covalent post-translational modifications including acetylation , methylation , and phosphorylation of the N- and C-terminal tails of histones contribute strongly to the structural organization of chromatin [35] . Chromatin structure is pivotal for the regulation of gene expression with a compact chromatin limiting the genes' accessibility to transcription factors . Of the regulatory posttranslational modifications of histone proteins , role of methylation and acetylation in the regulation of gene expression has extensively been studied [21] , [35] . Histone acetylation status is usually maintained by a dynamic equilibrium between the activity of histone acetyltransferases ( HAT ) and histone deacetylases ( HDAC ) . The ‘repressed state’ of chromatin is generally equated with strict nucleosome positioning and elevated HDAC activity . Acetylation of histones at specific residues precedes the onset of transcription which is accompanied by the loosening of nucleosome positioning [21] , [35] . Disruption of CgRSC3-A and CgRTT109 genes led to a marked redirection of the metabolism and many energy metabolism-related genes were found to be differentially expressed in the microarray experiments . Global expression studies of RPMI-grown and macrophage-ingested Cgrsc3-aΔ and Cgrtt109Δ cells revealed down-regulation of genes involved in mitochondrial respiration under normal growth conditions and induction of genes required for the generation of precursors of metabolites and energy upon macrophage internalization implicating chromatin organization in cellular energy homeostasis . A significant overlap between transcript expression profiles of internalized mutant cells , their impaired respiratory metabolism and an inability of several chromatin organization defective mutants to utilize alternative carbon sources suggest that chromatin remodeling may act as a link between metabolic adaptation and survival in macrophages . This notion is in accord with the STM screen findings wherein 12% of the identified mutants with reduced survival in THP-1 macrophages harboured Tn7 insertions in genes implicated in chromatin organization . Collectively , these data establish a pivotal role for the globally altered chromatin architecture in the metabolic adaptation and replication of C . glabrata cells in the macrophage milieu . Of the 212 genes showing a 2-fold or greater change in the transcript abundance , upon macrophage internalization , in C . glabrata wt cells , we could identify the CgRsc3 binding site ( Figure S9F ) in 59% ( 153 ) of the genes via in-silico analysis indicating a central role for CgRsc3 in the regulation of gene expression . Among 153 , 61 genes were up-regulated and 92 were down-regulated . Additionally , 5% , 4% , 3% and 3% of the differentially expressed genes in macrophage-internalized yeast were found to be under the control of CgMsn2 , CgSte12 , CgYap1 and CgSfp1 transcriptional factors , respectively ( data not shown ) . ATP-dependent remodeling of chromatin is critical for alterations in gene expression and regulation of several physiological processes . Our data suggest that post 6 and 12 h macrophage internalization , chromatin in C . glabrata cells is in the ‘heterochromatic state’ with elevated repressive histone methylation and diminished euchromatic acetylation marks ( Figure 2 ) . Our data also indicate that C . glabrata displays a similar epigenetic response under glucose-depletion conditions where sodium acetate is the sole carbon source ( Figure S10E ) . Based on our phenotypic , microarray and biochemical analyses , we propose that C . glabrata cells respond to THP-1 macrophage internal milieu in three distinct phases: an Early- , a Mid- and a Late-phase ( Figure 7 ) . In the Early-phase ( 0–2 h ) , soon after phagocytosis , C . glabrata cells encounter oxidative stress and activate DNA repair and DNA damage signaling as characterized by elevated ROS levels and an increase in the phosphorylation of H2A at serine-129 residue and the number of γ-h2AX foci ( Figures 2 , S1C , and 3D ) . Metabolically , C . glabrata cells shuts down translational machinery , down regulate glycolysis and up-regulate glyoxylate and citrate cycle ( data not shown ) presumably mimicking a cellular response to carbon starvation . The chromatin of Early-phase C . glabrata cells is probably still in active conformation and exhibits sensitivity to MNase digestion ( Figure 2A ) . During this early-phase , in response to C . glabrata infection , THP-1 macrophages induce production of ROS and a modest increase in the secretion of IL-4 ( Figures 1C and S1D ) , however , they are unable to nullify the C . glabrata-mediated prevention of phagosome acidification . In the Mid-phase ( 3–12 h ) , internalized C . glabrata cells , with a carbon metabolic profile similar to that of the early-phase , down-regulate the expression of genes implicated in iron and ergosterol biosynthesis and alter their chromatin architecture to the closed form ( Figure 7 ) . The chromatin of Mid-phase C . glabrata cells is resistant to MNase digestion ( Figure 2A ) and enriched for repressive H3 and H4 methylation marks with diminished H3 and H4 acetylation modifications at some specific residues ( Figures 2B and S4 ) . The Mid-phase may reflect the adaptive response of C . glabrata to macrophage environment which is characterized by presence of alternate carbon sources , high levels of ROS , and antimicrobial peptides . The Late-phase ( 13–24 h ) epitomizes the proliferating C . glabrata cells that have survived and adapted to the macrophage internal milieu ( Figure 7 ) . For the growth , cells must acquire nutrients from the host cell and activate the cell-growth related pathways . One of probable mechanisms that C . glabrata cells employ to proliferate in a macrophage cell is to restructure the chromatin back to open conformation which is consistent with the MNase sensitivity ( Figure S3 ) and the histone modification patterns indicative of an active transcriptional machinery ( reversal of major repressive heterochromatin marks and a modest increase in histone acetylation ) of the chromatin of 24 h macrophage-ingested cells ( Figure 2B ) . This remodeling of chromatin during survival and/or replication in macrophages may be important for survival of macrophage-elicited ROS-induced DNA damage , rewiring of transcriptional networks to adapt to the nutrient-poor intracellular milieu , activation of stress signalling cascades and maintenance of cellular energy homeostasis and cell wall integrity . A detailed study on the effects of changes in the chromatin structure of C . glabrata cells on key cellular processes such as DNA replication , silencing , transcription , and stress response will yield insights into the host antimicrobial response and fungal survival strategies . Lastly , nutritional cue-dependent reversible acetylation of major metabolic enzymes plays a pivotal role in metabolic adaptation [36] . Our data demonstrate that the total acetylation status of cellular proteins is reduced in macrophage-internalized C . glabrata cells which may imply a mechanism to globally regulate the activity of several key proteins at post-translational level . Although further experiments are necessary , it is plausible that deacetylation of cellular proteins including histones , in response to macrophage environment , serves three purposes; modulates the activity of metabolic enzymes , generates an acetate pool for mobilization to the generation of energy and other essential nutrients through acetyl CoA production , and modifies the chromatin to a closed , inactive form to suppress transcription , regulate cell cycle progression , and protect against DNA damage . Consistent with this , transcriptional activation of the genes , CgACS1 and CgACS2 , coding for acetyl-CoA synthetases was observed in response to macrophage environment . However , it remains to be investigated whether the acetate group removed by the deacetylases enters the metabolic pathways . In conclusion , we report for the first time that chromatin remodeling contribute largely to the capability of C . glabrata cells to survive , function , and replicate in macrophage milieu by maintaining energy homeostasis and genes implicated in chromatin organization are required for its virulence .
Experiments involving mice were conducted at VIMTA Labs Limited , Hyderabad in strict accordance with the guidelines of The Committee for the Purpose of Control and Supervision of Experiments on Animals ( CPCSEA ) , Government of India . The protocol was approved by the Institutional Animal Ethics Committee ( IAEC ) of the Vimta Labs Ltd . ( IAEC protocol approval number: PCD/OS/05 ) . Procedures used in this protocol were designed to minimize suffering . The human ( macrophage-like ) monocyte cell line THP-1 ( ATCC TIB202 ) was maintained in RPMI-1640 medium supplemented with 10% heat-inactivated FBS and 2 mM L-glutamine at 37°C under 5% CO2 . Briefly , 24-well plates were seeded with 1×106 THP1 cells and THP-1 cells were differentiated into macrophages for 16 h in the presence of 16 nM PMA followed by 12 h recovery period . For infection assays , 50 µl of overnight grown , 0 . 1 OD600 normalized , PBS-washed C . glabrata cell suspensions was added to PMA-activated THP-1 cells . A range of multiplicity of infection ( MOI = 1∶1 , 1∶5 , and 1∶10 ) was initially used to standardize the conditions for co-incubation and infection assay . At 2 h post-incubation , infected THP-1 cells were washed thrice with PBS to remove the non-phagocytosed yeast cells . At different times post-ingestion , PBS-washed THP-1 cells were lysed in water and the number of recovered yeast was determined by plate counts of suitable dilutions ( CFU assay ) . Yeast grown in RPMI+serum , the culture medium for THP-1 cells , were used as in vitro control . The strains and plasmids used are listed in table S3 . Procedures for ROS measurement and fluorescence microscopy are described in Supplemental Methods ( Text S1 ) . YPD-grown cultures ( 0 . 05 OD600 ) of each mutant pool ( 96 mutants , each carrying a unique signature tag ) were either inoculated in YPD medium ( input ) or were used to infect differentiated THP-1 cells ( 1×106 ) . After 2 h incubation , non-cell-associated yeast were removed by washing THP-1 cells with PBS . At 24 h post infection , THP-1 macrophages were lysed in water and the recovered yeast cells were used to infect THP-1 cells at a MOI of 1∶10 . Three rounds of macrophage infection for each pool were carried out to enrich for the desired mutants in the final population . The lysate of 3rd round infection was inoculated in YPD medium for overnight ( output ) . Cells were harvested , genomic DNA isolated from each input and output cell pellet and unique signature tags were PCR-amplified with P32 labelled α-dCTP using primers for the invariant region flanking each tag sequence . Labelled PCR products were denatured at 95°C for 10 min , chilled on ice and were hybridized to membrane filters , on which plasmids carrying the 96 unique tags were immobilized ( detailed in Supplemental Methods ( Text S1 ) ) , for 14–16 h at 42°C . The filters were washed twice and exposed to phosphorimager screen for 2–4 h . The counts for each spot were quantified using Image Quant and Fuji Multi Gauge V3 . 0 software . Relative percentage intensity for individual spot was calculated with respect to all the spots present on one membrane for both input and output hybridizations images . Nucleosomal-associated DNA was extracted from RPMI-grown and macrophage-internalized C . glabrata cells using EZ Nucleosomal DNA prep kit ( ZYMO Research ) . Infected macrophages were lysed in ice cold water , lysate centrifugated at 4000 rpm , 4°C for 5 min and internalized yeast were collected . Yeast cells were suspended in 50 µl of protein extraction buffer ( 320 mM ( NH4 ) 2SO4 , 200 mM Tris-Cl ( pH 8 ) , 20 mM EDTA ( pH 8 ) , 10 mM EGTA ( pH 8 ) , 5 mM MgCl2 , 1 mM DTT , 10% glycerol and protease inhibitors ) and disrupted using glass beads followed by centrifugation at 16000×g , 4°C for 15 min . 30 µg of total protein was resolved on a 15% SDS PAGE gel and immono-blotted with antibodies against mammalian histones/histone modifications ( table S4 ) . CgGapdh was used as a loading control . To exclude the possibility of any contribution of THP-1 proteins to the cell extracts prepared from recovered yeast , we performed three control experiments . First , probing the blots with antibodies specific for mammalian tubulin and actin yielded no signal . Second , proteinase-K treatment of lysates collected from infected macrophages , prior to the yeast pellet disruption , did not alter the epigenetic signature of C . glabrata cells . Lastly , probing of cell extracts , prepared from THP-1 macrophages and untreated and proteinase-K-digested C . glabrata cells recovered either after growth in RPMI-medium or from THP-1 macrophages , with an antibody raised against human nuclear matrix protein SATB1 gave no appreciable signal in any of the C . glabrata cell extracts suggesting that proteins extracted from intracellular yeasts are devoid of mammalian nuclear matrix proteins . Notably , SATB1 antibody did recognize a band of ∼ 100 kDa corresponding to the SATB1 protein of THP-1 nuclei . C . glabrata cells grown either in RPMI or harvested from THP-1 macrophages were collected , washed with DEPC treated water and were disrupted with glass beads in trizol . Total RNA was isolated using acid phenol extraction method and frozen at −80°C . The frozen RNA samples were sent to Ocimum Biosolutions Ltd . , Hyderabad ( http://www . ocimumbio . com ) . A 4×44 K GE Agilent array comprised of 10 , 408 probes representing 5 , 205 ORFs of C . glabrata , was used wherein average number of replicates for each probe was four to five . Feature Extraction software version 10 . 7 . 3 . 1 . ( Agilent ) and Quantile normalization was used for data analysis ( Text S2 ) . Hierarchical clustering was performed using Complete Linkage method , with Euclidean Distance as distance measure . Data is the average of two hybridizations from biological replicates for each sample and raw and normalized data sets for this study are available at http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE38953 . The protocol for qRT-PCR is described in Supplemental Methods ( Text S1 ) and the primers used are listed in table S5 . Experiments involving mice were conducted at VIMTA Labs Limited , Hyderabad ( www . vimta . com ) . Infection procedure is described in Supplemental Methods ( Text S1 ) . 20 µg protein samples , isolated from RPMI-grown and macrophage-internalized yeast , were taken to measure KDAC activity using HDAC Fluorimetric Assay/Drug Discovery Kit ( Enzo Life Science ) . | Hospital-acquired fungal infections pose a colossal health and economic challenge . Candida species are the leading cause of disseminated fungal infections and rank fourth among the most common nosocomial pathogens . C . glabrata , an emerging opportunistic fungal pathogen , is the second most frequently isolated Candida species after C . albicans from Intensive Care Unit patients world-wide . Limited information is available on the unique strategies that C . glabrata employs to evade and replicate in host phagocytic cells since it lacks the key virulence traits of C . albicans including hyphal formation and secreted proteolytic activity . In the current study , we have identified a total of 56 genes , via a functional genomics approach , which are required for survival and/or replication of C . glabrata in human macrophages . Our data demonstrates an essential role for chromatin remodeling in the intracellular survival of C . glabrata with ingested C . glabrata cells displaying transcriptionally active chromatin in early-phase , compact , closed chromatin in mid-stage , and open chromatin in the late-stage of macrophage internalization . Our findings identify novel fungal virulence determinants and potentially implicate epigenetic changes in the metabolic adaptation of fungal cells to the nutrient-poor host environment and the survival against oxidative stress-induced DNA damage . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome",
"expression",
"analysis",
"microbial",
"metabolism",
"functional",
"genomics",
"microbiology",
"host-pathogen",
"interaction",
"gene",
"function",
"emerging",
"infectious",
"diseases",
"epigenetics",
"molecular",
"genetics",
"microbial",
"growth",
"and",
"developm... | 2012 | Functional Genomic Analysis of Candida glabrata-Macrophage Interaction: Role of Chromatin Remodeling in Virulence |
The regulatory architecture of breast cancer is extraordinarily complex and gene misregulation can occur at many levels , with transcriptional malfunction being a major cause . This dysfunctional process typically involves additional regulatory modulators including DNA methylation . Thus , the interplay between transcription factor ( TF ) binding and DNA methylation are two components of a cancer regulatory interactome presumed to display correlated signals . As proof of concept , we performed a systematic motif-based in silico analysis to infer all potential TFs that are involved in breast cancer prognosis through an association with DNA methylation changes . Using breast cancer DNA methylation and clinical data derived from The Cancer Genome Atlas ( TCGA ) , we carried out a systematic inference of TFs whose misregulation underlie different clinical subtypes of breast cancer . Our analysis identified TFs known to be associated with clinical outcomes of p53 and ER ( estrogen receptor ) subtypes of breast cancer , while also predicting new TFs that may also be involved . Furthermore , our results suggest that misregulation in breast cancer can be caused by the binding of alternative factors to the binding sites of TFs whose activity has been ablated . Overall , this study provides a comprehensive analysis that links DNA methylation to TF binding to patient prognosis .
DNA methylation is a critical regulatory process that involves direct chemical modification of genetic material via the addition of a methyl moiety to the 5th carbon of Cytosine nucleotides . These covalent modifications occur most prevalently on CpG dinucleotides ( CpGs ) and are reversible , thus allowing the DNA methylome to achieve a balance of stability and plasticity . DNA methylation plays essential roles in X-chromosome inactivation [1] , genomic imprinting [2] , transposable elements silencing [3] , stem cell differentiation [1 , 4–6] , embryonic development [7 , 8] , and inflammation [9 , 10] . Considering these critical roles , aberrant DNA methylation patterning has been observed in nearly all cancer types and in a plethora of non-cancer diseases including autoimmune disorders [11 , 12] , neurological diseases [11 , 13] metabolic disorders [14] , and cardiovascular disease [15] . Furthermore , DNA methylation signatures and markers have been used to stratify cancer subtypes and predict patient prognosis [16–18] . Recently , the use of DNA methylation profiling to predict prognostic outcomes of diseased patients has gained popularity . In breast cancer , studies have shown that ER+ and ER- breast cancer cell lines could be distinguished by examining their DNA methylation patterns . Sun et al . identified 84 genes that were differentially methylated between ER+ and ER- cell lines [19] . Additionally , the TCGA consortium clustered 802 primary breast cancer samples based on their DNA methylation signals; this yielded 5 distinct clusters that comprised samples that exhibited varying molecular phenotypes [20] . In a recent study , Anjum et al . identified a BRCA1 mutation-associated DNA methylation signature in 144 case-control primary blood samples that was predictive of breast cancer incidence and patient prognosis [21] . Furthermore , Bullinger et al . applied a MALDI-TOF-MS based methylation analysis to identify a DNA methylation signature in 182 acute myeloid leukemia primary samples that was predictive of patient outcomes [22] . Several other studies have identified DNA methylation signatures and markers in primary breast tumor samples that were shown to predict patient outcome [23–26] . These studies have shown that understanding DNA methylation patterning and dissecting its functions provide valuable insight into its regulatory roles , which may ultimately introduce new avenues for developing efficacious breast cancer treatments . Despite recent focus on epigenetic based markers , the exact mechanism ( s ) by which DNA methylation regulates gene expression has yet to be elucidated but its interaction with transcription factors ( TFs ) have been shown to be a critical mechanism [27–31] . It has been suggested that 5-methyl-CpGs ( 5meCpGs ) physically impede the binding of TFs to their cognate sequences causing gene silencing [27] . Additionally , 5-meCpGs can indirectly control gene expression by modulating local chromatin structure via recruitment of histone remodeling factors such as histone deacetylases and histone methyltransferases [32–34] . Physical obstruction of TF binding and compaction of chromatin structure suggest that DNA methylation exerts a silencing effect; however , studies have shown TFs such as SP1 can bind 5-meCpGs and induce gene expression [35] . In a comprehensive and systematic genomics study , Hu et al . applied a protein microarray-based approach to identify which of 1321 TFs and 210 co-factors have the capacity to bind motifs containing methylated CpGs [36] . They reported 41 TFs and 6 TF co-factors that bound 5meCpG in a sequence-specific manner [36] . Indeed , it is possible that DNA methylation patterns are just passive markers of TF binding or gene regulation whereby CpGs in unbound chromatin are methylated and have no functional relevance [37 , 38] . Regardless of the exact mechanism , we postulate that changes in CpG signals near putative transcription factor binding sites ( TFBSs ) can reflect the activity of TFs and can be used to infer the underlying transcriptional machineries that drive the progression of several subgroups of breast cancer . We have previously integrated ENCODE [39] and TCGA [20] data to computationally examine the association between TF binding and DNA methylation levels in TFBSs ( i . e . ERα ) [31] . We found that there is a strong negative correlation between ERα activity and DNA methylation levels within ERα binding sites in breast cancer [31] . More importantly , differentially methylated CpGs between ER+ and ER- breast cancer are enriched in the DNA regions surrounding ERα binding peaks ( determined by ChIP-seq ) in a distance dependent manner—the closer to the center of binding peaks , the more differentially methylated the CpGs tend to be [31] . Conversely , given a set of differentially methylated CpG sites between ER+ and ER- samples , we would expect the binding site motif of ERα or other functionally related TFs to be enriched within the vicinity of these CpG sites . These findings suggest that DNA methylation patterns and their signals are informative for exploring transcriptional regulation mediated by TFs [14] . In this study , we aimed to utilize DNA methylation data derived from primary breast cancer samples to identify TFs that are associated with patient survival via their relationship with methylated CpGs . To achieve this we connected the DNA methylation-TF interactome to breast cancer patient survival using datasets derived from ENCODE and TCGA [20] . Specifically , we identified a list of CpG sites that were significantly correlated with patient survival time in their methylation level . We then determined which TF binding sites are enriched in DNA regions surrounding survival-associated CpGs to extrapolate TFs that are associated with patient survival . Interestingly , we ascertained that ERα TF binding motifs were significantly enriched in survival-associated CpG regions in ER- samples only , and p53 TF binding motifs were enriched in survival-associated CpGs regions in p53- samples only . Overall , our analysis framework demonstrates the intimate linkage between DNA methylation , TF binding , and breast cancer patient prognosis .
The ultimate goal of our analysis was to identify TFs that impact breast cancer patient prognosis via an association with CpG methylation . By identifying TF binding motifs enriched in regions containing differentially methylated CpGs or survival-associated CpGs , we were able to demonstrate a relationship between TF-DNA methylation mediated regulation and overall patient survival rates . Fig 1 depicts our integrated approach to dissecting epigenetic involvement in transcriptional regulation underlying breast cancer patient survival . First , we investigated TF binding motifs enriched in differentially methylated CpG regions to demonstrate that methylation data are informative for inferring transcriptional regulation in breast cancer ( Fig 1: Top ) . We successfully detected the enrichment of ERα TF binding motifs in DNA regions surrounding CpG sites that were differentially methylated between ER+ and ER- breast cancer patients . Most importantly , we applied motif enrichment analysis to survival-associated CpGs ( Fig 1: Bottom ) using all breast cancer samples and subsets of samples stratified based on histological , intrinsic , and CpG beta-value intensity phenotypes . We first identified a set of survival-associated CpGs by correlating the methylation levels of each CpG across TCGA breast cancer patient samples using a univariate Cox proportional hazards model [40] . Second , we defined a 102 bp genomic region centered at each CpG ( henceforth referred to as a CpG region ) and computationally searched for the presence of TF binding motifs in this region . Third , we systematically calculated TF binding motif enrichment in these CpG regions in different subtypes of breast cancer . To preliminarily demonstrate that TF-mediated transcriptional regulation could be inferred from DNA methylation signals , we investigated the relationship between CpGs with altered methylation levels and the presence of putative TF binding motifs vicinal to these CpGs . Specifically , we identified CpGs that were differentially methylated between estrogen receptor positive ( ER+ ) and estrogen receptor negative ( ER- ) breast cancer samples in TCGA data using a Student’s t-test , and then examined the occurrences of putative TF binding motifs within DNA regions surrounding these CpGs . We systematically calculated the enrichment levels of 703 TF binding motifs available from the TRANSFAC and JASPAR databases in differentially methylated CpG regions . Our analysis identified 60 TF binding motifs ( 38 TFs ) enriched in and 105 TF binding motifs ( 67 TFs ) depleted in hypomethylated ( ER+ < ER- ) CpG regions at a P<1E-15 significance threshold ( Fig 2A ) ( adjusted p-value using the Benjamini-Hochberg multiple testing correction method; hereafter , all reported p-values have been adjusted unless otherwise indicated ) . In addition , we identified 12 TF binding motifs ( 10 TFs ) enriched in and 50 TF binding motifs ( 35 TFs ) depleted in hypermethylated ( ER+ > ER- ) CpG regions at the same threshold ( Fig 2A ) . Similar results were obtained when a Wilcoxon ranked sum test was used to identify differentially expressed CpGs ( S1 Table ) . To validate the accuracy of our systematic analysis , we directed our attention to ERα TF binding motifs . Since stratifying breast cancer patient samples into ER+ and ER- groups is analogous to controlling for ERα activity , we hypothesized that ERα TF binding motifs would be significantly enriched in hypomethylated ( ER+ < ER- ) CpG regions or , alternatively , depleted in hypermethylated ( ER+ > ER- ) CpG regions . Our data set includes three ERα TF binding motifs: JA-ESR1 , TR-ER_Q6 , and TR-ER_Q6_02 for which we calculated their enrichment in the two CpG sets ( hypo- and hypermethylated sets ) . First , JA-ESR1 was depleted in hypermethylated CpG regions 0 . 93-fold ( P = 7 . 7E-4 ) and depleted in hypomethylated CpG regions 0 . 94-fold ( P = 0 . 005 ) ( Fig 2 ) . Second , we observed that the TR-ER_Q6 motif was also depleted in hypermethylated CpG regions 0 . 84-fold ( P = 3 . 1E-7 ) , but unlike JA-ESR1 , was enriched 1 . 11-fold in hypomethylated CpG regions ( P = 0 . 003 ) ( Fig 2B ) . Lastly , the TR-ER_Q6_02 motif was depleted in hypermethylated CpG regions 0 . 92-fold ( P = 0 . 006 ) and enriched in hypomethylated CpG regions 1 . 27-fold ( P = 3 . 8E-16 ) . These results indicate that TF binding activity can be inferred based on enrichment of their motifs in DNA regions near informative CpGs ( e . g . differentially methylated CpGs ) . Previously , we had established that DNA methylation within ERα binding sites was anti-correlated with ESR1 expression by integrating TCGA ( gene expression and DNA methylation ) and ENCODE ChIP-seq data [14] . Therefore , the depletion of all 3 ERα TF binding motifs in hypermethylated CpG regions is in accordance with our previous analysis confirming that ERα activity is associated with loss of binding site-specific DNA methylation [41] . In addition to ERα TF binding motifs , we also identified a number of other TF binding motifs that are known to be associated with ERα . Strikingly , hypomethylated CpG regions contained 18 TF binding motifs ( corresponding to 5 FOX family transcription factors , GATA1 , HFH8 , XFD2 , and XFD3 ) that exhibited greater than 2-fold enrichment , whereas hypermethylated CpG regions contained none ( Fig 2A ) . Operating under the passive model [42] , this suggests that loss of methylation is generally associated with enhanced binding activity of these transcription factors in ER+ breast cancer . In addition , we identified all GATA3 and FOXO1 TF binding motifs to be enriched in hypomethylated CpG regions and depleted in hypermethylated regions suggesting that these TFs are associated with ERα activity . Indeed , it has been experimentally shown that FOXA1 influences ERα function by modulating ER-chromatin interactions and FOXA1 deficiency results in loss of ERα activity ( S1 Table ) [43–45] . In addition , GATA3 has been shown to be necessary for estradiol stimulation of breast cancer cells and more recently , modulate ERα access to enhancer regions [46 , 47] . Overall , our motif enrichment results in differentially methylated CpG regions results are consistent with the known biological roles of our identified TFs . We were additionally able to verify several enriched TF binding motifs via a de novo motif search in hyper- and hypomethylated CpG regions . Specifically , we applied the Discriminative Regular Expression Motif Elicitation ( DREME ) algorithm to identify enriched DNA motifs in hyper- and hypomethylated CpG regions and then matched them to known motifs ( See Materials and Methods ) . We were able to detect the presence of the ESR1 motif in hypomethylated CpG regions but not in the hypermethylated CpG regions , which is consistent with the enrichment results for ERα . We were also able to confirm motifs including FOXO1 , SP1 , KLF4 , EGR2 , and E2F1 in hypomethylated CpG regions and TCF3 , NHLH1 , and HEB in hypermethylated CpG regions . After identifying enriched TF binding motifs in differentially methylated CpG regions , our next objective was to determine if TFs associated with patient survival could be inferred based on DNA methylation signals . Aberrant TF activity and DNA methylation changes have both been known to play a role in carcinogenesis and cancer progression . However , to our knowledge , there has been no other study that has systematically investigated survival-associated TF-DNA methylation relationships at the level of specific TF-CpG interaction . Thus , to proceed with this high-resolution analysis , we pinpointed CpGs whose methylation levels were significantly associated with breast cancer patient survival and calculated the enrichment of TF binding motifs in the regions surrounding these CpGs . CpGs with hazard ratios <1 were categorized as protective , CpGs with hazard ratio >1 were categorized as hazardous , and pooled protective and hazardous CpGs were simply categorized as survival-associated . We hypothesized that survival-associated fluctuations in CpG methylation intensities would be informative to the activity of specific survival-associated TFs . When survival analysis was implemented using all samples , we were able to identify 92 TF binding motifs ( 62 TFs ) enriched and 143 TF binding motifs ( 98 TFs ) depleted in protective CpG regions at significance level P<0 . 01 ( S2 Table ) . For hazardous CpG regions , we detected 11 TF binding motifs ( 9 TFs ) enriched and 2 TF binding motifs ( 2 TFs ) depleted at the same threshold , respectively . Fig 3 highlights four examples of TF binding motifs enriched in survival-associated CpG regions: p53 , ERα , HEB , and LRF . First , JA-ESR1 exhibited an enrichment score of 1 . 23 at P = 0 . 004 in hazardous CpG regions ( Fig 3A ) . This indicates that the effect ERα binding activity—it is known that ER status is a significant clinical factor for predicting survival of breast cancer patients—has on patient survival can be inferred from DNA methylation signals correlated with patient prognosis . Second , LRF is an oncogenic transcription factor involved in cell growth and differentiation , and is known to be overexpressed in breast cancer [48] . Our analysis shows that the TR-LRF_Q2 TF binding motif is enriched 1 . 18 times in hazardous CpG regions ( P = 0 . 005 ) ; additionally , it is depleted 0 . 93-fold in protective CpG regions ( P = 0 . 03 ) ( Fig 3A ) . In one study , Maeda et al . reported that LRF is necessary for embryonic fibroblast cells ( MEFs ) to undergo transformation even when other potent oncogenes such as H-Ras , T-antigen , and MYC are expressed [49] . Third , we identified HEB ( TCF12 ) to be enriched 1 . 36-fold in protective CpG regions ( P = 1 . 5E-29 , Fig 3B ) . HEB has been previously reported to correlate with colorectal cancer metastasis by inhibiting E-cadherin , thus verifying HEB as a potential oncofactor [50] . Fourth , it is well established that loss of p53 activity is detected in approximately 50% of all cancers [51] . In our analysis we were able to identify all p53 TF binding motifs to be enriched in survival-associated CpGs ( P<0 . 05 ) . To highlight , the TR-P53_02 motif was enriched 1 . 28-fold in protective CpG regions ( P = 6 . 8E-5 ) ( Fig 3A ) . This again indicates that DNA methylation levels provide information about the activity of key transcriptional regulators . To provide a brief summary , we show the top ten significant TF binding motifs enriched or depleted in protective CpG regions in Table 1 . Similarly , we illustrate the top ten TF binding motifs enriched or depleted in hazardous CpG regions in Table 2 . To demonstrate that survival-associated CpGs are informative for identifying clinically relevant TFs , we focus on two key breast cancer-related proteins: ERα and p53 . ERα and p53 are major proteins whose expression levels are typically measured in breast cancer cases to determine the molecular status of the tumor , and it is well-established practice to use this information for determining prognosis and treatment strategies . Here , we explored two major subtyping schemata by systematically calculating the enrichment/depletion of all TF binding motifs ( in particular ERα and p53 ) in survival-associated CpG regions for ER+ , ER- , p53+ , and p53- breast cancer subtypes ( S3–S6 Tables ) . First , we directed our focus to ERα TF binding in ER-stratified samples and were able to identify JA-ESR1 to be enriched 1 . 44-fold in protective CpG regions in the ER- subtype but not in the ER+ samples ( P = 2 . 6E-5 , Fig 4A ) . In fact , in ER+ samples , JA-ESR1 is significantly depleted 0 . 80-fold in protective CpG regions ( P = 2 . 2E-11 ) ( Fig 4A ) . When protective and hazardous CpGs are combined , JA-ESR1 is enriched 1 . 20-fold in survival-associated CpGs in ER- ( P = 0 . 001 ) samples only and depleted 0 . 83-fold ( P = 9 . 2E-9 ) in survival-associated CpGs in ER+ ( Fig 4A ) . Furthermore , TR-ER_Q6 was enriched 1 . 24-fold ( P = 0 . 04 ) in hazardous CpG regions and TR-ER_Q6_02 was enriched 1 . 18-fold ( P = 0 . 05 ) in survival-associated CpG regions ( protective and hazardous CpG regions combined ) in ER- samples . Next , we stratified patient samples into p53+ and p53- groups and calculated the enrichment of TF binding motifs with particular focus on p53 TF binding motifs ( Fig 4B ) . We identified JA-TP53 to be enriched 1 . 27-fold in protective CpGs regions only in p53- samples ( P = 0 . 02 ) and enriched 1 . 24-fold in survival-associated CpG regions ( P = 0 . 005 ) ( Fig 4B ) . Likewise , the TR-P53_01 motif was enriched 1 . 48-fold in protective CpG regions ( P = 3 . 8E-5 ) and 1 . 38-fold in survival-associated CpG regions ( P = 5 . 13E-6 ) in p53- samples only ( Fig 4B ) . In contrast to the other p53 motifs , TR-P53_02 was enriched 1 . 29-fold in p53+ samples ( P = 0 . 006 ) and 1 . 24-fold in p53- samples ( P = 0 . 05 ) , both in protective CpG regions ( Fig 4B ) . Taken together , our enrichment results from ER- and p53-stratified breast cancer patients sample show that the majority of ERα and p53 TF binding motifs are enriched in CpG regions in ER- and p53- samples , respectively . Presumably , the binding sites of key transcriptional regulators can become unbound and accessible to other factors once the activity of the key TF is ablated . Alternative factors may then bind to these open sites leading to misregulation of the associated genes , and contribute to cancer progression and clinical outcomes of patients . Breast cancer , like most other cancer types , exhibits a high degree of heterogeneity making it refractory to treatment . One approach to abrogate the effects of sample-to-sample variation is to classify tumors into subtypes , each with distinct genetic , molecular , and physiological features . Therefore , we aimed to resolve whether breast cancer subtypes determined by immunohistochemistry also exhibit differences in TF binding motif enrichment near survival-associated CpGs . First , we calculated TF binding motifs enriched/depleted in survival-associated CpGs in each histological subtype of breast cancer ( S3 , S4 , S7–S12 Tables ) . In summary , there are a total of 252 ( 178 TFs ) , 135 ( 85 TFs ) , and 247 ( 168 TFs ) TF binding motifs that are enriched in protective , hazardous , and survival-associated CpG regions , respectively in at least one subtype at significance level P<0 . 01 . In the opposite direction , 323 ( 217 TFs ) , 49 ( 41 ) , and 305 ( 208 TFs ) motifs were depleted in protective , hazardous , and survival-associated CpG regions , respectively in at least one subtype at the same significance level ( See S13 Table for more details ) . The large number of identified motifs suggests that a variety of TFs contribute to breast cancer development and each TFs activity may or may not be important drivers depending on the subtype . Second , we clustered the p-values ( P<0 . 05 ) of significantly enriched or depleted TF binding motifs in survival-associated CpGs and observed that PR+ and ER+ , which clustered together , exhibited enrichment patterns much different from that of the other subtypes ( Fig 5A ) . More specifically , these subtypes are enriched in TF binding sites that are depleted in the other subtypes and vice versa . This suggests that the TFs associated with survival in PR+ and ER+ samples may not be significant protein factors in the other subtypes . In addition , it is clear that significantly enriched/depleted TF binding motifs vary from subtype to subtype implying that each subtype exhibits distinct TF-DNA methylation interactions . This shows that unique enrichment signatures can differentiate between breast cancer subtypes by revealing transcriptional regulators most likely to exhibit altered activity . After showing global TF binding motif enrichment patterns of histological subtypes , we provide an example where the TR-NF1_Q6 motif is enriched 1 . 76-fold in protective CpG regions ( P = 1 . 1E-9 ) in the triple-negative subtype ( Fig 5B ) . Mutations in NF-1 have been implicated in the proliferation of triple-negative primary breast cancer tumors since it functions as an inhibitor of RAS and mTOR [52–54] . This suggests that DNA methylation within NF-1 binding sites is associated with longer survival times in patients with triple-negative breast cancer . To determine if different transcriptional regulators could also be identified in breast cancer subtypes based on molecular features , we classified our samples into 5 distinct intrinsic subtypes: luminal A , luminal B , HER2-enriched , and basal [55] . In some cases , intrinsic subtyping is more representative of the underlying molecular architecture in breast cancer and can be used to predict risk of cancer relapse after treatment with chemotherapy [56] . In our analysis , we first identified CpGs that were correlated with survival for each intrinsic subtype and determined which of 704 TF binding motifs were enriched in hazardous or protective CpG regions ( S14–S16 Tables ) . In summary , a total of 9 ( 6 TFs ) , 209 ( 80 TFs ) , 113 ( 62 TFs ) motifs were significantly enriched in protective , hazardous , and survival-associated CpGs , respectively ( in at least one subtype P<0 . 01 ) . Furthermore , 21 ( 16 ) , 31 ( 27 ) , and 40 ( 34 ) motifs were significantly depleted in protective , hazardous , and survival-associated CpGs , respectively ( in at least one subtype P<0 . 01 ) ( See S13 Table for more details ) . Second , we clustered the enrichment p-values of significant TF binding motifs ( P<0 . 05 ) in each intrinsic subtype and noticed that the luminal A subtype contained the largest number of significantly enriched/depleted TF binding motifs that yielded P<0 . 01 ( Fig 5C ) . Conversely , HER2-enriched samples contained no significant TF binding site enriched or depleted in survival-associated CpG regions . This disparity is most likely due to differences in statistical power resulting from unequal subtype sample sizes and/or longer average patient survival times associated with different subtypes ( Fig 5C ) . Despite this , it is clear that some enriched/depleted TF binding motifs are shared amongst luminal A , luminal B , and basal samples and some are not . Overall , this demonstrates global variation in TF binding site enrichment across intrinsic breast cancer subtypes . To explore individual TF binding motifs that are enriched in an intrinsic subtype , we illustrate JA-ELK1 as an example . JA-ELK1 is enriched 1 . 78-fold in hazardous CpG regions in the basal subtype ( P = 0 . 002 ) ( Fig 5D ) . ELK1 has been shown to be involved in up-regulation of Mcl-1 , a p53 inhibitor , and may contribute to survival of breast cancer cell lines [57] . Additionally , genome-wide studies in breast cancer cell lines have revealed that ELK1 is involved in the activation of c-Fos , a proto-oncogene that is implicated in tumorigenesis [58] . These studies verify that many TF binding motifs we find to be enriched in breast cancer subtypes are biologically meaningful in the context of breast cancer . When analyzing TF-DNA methylation relationships in breast cancer subtypes , we build upon conventional methods of cancer stratification . However , in order to analyze TF motif enrichment within a classification scheme focused on DNA methylation , we adopted a bottom-up approach by first classifying all CpGs into subtypes based on their intensity levels . Since many cancers show genome-wide changes in DNA methylation , this approach may be able to identify TFs that are directly related to distinct intensity levels of DNA methylation . Therefore , we created a class of subtypes based on the clustering of CpG β-values and calculated TF binding motif enrichment in these subtypes . Fig 6A shows CpGs organized into 5 clusters based on β-values , with high intensity clusters on top and low intensity clusters on the bottom . From C1 to C5 , the clusters are enriched in 68 ( 50 TFs ) , 45 ( 31 TFs ) , 6 ( 6 TFs ) , 6 ( 5 TFs ) , and 87 ( 59 TFs ) TF binding motifs , respectively ( P<0 . 05 ) ( Fig 6A ) . Furthermore , we identified 119 ( 80 TFs ) , 38 ( 24 TFs ) , 3 ( 3 TFs ) , 1 ( 1 TFs ) , and 10 ( 8 TFs ) TF binding motifs that were significantly depleted from C1 to C5 , respectively ( P<0 . 05 ) ( S18 Table ) . Like histological and intrinsic subtypes of breast cancer , certain TF binding motifs exhibit different levels of enrichment across CpG subtypes . To globally illustrate the variation in TF motif enrichment between CpG subtypes , we sorted significant motifs in cluster 1 ( C1 ) ( P<0 . 01 ) from most enriched to most depleted ( Fig 6B ) . We then ordered the TFs in the other 4 clusters relative to those belonging to cluster 1 ( Fig 6B ) . From this , it is clear that related clusters share common patterns of enrichment ( i . e . patterns in cluster 1 are more similar to that of cluster 2 than cluster 5 ) ( Fig 6B ) . Interestingly , cluster C1 , which contains highly methylated CpGs , is both enriched and depleted in TF binding motifs ( Fig 6A and 6B ) . In contrast , cluster C5 , which contains lowly methylated CpGs , is characterized mainly by TF binding motif enrichment events and few TF binding motif depletion events . This suggests that TF binding is generally associated with reduced methylation levels . Additionally , clusters C3 and C4 contain very few high-significance enriched/depleted TF binding motifs , suggesting that mid-intensity methylation are stochastic events and are not as informative for identifying important breast cancer-associated regulators . To provide an example , we illustrate TR-NFY_01 , which shows highest enrichment in C5 and lowest enrichment in C1 ( Fig 6C ) . It can also be observed that its enrichment level increases from C1 to C5 ( Fig 6C ) . This suggests that these CpG clusters have functional relevance in the context of NF-Y binding . NFY is known to be essential for proper cell cycle regulation and mutation of this protein can lead to inhibition of Cyclin A , RNR R2 , DNA polymerase , CDC2 , Cyclin B , and CDC25C [59] . Moreover , Agostino et al . showed that NF-Y facilitates gain-of-function p53 mutant binding to NF-Y promoters , resulting in cell cycle misregulation in breast cancer cell lines [60] . We also highlight TR_E47_01 , which exhibited highest enrichment levels in C1 and lower enrichment levels in clusters least similar to C1 , suggesting that E47 binding sites tend to be highly methylated in breast cancer ( Fig 6D ) . E47 ( also known was TCF3 ) is a repressor of E-cadherin and its activity has been implicated in epithelial-mesenchymal transition events in breast cancer [61] . In order to demonstrate differences in the regulatory interactomes of breast cancer subtypes , we constructed two TF-TF interaction networks for ER+ and ER- samples ( see Materials and Methods ) . Each network illustrates the first order partners of TFs whose motifs are significantly enriched ( depletion is excluded ) ( P<0 . 01 ) in ER+ and ER- samples ( Fig 7 ) . Interestingly , in ER- breast cancer , ESR1 ( ERα ) , RELA , SP1 , and AR exhibit the highest degree in the network ( Fig 7 ) . Consistent with our prior results , it can be observed that ESR1 is significantly enriched in protective and hazardous CpGs in the ER- network only ( Fig 7 ) . In addition to ESR1 , SP1 also exhibits high-degree in both ER+ and ER- networks; however , it is enriched in hazardous CpG regions in ER+ whereas , in ER- , it is enriched in protective CpG regions ( Fig 7 ) . This demonstrates that TF-DNA methylation relationships vary depending on disease context .
The effects of DNA methylation are widespread and vary according to genomic context and interactions with TFs . In this study , we proposed a novel method of inferring TF-DNA methylation relationships in breast cancer by utilizing both differential methylation and survival analysis to pinpoint informative CpGs . From these CpGs we were able to delineate TF involvement with methylation patterns and extend that to patient prognosis . Many mechanisms by which DNA methylation interacts with TFs have been proposed . It has been suggested that methylated CpGs can act as a direct physical hindrance to TF binding and thus interfere with its regulatory functions . Additionally , DNA methylation can recruit chromatin remodelers ( or proteins that then recruit chromatin remodelers ) to compact chromatin , assist in transcription elongation , or merely act as a passive marker of protein binding . This variety of potential mechanisms is key to understanding why the TF binding motifs of some experimentally verified oncofactors ( oncogenic TFs ) in our analysis were enriched in protective regions and vice versa for tumor suppressor motifs . For example , if methylated CpGs in the binding site of an oncofactor obstructs binding , then hypermethylation will ablate the oncofactor’s regulatory effect and promote survival . Alternatively , it is also possible for tumor suppressor TF binding motifs to be enriched in protective CpG regions if the role of the tumor suppressor is to recruit DNA methyltransferases to silence oncogenes . Additionally , the genomic location of CpGs may have an effect on its regulatory activity; methylated CpGs in promoters may inhibit gene expression but methylated CpGs in gene bodies may aid in transcriptional elongation [27] . Therefore , genomic context and prior knowledge of the TFs relationship with methylated CpGs must be established before reasonable conclusions can be made . As more experimental data is generated regarding these relationships , a TF-DNA methylation interactome network will be of greater use . Here , we have provided evidence that the application of motif enrichment , survival analysis , and differential methylation analysis can be integrated and used to define TF-DNA methylation interactomes in various subtypes of breast cancer . All in all , this study links TF-binding to DNA methylation to overall patient prognosis . By embracing the complexity of misregulation in various breast cancer subtypes , it may be possible to identify key players responsible for cancer subtypes and use that information to guide the development of treatment regimens in the clinic . Furthermore , our preliminary analysis shows that correlating gene expression with survival yields very few significant genes after multiple hypotheses testing correction ( S19 Table ) . This can be due to post-transcriptional modifications that affect the stability of mRNA transcripts , the fact that mRNA abundance is not always a good proxy for protein activity , and the short lifespans of cancer patients in our datasets . In light of this , the alternative use of DNA methylation signals can reveal significant CpGs even after testing corrections . This may be due to the fact that it is a binary chemical modification that is stable and can , in some cases , better reflect regulatory activity . The most striking result of our analysis is that the majority of ERα and p53 TF binding motifs are significantly enriched in survival-associated CpG regions in ER- and p53- samples , respectively . From these results , we propose that ERα TF binding motifs that are not bound by their respective TF ( as in the case of ERα in ER-samples ) may become bound by alternative factors that may cause misregulation of downstream genes and impact patient survival . Indeed , this may also be the case for p53 in p53- samples . The reasoning for this model begins with the observation that the TF binding motifs of a master regulator are enriched in CpG regions whose methylation is correlated with survival only in samples missing that regulator . If a TF is missing , why would a significantly large proportion of its binding sites be enriched in these informative CpG regions ? Therefore , we suspect that these motifs are open to binding by alternative factors whose binding events may simultaneously cause gene misregulation and detectable alterations in DNA methylation . Additionally , an alternative explanation may be that DNA methylation in p53 or ERα binding sites passively reflects the lack of p53 or ERα activity , respectively . This is valid in the case of p53- samples , where the loss of a key tumor suppressor would result in longer survival times compared to p53+ patients . However , in the case of ERα , its exact cancer regulatory roles are not as clear and thus difficult to interpret . To explore breast cancer TF-DNA methylation relationships in depth , we adopted three different classification schemes by which we divide samples based on histological , molecular , and methylation features . Since each subtyping method utilizes information at different levels ( i . e protein , gene expression , methylation ) it is sensible to adopt all three strategies . By calculating TF binding motif enrichment in different subtypes , we can effectively determine the similarities and differences in their TF-DNA methylation signatures . We also observe differential enrichment patterns between protective and hazardous CpG sets among histological and intrinsic breast cancer subtypes , suggesting that TF-DNA methylation relationships vary across subtypes . We also complemented our analyses by calculating whether TFs whose motifs were enriched in differentially methylated or survival-associated CpG regions were also differentially expressed in mRNA levels between ER+ and ER- , and between normal and tumor patient samples , respectively . We found that many of TFs with enriched motifs were also significantly differentially expressed , suggesting a greater biological role for these TFs ( S1–S12 and S14–S17 Tables ) . Additionally we validated many of our identified TF binding motifs in an independent DNA methylation dataset published by Dedeurwaerder et al . [62] . This dataset was generated from the Illumina HumanMethylation27K ( HM27K ) array which profiled ~27 , 000 CpG sites from 248 primary breast tumors . Particularly we found that the TR-ER_Q6 and TR-ER_Q6_02 motifs were both significantly enriched ~1 . 3-fold ( unadjusted P = 0 . 02 and P = 0 . 008 , respectively ) in hypomethylated ( ER+<ER- ) CpG regions between ER+ ( n = 132 ) and ER- ( n = 101 ) samples ( S1 Table ) . Moreover , we found that GATA3 and FOXO1 TF binding motifs were enriched in hypomethylated regions ( P<0 . 05 ) , which is consistent with results from the main TCGA dataset . We then extended our validation analysis to include the enrichment calculation of TF binding motifs in survival-associated CpG regions identified across all breast tumor samples . In particular , we sought to confirm our enrichment results for the top 20 protective and top 20 hazardous TF binding motifs in the Dedeurwaerder dataset , and were able to validate 19 out of the 20 protective and 16 out of the 20 hazardous TF binding motifs ( P<0 . 05 , S2 Table ) . To note , we did not perform the CpG filtering procedure in the Dedeurwaerder dataset since the number of CpGs interrogated by the Illumina HM27K array was substantially less than the Illumina HM450K array used by TCGA , resulting in a sizeable decrease in statistical power . Together , these results indicate that our analysis remains robust across independent datasets even when different genomic platforms are used . We concede that there are limitations to the informativeness and interpretation of our results . First , we used a 102 bp region to define a CpG region , which restricts our analysis to a local binned area . Even though most binding events only encompass ~100 bp , it may be possible that this sequence space may encompass the binding sites of TFs that are not associated with the CpG residue leading to false positives . On the other hand , the 102 bp region may be too small and not encompass the binding sites of TFs that do in fact interact with the CpG resulting in false negatives in our enrichment analysis . Overall , we experimented with varying sequence region sizes and determined that our results remain stable . Second , because we restricted our analysis to a local region , we do not take into account any potential long-range effects CpG sites may have on TF binding as a result of chromatin orientation . Third , limitations in platform technology must also be taken into account since only 450 , 000 out of a total of ~30 million CpGs are probed by the Illumina HumanMethylation 450K array . Fourth , we acknowledge that there may be differences in statistical power when conducting enrichment analysis in CpG sets ( e . g . protective , hazardous ) since the number of CpGs in each set may vary . Lastly , many of the analysis steps in our methodology including motif detection and setting significance criteria for survival-associated CpGs suffer from high false positive rates . However , we were able to overcome these obstacles by calculating relative enrichment of TF binding motifs . Therefore , even though we were able to identify significantly enriched TF binding motifs , our enrichment scores were ultimately biased towards the null ( RE = 1 ) . Overall , we maintain that our method has produced results that provide new insight into TF-DNA methylation relationships in breast cancer despite these limitations . In this study , we have developed a novel method for identifying transcriptional regulators involved in breast cancer in the context of patient survival by using DNA methylation data derived from primary breast tumor tissue . By doing so , we have provided insight into the complexity of TF-DNA methylation interactomes that underlie breast cancer across a wide variety of subtypes . Our analysis has revealed several informative results and , in addition , raises a manifold of new questions regarding cancer misregulation . Namely , we have identified transcriptional regulators that affect patient prognosis and proposed a new model whereby breast carcinogenesis may be driven via binding of alternative factors to unbound TFBSs . Additionally , we considered the heterogeneity exhibited by breast cancer tumors by characterizing TF-DNA methylation relationships in histological , molecular , and DNA methylation subtypes . In this analysis , we focused on well-defined TF binding motifs , but it is also possible to combine this analysis with de novo motif identification to identify novel motifs that are enriched in differentially methylated or survival-associated CpG regions . Such integration could also allow for an exhaustive and systematic identification of non-TF regulators that may also interact with methylated CpGs ( e . g . non-coding RNAs ) . Ultimately , our study has provided deep insight into the differential regulatory wiring of breast cancers that occur due to the divergent and combinatorial effects of diverse mutations .
Breast invasive carcinoma ( BRCA ) Level 3 DNA methylation datasets for the JHU-USC HumanMethylation450K platform , CpG annotation files , and clinical information were downloaded from the TCGA data portal [20] . CpG methylation signal intensities were represented as β-values in the datasets . In addition , Level 3 TCGA UNC AgilentG4502A_07 mRNA expression data was downloaded from the site data portal [20] . Subtype classification for all patient samples was derived from TCGA clinical information [20] . Breast cancer DNA methylation data and clinical information from Dedeurwaerder et al . was downloaded from the Gene Expression Omnibus ( GEO ) under the accession number GSE20713 . PWMs for human TFs were obtained from the TRANSFAC [63] and JASPAR [64] databases . In some cases there were multiple PWMs for a single TF . The TF-TF physical interaction data were compiled from two resources: an experimental dataset from Ravsi et al [65] containing 5238 TF-TF physical protein interactions across 1400 human TFs , and the human protein reference database [66] . Patient samples that were accompanied with histological information regarding ER status were split into ER+ ( 405 samples ) and ER- ( 122 samples ) groups . Differentially methylated CpGs were then identified using two-tailed student t-test . To achieve stringency while maintaining power for enrichment analysis , a Benjamini-Hochberg adjusted P<0 . 05 was chosen as the cutoff for differentially methylated CpGs . CpGs with p-values below the cutoff with a t-statistic >0 and <0 were categorized into hypermethylated ( ER+>ER- ) and hypomethylated ( ER+<ER- ) sets , respectively . A Wilcoxon ranked sum test was also implemented to identify differentially methylated CpG sites . RNA-seq data for 1154 breast cancers was downloaded from the TCGA data portal . Differentially expressed genes were identified using a Wilcoxon ranked sum test . A fold change >1 indicated gene up-regulation and a fold change <1 indicated down-regulation . Differentially expressed genes corresponding to TFs were included in all supplementary tables to complement TF enrichment information . The β-values for 376 , 667 CpGs across 562 samples and clinical data were used as input into a univariate Cox proportional hazards model [40] . Each CpG was considered individually and used as the covariate in the model . Significance of model coefficients was calculated using the Wald test . CpGs that yielded unadjusted P<0 . 02 and a hazard ratio of <1 or >1 were labeled as protective and hazardous , respectively . Samples were categorized into histological and intrinsic subtypes based on the clinical information downloaded from TCGA . The β-values for 376 , 667 CpGs across subtype-only samples were used as input into the Cox proportional hazards model where each CpG was considered individually . This allowed for the identification of survival-associated CpGs significant in the particular subtype . The 102 bp sequence region centered at each significant CpG was used as input into the FIMO software package [67] from the MEME suite to identify the existence of a motif in the region . Default parameters were used and a threshold cutoff of P<1E-4 was used to determine the presence of a motif . This yielded a matrix containing Boolean values indicating if a particular TFBS motif was present in a CpG region . The top 10 , 000 most significant differentially methylated CpG regions ( hyper- or hypo- ) were chosen as input into the Discriminative Regular Expression Motif Elicitation ( DREME ) algorithm ( MEME suite ) using default parameters , with the exception of the maximum motif size , which was set to 20 [68] . Identified motifs were then queried against the JASPAR vertebrate database using Tomtom ( MEME suite ) to identify their cognate TFs [69] . Because overlapping CpG regions can lead to over-counting of TF binding motifs , we filtered out all overlapping CpG regions in each chromosome for forward and reverse DNA strands . The filtering procedure proceeds as such: ( i ) CpGs located on different chromosomes were considered non-overlapping . ( ii ) CpGs that were located on different DNA strands were considered non-overlapping . ( iii ) Sort all CpGs based on their genomic coordinates and identify clusters of CpGs with overlapping 102bp regions . ( iv ) For each cluster , identify the CpG with the lowest p-value ( differentially methylated or survival-associated depending on analysis ) and set this as the reference CpG . ( v ) Identify all within-cluster CpGs whose regions do not overlap with that of the initial reference CpG and filter out the rest . ( vi ) Of the non-overlapping CpGs , select the one with lowest P-value and set as the new reference CpG ( vii ) Iterate until all CpGs have either been selected or filtered out . ( All “reference” CpGs are then included in the subsequent motif enrichment analysis . ) To compute enrichment of TFs in functional CpGs ( survival-associated , differentially methylated , or clustered CpGs ) , we applied a two-sided Fisher’s exact test for each TF binding motif to determine if it was overrepresented or underrepresented in a CpG set . The Fisher’s exact test involves calculating the hypergeometric probabilities of all possible matrices of a 2X2 contingency table while keeping the margin counts fixed . The probabilities of all possible fixed-margin contingency tables more extreme than the current table were summed to determine the probability of over-representation/under-representation of a motif to occur by random chance . The R function “fisher . test” was used to implement this computationally . Enrichment was calculated in protective , hazardous , and survival-associated ( protective & hazardous ) CpG sets for histological , intrinsic , and CpG ( CpG clusters ) breast cancer subtypes . Additionally , motif enrichment was implemented in differentially methylated CpGs between ER+ and ER- breast cancers . The Benjamini-Hochberg multiple hypothesis testing correction procedure [70] was used to adjust the P-values outputted by multiple Fisher’s exact tests . All P-values presented in the Results section had been adjusted for multiple testing . When comparing the distributions of TF binding motif enrichment values between hyper- and hypomethylated CpGs ( Fig 2A ) , we first controlled for the potential effects that sample size may have on the power of enrichment analysis . This issue may arise due to the unequal number of CpGs belonging to hyper- and hypomethylated CpG sets . Therefore , we took the top n most significant CpGs from each set , where n is the smallest number of CpGs between the two sets , and carried out enrichment analysis . TF binding motifs with P<0 . 01 in ER+ and ER- samples , and their first-order interacting partners were extracted from the TF-TF physical interaction network and used as input into Cytoscape to construct a regulatory network . TF network analysis was implemented using the “NetworkAnalyzer” function included in the software . The size of network nodes were mapped to the enrichment P-values of TFs represented in the network ( lower P-values correspond to larger nodes ) . The font sizes of TF names were mapped to node degree in the network ( larger font sizes correspond to higher degree ) . These mappings were implemented using Cytoscape’s VizMapper tools . If multiple motifs belonging to the same transcription factor fell below the significance threshold , their p-values were averaged . K-means clustering was applied to cluster CpGs based on their β-values . To determine the number of clusters , k-means was applied using 1–10 clusters and the total within-cluster sum of squares ( WCSS ) was calculated and graphed . Classification into 5 clusters yielded the last point where there is noticeable decrease in total WCSS . All statistical and computational analyses were implemented in the R statistical programming environment . | DNA methylation is a ubiquitous and simple covalent modification that occurs directly on genetic material whereby a simple methyl group ( CH3 ) is attached to Cytosine nucleotides in the context of CpG sites . Modifications of these sites have been postulated to function in gene regulation , potentially via interactions with transcription factors . In this study , we hypothesized that DNA methylation signals contain valuable information that can help infer transcription factors that may be associated with a given disease . Here , we utilize the vast repository of breast cancer data that is available in the public domain , and which contains a rich resource for DNA methylation and clinical data on breast cancer patients . In this guilt-by-association analysis , we postulated that conserved transcription factor binding motifs that are statistically enriched in regions near methylated CpG sites that are correlated with breast cancer patient survival would suggest that their cognate transcription factors would play a role in the initiation , growth , metastasis , or even suppression of the tumor . This integrative approach supports the claim that DNA methylation profiling of patient tumors in the clinic may contain valuable information that can guide the development of treatment regimens for individual patients; thus contributing to the progression of precision medicine . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Regulators Associated with Clinical Outcomes Revealed by DNA Methylation Data in Breast Cancer |
In this work we present a method for the differential analysis of gene co-expression networks and apply this method to look for large-scale transcriptional changes in aging . We derived synonymous gene co-expression networks from AGEMAP expression data for 16-month-old and 24-month-old mice . We identified a number of functional gene groups that change co-expression with age . Among these changing groups we found a trend towards declining correlation with age . In particular , we identified a modular ( as opposed to uniform ) decline in general correlation with age . We identified potential transcriptional mechanisms that may aid in modular correlation decline . We found that computationally identified targets of the NF-ΚB transcription factor decrease expression correlation with age . Finally , we found that genes that are prone to declining co-expression tend to be co-located on the chromosome . Our results conclude that there is a modular decline in co-expression with age in mice . They also indicate that factors relating to both chromosome domains and specific transcription factors may contribute to the decline .
Since the introduction of DNA microarrays over a decade ago , it has become possible to use genome-wide approaches to explore differences between two biological conditions , such as tumor versus healthy samples , mutant versus wild-type cells or old versus young tissues . The most common type of analysis , called differential expression analysis , looks for genes whose expression changes between two or more different groups . In addition to individual genes , differential expression analysis can also identify groups of genes or pathways that change expression levels in an experiment . For example , pathway analysis shows that genes involved in the electron transport chain show a general decrease in expression with age , even though individual genes in this pathway may not show a large effect [1] . A different type of analysis , differential network analysis , is to create a genome-wide network of genes , and then to look for changes that occur in the network . Gene co-expression networks give insight into how genes work together in particular pathways or systems across multiple microarray conditions . Because most biological processes arise from the complex interactions among multiple gene products , information about how genes function together can improve our understanding of the underlying biological mechanisms . For instance , Hughes et al . used DNA microarrays to profile expression of every gene in yeast in 300 different mutants and chemical treatments , and then calculated which genes were co-expressed with each other under these diverse conditions . This work clearly showed that genes could be grouped into cellular pathways based on co-expression , and provided a useful approach to categorize the function of unknown genes on a global scale [2] . Since this finding , gene co-expression networks have been constructed using worm , fly , mouse , and human microarray data [3]–[6] . In addition , the comparison of co-expression links between orthologous genes in multiple species allows one to search for relationships that are functionally conserved [6] , [7] . Looking at how gene co-expression relationships change between two networks is a potentially powerful way to obtain a holistic view of how gene co-expression relationships change between two states . However , searching for differences in networks requires great sensitivity to the initial choice of data . For example , the absence of a shared link in mouse and human co-expression networks does not necessarily indicate divergent function . Instead , differences in the mouse and human co-expression networks may indicate differences in the technical platforms or the experimental conditions used to build the networks . In this work , we present a novel method for differential co-expression network analysis . Past research has focused on differences in co-expression between networks . Ihmels et al . developed the differential clustering algorithm ( DCA ) to identify groups of co-expressed genes that differ between yeast species [8] . Choi et al . created a nonweighted co-expression network using a collection of published cancer arrays and compared it to a network composed of the control arrays [9] . Here , we describe a comprehensive and scalable methodology for differential co-expression network analysis and apply it to search for differences in gene co-expression networks during aging in mice . Aging affects a myriad of genetic , biochemical and metabolic processes , and thus it is attractive to use a network approach to globally characterize changes in old age . Genes do not work alone , but rather act within a functional group or pathway , such as a metabolic pathway or a regulatory network . One effect of aging may be to diminish the coherence in expression of gene pathways . In old animals , some genes in a pathway may not be fully activated in tissues that require the function of the pathway , and other pathway components may not be fully repressed in tissues in which the function of the pathway is not needed . In this case , old mice would show less correlation in expression for genes in the pathway than young mice . We test this possible effect of aging by comparing co-expression networks between young and old mice ( Figure 1A ) .
We developed a difference network framework to directly identify groups of genes that change correlation with age . In the difference network framework , every node represents a gene , and every weighted edge represents the change in correlation between old and young mice for the corresponding gene pair ( Figure 1B ) . The edge weights , , are scalars such that a negative represents a decrease in correlation and a positive represents an increase in correlation: ( 1 ) where and are the Spearman correlation coefficients between genes and in the 24 month and 16 month data sets respectively , and is the Fisher transformation . The Fisher transformation , when applied to a sample correlation coefficient , yields an approximately normally distributed estimator . Though changes in negative correlations could also be of interest , this choice of allows us to focus on changes in positive correlations . The most straightforward way to evaluate the difference network is to determine whether the average is positive or negative . However , in the difference network , there are a total of approximately 36 million weighted edges , a vast majority of which involve pairs of genes that are not co-expressed and thus have values that are near zero . The small fraction of edges that reflect true co-expression difference would likely be obscured by the large number of edges that do not change with age . We reduced the complexity of the difference network problem by focusing on groups of genes rather than individual gene pairs . We used two different classification methods to define gene groups: ( a ) genes that show strong co-expression in a related set of gene array experiments and ( b ) genes that are within the same function group as defined by the Gene Ontology ( GO ) classification . The first classification method involved finding groups of genes that are co-expressed . We could not use the gene array data from either the 16- or the 24-month-old mice because this would bias the analysis . For example , a gene set defined by co-expression in the 16-month data set naturally has higher co-expression than it would in the 24-month data set , and vice versa . For this reason , we used a separate gene expression data set , the AGEMAP compilation of data from 6-month-old mice . Like the 16- and 24-month AGEMAP data , the 6-month data included 10 gene arrays from 13 different tissues . Using the gene expression data from this set of gene arrays , we calculated the correlation in expression between all pairs of genes , then used average-linkage hierarchical clustering to group the genes into clusters ( see Materials and Methods ) . For this analysis , we defined gene clusters using a distance cutoff such that all resulting clusters had an average within-cluster Spearman correlation of at least . The hierarchical clustering established a list of 312 gene sets containing a minimum of five genes . For each group ( ) , we defined the test statistic ( ) as the mean of the intergroup edge weights in the difference network . For each gene group , we used a permutation test to count how many times the permuted test statistic exceeded the observed . We considered a gene group significantly decreasing if the sample value was below the first percentile of permutations ( two-sided -value ) . We considered a gene group significantly increasing if the sample value was above the 99th percentile of permutations . From the 312 gene clusters , we found nine clusters of genes that show decreasing correlation with age and one cluster that shows increasing correlation with age ( Table 1 ) . Because we tested 312 clusters , we would expect to identify 3 . 12 clusters of each type ( increasing and decreasing ) under the null hypothesis ( Figure 2A ) . We inspected the expression pattern of the genes for the ten clusters that change co-expression with age , and noticed an interesting pattern regarding expression in the gonads and adrenal glands . The genes from two clusters ( clusters 175 and 14 ) are expressed at high levels specifically in the gonads and adrenal glands , and the genes from four clusters ( clusters 78 , 73 , 68 and 178 ) are broadly expressed except for low expression in the gonads and the adrenals ( Table 1 ) . The gonads and adrenals both produce steroid hormones . The ovaries produce estrogen and progesterone , the testes produce testosterone and the adrenal glands produce corticosteroids such as ACTH . In all cases , steroid production decreases with age . In addition to grouping genes based on co-expression , we also grouped genes based on shared genetic functions using GO categories . We used gene sets that shared the same GO molecular function , associated cellular component , or biological process . GO categories are useful in that they provide a functional grouping of related genes , and genes with similar functions are often co-expressed . We used 395 GO categories containing 5 to 200 genes with minimal overlap to test for group-wide correlation changes with age . We identified the set of GO categories by looking at the categories at every level of distinction , and discarded any group that has more than 50% overlap with any other category that is smaller than it . As before , we used permutation to test for significance of within-group edge weight change . We found that nine GO categories decrease correlation with age , and zero categories increase correlation with age ( Figure 2A ) . Table 1 lists the GO categories that change significantly with age . The top GO category includes genes involved in memory , which clearly declines with age . Another interesting GO category is selenium binding . Selenium is a trace element that acts as a cofactor for reduction of antioxidant enzymes [11] . Several studies have suggested that low levels of selenium may be a risk factor for developing cancer in humans [12] . Changes in the genes responsible for selenium binding with age would have interesting implications for the role of antioxidants in aging . The observation that gene sets tend to show an overall decline in correlation of gene expression with age suggests that there may be densely-connected subgraphs of negative edges in the gene-correlation difference network . To find areas of the difference network that loosen with age , we clustered genes using the weighted edges of the difference network as a distance matrix . For example , to find a densely-connected subgraph of the difference network where all of the edges are negative ( i . e . , decreasing with age ) , we define the distance metric between two genes and to be . Thus , a gene pair that decreases correlation with age has a small clustering distance . If we cluster using the distance matrix composed of for and , then the resulting clusters of genes are chosen based on their shared correlation loss . Similarly , if we set the distance measurement to be , then two genes that increase correlation with age are separated by only a small distance . Clustering via will yield clusters in which the members increase correlation with one another . We clustered the genes in the difference network using and to locate clusters that increase and decrease respectively . We used average linkage hierarchical clustering with a set distance cutoff of to define clusters . We set the height cutoff to so that all of the resulting clusters had a mean average correlation difference of or for and , respectively . Using a minimum size cutoff of five probes , we found 54 clusters that decreased correlation with age and 18 clusters that increased correlation with age ( Figure 2B ) . As shown in Figure 2B , by using permutations to test for significance , we found that the number of clusters identified via and are both significant . For the true number of clusters fell in the top 0 . 4 percentile ( corresponding to two-tailed ) . For the true number of clusters far exceeded any of the permuted values . Although both clusterings are significant , there are more clusters that decrease in correlation than increase in correlation with age ( Figure S1 ) . From the difference network clusters obtained using and , we located interesting groups of genes that change correlation with age . Table 2 and Table 3 describe the clusters we found using and , respectively . Among the clusters that decreased correlation with age are many gene groups previously implicated in aging pathways , such as mitochondrial function , transcriptional regulation , and ribosome biogenesis . One cluster , enriched for DNA-damage genes , shows increasing correlation with age . Because DNA damage increases with age [13] , it is possible that DNA-damage pathways are more frequently triggered in old age , producing a more coordinated transcriptional response . One possibility is that there may be a uniform loosening of gene edges throughout the difference network . For example , there may be age-related damage to basic components of gene expression that are used for all genes , such as RNA polymerase II , or there may be damage to chromatin-modification enzymes . If such proteins become damaged in old age , expression of all genes may be affected and may result in decreased levels of gene correlation . Another possibility is that certain areas of the correlation difference network may show much greater loss of edge strength than other areas . This difference would appear modular , with entire groups of genes losing correlation relative to one another . For example , aging may affect a specific DNA-binding transcription factor , in which case the downstream target genes of the transcription factor would show large age-related losses in gene correlation . We investigated whether there was uniform or modular loosening of the correlation network with age using two parameters in the unweighted gene co-expression networks: the connectivity and the clustering coefficient . In an unweighted network , the connectivity is defined as the number of neighbors of a given gene . The clustering coefficient measures the degree to which genes cluster together ( Materials and Methods ) . The clustering coefficient ranges from zero ( none of a gene's neighbors is connected to any other ) to one ( all of the neighbors are connected to one another ) . Figure 3 shows a plot of the clustering coefficient against the connectivity for the 16- and the 24-month-old gene-correlation networks . We found that there are fewer genes in the upper right side of the plot in the older network than the younger network , implying that large interconnected gene groups tend to be lost as mice age . We developed simulation tests to determine whether the differences between young and old mice networks could be explained by uniform or modular loosening of gene expression edges . We then compared the observed data to each of the simulations to determine which showed the greatest resemblance . We simulated uniform loosening of the network using a node-based deletion . In the node based simulation we randomly selected nodes in the 16-month-old network and deleted all edges leading out of those nodes . We continued to delete edges until the number of edges in the simulated network was equal to the number of edges in the 24-month-old network . We repeated the simulation 100 times , and each time we drew the boundaries of the simulated networks on the scatter plot of the clustering coefficient versus the connectivity ( see Materials and Methods ) . As shown in Figure 4A , the boundaries of the simulated networks do not resemble the boundary for the 24-month-old network . This result indicates that loss of gene expression correlation in 24-month-old mice does not occur uniformly across the network . We also modeled networks in which the loss of gene co-epression during aging was modular . We simulated modular correlation loss using a cluster-based deletion , removing co-expressed clusters from the network derived from the young mice . First , we clustered the genes by co-expression in the 16-month-old network . We defined clusters using average linkage hierarchical clustering , with a distance metric of 1- , where is the Spearman correlation between genes and . Distinct clusters are formed by cutting the tree at a particular height ( see Materials and Methods ) . Next , we deleted all of the clusters from the 16-month-old network using a number of different values of . Figure S2 shows the simulation results for various height cutoffs . This figure shows that the correlation loss observed from young to old mice is consistent with removing clusters defined at a of ( Figure 4B ) . These results show that aging gene networks appear to loosen in a modular fashion . One possible mechanism of modular loosening of a gene expression cluster with age is if all of the genes in a cluster are targets of a specific transcription factor . If a transcription factor loses the ability to co-regulate a group of genes with age , we can expect to see a decline in correlation in expression between those genes as the animals age . To this end , we searched for transcription factors whose targets changed co-expression with each other between young and old mice . Transcription factors contain DNA binding domains that attach to a specific sequence of DNA . The Transfac database lists known sequence motifs to which a transcription factor binds . By identifying all genes that contain a conserved Transfac motif in their upstream regions , we obtain an estimate of genes that may be targets of a particular transcription factor ( See Materials and Methods ) . We downloaded binding information for 258 known conserved transcription factor binding sites from the Transfac database . From all of the transcription factor binding sites , we created 163 gene groups classified according to the presence of a conserved transcription factor binding site within 5000 bp upstream of the translation start site . We only used gene groups that contained five or more unique targets . To each group , we looked at the mean ( Equation 1 ) and assessed significance using the permutation method described in Materials and Methods . We found five transcription factors whose downstream targets decreased correlation with age , where are expected by chance ( Table 4 , Figure 5 ) . For each of these significant sets of genes , there is a subset of genes that strongly lose correlation in expression relative to one another with age . We did not find any transcription factors whose targets increased correlation in expression with age . Three of the most significant groups contain genes with binding sites for NF-B , AP2 , and MEF-2 . NF-B is involved in cellular inflammation . NF-B has a myriad of inducers , such as reactive oxygen species ( ROS ) , infection , and cytokines [14] , [15] . All of these factors increase with time and thus have been implicated in aging . The amounts of ROS , a by-product of cellular metabolism , has clearly been shown to increase in old animals . Perhaps because of the increase in ROS , NF-B is abnormally activated in the major lymphoid organs [16] . AP2 is involved in a variety of processes , including morphogenesis and development . Its involvement with aging primarily stems from its regulation of the aging-associated human helicase protein WRN [17] . The targets of MEF-2 also appear to lose correlation in expression with age . MEF2 is a muscle-specific transcription factor that has been shown to increase binding affinity with oxidative stress in human primary skeletal muscle cells [18] . Another mechanism that could account for the unevenness in the correlation loss of the gene co-expression network in old mice is changes in specific chromatin domains . Chromosomes have regions of open chromatin ( which are accessible to transcription factors and permit gene expression ) and closed chromatin ( which does not allow transcription factor binding and prevents gene expression ) . Domains that are open in young mice may become less open in old mice , and domains that are closed in young mice may become partially open in old mice . If so , genes that are fully expressed from open chromatin domains in young mice may become partially repressed in old mice , and genes that are not expressed in young mice because they are in closed chromatin domains may become partially derepressed in old mice . The net effect of such a loss of regulation could be to show lower correlation levels in old age . In this case , we would expect genes that tend to lose correlation with other genes in old age to be clustered together on the chromosome . We tested for a chromatin domain effect by determining whether genes that lose expression correlation with other genes in old age tend to be clustered together or randomly dispersed . We defined neighbors as the genes in the difference network that show a decreasing correlation in expression with the target gene above a set threshold . The number of such neighbors represents an age-related correlation loss score ( Figure 6A ) . Thus a gene that has a high age-related correlation loss score is a gene that loses correlation with many of the genes that it was previously correlated with at 16 months . We scanned the genome with a moving window and counted the number of windows that have two or more genes with a correlation loss score above the threshold . Using a threshold set at six genes and a window size of 80 kb , we identified 44 windows with two or more genes that lose expression correlation with age ( Figure 6B ) . To determine whether this number is statistically significant , we repeatedly scrambled the locations of the genes and recalculated the number of clusters . The results from 1000 permutations are presented in Figure S3A , which shows that the observed number of windows is greater than the number found by random permutation in all but two cases ( ) . This result indicates that genes that lose correlation in expression with their neighbors in old age tend to be clustered on the chromosome . We repeated this analysis using a number of different thresholds ( two to six genes ) and window sizes ( 10 to 200 kb ) . We found similar results for a range of parameters ( Figure S3B ) , including some even more significant than our original choices . Thus , genes that are sensitive to loss of regulation of expression with age occur in specific regions in the chromosome , perhaps because these regions correspond to chromatin domains that are affected by aging ( Table S1 ) .
Here we present a methodology to compare two biological states ( young versus old mice ) by performing a global comparison of changes in gene co-expression . There is an important difference between comparing changes in expression levels and comparing co-expression relationships . Traditional analysis focuses on finding genes or groups of genes whose expression levels differ between two states . On the other hand , differential co-expression analysis looks for changes in the co-expression relationships between genes . By comparing how the correlation in gene expression differs between two states , we can make inferences about changes in functional interconnectedness of those genes . Comparing network relationships is not a novel concept , however most such comparisons focus on finding similarities . For example , co-expression networks have been constructed for multiple species by identifying genes that show conserved co-expression with each other among large numbers of DNA microarray experiments [6] , [7] . Numerous algorithms have also been proposed to find similarities among different types of biological networks . For example , Walhout et al . combined co-expression data with protein-protein interaction and phenotypic data to obtain information about functional gene interactions in the Caenorhabditis elegans germline [19] . Similar approaches that integrate multiple high-throughput data types have been created for various microbes , yeast , worms , and humans [20]–[31] . The above approaches have successfully been used to pinpoint similarities between networks . Searching for differences is a more nuanced problem . In addition to our method , two previous studies have looked at differences in networks [8] , [9] . There are several key differences between our algorithm and the previous ones . Our method assigns a statistical significance to the changes in the gene clusters , it uses weighted networks and it allows for the unsupervised identification of changing clusters . Although the previous two algorithms were able to achieve many of these criteria , neither met all of them . When looking for network similarities , less attention can be given to the composition of the data from which the networks are constructed because similarities in differently constructed networks are likely to be biologically relevant . For example a similarity between an edge in the fly and worm gene co-expression network is likely to indicate a shared functional link between two genes . In contrast , when looking for network differences , more attention needs to be given to the input data so that the comparison shows biological differences rather than artifacts that reflect the manner in which the data were collected . A divergence in that network may be due to a trivial difference in the types of experiments being performed , the experimental platform , the lab that performed the experiment or the experimental design . For example , the differential clustering algorithm ( DCA ) identifies groups of genes that are co-expressed in one yeast species ( C albicans ) but not another ( S . cerevisiae ) , or vice versa [8] . However , the input expression data for the two yeast species are not closely matched . Thus , it is possible that some of the differences observed between S . cerevisiae and C . albicans arise from a bias in experiment selection rather than intrinsic differences in biological properties . Bias in the experimental input can be controlled by carefully matching the arrays for each condition . Choi et al . used this method to compare a non-weighted gene co-expression network from cancer samples to a similarly constructed network from normal tissues [9] . By matching each tumor type to the corresponding control from normal tissues , they minimized the potential for experimental bias in the construction of the cancer and normal networks . The AGEMAP data set provides a unique opportunity to create matching networks as the data set from the 16- month old mice is matched to the data set from the 24- month old mice . The mice were raised in the same facility , the data were collected by one lab using one experimental platform and identical experimental protocols . Since the only consistent difference between the 16- and 24-month old data sets is the age of the mice , it is more likely that the differences in gene correlations between the two networks reflect the effects of aging . We chose to use a network approach to compare young versus old mice because aging is a complex process involving the cumulative effects of many different genetic pathways in diverse tissues . Efforts to understand the underlying molecular basis of aging are often thwarted by the complexity of the aging process . DNA microarray analysis is well suited to aging research because it allows for simultaneous measurement of gene expression outputs from nearly all genes in the genome in parallel . However , by focusing on changes in expression of genes , most traditional differential analyses neglect the interactions in expression between the genes . For example , the AGEMAP publication described changes in expression levels between young and old mice for the 16 tissues separately [10] . Gene set enrichment analysis ( GSEA ) was also used to find groups of genes whose expression increased or decreased levels with age . This type of analysis successfully identified many pathways that were previously associated with aging , as well as many novel age-associated pathways . However , there is little overlap between the gene sets found in the AGEMAP paper based on changes in gene expression and the gene sets found in this work based on changes in co-expression interactions . By finding different pathways , both differential expression analysis and co-expression analysis can complement each other to generate a more complete overview of age-related changes . We used a differential co-expression network approach to show that there are large-scale changes in gene co-expression associated with the aging process . Previous work has shown that there is an increase in the variability in expression levels in old age . Using DNA microarrays , one study showed that expression levels typically show more variability in old versus young , when comparing different samples from either human or rat tissues [32] , [33] . Another study used single-cell PCR to show that aging was marked by increased cell-to-cell variation in gene expression in mouse cardiac myoblasts [34] . However , single cell analysis of mRNA levels in a variety of blood cell types did not replicate this finding [35] . These studies all show increased transcriptional instability with age , which is consistent with our finding of a decrease in gene co-expression in old mice . The loss of correlation in gene expression could be due to several different causes . For instance , transcriptional machinery may degrade with time , such that genes show weaker activation or repression in old tissues compared to young . It could be that there are changes in tissue specificity in old age , such that pairs of genes are co-expressed in specific tissues in young mice but show more general expression across many tissues in old mice . Another possibility is that certain pathways , such as inflammation , could become constitutively induced in old age . We found two possible mechanisms that could account for loss of gene correlation in old age . The first is that old age may affect the activity of transcription factor NF-B , such that the direct targets of NF-B may show strong co-regulation in young mice but weaker co-expression in old mice . Previous work has also implicated the NF-B transcription factor with aging . NF-B is involved with inflammation , which increases with age in all tissues [36] . Adler et al . used a combination of differential expression analysis and computational identification of transcription factor targets to identify transcription factors whose targets change expression levels with age [37] . They found that both NF-B and AP2 targets increased expression with age . Looking at arrays from three different mouse tissues and six different human tissues they found that their targets were age-induced in the majority of the tissues examined . NF-B activity increases with age and controls gene expression through its interaction with the sirtuin protein SIRT6 [38] . A reduction of NF-B activity in the skin of old mice caused a reversal of their gene expression aging profile [39] . These results suggest that high or constitutive activity of NF-B in old adults could be a molecular mechanism accounting for loss of co-regulation of the NF-B target genes in old age . Another possible mechanism for loss of gene co-expression in old age is deterioration of chromatin structure . Histone modifications in chromatin are responsible for both permitting and preventing gene expression [40] . If these histone modifications were to degenerate in old age , chromatin domains would become less well-defined . Genes that are completely repressed and strongly activated in young animals would show either high basal expression or low activated expression in old age . This would result in lower levels of gene co-expression with other genes in the network . In support for a role for chromatin domains in age-regulated changes in transcription , we found that genes that lose correlation with age tend to be clustered together on the chromosome . The strength of the connections between genes is an important system-wide property of a network that can be used to compare two states . Here , we compare young to old mice , but this approach could be used to compare many other states such as healthy versus disease or wild-type versus mutant . Differential network analysis , is applicable not only to co-expression networks derived from gene expression data , but can also potentially be applied to other types of biological networks , including networks constructed from protein-protein interactions , mutant phenotypes , or from integration of many types of gene interaction experiments . Differential network analysis could potentially be used to compare a network from one species to the network from another . Finally , this approach could be used to evaluate non-biological networks , including changes in social or economic networks over time .
We downloaded the AGEMAP data from the NCBI Gene Expression Omnibus ( accession GSE9909 ) [10] . The AGEMAP microarray collection contains microarrays for 16 different tissues for five male and five female mice aged both 16 and 24 months . We removed the liver , striatum , and bone marrow samples because they were missing multiple array experiments ( i . e . , they each had less than four biological repeats for either the males or females in a single age group ) . For each remaining array , we then calculated the mean correlation coefficient with each of the other four arrays in the same tissue , sex , and age class . For example , for an array taken from the kidney of a 16-month-old male , we calculated the correlation coefficients across all genes for the remaining four male kidney samples . We calculated the mean of those four correlations to determine how well the array agrees with other arrays in the same class . Figure S4 plots the distribution of mean correlation coefficients for each array . Thereafter , we removed any arrays with a mean correlation coefficient of less than 0 . 8 . In both the young and the old data set there were 2 such arrays . We substituted any missing or removed arrays with a pseudoarray calculated from the mean of the four arrays for the missing arrays' tissue , sex , and age class . The pseudoarray keeps the number of arrays equal for both age groups , ensuring that all tissues are represented equally in the resulting data . The presence of pseudoarrays could potentially bias gene-correlation coefficients toward a higher correlation . However , because there are equal numbers of pseudoarrays in both the young and the old data sets , we found this bias to be acceptable . Next , we normalized each array by subtracting the mean expression value over all genes for that array . After normalization , we removed all probes that had a low variance and low expression in both the 16-month-old and 24-month-old data sets . We tested the effect of removing low-expressing genes by studying the correlation between probes that match to the same genes . Each array contains 12 , 273 probes , which map to 8932 unique UniGene IDs [41] . If the mapping is correct , one would expect that the expression of two probes that match to the same UniGene ID would be highly correlated . Figure S5 plots the expression correlation for each pair of probes that map to the same UniGene ID . By discarding the probes that fall below a set mean ( ) and variance ( ) , we observed that thresholding reduces the proportion of matched probes with low correlation . By increasing the mean and variance thresholds , we decreased the total number of probes; however , the mean correlation between the matched probes increased . These results indicate that as we removed the genes that have low expression in all arrays , the amount of noise due to nonexpressing genes correspondingly decreased . We ultimately chose cutoffs of and , leaving 9104 probes that exceed these thresholds . If the neighborhood represents the set of directly connected neighbors of gene , then the connectivity of that node is the number of neighbors in the neighborhood:Given a gene's neighborhood ( ) the clustering coefficient is the fraction of links between the nodes in its neighborhood over the total possible number of edges between all genes in the neighborhood:The clustering coefficient can be considered a measure of modularity in the data . Networks that have a tendency toward high clustering coefficients contain many densely connected subgraphs . We used a permutation method to determine the significance for a number of test statistics . For each one , we generated permuted data sets such that each permuted data set contained randomly sampled data from the young and the old data . Because each data set is stratified into sex , mice , and arrays , we sampled from the individual male and female mice separately . For example , for the male mice we pooled the males , young and old . We then randomly chose mice from the pool and labeled them young in the randomized data set . The remaining mice were designated as old for the randomized data . For each mouse , all of the arrays and all of the probes went into the same data set . We repeated this procedure for the female mice . We then computed the test statistic on the permuted data and repeated more times . For each permutation of the data , we recalculated the test statistic and counted the number of times the permuted test statistic exceeded the observed value . For example , let be the set of test statistics generated from the total permutations . Then the p-value is calculated as:Because of the different stratifications , we end up with possible permutations . In each randomized data set the number of young and the number of old mice is not set to be equal . For example the randomized old data set may have young females and only old female . This type of permutation , called non-balanced permutation , has been found to be more accurate than it's balanced counterpart [42] . We downloaded the AGEMAP data for the 6-month-old mice from the NCBI Gene Expression Omnibus ( accession GSE9909 ) [10] . We normalized the data using the method described for the other AGEMAP data . We found gene clusters by performing a hierarchical clustering of the 6-month data , using a Spearman correlation–based distance metric and average linkage for merging nodes . In average linkage hierarchical clustering , each gene starts as its own cluster . Pairs of clusters are then successively merged according to their average distance . In this case , we used a Spearman correlation based distance metric , , to determine the distance between any two genes . Here is the Spearman correlation between two genes calculated across all available experiments in the 6-month data set . To determine the distance between two clusters for merging , we used the average distance between all of the genes in the two clusters . Clustering of this sort provides a hierarchical tree of clusters . By cutting the tree at an average distance of , we obtained distinct clusters . We discarded any clusters containing fewer than five genes . We obtained GO , KEGG and Interpro categories from the DAVID database [43] , [44] . For GO categories we looked at gene groupings based on GO molecular function , associated cellular component , and biological process . We discarded any categories with fewer than 5 or more than 200 genes . We determined overlap between a cluster and a functional category using the hypergeometric distribution . Here is the number of genes , is the number of genes that are both in the cluster and the functional category , is the size of the cluster and is the size of the functional categories . For the node-deletion simulation , we randomly selected nodes in the 16-month-old network and deleted all edges leading out of those nodes . We iterated the edge-selection and deletion process until approximately of the edges were deleted , i . e . , when the simulated network has the same number of edges as the 24-month-old network . Because it is not possible to exactly match the number of edges in the simulated network to the number of edges in the 24-month network , we allowed to be within of , where and are the total number of nodes in the 24-month and simulated networks , respectively . We repeated the simulation 100 times , and each time we drew the boundaries of the simulated networks on the scatter plot of the clustering coefficient versus the connectivity . We drew the boundaries by dividing the plot into a grid and binning the data points for each gene . For each simulation , we then drew a boundary around the outermost edge of the bins that contained at least one data point , such that all bins outside of the drawn boundary contained zero data points . We defined gene clusters using a range of cluster thresholds , and removed the clusters from the network simulating modular co-expression loss . Clusters were assigned using average linkage hierarchical clustering , with a distance metric of 1- , where is the Spearman correlation between genes and . Distinct clusters are formed by cutting the tree at a particular height . For a given , all genes in the resulting cluster have an average distance from one another of at least . For example , if , all the clusters that result from cutting the tree at have a mean distance of ( corresponding to a mean correlation of ) . In this way , sets the stringency for inclusion into a cluster: small values of result in clusters of genes that show higher levels of co-expression . We obtained predicted transcription start sites for all human RefSeq genes from the University of California , Santa Cruz ( UCSC ) human genome assembly ( hg17 ) . We then downloaded from the Transfac Matrix Database all of the conserved transcription factor binding sites ( TFBs ) found by Hinrichs et al . to be at . The database contains 258 transcription factors conserved in human , mouse , and rat at this threshold [45] . We then located all of the conserved TFBs within 5000 bp of the transcription start site of all RefSeq genes . Thus , we obtained a list of conserved TFBs within 5000 bp of a known human gene . To map these results to the mouse genome , we used the chained alignment of the mouse genome ( mm9 ) to the human genome ( hg17 ) supplied by the UCSC genome database [46] . For every RefSeq gene in mm9 , we assigned a TFB if the site in the human genome appeared in a conserved region within 5000 bp of the mouse gene . To search for chromosomally-clustered genes , we first defined the correlation loss score of a given gene to be the connectivity of that gene within the difference network . The connectivity here is defined by the number of neighbors a gene has with for , where is the edge weight of the difference network between two genes . We then mapped probes to the mouse genome using the mm9 assembly from the UCSC Genome Browser [47] . When more than one probe mapped to the same location , we averaged the connectivities of those probes . We used a moving window approach to define gene clusters as follows . First , we identified all genes whose connectivity was greater than or equal to a threshold . Then , we scanned a window across the genome and counted the number of windows containing more than one gene ( ) with . We defined a cluster as a window containing more than one gene that met or exceeded this threshold . For a given window size and threshold , we calculated the number of gene clusters on the chromosome , then permuted the data to determine whether the rate of clustering exceeded the rate expected by chance . For every chromosome , we permuted the genes' locations and recalculated the number of gene clusters . We defined the two sided p-value as , where is the number of gene clusters in the permuted set that exceeded the original number of clusters , and where is the number of permutations . | There is mounting evidence that mammalian aging is marked by increased gene transcriptional variation . This trend was shown not only by studying gene expression in single cells ( Bahar et al . 2006 ) , but at the coarse tissue resolution as well ( Somel et al . 2006; Li et al . 2009 ) . These led us to believe that looking at absolute changes in expression level alone may not tell the whole story of transcriptional changes in age . Instead the story may be in the more subtle changes in the coordination of expression among multiple genes . For this reason , we decided to look at changes in co-expression relationships with age . To this end , we developed a methodology for differential co-expression network analysis for the comparison gene co-expression on a global scale . We applied this methodology to compare co-expression between young ( 16-month ) and old ( 24-month ) mice . This allowed us to find both gene groups whose coordination appear to be affected by age and to propose potential mechanisms for the change . We believe our work is of broad importance because it represents a different paradigm for looking not only at aging but also at any complex condition or disease—away from changes in individual genes towards changes in gene relationships . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology/aging",
"genetics",
"and",
"genomics/bioinformatics",
"genetics",
"and",
"genomics/gene",
"expression"
] | 2009 | Aging Mice Show a Decreasing Correlation of Gene Expression within Genetic Modules |
A skeletal muscle fiber that is stimulated to contract and then stretched from L1 to L2 produces more force after the initial transient decays than if it is stimulated at L2 . This behavior has been well studied experimentally , and is known as residual force enhancement . The underlying mechanism remains controversial . We hypothesized that residual force enhancement could reflect mechanical interactions between heterogeneous half-sarcomeres . To test this hypothesis , we subjected a computational model of interacting heterogeneous half-sarcomeres to the same activation and stretch protocols that produce residual force enhancement in real preparations . Following a transient period of elevated force associated with active stretching , the model predicted a slowly decaying force enhancement lasting >30 seconds after stretch . Enhancement was on the order of 13% above isometric tension at the post-stretch muscle length , which agrees well with experimental measurements . Force enhancement in the model was proportional to stretch magnitude but did not depend strongly on the velocity of stretch , also in agreement with experiments . Even small variability in the strength of half-sarcomeres ( 2 . 1% standard deviation , normally distributed ) was sufficient to produce a 5% force enhancement over isometric tension . Analysis of the model suggests that heterogeneity in half-sarcomeres leads to residual force enhancement by storing strain energy introduced during active stretch in distributions of bound cross-bridges . Complex interactions between the heterogeneous half-sarcomeres then dissipate this stored energy at a rate much slower than isolated cross-bridges would cycle . Given the variations in half-sarcomere length that have been observed in real muscle preparations and the stochastic variability inherent in all biological systems , half-sarcomere heterogeneity cannot be excluded as a contributing source of residual force enhancement .
Stretching a contracting muscle produces a dynamic , non-linear force response . Some features of the response can be explained by classical cross-bridge theory [1] , [2] , but others apparently cannot . Residual force enhancement following stretch is a phenomenon widely believed to fall into the latter category [3] , [4] . The basic experimental finding was first described by Abbott & Aubert [5] . They found that a dogfish muscle stretched during active contraction produced greater steady-state force than if stretched to the same length prior to activation . ( A stretch protocol of this kind is depicted in Figure 1 . ) This was seen as a departure from the well known steady-state length tension relationship as the muscle could be made to produce different amounts of force at a given length by changing its history of movement . Since Abbott & Aubert's initial description , force enhancement has been observed in many other species , muscle types , and experimental preparations ( reviewed elsewhere [6] ) . Researchers have been studying residual force enhancement for nearly sixty years , but the underlying mechanism remains unknown and controversial . No existing models of muscle fully explain the observed behavior , and development of a model that does could therefore improve biomechanical analyses of musculoskeletal systems . Such a model would also be helpful to the growing number of investigators who seek to use mechanical measurements on muscle tissue to elucidate the integrative function of individual sarcomeric proteins [7] , [8] , [9] . If the effects of force enhancement are present in the data sets they collect but not in the mathematical models that they use to analyze them , the results are unlikely to be interpreted correctly . Several hypotheses have been advanced to explain residual force enhancement . The potential role of titin has been highlighted by Bagni and co-workers [10] , [11] , [12] , but classically the sarcomere length heterogeneity mechanism has received the greatest amount of attention [6] , [13] , [14] . This was developed by Morgan and co-workers following experimental studies and the development of mathematical theories [13] , [15] . Their hypothesis considers a muscle with an average sarcomere length on the descending limb of the steady-state length-tension curve , and assumes that there is some amount of heterogeneity in the initial length of each sarcomere . As the muscle is stretched , longer ( weaker ) sarcomeres lengthen more rapidly than shorter ( stronger ) ones , reach a yield point , and then lengthen uncontrollably until tension is determined almost entirely by passive elements . This sudden lengthening or ‘popping’ of weak sarcomeres allows serially connected sarcomeres to stretch less than they otherwise would . They thus produce more force because they remain at a sarcomere length that is shorter than the average sarcomere in the muscle , and are therefore positioned close to the isometric plateau . The net effect is that tension developed by the complete muscle system remains elevated following active stretch . While several studies have offered evidence of sarcomere heterogeneity in muscle preparations [14] , [16] , the claim that it is the principal source of residual force following stretch remains controversial . Two experimental results have been used against the sarcomere popping hypothesis: first , that enhancement is seen at muscle lengths on the ascending and plateau portions of the length-tension curve , and second , that enhanced force after stretch can exceed the isometric force at optimal length [3] . Neither of these results are possible under the sarcomere popping hypothesis as it stands [13] . Furthermore , the notion that residual force enhancement requires multiple sarcomeres has been directly challenged by new experiments showing enhancement in a single-sarcomere myofibrillar preparation [17] . To date , most computational models of interacting heterogeneous sarcomeres have been based on the classic isometric length-tension curve and Hill-type force-velocity relations [13] , [18] . In this work , we investigate residual force enhancement using a biophysically-detailed computational model of heterogeneous half-sarcomeres that was originally developed for a different purpose [19] . Rather than empirical length-tension and force-velocity relations , this model calculates force produced by each half-sarcomere from equations representing a population of actin myosin cross-bridges cycling through a strain-dependent kinetic scheme . The simulations show enhancement after stretch as a consequence of half-sarcomere inhomogeneity , but the effect is produced by a mechanism that is distinct from that of ( half- ) sarcomeres ‘popping’ to the passive length-tension curve .
A model of mechanically-coupled half-sarcomeres was previously described by one of us [19] . Full details of the model , including equations and parameter listings can be found in that publication and its accompanying supplementary information . Unless otherwise noted , the parameters used in the current simulations were the same as those obtained in previous work [19] by fitting simulated records to experimental data recorded using chemically permeabilized rabbit psoas fibers at 15°C [20] . The model is based on equations that define the length- and time-dependent force produced by individual half-sarcomeres . The total number of myosin cross-bridges available to participate in force production ( ) within a single half-sarcomere i having length is given by the equation ( 1 ) , the proportion of myosin heads at a given half-sarcomere length that are available to produce force , is based on the length-dependent overlap of thick and thin filaments . The average number of myosin heads per unit cross-sectional area in a half-sarcomere ( ) was 1 . 15×1017 m−2 and is scaled by , a variable that is randomly selected from a Gaussian distribution with a variance of α and a mean value of unity . Unless otherwise noted , α was set to 0 . 2 for all simulations . Because this quantity cannot be directly measured in experiments , sensitivity of model results to the value of α was assessed in a series of simulations ( see below ) . The only change in Equation 1 from the previously published model is the addition of the term , which represents the fraction of actin binding sites activated by Ca2+ at time t . This term , which is identical at any given time for all half-sarcomeres in the model , improves the numerical stability of calculations by allowing non-instantaneous activations . The force produced by cross-bridges in each half-sarcomere is determined by analyzing cross-bridge population distributions as originally described by Huxley [1] . The number of attached cross-bridges is defined by a kinetic model consisting of detached ( D ) , attached pre-powerstroke ( A1 ) , and attached post-powerstroke ( A2 ) states . Transitions between states have strain-dependent kinetic rates [20] . Passive force generated by a single half-sarcomere is assumed to follow an exponential dependence on half-sarcomere length as follows: ( 2 ) The parameters , , and L were determined by fitting passive length-tension measurements from rabbit psoas preparations [21] , [22] . Values are given in Table 1 . The term is the second addition to the previously published model , and represents a normally distributed random number having a mean value of unity and variance β . It can be used to introduce variability into the passive stiffness of half-sarcomeres in the same manner as α introduces heterogeneity in active properties . β was set to zero for all simulations , except where noted . The total force produced by a half-sarcomere is the sum of passive and active ( cross-bridge based ) forces . Most simulations use 50 half-sarcomeres connected in series , with exceptions noted in the text . The length of the complete fiber is set by the user for each time-point in the simulation while the lengths of the individual half-sarcomeres are calculated by the computer by balancing the force at each connection between half sarcomeres . The simulation procedures mimicked published experimental protocols [23] . Trials were run in pairs . At the beginning of the first simulation , was linearly increased from 0 to 1 over the first second to gradually activate the half-sarcomeres . The overall length of the network of coupled half-sarcomeres was held isometric at an initial length L0 during the period of increasing activation and for 1 second afterward , at which point the system was stretched at constant velocity to the final length , Lfinal . In the companion simulation , the muscle was activated at Lfinal and held at that length thereafter . The magnitude of the force enhancement was defined as the percentage increase above the isometric force , i . e . ( 3 ) where and are the forces produced in the ‘ramp and hold’ and isometric simulations , respectively , at time after the end of stretch . Since real muscle cannot be activated indefinitely without fatiguing ( in the case of electrically-excitable fibers ) or ‘running-down’ ( in the case of chemically permeabilized preparations ) , was set equal to 6 seconds after the conclusion of length change for all values of RFE reported herein . This is at least as large as the values used in many experiments ( e . g . a tss value of 4 seconds is used in ref . [24] ) . Unless noted otherwise , the length change protocol was an 8% stretch performed at 0 . 1 L0 s-1 , starting from an average half-sarcomere length of 1200 nm . Tension traces shown in each figure are the average of 60 simulations run with different random number seeds to produce distinct patterns of half-sarcomere variation . Averaging reduced the stochastic noise introduced by random assignment of the number of myosin heads per unit cross-sectional area in each half-sarcomere .
A typical simulated force response to a ramp-and-hold length change for a single half sarcomere is shown by the blue trace in Figure 1A . There is a large tension transient at the start of stretch , after which tension falls even though the fiber is still lengthening . When the stretch ceases , tension decays almost immediately to the isometric ( steady-state ) value . When 50 dissimilar half-sarcomeres are arranged in series and stretched , the tension response is altered greatly ( red trace , Figure 1A ) . After an initial rapid rise , tension continues to increase throughout the stretch . When the stretch stops , tension begins to drop but remains elevated above the corresponding isometric tension . Variation in the number of cycling cross-bridges per half-sarcomere caused half-sarcomere lengths to diverge substantially before and after stretch , with some lengthening rapidly during the applied length change ( Figure 1C , also see Video S1 ) . The slowly dissipating phase of the tension response seen in the heterogeneous case strongly resembles the residual force enhancement observed experimentally . When stretches of different velocities but equal magnitude were imposed on the heterogeneous model ( Figure 2 ) , residual force enhancement was nearly constant over the entire range of stretch velocities , in agreement with published reports [23] , [25] . The dependence of stretch magnitude was also examined , using a range of stretch sizes ( 2–16% L0 ) , each with a stretch velocity of 0 . 1% ( Figure 3 ) . RFE was proportional to the magnitude of stretch , a feature that has also been reported for real muscle preparations [5] , [25] , [26] . Isometric simulations at many half-sarcomere lengths were performed to construct a steady-state length tension curve for default model parameters . This defined half-sarcomere length intervals for the ascending limb , plateau , and descending limb of the length-tension relation ( Figure 4b ) . The stretch protocol was then performed at average half-sarcomere lengths of 1134 and 1160 nm ( ascending limb ) , 1210 and 1296 nm ( approximate plateau ) , and 1404 and 1512 nm ( descending limb ) . A small amount of enhancement was seen on the ascending limb of the length-tension relation ( 0 . 8–3 . 5% , Figure 4a ) , compared with 2–5% observed experimentally on the ascending limb in frog muscle [27] . 8–13% enhancement is produced from stretches ending at lengths on the plateau of the length-tension curve , demonstrating that the model is capable of producing residual force enhancement that exceeds the force at optimal length ( Figure 4b ) . Among the initial lengths tested , RFE was greatest for stretches on the descending limb . The above results were obtained using α = 0 . 2 ( ±20% variation in the active tension capacity of half-sarcomeres ) . However , this level of heterogeneity is not required for the appearance of enhancement behavior . Repeating the stretch protocol over a wide range of α values shows that RFE is highly sensitive to even small amounts of variation ( Figure 5 ) , with an α value of 0 . 02 producing more than 5% RFE . Enhancement essentially saturates at α values exceeding 0 . 1 ( Figure 5 ) . Simulations were also used to test whether variability in the passive elastic properties of half-sarcomeres ( quantified by the parameter β ) produced residual force enhancement . RFE was roughly half as sensitive to β as to α ( Figure 6 ) . Under the conditions of the standard stretch protocol , the maximal enhancement produced by passive tension variation was 7 . 6% . β in combination with α did not increase RFE substantially . With α at 0 . 2 , the addition of β = 0 . 2 raised RFE only slightly , from 12 . 9% to 13 . 8% . In intact muscle , structural proteins such as desmin are thought to link z-disks of adjacent myofibrils [28] . While this wouldn't be expected to affect half-sarcomere length variability , it could stabilize the lengths of whole sarcomeres across ( and thus potentially through a ‘second-order effect’ along ) fibers . In a series of additional simulations , we tested the possibility that stabilization among coupled myofibrils could eliminate residual force enhancement . The model size was increased to 6 myofibrils in parallel , each with 50 half-sarcomeres in series . kim was increased from 0 to 0 . 1 , effectively forcing z-lines to be registered across parallel myofibrils . Using an α value of 0 . 2 , the model was stretched ( 8% L0 over 4 seconds ) and the resulting force response showed an RFE of 12 . 7% ( Figure 7 ) . A wide range of half-sarcomere lengths can be seen throughout the record ( Figure 7C ) , just as in single myofibril simulations ( Figure 1C ) . At the same time , the relative variation in sarcomere length was much smaller than the variation in half-sarcomere length , and while a few of the whole sarcomeres elongate rapidly during stretch , they do not ‘pop’ to a length dominated by passive tension as some half-sarcomeres do ( Figure 7C ) .
We have used numerical simulations to discover a novel mechanism for residual force enhancement after stretch . Enhancement was an emergent behavior of a model system that assumed variation in the mechanical properties of series-connected half-sarcomeres . Force enhancement was not produced by sarcomere popping but by dynamic interaction between dissimilar half-sarcomeres which prevented rapid dissipation of strain energy stored in populations of cycling cross-bridges . The simulations suggest that observable levels of enhancement will occur if variation in half-sarcomere properties exceeds ±2% , a condition that is hard to discount in real biological systems . Hence , without excluding other contributing mechanisms , this work suggests that dynamic interactions among dissimilar half-sarcomeres are an important component of residual force enhancement . | Textbooks often state that the force produced by a contracting muscle depends on its length . Nearly 60 years ago , it was discovered that this length-tension relationship is violated if the muscle is stretched to a given length while contracting , in which case the muscle produces more force than if it was stretched to the given length prior to contraction . This effect is known as residual force enhancement , and its mechanism remains controversial . Understanding residual force enhancement is important because it potentially affects outcomes of many in vivo and in vitro experiments where contracting muscles or muscle preparations are stretched . In this work , we use a computational model of half-sarcomeres connected in series to show that residual force enhancement is an emergent behavior of the contractile system when the half-sarcomeres are not completely identical . Force enhancement in the model shares several key properties with the phenomenon observed in real muscle , including independence from the rate of stretch and proportionality to stretch magnitude . Enhancement in the model is produced by a previously undescribed mechanism in which complex interactions between the heterogeneous half-sarcomeres dissipate strain-energy from the imposed stretch at very slow rates , creating a long-lasting , enhanced level of force . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biophysic",
"al",
"simulations",
"biology",
"computational",
"biology",
"biophysics",
"simulations",
"biophysics",
"biomechanics"
] | 2011 | A Mathematical Model of Muscle Containing Heterogeneous Half-Sarcomeres Exhibits Residual Force Enhancement |
Inter-cellular communication with stromal cells is vital for cancer cells . Molecules involved in the communication are potential drug targets . To identify them systematically , we applied a systems level analysis that combined reverse network engineering with causal effect estimation . Using only observational transcriptome profiles we searched for paracrine factors sending messages from activated hepatic stellate cells ( HSC ) to hepatocellular carcinoma ( HCC ) cells . We condensed these messages to predict ten proteins that , acting in concert , cause the majority of the gene expression changes observed in HCC cells . Among the 10 paracrine factors were both known and unknown cancer promoting stromal factors , the former including Placental Growth Factor ( PGF ) and Periostin ( POSTN ) , while Pregnancy-Associated Plasma Protein A ( PAPPA ) was among the latter . Further support for the predicted effect of PAPPA on HCC cells came from both in vitro studies that showed PAPPA to contribute to the activation of NFκB signaling , and clinical data , which linked higher expression levels of PAPPA to advanced stage HCC . In summary , this study demonstrates the potential of causal modeling in combination with a condensation step borrowed from gene set analysis [Model-based Gene Set Analysis ( MGSA ) ] in the identification of stromal signaling molecules influencing the cancer phenotype .
Stromal tissue is a major component of solid tumors . It consists of extracellular matrix , connective tissue cells , inflammatory cells , and blood vessels . Stromal cells affect cancer development and progression by augmenting tumor cell proliferation , survival , motility and invasion [1 , 2 , 3] . Tumor and stromal cells can interact through both , direct cell-cell contact and secreted factors such as growth factors , cytokines , chemokines , and their cognate receptors [2 , 3] . Hepatocellular carcinoma ( HCC ) is one of the most prevalent and lethal malignant tumors worldwide . The major risk factor predisposing to HCC is hepatic cirrhosis . It arises through the activation of hepatic stellate cells ( HSC ) , myofibroblast-like cells that are responsible for the excessive hepatic matrix deposition seen in chronically damaged livers [4 , 5] . Moreover , HSCs infiltrate the stroma of liver tumors localizing around tumor sinusoids , fibrous septa , and capsules [4 , 1] . Conditioned medium collected from activated HSCs induces growth , migration and invasion of HCC cells in vitro [6 , 7 , 8 , 9] . Furthermore , HSCs promote aggressive growth of HCC cells in experimental in vivo models [4 , 6 , 9 , 10] and their presence predicts poor clinical outcome in HCC patients [11] . These data indicate that HSCs affect HCCs . Yet , the molecular mechanisms of this crosstalk are largely unknown . In functional assays , signaling pathways are analyzed through perturbation of the cellular systems . Unlike statistical associations in observational data , functional assays can directly distinguish between cause and effect . Their disadvantage is that they can be difficult to perform in high throughput . Recently , Maathuis and colleagues introduced a novel method to extract causal information from observational gene expression data [12] . In their IDA algorithm they combine local reverse network engineering using the PC-algorithm [13] with causal effect estimation [14 , 15] . These virtual functional assays predict lists of genes that will change expression if the expression of a query gene was perturbed experimentally . The method was successfully applied to predict the expression profiles of yeast deletion strains from observational data of wild type yeast only [16] . Here , we adapt the IDA framework to the problem of identifying agents of inter-cellular communication . We combine a specific experimental design with tailored causal discovery and data integration algorithms . In brief , HSCs obtained from n = 15 human donors were cultivated to generate conditioned media for stimulation of the established HCC cell line Hep3B . Gene expression was then measured in both , HSCs as well as stimulated and un-stimulated HCC cells and a list of genes that change expression in HCCs upon stimulation was established . First , we aimed at identifying gene pairs ( x , y ) where the expression of gene x in HSCs affects the expression of gene y in HCC cells . Next , we searched for a small set of HSC expressed genes that , together accounted for the majority of stimulation sensitive genes in HCC cells . This yielded a set of 10 HSC genes predicted to jointly influence 120 of 227 HCC cell genes affected by supernatant stimulation .
To study cell communication directed from stroma to cancer cells , we treated the HCC cell line Hep3B with 15 media conditioned by 24-hour cultivation with HSCs that had been isolated from different human donors . This design allows us to study the messages sent from HSCs to HCC cells independently from feedback messages that might be sent in the opposite direction from HCC cells to HSCs . The lack of feedback in this design is an indispensable prerequisite for our analytic approach . Genome-wide gene expression was measured in both , donor HSCs and HCC cells stimulated with conditioned media ( CM ) , yielding 15 pairs of gene expression profiles . The gene expression profiles of four un-stimulated HCC cell cultures served as controls . We identified a list of 227 genes with more than two-fold expression changes between stimulated and un-stimulated cells at an estimated false discovery rate ( FDR ) of 0 . 001 . Interestingly , 30 ( 13 . 2% ) of the 227 genes were among the top 200 genes with the highest variance in expression across the 15 stimulation assays ( Fig 1 ) . These genes reflect biological variation both across HSC donors and cancer cells stimulated by the HSCs . The genes that drive HSC induced neoplastic progression , including proliferation and migration in HCCs , are most likely among them [17] . In fact , testing for overrepresented Gene Ontology terms [18] pointed to several hallmarks of cancer: negative regulation of apoptosis ( anti-apoptosis , q < 10–4 ) , angiogenesis ( q < 10–4 ) , inflammation ( cellular response to lipopolysaccharide , q < 10–4 ) , positive regulation of cell migration ( q < 10–3 ) , and growth factor activity ( transforming growth factor beta receptor signaling pathway , q < 10–3 ) ( S1 Fig ) . Next , we searched for indications which pathways might be regulated by stromal signals in HCC cells . The CM sensitive genes were mapped onto the BioGRID interactome of established protein-protein and protein-gene interactions [19] and the largest regulated subnetwork was identified by the BioNet algorithm [20] . The regulated network comprises several interacting oncogenic signaling pathways including TGF-beta/SMAD3 , NFκB , JAK1 and MAP kinase signaling components ( Fig 2 ) . Another branch of the subnetwork can be attributed to anti-apoptotic signals with the highly induced BIRC3 gene ( ENSG00000023445 ) in its center . Amplification of the chromosomal region containing BIRC3 exons is frequently found in HCC and associated with chemotherapy resistance , metastasis and poor prognosis [21] . The strongest induced gene , RND1 ( log2 fold change of 4 . 9; ENSG00000172602 ) , a member of the Rho GTPase family [22] , belongs to yet another branch of the subnetwork that comprises genes involved in regulating rearrangements of the actin cytoskeleton and , thus , changes in cell adhesion and motility in response to extracellular growth factors [23] . So far , we have only described the HSC-mediated changes in the HCC cell transcriptome . We have not yet identified the HSC secreted proteins that actually stimulate receptors or otherwise directly interact with HCCs . In a naïve analysis , we might find many genes in HSCs that correlate with some of the genes that are regulated in HCCs; however , most of them will not cause these changes . In fact , if we counted the number of HCC genes a particular HSC gene correlates with ( absolute Pearson correlation > 0 . 7 ) , we would identify HSC-secreted POSTN ( ENSG00000119655 ) , PGF ( ENSG00000119630 ) , CSF1 ( ENSG00000184371 ) , NPC2 ( ENSG119655 ) and FGF5 ( ENSG00000138675 ) . The top 10 list also includes HGF ( ENSG00000019991 ) and is shown in S1 Table . Although this list points to potential stromal regulators , for some gene pairs correlation will be high due to a third factor that influences both of the correlated genes . To exclude the latter and to find true causal regulators instead , we use the “in silico perturbation framework” of the IDA algorithm [12] to filter for genes that are operative in stroma-to-tumor communication . Application of IDA comprises two steps . First , a partially directed network of regulatory interactions is constructed using the PC algorithm [13] . Second , causal effects are estimated from this network using Pearl’s Do-calculus [14] . To infer a potential effect of a stromal gene x on a cancer gene y , the Do-calculus needs the expression of y , x , and all genes x’ that generate spurious correlations between x and y ( e . g . common regulators ) . Adjusting for the expression of the x’ ( termed “parents of x” ) differentiates between true causal effects and spurious correlations . If x does not have parents in the network ( e . g . x10 in Fig 3 ) , the estimated causal effect is identical to the correlation coefficient . However , if there are parents , causal effects are different from correlation coefficients . In these cases interpreting correlation coefficients is misleading . Since HSCs were never in contact with HCC cells , parent genes of x must be of HSC origin . Hence , it is sufficient to confine the reconstruction of a regulatory network to the HSC expression profiles only . An illustration of the HSC network is shown in Fig 3 . To limit the computational burden resulting from genes that are not expressed in HSCs or that did not vary across HSCs from different donors , we only included the highest and most variably expressed genes ( see Material and Methods ) across the HSC samples in the analysis . The expression levels of HCC cell genes enter the model in the second step as y-genes , and the HSC network is used to derive causal effects of HSC on HCC genes ( represented by green dashed arrows in Fig 3 ) . For some genes , we have two expression values: one from the HSC sample that produced the CM , and one from the respective CM-stimulated HCC cell sample . For simplicity , we refer to these expression levels as the expression of the HSC gene and the HCC gene , respectively . For each of the 227 HSC-inducible HCC genes , we used IDA to screen for potential HSC genes that—when perturbed in expression—will have strong effects on the respective HCC gene . We limited our search for candidate HSC regulators to genes annotated as ‘secreted’ , ‘extracellular’ or ‘intercellular’ , but not ‘receptor’ by Gene Ontology and for which the gene product was detected in the conditioned media by HPLC/MS/MS . Gene products that are too small for detection , e . g . IGF1 ( ENSG00000017427 ) and IGF2 ( ENSG00000167244 ) were left in the analysis . This resulted in a final list of 186 HSC genes as candidate stromal regulators . The gene list with corresponding proteins can be found in S2 Table . Gene-pair-by-gene-pair , the HSC gene was “virtually repressed” by one standard unit and the expected change of the HCC gene was calculated . It is important to note that causal analysis will discover both direct and indirect effects of x on y , i . e . irrespective of potential mediators m , and discover effects of x and m if they are both secreted HSC genes . For example , in Fig 3 , x10 has a causal effect on y3 , although mediator node x11 also has a causal effect on y3 . To be robust against small perturbations of the data , the "virtual repression" was run in a sub-sampling mode , repeating the experiment 100 times each on a different subset of the samples . Within each run , secreted HSC genes were ranked by the size of their estimated effects on the 227 target HCC genes . We kept causal effects only if they appeared in the top ranks across the majority of sub-sampling runs ( see Material and Methods ) . This resulted in 96 HSC genes potentially regulating at least one of the 227 HCC genes . A flow-chart of our methodology is depicted in Fig 4 . Although all 186 HSC proteins have the potential to affect the expression of HCC genes , we postulate that a much smaller set of proteins is sufficient to activate HCCs . Thus we aimed at identifying a small set of HSC genes that jointly account for the wide spectrum of expression changes in HCC cells observed in response to stimulation with HSC-CMs . We have generated 227 lists of HSC regulators , one for each of the 227 CM sensitive HCC genes . Since many HSC genes were predicted to affect multiple HCC genes , these lists overlap . The lists can be reorganized by HSC genes instead of HCC genes . This resulted in 96 non-empty sets of HCC genes that are targeted by the same HSC gene . Model based gene set analysis [24] ( MGSA ) is an algorithm that aims at partially covering an input list of genes with as little gene ontology categories as possible . It balances the coverage with the number of categories needed . We modified this algorithm in such a way that it covered the list of 227 CM sensitive HCC genes with the 96 sets of HSC targets . This strategy identified sparse lists of predicted targets that covered most of the observed targets . By definition , every list corresponded to one secreted HSC protein . This analysis brings HSC genes in competition to each other: an analysis based on frequencies ( how many HCC genes does each HSC gene affect ) discovers redundant HSC genes that target the same HCC genes . Our approach strives for a maximum coverage of the target genes with a minimum number of HSC secreted genes . Both stability selection on the IDA algorithm and MGSA depend on the setting of a few parameters . Several studies have shown that hepatocellular growth factor ( HGF ) affects HCC cells [25] , and is highly expressed in HSCs [25 , 26] . We exploited this knowledge and calibrated the parameters such that HGF appeared in the list of predicted HSC genes . With these parameters , we identified 10 HSC secreted proteins . In addition to HGF the list included PGF , CXCL1 , PAPPA , IGF2 , IGFBP2 , POSTN , NPC2 , CTSB , and CSF1 ( Table 1 ) . With the exception of IGF2 all proteins were found in at least one of five CMs that were analyzed using LC/MS/MS . IGF2 is too small for successful detection [27] . Notably , the set of the most influential HSC regulators included several well-known tumor-promoting genes such as placental growth factor ( PGF ) [28] , and the chemokine CXCL1 , which promotes HCC angiogenesis and growth [29] . Periostin ( POSTN ) is a secreted cell adhesion protein whose expression levels are directly related to metastatic potential and poor prognosis of HCC [30] . High expression levels of the macrophage colony-stimulating factor 1 ( CSF1 ) are another indicator of tumor progression and poor survival in HCC patients [31] . Over-expression of cathepsin B ( CTSB ) , on the other hand , promotes HCC cell migration and invasion [32] . The role of Niemann-Pick Type C2 ( NPC2 ) protein in cancer is just beginning to be understood . NPC2 regulates intracellular cholesterol homeostasis via direct binding with free cholesterol . Perturbations of cholesterol metabolism affect cancer progression [33] . Elevated serum levels of NPC2 have been observed in patients with lung cancer [34] and , more recently , HCC [35] . Modulation of cholesterol homeostasis by NPC2 also affects activation of mammalian target of rapamycin ( mTOR ) [36] , a critical signaling cascade in several types of cancer including HCC [37] . Remarkably , we identified three genes of the insulin-like growth factor ( IGF ) axis . This signaling pathway regulates tumor progression in several types of tumors including HCC [38] . The key molecules in this pathway are the ligands IGF1 and IGF2 , IGF-binding proteins ( IGFBPs ) , and membrane-associated receptors ( IGF-I receptor ( IGF-IR ) , mannose-6-phosphate receptor/IGF-II receptor ( IGF-IIR ) ) . High expression levels of IGF2 are predictive of aggressive tumor growth and poor prognosis in HCC patients [39] . IGF2 binds to the receptor tyrosine kinases IGF1R ( ENSG00000140443 ) and IGF2R ( ENSG00000197081 ) on HCC cells and activates multiple intracellular signaling pathways , including the phosphatidylinositide-3′-kinase ( PI3K ) /Akt and MAP kinase signaling cascades [40] . IGFBPs bind to IGFs with higher affinity than IGF-receptors and , thereby , modulate local IGF concentrations and activities [40 , 41] . Unlike most IGFBP family members , which conduct antitumor activity , IGFBP2 promotes invasion , metastasis , and angiogenesis [41] . It is over-expressed in several tumor tissues including HCC [41 , 42] . The metalloprotease Pregnancy-Associated Plasma Protein A ( PAPPA ) is also a member of the IGF-axis . PAPPA is implicated in several biological functions [43] , including the regulation of local IGF1 bioavailability through cleavage of IGFBPs [44] . Its expression in the liver under both , physiological and pathological conditions , including HCC development and progression , has not been elucidated yet . The few available studies on other tumor entities located PAPPA expression to cancer rather than stromal cells [45] , and controversial roles of PAPPA regarding tumor progression have been reported in ovarian cancer [46] . Thus , we decided to focus our subsequent analysis on the role of PAPPA in HCC . In principle , parameters in our analysis could be set to different values and lead to different results . We evaluated the influence of gene pre-filtering and parameter settings in our analyses and found that the results were stable within the computationally feasible settings . Gene pre-filtering was necessary because network estimation is computationally very demanding with many genes . We evaluated our criteria for gene selection in a leave-one-out cross-validation and found that the selected genes are stable ( secreted HSC genes: 95 . 1% identical with standard deviation ( SD ) 0 . 7% , intracellular HSC genes: 86 . 6% identical with SD 1 . 3% , HCC genes: 97 . 2% identical with SD 1 . 4% ) . S3 Table shows an aggregation of results when varying parameters in the causal analysis and demonstrates that these results are also stable . Among others , PAPPA is always within the top 10 stromal regulators . The list of CM sensitive HCC genes includes various members of the NFκB pathway ( Fig 2; NFKB1 ( ENSG00000109320 ) , NFKB2 ( ENSG00000077150 ) , NFKBIZ ( ENSG0000014480 ) , NFKBIA ( ENSG00000100906 ) , RELB ( ENSG00000104856 ) ) and targets of the NFκB pathway previously collected by Compagno et al [47] , such as BIRC3 , EGR1 ( ENSG00000120738 ) , ICAM1 ( ENSG00000090339 ) , IL8 ( ENSG00000169429 ) , MAP3K8 ( ENSG00000107968 ) . Several of these genes were predicted to be targets of HSC secreted PAPPA by our causal analysis ( ICAM1 , MAP3K8 , NFKBIA , see S4 Table for the full list ) . Also the other predicted target genes are known to be regulated by the transcription factor NFκB or to affect this signal transduction pathway [48 , 49 , 50 , 51 , 52 , 53] . To test whether PAPPA might be indeed responsible for activation and auto-regulation of the NFκB pathway , we assessed NFκB activity in stimulated HCC cells and observed a striking correlation of PAPPA levels in conditioned medium ( CM ) from the 15 different HSCs with NFκB activity induced in HCC cells upon incubation with these different CMs ( Fig 5A ) . To verify a causal effect of PAPPA on NFκB activity in HCC , we stimulated Hep3B HCC cells with recombinant human PAPPA protein ( rPAPPA ) . We applied rPAPPA ( 25 ng/ml ) either alone or in CM of HSCs from two different donors containing endogenous PAPPA levels of 4 . 8 ng/ml and 6 . 2 ng/ml , respectively . In control medium , rPAPPA did not significantly affect IkB-α- and p65-phosphorylation , while together with CM both IkB- α- and p65-phosphorylation were higher than in CM-stimulated cells ( Fig 5B ) . Quantitative real time PCR analysis showed strong PAPPA mRNA expression in HSCs whereas no expression was detectable in 4 different human HCC cell lines including Hep3B ( S2 Fig ) . Concordantly , PAPPA protein levels ranged from approximately 5 to 35 ng/ml in supernatants of HSC cultures , while no PAPPA protein was detectable in the supernatants of the 4 different HCC cell lines ( Figs 6A and S3 ) . In the 15 different HSCs , we observed a significant correlation between mRNA and protein levels of PAPPA ( Fig 6B ) , indicating that secreted PAPPA levels are regulated at the transcriptional level . Next , we assessed PAPPA gene expression in HCC specimens from 52 patients and found a significant correlation with collagen type I ( COL1A1; ENSG0000010882 ) mRNA expression ( Fig 6C ) . This finding could be confirmed in the HCC cohort of The Cancer Genome Atlas ( TCGA , http://cancergenome . nih . gov ) ( S4 Fig ) . HSCs infiltrate and form the HCC stroma and collagen type I is specifically expressed by HSCs in HCC tissue [4 , 54 , 5] . Together , these findings indicate that HSCs are the major source of PAPPA in HCC . Histological staging of HCC is a prognostic factor of patient survival in HCC [54 , 55 , 56] . We found that PAPPA expression in human HCC specimens ( n = 52 ) was significantly lower ( p = 0 . 008 , one-way ANOVA ) in patients with low histological staging ( stage I; n = 12 ) compared to patients with stage II ( n = 19 ) and stage III ( n = 21 ) disease ( Fig 7 ) . In an independent data set , the HCC cohort of TCGA , PAPPA expression was also significantly lower in stage I patients ( n = 104 ) compared to stage II ( n = 56 ) and stage III ( n = 39 ) in a one-way ANOVA ( p = 0 . 0126 ) ( S5 Fig ) . Together , these findings indicate the clinical relevance of HSC secreted PAPPA for HCC progression .
Introductory statistical text books stress the difference between association and causation . For example , correlation between the expression levels of two genes does not imply that one gene regulates the other . They can as well be co-regulated by a third gene . The gold standard to infer causalities is experimental intervention . If a knock-down of the first gene changes the expression of the second , there is a functional relation between the two . In fact , the rationale of functional genetics is to understand the cell by breaking it . Functional assays that perturb biological networks experimentally shed light on cellular mechanisms . Causal inference from observational data is a more advanced statistical discipline [13 , 14] that only recently found its way into bioinformatics and systems biology after a statistical breakthrough paper by Maathuis et al . ( 2009 ) [12] . To date it has been used for the analysis of yeast deletion strains [16] , to predict genes regulating flowering time in Arabidopsis thaliana [57] , and for the prediction of miRNA targets [58] . Here , we add another biological application to this list: The identification of secreted proteins that drive inter-cellular communication in human cancer . State of the art statistical methodology does not allow for feedback mechanisms between the regulator and its target . This is an assumption that nature does not meet in many cases . In a tumor it is most likely that the communication between stromal and tumor cells is mutual . In our experimental setting however , feedback is blocked . Stromal and cancer cells grow in separate cultures . The stromal cells "talk" to the cancer cells via the CMs but there is no "reply" . Clearly , this does not give us a full picture of cellular communication; feedback mechanisms are blocked and so are signals mediated by cell-cell contacts . But it is this focus on unidirectional paracrine signaling that allows us to use causal modeling . The experimental design is tailored to the capabilities of the predictive model . In spite of these limitations our application to HCC demonstrates that the method can generate novel and potentially clinical relevant insights into the mechanisms of stroma-tumor communication . We unmasked PAPPA as a novel stroma secreted factor impacting the tumor phenotype . Notably , our 10 HSC secreted regulators did not only include PAPPA but two more genes of the IGF-axis . The IGF-axis is one of the molecular networks involved in the formation , progression and metastatic spread of many cancer types , including HCC . IGF2 and IGFBP2 are known to critically affect HCC development and progression . Still , most studies focused on autocrine effects of these two secreted proteins in cancer cells , while our data suggest a paracrine effect whereby HSC derived IGF2 and IGFBP2 influence IGF-signaling in HCC cells . The expression and function of PAPPA in normal and diseased liver were not known thus far . To date , PAPPA has been mainly used as a biomarker in prenatal screening for Down's syndrome [43] . More recently , PAPPA has been identified as a regulator of the bioavailability of IGFs through the cleavage of IGF binding proteins [43 , 59] . It has been suggested to exert a protumorigenic role in breast cancer , lung cancer , and malignant pleural mesothelioma [59] . In contrast , breast cancer cells have been reported to become more invasive after down-regulation of PAPPA [60] . Controversial roles of PAPPA have also been reported in ovarian cancer , with most ovarian cancer cell lines and primary tumors showing partial or complete loss of PAPPA expression [45] . Furthermore , PAPPA expression was shown to be consistently high in normal ovarian specimens , while it was suppressed by SV40 large T antigen [61] . In HCC , our data suggest PAPPA as a protumorigenic factor . We found significantly higher PAPPA expression levels in advanced stage tumors . On the mechanistic side , we found that PAPPA induces NFκB-activity in HCC cells . We observed a significant correlation between PAPPA levels in different conditioned media of HSCs and corresponding effects on NFκB activation in HCC cells in vitro . Interestingly , studies in ovarian , breast and lung cancer as well as malignant pleural mesothelioma revealed the cancer rather than the stromal cells as the cellular source of PAPPA . Here , in contrast , PAPPA expression was only detected in HSCs , but not in HCC cells . This makes PAPPA a promising therapeutic target in HCC , as tumor stromal cells are genetically more stable than cancer cells , which renders them less likely to evade therapy . Moreover , it has to be considered that the IGF-axis also plays a critical role in HSC activation and fibrosis [62] . Although the function of PAPPA in HSCs is unknown , it may be speculated that PAPPA inhibition may suppress the fibrogenic phenotype of HSCs . Since HCC mostly develops in cirrhotic liver tissue [1 , 4] , inhibition of PAPPA could not only affect HCC cells but also prevent the formation of a protumorigenic soil for cancer cells . Due to its central role in cancer progression , a variety of reagents have been developed to modulate IGF signaling including neutralizing antibodies against IGFs and IGF-receptors as well as associated receptor kinase inhibitors in aim for cancer treatment [63] . The structural similarities of the insulin and IGF-IRs complicate the development of specific agents that block IGF-IR signaling without affecting insulin signaling . This is particularly true with regards to treatment of liver cancer due to the central role of the liver in glucose metabolism and homeostasis . In contrast to the persistent and versatile physiological functions of other components of the IGF1 axis , PAPPA could not be detected in normal human liver and primary human hepatocytes ( S6 Fig ) . Therefore , PAPPA appears as a better therapeutic target for HCC with more tumor specificity and less risks of side effects as compared to other IGF1 axis components . Actually , genetic deletion of PAPPA extended lifespan of mice [59 , 64] . In conclusion , we have shown for the first time that causal modeling can be used to identify stromal signaling molecules that influence the cancer phenotype . Application of our modeling strategy unmasked PAPPA as a novel paracrine factor that shapes the tumor phenotype via activating the NFκB pathway .
Human liver tissues were obtained and experimental procedures were performed according to the guidelines of the charitable state controlled foundation HTCR ( Human Tissue and Cell Research ) , with the informed patients’ consents , and approval by the local ethics committee of the Ludwig-Maximilians University of Munich ( reference number 025–12 ) . All experiments involving human tissues and cells have been carried out in accordance with The Code of Ethics of the World Medical Association ( Declaration of Helsinki ) . The human HCC cell lines Hep3B ( American Type Culture Collection ( ATCC ) number HB-8064 ) , HepG2 ( ATCC; HB-8065 ) , PLC ( ATCC; CRL-8024 ) and Huh-7 ( Japan Collection of Research Bioresources ( JCR ) number B0403 ) were cultured as described [10 , 65] . Primary human hepatic stellate cells ( HSCs ) were isolated from 15 different human donors as described [10 , 66 , 67] . The isolation procedure and cell culture on uncoated tissue culture dishes led to the activation of HSCs as described [66 , 67] . For collection of conditioned medium ( CM ) , HSCs were seeded into T75 flasks ( 2 × 106 cells ) . One day after seeding cells were washed twice with serum-free DMEM , and then incubated for another 24 h with serum-free DMEM ( 15 mL/T75 ) . CM was centrifuged at 6 , 000 x g to remove cell debris , sterile filtered ( 0 . 45 μm pore size membrane filter ) , and stored in aliquots at −80°C until use . Serum-free DMEM incubated for 24 h in cell culture flasks without cells served as the control . For stimulation with HSC conditioned media , HCC cells were seeded into T25 flask ( 106 cells ) . One day after seeding , cells were washed with serum-free DMEM , and then incubated for another 12 h with serum-free DMEM . Subsequently , the medium was changed and cells were incubated with 3 mL of HSC-CM or control medium ( serum-free DMEM ) for 4 h . For individual experiments , CM was preincubated with recombinant PAPPA ( R&D Systems , Wiesbaden , Germany ) . HCC tissues were obtained from HCC patients undergoing surgical resection . Tissue samples were immediately snap-frozen and stored at -80°C until analysis . Isolation of total cellular RNA from cultured cells and tissues and reverse transcription were performed as described [10 , 65] . 300 ng of RNA were hybridized to Affymetrix Human Gene ST 1 . 0 arrays following the standard Affymetrix protocol ( Affymetrix , High Wycombe , UK ) . Hybridization and scanning were performed at an Affymetrix Service Provider and Core Facility , “KFB—Center of Excellence for Fluorescent Bioanalytics” ( Regensburg , Germany; www . kfb-regensburg . de ) . Quantitative real-time-PCR was performed applying LightCycler technology ( Roche , Mannheim , Germany ) and the following pairs of primers: human PAPPA ( forward: 5'-AGC CAG CAG CAT CCC AGG TGT-3'; reverse: 5'-CGC CCG GAG CCA AAA AGT GGT ) -3' and human collagen type I ( forward: 5'- CGG CTC CTG CTC CTC TT -3'; reverse: 5'-GGG GCA GTT CTT GGT CTC -3' ) . Amplification of cDNA derived from 18s rRNA ( forward: 5'-TCT GTG ATG CCC TTA GAT GTC C-3'; reverse: 5'-CCA TCC AAT CGG TAG TAG CG-3' ) was used for normalization . Protein extraction and western blotting analysis were performed as described [65] applying antibodies against phospho-NF-κB p65 ( ( Ser536 ) rabbit mAb #3033 ) and phospho-IκBα ( ( Ser32 ) ; rabbit mAb #2859 ) both from Cell Signaling Technology ( Danvers , MA , USA; all diluted 1:1 , 000 ) . Furthermore , an antibody against actin ( MAB1501 from Merck Millipore , Billerica , MA , USA; 1:1 , 000 ) was applied . Activated NF-κB was quantified in nuclear extracts with the ELISA based kit TransAm from Active Motif ( Rixensart , Belgium ) according to the manufacturer's instructions , as described [66] . Normalization of raw intensity values from CEL files was performed using variance stabilization ( VSN ) [68] . Median polish and a custom chip description file based on ensembl gene identifiers [69] were used to summarize individual probes to obtain an expression level per gene . Raw intensities and normalized gene expression data are available publicly at the NCBI Gene Expression Omnibus ( GEO , http://www . ncbi . nlm . nih . gov/geo/ ) under accession GSE62455 . Differential gene expression between Hep3B cells treated with different CMs and untreated Hep3B controls was estimated using limma [70] . All analyses were performed within the statistical programming environment R . Gene Set Analysis ( GSA ) was performed using hypergeometric tests implemented in the Bioconductor package HTSanalyzeR [71] . Genes meeting the FDR threshold of 0 . 001 and an absolute log2 fold change larger than one were selected for testing significant enrichment of Gene Ontology ( GO ) terms within the Biological Process ( BP ) branch . The Bioconductor package BioNet [20] was used to find the highest-scoring sub-network within the differentially expressed genes with FDR < 0 . 001 and an absolute log2 fold change larger than 0 . 7 . Aliquots of conditioned media ( 400 μL each ) were used for protein precipitation with 4 volumes of ice-cold acetone . After 2 h incubation at -20°C , samples were centrifuged at 20 , 000 x g for 10 min . Pellets were air-dried and stored at -20°C until further use . Combining the lists of proteins identified with gel-free and gel-based secretome analysis resulted in 305 proteins total . Protein pellets were dissolved in 0 . 5 M triethylammonium bicarbonate ( TEAB , Sigma Aldrich , St . Louis , MO , USA ) and denatured at 60°C for 1 hour . The exact protein concentration was determined employing a Bradford assay , using a serial dilution of bovine serum albumin ( BSA , Sigma Aldrich ) from 31 . 25 to 2000 μg/mL in 0 . 5 M TEAB for calibration . Disulfide bonds were reduced at 60°C for 1 hour by addition of 4 . 55 mM tris ( 2-Carboxyethyl ) phosphine hydrochloride solution ( TCEP-HCl , Sigma Aldrich ) , followed by alkylation with 8 . 7 mM iodo acetamide ( IAA , Sigma Aldrich ) at 24°C for 30 min . Protein digestion was performed overnight at 37°C using trypsin ( Promega , Madision , WI , USA ) at a ratio of 1:50 to the protein concentration . Digests were dried in a SpeedVac before adjusting peptide concentration to 1 μg/μL in 0 . 05% trifluoracetic acid ( TFA , Sigma Aldrich ) . The HPLC instrument was an UltiMate 3000 Nano LC system from Dionex ( Germering , Germany ) and the mass spectrometer was an LTQ Orbitrap XL from Thermo Scientific ( Waltham , MA , USA ) equipped with a nano-electrospray ion source . The spray was generated with 10 μm id and 360 μm o . d . fused silica tips from New Objective ( Woburn , MA , USA ) . Tryptic peptides were separated by nano-ion-pair reversed-phase ( IP-RP ) —HPLC at pH 2 . 0 on a 150 × 0 . 20 mm I . D . RP polymer monolith capillary column from Thermo Scientific using a 2-hour gradient of 0–40% acetonitrile in 0 . 05% aqueous trifluoroacetic acid at a flow-rate of 1 μL/min . The MS1 survey scans of the eluting peptides were executed in the LTQ Orbitrap XL with a resolution of 60 , 000 , recording a window between 450 . 0 and 2000 . 0 m/z . The three most intense precursor ions were selected for fragmentation with collision-induced dissociation ( CID ) . The normalized collision energy ( NCE ) was set at 35 . 0% for all scans . Data evaluation was performed with Proteome Discoverer ( Thermo Scientific ) and the open—source library OpenMS . Protein pellets were dissolved in 10 μL of LDS-sample buffer and separated on Invitrogen NuPAGE BisTris SDS-gels ( 4–12% , MOPS-buffer system ) with subsequent colloidal Coommassie staining . Lanes were cut into 30 slices of equal size and washed , carbamidomethylated and tryptically digested prior to nano-LC-QTOF-MS/MS analysis as published previously [72] . Tandem mass spectra were searched against the Uniprot database ( version 57 . 15 ) using the Mascot 2 . 2 search algorithm ( Matrix Science , London , UK ) applying the two-peptide-rule . To find HSC gene products that influence gene expression in HCC cells , we applied Intervention-calculus when the DAG ( directed acyclic graph ) is Absent ( IDA ) [12] . The algorithm consists of two parts: first , an equivalence class of DAGs is estimated from the observational expression data with the pc-algorithm [13] , before causal effects are derived using the graph and intervention calculus [14] . Prior to modeling , gene selection was performed as follows: Gene products secreted from HSC cells were defined as all genes with the terms ‘extracellular’ , ‘intercellular’ or ‘secret*’ in any Gene Ontology term or definition . This yielded 1919 genes . Next , genes coding for receptors were removed . The remaining genes were filtered based on expression level , excluding genes that had not been expressed at least in 1/15 CM-stimulated HSC samples at a level larger than the 40th percentile of expression values across all genes and HSC samples . Next , genes with low inter-quartile range , a robust estimate of the variance , across HSC samples were excluded ( lowest 20% ) , yielding 1024 genes annotated to be secreted or present outside of the cell . Next , the overlap between these genes and the gene products detected by mass spectrometry in the HSC-CM ( 305 gene products ) was generated , resulting in 153 gene products . Additionally , growth factors were retained even if they were not detected , as for example IGFs are too small to be monitored by mass spectrometry . This procedure led to a final number of 186 HSC-secreted proteins with a potential influence on HCC cell gene expression going into modeling . The list of HSC secreted gene products is provided in S2 Table . From the remaining HSC genes , only the genes with highest expression levels ( at least 3 samples above the 40th percentile ) and with highest inter-quartile range ( top 976 , such that the total number of HSC genes was 2000 ) were selected . These genes were supposed to build the network that regulates the secreted genes . On the HCC sample side , genes were selected for differential expression based on significance ( q < 0 . 001 ) , and on log2 fold change ( absolute log2 fold change > 1 ) to focus only on the strongest responses of the HCC cells . This resulted in 227 HCC genes . The filtering procedure is depicted in the left part of Fig 4 . Gene expression values were centered and scaled to standard deviation equal to one to make causal effects comparable across genes . From the 2000 HSC genes ( secreted and remaining genes ) , the equivalence class of DAGs was estimated and causal effects were derived from the secreted HSC genes on the selected HCC genes . IDA needs a single tuning parameter , α , which controls the neighborhood size of the graph . It was set to 0 . 2 as this resulted in the best balance between a not too sparse network and computational burden ( higher α values lead to longer running times ) . To find effects insensitive to small disturbances of the data , IDA was run in a sub-sampling approach adopted from Meinshausen & Bühlmann [73] . For a total of 100 times , 12 out of the 15 samples were drawn , the CPDAG was estimated and causal effects were derived for each DAG in the equivalence class . As a lower bound , the minimum effect of the individual DAGs was retained . The effects were then ranked across all outcome genes ( differentially expressed cancer genes ) by effect size for each sub-sampling run and the relative frequency of an effect being among the top 30% of effects across all runs was recorded . All effects with a relative frequency equal or above 0 . 7 were retained for further analysis and the median effect across all sub-samples was recorded . The steps of the causal analysis are schematically shown in the right part of Fig 4 . To gain insights into the most important HSC derived regulators of gene expression in HCC , Model-based Gene Set Analysis ( MGSA ) [24] was employed with the modification that gene sets were redefined as all genes targeted by a specific regulator . For example , the gene set ‘CXCL1’ was comprised of all HCC genes on which CXCL1 exerted a predicted causal effect . MGSA was then used to find a sparse set of regulators explaining the observed differentially expressed genes ( q < 0 . 001 , absolute log2 fold change > 1 ) . All predictor-target sets with a posterior probability > b were declared to be the most important regulators . The parameters within MGSA were left at default values , but the size of the gene sets ( controlled by the relative frequency cutoff in stability selection ) used as input of MGSA was calibrated such that HGF , a known true positive , was in the final list of secreted regulators . While this criterion did not give us unique parameter settings , the remaining genes in the lists resulting from different parameter settings that included HGF were almost identical ( S3 Table ) . Un-normalized RNA sequencing and clinical data of liver hepatocellular carcinoma ( LIHC ) patients was downloaded from The Cancer Genome Atlas ( TCGA , http://cancergenome . nih . gov ) and normalized using size factors calculated by the R package DESeq2 [74] ( function ‘estimateSizeFactorsForMatrix’ ) and log2-transformed with a pseudo-count of 1 to avoid missing values for samples with zero counts . For the analysis of association of PAPPA expression levels with staging , patients staged with the 7th edition of the AJCC ( American Joint Committee on Cancer ) that were classified into stages I , II or IIIA were used ( n = 199 ) . Stages IIIB , IIIC , IV , and IVA were omitted because of low sample sizes ( n<10 ) . For the correlation of PAPPA levels with COL1A levels , all LIHC patients were used ( n = 424 ) . | All living cells rely on communication with other cells to ensure their function and survival . Molecular signals are sent among cells of the same cell type and from cells of one cell type to another . In cancer , not only the cancer cells themselves are responsible for the malignancy , but also stromal ( non-cancerous ) cells and the molecular signals they send to cancer cells are important factors that determine the severity and outcome of the disease . Therefore , the identification of stromal signals and their influence on cancer cells is important for the development of novel treatment strategies . With a computational systems biology model of stroma-cancer cell communication , we have compiled a set of ten proteins secreted by stromal cells that shape the cancer phenotype . Most importantly , our causal analysis uncovered Pregnancy-Associated Plasma Protein A ( PAPPA ) as a novel paracrine inducer of the pro-tumorigenic NFκB signaling pathway . In liver cancer patients , higher levels of PAPPA protein indicate a more progressed tumor stage , confirming its clinical relevance . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Causal Modeling of Cancer-Stromal Communication Identifies PAPPA as a Novel Stroma-Secreted Factor Activating NFκB Signaling in Hepatocellular Carcinoma |
Monozygotic ( MZ ) twins do not show complete concordance for many complex diseases; for example , discordance rates for autoimmune diseases are 20%–80% . MZ discordance indicates a role for epigenetic or environmental factors in disease . We used MZ twins discordant for psoriasis to search for genome-wide differences in DNA methylation and gene expression in CD4+ and CD8+ cells using Illumina's HumanMethylation27 and HT-12 expression assays , respectively . Analysis of these data revealed no differentially methylated or expressed genes between co-twins when analyzed separately , although we observed a substantial amount of small differences . However , combined analysis of DNA methylation and gene expression identified genes where differences in DNA methylation between unaffected and affected twins were correlated with differences in gene expression . Several of the top-ranked genes according to significance of the correlation in CD4+ cells are known to be associated with psoriasis . Further , gene ontology ( GO ) analysis revealed enrichment of biological processes associated with the immune response and clustering of genes in a biological pathway comprising cytokines and chemokines . These data suggest that DNA methylation is involved in an epigenetic dysregulation of biological pathways involved in the pathogenesis of psoriasis . This is the first study based on data from MZ twins discordant for psoriasis to detect epigenetic alterations that potentially contribute to development of the disease .
Psoriasis is a common chronic inflammatory disease , which affects mainly the skin , but also the joints . The worldwide prevalence is reported to range between 1–11 . 8% depending on ethnicity , geographical area and method of assessment [1] . Psoriasis is known to have a strong genetic component with an estimated heritability of 66% [2] . Linkage peaks [3] , copy number variations ( CNVs ) [4]–[6] and genes identified by GWAS [7]–[13] are associated with psoriasis . However , the combined effect of these loci does not account for the genetic variation underlying the observed susceptibility to psoriasis , and indicates the involvement of additional genetic , epigenetic or environmental factors [7] , [14]–[15] . Further evidence for the role of epigenetic or environmental factors comes from the fact that the concordance rate among MZ twins is only 35–72% [2] , [14] , [16] . Phenotypic differences are a result of genetic and epigenetic variation , as well as environmental influences . The study of discordant MZ twins provides an attractive model to investigate epigenetic mechanisms in disease . Environmentally driven epigenetic changes are thought to contribute to development of autoimmune diseases through alteration of epigenetic profiles , but exactly how the environment modulates epigenetic states is not well understood . Besides twin discordance , accumulating evidence supports the contribution of epigenetics in the development of autoimmune diseases through dysregulation of the immune system [17]–[20] . Psoriasis is considered a T cell-mediated autoimmune disease , and T cell activation is a key event in the pathogenesis [21] . Antigen binding results in a complex downstream signaling pathway in which epigenetic mechanisms have an important role [21] , [22] . A likely scenario involves an abnormal activation and migration of T cells into the skin , followed by a cascade of events , which ultimately results in aggregation of the inflammatory cells and development of psoriasis . The development of psoriatic plaques is a result of an activation of cells in focal skin regions , which is mediated by CD4+ ( helper ) and CD8+ ( cytotoxic ) cells [23] . The aim of this study was to identify epigenetically dysregulated genes , which contribute to development of psoriasis . To do this , we assessed the extent of epigenetic and transcriptomic differences in CD4+ and CD8+ cells isolated from MZ twins discordant for psoriasis . This study design enabled identification of cell-type specific DNA methylation differences which correlate with gene expression , and thereby identification of genes where DNA methylation may have a functional role in development of psoriasis .
Since genome-wide patterns of DNA methylation are known to differ between cell-types [24]–[26] , and different cell-types in the immune system are implicated in the pathogenesis of psoriasis , we isolated and studied CD4+ and CD8+ cells to overcome the issue of epigenetic heterogeneity in whole blood . Both cell-types are relatively abundant in blood and have important functions in the immune system , thus they are good targets for epigenetic alterations which might contribute to the development of psoriasis . Comparisons of CD4+ and CD8+ cells revealed many significant differences for both DNA methylation ( 1 , 288 of the 26 , 690 CpG sites , 4 , 8% , n = 12 , Table S1 ) and gene expression ( 2 , 126 of the 37 , 846 transcripts , 5 , 6% , n = 13 , Table S2 ) in unaffected individual twins ( Figure 1 ) , which clearly demonstrate the importance of isolating specific cell-types . MZ co-twins are highly correlated for DNA methylation in both CD4+ cells ( n = 17 pairs , mean ρ = 0 . 98 , range = 0 . 96–0 . 99 ) and CD8+ cells ( n = 13 pairs , mean ρ = 0 . 97 , range = 0 . 95–0 . 98 ) ( Figure 2A and 2B ) . Both analyses of individual CpG sites and mean methylation per gene did not reveal any differentially methylated sites between unaffected and affected co-twins . Individual scatter plots of DNA methylation clearly demonstrate greater similarity among the MZ twins than among unrelated individuals ( Figure S1 ) . To ensure that the observed differences in DNA methylation between co-twins were genuine rather than technical artifacts , we ran internal replicates on a subset of the twins . Specifically , we replicated 7 pairs and calculated technical and biological differences between replicated samples ( self-self comparisons ) and between co-twins , respectively . The overall distributions of the technical differences in DNA methylation per CpG site were significantly smaller than the biological differences ( Kolmogorov-Smirnov test , two-sided , P-value<2 . 2×10−16 , Figure S2 ) . This clearly shows that the observed biological differences between unaffected and affected co-twins are genuine , although they are small . Similarly , differences in gene expression between MZ co-twins were small in both CD4+ cells ( n = 17 pairs , mean r = 0 . 99 , range = 0 . 97–0 . 99 ) and CD8+ cells ( n = 14 pairs , mean r = 0 . 99 , range = 0 . 98–0 . 99 ) . Although there are many small differences , we did not detect any genome-wide significant differences in DNA methylation or gene expression between co-twins discordant for psoriasis when analyzed separately ( Figure 2C and 2D ) . DNA methylation is essential for the regulation of gene expression . We reasoned that a combined analysis of DNA methylation and gene expression could select functional methylation sites involved in regulating gene expression . We therefore investigated if co-twin differences in DNA methylation and gene expression were correlated . To do this , we compared the differences ( between co-twins ) in mean β-values for the CpGs associated with each gene , with the log fold changes of the gene expression . Using Spearman's rho as a measure for correlation , we then ranked the genes according to the significance of the correlation coefficients . This combined analysis of DNA methylation and gene expression revealed cell-type specific differences , identifying genes known to be involved in immune response and associated with psoriasis , only in CD4+ cells . Table 1 shows the top 50 genes ranked according to the significance of the correlation in CD4+ cells . Genes associated with psoriasis ( shown in bold ) are overrepresented in this list ( Fisher's exact test , P = 3 . 3×10−5 , Table S3 ) . IL13 have been identified in GWAS [8] , [10] , [11] , [27] , and ALOX5AP [28] , PTHLH [29] and TNFSF11 [30] have all been linked to different aspects of the disease . The entire list of the 11 , 933 genes studied ranked according to the significance of the correlation in CD4+ and CD8+ can be found as Table S4 in supporting information . Scatter plots depicting the relationship of MZ co-twin differences in DNA methylation and gene expression of TNFSF11 demonstrate the strong correlation in CD4+ cells compared to a non-significant correlation in CD8+ cells ( Figure 3 ) . We used DAVID [31] , [32] to explore the potential of shared biological pathways among the genes in the list we generated from the combined analysis of DNA methylation and gene expression in CD4+ and CD8+ cells ( Table S4 ) . GO analyses identified significant enrichment of GO terms in CD4+ cells ( Table 2 ) , whereas the analysis did not detect any enriched terms in CD8+ cells . A significant portion of the top 1% of the genes ranked at the top of the list were found to be involved in the immune response ( GO: 0006955 , 12 . 5% , P = 0 . 042 ) , positive regulation of response to stimulus ( GO: 0048584 , 7 . 5% , P = 0 . 037 ) , immune system process ( GO: 0002376 , 15% , P = 0 . 043 ) and regulation of response to stimulus ( GO: 0048583 , 10 . 8% , P = 0 . 043 ) . All of these categories contain genes involved in the immune response , which are potentially important in autoimmune diseases . Interestingly , a subset of the genes identified in the GO analysis ( IL13 , IL23R , CCL1 , CCL5 , CSF2 , TNFSF11 , LTB and SF9 ) comprises part of the cytokine-cytokine receptor interaction pathway , which is essential in communication between cells in the immune system . Skewed cytokine levels of pro-inflammatory and anti-inflammatory cytokines characterize the pathogenesis of psoriasis . Cytokines and chemokines are essential in the communication between cells in the immune system . Whereas cytokines generally influence proliferation , differentiation and secretion of pro- or anti-inflammatory factors , chemokines primarily have an effect on the movement of cells [33] . Thus , pathways like the cytokine-cytokine receptor interaction are indeed relevant in the etiology of psoriasis . In this context , our results suggest that DNA methylation is important in regulation of the cytokine cascade and signaling pathways involved in psoriasis . Both CD4+ and CD8+ cells are known to be present in psoriatic plaques , and current evidence indicates that CD4+ cells play a more critical role than CD8+ cells [34] . Our data strongly suggest that CD4+ cells are important in the pathogenesis of psoriasis . However , in this context it is important to recognize the complexity of CD4+ cells , which consists of several subpopulations with specific roles in the immune system ( i . e . upon activation , naïve CD4+ cells develop into several lineages; Th1 cells , Th2 cells , Th22 and T regulatory cells ) . Recently , much attention has been drawn towards Th17 cells and the role in the pathogenesis of psoriasis [35] . It has also been speculated that CD4+ cells at different differentiation states may be present , which complicates the picture even further [36] . Distinctive compositions of these subpopulations can potentially contribute to the observed intra-pair differences . In addition , the complexity of the CD4+ and CD8+ cells could explain the small intra-pair differences by averaging out the level of DNA methylation and gene expression . Thus , a disease-associated change in DNA methylation and gene expression in a subset of cells can ultimately appear as an overall small difference . Recently , several genes and pathways associated with psoriasis have been identified in GWAS [7]–[13] . Many of these have an essential role in the immune system and this clearly demonstrates the importance of immune response regulation in the disease . The molecular mechanisms driving the inflammation in skin causing psoriasis are complex . Our findings identify new potential susceptibility genes and point to different plausible biological pathways in psoriasis that are under epigenetic regulation and suggest an epigenetic dysregulation of biological pathways implicated in immune function . It will be important to expand on these findings in larger twin and other non-twin cohorts .
The twins were recruited pair-wise from the Norwegian Twin Registry ( NTR ) . Altogether 105 pairs were invited to participate , 60 pairs were invited from the cohorts born 1967–1979 [37] , [38] and 45 pairs were invited from the cohorts born 1924–1960 [39] . The selection of discordant MZ pairs was based on a two-step procedure . Initial screening was conducted using self-reported data collected via questionnaires in earlier studies [37]–[39] . Pairs for which both twins consented to participate were called in to a clinical dermatology interview and skin examination at Oslo University Hospital , where additional information was also collected . Among the 105 pairs invited through the initial screening phase , 35 pairs consented and 27 pairs clinically evaluated to be discordant for psoriasis participated . The affected twins generally presented with a mild form of psoriasis , mainly affecting the scalp , knees and elbows . Scores for body surface area affected ( BSA ) were generally low and less than 10% for all the affected . The study was approved by the regional ethical committee and written informed consents were obtained . Lymphocyte subpopulations were isolated in a semi-automated way . Briefly , 60 ml EDTA-blood was diluted 1∶1 with RPMI , split into three aliquots and layered over Lymphoprep solution ( Axis-Shield ) in 50 ml centrifugation tubes . After centrifugation , PBMCs formed a distinct band that was harvested and washed twice to remove contaminating platelets . CD4+ cells and CD8+ cells were then sequentially isolated using positive and negative isolation kits from Miltenyi on an autoMACS Pro separator ( Miltenyi ) . CD8+ cells were positively isolated using CD8+ MicroBeads and the Possel program . CD4+ cells were then separated from the negative fraction by negative isolation ( i . e . by labeling all other cells but the CD4+ cells ) using CD4+ T Cell Isolation kit II and the Deplete program . The zygosity for all twin pairs was determined based on 13 microsatellites on chromosomes 13 , 18 , 21 , X and Y . CD4+ cells were cultured for 7 days in X-VIVO ( Lonza ) , exogenous rIL-2 ( 10 ng/µl ) and Dynabeads CD3/CD28 T Cell Expander ( Invitrogen ) using ½ bead per cell . CD8+ cells were cultured for 14 days under the same conditions . DNA was isolated from cultured cells on a Gentra autopure LS ( Qiagen ) using the 2–5×107 protocol . This resulted in high yields of pure DNA with an A260-A280 between 1 . 7 and 1 . 9 . Total RNA was prepared manually from cultured CD4+ and CD8+ lymphocytes using RNAqueous Small Scale phenol-Free Total RNA Isolation Kit ( Ambion ) according to manufacturers instructions . RNA quality was checked using an Agilent 2100 Bioanalyser ( Agilent Technologies ) . DNA methylation status was assessed using the Infinium HumanMethylation27 BeadChip , according to manufactures instructions ( Illumina ) . These arrays enabled detection of the methylation status of 27 , 578 individual CpGs predominantly distributed in the promoters of 14 , 475 coding genes throughout the genome . The fluorescence data were analyzed in BeadStudio ( Illumina ) to determine the β-values ( quantitative measurement of the methylation ) for each CpG and normalized in Bioconductor lumi package [40] . We selected 26 , 690 probes that unambiguously mapped to the genome ( hg18 ) with up to 2 mismatches as was done in Bell et al . [41] . In DNA methylation analysis , we excluded 1 out of 18 pairs in CD4+ cells from all analysis due to low bisulfite conversion . Gene expression profiling was performed using the HumanHT-12 v3 Expression BeadChip , targeting >25 , 000 annotated genes , according to manufactures instructions ( Illumina ) . The data were quartile normalized in BeadStudio ( Illumina ) . All statistical tests were conducted in R ( http://www . r-project . org/ ) . The significance of the differences in DNA methylation between CD4+ and CD8+ cells were calculated based on a paired t-test in unaffected twins with overlapping data from both cell-types . Correlation of DNA methylation in discordant MZ co-twins was computed based on a nonparametric Spearman rank correlation . To search for differentially methylated genes between unaffected and affected twins we used a paired t-test on the mean β-value on all CpG sites associated with each gene . In addition , we also searched for differentially methylated CpG sites based on the β-value per CpG separately . The FDR significance thresholds were calculated using stats ( R package ) , after raw P-values of a paired t-test had been computed . Nonparametric permutation tests were also performed , where P-values were calculated by comparing the t-statistic of the true data set with the t-statistics resulting from permutations of the affection status of the twins in all possible combinations . The results from these permutation tests produced similar results . To search for differentially expressed genes between CD4+ and CD8+ cells and between unaffected and affected twins , the data was first log2 transformed and an empirical Bayes moderated t-test was then applied using the Limma package [42] . Correlation of gene expression between MZ co-twins was computed based on the parametric Pearson correlation . All statistical tests were done two-tailed and a false discovery rate ( FDR ) below 5% was considered significant . Of the genes represented on the HumanMethylation27 BeadChip , 11 , 933 were also present on the HumanHT-12 Expression BeadChip , which enabled integrated analysis of the methylation status and gene expression . In the combined analysis of DNA methylation and gene expression , we compared the differences ( between co-twins ) in mean β-values for the CpGs associated with each gene , with the log fold changes of the gene expression . Using Spearman's rho as a measure for correlation , we then ranked the genes according to the significance of the correlation coefficients . GO analysis was conducted using the DAVID functional annotation tool [31] , [32] with the 120 most significantly correlated genes as input ( 1% ) and the 11 , 933 genes that was used in the combined analysis of DNA methylation and gene expression as background . Analysis was done with default parameters and results corrected for multiple testing by the method of Benjamini and Hochberg [43] . | Psoriasis is a common chronic inflammatory disease , which affects mainly the skin , but also the joints . It is considered a T cell–mediated autoimmune disease . Autoimmune diseases are in general due to a dysregulation of the immune system , and identification of genes involved in alterations of lymphocyte function is therefore essential . Although there is convincing evidence of a genetic basis underlying psoriasis , there is also evidence for the involvement of environmental and/or epigenetic factors . Here we use MZ twins discordant for psoriasis to search for disease-causing epigenetic and transcriptomic alterations in isolated lymphocyte subpopulations . We identified many genes involved in the immune response where changes in DNA methylation between unaffected and affected twins correlated with changes in gene expression . In addition , several of these genes had previously been identified in GWAS and linkage studies of psoriasis . These findings suggest that psoriasis involves an epigenetic dysregulation of immune-system genes . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"dermatology",
"genomics",
"molecular",
"cell",
"biology",
"immune",
"cells",
"cell",
"biology",
"genetics",
"epigenetics",
"biology",
"molecular",
"genetics",
"immunology",
"cellular",
"types",
"genetics",
"and",
"genomics",
"immune",
"system"
] | 2012 | DNA Methylation and Gene Expression Changes in Monozygotic Twins Discordant for Psoriasis: Identification of Epigenetically Dysregulated Genes |
Upon infection , many RNA viruses reorganize their capsid for release of the genome into the host cell cytosol for replication . Often , this process is triggered by receptor binding and/or by the acidic environment in endosomes . In the genus Enterovirus , which includes more than 150 human rhinovirus ( HRV ) serotypes causing the common cold , there is persuasive evidence that the viral RNA exits single-stranded through channels formed in the protein shell . We have determined the time-dependent emergence of the RNA ends from HRV2 on incubation of virions at 56°C using hybridization with specific oligonucleotides and detection by fluorescence correlation spectroscopy . We report that psoralen UV crosslinking prevents complete RNA release , allowing for identification of the sequences remaining inside the capsid . We also present the structure of uncoating intermediates in which parts of the RNA are condensed and take the form of a rod that is directed roughly towards a two-fold icosahedral axis , the presumed RNA exit point . Taken together , in contrast to schemes frequently depicted in textbooks and reviews , our findings demonstrate that exit of the RNA starts from the 3′-end . This suggests that packaging also occurs in an ordered manner resulting in the 3′-poly- ( A ) tail becoming located close to a position of pore formation during conversion of the virion into a subviral particle . This directional genome release may be common to many icosahedral non-enveloped single-stranded RNA viruses .
Human rhinoviruses ( HRV ) , members of the picornavirus family , genus Enterovirus , are the major causative agent of the common cold . Additionally , they play an important role in the exacerbation of asthma , cystic fibrosis , and chronic obstructive pulmonary disease [1] . Similar to other picornaviruses , the species HRV-A , -B , and -C , are composed of 60 copies each of four capsid proteins , VP1 , VP2 , VP3 , and the small myristoylated VP4 , arranged on an icosahedral T = 1 , P = 3 lattice . The diameter of the particle is roughly 30 nm . The viral genome is a single-stranded RNA molecule of positive polarity , about 7100 bases in length . It carries a covalently linked peptide ( VPg ) at its 5′-end and a poly- ( A ) tail of about 70 to 150 bases at its 3′-end [2] , [3] . The 5′-nontranslated region is approximately 650 bases in length , highly structured , and involved in cap-independent translation initiation and RNA replication [4] . Minor group rhinoviruses , exemplified by the prototype strain HRV2 , bind members of the low-density lipoprotein receptor ( LDLR ) family including LDLR and LDLR-related protein for entry via clathrin-dependent endocytosis [5] . Once in the endosome , the low pH leads to dissociation of the virus from the receptors as well as to structural changes in the viral capsid [6] , [7]; more specifically , the native virion sedimenting at 150S converts into the subviral A-particle sedimenting at 135S and devoid of the internal capsid protein VP4 [8] , [9] and exposure of amphipathic N-terminal sequences of VP1 renders it hydrophobic , thus allowing its direct attachment to the inner endosomal membrane [10] . These processes are accompanied by an expansion of the viral shell by about 4% along with the opening of symmetry-related channels . The largest channels are at the two-fold axes , whereas the smaller ones are located near the pseudo three-fold axes and at the base of the star-shaped vertices of the icosahedron [11] , [12 , and our unpublished resuts] . Finally , the RNA is released through one of these pores , most probably at a 2-fold axis as suggested from cryo-electron microscopy ( cryo-EM ) image reconstructions of the related poliovirus in which uncoating was induced by heating to 56°C [13] , [14] . The final product of this uncoating process is the empty capsid ( 80S B-particle ) . Most enteroviruses are believed to undergo similar conformational changes; however , with the exception of minor receptor group rhinoviruses , the process appears to be triggered by receptor binding and possibly assisted by low pH [7] , [15]–[17] . These structural modifications can be mimicked , at least partially , in vitro . Exposure to pH<5 . 8 converts native HRV2 preferentially into A-particles whereas incubation at 50°C–56°C in low ionic strength buffers favours conversion into B-particles ( empty capsids ) [18] . In vivo , and in the presence of liposomes in vitro [19] , both VP4 and N-terminal sequences of VP1 insert into lipid bilayers . They might contribute to formation of a pore connecting the virus interior with the cytosol of the host cell , thus allowing for the transit of RNA in its unfolded form ( reviewed in [20] , [21] ) . The necessity for unfolding was suggested by experiments with poliovirus , which demonstrated loss of the intercalating dye Syto 82 during RNA egress [22] . The mechanism of RNA exit is poorly understood . Energy would be required for breaking the hydrogen bonds of the double-stranded regions in the encapsidated RNA genome [23] , [24] in order to allow the RNA to thread through an opening only large enough to enable passage of a single strand [12]; however the source of this energy for in vivo uncoating is unknown . It appears highly likely that either the poly- ( A ) tail at the 3′-end or the VPg peptide linked to the 5′-phosphate of the RNA via a tyrosine ester , begins to emerge from the virion since other modes might be unproductive ( e . g . , simultaneous exit of both ends would be expected to impede complete uncoating and thus to be abortive ) . Directionality of this process may indicate that the RNA adopts a defined conformation inside the viral shell suggesting a well-organized process of assembly and uncoating . Here , we show that RNA exit does indeed occur in a specific and ordered manner , starting from the 3′-end .
Fluorescence correlation spectroscopy ( FCS ) allows measuring the diffusion time , and thus determining the diffusion coefficient , of fluorescent molecules at very low concentrations ( down to picomoles/l ) in very small sample volumes ( down to femtoliters ) [25] , [26] . Binding of a small labeled molecule to a substantially larger one gives rise to fluorescent complexes that diffuse slowly [27] . Deconvolution of the autocorrelation function of the free and the complexed component allows for calculation of their respective fractions present in the mixture . For instance , recombinant very-low density lipoprotein receptor fragments ( Mr = 12 to 80 kD ) were labeled with Cy3 at their N-termini and the change of their diffusion coefficient upon binding to HRV2 ( Mr = 8 . 5 MD ) was monitored [28] . Resolving the fraction of free and virus-bound receptor fragments ( at various virus concentrations ) allowed determination of the binding constants . Here we used fluorescently labeled oligonucleotides complementary to regions near the 3′ and the 5′-ends of the RNA molecule for hybridization to detect sequences becoming accessible during viral uncoating . In control experiments shown in Fig . 1A to C , the diffusion coefficients ( Table 1 and 2 ) were determined for each of the three reaction components ( see Methods section ) . Fig . 1A shows that the autocorrelation functions of fluorescent oligonucleotide probe , YOYO-labeled in vitro-transcribed HRV2 RNA , and Dylight-labeled virus are sufficiently different to allow for determination of their relative concentrations in the mixture; attempts at determining the autocorrelation function for virus with partially-released RNA were hampered by the heterogeneity of the samples obtained after the crosslinking step that was necessary for halting complete egress upon heating ( see below ) . As expected , the values were generally somewhat lower than those for native virus , however , they varied considerably . For this reason , all of the calculations below were carried out using the diffusion coefficient of the native virus instead . Further control experiments ( Fig . 1B and C ) demonstrated a rightward shift of the autocorrelation curve of the 5′ and 3′end-specific oligonucleotides upon binding to in vitro-transcribed viral RNA and a reversal on digestion with RNases . The diffusion coefficient was virtually identical for viral RNA obtained by heating HRV2 to 56°C for 20 min and in vitro-transcribed RNA ( not shown ) . Thus , the slightly different size of the poly- ( A ) tail , as well as the small VPg that is absent from in vitro-transcribed RNA , had no significant impact on the diffusion properties . The measured and calculated diffusion coefficients of the probes , the free RNA , and the virus are summarized in Table 1 and the values for the two different probes , either free or hybridized to viral RNA , prior and after digestion with RNases , are shown in Table 2 . HRV2 was incubated at 56°C in the presence of the respective probes for the times indicated in Fig . 2 and cooled on ice; the autocorrelation function of the oligonucleotides hybridized to cognate-accessible RNA sequences was then measured at ambient temperature . Cooling halted further exit of the RNA when returned to room temperature , whereas RNA egress continued at a markedly reduced rate when this step was omitted ( data not shown ) . For the 3′-specific probe ( Fig . 2A ) , the correlation time was increased at 7 min ( due to binding to partially-extruded RNA that was still connected with the protein shell ) but decreased at 20 min to a value corresponding to the probe now hybridized to free RNA . Conversely , in case of the 5′-specific oligonucleotide , it remained unchanged at 7 min and increased at 20 min , but did not exceed the value corresponding to free RNA ( Fig . 2B ) . Measurements were then repeated with additional incubation times . The percentages of free oligonucleotide , oligonucleotide bound to free RNA , and oligonucleotide bound to virus with part of its RNA having become exposed , were obtained by fitting the data to a three-component autocorrelation function using the diffusion coefficients of each constituent obtained in the previous experiment . We assumed that the value for virus and virus with partially-extruded RNA were negligibly different ( see above and Table 1 ) . Fig . 2C and D show a clear difference in the time-dependent change of the percentages of accessible 3′ and 5′ ends , free RNA , and RNA attached to the virion . At 3 min incubation at 56°C , the fraction of the free 3′-oligonucleotide ( O3 ) had strongly diminished . This was accompanied by an increase of O3 bound to virus with externalized RNA ( V-R ) . Between 10 and 15 min V-R diminished again , whereas O3 bound to free RNA ( R ) increased until attaining a plateau . Conversely , as shown in Fig . 2D , the fraction of free 5′-specific oligonucleotide ( O5 ) only started to decrease at about 12 min , with a concomitant increase of RNA-bound O5 . Apparently , part of the RNA molecules had been entirely released at that time resulting in the 5′-ends being accessible for hybridization with the probe . The plateau at about 20 min indicates that no further RNA became available for hybridization , suggesting that release was completed . There was no significant change in V-R over the observation time indicating absence of virus with exposed RNA harbouring the sequence complementary to O5 ( Fig . 2D ) . Control experiments showed that RNA exit was insignificant at ambient temperature during the timeframe of the experiment . Nevertheless , cooling to 4°C appeared important , as keeping the sample at room temperature for more than 10 min resulted in aggregation ( as indicated by the appearance of fluorescent species with very long diffusion times; not shown ) . Whereas hybridization to O3 plateaued at about 75% , hybridization to O5 only reached about 50% possibly due to the higher number of potential binding sites for the former ( the poly- ( A ) is between 50 and 200 nucleotides long whereas the region complementary to O5 is unique; see also Materials and Methods ) . In order to confirm and complement the FCS results , we determined the segment of the RNA remaining inside the virion . This was achieved by halting the uncoating reaction when only part of the RNA had left the virion and removing the exposed RNA by RNase digestion . Since psoralen crosslinking readily abrogated infectivity of the related poliovirus [29] , we reasoned that it may impede egress of the entire genomic RNA by preventing complete unfolding of double-stranded regions [30] during uncoating . This would lead to the accumulation of structures intermediate between full 135S particles ( A-particles with genomic RNA not yet released ) and empty capsids to levels high enough to allow for their characterization . Small molecules such as N-acetyl-aziridine [31] and Ribogreen [32] can diffuse into native HRV at physiologic temperature due to ‘breathing’ and bind to the RNA , rendering it fluorescent . Control experiments showed that incubation of purified HRV2 with 8-Methoxypsoralen ( 8-MOP ) for 4 h at 37°C followed by irradiation at 365 nm led to an approximately 90% loss of viral infectivity ( data not shown ) , indicating that this compound was also able to diffuse into the virion . We therefore assessed the effect of this treatment on the integrity of the virus and its subviral particles formed on incubation at 56°C by negative staining electron microscopy ( EM ) . As seen in Fig . 3 , psoralen crosslinking did not change the morphology of native virus ( compare Fig . 3Aa and 3Ab ) . When such samples were heated to 56°C , both native virus and crosslinked virus were converted into particles possessing an internal density with a rod-like appearance , but to a remarkably different extent . Without crosslinking , their proportion was low ( Fig . 3Ac ) , but it substantially increased upon crosslinking prior to heating ( Fig . 3Ad ) . Conversion from full to empty capsids was apparently halted at a stage where some RNA was still remaining inside the virion ( with part of it assuming a rod-like shape ) and relatively few empty particles were observed ( Fig . 3B ) . The time-dependent formation of these ‘rod-particles’ from psoralen-crosslinked virions was assessed by visual inspection of the micrographs and counting . The result of a representative experiment is depicted in Fig . 3C . Native virions ( present at t = 0 ) and ‘full-looking’ particles ( HRV2 remaining native plus A-particles present at t>0 ) decreased over time with a concomitant increase of ‘rod-particles’ and a minority of ‘empty-looking’ particles ( see inset in Fig . 3B for typical examples of these particles ) . Presumably , egress of the RNA ( single-stranded ) through the pore in the viral shell was arrested as soon as a crosslink was encountered . Apparently , the RNA adopts the form of a rod as a consequence of heat-triggered partial release . After 10 min , about 39% of the virus had converted into particles containing condensed RNA , indicated by either a typical ‘rod’ or a dot appearance in the center , depending on the orientation of the virion with respect to the plane of the grid . There were ∼5% empty particles and a remainder of 56% ‘native-like’ ( i . e . ‘full-looking’ ) virions . Further extension of the incubation time at 56°C led to accumulation of viral debris , suggesting that prevention of ordered uncoating may result in ( partial ) disruption of the virion ( see Suppl . Material , Fig . S2 ) . In order to exclude the possibility that the rod-like density inside the subviral particles observed in negative stain EM was a staining artefact , we also performed cryo-EM analysis . Samples were prepared as above but applied to microscope grids with holey carbon film , frozen , and images were recorded . As seen from the representative images in Fig . 4A , particles appearing full , containing rod-like density , and apparently empty particles were again observed . The dataset of downscaled particle images ( 64×64 pixels ) was submitted to maximum likelihood 3D ( ML3D ) classification [33] , [34] imposing three classes and particle images assigned to each class were analyzed individually for their relative RNA content by relating the density in the core ( mainly corresponding to the RNA ) to the density corresponding to the protein shell . Fig . 4B shows that the distribution of the RNA content in the entire dataset was bimodal with a peak and a broad shoulder . Separate analysis of the images corresponding to the respective classes yielded three partially-overlapping peaks; one representing high , but considerably variable , RNA content ( class1 ) , and two with intermediate ( class2 ) and low ( class3 ) RNA content , respectively . The reconstructions were then further refined by using particle images ( now at 128×128 pixels ) corresponding to the three respective classes; icosahedral symmetry was imposed ( top half of the models ) or not ( lower half of the models ) . These are depicted in Fig . 5 ( a ) – ( c ) as radially color-coded surfaces with the view down a 2-fold axis , as transverse sections ( d ) – ( f ) , and as central planes ( g ) – ( i ) . Despite the lower resolution of the asymmetric reconstruction ( 24 Å , 22 Å , and 22 Å , respectively ) , as compared to 14 Å , 13 Å , and 13 Å obtained when imposing symmetry , the icosahedral shape of the virus was quite well-defined in all three classes when rendered at a contour level of sigma = 1 above the mean density . Nevertheless , comparing the volumes reconstructed with and without imposing symmetry , the ( almost ) full particle had asymmetric features , whereas empty and ‘rod-containing’ particles were remarkably symmetric ( e . g . see Fig . S2B and compare the five-fold axes on the left of the class1 particle in Fig . 5a , black arrow indicating the distortion ) . The viral shell showed no obvious deviation from symmetry with respect to the orientation of the rod ( compare also to Fig . S2 ) . The spherically averaged radial density plots ( Fig . S1 ) established that the three models had similar diameters . All particles were by 4% larger than native virions; therefore , the ‘full’ particles are probably more similar to subviral A-particles than to native virions but contain variable amounts of RNA ( Fig . 4B ) . Determination of the long axis passing through the center of mass of the rod-like density allowed estimation of its approximate orientation with respect to the icosahedral axes of the virion . Fig . 5 ( j ) – ( l ) , show that the rod does not traverse the center of the virion but is slightly shifted aside . As a result , it contacts the protein shell at a position close to a 2-fold axis and close to a 3-fold axis on roughly the opposite face . As anticipated from visual inspection of the micrographs ( Fig . 4A ) , extending the number of classes in the ML3D analysis from 3 to 10 revealed a more heterogeneous dataset ( Fig . S2 ) . Four classes that were each populated by less than 5% of the total particle images corresponded to damaged or heavily distorted virions ( not shown ) , one class was represented by ( almost ) full virions , 2 classes corresponded to ( almost ) empty particles with one of them lacking a pentamer , and 3 classes constituted particles with significant ‘rod-like’ density . In keeping with the heterogeneity of the density of the rod seen in Fig . 4A , the three latter classes contained rods of different shape . Upon rendering at sigma = 1 . 7 ( class2 ) , sigma = 2 . 0 ( class3 ) , and sigma = 1 . 5 ( class4 ) all showed that the rod was contacting the inner face of the capsid close to a 2-fold axis and extending to the opposite side . Despite employing high underfocus we were unable to visualize any RNA outside of the particle . It is possible that traces of RNase remaining from viral purification ( see Methods section ) digested exposed RNA under the conditions of sample preparation employed for electron microscopy . To further assess the nature of these subviral particles , crosslinked HRV2 was incubated at 56°C for different times and subjected to capillary electrophoresis ( CE ) in the non-ionic detergent Thesit [35] . Native virus , A- and ( empty ) B-particles were identified according to their electrophoretic mobility as described previously [36] . As seen in Fig . 6A , non-crosslinked native virus used as a control ( N ) quickly converted into empty capsids ( E; to about 50% in 5 min and to almost 100% in 20 min under these particular conditions ) whereas crosslinked virus ( Nx ) gave rise , almost exclusively , to particles migrating as a broad peak with a substantially increased migration time ( Rx; Fig . 6B ) . Presumably , as a consequence of incomplete uncoating , Rx carry less than the entire complement of the RNA inside the shell , but the remainder is exposed , thereby modifying the surface charge-to-size ratio relevant for the migration behaviour . For quantification , samples were incubated as above , but collected at various time points and analysed . The data are summarized in Fig . 6C and D . The percentage of both non-crosslinked and crosslinked native virus decreased with increased incubation time at 56°C . For the non-crosslinked sample , a small transient increase in the base-line corresponding to heterogeneous RNA-containing intermediate particles ( R ) was observed; on further heating , these converted completely into empty capsids ( E ) . Conversely , crosslinked virus transformed virtually entirely into such heterogeneous intermediate particles ( Rx ) . The concentration of empty particles was below the detection limit . Note that R and Rx likely include intermediate particles with and without the typical ‘rod-like’ density ( as in Fig . 3 , 4 , and S2 ) , since the conformation of their RNA core is not expected to impact on their migration behaviour in capillary electrophoresis . In order to confirm the presence of exposed RNA in HRV2 subjected to crosslinking and heating as above , the resulting particles were incubated with mAB J2 , a monoclonal antibody specifically recognizing dsRNA [37] and again analyzed by CE ( Fig . 7 ) . There was a clear shift of the peak related to Rx in the presence of the antibody ( compare peaks marked with a blue arrow in Fig . 7A and B to the second internal standard , benzoic acid; B . A . ) . When the sample in Fig . 7A was treated with micrococcal nuclease ( MNase ) , the peak shifted to the position corresponding to 135S particles ( blue arrow in Fig . 7C ) , which have no exposed RNA [36] . Without incubation at 56°C , the crosslinked virus exhibited the electrophoretic mobility of native HRV2 [36] ( red arrow ) regardless of the presence of the antibody ( Fig . 7D ) . These experiments confirm that only part of the viral RNA can be released in the presence of psoralen crosslinks . This RNA remains connected to the subviral particle , refolds externally , and is therefore detected by dsRNA-specific antibodies . Crosslinking followed by incubation at 56°C for 10 min resulted in subviral particles that carried part of the viral genome exposed to the outside ( see above ) . Therefore , we determined which part of it remained inside the viral shell and thus protected from RNases . ‘Rod-particles’ were prepared as above at 56°C , accessible RNA was digested with MNase , the nuclease was inactivated with EGTA , and the sample subjected to native agarose gel electrophoresis followed by staining for RNA . As seen in Fig . 8A , crosslinked virus that had not been heated ( control; Nx ) showed a well-defined band close to the top of the gel . In contrast , the sample that had been heated and incubated with MNase migrated further ( Rx ) . The respective bands were cut out and analyzed for the presence of proteins by SDS polyacrylamide gel electrophoresis followed by silver staining ( Fig . 8B ) . The band stemming from crosslinked but not heated virus ( Nx ) contained VP1–VP4 , whereas crosslinked , heated , and MNase-treated virus ( Rx ) only contained VP1–VP3 indicating that it had been converted into subviral 135S A-particles devoid of VP4 . RNA was extracted from the remaining part of the gel slices , reverse transcribed and the cDNA amplified with two primer pairs each , selected to produce fragments derived from the 3′ and the 5′-ends , respectively ( see scheme on the bottom of Fig . 8C ) . Based on the gel analysis it becomes clear that cDNA was exclusively amplified from the 5′-end , indicating that at least about 1000 nucleotides of the RNA , including the extreme 3′-end , were lost due to digestion whereas at least one fourth of the genome remained inside the shell . This comprises roughly 35% of the total RNA estimated to be present in the subviral particles including ML3D class1 and class2 comigrating in CE ( see Methods section ) . Control experiments showed that in vitro-transcribed RNA and RNA extracted from band Nx in Fig . 8A yielded the same fragments at a ratio of about 1∶1 although the crosslinking reduced the overall yield of all fragments to some extent . This suggests that the crosslinks were introduced in a random fashion . A sample of the gel cut at position ‘C’ ( Fig . 8A , control ) was treated exactly as the other bands , verifying that the amplification reaction did not produce any signal-excluding contamination .
Because of its appealing and suggestive nature , schemes of enterovirus uncoating have long depicted the RNA as exiting at one of the five-fold axes; however , this model was never supported by experimental evidence . Exit at the 5-fold axis was brought into question by results of cryo-EM analysis of intermediate uncoating states of poliovirus [38] , and of metastable complexes between HRV3 and a soluble form of its receptor ICAM-1 [39] . These data suggested that structural changes led to thinning of the viral shell at the canyon floor pointing to the probable exit of the N-terminal extension , and even of the RNA , at or near this site . The original model was finally challenged by cryo-EM data of poliovirus heated to 56°C . In these more recent studies , external density was detected close to a two-fold axis and this external density was interpreted to be exiting RNA [13] , [14] . This view was supported by the X-ray analyses of the empty capsid of HRV2 [12] and of the converted natural empty particle of EV71 [40] , demonstrating that the transition from native virion to empty capsid is accompanied by the opening of channels at the two-fold axes . Additionally , the ‘plug’ of the ß-cylinder built from five copies of VP3 below the vertex at the five-fold axis was found to be virtually unchanged , blocking the small opening seen from the outside . The holes at the two-fold axes are now believed to be exit points for the genomic RNA of enteroviruses . In the present paper we demonstrate that RNA egress begins with the 3′-end . Our model is supported by two complementary findings . Firstly , we observed completely different kinetics for the exit of the 3′ and the 5′-ends of the RNA from the viral shell on incubation at 56°C . Deconvolution of the FCS autocorrelation data showed that the 3′-specific oligonucleotide first bound to molecular assemblies with diffusion properties expected for RNA that was still attached to the virus . At later time points , their diffusion rather corresponded to free RNA only , reflecting complete release from the virion . On the other hand , the oligonucleotide specific for the 5′-end never exhibited diffusion times corresponding to such virion-associated RNA . Secondly , in order to halt RNA transit at intermediate states , we used psoralen UV crosslinking . With capillary electrophoresis , we demonstrated that such crosslinked particles heated to 56°C indeed carried partially-externalized RNA that was accessible to a dsRNA-specific monoclonal antibody . We specifically digested this RNA , extracted the fraction that remained protected inside the virion , and , by RT-PCR , we exclusively found sequences close to the 5′-end . Fig . 9 shows our model of RNA release from HRV2 triggered by incubation at 56°C together with the results lending support to the model . In the course of our experiments , we discovered that subviral particles containing ‘rod-like’ density were produced on incubation of native HRV2 at 56°C along with apparently full and empty virions . Similar uncoating intermediates with internal density taking on various conformations had also been observed in heated poliovirus [14] . We found that the particles originating from HRV2 were greatly enriched upon crosslinking double-stranded regions of the RNA prior to incubation at 56°C . This allowed for cryo-EM 3D-reconstruction , revealing that the rods were pointing approximately towards a two-fold axis and extended roughly towards a three-fold axis on the opposite side of the shell . The structural basis of this particular orientation needs further investigation . We do not know whether the orientation of the rod-like structure with respect to icosahedral symmetry is related to exit of the RNA – in the form of a single strand – through one of the channels at the two-fold axes . Nevertheless , it is conceivable , and supported by our findings , that the release process is initiated normally , i . e . the RNA starts to emerge through one of these pores , but gets stuck as soon as a double-stranded region cannot unwind . Such a scenario might lead to the build-up of rods remaining connected to the exit point that resemble RNA ‘prolate ellipsoids’ as observed in molecular dynamics simulations and small angle X-ray analysis of large RNA molecules [41] . Remarkably , upon closer analysis , the particles visually appearing to be full turned out to be a heterogeneous mixture of substantially distorted virions with a generally higher and broader distribution of RNA content ( Fig . S2 and Table S1 ) . This could indicate that RNA release from these particles was halted at an earlier stage as compared to the ‘rod-particles’ . As previously observed in the presence of antiviral agents that inhibit breathing [42] , impediment of ordered release may result in strain , causing partial denaturation/deformation of the capsid . In keeping with this hypothesis , the particles that had released more of the RNA ( i . e . the psoralen crosslinks were closer to the 5′-end ) were more homogenous as also indicated by a somewhat higher resolution attained in the 3DR despite smaller numbers of particle images ( Fig . S2 and Table S1 ) . RNA detaching from the inner capsid wall , as is expected to occur on temperature elevation , may not be able to reform its original interactions upon cooling and thus be more prone to tangling . Alternatively , kinetic bottlenecks at regions of high secondary structures could be involved in the RNA condensation observed . ‘Rod-particles’ were only rarely observed on incubation of HRV2 at pH 5 in solution ( data not shown ) ; however , when HRV2 was bound to the surface of receptor-carrying liposomes and acidified some virions possessing internal density with a similar appearance were observed [19] . Thus , it is possible that membranes are involved in their formation at low pH , which will be a topic of future investigation . On transition from the native virion to the A-particle , a massive reorganisation of RNA-protein contacts is observed ( unpublished data ) . These contacts might be essential for ordered release , possibly explaining why exposure to 56°C disturbs this tightly-coordinated process . Ordered egress of RNA suggests that the viral genome becomes organized during packaging or assembly , which may occur co-transcriptionally [43] , [44] . Therefore , it is likely that the process of encapsidation begins when the 5′-end emerges from the replication complex or at least before the complete RNA has been synthesized . It is also possible that the same applies to other viruses with ssRNA of positive polarity . This would imply that in these viruses , the 3′-end becomes encapsidated last , remaining near the capsid wall presumably in close proximity to one of the holes poised to open upon uncoating , thus resulting in a ‘last-in-first-out’ process of assembly and uncoating .
Unless otherwise indicated , all chemicals were acquired from Sigma Aldrich . 9-methoxy-7H-furo[3 , 2-g][1]benzopyran-7-on ( i . e . 8-Methoxypsoralen , 8-MOP ) was dissolved at 2 mg/ml in DMSO . Oligonucleotides ( with and without fluorescent label ) were obtained from VBC-Biotech , Vienna , Austria . Dylight 488 and FAM dye were from Thermo Scientific and dissolved in DMSO at 10 mg/ml . YOYO-1 iodide509 was from Life Technologies and dissolved in DMSO at 1 mM . Oligonucleotides for RT-PCR were as follows ( denoted as under quotation marks ) “5′-end – reverse” 5′-dAAGGGTTAAGGTTAGCCACATTCAG-3′ “5′-end – forward” 5′-dGACCAATAGCCGGTAATCAG-3′ “ M-oligo – reverse” 5′-dAAGGTGTCAGTGTTATTTATTGGTACTAGGCTG-3′ “M-oligo – forward” 5′-dGCCCCATGTGTGCAGAGTTTTC-3′ “3′-end 3 - reverse” 5′-dCCACTCATGCAAAAGCAAATC-3′ “3′-end 3 - forward” 5′-dCCTTCCCTGAAGATAAATATTTGAATCC-3′ “3′-end 5 - reverse” 5′-dCTCTGGATCACATCCAACTGCTGATCCAG-3′ “3′-end 5 - forward” 5′-dGAGTTGACTTACCTATGGTCACC-3′ Their positions on the viral genome are indicated in the scheme in Fig . 8 . All were chosen so as to have similar melting temperatures ( 50–58°C ) . HRV2 was produced and purified as detailed elsewhere [11] , [12] and incubated at ∼1 . 5 mg/ml with 8-MOP ( 50 µg/ml ) for 4 h at 37°C in 50 mM Tris-HCl ( pH 8 . 0 ) . Samples were then pipetted onto parafilm on ice and irradiated with a UV lamp ( Wilber Loumat , TFP-20L 6×15W ) at 365 nm from a distance of about 5 cm for 10 min . Control samples were treated identically except that the preincubation with psoralen was omitted . Labeling of virus with DyLight 488 was performed following the protocol of the manufacturer . Viral RNA was transcribed in vitro by using the RiboMAX Large Scale RNA Production System – T7 ( Promega ) and purified by extraction with phenol chloroform followed by ethanol precipitation . For FCS measurements , in vitro-transcribed RNA was labeled with YOYO-1 iodide509; the dye was mixed with RNA ( 10 nM ) at a molar ratio of 400∶1 ( 5 base pairs per dye molecule ) in 50 mM Tris-HCl ( pH 8 . 3 ) , 75 mM KCl , 3 mM MgCl2 and incubated at room temperature for 15 min . With the exception of experiments involving RNase digestions , all samples were supplemented with 2U RNasin/µl ( Promega ) . FCS was carried out in a Zeiss Confocor 1 instrument using an Argon-ion laser with 488 nm wavelength . The fluorescence autocorrelation function was measured for the labeled oligonucleotides complementary to the HRV2 nucleotide numbers given in brackets: oligo-dT 25-mer ( O3; between 7102 and ∼7200 ) and 5′-end ( O5; 443–468 ) in the absence or presence of in vitro-transcribed viral RNA or heated HRV2 , as indicated in the Figures . O5 was selected to be complementary to a region of low secondary structure by using the Vienna RNA package version 1 . 8 ( http://rna . tbi . univie . ac . at/cgi-bin/RNAfold . cgi ) . Diffusion coefficients were related to the measured diffusion times by using Rhodamine6G as the standard ( DRh6G = 2 . 8×10−10 m2 sec−1 ) . For easier comparison , the autocorrelation functions were normalized to 1 at autocorrelation time zero ( i . e . to one particle in the observation volume ) . Ten to 50 consecutive measurements on at least three independent samples were carried out with the respective fluorescent analytes at 10 nM ( this gave the best signal to noise ratio ) and a data acquisition time of 10 sec each , in 50 mM Tris-HCl ( pH 8 . 3 ) , 75 mM KCl , 3 mM MgCl2 . The most favorable molar ratios between oligonucleotide and RNA were determined by mixing labeled oligonucleotide at 10 nM with in vitro-transcribed HRV2 RNA at concentrations between 0 and 1000 nM ( not shown ) . This revealed that at least a 25 fold excess ( i . e . 250 nM ) of heated virus/RNA over oligonucleotide was required for an optimal hybridization signal . The mixture was incubated for 10 min at 56°C and for at least 30 sec at 4°C prior to the measurments . To confirm that the observed change of the diffusion coefficient in the presence of viral RNA was due to hybridization , the sample was treated with a mixture of RNase A ( 1 mg/ml; Roche ) and RNase H ( 0 . 5 U/µg RNA; New England BioLabs ) for 15 min in 50 mM Tris-HCl ( pH 8 . 3 ) , 75 mM KCl , 3 mM MgCl2 , 1 mM DTT at 37°C prior to the measurement ( Fig . 1B , C ) . DyLight 488-labeled HRV2 was measured at 10 nM in 50 mM Tris buffer ( pH 8 . 3 ) . In all further experiments HRV2 at about 500 nM was mixed with 10 nM of the respective labeled oligonucleotide . The mixture was incubated at 56°C , samples ( 20 µl ) were withdrawn at the times indicated in Fig . 2 and measured as detailed above . Diffusion times and percentages of free oligo , oligo bound to free RNA , and oligo bound to RNA partially released from the virus were calculated with the FCS ACCESS software ( version 1 . 0 . 12 ) by using one- , two- , or three-component fit models as indicated . Hybridization equilibrium was reached almost instantaneously as deduced from the lack of change in the diffusion time within the measurement period ( 5 min ) . For comparison with the measured values , diffusion coefficients of probe and viral RNA were calculated for hypothetic conformational states expected to result in lowest and highest diffusion coefficients ( at 22°C , with the dynamic viscosity of water μ = 0 . 955×10−3 Pa·s ) . Assuming that they adopt a densely-packed sphere with mass density of 1 . 8 g/cm3 [45] , the calculated diameters were 2 . 52 nm and 16 . 2 nm for probe and viral RNA respectively . In case of a rod-like shape , the probe ( single-stranded DNA ) was estimated to have a diameter of 1 . 1 nm and a length of 10 . 75 nm ( 25 bases , 0 . 43 nm length each ) , which will give the lowest diffusion coefficient [46] . For the viral RNA , an overall length of 3 . 85 µm ( 7102 bases , 0 . 43 nm each ) and a persistence length of about 2 nm [47] were used and resulted in a calculated random coil diameter of 60 nm . For the viral particle , the diffusion coefficient was calculated on the basis of a perfect sphere with a radius of 15 nm . Table 1 lists those calculated values comparing them to the measured data . An automated HP3D Capillary Electrophoresis System ( Hewlett Packard , Waldbronn , Germany ) equipped with an uncoated fused-silica capillary ( Composite Metal Service Ltd . , 51 . 5-cm effective , 60 . 0-cm total length , 50 µm inner diameter ) packed in a standard Hewlett Packard cassette and thermostated at 20°C was used throughout . Injection was at 50 millibar pressure for 9 sec . Between all runs the capillary was conditioned by aspirating 100 mM NaOH , water , and the background electrolyte ( BGE , consisting of 100 mM Na-borate buffer ( pH 8 . 3 ) and 10 mM Thesit ) for 2 min each , applying 950 millibar pressure . Detector signals were recorded at 205 nm . For additional measurements at 260 nm , fast spectral scanning mode was employed . Positive polarity mode ( negative pole is placed at the capillary outlet ) with 25 kV was used for all experiments . HRV2 , either psoralen crosslinked or not , was incubated at 56°C for different times as indicated in the Figures . To detect viral RNA outside the virion but still connected to it , crosslinked virus either heat-treated or not ( control ) was incubated with 1 µl ( 1 µg ) /sample anti-dsRNA mAB J2 ( English & Scientific Consulting Bt . Szirák , Hungary ) for 20 min at room temperature . To remove accessible RNA , crosslinked and heat-treated ( 10 min at 56°C ) virus was incubated with micrococcal nuclease ( MNase , 100 units/µg RNA; New England BioLabs ) at 37°C for 20 min . MNase was inactivated by addition of 10 mM EGTA . Subviral particles were quantified as total protein by integration of the peak area at 205 nm and subtraction of the respective calibrated peak area at 260 nm ( viral protein = peak area ( at 205 nm ) −1 . 4×peak area ( at 260 nm ) ) . The RNA content of the particles was estimated by comparison with native virus ( set to 100% ) ; peak areas were recorded at 205 nm and 260 nm for native virus from 12 different preparations giving a ratio of 5 . 5±0 . 1 and after subtraction of the contribution of VP4 of 5 . 3±0 . 1 ( i . e . full 135S particles ) . For crosslinked , heated , and MNase-digested particles , this ratio was 12 . 6±1 . 0 ( 4 different electropherograms ) . Based on a ratio of the extinction at 205 nm/260 nm for pure RNA of 1 . 4 , the mean RNA content of these latter particles was estimated to amount to 34 . 7% . Assuming irreversible release of the RNA ( irreversible mass action ) , the evolution of the concentrations of the three components in time was modelled by using GEPASI v . 3 . 30 [48] based on a reaction model N→V−R→R+V ( shown as lines in Fig . 6 ) . Crosslinked HRV2 either incubated at 56°C for 10 min or not ( control ) was MNase treated as above and run on a 0 . 7% agarose gel in 50 mM Tris-HCl ( pH 8 . 3 ) , 10 mM EDTA . RNA was visualized by ethidium bromide-staining and extracted from the bands with the Zymoclean Gel RNA Recovery Kit ( Zymo Research ) . Reverse transcription ( RT ) was carried out with SuperScript III reverse transcriptase ( Invitrogen ) . PCR amplification was performed with Pfu DNA polymerase ( Promega ) by using the primer sets given above for 30 cycles . For protein analysis , aliquots of the excised bands were boiled in reducing sample buffer and run on a SDS polyacrylamide gradient gel ( 12–20% ) . Proteins were detected by silver staining [49] . Psoralen UV crosslinked HRV2 was incubated at 56°C for 10 min . Negative stain EM was carried out at 56 , 000× magnification ( 80 kV ) in a FEI Morgagni 268D equipped with an 11 megapixel CCD camera ( Morada from Olympus-SIS ) after staining with 2% phosphotungstic acid . For cryo-EM , samples ( 4 µl ) were deposited on Quantifoil 400 nm copper grids with a 1 . 2/1 . 3 holey carbon film ( glow discharged in a Bal-Tec SCD Sputter Coater ( 20 mA , 1 min; Scotia ) ) , blotted with an automatic Leica EM Grid Plunger , and shock frozen in liquid ethane cooled in liquid nitrogen . Image acquisition was done as follows: Subviral particles were viewed in a FEI Tecnai F30 Polara cryo-electron microscope ( 300 kV ) with C2 condenser aperture and objective aperture of 70 µm and 100 µm , respectively , at a magnification of 71 , 949 x . Images were acquired at underfocus between 2 . 0 µm and 4 . 2 µm with a Gatan Ultrascan 4000 4k CCD camera , using Leginon automatic image acquisition software [50] . The micrographs ( 383 finally used ) were corrected for the CTF with ctffind3 [51] and downscaled to 128×128 pixels at 3 . 76 Å/pixel by using xmipp 2 . 4 [52] , [53] . 3D-maximum likelihood classification into three classes was done with relion-1 . 1 [54] on 16 , 151 particle images ( downscaled to 64×64 pixel ) by using the cryo-EM 3DR of full HRV2 A-particles ( unpublished data ) as starting map . On using an angular sampling interval of 15° , 4581 images were classified as full particles , 5285 images as ‘rod-containing’ particles , and 6285 images as empty particles . Final maps were then refined by using particle images with 128×128 pixels as selected in the previous classification , with relion-1 . 0 with and without imposing symmetry as specified in the figures [54] . The resolution of the asymmetric 3DR ( between 22 and 24 Å , as determined from the spectral signal to noise ratio - SSNR∧MAP>1; see ref . [54] ) was sufficient to identify the orientation of the ‘rod’ with respect to icosahedral symmetry . The RNA content was estimated for each ( centered ) image after normalization to mean = 0 and standard deviation = 1 by using xmipp-2 . 4 . For each class , the density corresponding to the RNA ( within a radius of 113 Å i . e . 0–15 pixels on the 64×64 pixel images ) was related to part of the density of the protein shell ( between radius 113 and 143 Å ( i . e . 15–19 pixels; see Fig . S1 ) and displayed as a histogram plot . | Viral infection requires safe transfer of the viral genome from within the protective protein shell into the host cell's cytosol . For many viruses this happens after uptake into endosomes , where receptor-binding and/or the acidic pH trigger conformational modifications or disassembly of the shell , allowing the nucleic acids to escape . For example , common cold viruses are converted into subviral particles still containing the single-stranded positive sense RNA genome; subsequently , the RNA escapes into the cytoplasm , leaving behind empty capsids . We triggered this process by heating HRV2 to 56°C and found that 3′- and 5′-end emerged with different kinetics . Crosslinking prevented complete RNA egress and upon nuclease digestion only sequences derived from the 5′-end were protected . Part of the RNA remaining within the viral shell adopted a rod-like shape pointing towards one of the two-fold axes where the RNA is presumed to exit in single-stranded form . Egress thus commences with the poly- ( A ) tail and not with the genome-linked peptide VPg . This suggests that assembly and uncoating are well-coordinated to avoid tangling , kinetic traps , and/or simultaneous exit of the two RNA ends at different sites . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"respiratory",
"infections",
"microbiology",
"viral",
"structure",
"pulmonology",
"protein",
"structure",
"rhinovirus",
"infection",
"infectious",
"diseases",
"rna",
"structure",
"proteins",
"enterovirus",
"infection",
"biology",
"molecular",
"biology",
"biochem... | 2013 | Viral Uncoating Is Directional: Exit of the Genomic RNA in a Common Cold Virus Starts with the Poly-(A) Tail at the 3′-End |
Trypanosoma evansi is mechanically transmitted by biting flies and affects camels , equines , and other domestic and wild animals in which it causes a disease called surra . At least two types of Trypanosoma evansi circulate in Ethiopia: type A , which is present in Africa , Latin America and Asia , and type B , which is prevalent in Eastern Africa . Currently , no information is available about the drug sensitivity of any Ethiopian T . evansi type . This study was conducted with the objective of determining the in vivo drug sensitivity of two T . evansi type A and two type B stocks that were isolated from camels from the Tigray and Afar regions of Northern Ethiopia . We investigated the efficacy of four trypanocidal drugs to cure T . evansi infected mice: melarsamine hydrochloride ( Cymelarsan ) , diminazene diaceturate ( Veriben and Sequzene ) , isometamidium chloride ( Veridium ) and homidium chloride ( Bovidium ) . Per experimental group , 6 mice were inoculated intraperitoneally with trypanosomes , treated at first peak parasitemia by daily drug injections for 4 consecutive days and followed-up for 60 days . Cymelarsan at 2 mg/kg and Veriben at 20 mg/kg cured all mice infected with any T . evansi stock , while Sequzene at 20 mg/kg caused relapses in all T . evansi stocks . In contrast , Veridium and Bovidium at 1 mg/kg failed to cure any T . evansi infection in mice . We conclude that mice infected with Ethiopian T . evansi can be cured with Cymelarsan and Veriben regardless of T . evansi type . In contrast , Veridium and Bovidium are not efficacious to cure any T . evansi type . Although innate resistance to phenanthridines was previously described for T . evansi type A , this report is the first study to show that this phenomenom also occurs in T . evansi type B infections .
African trypanosomoses ( AT ) are neglected parasitic diseases of humans and animals caused by various subgenera of pathogenic trypanosomes ( Trypanozoon , Dutonella and Nannomonas ) . While human African trypanosomosis ( HAT ) has reached the point where elimination is being envisaged , animal African trypanosomosis ( AAT ) is still one of the major parasitic disease constraints to animal productivity in sub-Saharan Africa causing an estimated annual loss between 0 . 7 and 4 . 5 billion USD [1–4] . In Ethiopia , AAT has been described as a major impediment to livestock development and agricultural production , contributing negatively to development in general and to food self-reliance efforts of the country in particular . Both tsetse-transmitted ( TTAT ) and non tsetse-transmitted African trypanosomiasis ( NTTAT ) are endemic to the country . TTAT are due to Trypanosoma ( T ) . congolense , T . vivax , and T . brucei brucei , whereas NTTAT are due to mechanically transmitted T . evansi and T . vivax , and the sexually transmitted T . equiperdum [5–12] . Surra is the number one protozoan disease of camels and is caused by T . evansi . Infected camels and equines may die within 3 months after onset of the disease . Moreover , cattle , water buffalo , pigs , goat and sheep infected with T . evansi suffer from immunosuppression , resulting in increased susceptibility to other diseases and vaccination failure against classical swine fever and Pasteurella multocida [13–15] . The distribution of the disease mainly coincides with that of camels in the semi-desert areas of the country [5 , 7 , 16 , 17] . The control of surra relies mainly on the use of the trypanocidal drugs: the diamidine diminazene diaceturate , phenanthridines such as homidium salts ( homidium chloride and homidium bromide ) and isometamidium chloride , and the arsenical melarsamine hydrochloride [18–22] . Isometamidium chloride is mainly used as a prophylactic drug and provides on average 3 months protection against trypanosome infection . Homidium salts have limited prophylactic properties and are mainly used as therapeutic agent [23] . Diminazene diaceturate and melarsamine hydrochloride are exclusively used as therapeutic agents [24] . Control of AT through chemotherapeutics is challenged by the emergence of drug resistance [25 , 26] . Resistance of T . congolense to isometamidium treatment has been reported in various areas of Ethiopia [11 , 12 , 27] . Similarly , Hagos and co-workers reported on the resistance of T . equiperdum against diminazene diaceturate [28] . Till present , there is no published evidence for drug resistance in Ethiopian T . evansi . However , isometamidium treatment failures in T . evansi infections have been documented in Sudan , China , the Phillipines and Venezuela [22 , 29–33] . Ethiopian T . evansi stocks are composed of at least two types that are grouped into T . evansi type A and T . evansi type B based on the restriction enzyme profile of the kDNA minicircles [34 , 35] . T . evansi isolates with minicircle type A usually have the RoTat 1 . 2 variable surface glycoprotein ( VSG ) and are the most abundant in East and West Africa , Latin America and Asia [5 , 35–39] . In contrast , T . evansi type B is less common and so far has only been isolated from camels in Chad , Kenya and Ethiopia [5 , 34 , 35 , 39–41] . In a former study , we isolated T . evansi type A and type B stocks from camels in the Afar and Tigray regions in Northern Ethiopia [5 , 41] . The present study was undertaken to investigate the in vivo drug sensitivity profiles of some of these T . evansi stocks in mice with regard to diminazene diaceturate , isometamidium chloride , homidium chloride and melarsamine hydrochloride .
Handling and use of experimental mice was approved by the College of Veterinary Medicine , Mekelle University ( CVM-CRC/21/08 ) , in line with the National Research Ethics Review Guideline of the Ethiopian Ministry of Science and Technology , Addis Ababa , 2014 . For this study , we used two T . evansi type A ( MCAM/ET/2013/004 and MCAM/ET/2013/009 ) and two T . evansi type B ( MCAM/ET/2013/010 and MCAM/ET/2013/014 ) stocks , that we previously isolated from dromedary camel in Tigray and Afar , Northern Ethiopia [41] . All four stocks were typed as dyskinetoplastic trypanosomes based on absence of amplification of kDNA maxicircle targets . In addition , MCAM/ET/2013/009 is a natural akinetoplastic stock based on absence of kDNA minicircle amplification and loss of kinetoplast DAPI staining [41] . Trypanosome cryostabilates were thawed in a water bath at 37°C for 5 min , mixed with an equal volume of phosphate buffered saline glucose ( PSG; 7 . 5 g/l Na2HPO4 . 2H2O , 0 . 34 g/l NaH2PO4 . H2O , 2 . 12 g/l NaCl , 10 g/l D-glucose , pH 8 ) and checked for viability and motility of trypanosomes using microscopy . Swiss albino female mice of 6–8 weeks old and weighing between 25 and 30 g , obtained from the laboratory animal facility of the College of Veterinary Medicine of Mekelle University , were inoculated intraperitoneally ( IP ) with 0 . 2 ml of the trypanosome suspension . The parasitemia was monitored following the Matching Method , i . e . 5 μl of blood was transferred onto a microscope slide , covered with a 24x24 mm cover slip , examined at 40x10 magnification and the number of parasites per field of view were estimated and converted to parasites per ml of blood [42] . At peak parasitaemia , the mice were anaesthetised and exsanguinated by heart puncture with a heparinised syringe . Blood was diluted in PSG to a concentration of 2 trypanosomes per field ( about 8x107 trypanosomes/ml ) prior to use for in vivo drug sensitivity testing . Per experimental group , 6 mice were inoculated intraperitoneally ( IP ) with 2 . 5x107 living trypanosomes in PSG . Infection of each animal was confirmed individually by microscopy one day before treatment . Treatment started at day 4 post-infection and consisted of daily IP injections for 4 consecutive days with 0 . 1 ml/10g body weight ( BW ) 0 . 9% NaCl saline solution containing the appropriate concentration of drug . Five trypanocidal drugs were tested in this study . Melarsamine hydrochloride ( MelCy; Cymelarsan ) , isometamidium chloride hydrochloride ( ISM; Veridium ) , and diminazene diaceturate ( DIM; Veriben ) were procured in Europe . Diminazene diaceturate plus phenazone ( DIM-SEQ; Sequzene ) and homidium chloride ( HOM; Bovidium ) were procured from the local market in Shire Endaselasse , Western zone of Tigray regional state . All drugs , except MelCy , were assessed by the Animal Products , Veterinary Drug and Feed Quality Assessment Center in Addis Ababa ( Ethiopia ) for adherence to the physiciochemical characteristics stated by their manufacturers . The scientific name , trade name , origin and dosage of the drugs are presented in Table 1 . We tested the following doses: 0 . 125 mg/kg BW and 2 mg/kg BW MelCy , 1 mg/kg BW ISM , 20 mg/kg BW DIM or DIM-SEQ and 1 mg/kg BW HOM [28 , 43 , 46] . The control group consisted of infected mice that received 0 . 2 ml of saline solution [43] . Two days after the last treatment and subsequently once a week until day 60 post- treatment , each mouse was examined with the Matching Method for the presence of parasites . To detect subpatent parasitaemia , survivor mice were immunosuppressed with cyclophosphamide at 200 mg/kg BW ( Endoxan , Baxter , Lessines , Belgium ) 25 days post-treatment [47] . Relapsing mice were euthanised . At day 60 post-treatment , all mice that remained negative in microscopy , were tested by the microhaematocrit centrifugation technique ( mHCT , 4 tubes per mouse ) [48] . If negative in mHCT , all surviving mice were euthanised and their blood was collected on heparin by heart puncture . The blood of all mice from each group was pooled and run over a mini Anion Exchange Centrifugation Technique ( mAECT ) column to detect subpatent parasitaemia [49] . If negative in mAECT , the mice were considered to be cured .
Detailed data on the outcome of the mice after infection and treatment are given in S1 Table . All infected mice treated with 0 . 9% saline ( controls ) died between the onset of treatment and two days after treatment . Table 2 shows the observed number of relapses and the average day after treatment that relapses occurred . MelCy at 0 . 125 mg/kg BW cured only 2 out of 6 mice infected with MCAM/ET/2013/004 ( type A ) and none of the mice infected with MCAM/ET/2013/014 ( type B ) . Therefore , this dose was not administered to the mice infected with the two other stocks . MelCy at a higher dose ( 2 mg/kg BW ) and DIM at 20 mg/kg BW cured all mice infected with any T . evansi stock . Treatment with DIM-SEQ at 20 mg/kg BW caused relapses for all T . evansi stocks . HOM at 1 mg/kg BW failed to cure any mouse infected with any T . evansi stock , while ISM at 1 mg/kg BW cured 4 of the 6 mice infected with MCAM/ET/2013/10 and none of the mice infected with the other stocks . No particular difference was apparent in parasitemia during pretreatment and relapse , between the T . evansi type A and type B stocks .
We conclude that Ethiopian T . evansi can be treated in mice by diminazene and MelCy regardless of T . evansi type and presence of kinetoplast . However , measures should be taken by the Ethiopian Veterinary Drug and Animal Feed Administration and Control Authority ( VDAFACA ) to create market access to Cymelarsan , which is currently not registered in Ethiopia , and to ensure consistent quality of commercial drug formulations that are available from the local markets . A recent study found that 27 . 3% of the diminaze diaceturate formulations and 29 . 4% of the isometamidium chloride formulations failed to comply with quality requirements as assessed in HPLC [71] . Furthermore , the phenanthridines isometamidium chloride and homidum salts are DNA intercalating agents that raise serious concerns of mutagenicity and are not well tolerated by camels [72] . Unfortunately , they are still in use for treating animals that provide beef and milk for human consumption [18 , 59 , 73] . | Surra is a vector borne disease in camels , horses , water buffaloes , cattle and other domestic animals caused by Trypanosoma ( T . ) evansi . This protozoan parasite is transmitted by biting flies such as tabanids and stable flies and is endemic in many countries in Northern and Eastern Africa , Latin America and Asia . Surra is responsible for high economic losses due to mortality and morbidity of draught animals and leads to animal trade restrictions in endemic regions . Control of surra is mainly based on the treatment of sick animals presenting clinical symptoms . In Ethiopia two different types of T . evansi ( A and B ) have been described , yet no data existed about the drug sensitivity of any T . evansi type . In this study , we show for the first time that T . evansi type B is naturally in vivo resistant to the phenanthridine class of trypanocidal drugs , a phenonomen that was previously described for T . evansi type A . All Ethiopian T . evansi types are sensitive to melarsamine hydrochloride and diminazene diaceturate . Unfortunately , the most efficacious drugs are either not registered in Ethiopia or escape quality control of the active substance in commercial drug formulations . Furthermore , the inefficacious drugs remain accessible on the market despite their toxicity for animals . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"chemical",
"compounds",
"geographical",
"locations",
"vertebrates",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"mammals",
"animals",
"protozoans",
"pharmaceutics",
"ethiopia",
"cellular",
"structures",
"... | 2018 | Isometamidium chloride and homidium chloride fail to cure mice infected with Ethiopian Trypanosoma evansi type A and B |
The hair follicle system represents a tractable model for the study of stem cell behaviour in regenerative adult epithelial tissue . However , although there are numerous spatial scales of observation ( molecular , cellular , follicle and multi follicle ) , it is not yet clear what mechanisms underpin the follicle growth cycle . In this study we seek to address this problem by describing how the growth dynamics of a large population of follicles can be treated as a classical excitable medium . Defining caricature interactions at the molecular scale and treating a single follicle as a functional unit , a minimal model is proposed in which the follicle growth cycle is an emergent phenomenon . Expressions are derived , in terms of parameters representing molecular regulation , for the time spent in the different functional phases of the cycle , a formalism that allows the model to be directly compared with a previous cellular automaton model and experimental measurements made at the single follicle scale . A multi follicle model is constructed and numerical simulations are used to demonstrate excellent qualitative agreement with a range of experimental observations . Notably , the excitable medium equations exhibit a wider family of solutions than the previous work and we demonstrate how parameter changes representing altered molecular regulation can explain perturbed patterns in Wnt over-expression and BMP down-regulation mouse models . Further experimental scenarios that could be used to test the fundamental premise of the model are suggested . The key conclusion from our work is that positive and negative regulatory interactions between activators and inhibitors can give rise to a range of experimentally observed phenomena at the follicle and multi follicle spatial scales and , as such , could represent a core mechanism underlying hair follicle growth .
Hair is a characteristic feature of mammals and performs a variety of roles , such as thermal insulation , physical protection , camouflage , social interaction and sensory perception [1] . The relative importance of the different functions of hair depend on a host of factors ( e . g . local environment ) and it is often crucial that an individual can adapt its coat accordingly . Such control is perhaps most evident in the periodic shedding of fur in response to seasonal changes [2] . The base of a hair resides in an approximately cylindrically shaped , multicellular mini-organ called a hair follicle that is invaginated in the surface of the skin . Unlike the hair itself , which is composed of dead keratinocytes , hair follicles undergo a process of cyclical regeneration , regulated by an intrinsic clock as well as other extrinsic mechanisms [2] , that allows for the localised growth of individual hairs . The inner surface of the follicle is lined by epithelial cells and its rate of regeneration is ultimately controlled by the rate at which follicle stem cells exit their quiescent state and become activated . The follicle growth cycle is traditionally split into three phases: anagen and catagen , when growth and involution occur , respectively , and telogen , a quiescent phase when the follicle is either refractory or awaiting re-entry into anagen [1] , [3] . A follicle undergoes substantial morphological changes as the cycle progresses ( see Figure 1 ) : during telogen , the dermal papilla , a mesenchymal tissue at the proximal end of the follicle , is in close proximity to a stem cell niche that resides in a spatial region known as the follicle bulge . Upon anagen entry , stem cells in the bulge proliferate and generate transit amplifying cells , and the proximal end of the follicle ( including the dermal papilla ) extends proximally . As anagen progresses the transit cells differentiate and form the new hair shaft . Transition to catagen results in a rapid bout of apoptosis , the proximal end of the follicle involutes and the dermal papilla returns again to a position in close proximity to the follicle bulge . During telogen the morphological features of the follicle remain relatively conserved . Although it has been established that the hair follicle clock is controlled by interactions local to the hair follicle [2]–[5] and a large number of different extrafollicular signals ( e . g . hormones , neuropeptides , growth factors ) are known to impact upon follicle growth ( see Figure 2 ) , the fundamental interactions underlying the follicle clock remain elusive [1] , [2] , [6]–[9] . However , specific molecular pathways that become activated in different phases of the follicle cycle have been identified ( BMP , Wnt , Fgf and TGF ) and have been shown to , at least partially , control follicle growth dynamics [10] . For instance , using the transgenic mice KRT14-Wnt7a and KRT14-Nog , the Wnt and BMP pathways have been identified as activators and inhibitors of localised follicle growth , respectively [11] . These results were further corroborated using coated bead implants in wild-type mice . Notably , the Wnt and BMP pathways cycle out of phase with one another , with BMP activity high during refractory telogen and Wnt during anagen [11] ( see schematic diagram in Figure 3 ) . Whilst the close correlation between anagen/telogen and Wnt/BMP pathway activity has led to speculation that interaction between the Wnt and BMP pathways might provide a potential mechanism that governs the follicle cycle [8] , recent observations in which members of the Fgf and TGF signalling pathways have also been shown to perturb follicle growth suggest that regulation of the hair follicle cycle in mouse is mediated via multiple different molecular pathways . As illustrated in Figure 1 , a follicle has a number of physically distinct regions that can influence the proliferation of stem cells in the follicle bulge . For instance , the dermal papilla is a source of Wnt ligands but it is also maintained in anagen by Wnt3a and Wnt7a ligands [4] . Furthermore , when stabilized catenin is artificially elevated in resting stem cells , hair follicles are precociously induced to begin a new round of hair growth [8] , [12]–[14] . In contrast , cyclic BMP expression has been observed in adipocytes that reside in extrafollicular space [11] and it is thought that high levels of BMP signalling can maintain bulge stem cells in a quiescent state during telogen . For instance , Kobielak et al . have shown , via conditional ablation of a BMP receptor gene , that the BMP pathway inhibits the initiation of the hair cycle [15] . Moreover , BMP activity is stimulated by anagen progression which itself is stimulated by activator expression . Combined , these observations are suggestive of dermal papilla/Wnt mediated positive feedback signalling that results in the activation of stem cell proliferation in the bulge stem cell region and hence follicle growth , and a negative feedback loop in which anagen-inducing activators cause , perhaps indirectly , the production of telogen-inducing inhibitors , which themselves inhibit the activators of follicle growth . We again highlight that , although the influence of particular activators and inhibitors on hair growth is beginning to become better understood , precisely how the activators and inhibitors of follicle growth interact with one another has not yet been well characterised . At the multi follicle scale , it has been observed that follicles can either make the transition from telogen to anagen autonomously or via induction by neighbouring follicles that have themselves just entered anagen [11] . The former mechanism introduces stochasticity into follicle growth dynamics , as is evident from the random initiation sites of hair growth observed in vivo [11] . The latter mechanism allows coordinated behaviour amongst populations of follicles , resulting in the propagation of waves of hair growth ( see Figure 4 ) , and is thought to be mediated by the diffusion of activators and/or inhibitors [16] . Furthermore , long range signalling has been demonstrated in experiments where beads coated in activators/inhibitors have been shown to promote/inhibit follicle growth over extended spatial distances [11] . Whilst the growth cycle of a single hair follicle is dependent on coupled physical ( e . g . cell movement ) , biochemical ( multiple pathways ) and biological ( e . g . apoptosis and cell proliferation ) processes , the treatment of a single follicle as a functional unit has allowed Plikus et al . [11] , [16] to probe the nature of the follicle cycle in a ‘top down’ manner . For instance , quantification of the cycle resulted in the proposal that there are four functional stages in the follicle growth cycle: propagating anagen ( P ) , when a follicle can induce neighbouring follicles in telogen to enter anagen; autonomous anagen ( A ) , when follicles can no longer communicate with their neighbours but are still in the growth phase; refractory telogen ( R ) , when a follicle is no longer undergoing growth and neither influences nor is influenced by its neighbours; and competent telogen ( C ) , when a follicle can either enter anagen spontaneously or be induced to do so by neighbours in propagating anagen . The times individual follicles spend in the different phases of the hair cycle have been measured ( see Table 1 ) and these data used to parameterise a phase-structured model ( which will be denoted by PARC ) of the follicle growth cycle . As well as quantifying the excitable dynamics of individual follicles , Plikus et al . have exploited the coupling between anagen and the production of pigmentation [7] in order to obtain a spatial readout of the temporal dynamics of follicle activity ( the pigment is macroscopically observable on the surface of a clipped animal thus giving a readout of which follicles are in , or have recently been in , anagen ) . The results from such follicle population scale experiments can be seen in Figure 4 where the follicle growth patterns observed in Wnt over-expression ( KRT14-Wnt7a ) and BMP down-regulation ( KRT14-Nog ) mice models are compared . Simulations at the multi follicle scale have previously been modelled using cellular automata [11] , [16]–[18] . Halloy et al . [17] , [18] , considering human hair growth dynamics , originally proposed a follicular automaton model in which measurements of the functional phases of the hair cycle were used to parameterise a cellular automaton model . The phenomenon of inter follicle coupling was neglected as it is thought to play a negligible role in human hair growth dynamics . In contrast , communication between neighbouring follicles in mice is well established and , accordingly , Plikus et al . [11] , [16] developed a cellular automaton model of mouse follicles that accounted for local coupling between neighbouring follicles . Plikus et al . also used experimental measurements of times spent in different phases of the hair cycle to parameterise the automaton model and simulated how variation in the behaviour of individual follicles was manifest at the population scale . This approach provided a computational architecture in which to relate follicle scale quantities , such as the mean time spent in R phase , to emergent patterns at the population scale , both in individual organisms and across different species . Moreover , the model produced a range of patterns that exhibited many features in common with experimental observations: wave propagation , spontaneous excitation , border stability and instability ( under different conditions ) . When key parameters in the model , such as the probability of spontaneous excitation , were varied , the emergent patterns varied in similar ways to experimental observations . Whilst previous cellular automaton models of hair follicle growth provided a useful framework in which to integrate various experimental data and investigate hypotheses [11] , [16]–[18] , their primary limitation is that the automaton rules are chosen to simulate experimental observations and , hence , are not motivated by underlying mechanisms . Moreover , it can be difficult to meaningfully relate the automaton rules to the increasing amount of experimental data becoming available at the molecular scale . The goal of this study is to develop a model that can begin to bridge the three scales of observation ( molecular , single follicle and multi follicle ) in the hair follicle system . We demonstrate how populations of hair follicles can be described using a classical excitable medium framework [19] , with the mechanisms that control certain features of the follicle dynamics , such as the excitability threshold and length of the different phases in the PARC model , related to regulation by activators and inhibitors of follicle growth . The layout is as follows: firstly , we briefly introduce the well-established theory of excitable media and describe a minimal model of the hair follicle system; secondly , we present simulation results and compare them with experimental observations; thirdly , we describe how a number of model predictions could be tested experimentally; and , finally , we conclude with a summary and discussion .
Before discussing the specifics of the hair follicle system , we provide a brief introduction to the theory of excitable media , a field of study that is used to describe a disparate range of fundamental phenomena in biology , such as nerve signal propagation [20] , [21] , electrical activity in the heart [22] , calcium dynamics [23] and dictyostelium aggregation [24] . Whilst the underlying chemical/ionic equations for particular systems are often highly nonlinear , the essence of the phenomenon of excitability can be understood using much simpler models . For example , consider a two-variable activator-inhibitor system that has a single stable steady-state ( 1 ) where both activator and inhibitor activities are low ( see phase plane diagram in Figure 5 ) . Making the further assumption that the activator activity changes on a much faster time scale than that of the inhibitor , a perturbation of sufficient magnitude ( 12 ) can result in the fast activation of the activator and the system moves to a transient state of high activator activity ( ) . However , a consequence of high activator activity is that the inhibitor slowly gets activated ( ) and eventually causes a fast deactivation of the activator ( ) . The system remains in a refractory state until the inhibitor activity returns to steady-state levels ( ) , whence the excitable cycle is complete and competent for a further activation . The central tenet of this study can be described as follows: using an excitable medium framework , a follicle's state is represented by two variables , an activator and an inhibitor of follicle growth . The activator and inhibitor values are correlated with , but not explicitly representative of , the concentrations of known activators and inhibitors of follicle growth , such as members of the Wnt and BMP pathways , respectively . The dynamics of the activators and inhibitors can be described as follows: a follicle has a stable steady-state in which activator and inhibitor activities are low ( see Figure 5 ) . The follicle can become excited ( ) , either stochastically or by interaction with neighbours , and activator activity increases on a fast time scale . Activator activity corresponds to anagen so , whilst the activator activity is high , the follicle grows . However , the inhibitor activity increases on a slow time scale ( ) and eventually turns the activator off , thus follicle growth is halted ( ) . At this stage , inhibitor activity is still high and the follicle is in the refractory phase , i . e . it cannot be induced back into the growth cycle . The inhibitor then decays on the slow time scale ( ) and , eventually , the follicle returns to the competent phase , where upon another growth cycle can be induced upon appropriate perturbation . But is there experimental evidence in support of the aforementioned hypothesis ? At the multi follicle scale , it is clear that patterns of hair follicle growth share many features observed with patterns arising in excitable media ( e . g . wave propagation , spontaneous excitation , border stability and instability , thresholding ) . At the individual follicle scale , the functional phases of the hair follicle cycle described by Plikus et al . [11] ( propagating anagen , autonomous anagen , refractory telogen , competent telogen ) have the properties one expects from an excitable system ( excitability , propagation , refractoriness ) . Moreover , stochasticity in the hair follicle cycle occurs predominantly in competent telogen ( see Table 1 ) , which is precisely the behaviour one expects in an excitable system . At the molecular scale the picture is less clear , although the sequence of activations of the Wnt and BMP pathways ( see schematic illustration in Figure 3 ) is consistent with the dynamics of an activator and inhibitor in an excitable medium . Moreover , there is evidence , as described in the introduction , of positive feedback in the Wnt pathway dynamics and negative feedback between BMP and Wnt , interactions that might play a role in the emergence of excitability . In the modelling work that follows we will consider a scale of description at which the follicle is treated as a functional unit and develop a caricature description of activator and inhibitor dynamics at the single follicle scale .
In this section we suggest a number of further experiments which could help to further determine the excitable properties of the hair follicle system .
Recent experimental work in the hair follicle system has allowed the gathering of information across a range of spatial scales: at the molecular scale , numerous pathways have been shown to activate and inhibit follicle growth; at the single follicle scale , hair plucking assays have allowed quantification of the time spent in the different phases of the follicle cycle; and at the multi follicle scale , hair clipping assays have allowed the characterisation of population scale behaviours , such as wave propagation . The interdependence between the different scales is only beginning to become understood . A previous model of mouse hair follicle growth proposed by Plikus et al . related the individual and multi follicle population scales [11] , [16] . The hair plucking assay data were used to parameterise the PARC model and the simulation of populations of follicles allowed investigation of the interplay between the characteristic times spent in the different phases of the clock cycle and emergent patterns . Similarly , Halloy et al . [17] , [18] previously considered a stochastic follicle automaton model of hair growth in humans . Notably , human hair growth patterns do not exhibit the same wave-like growth patterns as mice and it is thought that inter-follicular coupling is either not present or at least much weaker than in mice . Given recent advances in understanding of the molecular regulators of hair follicle growth , a disadvantage with cellular automaton frameworks is that it is not obvious how to relate the PARC phase times to observations at the molecular scale . In this study we propose a stochastic , two-variable , activator-inhibitor model of mouse hair follicle growth dynamics . An important feature of the model is that the functional phases of the hair follicle cycle are emergent , thus allowing us to relate hair plucking measurements at the single follicle scale to underlying molecular regulation . Whilst the two-variable description of molecular events is undoubtedly an abstraction , we believe it is justified in the present case for the following reasons: ( a ) although the molecular pathways underlying follicle growth are becoming increasingly better understood , the current level of description is qualitative at the molecular scale , making the parameterisation of detailed molecular models difficult; ( b ) the model is tractable and can thus help to develop insight into how measured effects at different spatial scales are inter-related; and ( c ) the model can be formulated in a manner allowing comparison with both previous models and experimental observations . We anticipate that increasing quantification at the molecular scale will enable our description of underlying molecular interactions to be fine-tuned in future iterations of the model . After developing an excitable , stochastic model of a single follicle , we considered a two-dimensional field of diffusively-coupled follicles , as previously suggested by Plikus et al . [11] . Model simulations exhibited many features in common with experimental observations: activator-inhibitor dynamics in qualitative agreement with known activators and inhibitors of follicle growth; stochastic , spontaneous initiations causing a single follicle to pass through the excitability threshold; propagation through the excitable medium of single waves originating from a single excitation; border stability when an excitation occurs close to a refractory region; border instability when a border separates a region of excited and competent follicles; and the emergence of regions of localised activation upon simulation of an activator-coated bead . We propose that an advantage of the current framework is that it allows one to investigate how changes at the molecular scale might give rise to different patterning phenotypes . In the KRT14-Wnt7a mouse , the activator Wnt7a is constitutively over-expressed and Plikus et al . observed decreased refractory and competent phase times , an increased spontaneous initiation rate , faster excitation waves and the emergence of target-like patterns . Notably , the constitutive over-expression of Wnt7a did not destroy the follicle cycle as the different functional phases were still distinguishable . To the best of our knowledge , there is no well-understood mechanism describing why the observed patterns arise in this particular mutant . We set about trying to investigate the KRT14-Wnt7a phenotype within the proposed framework and found that an increase in the activator positive feedback strength resulted in decreased refractory and competent telogen times at the follicle scale . We then investigated the effect of the increased production rate at the follicle population scale and found an increased wave velocity and greater propensity for stochastic excitations . Intriguingly , the population scale patterns changed from being single waves of excitation to target patterns . Notably , the target patterns arise as a consequence of diffusive coupling acting over multiple follicles , a behaviour that represents a significant deviation from the previous PARC model , where coupling only occurred between neighbouring competent and anagen follicles . Furthermore , we have demonstrated that coincident increased activator and decreased inhibitor production rates yield a shorter refractory phase and an oscillatory follicle , and have suggested that such a change in follicle stability might be responsible for population scale observations in the KRT14-Nog mouse . A notable feature of our simulations is that competent telogen times must be of the order of days such that the frequency of stochastic excitations across a population of follicles is comparable to the population scale patterns measured by Plikus et al . However , at the single follicle scale Plikus et al . have measured competent telogen times in the range of 0–60 days . In fact , when we used these much shorter competent telogen times the simulations are dominated by stochastic excitations in a manner inconsistent with population scale measurements from wild-type mice ( data not shown ) . In order to resolve this apparent conflict we highlight that: ( a ) the mouse system is dominated by the nearest-neighbour propagation mechanism of anagen initiation , thus placing an upper bound on the observation range for stochastic excitations ( i . e . after 60 days a propagating wave has excited a given follicle , thus biasing the observation range of stochastic excitation events ) ; and ( b ) the proposed model predicts an exponential distribution of competent telogen times , hence it is exponentially less likely to observe longer competent telogen times than shorter ones . In summary , viewed through the theoretical framework proposed in this study , competent telogen in mice should be much more stable than one might immediately infer from previous measurements . This prediction could be investigated in experimental work in which the inter-follicular communication mechanism is disrupted . The model presented in this study has a number of limitations . Firstly , we do not have direct estimates of molecular parameters , such as decay rates and cross-activation and -inhibition rates of activators and inhibitors . Secondly , we note that in vivo geometries have both periodicity and boundary effects that might influence emergent patterns in real systems . Thirdly , in the mouse mutants the time spent in anagen does not vary while in our model this depends strongly on parameters such as the inhibitor production rate ( data not shown ) . Fourthly , diffusion and stochastic effects have been modelled only for the activator dynamics . These details could also be introduced into the inhibitor dynamics but at the expense of further model complication . Finally , our model does not explicitly account for biophysical changes that occur in a follicle as the hair cycle progresses or sub-follicular structures such as dermal papillae . However , whilst it will be important to account for the aforementioned limitations in future iterations of the model , it is our belief that the central thesis of this study , that populations of hair follicles can be treated as an excited medium , will remain unchanged once these limitations have been addressed . We envisage that in the same manner as the Fitzhugh-Nagumo equations can be used as a caricature description of the dynamics of action potential propagation in cardiac tissue , the model proposed in this study might provide a caricature description of hair follicle growth propagation . Before the proposed model is embellished to account for further details of underlying hair follicle biology , there are a number of conceptually simpler experiments that could allow us to further validate the central thesis of this study . Firstly , the separation of time scales in the model allows a clear distinction between excited and competent phases of the cycle . In our model , the activator changes on a much faster time scale than the inhibitor and this is observable by much larger spatial gradients in the activators . Secondly , a ubiquitous feature of excitable media is the presence of spiral waves . If the hair follicle system is an excitable medium , one would expect that particular initiations of oscillators would result in the development of propagating spirals . Finally , excitable media typically exhibit a thresholding property whereby a stimulus of a sufficiently large magnitude is required to excite a given follicle . Hence , one would expect that such a threshold could be identified by examining the behaviour of beads coated with different activator concentrations . On a concluding note , the regulation of regeneration and renewal is a key characteristic of any homeostatic biological system . In this study we have coupled hair follicle growth to the activity of activator in an excitable medium , a hypothesis that seems particularly attractive given that growth occurs only on a transient time scale . We expect that if further substantiated in the hair follicle system , there may be other instances where an excitable medium framework can be used as a mechanism for regulating regeneration in homeostatic systems . | Although the molecular interactions that regulate the follicle growth cycle have begun to be uncovered , the fundamental interactions that regulate periodicity remain elusive . In this study we develop a model in which we neglect biophysical effects ( and hence morphological changes ) by treating each follicle as a functional unit . We then describe caricature interactions at the follicle scale which have the property that a field of coupled follicles can be treated as an excitable medium . We perform a range of simulations that demonstrate qualitative agreement with experimental observations . Furthermore , the modelling results suggest a regulatory mechanism that might represent a key underlying principle in the regulation of hair growth . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"systems",
"biology",
"biochemical",
"simulations",
"mathematics",
"theoretical",
"biology",
"biology",
"computational",
"biology",
"nonlinear",
"dynamics"
] | 2012 | Modelling Hair Follicle Growth Dynamics as an Excitable Medium |
In 404 Lepob/ob F2 progeny of a C57BL/6J ( B6 ) x DBA/2J ( DBA ) intercross , we mapped a DBA-related quantitative trait locus ( QTL ) to distal Chr1 at 169 . 6 Mb , centered about D1Mit110 , for diabetes-related phenotypes that included blood glucose , HbA1c , and pancreatic islet histology . The interval was refined to 1 . 8 Mb in a series of B6 . DBA congenic/subcongenic lines also segregating for Lepob . The phenotypes of B6 . DBA congenic mice include reduced β-cell replication rates accompanied by reduced β-cell mass , reduced insulin/glucose ratio in blood , reduced glucose tolerance , and persistent mild hypoinsulinemic hyperglycemia . Nucleotide sequence and expression analysis of 14 genes in this interval identified a predicted gene that we have designated “Lisch-like” ( Ll ) as the most likely candidate . The gene spans 62 . 7 kb on Chr1qH2 . 3 , encoding a 10-exon , 646–amino acid polypeptide , homologous to Lsr on Chr7qB1 and to Ildr1 on Chr16qB3 . The largest isoform of Ll is predicted to be a transmembrane molecule with an immunoglobulin-like extracellular domain and a serine/threonine-rich intracellular domain that contains a 14-3-3 binding domain . Morpholino knockdown of the zebrafish paralog of Ll resulted in a generalized delay in endodermal development in the gut region and dispersion of insulin-positive cells . Mice segregating for an ENU-induced null allele of Ll have phenotypes comparable to the B . D congenic lines . The human ortholog , C1orf32 , is in the middle of a 30-Mb region of Chr1q23-25 that has been repeatedly associated with type 2 diabetes .
Type 2 diabetes ( T2D ) afflicts ∼246 million people worldwide , including ∼21 million in the United States ( 7% of the population ) ; another 54 million Americans are pre-diabetic . If the incidence of T2D continues to increase at the present rate , one in three Americans , and one in two minorities born in 2000 will develop diabetes in their lifetimes [1] . Direct medical costs associated with diabetes in the United States exceed $132 billion a year [2] , and consume ∼10% of health care costs in industrialized nations . Peripheral hyporesponsiveness to insulin increases metabolic demands on the insulin-producing β-cells of the pancreatic islets . Many obese individuals are insulin-resistant , but do not become overtly diabetic provided that the increased demand for insulin is effectively met [3] , [4] . However , if β-cell mass and/or function are insufficient to meet this requirement , overt hyperglycemia and T2D ensue [5] . In autopsy series of subjects with T2D , total β-cell mass is decreased [6] , [7] . Primary reductions of β-cell mass predispose to diabetes in rodent models [8] , [9] , [10] and in autosomal dominant forms of diabetes ( e . g . , MODY; maturity onset diabetes of youth ) [11] . Such primary reductions might predispose to some instances of T2D . Susceptibility to T2D is strongly inherited as evidenced by the >80% concordance rates in monozygotic twins [12] , [13] , [14] , [15] , familial aggregation , and ethnic predispositions [16] . Heritability of sub-phenotypes related to T2D , e . g . insulin resistance and β-cell hypofunction is even higher [17] . Environmental factors are also important [17] , [18] . Although several genes for relatively rare monogenic forms of diabetes , including MODY , syndromic ( Wolfram syndrome ) , lipoatrophic , and mitochondrial-inherited diabetes have been identified [2] , [19] , the underlying genetic bases for the genetically complex T2D , accounting for >95% of diabetes patients , have remained elusive . The identification of susceptibility genes is made difficult by the polygenic nature of the phenotype [20] , its reflection of convergent , distinct metabolic processes producing identical phenotypes ( phenocopies ) , and the potent gene-gene and gene-environment ( e . g . obesity ) interactions that characterize the disease . Clear genetic influences on the endophenotypes ( intermediate phenotypes ) of β-cell mass/function and insulin resistance have been shown , and vary among racial groups . [21] , [22] , [23] , [24] . Some notable earlier successes ( e . g . PPARG , CAPN10 ) , and a recent series of genome-wide association studies of large numbers of well-phenotyped subjects [25] , [26] , [27] , [28] , [29] , [30] , [31] have identified T2D susceptibility loci/genes ( e . g . TCF7L2 ) whose functions with regard to the implicated phenotypes are uncertain . As no single implicated gene or allele accounts for more than a small fraction of risk to develop T2D , there are still many genes/molecular mechanisms awaiting identification . In mice , there is striking strain-dependent susceptibility to T2D in the context of obesity [32] . We exploited the differential diabetes susceptibilities of the B6 and DBA strains segregating for the obesity mutation Lepob [32] to identify a diabetes susceptibility QTL in B6xDBA progeny and then used congenic lines derived from the implicated interval to clone a candidate gene accounting for the QTL . Similar strategies have been used to identify QTLs ( and responsible genes ) for other complex phenotypes in mice [33] such as type 1 diabetes [34] , diet-induced obesity [35] , tuberculosis susceptibility [36] , atherosclerosis [37] , epilepsy [38] , schizophrenia [39] and , also , T2D [40] , [41] , [42] , [43] . We identified , “Lisch-like” ( Ll ) , a novel gene , encoding multiple , tissue-specific transcripts in brain , liver and islets . The functional consequences of the hypomorphic DBA allele ( diabetes-prone ) in Lepob/ob mice appear to be late embryonic to early postnatal reductions in β-cell mass due to diminished rates of β-cell replication , some “catch-up” of β-cell mass by 2–3 months , followed by mild glucose intolerance at >6 months of age . These phenotypes are recapitulated in mice with an ENU-induced null allele of Ll .
We identified a QTL for diabetes-related phenotypes in obese F2 and F3 progeny of an intercross between diabetes-resistant C57BL/6J ( B6 ) and diabetes-susceptible DBA/2J ( DBA ) mice segregating for Lepob . Phenotypes including fasting blood glucose , HbA1c and islet histology mapped with LOD >8 around D1Mit110 on distal Chr 1 at 169 . 6 Mb ( details in Methods: Mapping T2D-related Phenotypes ) . By producing congenic and sub-congenic B6 . DBA lines also segregating for Lepob , we refined the interval to 5 . 0 Mb between rs31968429 at 168 . 1 Mb and rs31547961 at 173 . 1 Mb where all four congenic lines overlap for DBA ( Figure 1; details in Methods: B6 . DBA Congenic Lines: Creation and Fine Mapping ) . We further restricted the search ( Figure 1 ) by identifying a haplotype block [44] conserved between B6 and DBA that extends 3 . 2 Mb from rs30708865 at 169 . 9 Mb to rs31547961 at 173 . 1 Mb . Only eleven unvalidated B6 vs . DBA single nucleotide polymorphisms ( SNPs ) in this interval are listed in the Mouse SNP database ( www . ncbi . nlm . nih . gov/SNP/MouseSNP . cgi ) ; however , among fragments we could amplify containing nine of these putative SNPs , we detected no sequence variants . Moreover , we found no coding sequence/expression difference between B6 and DBA among all genes and transcripts in the “conserved” interval by computation , direct sequencing , and quantitative mRNA expression analysis . Thus , it is unlikely that the variant ( s ) in the genetically-defined interval with peak at 169 . 6 Mb mediating differential diabetes susceptibility between these two strains is within the “conserved region . ” We sequenced the 3 kb interval between rs31968429 and rs33860076 at the centromeric end of subcongenic line 1jcdt and detected no variants between the two strains . Therefore , we focused our efforts on the 1 . 8 Mb B6 vs . DBA “variable” interval , between rs33860076 at 168 . 1 Mb and rs30708865 at 169 . 9 Mb . The congenic/sub-congenic lines shown in Figure 1 displayed phenotypes of hypoinsulinemic hyperglycemia in association with relative reductions in β-cell mass due to reduced β-cell proliferation ( see Islet Morphology and β-cell Replication and Apoptosis ) . Phenotypes were generally more salient in male animals . Genotype in the congenic interval ( B6 or DBA ) per se did not affect their body weight or composition . Supporting experiments are described below . By 4 weeks of age , fasting plasma glucose was elevated in Lepob/ob males who were D/D ( DBA/DBA ) for the congenic interval 1jcd and fed standard ( 9% fat ) chow; glucose concentrations were higher up to 120 days . After 120 days , there were no significant differences in fasting glucose between D/D ( DBA/DBA ) and B/B ( B6/B6 ) mice ( Figure 2A ) . The decline in pre-prandial blood glucose levels in Lepob/ob males between 90 and 200 days is probably attributable to a slight expansion of β-cell mass in response to transient insulin resistance occurring as a normal consequence of sexual maturation ( ∼60 days of age ) [9] , [45] . To examine diabetes susceptibility in D/D animals that were obese independent of leptin deficiency , we fed lean ( Lep+/+ ) 1jcd males a high-fat diet ( 60% kcal from fat ) for 13 weeks , starting at 7 weeks of age . These mice became more hyperglycemic than B/B mice ( Figure 2B ) , showing a persistence of this difference – similar to the animals in 2A – up to age ∼140 days when the study ended . To delineate differences in acute glucose handling in D/D vs . B/B animals , we used intraperitoneal glucose tolerance testing ( ipGTT ) . At 60 days ( Figure 2C ) , and even up to 200 days , when the study ended ( Figure 2D ) , Lepob/ob 1jcdc males were less glucose tolerant than B/B . The relative reduction in glucose tolerance in D/D vs . B/B animals that are not overtly diabetic is likely related to reduction in the number of islets . The occurrence of the diabetes-related phenotype is independent of Lepob , since 100-day old Lep+/+ 1jc D/D males fed the Surwit ( high fat , high sucrose ) diet for 10 weeks were also less glucose tolerant than littermate B/B males ( Figure 2E ) . Hyperglycemia due to relative hypoinsulinemia , was evident in 1jc Lepob/ob D/D animals fed a chow diet as early as 4 weeks ( Figure 3A ) . At mean ages of 30- and 62-days , age-adjusted plasma insulin concentrations per mg blood glucose were lower in D/D than in B/B animals . This difference was due to lower plasma insulin in D/D ( p = 0 . 0004 ) and not higher blood glucose in D/D ( p = 0 . 916 ) . Consistent with these ratios , D/D Lep+/+ males showed a 40% decrease in insulin secretion when clamped at a blood glucose level of 250 mg/dl for an hour ( Figure 3B ) . No difference in insulin sensitivity was detected by euglycemic – hyperinsulinemic clamping ( data not shown ) . Consistent with their hypoinsulinemic hyperglycemia , 21-day old 1jcd D/D males had smaller islets than their B/B counterparts ( Figure 4A ) . A qualitative cell-autonomous β-cell defect in insulin secretion , however , is unlikely to be the primary functional defect in D/D animals , since islets isolated from 28-day old 1jcd D/D males responded to graded glucose concentrations ( 2 . 8 mM–16 . 8 mM ) or 10 mM arginine by secreting amounts of insulin comparable to age- and sex-matched B/B littermates ( Figure 4B ) . Also consistent with insulin/glucose ratios and hyperglycemic clamp results , isolated islets from 60-day old 1jc Lepob/ob males fed normal chow and 100-day old 1jc Lep +/+ on the Surwit diet showed reduced insulin secretion at 2 . 8 mM and 5 . 6 mM [glucose] in D/D vs . B/B littermates . For reasons indicated below , the early glucose intolerance of D/D mice is probably due , in part , to a deficiency of β-cell mass . The fractional area of the pancreas accounted for by β-cells [46] in Lepob/ob 1jcd males was examined in 20- , 60- and 150-day old mice . By 60 days a trend to reduced β-cell area was apparent in D/D , and by 150 days of age , β-cell mass of the 1jcd D/D sub-congenics was about half that of B/B littermate controls . B/D animals had β-cell masses that were about two-thirds of B/B littermate controls ( Figure 5A ) . These findings are consistent with in vivo data showing onset of elevated blood glucose ( see Figure 2A ) and lower circulating insulin concentrations ( relative to glucose ) in D/D sub-congenics at ∼60 days of age ( see Figure 3A ) , and persistence of decreased glucose tolerance at 200 days of age . The lower relative β-cell mass in D/D animals reflects fewer numbers of β-cells , rather than smaller sized β-cells . There were no differences in pancreatic weight between D/D and B/B male animals . To assess the basis for the difference in β-cell mass by 60 days , we measured rates of β-cell replication and apoptosis . We co-stained pancreatic sections in 1jcd congenic 1- and 21-day old Lepob/ob male mice with antibodies to insulin and Ki67 antigen , a nuclear marker of proliferation expressed during all stages of the cell cycle except G0 [47] . To estimate the proportion of dividing β-cells , we normalized the number of Ki67 positive β-cells to the total number of insulin positive cells . Groups consisted of 4 B/B and 4 D/D 1-day old mice or 4 B/B , and 8 D/D 21-day old mice . In 1-day old D/D males , the rate of β-cell replication was ∼1/3 that of B/B littermates , whereas there was no difference in 21-day old animals due to normally reduced β-cell replication by the time of weaning ( Figure 5B ) [48] , [49] , [50] . The proportion of small islets ( 250–2000 µm2 ) in 21-day old Lepob/ob males was greater in D/D ( 1jc and 1jcd ) mice ( 73% ) than in B/B ( 60% ) ; whereas the proportion of large islets ( 10 , 000–50 , 000 µm2 ) was lower ( 9% in D/D and 14% in B/B ) . This finding is consistent with the β-cell replication studies in P1 mice ( Figure 5B ) , and recently reported evidence that new β-cells are derived from replication of pre-existing β-cells [51] . In 13-day old 1jc mice , when β-cell apoptosis is active [52] , we did not detect significant differences between B/B and D/D islets in β-cell apoptosis using a TUNEL assay [53] and caspase-3 staining [54] ( data not shown ) . Thus , the lower number of β-cells in D/D mice is primarily a result of lower rates of proliferation of β-cells in the perinatal period . To identify all genes in the minimal DBA variable interval , ( see above for definition ) we screened 277 genes and transcripts , computationally predicted by GenScan , TwinScan , FGeneSH , Otto , or SGP2 that map to the interval . We excluded 50 single-exon transcripts ( probably pseudogenes [55] ) that did not belong to a transcript cluster and were not homologous to transcripts in the syntenic human interval , and 16 ribosomal gene transcripts , unique to this interval , that could not be specifically amplified due to their genomic redundancy , and manually curated the remaining 211 predicted transcripts . We rejected 63 that did not amplify in RNA/cDNA pools from multiple organs/ages of B6 and DBA mice ( see Methods: Testing for Predicted Transcripts in cDNA Pools ) and , using BLASTn , clustered the remaining 148 transcripts into 14 groups . These , correspond to 11 known genes and 3 predicted genes that we validated by amplification in cDNA pools . A map of the “variable” interval shows 14 genes , flanked by Mael and Pbx1 ( Figure 6 ) . We analyzed all transcripts in the entire “variable” region . The genetic variation accounting for differential diabetes-susceptibility in mice segregating B/B vs . D/D in the congenic intervals could be due to: 1 ) coding sequence variant ( s ) that alter the amino acid sequence of a protein ( s ) ; 2 ) regulatory variants , including anti-sense transcripts that affect expression and stability , and 3′ untranslated region ( UTR ) variants; or 3 ) splicing variants . We investigated all hypotheses . To identify all non-synonymous B6 vs . DBA sequence variants in the “variable” interval , we collected genomic sequence for B6 and DBA strains from databases at NCBI and Celera [56] , filled gaps using bi-directional sequencing to achieve 100% coverage of all coding sequences in both strains , and validated coding sequence variants by bi-directionally re-sequencing gene fragments encompassing each variant in both B6 and DBA strains . Consequently , we identified five non-synonymous single nucleotide variants: one in each of three FMO-like ( flavin mono-oxygenase ) genes , and two in chr1 . 1224 . 1 ( Figure 6 ) . The latter gene , we designated “Lisch-like” ( Ll ) because of its sequence similarity to a gene in mouse and rat , formerly known as Lisch7 ( http://rgd . mcw . edu/ ) , but now known as Lsr ( lipolysis stimulated receptor ) . Computational analysis of LL and the three FMO-like proteins using SNAP [57] , PolyPhen [58] , SIFT [59] , PAM250 matrix substitution weights [60] and PROFacc [61] predicted that all of the amino acid substitutions were benign with respect to function . The SNAP scores obtained for our variant alleles , -1 ( FMO13 , K282E ) , -2 ( FMO12 , V239I ) , -3 ( LL , A647V ) , and -6 ( LL , T587A; FMO9 , Q5R ) , indicate that there is a ∼60% , ∼69% , ∼79% , and ∼90% respective chance of the non-synonymous variants being neutral . Similarly , PolyPhen classified all variations as “benign” and SIFT scores were well above 0 . 05 ( neutral ) . PAM weights of 0 and above suggest interchangeability of the respective amino acids throughout evolution . The % differences were low , suggesting that the DBA and B6 variants are equally likely to occur in related sequences ( see Methods: Computational Methods for Evaluating Effects of nsSNPs ) . We used Affymetrix microarrays to quantify those transcripts in the minimum congenic interval that we had validated by PCR-amplification ( see Methods: Testing for Predicted Transcripts in cDNA Pools ) . We examined hypothalamus , islets , liver , soleus and EDL ( extensor digitorum longus ) skeletal muscle from DD and BB Lepob/ob congenic animals ( see Methods: Microarray Gene Expression Analysis ) . These arrays did not contain elements for all of the 14 genes we confirmed in the interval: missing from the array were the 3 FMO genes . Therefore , we also used real-time qPCR , to quantify expression of each gene and confirmed transcript in tissues and organs central to diabetes ( pancreatic islets , liver , skeletal muscle , adipose tissue and hypothalamus ) in 90-day old male Lepob/ob 1jc D/D and B/B animals ( see Methods: real time qPCR ) . Results of the microarray and qPCR experiments are shown in Table 1 and summarized in Figure 7A . Among genes in the region , including Lmx1a [62] , and Rxrg [63] , that constitute candidates for susceptibility to T2D , we identified no non-synonymous SNPs ( nsSNPs ) and no multi-organ differences in expression levels between B/B and D/D animals . The most prominent and consistent differences in expression we did observe were for chr1 . 1224 . 1 ( Ll ) , which was two to four-fold lower in 21-day old Lepob/ob D/D mice than in B/B mice in the diabetes-relevant tissues/organs by microarray analysis and up to twenty-fold lower by qPCR ( Figure 7A ) . ( We later show that Ll protein in hypothalamus is strikingly reduced in 1jc D/D vs . B/B; see Figure 11A ) . The difference in Ll gene expression in liver persists with age ( Figure 7B ) as does the difference in glucose tolerance in response to overt glucose challenge ( see Figure 2D ) . Whether the differences in hepatic Ll expression are mechanistically related to differences in glucose homeostasis are unknown at this point; LL may influence hepatic gluconeogenesis , or the hepatic differences could simply mirror parallel and more physiological relevant changes in β-cells . We also detected ( by PCR ) Ll transcripts in e7 , e11 , e15 , and e17 whole mouse embryos , and in testis , kidney , heart , lung , uterus , eye , thymus and spleen . For the anti-sense interval between intron 9 and intron 7 ( see below and Figures 1 and 8 ) , we found higher expression levels in liver and hypothalamus of D/D v . B/B animals . This difference is consistent with a possible suppressive role for the D/D anti-sense transcript ( see below ) . The Aldh9a1 gene , known to be highly expressed in human embryonic brain and involved in glycolysis and fatty acid metabolism , showed qualitative changes comparable to those seen in Ll . The mapping experiment that identified the interval of mouse Chr1 containing statistical signals related to T2D phenotypes would be expected to enrich for regions in which several genes might contribute to the phenotypes . Although Aldh9a1 may be such a gene , we chose to focus initially on Ll , since it showed the most striking quantitative differences in expression between D/D and B/B animals . From the Ensembl database , we identified zebra fish orthologs of Ll and Lsr . The clustalW pair-wise similarity scores for the predicted protein coded for by the zebra fish gene zgc:114089 ( Lsr ortholog ) is 42 vs , the mouse LSR protein , and 29 vs . the mouse LL protein . The similarity scores for the predicted protein coded for by the zebra fish gene zgc:110016 ( Lisch-like ortholog ) are 36 vs . LL and 28 vs . LSR . We performed clustalW analysis ( Figure 9 ) between the mouse LL-iso1 protein and three related proteins: 1 ) the human C1orf32 protein at 1q24 . 1 ( chr . 1 165 , 154 , 620–165 , 211 , 185; NCBI Build 36 . 1 ) , which is the product of a gene highly expressed in the developing human retina and brain [64]; 2 ) the predicted protein sequence for the zebra fish Lisch-like ortholog , zgc:110016 located on zebra fish chromosome 9 at 31 . 6 Mb; and 3 ) the mouse LSR protein , transcribed from a gene on chromosome 7 at 30 . 7 Mb . Pair-wise similarity scores for the intact proteins and major domains are shown in the legend . The human homolog is similar throughout , but diverges slightly in the putative ICD . The zebra fish Lisch-like ortholog and mouse LSR proteins are most alike in the TMD , less so in the Ig-like domain , and most dissimilar in the ICD . The Lsr protein has a short extension to exon 6 , and no exon 8 equivalent . Ll and Lsr also have splicing patterns similar to the mouse Ildr1 ( Ig-like domain receptor 1 ) gene [65] , and the proteins they encode all belong to the Lisch7 family ( IPR008664; www . ebi . ac . uk/interpro ) . To assess the function of Ll in islet/β-cell ontogenesis , we examined expression patterns and the effects of morpholino-mediated knockdown in zebra fish embryos . Morpholinos are modified anti-sense oligonucleotides that produce a strong hypomorphic “knockdown” phenotype [66] either by inhibiting proper splicing of the pre-RNA transcript [66] or by ATG-blocking of translation [67] . Morpholino knockdown has been used to demonstrate a role for the endocrine hormones GnRH , GHRH and PACAP during development [68] , [69] , [70] , [71] . Many of the molecular mechanisms regulating pancreas development appear to be conserved among zebra fish and other vertebrates [72] , and the single zebra fish islet provides an excellent model of vertebrate development . Using whole mount in situ hybridization ( Figure 10A ) , we observed that the Lisch-like ortholog zgc:110016 was expressed in the brain and otocyst by 48 hours post fertilization ( hpf ) , and by 72 hpf expression was evident in the intestine . The Lsr ortholog zgc:114089 , located on Chr 15 at 39 . 0 Mb , was expressed in pancreas at 48 and 72 hpf , ( similar to our postnatal observations in mouse with Ll ) , intestine , liver , pharynx , pronehphros and otocyst for 48 hpf ( 72 hpf not shown ) , and , at 34 hpf , in both pancreatic buds . Since the anterior bud gives rise to exocrine tissue , pancreatic duct , and a small number of endocrine cells , while the posterior bud gives rise only to endocrine tissue [69] , expression of the Lsr-like paralog throughout this stage is consistent with a role in the ontogeny of pancreatic endocrine tissue . The close structural similarities among Lisch-related genes ( see Figure 9 ) suggested that functional data on both zebra fish genes could be physiologically relevant and , therefore , we studied the involvement in islet development of both orthologs . We injected ( in separate experiments ) morpholinos for both genes into embryos homozygous for the gut-GFP ( green fluorescent protein ) transgene to visualize developing endodermal organs ( Figure 10B ) [73] . We assessed β-cell development with an anti-insulin antibody at 48 hpf or by insulin in situ hybridization at 24 hpf ( not shown ) . To assess morpholino specificity , we analyzed the effects of two separate , non-overlapping morpholinos for each gene . Both morpholinos for each ortholog independently produced similar phenotypes , providing evidence that the effects ( described below ) were the result of specific gene knockdown and not due to nonspecific morpholino-related effects . Figure 10B shows that both Lsr-like and Ll morpholinos injected at 15 ng/embryo produced general developmental delay in the endodermal organs , evidenced by a smaller liver , a smaller , straighter intestine , and a smaller pancreas that does not extend as much as in wild-type . The Lsr-like morpholinos disrupt β-cells more severely ( note ectopic insulin-positive cells in the cephalad region of the pancreas ) than do the Ll morpholinos ( note the milder local dispersion of insulin-positive cells ) ; 48/72 and 25/144 embryos injected with morpholinos targeting Lsr-like and Ll , respectively , displayed a scattered β-cell phenotype . These effects were rarely observed in uninjected sibling embryos ( 0/25 ) or embryos injected with a control morpholino ( 1/35 ) . Lower doses of Lsr-like and Ll morpholinos ( ∼7–10 ng ) resulted in a lower frequency of β-cell scattering and higher doses ( ∼20–25 ng ) resulted in embryonic toxicity , which is common with high doses of morpholinos . The efficacy of the splice-blocking Lsr-like and Ll morpholinos was assessed via RT-PCR and all were found to strongly and specifically inhibit proper splicing of their respective target transcripts at the 15 ng dose ( not shown ) . In combination , the expression analyses and morpholino knockdown studies provide support for a role of Lisch gene family members in endodermal development , and suggest specific effects on the embryonic β-cell . The relevance of such zebra fish studies to mammalian pancreas development has been shown earlier for Ptf1a [74] , [75] and for Pdx1 [76] . To examine phenotypes of mice segregating for a null allele for Ll , we screened a repository of ENU-generated ( N-ethyl-N-nitrosourea ) mutant sperm DNAs from 18 , 000 C3HeB/FeJ G1 males ( Ingenium; http://www . ingenium-pharmaceuticals . com/ ) for mutations in Lisch-like [77] . We detected a G/A substitution that encodes an amber stop mutation at threonine-87 [W87*] and also creates an EcoN1 cleavage site , which we used to genotype for the mutation . By in vitro fertilization , we generated W87* heterozygotes on the C3HeB/FeJ background , and bred these animals to generate progeny that were homozygous wild-type ( +/+ ) , homozygous mutant ( −/− ) or heterozygous ( +/− ) for the W87* mutation . Progeny were born at the anticipated Mendelian ratios , and the −/− animals did not appear grossly compromised . To verify that the W87* homozygous mutant was hypomorphic for LL protein , we compared a Western blot of hypothalamic extracts prepared from C3HeBFeJ wild-type ( +/+ ) and mutant ( −/− ) mice , with a second blot of hypothalamic extracts prepared from B/B and 1jc-D/D congenic mice . We probed both sets of filters with a polyclonal rabbit antibody generated to a conjugated polypeptide , corresponding to exons 7 and 8 of isoform 1 , in the predicted ICD of LL . As anticipated , LL protein was greatly reduced in the brains of D/D vs B/B congenics and in the ENU-treated W87* homozygotes vs . the wild-type animals ( Figure 11A ) . In mice at 14 days of age we can detect reductions in β-cell replication rates that are similar to those seen in the DD congenic lines ( Figure 5B ) There is a >2-fold difference in the proportion of Ki67-positive β-cells in 14-day old wild-type ( 3 . 75% ) vs . homozygous W87* mice ( 1 . 75% ) , with heterozygotes intermediate ( 2 . 5% ) ( Figure 11B ) . Plasma insulin concentrations in Ll W87* homozygotes are reduced by the time of sexual maturation ( Figure 11C ) and , consistent with this difference , at 50 days of age , homozygous W87* males show an increased glucose AUC during iPGTT ( Figure 11D ) . A significant decrease in β-cell mass is also detected in W87* homozygotes ( 1 . 05%± . 117 , n = 3 , p = . 0113 ) v . +/+ littermates ( 2 . 74± . 364; n = 3 ) at 150 days of age . It is important to note that these phenotypes were detected despite the segregation of the mutation on a different background strain ( C3HeB/FeJ ) than our congenics ( C57BL/6J ) , and in the absence of co-segregation of the Lepob . These preliminary data strongly support the candidacy of Ll as the gene accounting for the diabetes-related phenotypes of the DD congenic lines .
Insight into the function ( s ) of the mouse Lisch-like protein may be gained from similarities in structure , expression , and cellular location with the human paralog , C1orf32 , and with genes encoding related trans-membrane receptors , Ildr1 [65] and Lsr [91] . Splicing patterns of these genes generate isoforms , similar to those of Ll . Each gene's largest isoform includes an extra-cellular Ig-like domain , a single TMD , and a similar set of ICDs in related order . In one isoform of each protein , the TMD and cysteine-rich domains are absent . An evolutionary , regulatory relationship is suggested by the observation that the Ll-paralog and lldr1 are adjacent in the zebra fish genome ( Zv6 assembly , UCSC Genome Browser ) . All three genes are abundantly expressed in the brain , liver and pancreas ( and islets , where studied ) , and all are predicted to have 14-3-3 interacting domains ( thus far experimentally verified for the human LSR ) [94] . Although 14-3-3 interacting domains may be present on as many as 0 . 6% of human proteins , their occurrence on all of these Lisch-related proteins is notable , since among known 14-3-3-interacting proteins is phoshodiesterase-3B , which is implicated in diabetes and pancreatic β-cell physiology [95] , [96] , [97] , and others , such as the Cdc25 family members , important in regulating cell proliferation and survival [98] , [99] . The human ortholog of Ll , C1orf32 , which is 90% identical to Ll at the amino acid level , maps to a region of Chr1q23 that has been implicated in T2D in seven ethnically diverse populations including Caucasians ( Northern Europeans in Utah ) [100] , Amish Family Study [101] , [102] , United Kingdom Warren 2 study [103] , French families [104] , and Framingham Offspring study [105] , Pima Indians [106] , and Chinese [96] with LOD scores as high as 4 . 3 . The mouse congenic interval examined here is in the middle of , and physically ∼10× smaller than , the 30 Mb human interval . Recent analysis of the broad interval ascertained in Utah identified two peaks , one of which , at D1S2762 ( at 163 . 6 Mb ) , is just 12 kb telomeric to the 5′ end of C1orf32 [107] . The genes , and gene order , are generally conserved between mouse and human in the region syntenic to the congenic interval . The metabolic phenotypes documented in human subjects with T2D linked to 1q23 resemble diabetic phenotypes observed in congenic mice segregating for the DBA interval in B6 . DBA congenics examined here [108] , suggesting that the diabetes-susceptibility gene in congenic mice and human subjects may be the same gene , or among the genes , acting in the same genetic pathway ( s ) . The syntenic interval in the Goto-Kakizaki ( GK ) rat also correlates with diabetes-susceptibility [109] . We report the molecular cloning and preliminary characterization of a candidate gene for a mouse QTL modifying T2D phenotypes in mice . The gene , Lisch-like , is novel in structure among diabetes susceptibility genes , and appears to modify β-cell development . Amino acid sequence analysis is consistent with the possibility that hypomorphism for this gene could affect β-cell development by a number of possible molecular mechanisms . Proof of the role of this gene in the imputed phenotypes and molecular processes awaits its further analysis in transgenic animals and cell-based systems .
Mice were housed in a barrier facility in ventilated Plexiglas cages under pathogen-free conditions at room temperature ( 22±1°C ) with a 12 h light/dark cycle . Mice were weaned at 21 d and given ad libitum access to water and 9% Kcal fat Picolab Rodent Chow 20 ( Purina Mills; www . purinamills . com/ ) . The high fat diet protocol used in some animals is described below . Columbia University's Institutional Animal Care and Use Committee ( IACUC ) approved all protocols . After a 4 h morning fast , mice were sacrificed by carbon dioxide asphyxiation and phenotyped for weight , naso-anal length , and glycosuria . Blood was collected by cardiac puncture and aliquoted into microfuge tubes containing an anticoagulant cocktail of 10 µl of 1 mM EDTA and 1 . 5 mg/ml aprotinin ( Sigma A-6279 ) . Plasma and red blood cell pellets were used to measure glucose , insulin , and glycosylated hemoglobin as previously described [110] . Tissues ( skeletal muscle , pancreas/pancreatic islets , liver , brain , hypothalamus , kidney , spleen , heart , visceral fat , retroperitoneal fat ) were collected and immediately frozen in liquid N2 , and stored at −80°C for further studies . Pancreata were dissected under stereoscope , weighed , and fixed in Z-fix zinc-formalin fixative ( Anatech; www . anatechltdusa . com/ ) . Liver tissue or tail tips were used for genomic DNA isolation according to standard procedures [111] . A mutation-specific assay was used to confirm that all phenotypically obese animals were Lepob/Lepob and all lean animals either +/+ or heterozygous at the Lep locus [112] Animals were genotyped using MapPairs Microstaellite Markers ( Invitrogen; www . invitrogen . com/ ) as previously described [113] . Maps were created using MapMarkerQTL ( www . broad . mit . edu/genome_software/other/qtl . html ) on a dataset representing 404 obese F2 progeny of a B6xDBA cross segregating for Lepob at 120–150 days of age . The QTL for T2D was most significantly associated with fasting blood glucose , glycosylated hemoglobin , and islet histology in male mice to a region of Chr1 , with peak statistical significance at D1Mit110 at 169 . 6 Mb from the centromere ( p<10−8 ) ( Figure 12 ) . Other QTLs were identified on other chromosomes ( for example Chr5 at 78cM ) , but none had as great an effect on the phenotype or demonstrated consistent effects on all aspects of the phenotype . We tested for interactions for QTLs and identified a modest interaction between the locus on Chr1 and a second locus at D4Mit286 ( p = 0 . 008 ) . B6 . DBA congenic mice were generated by intercrossing Lepob/Lep+ B6 X DBA mice from Jackson Laboratory ( www . jax . org/ ) to generate F1 progeny , followed by backcrossing to the recurrent B6 strain using a “speed congenic” approach in subsequent generations [114] . At the eighth backcross , a genome scan was performed in all breeders using polymorphic markers at 20 cM intervals . In the mouse line that was continued , all non-contiguous markers outside the DBA interval were homozygous B6 . Over the next two generations , there were two recombination events , one that eliminated a telomeric portion of the DBA interval ( line 1jc ) and one that preserved approximately half of the originally defined DBA interval ( line 1jcd ) . The 1jcd mouse was bred repeatedly to B6 mice , giving rise , by meiotic recombination , to two additional subcongenic lines ( 1jcdt and 1jcdc ) ( see Figure 1 ) . Preservation of the phenotypes present in the original B6xDBA and DBAxB6 F2/F3 progeny was assessed by longitudinal and end-point measurements of fasting glucose , insulin , glycosylated hemoglobin and islet morphology . At N12 , Lepob/+ mice B6/DBA ( B/D ) for the respective congenic intervals were intercrossed to produce N12F1 progeny . Obese progeny were used for fine mapping and phenotyping experiments . Lepob/+ animals D/D for the congenic interval were recurrently intercrossed or crossed to B6 Lepob/+ animals to generate ob/ob Lepob/Lepob animals with D/D and B/D genotypes for the Chr1 interval , respectively . For longitudinal phenotyping studies , mice were fasted for 4 h and restrained for blood collection by a trained individual . Blood was collected from unanesthetized animals by capillary tail bleed into heparinized tubes and stored at −80°C . Glucose was measured with a FreeStyle Flash Blood Glucose Monitor ( Abbott; www . abbottdiabetescare . com/ ) . Insulin was measured by ultra-sensitive rat insulin ELISA ( ALPCO; www . alpco . com/ ) . HbA1c was measured by affinity chromatography ( Mega Diagnostics; www . mega-dx . com/ ) . Urine ketones were measured using Chemistrip Test Strips ( Roche Diagnostics; http://us . labsystems . roche . com/index . shtml ) . For ipGTT , mice were fasted overnight and 0 . 5 g/kg body weight of 50% dextrose was administered intra-peritoneally at time 0 . Plasma glucose was measured at 15–30 min intervals for 3 h , as above . Terminal phenotypic characterization consisted of measurements of fasting glucose , insulin , glycosuria , and glycosylated hemoglobin as previously described [110] . To control for stress-induced hyperglycemia at the time of sacrifice , tail blood glucose was also measured by glucometer one day prior to sacrifice . High fat chow pellets ( #D12492i: 60% kcal from fat , 20% kcal from protein , 20% kcal from carbohydrate ) and “Surwit” [115] ( #D12331i; 58% kcal from fat , 16 . 4% kcal from protein , 25 . 5% kcal from carbohydrate ) ( Research Diets; www . researchdiets . com/ ) were used as described in the text . Pancreatic tissues were dissected under stereoscope to avoid contamination with adipose tissue , and weighed . Non-overlapping images of longitudinal pancreatic sections were acquired and analyzed using ImageProPlus software version 5 . 0 ( Media Cybernetics; www . mediacy . com/ ) to calculate insulin-positive area , insulin-positive area as % total area , and number of islets ( defined by an area containing a minimum of 8 contiguous insulin-positive cells ) . For β-cell replication studies , we recorded the number of Ki67-positive or negative , insulin-positive cells . Replication of β-cells was expressed as % of cells ( Ki67-positive and insulin-positive ) / total insulin-positive . For replication studies , ∼100 islets were examined per animal from several different non-overlapping sections through the pancreas . ImageProPlus or Image J ( 1 . 37 V; NIH ) were used to determine the relative area of each section occupied by β-cells or the actual of number of β-cells for each representative longitudinal pancreatic section ( 50 µm apart ) that had been immunochemically stained for insulin as previously described [116] . We analyzed 5–7 sections from different regions of the pancreas . Apoptosis rates were assessed using the DeadEnd Fluormetric TUNEL System G3250 ( Promega; www . promega . com/ ) TUNEL assay and cleaved Caspase-3 ( Asp175 ) Antibody 9661S ( Cell Signaling Technology; www . cellsignal . com/ ) . Pancreatic perfusion and islet collection were performed as previously described [117] . Each pancreas was perfused via the bile duct with 1 . 5 mg/ml collagenase P ( Roche Applied Science; www . roche-applied-science . com/ ) and incubated at 37°C for 17 min . Following disaggregation of pancreatic tissue , pancreata were rinsed with M199 medium containing 10% NCS . Islets were collected by density- gradient centrifugation in Histopaque ( Sigma-Aldrich; www . sigmaaldrich . com/ ) [117] , and washed several times with M199 medium . For glucose-stimulated insulin release studies [118] , [119] , islets were incubated overnight in RPMI medium 1640 ( Invitrogen ) . The GSIS procedure has been described previously [120] . Islets were hand-picked into tissue culture dishes containing cold Kreb's buffer ( 118 . 5 mM NaCl , 2 . 54 mM CaCl2 , 1 . 19 mM KH2PO4 , 1 . 19 mM MgSO4 , 10 mM HEPES , pH 7 . 4 ) , and 2% BSA ( Sigma-Aldrich ) , 5 . 5 mM glucose , and incubated overnight at 37°C . Islets were hand-picked and incubated another 15 min . in Kreb's buffer+BSA , containing 11 . 2 mM glucose . Hand-picked islets are then resuspended in Kreb's buffer plus BSA , supplemented with 2 . 8 mM glucose , and shaken at 37°C for 15 min . The pellet was spun down gently and resuspended in triplicate ( 5–10 islets each ) in 500 µl Kreb's buffer , supplemented with glucose at final concentrations of 2 . 8 mM , 5 . 6 mM , 11 . 2 mM or 16 . 8 mM , or supplemented with 10 mM arginine and incubated for 1 h in a water bath at 37°C with constant shaking ( 300 rpm ) . After 1 h incubation , islets were gently pelleted and the supernatant collected and assayed for insulin by ELISA . Islet pellets were dissolved in high salt buffer ( 2 . 15 M NaCl , 0 . 01 M NaH2PO4 , 0 . 04 M Na2HPO4 , EDTA 0 . 672 g/L , pH 7 . 4 ) and sonicated at 4–5 W for 30 s and DNA concentration was measured using a TKO100 fluorometer ( Hoefer; www . hoeferinc . com/ ) with Hoechst #33258 dye ( Polysciences; www . polysciences . com ) . Results were expressed as concentration of secreted insulin/[DNA]/h . Putative transcripts , identified from public annotation and local sequencing , were validated by PCR-amplification from tissue-specific cDNA pools prepared from male and female B6 mice . Two cDNA pools were used: 1 . An inclusive cDNA pool was prepared from E7 and E20 fetuses and P1 pups , and included the following tissues of 60-day old mice: eyes , large intestine , skin , tongue , spinal cord , kidney , testes/ovaries , pancreatic islets , whole brain , hypothalamus , skeletal muscle , and liver . This pool was used for transcript validation . 2 . A diabetes-relevant cDNA pool , from 90-day old mice , was comprised of only the following tissues and organs: pancreatic islets , whole brain , hypothalamus , skeletal muscle , liver , and adipose tissue . This pool was used to quantify transcripts identified by computational approaches and the microarrays . Nominal intron-spanning primers were generated using the Primer3 program ( www . genome . wi . mit . edu/cgi-bin/primer/primer3_www . cgi ) . Amplification was first performed on the diabetes-relevant pool at an annealing temperature of 60°C . If we detected no PCR-product , we performed gradient temperature PCR on the same pool using eight different annealing temperatures from 58–68°C . Gradient temperature PCR was then used to amplify the inclusive cDNA pool . If no product was detected in this pool , a 2nd set of intron-spanning primers was used before we interpreted negative amplification as failure to substantiate a predicted transcript . Positive amplification products of predicted sizes , and those that did not match the expected sizes , were gel-purified and sequenced for confirmation . The final set of primer-pairs is listed in Real-time qPCR . RNA extraction , purification , labeling , hybridization and analysis were performed as described [121] . 10 BB and 10 DD 21-day old Lepob/ob 1jc males were dissected and RNA was extracted from hypothalamus , liver , isolated islets , EDL muscle , and soleus muscle . Individually labeled RNA ( by mouse and organ ) was interrogated with Affymetrix MOE-430A expression arrays . For further details , see legends to Table 1 and Figure 7 . For all transcripts in the region of interest , where possible , only probes that spanned multiple exons and clearly represented each of the 14 genes in the interval were used . If >1 probe met these conditions , we used only , the probe that gave the strongest signal . Organs were grouped into two groups by genotype and were compared using a two tailed T-test . The Affymetrix probe IDs selected for this analysis are shown in Table S3 . Effects of the DBA congenic interval on the levels of confirmed transcripts expressed in diabetes-relevant organs were assessed on an organ-specific basis . We made separate pools from 90-day old Lepob/ob 1jc D/D and B/B mice for each of the diabetes-relevant organs ( see above ) . Each individual organ pool was generated on 2 occasions from 5 mice . RNA was extracted from organs with TRIzol acid-phenol reagent ( Invitrogen ) . 2 µg of RNA were reverse-transcribed using SuperScript III reverse transcriptase ( cDNA First Synthesis Kit , Invitrogen ) with random hexamer priming . The cDNA was diluted 4-fold using nuclease-free water ( QIAGEN; www . qiagen . com ) . 2 µl of diluted cDNA were amplified by PCR in Roche LightCycler . A standard curve for each transcript was generated using cDNA diluted 1∶1 , 1∶10 , and 1∶100 . We assessed the number of mRNA molecules in each sample using the slope and intercepts of PCR product appearance during the exponential phase of the PCR reactions optimized for transcript-specific product using specific primers . Each sample was run in triplicate in the same LightCycler run . Using LightCycler Software , we calculated the crossing point ( CP ) for each sample . The CP is the first maximum of the second derivative of the fluorescence curve , and is equivalent to the number of cycles at which the fluorescence first exceeds background . In the exponential phase , the relationship between CP and initial transcript concentration is linear . We calculated relative concentration ratios , normalized to actin , as follows: In this expression , ΔCPgene is the CP of the gene in the sample minus the CP of the gene in the relevant reference; ΔCPhg is the CP of the housekeeping gene in the sample minus the CP of the housekeeping gene in the reference ( “ref” ) sample; and η is the efficiency ( where 2 is perfectly efficient ) as determined by the negative slope of the plot generated when CP is plotted as a function of the log of initial concentration determined in the standard curve . Each CP listed is the mean of CP values of the triplicates for each sample . Results are summarized in Table 1 . Primers used are listed in Table S4 ( A ) . We amplified full-length Ll cDNAs from either B6 islets ( isolated by us ) or from Clontech MTC Panels 1 #636745 and 3 #636757 , containing pooled multiple tissue cDNAs from 8–12 week old BALB/c mice and from Swiss Webster embryos . In a final volume of 50 µl , we added 0 . 5 µl LA Taq ( TaKaRa; www . takara-bio . com/ ) to a cocktail containing TaKaRa GC Buffer II , 400 µm each dNTP , 1 µl cDNA and 1 µl each primer ( 300 ng/µl ) . Primers are listed in Table S4 ( B ) . Samples were cycled in an MJ Tetrad Thermalcycler ( BioRad; www . bio-rad . com ) using a Touchdown protocol of a 2 min . extension and decreasing annealing temperature from 60°C to 55°C for 10 cycles , followed by 25 cycles with an annealing temperature of 55°C . Each sample was TOPO TA cloned ( Invitrogen ) and plated . From all three libraries , a total of 140 colonies were picked and grown overnight in LB buffer . Inserts were amplified by colony PCR and sized by gel-fractionation . Inserts representing each unique size were then sequenced . The isoforms and the exons deleted ( Δ ) : iso1 ( intact 10 exons ) ; iso2 , Δ6; iso3 , Δ4 , 5 , 6; iso4 , Δ4; iso5 , Δ5 , 6; iso6 , Δ9; iso7 , Δ5 , 6 , 7 , 8 , 9 . We used five methods to compute the likelihood of a functional change due to single amino acid substitutions ( see Figure 9 ) . SNAP , PolyPhen , and SIFT predict changes in protein function due to a single amino acid substitution . SNAP [57] is a neural-network based method that considers protein features predicted from sequence ( e . g . , residue solvent accessibility and chain flexibility ) . Scores from −9 to +9 are estimates of accuracy of prediction , computed using a testing set of ∼80 , 000 mutants . A low negative score indicates confidence in prediction of neutrality ( functional change absent ) , whereas a high positive score indicates confidence in prediction of non-neutrality ( functional change present ) . Accuracy was computed for neutrals using the equation below: PolyPhen considers structural and functional information and alignments . Predictions are sorted into 4 classes: benign , possibly damaging , probably damaging , and unknown . SIFT predictions . SIFT [59] is a statistical method that only considers alignments . Scores range from 0 to 1 . Scores >0 . 05 indicate neutrality of a substitution . PAM250 matrix substitutions . PAM matrix [124] ( Percent Accepted Mutations ) reflects frequency of amino acid interchange throughout evolution ( by evaluating alignments of proteins in a family ) . Scores range from a low of −8 for rare substitutions ( e . g . W to C ) to a high of 17 ( same residue found in almost all proteins in alignment ) . Percentage in alignment ( PROFacc ) . The score is reported as the difference in observed percentages of wild-type and mutated residues in alignments against a non-redundant UniProt [125] and PDB [126] database ( at 80% sequence identity ) . Scores range from −100 ( if the mutant is observed in all instances ) to +100 ( if the wild type is observed in all instances ) ; 0 if the mutant is observed as often as the wild type . Scores near 0 favor the likelihood of a mutation being neutral . BAC 95f9 DNA ( 5 µg ) was fragmented to 1–5 kb using a nebulizer supplied with the TOPO Shotgun Subcloning kit ( Invitrogen ) and checked for size and quantity on an agarose gel . The shotgun library was constructed with 2 µg of sheared DNA . Blunt-end repair , dephosphorylation , ligation into PCR 4Blunt-TOPO vector , and transformation into TOP10 Electrocompetent E . coli were performed with the TOPO Shotgun Subcloning kit , following the manufacturer's protocol . Phenol∶chloroform extraction of the dephosphorylated DNA was replaced with Qiagen QIAquick PCR Purification spin columns ( QIAGEN ) . Recombinant colonies were selected by blue/white screening and incubated in LB medium supplemented with 50 µg/ml ampicillin for 20 h at 37°C in 96-well deepwell plates . Plasmid miniprep was conducted in 96-well plates using QIAGEN Turbo Miniprep kits on a QIAGEN BioRobot 9600 . DNA sequencing was performed on a 3730xl Genetic Analyzer ( Applied Biosystems; www . appliedbiosystems . com/ ) using BigDye® Terminator v3 . 1 Cycle Sequencing Kits with M13 forward and reverse sequencing primers . ANOVA and ANCOVA were used to assess effects of genotype in congenic interval . Comparisons at individual time points , or pairs of means were performed using Student's t-test . P values are 2-tailed . The Statistica package ( StatSoft; www . statsoft . com/ ) was used for ANOVAE; Excel ( Microsoft , http://office . microsoft . com/en-us/default . aspx ) for t-testing . Hypothalamic extracts were prepared using M-PER Mammalian Protein Extraction Reagent ( Pierce Biotechnology , www . piercenet . com/ ) . Hypothalamic extracts ( 85 mg for B/B and D/D congenics and 175 mg for wild-type and mutant ENU mice ) were resolved by 8% SDS-PAGE , transferred to nitrocellulose membrane ( Invitrogen ) . We generated a set of polyclonal rabbit antibodies ( Covance Research Products; www . covance . com ) against the predicted ICD , spanning residues 298–401 ( exons 7 , 8 ) and verified that the α-ICD rabbit antibodies detected the appropriate fusion proteins , with only minor cross-reactivity in cultured cells . We hybridized the blot with anti-LL anti-sera at a dilution of 1∶5 , 000 in TBS/0 . 05%Tween/5% milk ( TBSTM ) or with blocked anti-LL anti-sera diluted 1∶10 , 000 in TBSTM . To prepare blocked anti-sera , liver sections from C3HeB/FeJ knock-out mice were fixed overnight in phosphate-buffered paraformaldehyde at 4°C and rinsed in PBS . Sections equivalent to one-third of a liver were fragmented and mixed with 1 ml anti-sera diluted 1/1000 in PBS/0 . 1% Triton . Liver fragments were spun out and the supernatant was used to probe filters from ENU mice . We detected bound antibody with horseradish peroxidase-coupled antibody against rabbit IgG ( Amersham Biosciences; www . amershambiosciences . com ) at a dilution of 1∶5 , 000 using the SuperSignal West Pico Chemiluminescent Substrate kit ( Pierce Biotechnology ) . Genbank ( www . ncbi . nlm . nih . gov/ ( Genbank ) accession numbers for the M . musculus genes: Lisch-like , lipolysis-stimulated remnant receptor-related ( XM_001473525 ) ; Lsr ( NM_017405 ) ; Ildr1 ( NM_134109 ) ; Tada1l SPT3-associated factor 42 ( NM_030245 ) ; Pogk pogo transposable element with KRAB domain ( NM_175170 ) ; FMO13 , flavin-containing monooxygenase family; FMO-like ( XM_136366 ) ; FMO9 , flavin-containing monooxygenase family; FMO-like ( NM_172844 ) FMO12 , flavin-containing monooxygenase family; FMO-like ( XM_136368 ) ; C030014K22Rik , unknown ( NM_175461 ) ; Uck2 , uridine monophosphate kinase ( NM_030724 ) ; Tmco1 , membrane protein of unknown function ( NM_001039483 ) ; Aldh9a1 , aldehyde dehydrogenase 9 , subfamily A1 ( NM_019993 ) ; Mgst3 , microsomal glutathione-S-transferase 3 ( NM_025569 ) ; Lrrc52 , leucine-rich repeat ( LRR ) protein of unknown function ( NM_00103382 ) ; Rxrg , retinoid X receptor , gamma ( NM_009107 ) ; Lmx1a , LIM homeobox transcription factor 1 , α ( NM_033652 ) ; Pbx1 ( NM_008783 ) ; H . sapiens C1orf32 ( NM_199351 ) ; LSR ( NM_015925 ) ; ILDR1 ( NM_175924 ) ; D . rerio Ll ortholog , zgc:110016 ( NM_001030192 . 1 ) ; D . rerio Lsr ortholog , zgc:114089 ( NM_001025472 . 1 ) ; R . rattus Lsr ( NM_032616 ) . The Genbank accession numbers for protein sequences: M . musculus Lisch-like ( amino acid residues 150–795 , XP_001473575 ) ; ( Lsr ) ( NP_059101 ) ; Ildr1 ( NP_598870 ) ; H . sapiens C1orf32 ( NP_955383 ) ; LSR ( NP_057009 ) ; ILDR1 ( NP_787120 ) ; D . rerio Lisch-like ( NP_001025363 ) ; D . rerio Lsr ( NP_001020643 ) ; R . rattus Lsr ( NP_116005 ) | Type 2 diabetes ( T2D ) accounts for over 90% of instances of diabetes and is a leading cause of medical morbidity and mortality . Twin studies indicate a strong polygenic contribution to susceptibility within the context of obesity . Although approximately ten genes making important contributions to individual risk have been identified , it is clear that others remain to be identified . In this study , we intercrossed obese , diabetes-resistant and diabetes-prone mouse strains to implicate a genetic interval on mouse Chr1 associated with reduced β-cell numbers and elevated blood glucose . We narrowed the region using molecular genetics and computational approaches to identify a novel gene we designated “Lisch-like” ( Ll ) . The orthologous human genetic interval has been repeatedly implicated in T2D . Mice with an induced mutation that reduces Ll expression are impaired in both β-cell development and glucose metabolism , and reduced expression of the homologous gene in zebrafish disrupts islet development . Ll is expressed in organs implicated in the pathophysiology of T2D ( hypothalamus , islets , liver , and skeletal muscle ) and is predicted to encode a transmembrane protein that could mediate cholesterol transport and/or convey signals related to cell division . Either mechanism could mediate effects on β-cell mass that would predispose to T2D . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"physiology/endocrinology",
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/gene",
"discovery",
"diabetes",
"and",
"endocrinology/type",
"2",
"diabetes"
] | 2008 | Positional Cloning of “Lisch-like”, a Candidate Modifier of Susceptibility to Type 2 Diabetes in Mice |
Aerobic organisms have a tricarboxylic acid ( TCA ) cycle that is functionally distinct from those found in anaerobic organisms . Previous reports indicate that the aerobic pathogen Mycobacterium tuberculosis lacks detectable α-ketoglutarate ( KG ) dehydrogenase activity and drives a variant TCA cycle in which succinyl-CoA is replaced by succinic semialdehyde . Here , we show that M . tuberculosis expresses a CoA-dependent KG dehydrogenase activity , albeit one that is typically found in anaerobic bacteria . Unlike most enzymes of this family , the M . tuberculosis KG: ferredoxin oxidoreductase ( KOR ) is extremely stable under aerobic conditions . This activity is absent in a mutant strain deleted for genes encoding a previously uncharacterized oxidoreductase , and this strain is impaired for aerobic growth in the absence of sufficient amounts of CO2 . Interestingly , inhibition of the glyoxylate shunt or exclusion of exogenous fatty acids alleviates this growth defect , indicating the presence of an alternate pathway that operates in the absence of β-oxidation . Simultaneous disruption of KOR and the first enzyme of the succinic semialdehyde pathway ( KG decarboxylase; KGD ) results in strict dependence upon the glyoxylate shunt for growth , demonstrating that KG decarboxylase is also functional in M . tuberculosis intermediary metabolism . These observations demonstrate that unlike most organisms M . tuberculosis utilizes two distinct TCA pathways from KG , one that functions concurrently with β-oxidation ( KOR-dependent ) , and one that functions in the absence of β-oxidation ( KGD-dependent ) . As these pathways are regulated by metabolic cues , we predict that their differential utilization provides an advantage for growth in different environments within the host .
Despite the identification of Mycobacterium tuberculosis as the causative agent of tuberculosis ( TB ) over 125 years ago , two billion people worldwide are infected with this potentially lethal pathogen [1] . Each year , nearly ten million individuals will develop active TB; of these , approximately two million will die . Efforts to control the TB pandemic are now being threatened by the increasing prevalence of M . tuberculosis strains that are resistant to many or all available antimycobacterial drugs [2] . Understanding the biology of M . tuberculosis will facilitate the identification of targets for novel therapeutic approaches to preempt this persistent pathogen . Determination of the full genome sequence of M . tuberculosis has enabled the prediction and assembly of conserved metabolic networks [3]–[5] . While such models are valuable for understanding the metabolic architecture of an organism , discrepancies between genome-based predictions and data from genetic and biochemical analyses occasionally arise . For example , of the ten M . tuberculosis genes predicted to encode subunits for α-ketoglutarate ( KG ) and pyruvate dehydrogenases , only two have been shown to possess the corresponding function [6] , [7] . Indeed , biochemical surveys of enzymes of the tricarboxylic acid ( TCA ) cycle indicate that M . tuberculosis does not utilize a conventional KG dehydrogenase [8] . This disjunction at the conversion of KG to succinyl-CoA suggests either that this activity is non-essential for cellular metabolism , or that conversion of KG proceeds by means of a novel pathway . In support of the latter , it was recently shown that M . tuberculosis encodes enzymes capable of catalyzing a variant TCA cycle which uses succinic semialdehyde ( SSA ) rather than succinyl-CoA [8] . In this novel cycle , KG decarboxylase ( KGD ) catalyzes the thiamine pyrophosphate ( TPP ) dependent decarboxylation of KG to form SSA [8] . Subsequently , SSA dehydrogenase oxidizes SSA to succinate with the reduction of NADP+ to NADPH [8] . Similar to the canonical cycle , this cycle enables the extraction of reducing power to drive reductive processes , while still directing KG to succinate . However , similar to the glyoxylate shunt , this pathway bypasses the synthesis of ATP via succinate thiokinase . This bypass requires that pools of succinyl-CoA for synthesis of methionine , diaminopimelate , sulfolipids and heme be derived in an energy-dependent manner , either from succinate at the expense of ATP or from methylmalonyl-CoA via methylmalonyl-CoA mutase [9] . Despite this apparent inefficiency , KGD is predicted to play an important role in growth of the Mycobacteria on carbohydrates as the sole carbon and energy source [10] . In most aerobic organisms , the unidirectional oxidative decarboxylation of KG to succinyl-CoA is catalyzed by a ternary complex consisting of dihydrolipoyl dehydrogenase , dihydrolipoyllysine-residue succinyltransferase , and succinyl-transferring KG dehydrogenase . Interestingly , microaerophilic and strictly anaerobic organisms often utilize an alternative enzyme , KG: ferredoxin oxidoreductase ( KOR ) , which can couple the interconversion of KG and succinyl-CoA to the reduction/oxidation of ferredoxin . KOR and other α-ketoic acid: ferredoxin oxidoreductase family members are typically composed of a CoA-coordinating α/γ subunit , and a TPP and iron-sulfur cluster containing β-subunit [11] . Measurement of this activity requires anaerobic conditions , both because these enzymes are irreversibly inactivated by O2 , and because the commonly used chromogenic reporter substrate for the assay is rapidly oxidized under aerobic conditions [12]–[14] . These oxidoreductases have been identified in anaerobes and microaerophiles belonging to all three domains of life [15] , suggesting their presence in the last universal common ancestor . In most cases , KOR is utilized for the reductive carboxylation of succinyl-CoA to KG [16] , [17] . Yet , it has been suggested that the hyperthermophilic anaerobe Thermococcus litoralis might utilize KOR for the generation of succinyl-CoA to support biosynthetic reactions [13] . Here , we demonstrate that while M . tuberculosis can drive a TCA cycle with the canonical intermediates , it does so in an unconventional way using an anaerobic-type KOR . As homologs of KOR are broadly distributed throughout the Actinobacterial class , with the exception of the Corynebacterial and Bifidobacterial families , this enzyme likely plays a greater role in oxidative metabolism than was previously thought . In addition , we find that KOR is dispensable for growth of M . tuberculosis under conditions that promote the utilization of the variant SSA-containing TCA cycle , revealing that these cycles are regulated by different environmental cues . These studies indicate that the KOR pathway operates concurrently with β-oxidation , while the KGD pathway operates under conditions that do not favor fatty acid catabolism . These pathways likely endow M . tuberculosis with metabolic plasticity required for growth on diverse host-derived carbon and energy sources . Since a growing body of evidence indicates that lipids ( for example cholesterol and fatty acids ) are a predominant growth substrate for M . tuberculosis during infection [18]–[22] , we speculate that flux through KOR represents an important step in intermediary metabolism in vivo .
While all other activities of the TCA cycle have been measured from cellular extracts of M . tuberculosis , detection of KG dehydrogenase has remained elusive [6] , [8] . Several anaerobic and microaerophilic organisms encode various evolutionarily related ferredoxin-dependent oxidoreductases that can interconvert specific acyl-CoA thioesters and their cognate α-ketoic acids , such as KG , pyruvate , indolepyruvate and isovalerate [23] . These oxidoreductases are of distinct ancestry from aerobic-type α-ketoic acid oxidoreductases such as pyruvate dehydrogenase and KG dehydrogenase [15] . Surprisingly , while M . tuberculosis does not express measurable KG dehydrogenase activity [6] , it does encode a putative anaerobic-type α-ketoic acid: ferredoxin oxidoreductase ( Figure 1A ) . Products of this locus include a fused α- and γ-subunit encoded by Rv2455c which shows a conserved binding site for coordination of CoA ( GXXGXG ) , and a β-subunit encoded by Rv2454c which shows the highly conserved motif involved in TPP and iron-sulfur cluster binding ( CXGCGXnGDGXnC ) [11] . Based on intergenic distance and an extensive assessment of correlative expression data , Rv2454c and Rv2455c are likely organized in an operon with Rv2452c and Rv2453c [24] . While Rv2452c is of unknown function , Rv2453c encodes a putative molybdopterin-type dinucleotide biosynthesis protein . It was not possible to predict the substrate for this oxidoreductase based on sequence homology alone , thus , α subunits of several biochemically characterized α-ketoic acid: ferredoxin oxidoreductases were aligned to reconstruct a phylogenetic tree ( Figure 1B ) . In most cases , the resulting clades coincided with the preferred α-ketoic acid substrate . The M . tuberculosis homolog grouped within the clade for KG oxidoreductases ( KOR ) . Further , this gene cluster was identified in 23 out of 43 individual Actinobacterial species for which the complete genome sequences were available . A phylogram derived from alignment of these homologs revealed a cladotypic pattern ( Figure 1C ) suggesting that the locus was present in the Actinobacterial ancestor . It is interesting to note that these genes were not identified in Bifidobacterial and Corynebacterial species . Indeed , C . glutamicum has been show to express a bona fide aerobic-type KG dehydrogenase , the sequence of which is highly similar to KGD of M . tuberculosis [25] . French pressure cell lysates prepared from aerobic cultures of M . tuberculosis , M . bovis BCG and M . smegmatis were assessed for methylviologen ( MV , an artificial chromogenic electron acceptor ) reductase activity using various electron donors . Enzymatic assays were performed under anaerobic conditions to prevent potential oxidation of the KOR complex , and to prevent the reoxidation of MV by O2 and components of the aerobic respiratory chain . Consistent with the presence of an anaerobic-type KOR , KG served as an electron donor for reduction of MV ( Table 1 ) . This activity was dependent on the presence of CoA and Mg2+ ( data not shown ) . Unlike some α-ketoic acid: ferredoxin oxidoreductases , the mycobacterial activity was not stimulated by addition of TPP to the reaction mixture . However , a CoA-independent KG dehydrogenase activity was observed in the presence of TPP ( data not shown ) , consistent with the previous report of KGD [8] . To determine whether the KOR product was succinyl-CoA , the reaction mixture was separated by ion exchange chromatography and eluted material was analyzed by UV absorbance at 260 nm . The absorbance profile of the eluted material was compared to those of CoA and succinyl-CoA standards . As shown in Figure 2 , the reaction mixture prepared using cell extract from M . tuberculosis strain mc27000 ( H37Rv ΔpanCD ΔRD1 [26] ) revealed a peak that coincided with that of succinyl-CoA , indicating that a significant fraction of CoA was activated to succinyl-CoA in the reaction mixture . Assays in which pyruvate ( Table 1 ) , glyoxylate , oxaloacetate and 3-indole pyruvate ( data not shown ) were used as electron donors did not yield measurable MV reduction , indicating that the observed α-ketoic acid oxidation is likely specific to KG in M . tuberculosis . In contrast , M . smegmatis extracts could catalyze pyruvate-dependent reduction of MV ( Table 1 ) . Yet , a large fraction of this activity was found to be CoA-independent ( Table 1 ) . Unlike M . tuberculosis and M . bovis , M . smegmatis encodes an additional α-ketoic acid: ferredoxin oxidoreductase homolog , which is likely responsible for this activity . In addition , there was no measurable reduction of other electron carriers , such as NAD+ , NADP+ , FMN , FAD or menadione , when KG was used as an electron donor ( data not shown ) . Due to the presence of a solvent exposed iron-sulfur cluster , most α-ketoic acid: ferredoxin oxidoreductases are rapidly inactivated when exposed to O2 [13] , [14] . Thus , the utility of these enzymes is usually restricted to anaerobic and microaerophilic environments . To determine whether the M . tuberculosis KOR was tolerant to air exposure , cell extracts were incubated under a normal atmosphere at room temperature . At various intervals the extracts were assayed for remaining KOR activity . As controls , air-exposed B . fragilis extracts were assayed for pyruvate: ferredoxin oxidoreductase ( POR ) and KOR . Similar to that which has been described for B . thetaiotaomicron , the B . fragilis POR was rapidly inactivated following air exposure ( Figure 3 ) [14] , as was the B . fragilis KOR ( Figure 3 ) . In contrast , when M . tuberculosis lysates were exposed to air the KOR activity was remarkably stable ( Figure 3 ) , indicating that M . tuberculosis KOR remains functional under aerobic conditions . To determine the contribution of Rv2454c-Rv2455c ( herein referred to as korAB ) to KOR activity , the respective genes were deleted using specialized transduction [27] . Consistent with a previous study of gene essentiality in M . tuberculosis , mutations in Rv2454c-Rv2455c could be tolerated [28] . Codon 50 of korA through codon 334 of korB were replaced with a hygromycin resistance cassette ( Figure 1A ) . While cell extracts from the resulting strain had wild type levels of NADH: MV oxidoreductase activity , there was no measurable reduction of MV using KG as an electron donor ( Table 1 ) and succinyl-CoA production was diminished to <5% of that from the wild type extract ( Figure 2 ) . Importantly , introduction of an intact copy of korAB restored both KOR activity and succinyl-CoA production ( Table 1 , Figure 2 ) . These results demonstrate that the M . tuberculosis korAB gene cluster codes for a KOR that is expressed and stable under fully aerobic conditions . Based on the lack of detectable KG dehydrogenase activity [6] and the presence of KGD [8] , it has been proposed that M . tuberculosis catalyzes a variant TCA cycle in which succinyl-CoA is replaced by SSA [8] . However , as KOR is active in aerobically grown M . tuberculosis , it is possible that this enzyme can also functionally replace KG dehydrogenase in the TCA cycle . Consistent with a role for KOR in oxidative metabolism in M . tuberculosis , the ΔkorAB strain was incapable of growth on solid medium unless the atmosphere was supplemented with abundant CO2 ( Figure 4A ) . Indoor ambient CO2 levels were found to range from 0 . 078% to 0 . 084% during the course of these experiments . This CO2-dependent phenotype was also observed for M . tuberculosis strains H37Ra , CDC1551 and M . bovis BCG in which korAB was deleted ( data not shown ) . In liquid medium , growth of the KOR-deficient strain was similar to that of the wild type strain when abundant CO2 was supplied ( Figure 4B ) , whereas this strain was retarded under ambient air ( Figure 4C ) and fully inhibited upon further CO2 restriction ( Figure 4D ) . Importantly , introduction of a cosmid containing Rv2425c-Rv2456c abolished this CO2 dependency ( Figure 4A , C , D ) . This graded response to CO2 indicates that KOR-dependent decarboxylation of KG is an important source of CO2 in M . tuberculosis metabolism . It is predicted that the KOR-deficient strain is capable of growth with a broken TCA cycle due to the presence of the glyoxylate shunt . While this mode bypasses the production of CO2 , it permits the extraction of reducing equivalents and production of biosynthetic precursors from two carbon units that enter the cycle . To determine whether the glyoxylate shunt is essential for growth in the absence of KOR , strains were plated on medium containing the isocitrate lyase ( ICL ) inhibitor 3-nitropropionate ( 3NP ) [21] , [29] . Surprisingly , 3NP was found to alleviate the CO2 requirement of the KOR mutant strain ( Figure 5A ) . Moreover , this heightened CO2 dependency was also diminished by exclusion of fatty acids , namely oleic acid and Tween 80 ( an oleic acid-polyethylene ester used to prevent cell aggregation ) , which are standard components of mycobacterial growth media ( Figure 5B ) . Thus , while KOR is important for CO2 metabolism in the presence of exogenously supplied fatty acids , suppression of fatty acid utilization appears to promote activity of a compensatory pathway . As glyoxylate , a product of ICL , can inhibit SSA dehydrogenase [30] , it is possible that the variant TCA cycle proposed by Tian et al [8] is favored under conditions in which catabolism of exogenous fatty acids is dampened . To determine whether KOR is essential for growth of M . tuberculosis on fatty acids as the sole carbon source , the ΔkorAB mutant was grown in medium containing Tween 80 , which can be hydrolyzed by mycobacteria to form oleic acid and an inert non-metabolizable ethylene polymer . Similar to that which was observed for growth on mixed carbon sources , the ΔkorAB mutant grew nearly as well as the wild type strain on Tween 80 when the atmosphere was supplemented with 5% CO2 ( Figure 5C ) . As expected , 3NP inhibited growth of the wild type strain , indicating that the glyoxylate shunt is essential for growth on this source of oleic acid . Further , when the ΔkorAB strain was grown under an atmosphere with ambient levels of CO2 , there was a modest growth defect that could not be reversed by supplementation with succinate ( Figure 5D ) . Unlike that which was observed with mixed substrates , the ΔkorAB strain grew poorly under further CO2 restriction ( Figure 5E ) , suggesting that gluconeogenesis might provide enough additional CO2 to support growth in the absence of KOR . To determine whether KGD is important for intermediary metabolism , kgd ( Rv1248c ) was deleted in M . tuberculosis mc27000 and in the KOR deficient strain . ΔRv1248c mutants were readily obtained despite the prediction that Rv1248c is essential for growth of M . tuberculosis on standard medium [31] . In medium containing both carbohydrates ( dextrose and glycerol ) and fatty acids ( Tween 80 ) , under a CO2 enriched atmosphere , growth of the Δkgd strain was indistinguishable from that of the wild type strain regardless of the presence of 3NP ( Figure 6A , B ) . Thus , under these conditions KOR is sufficient to maintain flux through the TCA cycle . However , the ΔkorAB Δkgd strain showed a slower growth rate relative to the wild type and single mutant strains ( Figure 6A ) , indicating that either pathway can function to some degree in the M . tuberculosis TCA cycle . As growth of the ΔkorAB Δkgd strain was fully inhibited by the presence of 3NP ( Figure 6B ) , blockade of all three pathways results in arrest of intermediary metabolism . When the Δkgd strain was cultivated in medium containing carbohydrates as the sole carbon source in the presence of 5% CO2 , there was a marked defect compared to the wild type and ΔkorAB strains ( Figure 6C ) . Interestingly , growth of the ΔkorAB Δkgd strain was similar to that of the Δkgd single mutant ( Figure 6C ) , indicating that KOR contributes minimally during growth on carbohydrates as the sole carbon source . This notion is further supported by the observation that growth of the Δkgd and ΔkorAB Δkgd strains was fully inhibited by the presence of 3NP , whereas growth of the wild type and ΔkorAB strains was unaffected ( Figure 6D ) . The growth defects observed for both the Δkgd and ΔkorAB Δkgd strains were exacerbated by incubation under an atmosphere with an ambient level of CO2 ( Figure 6E ) . As the growth defect of the ΔkorAB Δkgd strain was more severe than that of Δkgd alone , KOR appears to have a minimal contribution to intermediary metabolism under these conditions . Growth of these strains was markedly improved by supplementation with succinate ( Figure 6E ) , indicating that the growth defects of these strains are due both to limitations in generation of succinate and CO2 . These observations indicate that KGD plays a predominant role in growth on carbohydrates as the sole carbon source . Growth of the Δkgd strain on Tween 80 as the sole carbon source was similar to that of the wild type strain regardless of the presence of CO2 ( Figure 6F ) , consistent with a primary role for KOR in the TCA cycle under conditions that favor β-oxidation . In contrast , the ΔkorAB Δkgd strain was significantly more retarded for growth on Tween 80 than was either ΔkorAB or Δkgd alone ( Figure 5C , 6F ) . Thus , while KOR is the major mediator for conversion of α-ketoglutarate during growth on fatty acids , KGD can also contribute to a minimal degree .
A previous report indicates that M . tuberculosis lacks a canonical TCA cycle , as CoA-dependent KG dehydrogenase activity was undetectable in crude cellular extracts [6] . Until recently it was unclear whether mycobacteria require an intact TCA cycle as they can produce TCA cycle-derived biosynthetic precursors via the glyoxylate shunt [9] , although doing so would require that succinyl-CoA be formed in an energy dependent manner . Since isocitrate lyase ( ICL ) is dispensable for growth on carbohydrates as a carbon source [20] , [21] , it is likely that a TCA cycle of some form exists in M . tuberculosis . Based on biochemical studies of KG decarboxylase ( KGD ) and SSA dehydrogenase , it was recently proposed that SSA replaces succinyl-CoA in the M . tuberculosis TCA cycle [8] . While this pathway should support growth on carbohydrates when the glyoxylate shunt is inoperable ( Figure 7 ) , it still requires that succinyl-CoA be produced by alternate means . In addition to enzymes of this alternate pathway , M . tuberculosis and other mycobacterial species encode an anaerobic-type α-ketoic acid: ferredoxin oxidoreductase homolog that is most closely related to those that interconvert KG and succinyl-CoA . Here , we demonstrate that M . tuberculosis contains a KOR activity that results in the formation of succinyl-CoA , and requires the korAB gene cluster . As we were unable to identify the physiologic electron acceptor for this enzyme , it is currently unclear how KOR feeds into the cellular reduction/oxidation pools . Yet , since the M . tuberculosis KOR α subunit contains two hydrophobic stretches ( from amino acids 249–273 and 304–340 ) , it is possible that the complex is membrane associated and is reoxidized following interaction with another membrane associated redox partner . KOR was probably overlooked in previous studies because of the requirement for anaerobic assay conditions resulting from the use of an O2-reactive reporter dye . In contrast to similar enzymes from the obligate anaerobe B . fragilis , the M . tuberculosis KOR activity was stable during extensive air exposure . Interestingly , sequence of the predicted iron-sulfur cluster coordination site of the M . tuberculosis enzyme was similar to that of other O2-sensitive KORs and does not contain a recognizable stabilization domain found in some δ-proteobacterial PORs [32] . While it is possible that the mycobacterial KOR is intrinsically aerostable , it is also possible that there might be an unidentified stabilization partner . Despite the peculiarity of finding such an oxidoreductase in an obligately aerobic organism , KOR is conserved in many other Actinobacteria . As this class of eubacteria includes both aerobes and anaerobes , it is possible that the Actinobacterial ancestor was a facultative anaerobe , and that more stringent O2 requirements may have arisen with divergence of the various clades . Indeed , mycobacteria possess a suite of genes commonly associated with anaerobic metabolism , such as an anaerobic-type ribonucleotide reductase [33] and a respiratory nitrate reductase [34] . Although conditions for anaerobic cultivation of mycobacteria have yet to be defined , it has recently been demonstrated that M . tuberculosis can grow under an atmosphere containing as little as 1 . 3% O2 when provided with supplemental CO2 [35] . Thus , dampening oxidative metabolism results in a CO2 deficit , and likely leads to defects in lipid , arginine , adenine and uracil biosynthesis [36] . Consistent with its role in oxidative metabolism , we find that KOR is essential for growth of M . tuberculosis under an ambient atmosphere where the CO2 concentration is relatively low . As outlined in Figure 7 , KOR functions concurrently with the glyoxylate shunt , likely to provide both succinyl-CoA and CO2 , as well as reducing equivalents . Interestingly , KOR is dispensable for growth when the glyoxylate shunt is inoperable . Further genetic analysis revealed that KGD is essential for this bypass , indicating that the SSA pathway operates under conditions where utilization of exogenous fatty acids is minimal ( Figure 7 ) . Along these lines , we find that while KGD plays a critical role during growth on carbohydrates , it contributes little during growth in medium containing fatty acids . These observations suggest the presence of a metabolic regulatory cascade that is responsive to β-oxidation . It was recently shown that the forkhead-associated protein GarA inhibits KGD activity in M . smegmatis [10] and its homolog OdhI inhibits KG dehydrogenase in Corynebacterium glutamicum [25] . In both cases , the serine threonine kinase PknG was found to modulate GarA activity via phosphorylation , although the signal for this regulatory cascade has yet to be described . In M . smegmatis , constitutive inhibition of KGD via unphosphorylated GarA results in a profound growth defect on dextrose and glycerol [10] , indicating that this decarboxylase is predominantly utilized for growth on carbohydrates . Based on these findings , we predict that glyoxylate serves as the metabolic trigger for inhibition of KGD , likely through inhibition of PknG-mediated GarA phosphorylation ( Figure 7 ) . Furthermore , since we find that KOR cannot compensate for the loss of KGD during growth on carbohydrates alone , KOR must be subject to some means of regulation that has yet to be identified . While a comprehensive analysis of the genetic requirements for growth and survival of M . tuberculosis indicates that genes linked to carbohydrate metabolism might be important early during infection [22] , a large body of evidence indicates that lipids represent the major carbon source for growth and persistence of M . tuberculosis in vivo [18]–[22] . When lipids are processed via β-oxidation pathway , they are broken down in a series of steps into acetate and propionate ( reviewed in [37] ) . Studies of ICL ( for acetate utilization ) and methylcitrate lyase ( MCL , for propionate utilization ) suggest that fatty acids might be important lipidic carbon sources for M . tuberculosis during infection [20] , [21] . Yet , it has recently been demonstrated that sterols , such as cholesterol , can be also be catabolized by M . tuberculosis both in vitro and in vivo [18] , [19] . As cholesterol catabolism also results in the formation of propionate and acetate , ICL and MCL are predicted to be required for use of sterols as well [38] . Given the concurrent function of the KOR pathway with the glyoxylate shunt , we predict that flux of KG runs largely through KOR , rather than KGD , during growth in vivo . The availability of these mutant strains will allow us to distinguish between these possibilities .
M . tuberculosis strains used in this study ( Table 2 ) were derived from strain H37Rv and were routinely cultivated using Middlebrook 7H9 and 7H10 media ( Difco , Sparks , MD ) supplemented with NaCl ( 0 . 85 mg ml−1 ) , oleic acid ( 60 nl ml−1 ) , bovine albumin-fraction V ( 5 mg ml−1 ) , dextrose ( 2 mg ml−1 ) , and glycerol ( 5 mg ml−1 ) . As indicated in the text , Tween 80 ( 0 . 5 mg ml−1 or 5 mg ml−1; a non-ionic surfactant ) or tyloxapol ( 0 . 5 mg ml−1; a non-metabolizable non-ionic surfactant ) were added to the growth medium . 100 µg ml−1 pantothenic acid was added for growth of pantothenic acid auxotrophic strains [26] . M . smegmatis strain mc2155 [39] was cultivated with 7H9 and 7H10 supplemented with dextrose ( 2 mg ml−1 ) and tyloxapol ( 0 . 5 mg ml−1 for liquid media ) . E . coli strain HB101 , used for plasmid , cosmid and phasmid manipulation , was cultivated using LB medium . B . fragilis strain ATCC 25285 was cultivated using brain heart infusion medium supplemented with 5 µg ml−1 hemin and 5 mg ml−1 yeast extract [40] , or using modified anaerobic minimal medium [41] . Bactoagar ( 1 . 5% ) was added to media when necessary . 3-nitropriopionate ( 200 µM ) , carbenicillin ( 50 µg ml−1 ) , hygromycin ( 50 µg ml−1 for M . tuberculosis , 150 µg ml−1 for E . coli ) and kanamycin ( 20 µg ml−1 for M . tuberculosis , 40 µg ml−1 for E . coli ) were added to the growth media as needed . For growth experiments involving modified atmospheres , cultures were incubated in a controlled atmosphere chamber ( Coy Laboratory Products , Grass Lake , MI ) or in sealed 2 . 5 L AnaeroPack boxes ( Mitsubishi Gas Chemical Co . , Inc . ) . Atmospheric CO2 supplementation was regulated using an AC100 CO2 controller ( Coy Laboratory Products ) . Atmospheric CO2 restriction was achieved by absorption with 40 ml of a 5% ( w/v ) solution of KOH [42] . Under such conditions , atmospheric CO2 was maintained at a level of 0 . 04% . Atmospheric CO2 determinations were made using a Bacharach model 2810 CO2 analyzer ( New Kensington , PA ) , or a Gray Wolf DirectSense™ indoor air quality monitor ( Trumbull , CT ) . Purification of cosmids , plasmids and PCR products were performed using Qiagen products following the manufacturer's suggestions . Genetic manipulations of mycobacterial species were performed as described in [43] . M . tuberculosis strains mc27010 ( ΔkorAB ) and mc27012 ( Δkgd ) were constructed by allelic exchange using specialized transduction . The allelic exchange phasmids were constructed by amplifying 1 kb regions flanking korAB and kgd with primers described in Table 2 . Purified DNA fragments were digested with indicated restriction enzymes . Fragments were ligated with Van91I fragments of p0004S ( T . Hsu , unpublished ) using T4 DNA ligase ( NEB ) . The resulting allelic exchange substrates were digested with PacI , ligated to phAE159 [44] and packaged with MaxPlax packaging extract ( Epicenter Biotechnologies ) for propagation as shuttle phasmids in E . coli . Phasmids were electroporated into M . smegmatis mc2155 for phage propagation . Allelic exchange substrates were delivered to M . tuberculosis as previously described [27] . Mutant strains were confirmed by PCR . For construction of the double mutant strain , it was necessary to first resolve the hygromycin resistance cassette ( which was flanked by γδ resolvase recognition sequences ) in mc27010 using the γδ resolvase containing vector pJH532 . Cosmid pBH33K used for complementation of strain mc27010 contains base pairs 2721932–2757635 of the M . tuberculosis H37Rv genome . Enzymatic assays were performed using French pressure cell lysates prepared from mycobacteria grown in aerobic supplemented 7H9 medium , or from B . fragilis grown in modified anaerobic minimal medium . Cells were harvested by centrifugation at 3 , 500 X g for 10 min at 4°C . The following steps were performed in an anaerobic chamber ( <1 ppm O2 , Coy Laboratory Products ) containing a gas mixture of 10% H2 , 5% CO2 and 85% N2 . Cells were washed in an equal volume of chilled anaerobic 100 mM sodium phosphate ( pH 7 . 2 ) , and resuspended in 0 . 1 volume chilled anaerobic phosphate buffer . Anaerobic phosphate buffer was prepared in the anaerobic chamber using distilled H2O that had been degassed by boiling immediately before introduction into the anaerobic atmosphere . Samples were removed from the chamber and cells were lysed under 70 kg cm−2 using a French pressure mini-cell and maintained under a stream of N2 to exclude O2 . Extracts were clarified by centrifugation at 13 , 000 X g for 10 min at 4°C , flash frozen with liquid N2 and stored at −80°C until ready to use . Reagents for enzymatic assays were purchased from Sigma-Aldrich and were dissolved in anaerobic phosphate buffer . Reaction mixtures were prepared under anaerobic conditions using anaerobic phosphate buffer containing 2 . 5 mM MgCl2 in QS-517-S quartz screw top cuvettes ( Nova Biotech , El Cajon , CA ) . α-ketoic acid: MV oxidoreductase assays were performed as described [13] , with the following modifications: 0 . 5 mM coenzyme A , 5 mM MV and 25 mM α-ketoic acid ( pyruvate or KG ) were added , and TPP was excluded . NADH: MV oxidoreductase activity was measured using 0 . 2 mM NADH and 2 . 5 mM MV . Reactions were started by addition of cell extract and MV reduction was measured spectrophotometrically at room temperature . Reduction values were based on an absorption coefficient of 12 , 000 M−1 cm−1 at 600 nm [45] . Protein estimations were made using the BioRad protein assay reagent . For analysis of succinyl-CoA production , reaction mixtures were prepared as described above , however , 100 µM CoA and 50 mM MV were used . Reaction mixtures were passed through a 0 . 2 µm filter and incubated under anaerobic conditions for 30 minutes . Samples were stored at −80°C until time of analysis . CoA species were separated by ion exchange chromatography using an Äkta Explorer ( Amersham Biosciences ) . CoA standards and reaction mixtures ( 500 µl ) were injected onto a Mono Q HR 5/5 column in 50 mM potassium phosphate ( pH 6 . 5 ) 50 mM NaCl . Nucleotides were eluted from the column with a linear gradient from 50 to 350 mM NaCl in 10 column volumes at a flow rate of 2 ml/min , and detected by UV absorbance ( 260 nm ) . Sequences for phylogenetic analyses were acquired from the NCBI protein database ( www . ncbi . nlm . nih . gov ) . Accession numbers corresponding to Figure 1B are as follows: Pyrococcus abyssi , NP_127041; P . horikoshii , NP_142630; P . furiosus , Q51801; Thermococcus kodakarensis , YP_184393; P . furiosus , Q51804; Thermotoga maritima , O05651; Helicobacter pylori , AAC38206; Pantoea agglomerans , X78558; Klebsiella pneumoniae , CAA31665; Rhodospirillum rubrum , X77515; Desulfovibrio africanus , CAA70873; Trichomonas vaginalis , U16822; T . vaginalis , U16823; Halobacterium salinarum , CAA45825; Hydrogenobacter thermophilus , BAB21494; M . tuberculosis , CAA16032; Sulfolobus sp . , BAA10898; H . pylori , AAC38211; Methanothermobacter thermautotrophicus , NP_276168; P . horikoshii , NP_142702; P . furiosus , NP_578262; P . abyssi , NP_126972; T . kodakarensis , YP_182549 . Those corresponding to Figure 1C are as follows: M . tuberculosis , CAA16032; M . bovis , NP_856129; M . ulcerans , ABL05848; M . avium , YP_880944; M . gilvum , ABP45093; M . vanbaalenii , ABM14742; M . smegmatis , YP_888909 . 1; Mycobacterium sp . KMS , YP_939624; pRHL2 , Rhodococcus jostii , ABG94195; Nocardia farcinica , BAD58508; Saccharopolyspora erythraea , CAM03198; Salinispora arenicola , ABV96568; S . tropica , ABP53176; F . alni , CAJ59716; Frankia sp . CcI3 , ABD09939; Frankia sp . EAN1pec , ABW15421; Acidothermus cellulolyticus , ABK52058; Streptomyces coelicolor , CAB60189; S . coelicolor , CAC08296; S . avermitilis , BAC72589; Thermobifida fusca , AAZ56707; Propionibacterium acnes , EEB67312; Nocardioides sp . JS614 , ABL80079; Janibacter sp . HTCC2649 , EAQ00280; F . alni , CAJ65422; Frankia sp . CcI3 , ABD13820; Frankia sp . EAN1pec , ABW16201; M . thermautotrophicus , NP_276168; H . pylori , AAC38211; H . salinarum , CAA45825; H . thermophilus , BAB21494; Sulfolobus sp . , BAA10898 . R . jostii pRHA1 KorA was assembled from translated sequence derived from NC_008270 . | Knowledge of the basic biology of Mycobacterium tuberculosis is essential to identifying novel ways to combat the emerging threat of drug-resistant tuberculosis . Since the tricarboxylic acid ( TCA ) cycle is a cornerstone of metabolism and M . tuberculosis does not possess a “typical” TCA cycle enzyme set , much effort has been focused on elucidating the components of this pathway . Previous reports indicate that M . tuberculosis possesses a variant TCA cycle in which succinic semialdehyde replaces succinyl-CoA . Since this pathway does not conserve as much metabolic energy as the canonical pathway , we considered an alternative hypothesis: that M . tuberculosis might possess an anaerobic type α-ketoglutarate dehydrogenase . In this manuscript , we investigate this previously unknown activity for mycobacteria using a combination of genetic and biochemical approaches , and demonstrate that M . tuberculosis is capable of driving a conventional TCA cycle in an unconventional way . We also validate the existence of the previously described variant pathway and provide evidence that these two pathways are differentially utilized in response to a metabolic signal , fatty acid catabolism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microbiology/microbial",
"physiology",
"and",
"metabolism"
] | 2009 | An Anaerobic-Type α-Ketoglutarate Ferredoxin Oxidoreductase Completes the Oxidative Tricarboxylic Acid Cycle of Mycobacterium tuberculosis |
It is believed that schistosomes evade complement-mediated killing by expressing regulatory proteins on their surface . Recently , six homologues of human CD59 , an important inhibitor of the complement system membrane attack complex , were identified in the schistosome genome . Therefore , it is important to investigate whether these molecules could act as CD59-like complement inhibitors in schistosomes as part of an immune evasion strategy . Herein , we describe the molecular characterization of seven putative SmCD59-like genes and attempt to address the putative biological function of two isoforms . Superimposition analysis of the 3D structure of hCD59 and schistosome sequences revealed that they contain the three-fingered protein domain ( TFPD ) . However , the conserved amino acid residues involved in complement recognition in mammals could not be identified . Real-time RT-PCR and Western blot analysis determined that most of these genes are up-regulated in the transition from free-living cercaria to adult worm stage . Immunolocalization experiments and tegument preparations confirm that at least some of the SmCD59-like proteins are surface-localized; however , significant expression was also detected in internal tissues of adult worms . Finally , the involvement of two SmCD59 proteins in complement inhibition was evaluated by three different approaches: ( i ) a hemolytic assay using recombinant soluble forms expressed in Pichia pastoris and E . coli; ( ii ) complement-resistance of CHO cells expressing the respective membrane-anchored proteins; and ( iii ) the complement killing of schistosomula after gene suppression by RNAi . Our data indicated that these proteins are not involved in the regulation of complement activation . Our results suggest that this group of proteins belongs to the TFPD superfamily . Their expression is associated to intra-host stages , present in the tegument surface , and also in intra-parasite tissues . Three distinct approaches using SmCD59 proteins to inhibit complement strongly suggested that these proteins are not complement inhibitors and their function in schistosomes remains to be determined .
Schistosomiasis is an important parasitic disease , caused by trematode worms of the genus Schistosoma , affecting more than 200 million people worldwide , with a further 650 million individuals living at risk of infection , remaining a major public health problem in many developing countries [1] . Adult worms are able to survive for decades in the hostile blood environment of their vertebrate host , apparently unharmed by circulating leukocytes , antibodies and the complement system . Thus , the parasite must have developed strategies to evade the host's immune defenses . One of the most important modifications is the new tegument surface organization that develops immediately after penetration of cercariae into the skin and their transformation into schistosomula . The tegument is a thin syncytial layer that covers the whole parasite , limited by a basal membrane and a multilaminate surface membrane complex , which constitutes the major host–parasite interface [2] . Schistosomula are at first sensitive to complement killing , but rapidly become highly resistant to complement attack by both the Classical [3] , [4] and Alternative Pathways [5] , [6] . The later developmental stages of the parasite , i . e . , the lung schistosomula and adult worms , have also been shown to be refractory to both pathways of complement [3] , [6] , [7] , [8] . Intravascular parasites must also be resistant to a third complement activation pathway - the Lectin Pathway [9] . However , the precise mechanisms of complement resistance have not been fully characterized . The strategies used by the parasite to subvert complement attack are most likely related to the function and composition of the tegument ( reviewed by [10] ) . The apical membrane is considered poorly immunogenic due to the limited number of exposed proteins and the acquisition of a variety of host molecules , which may mask important surface proteins [11] , [12] , [13] . Shedding of the outer coat has been considered as a mechanism to eliminate surface-bound complement and immune-complexes [14] . There are also reports of serine proteases in the surface of freshly transformed schistosomula , lung stage and adult worms , that may play a role in complement resistance by cleaving immuno-complexes and complement proteins [15] , [16] . One of the most intriguing strategies that has been proposed as a mechanism by which schistosomes escape complement attack involves the presence in the tegument of several complement regulatory proteins , including two of host origin ( reviewed by [17] , [18] ) . The binding activity to the complement components C1 [19] , [20] , C3 [21] , [22] , [23] and C8/C9 [24] , have been reported to be present at the surface of schistosome parasites . It is important to note that while some of these parasite proteins ( e . g . paramyosin ) have been isolated and shown to interact with complement components in vitro , the definitive protective role of these proteins in vivo has not yet been demonstrated . In the last seven years , several proteomics studies attempted to define the tegument protein composition , as well as its surface exposed molecules [25] . From these studies , one surprising finding was the detection of the host complement components C3 and C4 , but not those required for formation of the Membrane Attack Complex ( MAC ) , i . e . C5b to C9 molecules , as revealed by biotinylation studies of the tegument surface [12] . A reasonable explanation is that the complement fixation is initiated , but then inhibited to prevent MAC formation . A mouse C3 complement regulatory ( Crry ) -like molecule has also been detected on the tegument surface by proteomics [12] . Host cells are protected from MAC-mediated lysis mainly by CD59 , a 18–21 kDa glycosylphosphatidyl-inositol-linked membrane glycoprotein that inhibits polymerization of C9 by binding to C8α and C9 [26] , thus preventing the formation of the cytolytic MAC . Earlier studies indicated that the exposed form of the schistosome muscle protein paramyosin , was able to inhibit the assembly of C5b-9 by binding to C8 and C9; additionally , this protein was reportedly recognized by rabbit anti-human CD59 antiserum [27] . However , the in vivo significance of paramyosin-complement interactions still awaits further clarification . Recently , Wilson and Coulson [28] identified in the schistosome genome six homologues of human CD59 , containing 20–30% amino acid identity which rise to >40% if conservative amino acid substitutions are included . One of these molecules ( CD59b , formerly Dif 5 ) was described by our group as a vaccine candidate , due to its up-regulated expression in the schistosomulum stage [29] . Furthermore , in another approach to select vaccine candidates , two members of this family ( CD59a and CD59b ) were identified within a group of molecules exposed on the parasite's tegument by proteomics and molecular shaving with phosphatidylinositol-specific phospholipase C ( PiPL-C ) treatment of live adult worms [13] . More recently , two other isoforms similar to CD59 ( Smp_166340 and Smp_081920 , GeneDB , ( http://www . genedb . org/Homepage/Smansoni ) were reported as membrane-associated tegumental proteins by proteomic analysis [25] . Therefore , it is tempting to speculate whether these six homologues could act as CD59-like complement inhibitors in schistosomes as part of an immune evasion strategy , especially because two of them were found on the tegument surface . The CD59 family possesses the Three-Finger Protein Domain fold ( TFPD ) [30] , that is also a feature of proteins with several distinct sequence and structural attributes , such as the receptors of activins , bone morphogenetic proteins , Mullerian inhibiting substance , transforming growth factor-β receptor II , C4 . 4a ( a structural homologue of the urokinase receptor ) , urokinase/plasminogen activatory receptor ( uPAR ) and several members of Ly6 family . The Ly6 molecules ( lymphocyte differentiation antigens ) were among the first cell surface molecules identified in mouse [31] and there is emerging evidence showing their role in cell signaling , cell adhesion and cellular activation [32] . The TFPD superfamily is characterized by the structural conservation of at least six half-cystines forming three disulfide bridges ( B1 , B2 and B4 ) , five β-strands and one asparagine adjacent to the N-terminal of the last half-cystine from the last disulfide bridge B4 [33] . A very striking characteristic of this domain is the finger-shaped spatial conformation that the amino acid backbone acquires between the two half-cystines of the same disulfide bridge [30] . In the current work , we describe the molecular characterization of seven putative SmCD59-like genes from genome assembly version 5 ( http://www . genedb . org/Homepage/Smansoni ) and attempt to address their putative biological function . Our data confirms up-regulation in the transition to intra-host stages . However , the functional studies performed with the two CD59-like members ( CD59a and CD59b , named in this study as SmCD59 . 1 and SmCD59 . 2 ) , previously identified at the host-parasite interface , did not show any complement inhibition activity . Therefore , the function of these seven proteins in schistosomes remains to be established .
The life cycle of S . mansoni ( BH strain ) was maintained in the laboratory by routine passage through mice and the intermediate snail host Biomphalaria glabrata . S . mansoni eggs were extracted from infected mouse livers and miracidia were hatched from S . mansoni eggs , as previously described [34] . Schistosomula were cultivated in culture medium after transformation of cercariae , as previously described [35] . Adult worms were obtained by perfusion of the portal hepatic and intestinal veins from hamsters , 7–8 weeks after infection with approximately 100 cercariae . The procedures involving animals were carried out in accordance with the Brazilian legislation ( 11790/2008 ) . All animals were handled in strict accordance with good animal practice and protocols were previously approved by the Ethical Committee for Animal Research of Butantan Institute ( CEUAIB , São Paulo , Brazil ) , under the license number 603/09 . The SmCD59 nucleotide sequences were identified searching the v5 . 0 of S . mansoni genome assembly ( GeneDB ) ( http://www . genedb . org/genedb/smansoni/ ) . The search for conserved domains was performed using SMART ( http://smart . embl-heidelberg . de/ ) . The molecular weight ( MW ) and isoelectric point ( pI ) were calculated with the Compute pI/Mw tool ( http://www . expasy . ch/tools/pi_tool . html ) . Post-translational modification predictions: the signal peptide prediction was performed using the SignalP 4 . 0 server ( http://www . cbs . dtu . dk/services/SignalP/ ) [36]; potential GPI-modification sites were analyzed by big-PI Predictor ( http://mendel . imp . ac . at/sat/gpi/gpi_server . html ) [37]; and N-glycosylation sites with the NetNGlyc version 1 . 0 algorithm . Protein sequences alignments were performed using the ClustalX 2 software . Homology modeling was done with MODELLER [38] using the crystal structure of human CD59 ( hCD59 ) ( 2UWR ) as a template . The modeled structure was visualized with PyMOL ( The PyMOL Molecular Graphics System , Version 1 . 5 . 0 . 4 Schrödinger , LLC ) and the stereochemical quality of the model was examined using the program PROCHECK [39] , which evaluates the geometry of residues in the model when compared with the stereochemical parameters from the template . Additional algorithms such as WHAT_CHECK , ERRAT , VERIFY_ 3D , PROVE and CRYST1 record matches ( available from: http://services . mbi . ucla . edu/SAVES/ ) were also used to assess the quality of the model generated . We searched the Schistosoma Genomic Resources SchistoDB ( V3 . 0 ) ( http://schistodb . net/schisto/ ) and the GenBank ( http://www . ncbi . nlm . nih . gov/ ) to identify similar sequences in S . mansoni , S . japonicum and S . hematobium , using SmCD59 . 1 and SmCD59 . 2 as queries . Additionally , BLAST and PSI-BLAST searches against the non-redundant protein sequence database were used to identify similar sequences in other Platyhelminthes ( http://smedgd . neuro . utah . edu/blast . php and http://bioinfosecond . vet . unimelb . edu . au ) , as well as in representative mammals . For phylogenetic analysis , alignments of protein sequences were performed using the ClustalX 2 software . The tree was constructed using ClustalX 2 using the Neighbor-Joining method . The numbers represent the confidence of the branches assigned by bootstrap ( in 1000 samplings ) . The TreeView program [40] was used to visualize and analyze the tree . In order to establish the level of expression of each SmCD59 gene throughout the parasite life cycle , total RNA was extracted from adult worms using TRIzol ( Life Technologies ) and from schistosomula , cercariae , miracidia and eggs using the Kit RNAspin mini ( GE Healthcare , USA ) , as per the manufacturer's recommendations . The RNA was quantified by spectrophotometry ( NanoDrop 1000 , Thermo Fischer Scientific ) and the quality was analyzed in the Agilent 2100 Bioanalyzer . The cDNA synthesis and the quantitative Real-Time PCR ( qRT-PCR ) reactions using SYBR Green ( Life Technologies ) were performed according to [41] . The primers were designed in the software Primer Express ( Applied Biosystems ) to span exon/exon boundaries avoiding amplification of contaminating genomic DNA ( Table S1 ) . S . mansoni alpha-tubulin ( Smp_090120 . 1 ) was chosen as normalizing gene . Quantitation of relative differences in expression between the stages was calculated by the comparative 2−ΔΔCt method [42] , using the parasite stage with lowest gene expression as calibrator for each gene independently . In an attempt to compare the levels of gene expression among the seven different SmCD59 isoforms , we performed qRT-PCR using TaqMan Gene Expression Assay ( Life Technologies/Applied Biosystems , CA ) as previously described [43] . The life cycle stages examined were cercariae , schistosomula cultured for 11 days and adults . RNA was extracted from each life cycle stage using the Trizol method ( Life Technologies , CA ) and the cDNA was synthesized using 1 µg of high quality total RNA , pre-treated with TurboDNAse ( Life Technologies ) , oligo ( dT ) and Superscript reverse transcriptase III ( Life Technologies ) . qRT-PCR was performed using cDNA equivalent to 50 ng total RNA . The set of primers and MGB reporter probe , labeled with 6-carboxyfluorescein ( FAM ) specific for the detection of each SmCD59 were custom synthesized by Applied Biosystems ( Life Technologies ) and are shown in Table S1 . Primers/probe positions were designed to span exon/exon boundaries to minimize detection of any contaminating genomic DNA . The qRT-PCR reactions were run in triplicate and underwent 45 amplification cycles on the StepOne Plus System Instrument ( Applied Biosystems ) . The 2−ΔΔCt method was employed for relative quantification [42] with S . mansoni triose phosphate isomerase ( SmTPI , Smp_003990 ) as the normalizing gene . We used the expression of SmCD59 . 6 from adult worm stage as calibrator to calculate the relative expression of all other SmCD59 analyzed , because this gene had the lowest expression in adult worms . The sequence from S . mansoni EST assembled contig SmCD59 . 2 ( Smp_105220 , GeneDB ) was redesigned excluding the signal peptide sequences and manufactured by DNA 2 . 0 , Inc . USA ( https://www . dna20 . com/ ) using DNA2 . 0 optimization algorithms for expression in Pichia pastoris ( Table S1 ) . The fragments corresponding to the mature protein sequences for SmCD59 . 1 ( Smp_019350 , GeneDB ) ( from H28 to F126 ) and SmCD59 . 2 ( from K21 to A97 ) were digested with EcoRI and XbaI to generate inserts with overhang ends that were purified and cloned into the same sites for the expression vector pPICZαA ( Life Technologies ) , to produce a protein that contained a C-terminal hexa-histidine tag . The resulting constructs were sequenced to confirm their identity . The SmCD59 . 2 was also expressed and purified from E . coli . The 5′ and 3′ oligonucleotides were designed using the S . mansoni genome assembly sequence ( Smp_105220 ) . The SuperScriptTM First-Strand Synthesis System for RT-PCR ( Life Technologies ) was used to amplify a fragment from C27 to H101 of the mature SmCD59 . 2 protein . The PCR fragments were purified from agarose gel electrophoresis and digested with XhoI and KpnI to be cloned into pAE-6His vector [44] and sequenced to confirm identity . The plasmids containing the gene fragments pPICZ-α-SmCD59 . 1 and pPICZ-α-SmCD59 . 2 ( optimized sequence ) , were linearized with SacI and used to transform P . pastoris strain GS115 ( Life Technologies ) by electroporation . Putative multi-copy recombinants were selected following the instructions of the manufacturer . To verify production of the relevant proteins , initial studies were done in small-scale expression conditions , followed by Western blot with anti-His-tag antibody ( GE ) . Fermentation conditions for selected clones were carried out in BMGY media ( 15 mL ) at 28–30°C in a shaking incubator until cultures reached an OD600 = 2 . 0 ( approximately 16–18 h ) , as per manufacturer's recommendations . Induction was performed by addition of methanol to a final concentration of 0 . 5% every 24 h; expression was monitored at 48 and 96 h time points . The supernatants and cell pellets for 10 colonies of each SmCD59 were analyzed for protein expression by Western blot . The colonies that presented the highest expression level were selected for scale-up fermentation . For protein expression and purification , selected clones were scaled-up for growth in 300 mL in 2 . 0 L baffled flasks under the same conditions . Cells were harvested after 96 h by centrifugation . The culture medium containing the secreted proteins were filtered through a 0 . 22 µm membrane , and diluted with 3 volumes of equilibration buffer ( 20 mM sodium phosphate , 300 mM NaCl , 10 mM imidazole , pH 7 . 4 ( for rSmCD59 . 1 ) and pH 8 . 0 ( for rSmCD59 . 2 ) . The recombinant proteins were then purified by metal affinity chromatography using the ÄKTAprime system ( GE Healthcare ) under native conditions . Briefly , the sample was loaded onto a Ni2+-NTA column ( 5 mL bed volume ) pre-equilibrated with the same buffer . The column was washed with 20 bed volumes of the equilibration buffer and then eluted with a 20–500 mM imidazole linear gradient . Fractions encompassing the main peak were characterized by sodium dodecyl sulfate polyacrylamide gel electrophoresis ( SDS-PAGE ) . Eluted fractions containing the recombinant protein in near pure form were pooled and submitted to extensive dialysis in Phosphate Buffer Saline pH 7 . 4 ( PBS ) . This sample was analyzed by SDS-PAGE and stained with Schiff's reagent ( Sigma ) for detection of glycoproteins as per the manufacturer's recommendations . Bovine Serum Albumin ( BSA ) ( Bio-Rad ) ( non-glycosylated protein ) and rSmVAL4 ( glycosylated protein ) were used as controls for the specificity of the reaction [45] . The rSmCD59 . 1 and rSmCD59 . 2 proteins expressed in P . pastoris were glycosylated ( products between 14 . 4 kDa and 20 . 1 kDa ) , based on their staining with Schiff's reagent and increased size in SDS-PAGE ( as can be observed by the protein band shift in comparison to the rSmCD59 . 2 expressed in E . coli ( Figure S1 ) . Both recombinant proteins were used in the hemolytic assay and to generate polyclonal antibodies in rats . The pAE-SmCD59 . 2 was transformed into E . coli BL21 ( SI ) ( Life Technologies ) and the transformed cells were grown in 300 mL LB ON plus ampicillin ( 100 µg/mL ) until they reached an OD600 = 0 . 7 , after which induction was performed by addition of 300 mM sodium chloride ( NaCl ) for another 4 h at 30°C . Harvested cells resuspended in 30 mL of lysis buffer ( 20 mM Tris pH 8 . 8 , 150 mM NaCl ) were lysed in a French Press . The pelleted inclusion bodies obtained by centrifugation at 20 , 000× g for 30 min were washed twice with wash buffer ( lysis buffer , 2% Triton X-100 , 2 M urea ) , and finally resuspended in solubilization buffer ( lysis buffer , 10 mM imidazole , 8 M urea ) . The recombinant protein was refolded from the inclusion bodies by diluting 200-fold into equilibration buffer ( solubilization buffer without urea ) . The recombinant protein was then purified by metal-affinity chromatography using the ÄKTAprime system under native conditions . Briefly , the sample was loaded onto a Ni2+-NTA column pre-equilibrated with equilibration buffer . The column was washed with 10 bed volumes of the equilibration buffer and then eluted with 10–500 mM imidazole linear gradient . The main peak was pooled and the protein purity of fractions was assessed by SDS-PAGE . Before its use the protein was dialyzed against PBS , pH 7 . 4 . This sample was used in the hemolytic assay and to generate polyclonal antibodies in rats . Polyclonal rat antiserum was produced against the preparations of rSmCD59 . 1 ( P . pastoris ) and rSmCD59 . 2 ( E . coli ) . Rodents were inoculated three times subcutaneously , at 15-day intervals with 100 µg of protein mixed with TiterMax adjuvant ( CytRx Corporation; first dose ) or PBS ( in subsequent doses ) . Fifteen days after the last inoculation , rodents were exsanguinated . Total protein extracts from eggs , miracidia , cercariae , 7 day-old schistosomula and adult worms of S . mansoni were prepared as previously described [41] . The tegument extract was obtained using a freeze/thaw/vortex procedure [46] . Tegument surface membranes ( Tsm ) and tegument-extract without-surface membranes ( Twm ) were obtained after a low speed centrifugation ( 100× g , 30 min ) ( adapted from [46] ) . Additionally , soluble ( Sol ) and insoluble ( Ins ) fractions of stripped worms after tegument removal were prepared as previously described [47] . The protein extract concentrations were determined with a RC DC Protein Assay Kit ( Bio-Rad , CA , USA ) . Purified rSmCD59 . 1 or rSmCD59 . 2 ( 100 ng ) and different parasite extracts ( 20 µg ) were subjected to SDS-PAGE . The gel was electroblotted onto PVDF membrane , which was blocked with 0 . 02 M Tris ( pH 7 . 5 ) and 0 . 3% Tween 20 containing 5% dry milk for 16 h at 4°C . The membranes were incubated in 1∶2 , 000 or 1∶5 , 000 dilution of anti-rSmCD59 . 1 and anti-rCD59 . 2 primary antibody , respectively in blocking buffer plus 150 mM NaCl for 3 h at room temperature . After three washes using 150 mL of 10 mM Tris ( pH 7 . 5 ) , the membranes were incubated in a 1∶5 , 000 dilution with secondary goat anti-rat IgG conjugated to horseradish peroxidase ( Sigma ) for 1 h , followed by another three washes using the same buffer . Antibody reactivity was developed with ECL reagent ( GE Healthcare ) according to the manufacturer's instructions and imaged using Hyperfilm or Image Quant LAS ( GE Healthcare ) . Immunocytochemistry on whole adult worms followed a previously described protocol [48] . Briefly , adult worms were fixed in 4% paraformaldehyde for 4 h , washed in PBS ( 0 . 1 M , pH 7 . 4 ) for 1 h and then transferred to a fresh fixative for another 3 h . After permeabilization with 1% Triton X-100 , 0 . 1% SDS , 10% ( heat-inactivated ) rabbit serum , 0 . 1% NaN3 in PBS overnight at 4°C , the worms were incubated with primary antibody , diluted 1∶200 , for 96 h at 4°C . After extensive washes , the worms were incubated for 48 h with 100 ng/mL of Phalloidin-rhodamine ( Molecular Probes , Life Technologies ) , to stain the musculature of the parasite , and with Alexa Fluor 488-labeled rabbit-anti-rat antibody ( 1∶200 , Molecular Probes , USA ) in PBS containing 0 . 1% Triton X-100 , 1% BSA , 0 . 1% NaN3 and 10% rabbit serum at 4°C . After several rinses , the worms were visualized with a LSM 510 Meta confocal microscope ( Zeiss ) , attached to a Zeiss Axiovert 100 microscope . For cryosection analysis , perfused adult worms were embedded in OCT medium ( Tissue-Tek , Sakura ) in a pre-cooled beaker of isopentene , frozen in liquid N2 . Eight-micrometer cryostat adult worm sections were obtained and adhered to silanized glass slides ( DakoCytomation , USA ) and fixed in acetone for 30 min at −20°C before blocking with 1× PBS , 10% Naive rabbit serum and 0 . 1% Tween 20 ( PNT ) overnight at 4°C . They were then incubated with anti-rSmCD59 . 1 and anti-rSmCD59 . 2 anti-serum diluted 1∶100 in PNT for 4 h at room temperature . After five washes with PBS 0 . 1% Tween 20 , pH 7 . 4 ( PBS-T ) , an Alexa Fluor 488 conjugated anti-rat IgG ( 1∶200 ) ( Life Technologies ) and 20 µM DAPI ( 4′ , 6-diamidino-2-phenylindole dihydrochloride , Molecular Probes ) to visualize nuclei were added to the PNT solution and samples were incubated for 1 h at room temperature . Sections were washed five times , and then mounted in Fluorescent Mounting Medium ( DakoCytomation ) . Rat pre-immune sera were used as negative control . Images were acquired as described above . Briefly , to evaluate Alternative Pathway activity , washed rabbit erythrocytes ( ERs ) were diluted in AP-CFTD ( 144 mM NaCl , 0 . 96 mM sodium barbital , 2 . 48 mM barbituric acid , 1 . 4 mM MgCl2 , 10 mM EGTA ) and then added to Normal Human Serum ( NHS ) ( serial dilutions 1∶4 to 1∶256 ) . After incubation at 37°C for 30 min , VBS−EDTA ( 144 mM NaCl , 0 . 96 mM sodium barbital , 2 . 48 mM barbituric acid , 20 mM EDTA ) was added to stop lysis . After centrifugation ( 0 . 8× g for 10 min at 4°C ) the supernatant absorbance was measured at 405 nm and percentage hemolysis was calculated using ERs lysed by water as the 100% reference . The serum volume that produced 50% lysis of ERs was determined and used in the inhibition assays [49] , [50] . To evaluate the hemolytic activity mediated by the Classical Pathway , washed antibody-sensitized sheep erythrocytes ( EAs ) were diluted in VBS++ ( 144 mM NaCl , 0 . 96 mM sodium barbital , 2 . 48 mM barbituric acid , 0 . 83 mM MgCl2 , 0 . 25 mM CaCl2 ) and were added to NHS ( serial dilutions 1∶20 to 1∶500 ) . After incubation at 37°C for 30 min , samples were centrifuged ( 0 . 8× g for 10 min at 4°C ) and the supernatant absorbance at 405 nm was measured . The hemolysis percentage ( relative to EAs suspension completely lysed by water ) was calculated . The volume of serum necessary to promote 50% lysis of EAs was determined and used in the inhibition assays [50] . Inhibition assays were performed to evaluate if the proteins rSmCD59 . 1 and rSmCD59 . 2 ( produced in P . pastoris and in E . coli ) were able to protect ERs from the lysis triggered by the Alternative Pathway , or EAs from the lysis by the Classical Complement Pathway . BSA ( Sigma-Aldrich ) was used as a negative control . Different amounts of the proteins ( 0 , 2 , 5 and 10 µg for the Alternative Pathway , and 0 , 1 , 2 and 4 µg for the Classical Pathway ) were pre-incubated with NHS ( corresponding to 50% lysis ) for 20 min at 37°C . Washed ERs ( 3 . 5×106 ) or EAs ( 6 . 7×104 ) were incubated with treated NHS for 30 min at 37°C . After centrifugation ( 0 . 8× g for 10 min at 4°C ) , the supernatant absorbance at 405 nm was measured and % hemolysis was calculated . CHO cells were obtained from American Type Culture Collection ( Manassas , VA ) and cultured as monolayers in 100 mm cell culture dishes using DMEM/F12 medium ( Life Technologies ) supplemented with 10% fetal calf serum ( FCS ) , 2 mM L-glutamine , 100 units/mL penicillin and 50 µg/mL streptomycin sulfate . Cells were maintained at 37°C and 5% CO2 and reseeded twice a week using 0 . 05% trypsin . The full size SmCD59 . 1 and SmCD59 . 2 coding regions , including the putative domains for the signal peptide and GPI anchor , respectively , were codon optimized for human and hamster codon preferences , synthesized ( Genscript USA Inc . , Piscataway , NJ ) ( Table S1 ) and cloned into pcDNA-3 . 1 ( + ) via BamHI/XhoI restriction sites . A full length hCD59 cDNA obtained from Open Biosystems/Thermo Scientific , AL was sub-cloned into pcDNA to serve as a positive control for protein expression and complement inhibitory studies in CHO cells . The recombinant plasmids were transiently introduced into 70–80% confluent CHO cells cultured in 6 well-plates using the polymer-based DNA transfection agent jetPEI ( Polyplus , France ) according to manufacturer's instructions . Transfected cells were cultured for an additional 48 h before harvesting . CHO cells transfected with empty vector were included as negative control . CHO cells transiently transfected with SmCD59 . 1 and SmCD59 . 2 were tested for membrane protein expression . Cells expressing hCD59 , or containing empty pcDNA , were included as positive and negative controls , respectively . For fluorescent microscopy and flow cytometry analysis , live cells were washed twice with PBS without CaCl2 and MgCl2 and detached from the plate with a non-enzymatic cell dissociation solution ( Sigma ) . Cells were washed once with DMEM/F12 containing 0 . 2% BSA ( DMEM-BSA ) by centrifugation at 100× g for 5 min and resuspended to 106 cells/mL in the same medium . Cells were incubated for 30 min at room temperature with the respective primary antibodies , i . e . CHO-SmCD59 . 1 with rat polyclonal serum anti-rSmCD59 . 1 at 1∶500 dilution; CHO-SmCD59 . 2 with an anti-rSmCD59 . 2 at 1∶500 and CHO-hCD59 with a rat monoclonal antibody anti-hCD59 at 1∶2 , 000 . After 2 washes with DMEM-BSA , cells were incubated with a secondary FITC-labeled goat anti-rat IgG ( Invitrogen , Life Technologies ) . Cells were immediately visualized under an inverted fluorescent microscope and positive cells were quantified by flow cytometry after subtracting fluorescent background from cells transfected with empty vector . For analysis by Western blot , detached cells were washed twice with PBS and membrane extract was prepared using the ProteoExtract Native Membrane Protein Extraction Kit ( EMD Millipore , Billerica , MA ) , following manufacturer's instructions . The membrane fractions were incubated with each corresponding primary antibody followed by HRP-labeled goat anti-rat IgG ( Invitrogen ) and developed using the ECL Western Blot Developing System . To confirm that SmCD59 is GPI-anchored in transfected CHO cells , we followed previously described methodology for hCD59 [51] . Briefly , transfected cells were removed from culture plate wells , washed in Hank's Balanced Salt Solution ( HBSS ) and exposed to 1 Unit/mL of PiPL-C ( Sigma ) in HBSS or kept in buffer only as control . After incubation for 1 h at 37°C , cells were stained with the antibody anti-rSmCD59 . 1 and the percentage of fluorescent cells was quantified and compared with the control sample not treated with PiPL-C using flow cytometry . CHO cells expressing SmCD59 . 1 and SmCD59 . 2 were tested in a complement mediated cell damage assay to determine whether these proteins have complement regulatory activity . This approach was performed according to previously reported methodologies described for hCD59 [51] , [52] , with some modifications . Briefly , CHO cells transiently transfected with SmCD59 . 1 , SmCD59 . 2 , hCD59 and empty vector were detached from 6-well culture plates as described in the previous section , washed and incubated with 5% rabbit anti-CHO membrane serum in DMEM-BSA for 30 min at room temperature , washed twice with DMEM-BSA and incubated with 10% NHS ( Complement Technology , USA ) in DMEM-BSA as source of human complement factors . Following 1 h sample incubation at 37°C , cell viability was determined by adding the vital dye propidium iodide at 5 µg/mL and the population of live cells ( not stained by propidium iodide ) were identified by flow cytometry . Samples with NHS heated at 56°C for 30 min to inactivate complement were used to measure background cell mortality . Gene-specific siRNAs were commercially synthesized ( Integrated DNA Technologies , Inc . , IA ) and used to induce gene expression knockdown of SmCD59 . 1 and SmCD59 . 2 , respectively , by electroporation as described [53] . The DNA sequence for SmCD59 . 1 siRNA is 5′-CTACAAGTGACTAGTCGTAGTTGTG-3′ , spanning coding DNA positions 193–217 . SmCD59 . 2 siRNA sequence is 5′-GGTAAAGCTGGCTTAGTAACTGAAT-3′ spanning coding DNA positions 241–265 . The negative control siRNA was acquired from IDT and has been previously tested by our own group [53] . Briefly , freshly transformed schistosomula were placed into electroporation buffer ( Bio Rad , CA ) containing a mixture of SmCD59 . 1 and SmCD59 . 2 siRNAs at 5 . 6 µM each or control siRNA ( 11 . 2 µM ) or no siRNA . Parasites were immediately electroporated and then cultured for 5 days in DMEM/F12 medium containing 10% FCS at 37°C , in an atmosphere of 5% CO2 . Reduction in target transcript levels was measured by qRT-PCR relative to transcript levels in parasites treated with control irrelevant siRNA ( 100% gene expression ) . Decrease in SmCD59 . 1 and SmCD59 . 2 protein levels was measured in whole parasite lysates by Western blots . Schistosomula lysates were prepared by adding 80 µL of RIPA buffer containing a cocktail of protease inhibitors followed by incubation on ice for 1 h . The protein content in each extract was estimated using the BCA Protein Assay Kit ( Pierce , IL ) according to the manufacturer's instructions . Soluble protein ( 800 ng ) was subjected to SDS-PAGE , blotted onto PVDF membrane and blocked with PBS containing 0 . 1% Tween 20 and 5% milk for 1 h at room temperature . The membrane was then probed overnight at 4°C with anti-rSmCD59 . 1 or anti-rSmCD59 . 2 immune rat serum at 1∶500 dilution or rabbit antibody directed against the schistosome protein aquaporin ( SmAQP ) as loading control at 1∶250 dilution . Bound primary antibodies were detected by appropriate secondary antibodies conjugated to horseradish peroxidase and exposure to ECL substrate . The same membrane was probed three times to detect SmCD59 . 1 , SmCD59 . 2 and the control protein , SmAQP . For each re-use , the membrane was stripped with Western Blot Stripping Buffer ( Thermo Scientific , IL ) following manufacturer's instructions to remove bound antibody . SmCD59 . 1 and SmCD59 . 2 suppressed schistosomula were tested in a complement killing assay modified from Deng et al . [27] . Briefly , 200 control and siRNA-suppressed parasites cultured for 5 days were washed in DMEM/F12 and incubated for 30 min at 37°C with serum ( 1∶4 dilution ) from rats infected twice with cercariae ( IRS ) or with naive rat serum ( NRS ) , both heat-inactivated . Parasites were washed with medium and incubated overnight at 37°C with undiluted NHS as a source of complement . Suppressed and control schistosomula were also tested for complement killing in the absence of rat serum . After 18–24 h , schistosomula were exposed to 1 µg/mL of DNA binding stain Hoechst 33258 dye for 5 min and examined with an inverted fluorescent microscope to count the number of dead fluorescent parasites under ultraviolet light ( 352 excitation/455 emission ) [54] . Dye uptake only occurs if there is complement induced-tegumental damage . Samples were run in duplicates in three independent experiments . Percent net mortality was calculated by subtracting background mortality of parasites treated with heat-inactivated NHS . The same complement killing assay protocol was used to measure mortality of freshly transformed schistosomula cultured for 3 h in DMEM/F12 plus 10% FCS .
BlastP comparisons of SmCD59 . 1 and SmCD59 . 2 sequences to GenBank revealed low similarity with orthologs of CD59 proteins belonging to the uPAR/Ly6/CD59/snake toxin-receptor superfamily ( E-value , ranging from 3×10−10–4×10−4 ) . Because this is the first article to deal specifically with this schistosome gene family , herein we propose numbering of these genes as shown on Table 1 . In addition to the six previously reported S . mansoni sequences , SmCD59 . 1-6 , it was possible to identify a new S . mansoni member ( SmCD59 . 7 – Sm_125250 ) ( Table 1 ) , which displays a slightly different distribution of Cys 10 ( Figure 1 ) . Furthermore , nine sequences from S . japonicum and six from S . hematobium could be identified , and all branched together with the S . mansoni members representing possible orthologs ( Figure S2 ) . In spite of the low bootstrap value , all these sequences did not branch together with mammalian Ly6 or CD59 sequences . Additionally , searching other databases , 13 non-characterized proteins from other Platyhelminthes species were identified ( E-value , ranging from 9×10−19–2×10−7 ) . Noteworthy , was the identification of a member in the turbellarian S . mediterranea ( Figure S2 ) . Analyses of the primary sequences revealed the presence of ten conserved cysteine residues following the same pattern of distribution in Platyhelminthes ( Figure 1 ) . A very intrinsic characteristic of the TFPD fold is the presence of at least four conserved cystines among all the different classes of proteins having this domain [30] . Nonetheless , this pattern differs from the CD59s of mammals in the 3rd and 4th cysteines , which are separated by a larger stretch of amino acids in the Platyhelminthes sequences and are located at a distance of exactly two amino acids from their closest cysteines ( 2nd and 5th Cys ) . These distances ( from the 2nd to the 3rd and from the 4th to the 5th Cys ) in mammals are more variable . However , the most important difference of these molecules when compared to human CD59 , is that they do not have the conserved amino acid residues ( D24 , W40 , R53 and E56 in hCD59 ) involved in complement recognition [30] , [55] , as well as the “hydrophobic groove” ( C39 , W40 and L54 in hCD59 ) , highly conserved in all CD59s [55] . The sequence belonging to Herpesvirus Saimiri Protein ( Sah-CD59 ) was included as a homolog from human CD59 . Similarly , a glycine adjacent to the sixth cysteine ( characteristic of toxins ) is not present in the sequences [33] ( Figure 1 ) . In order to explicitly visualize the putative domain features and also provide insight into its folding , 3D-modeling of SmCD59 . 2 ( Smp_105220 ) was performed using human CD59 as a template ( 2UWR ) ( Figure 2 ) . Despite the low similarity between the two sequences ( ∼37% ) , it was possible to align them in order to perform the modeling . Out of the 124 residues , 73 of them were modeled ( beginning at V25 , which becomes V1 in the mature protein , and ending at A97 , becoming A73 ) . The last 27 residues at the C-terminal remained out of the model , since they did not show any similarity to 2UWR ( Figure 1 ) . RMSD was of 1 . 24 Å between template and models in the common core ( 36 alpha-carbon atoms ) . In this model , the finger-shaped backbone is present within the structure , emerging from a hydrophobic palm , clearly seen in Figure 2A . Additionally , the four conserved cystines can be observed ( Figure 2B ) and would be formed by the following pairs: Cys3-Cys27 , Cys24-Cys45 , Cys49-Cys65 and Cys66-Cys71 . There are another two conserved cysteines in the Schistosoma spp . ( Cys6 and Cys9 in SmCD59 . 2 ) that may not form a disulfide bridge . However , the possibility of a fifth cystine formed by these two cysteines cannot be ruled out , since the model generated is not necessarily a real portrait of the native molecule's fold . All these spatial conformations are possible since the TFPD signature is present in the sequence ( Figure 1 ) . The geometry of the final refined model was evaluated with a Ramachandran plot , which showed that 97% of the amino acid residues were positioned in the “allowed” regions ( data not shown ) . In order to establish the level of expression in different parasite stages , including cercariae , in vitro cultured 7-day old schistosomula , adult worms , eggs and miracidia , qRT-PCR was performed using SYBR Green ( Life Technologies ) . Gene expression fold changes of each SmCD59 were calculated relative to the less expressed stage after normalization to the alpha-tubulin housekeeping gene . The results show that the SmCD59 genes , with the exception of SmCD59 . 6 , display increased gene expression in the schistosomulum and adult worm stages ( Figure S3 ) . To evaluate the relative expression levels between the different SmCD59 isoforms , qRT-PCR was performed using the Taqman system on three stages , the free living infective stage ( cercaria ) , in vitro cultured 11-day old schistosomula and adult worms . SmCD59 . 1 , SmCD59 . 3 and SmCD59 . 4 showed the highest levels of expression with higher expression in the schistosomula and adult stages , while SmCD59 . 2 , SmCD59 . 5 and SmCD59 . 7 showed intermediate levels of expression and SmCD59 . 6 was found at very low levels ( Figure 3 ) . Samples prepared from cercariae , schistosomula , adult male and female worms , eggs and miracidia stages from S . mansoni , and tegument isolated by the freeze/thaw method , were all separated by 15% SDS-PAGE . Immunoblotting was performed using mouse anti-rSmCD59 . 1 and anti-rSmCD59 . 2 antisera . The protein expression profile of SmCD59 . 1 generally correlated with the Real Time RT-PCR data , revealing low expression in miracidia and cercariae , increasing in schistosomula and adult worms ( Figure 4A ) . A lower intensity band of higher molecular mass may be attributed to some cross reactivity with other SmCD59 members ( possibly glycosylated ) ( Figure 4A ) , as suggested by the in silico predicted N-glycosylation sites ( Figure 1 ) . In the case of SmCD59 . 2 , a similar expression profile with higher levels in schistosomula was observed ( Figure 4B ) . Notably , after the separation of the tegument from the worm body , the SmCD59 . 1 protein was detected only in the insoluble fraction of stripped worms ( Figure 4C ) . We cannot rule out the presence of this protein in the tegument fraction; however , this data suggested that the protein is much more abundant in denuded worms than in the tegument . The SmCD59 . 2 protein was found to be more abundant in stripped worms , with similar proportions in the soluble and insoluble fractions ( Figure 4D ) . Furthermore , analysis of the tegument surface membrane fraction revealed that the small amount of protein present in the tegument is membrane associated , with no protein detected in the soluble supernatant ( soluble syncytial proteins ) ( Figure 4D ) . Immunolocalization studies on whole adult worms using rat serum raised against rSmCD59 . 1 and rSmCD59 . 2 revealed that both proteins were expressed on the surface of S . mansoni adult male worms , as shown by the green staining apparent around and over their dorsal tubercules ( Figure 5A and B ) . However , analysis of adult worm sections revealed that SmCD59 . 1 protein is also expressed at significant levels in the parenchyma cells of male and female adult worms ( Figure 5D ) , while SmCD59 . 2 protein shows even a broader localization , including muscle cells of male and female adult worms ( Figure 5E ) . In order to investigate the involvement of SmCD59 isoforms in the inhibition of the complement cascade , NHS was pre-incubated with various amounts of rSmCD59 . 1 and rSmCD59 . 2 proteins , and then added to rabbit erythrocytes ( Alternative Pathway ) or antibody-sensitized sheep erythrocytes ( Classical Pathway ) . Our data revealed that the rSmCD59 isoforms and the negative control BSA did not inhibit hemolysis triggered by the Alternative or Classical Complement Pathways ( Figure 6 ) . The NHS used in these assays promoted 50% erythrocyte lysis and addition of the recombinant schistosome proteins or BSA did not alter this percentage . These results therefore suggest that the rSmCD59 isoforms are not involved in the inhibition of complement deposition by the Classical or Alternative Pathways . It has been shown that the membrane form of hCD59 ( GPI-anchored ) is about 5 to 10 times more active at inhibiting the complement system as compared to the soluble form lacking the GPI anchor [56] , [57] . Thus , CHO cells were transfected with the codon optimized full-length cDNA of SmCD59 . 1 ( CHO-SmCD59 . 1 ) and SmCD59 . 2 ( CHO-SmCD59 . 2 ) to measure complement resistance of cells expressing these membrane-anchored proteins . Cells transfected with hCD59 ( CHO-hCD59 ) and empty vector ( CHO-pcDNA ) were included as positive and negative controls , respectively . Expression of SmCD59 . 1 in the CHO cell surface was demonstrated by positive staining of the membrane of live cells with an anti-rSmCD59 . 1 antibody . Flow cytometer analyses confirmed expression of SmCD59 . 1 in about 20% of the transfected cells ( Figure S4A ) . Similar results were obtained with hCD59 ( Figure S4B ) . Expression of SmCD59 . 2 in the cell membranes was detected by Western blot of a membrane preparation of transfected cells ( anti-rSmCD59 . 2 rat serum did not react with endogenous CD59 protein in live CHO cells – data not shown ) . To confirm that SmCD59 . 1 is GPI-anchored , transfected CHO cells were treated with PiPL-C , labeled with anti- rSmCD59 . 1 antibody and analyzed by flow cytometry . As shown in Figure S4C ( top panels ) , there was a pronounced reduction in antibody labeling anti-rSmCD59 . 1 after treatment with PiPL-C ( 6 . 8% ) as compared to untreated cells ( 25 . 8% ) . Results were similar for CHO-hCD59 cells included as positive control ( bottom panels ) . CHO-SmCD59 . 1 and CHO-SmCD59 . 2 were tested for complement resistance when exposed to antibodies reactive to CHO membrane proteins and to complement from NHS . Background cell mortality was monitored by including samples treated with heat-inactivated complement ( iNHS ) . Complement-treated cells were analyzed by flow cytometry to quantify the live cell population , i . e . cells of reduced fluorescence that are not stained by propidium iodide ( Figure 7 ) . The number of viable cells in the CHO-hCD59 sample was considerably higher than the number of live cells in the CHO-pcDNA sample after addition of NHS ( Figure 7A , first peak in left panel ) . T-test analysis of three independent experiments ( Figure 7D ) showed that CHO-hCD59 cells were significantly more resistant to complement ( 73 . 4%±9 . 1 ) than CHO-pcDNA cells ( 37 . 8%±9% ) ( p<0 . 05 ) . When cells were treated with iNHS , the number of live cells in both samples was nearly identical and close to 95% ( Figure 7A , first peak in right panel and Figure 7D ) , confirming the complement inhibitory activity of hCD59 in the transfected cells . The same cell survival analysis was performed on CHO-SmCD59 . 1 and CHO-SmCD59 . 2 samples , but there was no difference in cell viability in both samples compared to CHO-pcDNA sample after exposure to NHS ( Figure 7B and 7C respectively ) . These results strongly indicate that SmCD59 . 1 and SmCD59 . 2 are not complement inhibitory proteins . To further assess whether SmCD59 . 1 and SmCD59 . 2 protect schistosomes from the complement attack , freshly transformed schistosomula were treated with a mixture of SmCD59 . 1 and SmCD59 . 2 specific siRNAs and tested for complement susceptibility 5 days later . Transcript levels measured by qRT-PCR were about 60% lower in SmCD59-suppressed parasites than in control parasites ( Figure 8A ) . Schistosomula were suppressed immediately after cercarial transformation because SmCD59 mRNA levels are undetectable by qRT-PCR at that early time point after which expression increases significantly ( data not shown ) . This approach was expected to ensure that SmCD59 gene knockdown would substantially inhibit protein production in siRNA-treated parasites , while control parasites would have abundant protein levels . Indeed , Western blot analysis in Figure 8B confirmed dramatic reduction of SmCD59 . 1 ( top panel ) and SmCD59 . 2 ( middle panel ) proteins in siRNA-suppressed parasites compared to lysates of control parasites . In the bottom panel , the SmAQP control protein was detected in all lysates , demonstrating that comparable levels of protein were present in each lane . The significant decrease of protein levels in siRNA-suppressed schistosomula did not enhance in vitro parasite killing by human complement beyond 20% , even in the presence of immune serum when compared to the control groups in three independent experiments ( Figure 8C ) . Finally , data in figure 8D was obtained to validate the complement assay and to confirm previous results [3] , [4] , [7] in which , contrary to older schistosomula , 3 h-cultured parasites are more susceptible to complement killing , particularly when the killing is antibody mediated . Thus , our results indicate that SmCD59 . 1 and SmCD59 . 2 are not complement regulatory proteins in schistosomes .
Early analysis of the S . mansoni sequences , SmCD59 . 1 and SmCD59 . 2 , revealed similarity to human CD59 [13] , [28] , [29] . In the present report , we have described members of the TFPD family in Platyhelminthes and compared their structure and folding to the human CD59 protein [58] . This group of proteins contains the TFPD fold , since the characteristic cysteines of this family are present in their sequences . Despite the fact that the cysteines are conserved , there are two important differences that are evident in the alignment: i ) the 3rd and 4th cysteines are in different positions when aligned with their mammalian counterparts; nevertheless , all the Platyhelminthes sequences maintain the two-amino acid distance pattern between the 2nd/3rd and 4th/5th cysteines; ii ) the Platyhelminthes sequences lack the essential amino acids in the active site required to block the formation of the MAC . However , it is reasonable to expect that not all these features would be totally conserved in trematodes and mammals , since these proteins may have hypothetically evolved as complement inhibitors through convergent evolution . For example , there is a fish homologue of CD59 , which was shown to have fish complement inhibitory activity , but does not have many of the so-called ‘critical’ residues conserved [59] . Furthermore , the presence of at least one CD59 member in the non-parasitic flatworm , S . mediterranea , suggests that these proteins are also participating in aspects of non-parasitic Platyhelminth biology . Moreover , it is at least intriguing how this pattern remained conserved in inferior taxa , while in mammals it seems to be more variable . The transcription of most SmCD59 genes is up-regulated following the transition from free-living cercaria to the parasitic schistosomula stage; some of these genes continue to show increasing transcription into the adult worm stage . These data confirm previous microarray results showing up-regulation of SmCD59 . 1-5 expression in 3-day old schistosomula in comparison to germ-balls and cercariae [60] . Since SmCD59 . 1 and SmCD59 . 2 are the only representatives of this family confirmed to localize at the host-parasite interface by PiPL-C treatment , they were chosen for additional characterization , including functional studies . Western blot analysis of SmCD59 . 1 and SmCD59 . 2 at different life cycle stages confirms that protein expression of these genes also increases during the transition to the mammalian stage . This coincides with the transition of the parasite from a complement-sensitive state to one of resistance [5] . Confocal microscopy immunolocalization studies using anti-rSmCD59 . 1 and anti-rSmCD59 . 2 antibodies confirmed previous data obtained by PiPL-C shaving showing that the proteins were surface-exposed on the tegument . On the other hand , it was clear that the proteins were also present in considerable amounts inside the parasite , which is in accord with the Western blot experiments on tegument extracts and stripped worms . On the whole , these results further support the concept that these proteins are probably involved in aspects of non-parasitic schistosome biology . Due to the sequence similarity , we also investigated whether polyclonal antibodies directed to human CD59 ( Abnova ) could recognize SmCD59 . 1 and SmCD59 . 2 and whether rat anti-rSmC59 . 1 and anti-rSmCD59 . 2 polyclonal antibodies recognize human CD59 ( Abnova ) . However , no cross reactivity could be observed in any of these studies ( data not shown ) . To demonstrate whether SmCD59 . 1 and SmCD59 . 2 are functional homologues of human CD59 , soluble rSmCD59 proteins produced in P . pastoris and E . coli were used to examine their complement-inhibitory activity in an in vitro hemolytic assay designed to favor the activation of the Alternative ( rabbit erythrocytes ) or the Classical ( antibody-sensitized sheep erythrocytes ) Pathways , respectively . Soluble forms of CD59 lacking the GPI-anchor domain have been expressed in mammalian cells , insect cells and yeast and shown to have MAC-inhibitory activity in vitro [57] , [61] , [62] , indicating that glycosylation by Pichia pastoris does not inhibit the protein function . However , in our study no inhibition of erythrocyte lysis was observed when NHS was pre-incubated with rSmCD59 proteins . Membrane-targeted forms of CD59 have been shown to be more potent in inhibiting complement than the soluble forms [54] , [59] . Thus , as an alternative approach , SmCD59 . 1 and 2 were expressed as membrane proteins ( complete coding region ) in CHO cells followed by treatment of transfected cells with human complement . Membrane protein expression in CHO cells may also contribute to produce rSmCD59 proteins as close to the native state as possible , which is essential for functional studies . In addition , eukaryotic cells transfected with complement regulatory proteins represent a more physiologically relevant target for in vitro complement experiments as compared to erythrocytes pre-treated with these proteins [63] . Despite our results showing proper localization of rSmCD59 . 1 and rSmCD59 . 2 at the plasma membrane and proper GPI-anchorage of rSmCD59 . 1 , transfected CHO cells were not resistant to killing by human complement . Taken together , these functional analyses suggest that SmCD59 . 1 and SmCD59 . 2 do not possess inherent complement inhibitory activities . In a third attempt to investigate the potential of SmCD59 . 1 and 2 to protect schistosomes from complement attack , the parasites were treated with target-specific siRNAs to induce gene expression knockdown for both targets simultaneously . Despite successful suppression of SmCD59 . 1 and 2 transcription and translation in schistosomula , these parasites did not become more susceptible to complement killing , by either the Alternative or the Classical Pathways as compared to control parasites . We believe that these results , together with the other complement assays and the sequence and structural comparison , strongly support the conclusion that SmCD59 . 1 and 2 do not have a complement regulatory function . These schistosome CD59-like proteins probably have another function in the parasite that is unrelated to complement evasion , although it is quite likely that they are involved in some kind of molecular interaction . This hypothesis is acceptable since TFPDs commonly bind molecules , either as ligands ( e . g . toxins ) or membrane-attached receptors , like CD59 or urokinase/plasminogen activator receptor , uPAR [33] . Gathering all the evidence , we conclude that these CD59-like proteins do not have a complement regulatory role in schistosomes . Thus , it would be more appropriate to rename this class of proteins . However , since several papers have been published on this gene family as SmCD59 and the protein function is still unknown , it is reasonable to change the family name only when their function is elucidated . Further studies should focus on resolving the 3-D structure of the proteins and deriving their function within the biological context of the host/parasite relationship . | Schistosomes are parasites that reside for many years in the blood stream , demanding efficient mechanisms of evading immune response effectors such as complement deposition . A group of genes similar to human CD59 , an important complement inhibitor in mammals , were identified in the schistosome genome . Computer predictions of protein structure indicated substantial similarity of the schistosome proteins and the mammalian CD59 family of proteins , which due to their three-finger-shaped spatial conformation are members of the Three-Finger Protein Domain fold superfamily ( TFPD ) . Members of this family of schistosome proteins were also shown to be expressed predominantly during the mammalian stages when worms are exposed to complement and found to be present at the host-interactive surface of schistosomes . Three different methods were employed to test the possible involvement of these proteins in complement inhibition . Our results strongly suggest that these proteins are not involved in the inhibition of complement and that further studies are needed to establish their functional role ( s ) in schistosomes . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | On the Three-Finger Protein Domain Fold and CD59-Like Proteins in Schistosoma mansoni |
The proapoptotic PB1-F2 protein of influenza A viruses has been shown to contribute to pathogenesis in the mouse model . Expression of full-length PB1-F2 increases the pathogenesis of the influenza A virus , causing weight loss , slower viral clearance , and increased viral titers in the lungs . After comparing viruses from the Hong Kong 1997 H5N1 outbreak , one amino acid change ( N66S ) was found in the PB1-F2 sequence at position 66 that correlated with pathogenicity . This same amino acid change ( N66S ) was also found in the PB1-F2 protein of the 1918 pandemic A/Brevig Mission/18 virus . Two isogenic recombinant chimeric viruses were created with an influenza A/WSN/33 virus background containing the PB1 segment from the HK/156/97: WH and WH N66S . In mice infected with WH N66S virus there was increased pathogenicity as measured by weight loss and decreased survival , and a 100-fold increase in virus replication when compared to mice infected with the WH virus . The 1918 pandemic strain A/Brevig Mission/18 was reconstructed with a pathogenicity-reducing mutation in PB1-F2 ( S66N ) . The resultant 1918 S66N virus was attenuated in mice having a 3-log lower 50% lethal dose and caused less morbidity and mortality in mice than the wild-type virus . Viral lung titers were also decreased in 1918 S66N–infected mice compared with wild-type 1918 virus–infected mice . In addition , both viruses with an S at position 66 ( WH N66S and wt 1918 ) induced elevated levels of cytokines in the lungs of infected mice . Together , these data show that a single amino acid substitution in PB1-F2 can result in increased viral pathogenicity and could be one of the factors contributing to the high lethality seen with the 1918 pandemic virus .
Influenza A virus causes 300 , 000–500 , 000 deaths worldwide each year , and in pandemic years , this number can increase to 1 million ( in 1957–1958 ) or as high as 50 million , as was seen in 1918–1919 [1–3] . More recently , H5N1 highly pathogenic avian influenza viruses have generated great concern regarding their potential to cause a pandemic . H5N1 infections in humans were seen in Hong Kong in a small outbreak in 1997 that resulted in 18 human infections and six fatalities , and since 2003 , 309 human cases of H5N1 have been confirmed with a 61% fatality rate ( 6/1/07 ) [4–7] . Recent work on these viruses has aimed to elucidate the virulence factors that account for the severe illness observed in humans and mice [4 , 8–12] . The viral PB1 segment is of particular interest , since , in addition to the glycoprotein genes , the PB1 gene was the only other segment that was exchanged in the pandemic viruses of 1957 and 1968 [13] . Introduction of a novel PB1 gene into the 1998 swine reassortant viruses further implicates the role of this gene in the pathogenesis of ( animal ) influenza [14] . Moreover , while changes in the surface glycoproteins allow the viruses to overcome the preexisting humoral immune response , they may not be solely responsible for the high virulence of the pandemic influenza viruses . In particular , the 1918 pandemic was associated with significantly higher morbidity and mortality than the subsequent pandemics [15] . Recent reconstruction of the 1918 virus has confirmed that the viral polymerase from the 1918 influenza virus is required for full pathogenicity of the recombinant 1918 virus in mice [16] . In fact , substitution of the viral polymerase genomic segments with those of the modern H1N1 strain severely attenuated the virus in mice [16] . Recent identification and characterization of a novel influenza virus protein called PB1-F2 encoded by the PB1 gene introduced a potential virulence factor that could play a role in pathogenesis of infection with pandemic influenza viruses and explain the selection of the PB1 gene in these viruses [17] . The influenza virus PB1-F2 is a 90–amino acid ( aa ) protein that is associated with the induction of cell death . The protein directly permeabilizes mitochondria , resulting in the dissipation of the mitochondrial membrane potential and the release of cytochrome c [17–19] . We have previously shown that PB1-F2 contributes to viral pathogenesis in the mouse model and wanted to further investigate whether the PB1-F2 proteins encoded by highly pathogenic viruses have conserved mutations in their aa sequence that are associated with pathogenicity [20] . We chose to study the PB1-F2 proteins of the Hong Kong 1997 H5N1 viruses that caused an outbreak in humans . Characterization of the isolated viruses in mice revealed that the viruses could be subdivided into three different groups based on the pathogenicity phenotype: high-virulence , intermediate-virulence , and low-virulence [21] . Further studies provided molecular correlates of pathogenicity in the high-virulence group , though such studies were not conducted for the PB1-F2 protein [22] . Herein , we assess the contribution of the PB1-F2 protein to the pathogenicity of a highly pathogenic H5N1 virus and the 1918 pandemic strain virus . An alignment of the aa sequences of isolates from the Hong Kong 1997 H5N1 outbreak revealed that a mutation , N66S , was associated with high pathogenicity phenotype in mice . Using a recombinant A/WSN/33 virus with the PB1 segment of A/HK/156/97 , we observed increased morbidity and mortality of mice infected with a virus that contained the N66S mutation . In addition , infection with the reconstructed A/Brevig Mission/18 virus , which has an S at position 66 , resulted in increased pathogenicity when compared with a reconstructed A/Brevig Mission/18 virus in which position 66 was changed to N [16] . We thereby show that PB1-F2 proteins from highly virulent viruses can contribute to pathogenicity , and identify a single aa change that confers a virulent phenotype in mice .
Madin Darby canine kidney ( MDCK ) , 293T , and A549 cells were obtained from ATCC ( http://www . atcc . org/ ) and were maintained in MEM and DMEM culture media ( Gibco , http://www . invitrogen . com/ ) , respectively , supplemented with 10% fetal calf serum ( Hyclone , http://www . hyclone . com/ ) and penicillin/streptomycin ( Gibco ) . The pPolI vectors encoding viral genomic RNA of the WSN strain have been described previously [23] . The PB1 gene of the A/HK/156/97 virus was reverse transcribed from purified genomic RNA , amplified by PCR with PB1 segment-specific primers , and cloned into the pPolI vector . The cloning of genes for A/Brevig Mission/18 has been described previously [16] . To generate pPolI vectors encoding the N66S PB1-F2 mutants , the pPolI vectors encoding the A/HK/483/97 PB1 or A/Brevig Mission/18 PB1 were subjected to site-directed mutagenesis using the Stratagene Quick-Change mutagenesis kit ( Stratagene , http://www . stratagene . com/ ) . Sequences of each construct were confirmed by automated sequencing performed at the Mount Sinai sequencing core facility . The reverse genetics technique for the generation of recombinant influenza viruses has been described previously [23] . Briefly , 293T cells were transfected with eight pPolI vectors encoding the viral genomic RNA segments and four pCAGGS protein expression vectors encoding the subunits of viral polymerase and the nucleocapsid protein . The transfected 293T cells were cocultured with MDCK cells , and virus released into the supernatant was isolated by plaque purification on MDCK cells . The presence of the introduced mutations was confirmed by reverse transcription and sequencing of the PB1 genes of the newly generated viruses . Viruses possessing 1918 genes were generated under biosafety level 3 ( BSL-3 with enhancements ) containment [24] to ensure the safety of laboratory workers , the environment , and the public . All subsequent laboratory and animal work with live virus containing A/Brevig Mission/18 genes also was performed under these high-containment conditions . Female C57/BL6 mice 6 to 7 wk old ( Jackson Laboratories , http://www . jax . org/ ) were anesthetized with intraperitoneal injection of 0 . 07 ml of ketamine/xylazine ( 0 . 15 mg ketamine and 0 . 03 mg xylazine ) , and infectious virus was diluted in PBS/BSA/PS ( phosphate-buffered saline/bovine serum albumin/penicillin and streptomycin ) and inoculated intranasally in a volume of 30 μl . To assess virus pathogenicity , groups of four mice were inoculated with appropriate dose and were monitored daily for weight loss over 8 d . Mice that lost more than 25% of their initial body weight were killed according to institutional guidelines and scored as dead . To determine viral replication in the lungs , lungs were collected on days 1 , 2 , 3 , 5 , 7 , and 8 after infection from 2 ( days 1 and 2 ) or 4 ( days 3 , 5 , 7 , and 8 ) mice from each group and two mice in the PBS group . The lungs were homogenized in PBS using a Dounce homogenizer and processed for virus titering . Virus titers in the supernatant of lung homogenates were determined by plaque assay in MDCK cells . For 1918 recombinant virus infections , female BALB/c mice , 6 to 7 wk old ( Jackson Laboratories ) were anesthetized with an intraperitoneal injection of 0 . 2 ml of 2 , 2 , 2-tribromoethanal in tert-amylalcohol ( Avertin; Aldrich Chemical Co . , http://www . sigmaaldrich . com/ ) , and 50 ul of infectious virus diluted in PBS was inoculated intranasally . The 50% lethal dose ( LD50 ) titers were determined by inoculating groups of three mice intranasally with serial 10-fold dilutions of virus . LD50 titers were calculated by the method of Reed and Muench [25] . Individual body weights from eight mice were recorded for each group daily and monitored daily for disease signs and death for 14 d after infection . For determination of lung virus titers , 18 additional mice were infected intranasally with the intermediate inoculating dose ( 104 plaque-forming unit [PFU] ) of virus . On days 1–3 and 5–8 after infection , three mice from each group were killed , and whole lungs were removed aseptically and homogenized in 1 ml of sterile PBS . Homogenates were titrated for virus infectivity using a standard plaque assay . The statistical significance of virus titer data was determined by using analysis of variance . To determine the in vivo levels of cytokine supernatants from the lung , homogenates of the lungs of WH-infected mice were assayed for IFN-γ and TNF-α ( assay sensitivity , 2 pg/ml ) by use of enzyme-linked immunosorbent assay kits purchased from R&D Systems ( http://www . rndsystems . com/ ) . For high-containment laboratory work with 1918 recombinant viruses , the in vivo levels of cytokine proteins were determined from three individual mice per group . On day 4 after infection , mice were exsanguinated from the axilla and killed , and lung tissues were removed from naive and infected mice . Individual whole-lung samples were immediately frozen at −70 °C . On the day of analysis , tissues were thawed , homogenized in 1 ml of cold PBS , and centrifuged at 150g for 5 min . Cytokine protein levels were measured from clarified lung homogenates by the Bioplex Protein Array system [26] ( Bio-Rad , http://www . bio-rad . com/ ) using beads specific for mouse IL-1β , IFN-γ , and TNFα . Cytokine protein levels were measured according to the manufacturer's instructions by fluorescently conjugated monoclonal antibodies in duplicate against a standard curve .
It has been previously shown that H5N1 viruses from the Hong Kong 1997 outbreak fall into three separate pathogenicity phenotypes: low , intermediate , and high [22] . The intermediate phenotype had aa sequence identity with the high-pathogenicity phenotype and all of the previously identified molecular correlates of a high pathogenicity phenotype , but caused a less severe disease in mice . However , PB1-F2 was not examined in the study by Katz et al . because it was not known at the time [22] . Alignment of the 1997 human H5N1 PB1-F2 sequences available in the National Center for Biotechnology Information database ( http://www . ncbi . nlm . nih . gov/ ) revealed several aa changes that separated the high-virulence from the low-virulence groups . These were ( low-virulence versus high-virulence ) E6D , R53K , N66S , and R75H . These changes were silent in the open reading frame of the PB1 gene . Alignment of the proteins with other PB1-F2 sequences available in the database revealed that with the exception of the N66S substitution , all of the described mutations were previously present in other influenza viral strains . The N66S mutation was of a particular interest , since it was found only to be present in the highly virulent 1997 H5N1 group , in the PB1-F2 proteins of some avian isolates , and in the 1918 A/Brevig Mission/18 PB1-F2 ( Figure 1 ) . Interestingly , the A/HK/156/97 virus ( from the intermediate-virulence group ) , which was previously shown to possess all of the molecular signatures of the high-virulence group , possesses N at position 66 of the PB1-F2 protein . aa residue 66 resides in the C-terminal α-helical region of PB1-F2 . This region is the interacting domain for ANT3 and VDAC1 and contains the mitochondrial targeting sequence , making the C-terminal region essential for the function of PB1-F2 [19 , 27] . The location of the N66S mutation in the structure of PB1-F2 and its presence in the C-terminal region supports the hypothesis that this aa change could impact PB1-F2′s effects in vivo . We hypothesized that this aa substitution may be responsible for the decreased pathogenicity phenotype observed for the A/HK/156/97 virus . Given these findings , we proceeded to determine whether the PB1-F2 mutation in position 66 ( N66S ) in the 1997 H5N1 viruses contributed to viral pathogenicity . To better understand the increased pathogenicity in the infected mice , we examined the levels of TNF-α and IFN-γ in the lungs . IFN-γ levels were observed to be higher in mice infected with WH N66S virus , especially at days 7 and 8 after infection , when the levels were approximately two times higher than the levels in the WH virus-infected mice ( Figure 5A ) . Levels of TNF-α in the lung also showed significant differences late in infection . At days 7 and 8 after infection , TNF-α levels in mice infected with WH N66S virus had a two times higher increase over levels in WH-infected mice ( Figure 5B ) . Individual lung tissues were also collected on day 4 after infection from 1918 virus–infected mice . A single timepoint ( day 4 after infection ) was chosen because it was previously determined that maximal lung cytokine/chemokine levels occurred at this time among mice infected with highly virulent influenza strains [27 , 28] . Tissues were homogenized and lysates were assayed for cytokines by the Bioplex Protein Array system . Determination of IL-1α , IFN-γ , and TNF-α levels demonstrated that these cytokines were produced above their constitutive levels 4 d after infection with both 1918 S66N mutant and wild-type virus ( Figure 5C ) . All three cytokines were detected at significantly higher levels ( p ≤ 0 . 5 , analysis of variance ) in 1918 wild-type–infected than in 1918 S66N mutant–infected mice . Together , these data indicate that PB1-F2 may play a role in immunomodulation , especially later in infection during viral clearance .
Previous studies by our lab have shown that PB1-F2 contributes to the pathogenesis of the influenza A virus [20] . When expression of PB1-F2 was knocked out of a moderately virulent virus in mice , there was a significant loss in pathogenicity , indicating that PB1-F2 plays an important role in virulence [20] . In the present study , we show that a single aa change in PB1-F2 from highly virulent viruses increases pathogenicity in mice and modulates the immune response . It has been proposed that PB1-F2 causes apoptosis of immune cells , which may lead to decreased antigen presentation and a decrease in the adaptive immune response [17] . Humans infected with highly pathogenic viruses consistently have decreased lymphocytes and impaired immune response to influenza virus infection [4 , 8 , 10 , 21 , 29 , 30] . We wondered if these effects could be caused in part by PB1-F2 . In this study , we provide evidence that PB1-F2 does contribute to the high pathogenicity phenotype and that the N66S mutation , also found in the 1918 H1N1 virus , contributes to virulence in highly pathogenic viruses . After aligning the PB1-F2 sequences from H5N1 viruses that exhibited high- and low-pathogenicity phenotypes , a single aa change was found to correlate with high pathogenicity . The location of the N66S mutation also made it an excellent candidate for affecting the proapoptotic function of PB1-F2 . Position 66 is in the α-helical structure of PB1-F2 , in the mitochondrial targeting sequence . The location of aa 66 in the C-terminal mitochondrial targeting sequence of the protein could affect PB1-F2 interactions with ANT3 and VDAC1 , potentially increasing the induction of apoptosis by PB1-F2 [19] . Recombinant A/WSN/33 viruses were created to specifically examine the effects of the N66S mutation during viral infection . The recombinant virus WH has decreased pathogenesis in mice compared with that of A/WSN/33 ( unpublished data ) , likely due to the mismatched polymerase genes , resulting in less efficient replication in the host . The N66S mutation within the PB1-F2 protein partially reversed this attenuating effect . Within a natural setting , the presence of a “virulent” PB1-F2 may be important when influenza viruses cross species barriers or when new pandemic strains are generated by reassortment . In fact , the PB1 gene has been one of the segments found to reassort to create the pandemic strains of 1957 and 1968 , potentially giving these viruses a more pathogenic PB1-F2 and thus a higher virulence [13] . It is possible that the PB1-F2 protein could allow a newly reassorted virus to replicate in a new host efficiently enough to spread , and develop mutations to create a more efficient polymerase complex . In addition , influenza surveillance data shows that in recent history ( 1970 onward ) , H3N2 infections cause almost 14 times the number of influenza related deaths than H1N1 infections and are associated with a higher epidemic severity index ( as measured by the rate of increase in pneumonia and influenza mortality ) [31–33] . Interestingly , recent H1N1 isolates contain a truncated PB1-F2 , which possibly plays a role in their decreased virulence [19 , 34] . The mutation we investigate here is not currently found in recent H5N1 isolates; however , it is possible for those viruses to acquire the mutation either through the error-prone RNA polymerase or through reassortment with a virus that contains the N66S mutation . The observation that the WH N66S virus grew to higher titers in the lung and persisted at high titers for a longer time than the WH virus supports the role of PB1-F2 in allowing for increased replication . This may also explain the impairment of viral clearance in the mice infected with WH N66S . In addition , the 1918 wt virus showed higher lung titers and slower viral clearance when compared with the 1918 S66N virus . We suspect that the delay in viral clearance due to expression of PB1-F2 protein may allow for prolonged viral replication and development of irreversible pulmonary immunopathology , the findings observed with highly pathogenic influenza strains . CD8+ T cells are mainly responsible for viral clearance in the host , and it is possible that their function could be impaired by PB1-F2 [35 , 36] . In support of this , we observed that the WH N66S and wt 1918 viruses caused a significant increase in IFN-γ and TNF-α cytokine production over the WH and 1918 S66N viruses , respectively . Whether this change in cytokine levels is through the direct action of PB1-F2 or through its impact on viral replication in the lung is difficult to determine . However , the cytokine dysregulation is of special interest because it has been associated with both H5N1 and 1918 H1N1 virus infections . In previous studies , cytokine dysregulation was associated with high virulence and death in animal models [29 , 37] . Our study supports these findings and suggests that PB1-F2 could be one of the factors contributing to the cytokine dysregulation seen in H5N1 virus–infected patients and 1918 H1N1 virus–infected animals [4 , 37] . | PB1-F2 is the most recently discovered protein produced by the influenza A virus . It has been previously shown that PB1-F2 is present in the mitochondria , where it induces cell death; our laboratory has demonstrated that PB1-F2 is a contributor to pathogenesis in the mouse model of infection . To study PB1-F2 further , we examined highly pathogenic strains of avian influenza virus and located an amino acid change that seemed to be associated with increased death in mice . We studied this amino acid change in PB1-F2 at position 66 in two different viruses . A recombinant virus that has a PB1 gene from an H5N1 virus was used as well as a fully reconstructed 1918 pandemic virus . In this study , we show that a mutation in PB1-F2 found in highly pathogenic influenza A virus isolates causes nonpathogenic viruses to induce disease in mice . In addition , we show that the increased pathogenicity is associated with higher levels of virus and cytokines in the lungs . We conclude that PB1-F2 does affect pathogenicity , and that position 66 seems to play an important role in contributing to the effects of PB1-F2 in the mouse model . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"viruses",
"infectious",
"diseases",
"none",
"virology",
"in",
"vitro"
] | 2007 | A Single Mutation in the PB1-F2 of H5N1 (HK/97) and 1918 Influenza A Viruses Contributes to Increased Virulence |
In the oocytes of many animals including humans , the meiotic spindle assembles without centrosomes . It is still unclear how multiple pathways contribute to spindle microtubule assembly , and whether they are regulated differently in mitosis and meiosis . Augmin is a γ-tubulin recruiting complex which “amplifies” spindle microtubules by generating new microtubules along existing ones in mitosis . Here we show that in Drosophila melanogaster oocytes Augmin is dispensable for chromatin-driven assembly of bulk spindle microtubules , but is required for full microtubule assembly near the poles . The level of Augmin accumulated at spindle poles is well correlated with the degree of chromosome congression . Fluorescence recovery after photobleaching shows that Augmin stably associates with the polar regions of the spindle in oocytes , unlike in mitotic cells where it transiently and uniformly associates with the metaphase spindle . This stable association is enhanced by γ-tubulin and the kinesin-14 Ncd . Therefore , we suggest that meiosis-specific regulation of Augmin compensates for the lack of centrosomes in oocytes by actively biasing sites of microtubule generation within the spindle .
Spindles can assemble without centrosomes naturally in oocytes and artificially in mitotic cells [1]–[3] . This report asks whether the meiotic spindle is simply a mitotic spindle without centrosomes , or if oocytes have developed specific mechanisms which compensate for the absence of centrosomes . Centrosome-dependent and chromosome-dependent pathways are two well documented pathways which generate spindle microtubules [4] . Recently another pathway has been described . This pathway generates new microtubules from the side of existing spindle microtubules [5] , and therefore depends on other microtubule assembly pathways . A new microtubule branches at a low angle and with the same polarity as the original filament , and is mediated by the 8-subunit γ-tubulin recruiting complex Augmin in mitosis [6] , [7] , [8] . Augmin subunits are functionally interdependent , and require each other for protein stability [6] , [9] . The Augmin complex was originally identified in Drosophila , but shown to be conserved widely among higher eukaryotes [6] , [7] , [10]–[15] . The Augmin complex is associated uniformly with spindle microtubules , recruits γ-tubulin onto spindle microtubules , and increases both the density of spindle microtubules during mitotic metaphase as well as the density of central spindle microtubules during mitotic anaphase [6] , [9] , [10] , [16] . Therefore , Augmin's function in mitosis is proposed to amplify microtubules by generating new microtubules along existing spindle microtubules [6] . The localisation of Augmin to centrosomes or centrosomal regions is shown in mammalian interphase and in Drosophila mitotic prophase , although the significance of these localisations has not been demonstrated [7] , [11] , [17] , [18] . The microtubule binding activity of the human Augmin complex is regulated by phosphorylation by Plk1 and Aurora A [19] , [20] . Mutations of Augmin subunits have been isolated in Drosophila and lead to female sterility [9] , [16] . In mutant oocytes , the spindle microtubules are robustly assembled with chromosomes misaligned and homologous centromeres further apart [9] . This suggests that in oocytes Augmin plays a different role from that in mitosis . Here we report meiosis-specific regulation of this γ-tubulin recruiting complex that may function to substitute for the lack of the centrosomal activity in oocytes .
We have previously isolated a null mutant of the core Augmin subunit Wac , in which other subunits of the Augmin complex are also destabilised [9] . In the mutant oocytes , chromosomes are misaligned with homologous centromeres mostly bi-oriented , but further apart within the robustly assembled meiotic spindle . In later stages , chromosomes are mis-segregated at a very high frequency , leading to female sterility [9] . To establish the precise role of Augmin , we followed chromosome behaviour and microtubule assembly in live wild-type and mutant oocytes from nuclear envelope breakdown onwards ( Figure 1A , B; Movies S1 , S2 ) . When the nuclear envelope breaks down , chromosomes were clustered together in both wild-type and mutant oocytes . Then , in wild-type oocytes , individual chromosomes moved within a limited range ( up to 8 µm ) along the spindle axis for about 30 minutes , before chromosomes were congressed to the equator of the spindle . In contrast , in most mutant oocytes , chromosomes tended to cover a wider range than wild type and spread widely along the spindle axis from the start of spindle formation . Chromosomes eventually congress in most , but not all , older oocytes ( Figure S1 ) . Therefore Augmin is required for limiting the range of chromosome movement to keep chromosomes at the spindle equator especially in early stages . Previous studies showed that in mitosis , Augmin assembles the majority of spindle microtubules independently of centrosomes [6] , [9] , [16] , [17] . In contrast , our previous observation from fixed oocytes indicated that metaphase I spindles in a wac null mutant show no significant difference in microtubule density from those in wild type [9] . To investigate the kinetics of meiotic spindle assembly , live imaging was carried out in mutant and wild-type oocytes expressing GFP-α-tubulin . The timing of the appearance of the first spindle microtubules after nuclear envelope breakdown showed only a marginal difference between wild type and the mutant ( Figure 2A; Figure S2 ) . Taken together , these results indicate that unlike in mitosis of S2 cells , Augmin is largely dispensable for bulk spindle microtubule assembly in oocytes . In mitotic metaphase , Augmin is associated with spindle microtubules uniformly , and in cells depleted of Augmin , microtubule density is uniformly reduced . Therefore , it was originally proposed that Augmin amplifies ( augments ) microtubules by generating new microtubules alongside existing ones [6] . In fixed oocytes we previously showed that an Augmin subunit is concentrated at the polar regions of the meiotic spindle in oocytes [9] . We hypothesise that in oocytes , Augmin may generate microtubules within the metaphase spindle in a spatially biased manner . To test this hypothesis , we quantified the spatial distribution of the microtubule density within the spindle in wild-type and mutant oocytes . Oocytes were fixed and immunostained with tubulin , and the signal intensity was measured along each spindle from one pole to the other . After normalisation , the tubulin intensity was plotted along the spindle length ( Figure 2B ) . In the wac mutant , the tubulin intensity near the polar regions relative to the spindle equator is significantly lower than in wild type ( p<0 . 01 ) . This can be explained by a decrease of microtubule density near the poles or an increase at the spindle equator in the wac mutant . As Augmin is known to generate microtubules , we favour the first interpretation . To further confirm this observation from fixed samples , we examined live oocytes using GFP-tubulin . We found a significant increase in the frequency of spindles with a missing or weak spindle pole in mutant oocytes ( Figure 2C , D ) . Although these morphologies were not obvious in fixed samples either due to fixation or the absence of GFP-tagged protein , results from live oocytes further support the possibility of an underlying reduction in microtubules near the polar regions of the meiotic spindle . Therefore , our evidence suggests that Augmin is required for full assembly of microtubules near spindle poles in oocytes , rather than to simply amplify the existing microtubules as is seen in mitotic metaphase . We previously showed by immunostaining that in oocytes , the Augmin subunit Dgt2 concentrates at acentrosomal spindle poles in meiotic metaphase I . This is in clear contrast to the uniform localisation of Dgt2 along spindle microtubules in mitotic metaphase . To further confirm this localisation , we generated transgenic flies which express Dgt2 tagged with GFP ( GFP-Dgt2 ) in oocytes . Live analysis shows that GFP-Dgt2 signal was enriched in the polar regions of meiotic spindles ( Figure 3A ) . To exclude the possibility that the Dgt2 subunit on its own , but not the Augmin complex , localises to the poles , the localisation of other Augmin subunits was examined in oocytes . Wac tagged with GFP also showed similar enrichment at the spindle poles ( Figure 3B ) . Furthermore , immunostaining showed that Dgt6 was also concentrated to the polar regions of the spindle ( Figure 3C ) . This behaviour of Augmin as a complex is consistent with our previous observation that the amount of Dgt2 is greatly reduced in the absence of Wac in ovaries [9] . Therefore we conclude that the Augmin complex is concentrated at the polar regions of spindles in oocytes . We noticed that the intensity of Augmin in the polar regions of the spindle was variable from one oocyte to another . To test whether the polar accumulation of Augmin promotes chromosome congression , we measured the intensity of the GFP-Dgt2 signal at spindle poles and the degree of chromosome congression ( as the length of the chromosome mass along the spindle axis ) for each spindle from oocytes with various ages . We found a strong positive correlation between the intensity of Dgt2 signals in the polar region and the degree of chromosome congression ( Figure 3D ) . In S2 cells , it has been previously shown that Augmin on the mitotic spindle turns over very rapidly ( t1/2 of 4 seconds; [6] ) . To establish the molecular dynamics of Augmin on the spindle in wild-type oocytes , fluorescence recovery after photobleaching ( FRAP ) was used to measure the turnover rates of GFP-Dgt2 associated with the meiotic metaphase spindle in mature oocytes . After the spindle was photobleached by laser , recovery of the fluorescent signal was monitored over time . Surprisingly , the recovery was much slower in oocytes than that reported in mitotic metaphase of S2 cells . To quantify the turnover rate , recovery time of fluorescent signals were pooled together and plotted ( Figure 4A ) . A one-population model , or a two-population model with only one turnover population did not fit to the data satisfactorily . We found that a two-population model with two distinct turnover populations fit to the observation very well . It is estimated that 85% of GFP-Dgt2 belongs to the slow population with a half turnover time ( t1/2 ) of 5 minutes , and 15% belongs to the fast population with t1/2 of 8 seconds . This turnover rate is much slower than an estimated value from the published data in mitotic metaphase using S2 cells [6] . However , a direct comparison is difficult due to the difference in the experimental conditions . Additionally , S2 cultured cells may not accurately represent mitotic cells in flies . Spindle defects in S2 cells depleted of Augmin appear stronger than that of neuroblast mitosis in mutants lacking Augmin which are in fact viable [6] , [9] , [10] , [16] , [17] , [21] , [22] . Therefore , we carried out FRAP experiments on mitotic spindles in syncytial embryos laid by the GFP-Dgt2 expressing flies ( Figure 4B ) . The recovery curve showed that GFP-Dgt2 on metaphase mitotic spindles consists of 85% of the fast population with t1/2 of 15 seconds and 15% of non-turnover or very slow turnover population . We found that Wac-GFP also showed a slow turnover in oocytes comparable to GFP-Dgt2 confirming that this is the behaviour of the Augmin complex ( Figure S3 ) . Furthermore , FRAP of GFP-tubulin indicated that microtubule dynamics is comparable between mitotic spindles in syncytial embryos and meiotic spindles in oocytes , excluding a possibility that a slow turnover of Augmin in oocytes is a mere reflection of slow microtubule turnover in oocytes ( Figure S4 ) . Therefore we conclude that the Augmin complex is associated with spindle microtubules much more stably in oocytes than in mitosis . This indicated the existence of an oocyte-specific mechanism which stabilises the association of Augmin to the polar regions of the spindle . Augmin is thought to recruit the γ-tubulin complex to nucleate new microtubules on existing microtubules . It is possible that the stable interaction of Augmin with the meiotic spindle is caused by its attachment to the nucleated microtubule and for some reason nucleation of microtubules is much more efficient in oocytes ( “nucleate to stabilise” model; Figure 4F ) . In this case , a reduction of γ-tubulin would greatly decrease the population of stably attached Augmin . To test this possibility , γ-tubulin37C , the major γ-tubulin in oocytes , was depleted by RNA interference ( RNAi ) . Depletion was confirmed by Western blotting using an antibody that recognises all γ-tubulin isoforms ( Figure 4E ) , and immunostaining showed a spindle defect similar to the previous report [23] . FRAP of GFP-Dgt2 showed that there is no dramatic decrease in the proportion of the slow-turnover population ( 85% to 74%; Figure 4C ) . The morphology of the spindle polar region was not significantly changed during FRAP . The most significant difference is the increase in the turnover rate of the slow population ( t1/2 = 5 minutes to 2 minutes 50 seconds ) , although it is still much slower than the fast-turnover population ( t1/2 = 8 seconds ) . However , as Western blot showed residual γ-tubulin either from incomplete depletion by RNAi or the other γ-tubulin isoform ( γ-tubulin23C ) , the role of γ-tubulin may be bigger . Nevertheless , our quantitative study suggests that interaction with γ-tubulin is not the major cause of the stable population , but further stabilises the already stable population ( “stabilise , then nucleate”; Figure 4F ) . Next we looked into the involvement of Ncd ( a kinesin-14 ) , a minus-end directed motor which cross-links spindle microtubules . Ncd is not essential for viability but is important for pole focusing of acentrosomal spindles and accurate chromosome segregation in oocytes [24] . It could potentially anchor Augmin onto microtubules , or cross-link newly nucleated microtubules to existing microtubules and further stabilise the microtubule interaction of the stable Augmin population . FRAP of GFP-Dgt2 in ncd mutant oocytes showed that there is little change in the proportion of the slower population ( 85% to 82%; Figure 4D ) . The morphology of the spindle polar region was not significantly changed during FRAP . The most significant difference is the decrease in t1/2 of the slower population ( 5 minutes to 3 minutes 10 seconds ) . This suggests that interaction with Ncd does not play a major role in stabilising Augmin onto the spindle in the first place , but further stabilises the already stable population possibly by cross-linking the newly nucleated microtubule to an existing microtubule ( Figure 4F ) .
Oocytes form the spindle without centrosomes in many animals including humans and Drosophila [1] . We propose that oocytes have specific mechanisms which compensate for the lack of centrosomal activity in meiotic metaphase ( Figure 5 ) . In mitotic prophase , Augmin concentrates to centrosomal regions [17] . In mitotic metaphase , it localises uniformly and transiently to the spindle , and generates the majority of centrosome independent microtubules ( Figure 5; [6] , [9] , [16] , [17] ) . In oocytes , Augmin associates with the poles of the metaphase I spindle in a stable manner , and biases microtubule assembly near the poles which lack centrosomes ( Figure 5 ) . The microtubules generated from the spindle poles may be important to congress chromosomes towards the spindle equator . The position of chromosomes is thought to be determined by a balance of multiple forces acting on chromosomes . The main source of the forces is the kinetochore-microtubule interaction . This interaction mainly pulls , but also can push , chromosomes , and is sensitive to lack of tension [25] , [26] . In addition , other forces acting on chromosome arms , often called polar ejection forces , push chromosomes to the spindle equator [25] . These forces are generated by an interaction between chromosome arms and spindle microtubules . Proposed origins of polar ejection forces include motor activities on chromosomes ( chromokinesins ) , chromosomal proteins tracking microtubule ends and simple collision of chromosomes with polymerising microtubules [25] , [27] . It is still poorly understood how chromosomes find the equator of the spindle by a balance of these forces . The organisation of spindle microtubules results in spatial differences of some forces , and chromosomes are positioned where poleward forces and anti-poleward forces are balanced . It is proposed that chromosomes are congressed at the equator as polar ejection forces are strongest near the poles and gradually decrease at the equator , which reflects the density and polarity of microtubules [25] . Polar ejection forces may play even more important roles in chromosome congression in oocytes than mitosis , but crucial differences of spindles between the two modes of divisions make it more challenging to understand . As the spindle is formed without centrosomes in oocytes [28] , microtubule density is lowest near the poles [29] . This difference of the spindle geometry potentially may affect the spatial distribution of polar ejection forces and other forces . Another difference is that centromeres/chromosomes are clustered together even before the nuclear envelope breaks down [29] . After the spindle elongates from the clustered chromosomes , chromosomes initially become spread along the spindle , and then congressed . The third difference is that , without centrosomes , the position of chromosomes heavily influences the length of the spindle and the distribution of microtubules within the spindle . Furthermore , the age of the oocytes or the length of metaphase arrest increases chromosome congression and decreases spindle length [30] . We know from the observation of other mutants that more spreading of chromosomes results in longer spindles , but do not know whether the converse is correct . This intimate relationship makes it difficult to disentangle the causal relationship between spindle length/morphology and chromosome congression . In the wac mutant , chromosomes are more widely spread in the meiotic spindle with homologous centromeres further apart , even relatively to the spindle length [9] . This chromosome spreading is particularly prominent in the wac mutant at the early stages of spindle formation , and becomes less prominent at the later stages , suggesting that an Augmin-independent polar ejection force gradually takes over though not completely . The failure or delay of proper chromosome congression in the wac mutant is better explained by a decrease in the pushing force acting on chromosome arms ( polar ejection force ) rather than an increase in the pulling force acting on kinetochores . In wac meiotic spindles , we found that the density of microtubules in sub-polar regions is relatively reduced , consistent with the polar concentration of Augmin in wild-type spindles . These microtubules generated near poles may be important for polar ejection forces by interacting with chromosomes directly , or through motors or microtubule end tracking proteins . We found a strong correlation between Augmin accumulation in the polar regions and chromosome congression , although the correlation does not necessarily imply a causal relationship . These findings suggest that polar microtubules generated by Augmin at the spindle poles are important for the polar ejection force which congresses chromosomes in oocytes . However , we cannot exclude alternative possibilities , such as that a wac mutation more indirectly affects chromosome positioning by altering the general organisation of the meiotic spindle , or the localisation of factors that influence chromosome movement or architecture . We examined thread-like projections thought to be connecting homologous heterochromatin using a phospho-H3 antibody [23] , [31] , but no noticeable differences were observed . Furthermore , although we did not see Augmin at the central spindle region higher than cytoplasmic background during spindle formation , we cannot exclude a possibility that Augmin may play a role there . As we previously showed [9] , wac mutant oocytes exhibit an elevated frequency of mono-orientation of homologous centromeres ( ∼10% of X chromosomes ) , and further show a very high incidence of chromosome mis-segregation in both stages of meiosis ( ∼80% and ∼70% of meiosis I and II , respectively ) . Similarly , in nod mutant oocytes which show reduced congression of achiasmatic chromosomes , it was genetically estimated that about 50% of achiasmatic X chromosomes and about 2% of chiasmatic ones mis-segregate [32] . This evidence suggests that chromosome congression or a force behind chromosome congression is crucial for accurate chromosome segregation in oocytes . Additionally , a nod mutant frequently shows detachment of achiasmatic chromosomes from the spindle [29] , while such detachment is seen less often in the wac mutant . Thus Nod may play a role in anchoring chromosomes to microtubules , as well as generating a polar ejection force . The Augmin complex is much more stably associated with spindle microtubules in oocytes than in mitotic cells . This difference is not simply explained by a difference in spindle dynamics , as the turnover of spindle microtubules is fast in both oocytes and syncytial mitosis . Interestingly , the turnover of Augmin on spindle microtubules in oocytes is significantly slower than the turnover of spindle microtubules itself . At first glance , this seems contradictory or puzzling . However , Augmin freed from one depolymerised microtubule can easily be re-captured by other microtubules in proximity rather than diffusing into the cytoplasm . Therefore the Augmin turnover between spindle and cytoplasm can be much slower than the turnover of individual microtubules . Alternatively , as our data showed a small proportion of microtubules are less dynamic , Augmin somehow could preferentially bind to these stable microtubules . We also found that γ-tubulin and Ncd further stabilise the slow-turnover population of Augmin . One explanation is that Augmin already anchored to a microtubule is further stabilised by combined actions of γ-tubulin which nucleates a new microtubule and Ncd which crosslinks this new microtubule to existing microtubules . However , we could not exclude possibilities that these indirectly affect the turnover of Augmin by altering the spindle dynamics or organisation . In Xenopus egg extract , bipolar spindles can be formed in the presence and absence of centrosomes [33] . It has also been shown that a bipolar spindle can be assembled without centrosomes in mitotic cells [2] , [3] . Therefore , it is often assumed that the only difference between a mitotic and a meiotic spindle is the presence of centrosomes . But our study clearly demonstrated that a meiotic spindle in oocytes is more than a mitotic spindle without centrosomes , and meiosis-specific regulation of Augmin is a crucial part of this difference . The Augmin complex was originally identified in Drosophila , but shown to be conserved widely among higher eukaryotes [6] , [7] , [11] , [12] , [13] , [14] , [15] . Therefore our finding of meiosis-specific Augmin regulation has an important implication in our understanding of chromosome segregation and mis-segregation in human oocytes .
Standard DNA manipulation and immunological techniques were used throughout [34] , [35] . Full-length Rcc1 was fused to mCherry and cloned into pUASp expression vector . A mutation ( AAAATG to TAAATG ) was introduced upstream of the ATG to reduce the expression level . Dgt2 and Wac were cloned into the Gateway expression vectors pPGW and pPWG respectively . The plasmids were injected into w1118 embryos by Genetic Research Inc . For western blot , γ-tubulin ( GTU-88; Sigma ) primary antibody was used ( 1/500 ) . Fluorescently labelled secondary antibodies were detected by Odyssey ( Licor ) . Standard techniques of fly manipulation were followed [36] . All stocks were grown at 25°C in standard cornmeal media . w1118 was used as wild type . Heterozygous flies for P[TRiP . HMS00517]attP2 and nanos-Gal4 maternal driver were used for γ-tubulin37C RNAi . Details of mutations and chromosome aberrations can be found in [37] or at FlyBase ( http://flybase . org ) [38] . Immunostaining of non-activated oocytes was carried out as previously described [39] . Antibodies against α-tubulin ( DM1A; Sigma ) and Dgt6 ( 1/50; [17] ) were used . Live-imaging of meiosis I spindle was carried out as described [40] . Adult flies expressing Rcc1-mCherry ( except for FRAP ) were matured 3 to 5 days at about 21°C before dissection . Series of z-sections covering the entire spindle ( separated by 1 µm ) were taken every 1 to 2 . 5 minutes . The following transgenes were used as heterozygotes: UASp-GFP-α-tubulin , GAL4 under the maternal nanos promotor , and UASp-Rcc1-mCherry on the third chromosome , and UASp-GFP-Dgt2 and UASp-Wac-GFP on the second chromosome . The images were taken with a laser scanning confocal microscope for immunostaining and FRAP experiments , and with a spinning disc confocal for other live samples , as previously described [8] , [30] . Z sections were separated by 1 µm . Images were presented as a maximum intensity projection of the z-stacks , were processed using Photoshop/ImageReady ( Adobe ) , imageJ or Volocity , and the brightness and contrast were uniformly adjusted for the whole field without changing features of the images . Measurement of tubulin intensity throughout fixed spindles ( Figure 2B ) was carried out on the maximum intensity projection . Pixel intensity was measured along a line from one pole to the other . The position along the line was normalised to the spindle length . The fluorescence intensity was normalised to the maximum intensity . The spindle length was divided into 10 and the medians of pixel intensities for each division in each spindle were box-plotted . The signal intensity of sub-polar regions ( divisions 3 , 8 ) relative to the equator ( 5 , 6 ) was significantly different between wild type and the mutant ( p<0 . 01 ) . p-values were calculated using the Wilcoxon test ( Figure 2B , 3D ) , the Student's t-test ( Figure 2A ) and Chi square test ( Figure 2C ) . To compare Dgt2 accumulation at the poles with the spread of chromosomes , the sum of intensity ( I ) was measured on maximum intensity projection in two regions of interest ( ROI ) . The first ROI ( I1 , N1 = 100 , 000 to 160 , 000 pixels ) includes both spindle and cytosol and the second region ( I2 , N2 = 500 to 3000 pixels ) is the sum of the two spindle poles ( including as little cytosol as possible ) . The following equation was then used to substract the background intensity ( cytosol intensity ) : intensity per pixel at the poles = I2- ( I1-I2 ) / ( N1-N2 ) . For the FRAP in wild type , γ-tubulin37C RNAi and ncdD mutant oocytes , the females were matured at 25°C for five days before dissection . For all the oocytes , an identical region of interest ( ROI = 300×125 pixels; the pixel size is 0 . 1 µm ) covering the whole spindle was bleached . Pictures of one pole ( 3z sections 1 µm apart ) were taken every 5 seconds for 165 seconds . Three images were taken before bleaching . The analysis was performed on the maximum intensity projection . The fluorescence intensity was measured in an ROI of 10 pixels diameter within the pole . The average intensity for each oocyte ( F ( t ) ) was corrected for the acquisition-induced photobleaching using the average intensity ( FControl ( t ) ) of non-FRAPed spindles using equation 1 . The corrected intensity ( FCorr ( t ) ) was then normalised ( FNorm ( t ) ) between the pre-bleach and the post-bleach intensities ( t = 0 ) using equation 2 . A model of two recovering populations ( equation 3 ) fits well to the observations . ( 1 ) ( 2 ) ( 3 ) where f and s are the proportion of the fast and slow populations , and tf1/2 and ts1/2 are the half-recovery times of the two populations . For the FRAP in wild-type embryos , embryos were collected after 1 . 5 hour aging , manually dechorionnated and covered with halocarbon oil . For GFP-Dgt2 FRAP , an ROI ( 210×125 pixels ) covering only one half spindle was bleached as in oocytes; for both GFP-tubulin and Wac-GFP FRAP , the ROI ( 210×125 ) covered the whole spindle . The analysis was performed as for oocytes except that the reduction of fluorescence in a non-FRAPed spindle in the same embryo was used to correct for acquisition-induced photobleaching of corresponding FRAPed spindles ( the average of ≤25% after 80 s ) . | Although centrosomes are the main sites of microtubule assembly in mitotic cells , the meiotic spindle assembles without centrosomes in the oocytes of many animals including humans . It has also been shown that bipolar spindles can be assembled in mitotic cells even when functional centrosomes are artificially eliminated . It is still unclear how spindle microtubules assemble without centrosomes , and whether microtubule assembly is regulated differently in mitosis and meiosis . Here we investigated the role and regulation of the conserved protein complex Augmin , which recruits γ-tubulin to microtubules to generate new microtubules , in Drosophila oocytes . We found that meiosis-specific regulation of Augmin substitutes for a lack of centrosomal activity in oocytes by biasing microtubule assembly towards poles . As Augmin is conserved widely in higher eukaryotes , our finding has an important implication in our understanding of chromosome segregation and mis-segregation in human oocytes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"models",
"mitosis",
"meiosis",
"cellular",
"structures",
"drosophila",
"melanogaster",
"model",
"organisms",
"cell",
"division",
"genetics",
"biology",
"molecular",
"cell",
"biology",
"cytoskeleton"
] | 2013 | Meiosis-Specific Stable Binding of Augmin to Acentrosomal Spindle Poles Promotes Biased Microtubule Assembly in Oocytes |
Multiple mutations in the voltage-gated sodium channel have been associated with knockdown resistance ( kdr ) to DDT and pyrethroid insecticides in a major human disease vector Aedes aegypti . One mutation , V1016G , confers sodium channel resistance to pyrethroids , but a different substitution in the same position V1016I alone had no effect . In pyrethroid-resistant Ae . aegypti populations , V1016I is often linked to another mutation , F1534C , which confers sodium channel resistance only to Type I pyrethroids including permethrin ( PMT ) , but not to Type II pyrethroids including deltamethrin ( DMT ) . Mosquitoes carrying both V1016G and F1534C exhibited a greater level of pyrethroid resistance than those carrying F1534C alone . More recently , a new mutation T1520I co-existing with F1534C was detected in India . However , whether V1016I or T1520I enhances pyrethroid resistance of sodium channels carrying F1534C remains unknown . V1016I , V1016G , T1520I and F1534C substitutions were introduced alone and in various combinations into AaNav1-1 , a sodium channel from Aedes aegypti . The mutant channels were then expressed in Xenopus oocytes and examined for channel properties and sensitivity to pyrethroids using the two-electrode voltage clamping technique . The results showed that V1016I or T1520I alone did not alter the AaNav1-1 sensitivity to PMT or DMT . However , the double mutant T1520I+F1534C was more resistant to PMT than F1534C , but remained sensitive to DMT . In contrast , the double mutant V1016I+F1534C was resistant to DMT and more resistant to PMT than F1534C . Furthermore , V1016I/G and F1534C channels , but not T1520I , were resistant to dichlorodiphenyltrichloroethane ( DDT ) . Cryo-EM structures of sodium channels suggest that T1520I allosterically deforms geometry of the pyrethroid receptor site PyR1 in AaNav1-1 . The small deformation does not affect binding of DDT , PMT or DMT , but in combination with F1534C it increases the channel resistance to PMT and DDT . Our data corroborated the previously proposed sequential selection of kdr mutations in Ae . aegypti . We proposed that mutation F1534C first emerged in response to DDT/pyrethroids providing a platform for subsequent selection of mutations V1016I and T1520I that confer greater and broader spectrum of pyrethroid resistance .
Pyrethroid insecticides are synthetic analogs of naturally occurring pyrethrins from Chrysanthemum spp . [1] . Due to their low mammalian toxicity , high insecticidal activity and fast action , pyrethroids are currently a dominant class of insecticides used globally against mosquitoes and other human disease vectors . However , intensive use of pyrethroids has led to selection of resistant mosquitoes around the world . Pyrethroid resistance is currently a major obstacle in mosquito control [2] . Pyrethroids target voltage-gated sodium channels in insects . The pore-forming α1 subunit of sodium channels has four homologous repeat domains ( I-IV ) , each containing six transmembrane segments , S1-S6 ( Figs 1A and 2A ) . In each repeat , segments S1-S4 constitute a voltage-sensing domain . Eight segments S5 and S6 along with four membrane-reentrant P-loops , which connect S5s and S6s , form the pore domain . Voltage-gated sodium channels are responsible for initiation and propagation of the action potential in almost all excitable cells [3] . In response to membrane depolarization , sodium channels open ( activate ) and allow sodium ions to flow into the cell , causing rapid membrane depolarization , the rising phase of action potentials . A few milliseconds after channel activation , sodium channels immediately undergo fast inactivation , which is critical for the action potential termination , preventing excessive membrane depolarization [4] . In response to prolonged depolarization ( seconds to minutes ) sodium channels progressively enter into more stable slow-inactivated states . Slow inactivation is important for regulation of membrane excitability , action potential firing patterns , and spike frequency adaptation [5] . Pyrethroids , as well as DDT which was used intensively before 1960s , prolong the opening of sodium channels by inhibiting their inactivation and deactivation [6–9] . Pyrethroids are categorized into two groups based on their poisoning symptoms , chemical structures and effects on sodium channels . Type II , but not Type I pyrethroids have an α-cyano group next to the phenylbenzylalcohol moiety . Type II pyrethroid , such as deltamethrin ( DMT ) , inhibit deactivation of sodium channels to a greater extent than Type I pyrethroids , such as permethrin ( PMT ) , inducing much slower decay of tail currents associated with repolarization . Our understanding of pyrethroid interaction with sodium channels at the molecular level began with identification of mutations that confer resistance to DDT and pyrethroids , known as knockdown resistance ( kdr ) mutations , in a wide variety of insect species including Aedes aegypti , a vector of viruses causing dengue fever , chikungunya , Zika fever , yellow fever and other diseases . Eleven sodium channel mutations , V410L [10] , G923V [11] , L982W [11] , S989P [12] , V1016G/I [13–15] , I1011M/V [16] , T1520I [17] , F1534C [15] and D1763Y [18] were found to be associated with pyrethroid resistance in Ae . aegypti . Co-occurrence of multiple mutations appears to be a common phenomenon in populations of pyrethroid-resistant Ae . aegypti [19] . Examples include V1016G/S989P [12 , 20 , 21] , V1016G/F1534C [22] and V1016I/F1534C [14 , 23] . Amino acid positions of these and other mutations in this study are numbered based on the house fly sodium channel protein ( Genbank accession number: AAB47604 ) . Mutation V1016G located in IIS6 was first identified in PMT- and DDT-resistant Ae . aegypti from Indonesia , Vietnam and Thailand [11] . V1016G was often found associated with S989P in Thailand [12] , Malaysia [21] , Saudi Arabia [24] and other countries in the south-east Asia . S989P is located in the extracellular loop that connects segments IIS5 and IIS6 . Another mutation in the same position , V1016I , was found in thirty Ae . aegypti populations in Latin America [25] . V1016I was always found co-existing with F1534C ( segments IIIS6 ) in pyrethroid-resistant populations in South and North Americas [23 , 26] , Brazil [14 , 27] , Mexico [28 , 29] , and the USA [30] . In contrast , although co-existing mutations F1534C/V1016G were found in the DMT-treated Ae . aegypti populations in Singapore [22] , they also occurred in separate haplotypic populations [31] . Interestingly , F1534C alone was found in many DDT- and PMT-resistant Ae . aegypti populations in Thailand , Vietnam [12] and Venezuela [23] . More recently , a new mutation T1520I at the extracellular N-end of IIIS6 was found to coexist with F1534C in India [17] . Previous studies demonstrated that mutation F1534C reduced sensitivity of the Ae . aegypti sodium channel , AaNav1-1 , and the cockroach sodium channel , BgNav1-1a , to Type I pyrethroids ( e . g . , PMT and bioresmethrin ) in the Xenopus oocyte expression system [32–34] . V1016G was likely selected under the pressure of pyrethroids because it conferred resistance to both Type I and Type II pyrethroids [32 , 33] . However , whether V1016I or T1520I enhances the F1534C-mediated pyrethroid resistance remain unknown . In this study , we introduced V1016I , F1534C and T1520I alone and in various combinations in the AaNav1-1 channel , expressed the mutant channels in Xenopus oocytes , and examined their gating properties , pyrethroid and DDT sensitivity . We found that ( i ) like V1016I [26] , T1520I did not alter the sensitivity of AaNav1-1 channels to PMT or DMT , ( ii ) the T1520I+F1534C channel was resistant to PMT , but sensitive to DMT , ( iii ) the V1016I+F1534C channel was resistant to DMT and more resistant to PMT than the F1534C channel , and ( iv ) V1016I/G and F1534C channels , but not T1520I , were resistant to DDT . Cryo-EM structures of sodium channels suggest that T1520I in AaNav1-1 allosterically distorts the pyrethroid receptor site PyR1 . The small distortion per se does not affect action of insecticides , which we studied here , but in combination with F1534C it affects binding of PMT and DDT . Our results corroborated sequential selection of kdr mutations in Ae . aegypti: F1534C emerged first in response to DDT and/or pyrethroids , whereas V1016I and T1520I appeared later under more intensive selection from pyrethroid use .
( 1R , 3R , α-S ) -deltamethrin and isomer-mixed cypermethrin were purchased from Sigma-Aldrich ( Sigma-Aldrich , St . Louis , MO , USA ) . ( 1R ) -cis-permethrin and bifenthrin were purchased from Chem Service ( Chem Service , West Chester , PA , USA ) . β-cyfluthrin was purchased from Fluka ( Fluka , Ronkonkoma , NY , USA ) . ( 1R ) -cis-NRDC 157 , which is structurally similar to deltamethrin , but lacks the α-cyano group next to the phenylbenzylalcohol moiety , was a gift from Bhupinder Khambay ( Rothamsted Research , Harpenden , United Kingdom ) . The purities of these compounds were above 98% . Stock solutions of the compounds ( 100 mM ) were dissolved in dimethyl sulfoxide ( DMSO ) . The working solution was prepared in ND96 recording solution immediately prior to experiments . The concentration of DMSO in the final solution ( < 0 . 5% ) had no effect on the function of sodium channels in the experiments . V1016I , V1016G and F1534C channel constructs in the background of AaNav1-1 , a pyrethroid-sensitive sodium channel from Ae . aegypti , were available from a previous study [32] . In this study , we generated a double mutant V1016I+F1534C by introducing V1016I into the F1534C construct . We introduced T1520I into AaNav1-1 or F1534C to generate T1520I and T1520I+F1534C mutant channels , respectively . Site-directed mutagenesis was performed by PCR using Phusion High-Fidelity DNA Polymerase ( NEB , Ipswich , MA ) . The sequence of the primers used in mutagenesis to introduce T1520I mutation were “CAGCCGATTCGCGAGATCAACATCTACATGTACC” ( forward primer ) and “GGTACATGTAGATGTTGATCTCGCGAATCGGCTGC” ( reverse primer ) . The mutant clones were verified by DNA sequencing . AaNav1-1 and mutants were expressed in the Xenopus oocytes , Ovaries from oocyte-positive female Xenopus laevis purchased from Xenopus 1 ( Dexter , MI ) . The procedures for oocyte preparation , cRNA synthesis and injection were identical to those described previously [35] . cRNA was prepared by in vitro transcription with T7 polymerase using the mMESSAGE mMACHINE high yield capped RNA kit ( Ambion , Austin , TX ) . To enhance expression of AaNav1-1 and mutant channels , their cRNAs were co-injected into oocytes with Ae . aegypti tipE cRNA in the 1:1 molar ratio [36 , 37] . Sodium currents were recorded by using the oocyte clamp instrument OC-725C ( Warner Instrument , Hamden , CT ) , Digidata 1200A , and pCLAMP 6 software interface ( Axon Instruments Inc . , Foster City , CA ) . Methods for electrophysiological recording and data analysis were similar to those described previously [38] . The peak current was recorded by -10 mV test pulse from a holding potential of -120 mV . The peak sodium current was limited to 2 . 0–3 . 0 μA to achieve better voltage control . This was achieved by adjusting the amount of cRNA and the incubation time after injection . The voltage dependence of sodium channel conductance ( G ) was calculated by measuring the peak current at test potentials ranging from -80 to +65 mV in 5 mV increments and divided by ( V-Vrev ) , where V is the test potential and Vrev is the reversal potential for sodium ions . The peak conductance values were normalized to the maximal peak conductance ( Gmax ) and fitted with a two-state Boltzmann equation: G/Gmax=[1+exp ( V‐V1/2 ) /k]−1 where V1/2 is the voltage of half-maximal activation , and k is the slope factor . The voltage dependence of sodium channel inactivation was determined by using 100 milliseconds prepulses ranging from -120 to -10 mV in 5 mV increments from a holding potential of -120 mV , followed by test pulses to -10 mV for 20 milliseconds . The peak current amplitude during the test depolarization was normalized to the maximal current amplitude and plotted as a function of the prepulse potential . Data were fitted with a two-state Boltzmann equation: I/Imax=[1+ ( exp ( V‐V1/2 ) /k ) ]−1 where I is the peak sodium current , Imax is the maximal current evoked , V is the potential of the voltage prepulse , V1/2 is the half maximal voltage for inactivation , and k is the slope factor . Recovery time from fast inactivation was measured by a 100 milliseconds depolarizing pulse to -10 mV , then repolarization to -120 mV for an interval of variable duration , followed by a 20 milliseconds test pulse to -10 mV . The peak current during the test pulse was divided by the peak current during the inactivating pulse and plotted as a function of duration time between the two pulses . To determine the time constant for recovery , the curve was fitted by double exponential function: I=1‐[A1×exp ( ‐t/T1 ) +A2×exp ( ‐t/T2 ) ] where A1 and A2 are the relative proportions of current recovering with time constants Ƭ1 and Ƭ2 , and t is the recovery interval . The voltage dependence of slow inactivation was measured with 60 milliseconds conditioning pulses ranging from -100 mV to 0 mV in 10 mV increments , followed by repolarization to a holding potential of -120 mV for 100 milliseconds to remove fast inactivation , and at last a -10 mV test pulse for 20 milliseconds . The peak current amplitude during the test depolarization was normalized to the maximal current amplitude and plotted against the pre-pulse potential . Data were fitted with a two-state Boltzmann equation as above for recovery from fast inactivation . Recovery from slow inactivation was tested by a pre-pulse to -10 mV for 60 seconds to drive sodium channels into the slow inactivated state , followed by repolarization to -120 mV for 0 to 30 seconds , and finally a test pulse to -10 mV for 20 milliseconds . The peak current during the test pulse was divided by the peak current , which has a repolarizing duration of 30 seconds , and plotted as a function of duration between the prepulse and test pulses . Recovery from slow inactivation was fitted by a double exponential function as that used for recovery from fast inactivation . The method of pyrethroids application in the recording system was identical to that described previously [35] . Effects were measured 10 minutes after the pyrethroids application . Insecticide-induced tail currents were recorded by using a 100-pulse train of 5 milliseconds step depolarization from -120 to 0 mV with 5 milliseconds inter-pulse intervals [39] . The percentage of sodium channels modified by pyrethroids was calculated using the following equation [40]: M={[Itail/ ( Eh‐ENa ) ]/[INa/ ( Et‐ENa ) ]}×100 where Itail is the maximal tail current amplitude , Eh is the potential to which the membrane is repolarized , ENa is the reversal potential for sodium currents determined from the current-voltage curve , INa is the amplitude of the peak current during depolarization before exposure to insecticides , and Et is the potential of the step depolarization . The inhibitory effect of DDT on the sodium channel inactivation was assayed by measuring the remaining current at the end of a 20 ms depolarization to -10 mV from a holding potential of -120 mV and normalizing it to the peak current . The DDT application and data analysis method were identical to those reported previously [41] .
Sodium channels are critical for electrical signaling in the nervous system . Since kdr mutations could reduce insect fitness , we examined the functional properties of mutant channels V1016I , F1534C , T1520I , V1016I+F1534C and T1520I+F1534C expressed in Xenopus oocytes . All these channels generated sodium currents comparable to those in the wild-type channel . None of the mutants had any detectable changes in the voltage dependence of activation or fast inactivation ( Table 1 , Figs 1B , 1C , 2B and 2C ) . We further examined effects of mutations V1016I and T1520I on slow inactivation , and recovery from slow and fast inactivation . Neither V1016I nor T1520I significantly altered development of slow inactivation ( Table 2 , Figs 1D and 2D ) . The time courses of recovery from fast inactivation ( Figs 1E and 2E ) and slow inactivation ( Figs 1F and 2F ) of V1016I and T1520I channels were essentially the same as those of the AaNav1-1 channel . Thus , neither V1016I nor T1520I modified any measured functional properties of AaNav1-1 channel expressed in Xenopus oocytes . Mutations V1016I and F1534C were examined previously [32–34] . Here , we compared pyrethroid sensitivities of double mutant V1016I+F1534C and single mutation mutants V1016I and F1534C . We expressed the wild-type AaNav1-1 and the mutants in Xenopus oocytes , measured tail currents induced by pyrethroids ( Fig 3 ) using a multiple short-depolarizations protocol [39] , and determined the percentage of channels modified by pyrethroids . The channel modification by pyrethroids increased in a dose-dependent manner ( Fig 3 ) . Consistent with earlier findings [32 , 34] , mutation F1534C conferred the channel resistance to PMT , but not to DMT ( Fig 3 ) , whereas mutation V1016I did not reduce the channel sensitivity to either PMT or DMT ( Fig 3 ) . However , V1016I+F1534C channel was not only more resistant to PMT than F1534C , but also showed resistance to DMT ( Fig 3 ) . Thus , although V1016I alone had no effect on the channel sensitivity to PMT or DMT , it enhanced the F1534C-mediated resistance to both PMT and DMT . To evaluate the role of T1520I in pyrethroid resistance , we introduced T1520I alone or together with F1534C . Like V1016I , T1520I alone did not reduce the channel sensitivity to PMT or DMT ( Fig 4A and 4B ) . However , T1520I+F1534C channel showed greater resistance to PMT , compared to that of F1534C channel ( Fig 4A ) , but were not resistant to DMT . Thus , unlike V1016I , T1520I enhanced the resistance of F1534C channel to PMT , but not to DMT . Previously the F1534C channel was found resistant to Type I , but not Type II pyrethroids [26 , 28] . We further explored resistance of the double mutant channel T1520I+F1534C , to two Type I pyrethroids ( bifenthrin and NRDC 157 ) and Type II pyrethroids ( cypermethrin and β-cyfluthrin ) . β-cyfluthrin , cypermethrin and bifenthrin are commonly used in mosquito control . Consistent with data on PMT and DMT , T1520I+F1534C channel was resistant to Type I , but not Type II pyrethroids . Although structurally distinct from pyrethroids , DDT also inhibits inactivation and deactivation of sodium channels [42 , 43] and shares with pyrethroids two binding sites in insect sodium channels [32] . DDT induces extremely small and fast decaying tail currents in sodium channels and inhibits fast inactivation [41] . Therefore , we assessed the AaNav1-1 sensitivity to DDT by measuring DDT-induced non-inactivating current . Fig 5A shows representative current traces from the AaNav1-1 and mutants V1016G , V1016I , F1534C , T1520I , V1016I+F1534C and T1520I+F1534C . DDT caused smaller inhibition of fast inactivation in V1016G and V1016I than in the wild-type AaNav1-1 , indicating that the mutants were resistant to DDT . The V1016G channel was more resistant to DDT than the V1016I channel . In contrast , the T1520I channel was not resistant to DDT ( Fig 5 ) . Previously we have shown that F1534C slightly reduces sensitivity of AaNav1-1 to DDT [41] . Here we compared the DDT sensitivity of double mutants of T1520I+F1534C and V1016I+F1534C with that of F1534C . Channel T1520I+F1534C was more resistant to DDT than the wild-type channel , but there was no difference between the T1520I+F1534C and F1534C channels . However , the double mutant V1016I+F1534C conferred 2-fold more resistance to DDT than V1016I or F1534C channels . To describe homologous residues in various sodium channels we use labels , which are universal for P-loop channels [32 , 44] . A label refers to the channel repeat ( 1–4 ) , channel segment ( "i" for the inner helix S6 and "p" for the P-loop ) , and relative position of the residue in the segment ( Fig 6A ) . For example , following this nomenclature V1016 becomes 2i18 . To facilitate recognition of residues in Ae . aegypti , we use both residues numbers and labels . In homology models of AaNav1-1 , which are based on the X-ray structures of open eukaryotic potassium channel Kv1 . 2 [45] or open prokaryotic sodium channel NavMs [46] , residues V1016/2i18 and F1534/3i13 are located , respectively , in helices IIS6 and IIIS6 and contribute to the pyrethroid receptor site PyR1 in the II/III repeat interface [26] . The cryo-EM structures of cockroach sodium channel NavPaS [47] and electric eel sodium channel Nav1 . 4 [48] confirm that residues in positions 2i18 and 3i13 are located in the II/III repeat interface ( Fig 6B and 6C ) . F3i13 and M2i18 in EeNav1 . 4 ( Fig 6C ) are closer to each other than F3i13 and V2i18 in NavPaS ( Fig 6B ) . It should be noted that IIS6 in the non-functional NavPaS is distorted likely due to a π-helix bulge above I2i18 . Such bulges are not seen in IIS6 of EeNav1 . 4 or in the inner helices of prokaryotic sodium channels . In NavPaS , T3 ( i-1 ) ( i . e . , T1520 ) in the extracellular loop IIIP2-S6 is far from F3i13 and V2i18 and cannot directly interact with pyrethroids bound to the latter residues . In NavPaS , T3 ( i-1 ) donates an H-bond to the backbone carbonyl of D3p62 in the AID motif at the C-terminus of helix IIIP2 ( Fig 6B ) . This H-bond stabilizes the loop conformation . Channel AaNav1-1 has the same-length loop IIIP2-S6 and the same AID motif ( Fig 6A ) . Mutation T3 ( i-1 ) I eliminates the H-bond , thus destabilizing the loop conformation . The loop contains isoleucine I3i1 , which forms tight inter-repeat contacts with residues L2p37 and H2p38 in the P-loop helix IIP1 . AaNav1-1 has the same residues I3i1 , L2p37 and H2p38 ( Fig 6A ) , which are likely involved in the inter-repeat contact . The IIIP2-S6 loop destabilization due to mutation T3 ( i-1 ) I would cause small shifts of IIIS6 , IIP1 , as well as IIS5 and IIS6 , which form tight contacts with IIP1 . Possible consequences of this changes are discussed in a later section .
Mutations V1016I and T1520I alone with F1534C cause high levels of pyrethroid resistance in Ae . aegypti field populations [13 , 17 , 49] , but their effects on the sodium channel function and sensitivity to pyrethroids were unknown . Here we demonstrated that although neither mutation alone conferred pyrethroid resistance of mosquito sodium channels , they enhanced pyrethroid resistance caused by a common Ae . aegypti kdr mutation , F1534C . Specifically , V1016I+F1534C caused a high resistance to both a Type I pyrethroid , PMT , and a Type II pyrethroid , DMT , whereas T1520I+F1534C increased resistance to PMT , but not to DMT . Our electrophysiological analysis in Xenopus oocytes established that F1534C confers sodium channel resistance only to Type I pyrethroids , but not to Type II pyrethroids including DMT , cypermethrin and cyfluthrin [32 , 34] . These results are consistent with bioassay results from several field Ae . aegypti populations carrying the F1534C mutation [21 , 50 , 51] . For example , a population ( Kota Bharu ) from Malaysia was resistant to PMT , but not to DMT [21] . Populations with the homozygous F1534C mutation were susceptible to DMT [50] and λ-cyhalothrin , another Type II pyrethroid [51] . However , some field populations carrying the F1534C mutation were found to be resistant to DMT [52] . Similarly , although we show that the T1520I+F1534C channel is not resistant to Type II pyrethroids , populations in which T1520I+F1534C were detected were also resistant to DMT [17] . It is possible that in these populations of Ae . aegypti additional kdr mutation ( s ) in the sodium channel or other pyrethroid-resistance mechanisms , such as enhanced metabolic detoxification , contribute to resistance to DMT . Since V1016I or T1520I alone do not confer any resistance to pyrethroids , we suggest that likely they have been selected in populations with the background kdr mutation F1534C established in the field ( Fig 7 ) . Our results support the hypothesis on sequential evolution of F1534C and V1016I proposed in 2015 by Vera-Maloof et al . [49] . The hypothesis was based on a linkage disequilibrium analysis of the two mutations in Ae . aegypti collected in Mexico from 2000 to 2012 . It is also consistent with the observation that in natural populations frequencies of F1534C are higher and increase more rapidly than frequencies of V1016I [14 , 49 , 53] . V1016I likely has emerged in the F1534C background in response to intensive use of pyrethroids in mosquito control . Interestingly , while the F1534C channel is resistant to only Type I pyrethroids , V1016I+F1534C is highly resistant to both Type I and Type II pyrethroids . Thus , V1016I enhanced F1534C-mediated pyrethroid resistance to both Type I and Type II pyrethroids . We suggest a similar evolution path ( Fig 7 ) for the emergence of T1520I in the background of Ae . aegypti populations carrying F1534C in India [17] . Another mutation , V410L in IS6 , was recently found to co-occur with F1534C and V1016I in pyrethroid resistant populations in Brazil [10] , Mexico [54] and Colombia [51] . Unlike V1016I and T1520I , V410L alone confers resistance to both Type I and Type II pyrethroids . V410L could be selected independent of F1534C . However , while the frequency of the triple mutation increased drastically from 2000 to 2006 , heterozygote/homozygotes of V410L or V1016I without F1534C were detected at extremely low frequencies [54] . These results suggest that in these populations , strong selection pressure favors the haplotype carrying all three mutations , which likely confer the greatest level of pyrethroid resistance . Whether the concurrence of V410L+V1016I+F1534C provides fitness advantages remains to be determined . The pyrethroid resistance augmentation caused by mutations V1016I and T1520I in Ae . aegypti is reminiscent to that of kdr mutations in Anopheles gambiae mosquitoes and German cockroach Blattella germanica [35 , 55 , 56] . For example , mutation N1575Y detected in pyrethroid-resistant An . gambiae populations [57] , has no effect on the action of pyrethroids , but enhances pyrethroid resistance caused by mutations L1014F/S/W [56] . In cockroaches , mutations E435K and C785R alone did not reduce sodium channel sensitivity to pyrethroids . However , concurrence of either E435K or C785R with other kdr mutations , V410M in IS6 or L1014F in IIS6 , significantly increases pyrethroid resistance [42] . Collectively , these data reflect complexity of interactions of pyrethroids with sodium channels and indicate sequential evolution of insect resistance to pyrethroids . It is unknown whether or not the kdr mutations explored in this study were selected in populations due to intensive use of DDT prior to introduction of pyrethroids in the 1970s . The first case of insect resistance to DDT was documented by Busvine in 1951 [58] . The resistance emerged after widespread house-spraying with insecticides including DDT during 1947–50 to eradicate Anopelines . Furthermore , one DDT resistant strain was also cross-resistant to pyrethrins , naturally occurring prototypes of pyrethroids [58] . Although the DDT resistance was detected in the field [59 , 60] , decades passed before the first evidence that resistance to DDT could be caused by the sodium channel kdr mutations , which confer resistance to pyrethroids . For example , V1016G and F1534C ( Fig 5 ) , as well as several other kdr mutations , such as L1014F , confer sodium channel resistance to DDT [61] . Therefore , it is possible that some kdr mutations , such as V1016G and F1534C , could have emerged due to intensive DDT use in eradication of malaria and other arthropod pest management programs in the 1950s and 1960s , before pyrethroids were introduced [62 , 63] . It would be interesting to examine in future whether insect specimens collected in the 1950s and 1960s carry some kdr mutations . Some kdr mutations likely have appeared due to pyrethroid selection . For example , recently detected T1520I does not confer pyrethroid or DDT resistance by itself , nor does it increase F1534C-mediated resistance to DDT ( Fig 7 ) . How mutation T1520/3i ( -1 ) I may augment pyrethroid resistance of channel F1534/3i13C ? In NavPaS T1520/3i ( -1 ) is far above F1534/3i13 ( Fig 6B ) and cannot directly interact with PyR1-bound pyrethroids or DDT , consistent with our data that point mutation T1520/3i ( -1 ) I alone does not affect action of pyrethroids or DDT ( Fig 4 and Fig 5 ) . The fact that T1520/3i ( -1 ) I increases the channel resistance to PMT , which is caused by F1534/3i13C ( Fig 4 ) , indicates that T1520/3i ( -1 ) I allosterically induces a small deformation of PyR1 . Double mutant T1520/3 ( i-1 ) I+F1534/3i13C was much more resistant to PMT than F1534/3i13C , but remained sensitive to DMT ( Fig 4 ) . This concords with our data that point mutation F1534/3i13C decreases action of PMT , but not that of DMT ( Fig 3 ) . A possible cause is that DMT , but not PMT has a nitrile group , which is proposed to accept an H-bond from threonine T2o10 in IIS5 [64] . This H-bond would attract DMT closer to IIS5 and therefore shift farther from IIIS6 , making the DMT action insensitive to mutation F1534/3i13C . Substitution of the hydrophobic beta-branched V1016/2i18 with the hydrophobic beta-branched isoleucine has no impact the channel resistance to PMT or DMT ( Fig 3 ) . However , the fact that channel V1016/2i18I+F1534/3i13C is resistant to both DMT and PMT suggests that I1016/2i18 , which is larger than V1016/2i18 , shifts the PyR1-bound ligand closer to helix IIIS6 making action of pyrethroids sensitive to mutation in position i13 , which is located against position 2i18 in the II/III repeat interface ( Fig 6B and 6C ) . In our Kv1 . 2-based model of mosquito sodium channel [41] , DDT directly interacts with F3i13 in the PyR1 site and is less than 5 Å from V2i18 , which also contributes to the PyR1 site . This model is consistent with our data that mutations V1016/2i18G/I , F1534/3i13C and T1520/3 ( i-1 ) I+F1534/4i13C increase the channel resistance to DDT ( Fig 5 ) . Furthermore , our data that point mutation T1520/3 ( i-1 ) I does not affect the channel resistance to DDT ( Fig 5 ) is consistent with our proposition that this mutation induces a small change in the PyR1 site geometry . This change , however , increases the DDT resistance in the double mutant T1520/3 ( i-1 ) I+F1534/3i13C as it does in case of PMT . Thus , the experimental data of this study can be explained in view of our previous models of mosquito sodium channel and recent cryo-EM structures of eukaryotic sodium channels . Furthermore , the latter structures suggest an allosteric mechanism by which mutation T1520/3 ( i-1 ) I induces a small change of the PyR1 site geometry . In conclusion , our functional characterization revealed common and unique effects of V1016I and T1520I mutations on the channel sensitivity to DDT and Type I and Type II pyrethroids . Our results supported the hypothesis on sequential selection of kdr mutations in Ae . aegypti: F1534C ( and probably V1016I ) appeared first in response to DDT , while T1520I emerged later under the continued pyrethroid selection . Both V1016I and T1520I can be established in kdr populations that possess the F1534C mutation . | Intensive use of pyrethroids has led to the selection of resistance in mosquitoes , and knockdown resistance ( kdr ) is one of the major mechanisms of pyrethroid resistance . So far , eleven kdr mutations were identified to be associated with pyrethroid resistance in Aedes aegypti . Among the mutations , the V1016I and T1520I substitutions were found to be associated with F1534C but rarely found alone . F1534C confers sodium channel resistance to Type I pyrethroids including permethrin ( PMT ) . However , whether V1016I or T1520I enhances the F1534C-mediated sodium channel resistance remain unknown . In this study , our electrophysiological results confirmed their involvement in kdr and corroborate the previously proposed sequential selection of kdr mutations in Ae . aegypti: F1534C likely emerged first in response to DDT and/or pyrethroids , whereas V1016I and T1520I appeared later under more intensive selection from pyrethroid use . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"depolarization",
"membrane",
"potential",
"vertebrates",
"electrophysiology",
"neuroscience",
"animals",
"xenopus",
"animal",
"models",
"mutation",
"ion",
"channels",
"ddt",
"model",
"organisms",
"amphibians",
"ex... | 2019 | Molecular evidence of sequential evolution of DDT- and pyrethroid-resistant sodium channel in Aedes aegypti |
Cytosine methylation of DNA is an important epigenetic gene silencing mechanism in plants , fungi , and animals . In the filamentous fungus Neurospora crassa , nearly all known DNA methylations occur in transposon relics and repetitive sequences , and DNA methylation does not depend on the canonical RNAi pathway . disiRNAs are Dicer-independent small non-coding RNAs that arise from gene-rich part of the Neurospora genome . Here we describe a new type of DNA methylation that is associated with the disiRNA loci . Unlike the known DNA methylation in Neurospora , disiRNA loci DNA methylation ( DLDM ) is highly dynamic and is regulated by an on/off mechanism . Some disiRNA production appears to rely on pol II directed transcription . Importantly , DLDM is triggered by convergent transcription and enriched in promoter regions . Together , our results establish a new mechanism that triggers DNA methylation .
DNA methylation at the 5th position of cytosine to form 5-methylcytosine ( 5mC ) is an important epigenetic gene silencing mechanism conserved from plants , fungi to animals [1] , [2] . Even though most of the DNA methylation is relatively stable , dynamic DNA methylation has been observed during specific stages of animal development [3] . DNA methylation occurs in three different nucleotide sequence contexts: CG , CHG , and CHH ( where H is C , A , or T ) . Small non-coding RNAs have been shown to be involved in the establishment and maintenance of heterochromatin formation in different organisms . In the fission yeast Schizosaccharomyces pombe , small RNAs and the RNAi pathway mediate histone H3 lysine-9 methylation at the centromeric regions [4]–[6] . In plants , the asymmetrical CHH methylation is maintained by de novo DNA methylation mediated by 24-nt small interfering RNAs ( siRNAs ) [review in 7] , [8] . In mammalian germ cells , the Dicer-independent Piwi-interacting RNAs ( piRNAs ) are thought to be involved in DNA methylation [9] , [10] . In the filamentous fungus Neurospora crassa , about 2% of cytosines in the genome are methylated [11] . Nearly all of the known methylation sites are within transposon relics of repeat-induced point mutation ( RIP ) and the repetitive ribosomal DNA locus [11]–[13] . RIP is a genome defense mechanism that results in mutation and methylation of duplicated DNA sequences during the sexual cycle [11] , [14] . All previously known Neurospora DNA methylation is dependent on the histone methyltransferase , DIM-5 , which meditates trimethylation of histone H3 at the lysine9 ( H3K9me3 ) [15] , [16] . Heterochromatin protein 1 ( HP1 ) recognizes H3K9me3 and recruits DIM-2 , the only confirmed DNA methyltransferase in Neurospora , to methylate DNA [17]–[19] . For the natural RIP'd sequences , DNA methylation is more or less stable and is generally not required for the maintenance of H3K9 methylation [12] . In addition , the known DNA methylation events are not dependent on the canonical RNAi pathway [20] . Neurospora produces many types of small RNAs , including microRNAs , siRNAs , QDE-2 interacting RNA ( qiRNAs ) , and dicer-independent siRNAs ( disiRNAs ) , through diverse small RNA biogenesis pathways [reviewed in 21] . disiRNAs are a distinct class of small RNAs as they symmetrically are mapped to both strands of DNA and their production is independent of the known canonical RNAi components , including Dicer [22] . The disiRNA loci range from a few hundred base pairs to more than 10 kilobases in size and are located in gene-rich regions of the genome . The function of disiRNA is unknown . In this study we identified a new mechanism of DNA methylation that is associated with disiRNA loci . Our results showed that this type of DNA methylation , which we call disiRNA loci DNA methylation ( DLDM ) , is very different from the previously known DNA methylation in Neurospora . DLDM is highly dynamic , depends on transcription at disiRNA loci , and is triggered by convergent transcription in some loci .
DNA methylation is an important regulatory mechanism of transcription silencing . Therefore , we examined whether in Neurospora the disiRNA loci are associated with DNA methylation using the methylation-sensitive restriction enzyme-based PCR ( MSP ) assays [23] . The Neurospora genomic DNA from a wild-type strain was digested with the isoschizomers DpnII or BfuCI . Both enzymes digest unmethylated GATC sites , but only DpnII can cut at sites when C is methylated . Primer sets were then used for semi-quantitative and quantitative PCR ( qPCR ) . As shown in Figure 1A , a non-disiRNA locus ( NCU06312 ) was not methylated , as indicated by the lack of PCR amplification product after digestion by BfuCI , whereas a PCR product was readily detected for the ζ-η region , a relic of RIP that forms constitutive heterochromatin and was previously shown to carry DNA methylation [14] , [24] . For the 9 selected disiRNA loci with high level of disiRNA and with size larger than 5 kb , PCR products were detected in all samples after BfuCI digestion , indicating that these disiRNA loci are methylated . From quantitative PCR ( qPCR ) analyses , we estimated that percentages of methylation at the DpnII sites in these loci ranged from ∼3 . 5% to 28 . 8% , levels much lower than that of the ζ-η region ( 58 . 4% ) . We then applied qPCR to MSP to further determine the DNA methylation profile in disi-47 , disi-6 , and disi-29 loci , which are three large loci with high levels of disiRNAs ( Figure S1 ) . The DNA methylation profiles at these loci are correlated with disiRNA profiles: DNA methylation peaked at regions of peaked disiRNA expression , and there was little or no DNA methylation outside of the disiRNA loci ( Figure 1B and Figure S2 ) . Since the MSP results only reflect the methylation status of cytosines within the cleaved GATC sites , the detected methylation levels might be biased . To exclude this possibility , we performed methylated DNA immunoprecipitation ( MeDIP ) to detect DNA methylation based on 5mC density in disiRNA loci . For 13 loci ( disi-6 , 8 , 22 , 23 , 29 , 34 , 35 , 39 , 42 , 47–50 ) with relatively high disiRNA level and with size larger than 5 kb , all of them harbor DNA methylation ( data not shown ) . We further analyzed disi-6 , disi-29 and disi-47 loci with more primer sets for better resolution . As expected , the levels of DNA methylation were high in regions of high disiRNA expression and were low or absent outside of the disiRNA loci ( Figure 1C and Figure S3 ) . In constrast , no DNA methylation was detected at two negative control loci al-1 and NCU06312 . Since the previously known DNA methylation in Neurospora is located in regions derived from RIP , we calculated the RIP indices of the methylated disiRNA loci [25] . We observed no significant difference in RIP indices between that of the whole genome ( 1 . 11±0 . 33 ) and all of the 50 identified disiRNA loci ( 1 . 13±0 . 32 ) . Moreover , for each disi-6 , disi-29 and disi-47 loci , the lowest RIP indices are well above the threshold ( 0 . 7 ) for RIP-induced DNA methylation [25] ( Figure S4 ) . These data suggest that the DNA methylation in these disiRNA loci is not a result of RIP . We therefore named this type of DNA methylation DLDM ( disiRNA loci DNA methylation ) to distinguish it from the RIP-induced DNA methylation . To determine the nature of DLDM , we performed bisulfite sequencing of selected genomic regions . As controls , we first sequenced the ζ-η region , a relic of RIP process and the am locus , which was previously shown to be unmethylated [26] . As expected , every clone of the DNA at the ζ-η region was methylated to various degrees at both symmetric and non-symmetric cytosine sites with an average methylation frequency of 36% ( Figure 2A ) , similar to the estimated frequency ( 42% ) determined in the MSP assay . In contrast , the genomic DNA was unmethylated at the am locus . In our experiment , 99 . 5% of all cytosines were converted to uracils . This percentage was similar to the conversion rate ( 99 . 8% ) of an unmethylated PCR fragment after the bisulfite treatment , indicating that our bisulfite conversion was complete ( data not shown ) . To our surprise , bisulfite sequencing of the disiRNA regions with peak DNA methylation revealed that vast majority of the clones were not methylated , as indicated by the nearly 100% conversion of all cytidines into uracils . For a few of the disiRNA loci clones , however , almost all cytosines were maintained after the bisulfite treatment ( data not shown ) . To rule out the possibility that these highly methylated clones were due to incomplete bisulfite conversion , we first treated the genomic DNA with bisulfite , then performed PCR and subcloned the fragments of either the disi-6 or the am locus into plasmids . PCR was then performed by using the plasmids as templates , and the resulting PCR products were subjected to DpnII digestion , which can only digest the PCR products if GATC site remains intact after bisulfite conversion due to the protection of 5mC ( Figure 2B ) . As expected , all 111 clones of DNA from the am locus examined were resistant to the DpnII digestion , indicating that the bisulfite treatment of the genomic DNA was complete . For the disi-6 locus , however , 18 out of 123 DNA clones were cleaved by DpnII . Sequencing of the DpnII digestible ( “cut” ) and resistant ( “noncut” ) clones of disi-6 locus showed that in all “noncut” clones nearly 100% of cytosines were converted to uracils , indicating that the initial genomic DNA carried no 5mC ( Figure 2C ) . In contrast , most of the cytosines in the DpnII digestible clones were methylated . Similar results were obtained from the disi-47 and disi-29 loci ( Figure S5 ) . Since Neurospora is haploid , these results indicate that the DNA methylation at disiRNA loci is actively regulated and is either on or off within each nucleus: disiRNA loci are either not methylated or are heavily methylated once DNA methylation process is switched on . To confirm this finding , we used Southern blot analysis to visualize DNA methylation at several disiRNA loci . As shown in Figure 2D , both DpnII and BfuCI resulted in the same digestion pattern of the am locus , consistent with the lack of DNA methylation . For the ζ-η region , the BfuCI digestion resulted in the disappearance of one DNA fragment and the appearance of several additional higher molecular weight DNA fragments , consistent with all DNA molecules from this locus being methylated to some degree . The largest DNA fragment from BfuCI digestion was less than 2 kilobases ( kb ) , indicating that methylation in the ζ-η region is limited to a small region . For the disi-47 and disi-29 loci , however , BfuCI digestion did not significantly change the levels or the relative ratios of the DpnII-digested bands , suggesting that most of the DNA lack methylation . We did observe , however , a ladder of high molecular weight ( up to 8–10 kb ) DNA fragments in the BfuCI-digested DNA , indicating that when methylated , the DNA at the disiRNA loci are heavily methylated across a large DNA region . Taken together , these results suggest that DLDM is highly dynamic and is regulated differently from the known DNA methylation events within relics of RIP in Neurospora . The histone modification of H3K9me3 is mediated by the H3K9 methyltransferase DIM-5 . This enzyme is essential for DNA methylation at relics of RIP , but H3K9 methylation at these loci is generally maintained in the absence of DNA methylation [12] , [27] . To determine whether DLDM is mediated by H3K9me3 modification , we performed H3K9me3 chromatin immunoprecipitaton ( ChIP ) assays . Our results showed that the disiRNA loci are enriched by histones containing the H3K9me3 mark ( Figure 3 and S6 ) . However , in contrast to the ζ-η locus , in which the levels of H3K9me3 were not affected by the deletion of dim-2KO , the levels of H3K9me3 at the disi-47 , -6 and -29 loci decreased dramatically in the dim-2KO mutant . These results indicate that DLDM requires DIM-2 for the maintenance of H3K9me3 at the disiRNA loci . As disiRNA loci are gene rich , we examined whether transcription was important for disiRNA and DLDM . Our previous EST analyses showed that at least some disiRNA loci harbor fully or partially overlapped antisense transcripts , suggesting that convergent transcription might be a trigger of disiRNA production and/or DLDM [22] . We chose disi-47 locus to test this hypothesis since it harbors the well-studied circadian clock gene frequency ( frq , NCU02265 ) , which is known to produce the sense frq transcript and an overlapping antisense transcript qrf [28] , [29] . The frq gene encodes a core component of the Neurospora circadian clock . MeDIP and MSP assays showed that the promoter region of the frq gene was methylated; however , only low levels of DNA methylation were detected in the frq coding region ( Figure S3 ) . The transcription of both frq and qrf is activated by light in a process that mediated by the WHITE COLLAR ( WC ) complex , which consists of two PAS domain-containing transcription factors WC-1 and WC-2 [29] . Previous studies showed that the expression of both frq and qrf are high in constant light ( LL ) and are low in constant darkness ( DD ) [28] . Our qRT-PCR assays also confirmed that the level of frq mRNA was significant higher in LL than in DD ( Figures 4A ) . Importantly , our mRNA deep sequencing [30] and RT-qPCR assays also demonstrated the existence of light-dependent expression of transcripts from the promoter region of frq ( Figure 4B and S7 ) , where disiRNA level is high . In a wc-2KO mutant , frq mRNA and the transcripts originating from the promoter region were both abolished , indicating that , like frq mRNA , the promoter-specific transcription requires WC-2 . MeDIP assays showed that the level of DNA methylation in the promoter of frq was significantly higher in LL than that in DD in the wild-type strain and was completely abolished in the wc-2KO mutant ( Figure 4C ) . Together , these results suggest that DNA methylation at the frq promoter is dependent on WC-2-mediated transcription and that transcription at promoter region may be necessary for the DNA methylation process and disiRNA production . Similarly , divergent promoter transcripts are also clearly seen in other disiRNA loci such as disi-29 ( Figure S7 ) , suggesting that this architecture of transcription might be the reason of DLDM . To determine whether disiRNA production is dependent on the WC-dependent transcription , we performed small RNA deep sequencing analyses in wc-2KO strain . As shown in Figure 4D , the disiRNA abundance in the mutant is completely abolished at the disi-47 locus . This result is consistent with the loss of DLDM in the wc-2KO strain ( Figure 4C ) and suggests that disiRNA and DLDM are tightly linked and both triggered by pol II-dependent transcription .
Neurospora is a well-established model system for DNA methylation . All previously known DNA methylation in Neurospora occurs in relics of RIP [12] , [32] . RIP is a process that silences repetitive DNA sequences during sexual stage ( prior to meiosis ) by converting cytosine to thymine in target sequences and occurs mostly at CpA dinucleotide context [33] . The resulted A/T rich region then serves as a signal that induces methylation of the nearby region to silence gene expression . In this study , we showed that DLDM is established and maintained very differently from the RIP-induced DNA methylation . First , DLDM occurs in the gene-rich disiRNA loci that contain no relics of RIP or other repetitive elements . Second , in contrast to RIP'd regions in which DNA methylation is more or less constitutive and occurs in all alleles , most of the alleles are not methylated at disiRNA loci . In disiRNA loci , only a small percentage of alleles are extensively methylated with most of cytosines modified over a region that extends several kilobases . The dense cytosine methylation is similar to recently demonstrated dense methylation/hydroxylmethylation of cytosines in mouse embryonic stem cells [34] . The on-off pattern of DLDM indicates that DLDM is highly dynamic and that there is an inducible mechanism that mediates the establishment of DLDM . On the other hand , a de-methylation process may also exist to convert methylated alleles back into unmethylated alleles . Third , unlike the DNA methylation in the RIP'd regions , which is generally not required for maintenance of H3K9 methylation , DLDM is required for the maintenance of H3K9 methylation at the disiRNA loci . It suggests that an unknown mechanism should exist to recognize DNA methylation and in turn trigger H3K9me3 . Finally , DLDM is dependent on transcription . We demonstrated that DLDM is induced by convergent transcription from artificial constructs expressed in Neurospora . This conclusion is in agreement with the fact that most of disiRNA loci are known to produce sense and antisense RNAs [22] . Interestingly , the peaks of DNA methylation occurred at the promoter regions , where disiRNA expression also peaked . In addition , promoter-specific RNA transcripts were detected , and levels of these transcripts correlated with the levels of DNA methylation , suggesting that these non-coding RNA transcripts are involved in DLDM and are the precursors for disiRNAs . The induction of DLDM by transcription may explain its on/off pattern and suggests that a certain threshold level of transcription may be required for the establishment of DNA methylation . These results indicate that DLDM differs substantially from the typical DNA methylation in RIP'd DNA regions . It should be noted that transient DNA methylation was recently reported at the frq locus and was shown to be involved in setting the proper phase of circadian clock during the preparation of this paper [35] . In addition , the distribution of DNA methylation induced by convergent transcription is mainly accumulated upstream and peaks at about 3 kb from the TSS , suggesting that DLDM might also be involved in suppressing the promiscuous transcription at promoter region during transcriptional initiation . Indeed , recent studies suggest that pol II-directed gene transcription may adopt a gene loop structure by tethering promoter and terminator sequence , which enhances the transcriptional directionality toward the gene body [36]–[38] . Therefore , it is possible that the DNA methylation is a result of complex interaction between both ends of the gene for transcription initiation and termination , which strengthens the directionality of both sense and antisense transcription .
In this study , FGSC 4200 ( a ) was used as wild-type ( WT ) strain . Mutant strains wc-2KO , qde-1KO , qde-2RIP , qde-2RIP;sms-2RIP double mutant , and dcl-1RIP;dcl-2KO double mutant ( dclDKO ) were generated in previous studies [31] , [39] , [40] . The dim-2KO and dim-5KO strains were generously provided by Dr . Qun He [23] . Liquid cultures were grown in minimal medium ( 1× Vogel's , 2% glucose ) at 30°C overnight and then at room temperature with shaking at 130 r . p . m . for 24 h [41] . For liquid cultures containing QA , 0 . 01 M QA , pH 5 . 8 , was added to the liquid culture medium containing 1× Vogel's , 0 . 1% glucose and 0 . 17% arginine . To make the his-3 targeting Pqa-2:cul:1-gccP constructs , a PCR fragment containing the promoter of ccg-1 was inserted into the plasmid pDE3dBH-Pqa-2 [42] to generate Pqa-2::1-ccgP . Then a PCR fragment of luciferase gene ( luc ) was inserted between the two promoters , with the luc sense transcripts and antisense transcripts driven by ccg-1 and qa-2 promoter , respectively . The control construct , Pqa-2:cul , was created by inserting the luc gene into pDE3dBH-Pqa-2 , with qa-2 promoter driven antisense transcription of luc . The resulting constructs were introduced into the his-3 locus of dclDKO , his-3 strain [39] , a his-3 strain and an eri-1lKO , his-3 recipient strain . Approximately 10 µg genomic DNA was digested with BfuCI or DpnII , fractioned in 1 . 0% agarose gels , and transferred to nylon membrane . Hybridization probes were prepared from PCR products of interest ( primer sequences in Table S1 ) with Rediprime II DNA Labeling System ( GE Healthcare ) . Approximately 10 µg genomic DNA was sonicated into small fragments ( size ∼300–1000 bp ) . In each reaction , 1 µg of the 5-methylcytosine monoclonal antibody ( Epigentek ) was used to perform the MeDIP assay as previously described [12] , [43] . MeDIP samples were analyzed with qPCR with corresponding primers listed in Table S1 . In order to compare methylation in different regions , relative enrichment of DNA was calculated as the ratio of MeDIP sample over its input ( set as 1 ) , and the qPCR result of a primer pair of the am locus was used for normalization to correct for possible primer efficiency bias [44] . To compare MeDIP results of different samples or treatments , we performed the MeDIP at the same time with same batch of anti-5mC antibodies , due to the variation of MeDIP efficiency for different batches of antibodies . The ChIP assay was performed as previously described [45] . The immunoprecipitation was performed with an H3K9me3 antibody ( Abcam ab8898 ) . The relative enrichment was calculated as the MeDIP assay and the qPCR result of a primer pair of the am locus was used for normalization . Methylation specific PCR ( MSP ) was performed as previously described [23] . The methylation rate , determined by quantitative PCR , was calculated as the ratio of BfuCI-digested DNA signal to its input . A primer pair for ( 113–114 ) , whose PCR product carries no BfuCI/DpnII recognition site ( GATC ) , was used to normalize for loading and primer efficiency . The bisulfite PCR methylation analysis was carried out in three steps: 1 ) The bisulfite treatment of genomic DNA was carried out as described in the manual of EpiTech Bisulfite Kit ( Qiagen ) except that we used a modified thermal cycler condition: 99°C for 5 min followed by 60°C for 25 min; 99°C for 5 min followed by 60°C for 85 min repeated 3 times; and 99°C for 5 min followed by 60°C for 90 min . 2 ) Two rounds of nested PCR were performed; the PCR product of first round was diluted 10–100 fold and 1 µL was used for second round of PCR . The second round PCR products of the expected size were cloned into TOPO clone kit ( Invitrogen ) and individual clones were sequenced . Two strategies were used to examine DNA methylation in disiRNA loci . Strategy 1 , shown in Figure 2 , used plasmids as templates to amplify the cloned fragment . The genomic counterpart of the cloned fragment carries a GATC site . If the site was methylated , the fragment would be resistant to bisulfite treatment , whereas the unmethylated sites were converted into uridine and no longer recognized by DpnII or BfuCI . By identifying whether the PCR product was resistant to DpnII or not , we could distinguish whether the cloned fragment was methylated in the GATC sequence . Strategy 2 , used in experiments in Figure S5 , is similar to the first strategy except that one aliquot of genomic DNA was treated and one was not treated with BfuCI before bisulfite treatment , PCR , cloning , and sequencing . The average methylation rate was calculated by dividing total number of 5-methylcytosines by the total number of cytosines in the amplified sequence . The primers used for bisulfite sequencing are shown in Table S2 and S3 . Total RNA was extracted with TRIzol ( Invitrogen ) , digested with Turbo DNase ( Ambion ) and reverse transcribed into cDNA with SuperScript II ( Invitrogen ) . β-tubulin transcripts ( primer pair tub ) were used as loading control for quantitative PCR . Total RNA of wc-2KO strain , wild-type strain and dicerDKO Pqa-2:cul:1-gccP strain were extracted with the TRIzol reagent ( Invitrogen ) and small RNAs were enriched with 5% polyethylene glycol ( MW8000 ) and 500 mM NaCl as previously described [31] . Library construction and small RNA sequencing was performed by the Beijing Genomic Institute ( Shenzhen , China ) with Illumina standard protocol . All small RNA analyses were performed as described previously [22] except that an alignment tool Bowtie ( ver 0 . 12 . 7 ) was used to map the small RNAs onto the N . crassa genome . In order to compare the density of small RNAs between samples , a standard normalization method was applied by scaling total reads of different samples to those of the same library size [46]–[48] . To correct bias induced by ribosomal RNA degradation products , we filtered out the reads matching rDNA regions from the total reads and used the remaining reads for scaling . The density of small RNA is presented as the relative number of small RNAs in a 100 nt non-overlapping sliding window along the Watson or Crick strand of each chromosome . The sRNA sequencing data was visualized with Generic genome browser ( version 1 . 70 ) [49] . The NCBI accession number of the sRNA deep sequencing data reported in this study is GSE47666 . | DNA methylation in eukayrotes refers to the modification of cytidines at 5th position with methyl group ( 5mC ) . Though absent in some species , DNA methylation is conserved across fungi , plants and animals and plays a critical role in X chromosome inactivation , genomic imprinting , transposon silencing etc . In addition , DNA methylation also occurs at the promoter sequence to regulate gene expression . Filamentous fungus Neurospora crassa has a well-known mechanism of DNA methylation for genomic defense . During sexual stage repetitive sequences ( e . g . transposons ) are recognized and point mutations are introduced . During vegetative stage these mutations serve as signals for establishing static DNA methylation to silence all copies of the sequences . In this study , we report a new type of DNA methylation in Neurospora . It is tightly linked to a type of non-coding small RNA termed dicer-independent siRNA ( disiRNA ) and therefore was termed disiRNA loci DNA methylation ( DLDM ) . DLDM is dynamic regulated and shows an on/off pattern , i . e . most alleles contain no 5mC but some are densely methylated . Interestingly , DLDM can be triggered by convergent transcription and is accumulated at promoter regions . In summary , our findings demonstrate a new type of dynamic DNA methylation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Convergent Transcription Induces Dynamic DNA Methylation at disiRNA Loci |
The relationship between malaria and undernutrition is controversial and complex . Synergistic associations between malnutrition and malaria morbidity and mortality have been suggested , as well as undernutrition being protective against infection , while other studies found no association . We sought to evaluate the relationship between the number of malaria episodes and nutritional statuses in a cohort of children below 15 years of age living in a rural community in the Brazilian Amazon . Following a baseline survey of clinical , malaria and nutritional assessment including anthropometry measurements and hemoglobin concentration , 202 children ranging from 1 month to 14 years of age were followed for one year through passive case detection for malaria episodes . After follow-up , all children were assessed again in order to detect changes in nutritional indicators associated with malaria infection . We also examined the risk of presenting malaria episodes during follow-up according to presence of stunting at baseline . Children who suffered malaria episodes during follow-up presented worse anthropometric parameters values during this period . The main change was a reduction of the linear growth velocity , associated with both the number of episodes and how close the last or only malaria episode and the second anthropometric assessment were . Changes were also observed for indices associated with chronic changes , such as weight-for-age and BMI-for-age , which conversely , were more frequently observed in children with the last or only episode occurring between 6 and 12 months preceding the second nutritional assessment survey . Children with inadequate height-for-age at baseline ( Z-score < -2 ) presented lower risk of suffering malaria episodes during follow-up as assessed by both the log-rank test ( p =0 . 057 ) and the multivariable Cox-proportional hazards regression ( Hazard Ratio = 0 . 31 , 95%CI [0 . 10; 0 . 99] p=0 . 049 ) . Malaria was associated with impaired nutritional status amongst children in an endemic area of the Western Brazilian Amazon where P . vivax predominates . Our data all supports that the association presents differential effects for each age group , suggesting distinct pathophysiology pathways . We were also able to demonstrate that undernourishment at baseline was protective to malaria during follow-up . These findings support an intriguing interaction between these conditions in the rural Amazon and the need for a more integrative approach by health systems in endemic areas .
Malaria is one of the most serious public health problems in the world , with 3 . 3 billion people at risk of contracting the disease and almost one million deaths annually , primarily in children under five years of age [1] . Although Latin American countries have experienced important reductions in malaria incidence , 490 , 545 cases were still reported in 2011 , of which 363 , 948 ( 76 . 7% ) were caused by Plasmodium vivax . Of these , 263 , 767 cases were reported in Brazil , mainly in the Amazon region , also with a vast majority of P . vivax ( 87 . 8% ) [1 , 2] . It has long been acknowledged that populations residing in malaria-endemic areas generally live under socio-economic conditions leading to poor nutritional status . The groups at highest risk for the adverse effects of malaria , children and pregnant women , are also those most affected by poor nutrition . Although it has been suspected that nutrition may influence susceptibility to the disease or alter its course , there have been comparatively few efforts to comprehensively examine such interactions [3 , 4] . It has been gradually recognized that any infection is associated with a risk of worsening the nutritional status [5] . Whereas some studies have suggested that undernutrition is protective for presenting symptomatic malaria [4 , 6] , other studies have shown that undernutrition or the worsening of nutritional status may result in clinical complications and severe malaria by modifying the immune response [7–9] . On the other hand , some studies [10–12] , have not been able to show a correlation between infection by Plasmodium and undernutrition [13 , 14] , with some studies even describing an antagonistic association between both entities [15] . The majority of studies addressing the relationship between malaria and nutritional status come from Africa , where P . falciparum is the main species causing malaria . In Latin America , there is scant information regarding the association between infectious diseases and nutrition . A survey performed in Central American countries showed that diarrhea and respiratory diseases worsened the nutritional status of children [5] . This nutritional impairment is most likely the result of multiple infections , poor socioeconomic status and inadequate diet , which in combination result in growth limitation . Previously conducted cross-sectional and case-control studies from Brazil have demonstrated an association between malaria and undernutrition in adults [16 , 17] . In Colombia , a positive association was found between malaria and undernutrition prevalence in children [18 , 19] . These studies however , do not allow for temporal analysis and a more conclusive causal association or the evaluation of dynamic parameters including growth velocity . In the Peruvian Amazon , children with vivax malaria disease were found to have a delay in linear growth and weight [20] . A recent study has found intriguing results of nutritional status and cognitive function after deworming in a malaria-endemic area in Cote d’Ivoire [21] , with still unclear conclusion on the joint effect of helminths and malaria on nutritional status [22] . In the present study , we sought to analyze the occurrence of malaria episodes and undernutrition in a cohort of children living in a rural Amazonian community where malaria ( caused by both P . vivax and P . falciparum ) is endemic . The aim was to evaluate the effect of malaria episodes on the anthropometric nutritional status indicators change and to explore the influence of baseline height-for-age z-scores ( HAZ ) , which more accurately reflects chronic undernutrition with less interference from recent episodes of malaria or other acute conditions ) on the risk of subsequent malaria incidence .
The study was approved by the Ethics Committee Board of the Fundação de Medicina Tropical Dr . Heitor Vieira Dourado ( 1899/2008 and 918/2010 approvals ) . Each participant and his/her parents or legal guardians signed written informed consent forms . The study was undertaken in two rural communities located in two recently colonized areas devoted to agriculture ( Panelão and Castanho Sítio Communities ) , from May 2008 to May 2011 . These settlements are located in the Municipality of Careiro , in the Amazonas State . The municipality has an area of 6 , 124 , 300 km2 and has 31 , 063 inhabitants . The climate is tropical and humid , with rainfalls ranging from 2 , 100 to 2 , 400 mm per annum . The municipality is connected to the capital of the state , Manaus , through a federal road ( 112 km of distance ) . Malaria is endemic in this area . The major economic activities are family farming , hunting and fishing . Drinking water comes from rainwater reservoirs or creeks . Garbage collection and sanitation are absent . Two health agents in each community are responsible for health care . The total population of both communities is 790 people , according to the census performed before the beginning of the study , including 300 children ranging from 1 month to 14 years of age . A cohort of 248 children ranging from 1 month to under 15 years of age was recruited for the present study , with 202 completing follow-up . As part of the baseline survey , a questionnaire was completed to collect socio-demographic information , including health history , maternal education , housing , and the number of assets ( television sets , bicycles , fridge , motorcycle and cars ) . The housing characteristics were combined using principal component analysis ( PCA ) as previously described [23] , in order to generate socioeconomic status indicators to classify individuals into one of three categories: the richest quintile , the three middle quintiles or the poorest quintile ( Table 1 ) . Children were recruited from May 2008 to May 2010 . All children under 15 years of age were eligible and included if their guardians provided informed consent . Children lost to follow-up were excluded from the analysis . At the baseline survey , children were weighed and measured , and then subjected to thick blood smear , hemoglobin measurement and stool examination and were presumptively treated with mebendazole ( 100mg bid for 3 days ) irrespectively of the stool results . Children were followed-up for 12 months through passive case detection of malaria based on microscopy in case of fever occurring at any time during the study . Anthropometric and hemoglobin measurements and thick blood smear ( to diagnose asymptomatic parasitaemia ) were repeated at the conclusion of the follow-up period , when no stool samples were repeated . Fig 1 illustrates the study design , showing 5 possible patients with different number/timing of malaria episodes . To evaluate whether malaria affects nutritional status , the following outcome variables were used: growth velocity and the weight for height z-scores ( WHZ ) , weight for age z-scores ( WAZ ) , height for age z-scores ( HAZ ) and body mass index for age ( BMI-Z ) nutritional index scores , all of them considering the WHO standard cut-offs [24] . Malaria was ascertained by the presence of fever and positive parasitemia with additional reporting on the number of malaria episodes experienced by each individual and time between each of the episodes in relation to the first and second visits . Therefore we considered only disease caused by Plasmodium and did not systematically examined patients for detection of asymptomatic infection . Nutritional assessment through anthropometric measurements was performed during the cross-sectional evaluations , at the beginning and at the end of the follow-up . Anthropometric measurements were performed with minimal clothing and no shoes , according to guidelines by the Brazilian Ministry of Health [25] . Weight and height were obtained by internationally recommended methods [26] . Weight was measured using a digital weight balance for children above 2 years of age and a mechanical paediatric balance for younger children , both in the grams scale , while height was assessed by a single observer with the use of a portable anthropometer stadiometer with lateral scale in centimeters . All instruments were adjusted and verified after each measurement and fully calibrated every three months . The measurement techniques were harmonized according to the mentioned procedures [25–27] , with all team members following the same standardized protocol . Body mass index ( BMI ) was calculated using the program ANTHRO and ANTHRO PLUS [24] . BMI for age Z-scores below -2 were defined as undernutrition . Growth velocity was measured in cm/year , defined as the difference between the final and initial height in the period of 12 months . The classification of each individual as having adequate or inadequate growth was made according to the WHO standards for each age [28] Children were categorized into three age groups for analyses , according to WHO guidelines for use of indicators WAZ , HAZ , WHZ and BMIZ [32]: ages below 5-years-of-age , between 5 and 10 , and between 10 and 14 ( for which WAZ and WHZ were not applied ) . In order to evaluate the impact of malaria on changes on the nutritional status , two main approaches were used . First , we evaluated the nutritional status at end of follow-up according to occurrence of any malaria episode , followed by a further categorization of malaria regarding number of episodes during the study period . In order to investigate the influence of timing of malaria episode on the impact over the anthropometric changes , we categorized the interval between the last or only malaria episode and the second measurement ( IME in Fig 1 ) . Chi-squared test comparing the proportion of children with adequate and inadequate status was performed for each anthropometric variable according to the different malaria exposure variables ( presenting malaria , number of malaria episodes and timing of last malaria episode ) . Univariable and multivariable logistic regression were performed for each age group; the latter after adjustment with the a priori defined variables age , gender , maternal education , and socioeconomic status . This association was tested against the occurrence or not of malaria episodes during the study period , number of malaria episodes and malaria categorization of time from last or only malaria episode , providing odds ratios ( OR ) and Wald test p-values , with children with no malaria serving as reference in all analyses . Additionally , the interaction between number of malaria episodes and time from last or only episodes was tested for each variable of interest by evaluating the addition of a linear interaction term to the multivariable logistic regression models for each outcome of interest . A second analysis was performed considering the change on the Z-score of malaria indicators from the first to the second assessments according to malaria status using univariable and multivariable linear regression following the same approach as previously described . Statistical significance was considered if p<0 . 05 The association between stunting ( height for age below -2 Z-scores ) at baseline and the risk of development of malaria was assessed considering time from enrollment to the first or only malaria episode , using survival analysis techniques with inspection of the Kaplan-Meier survival curve , with performance of log-rank test and computation of adjusted Cox proportional hazard ratios [33] . This was the only index for which this association was analyzed as it is an indicator of chronic malnutrition and also for it being less likely to suffer influence of recent acute conditions ( i . e . malaria and other acute febrile illnesses ) , which could interfere and confound the analysis . All analyses were performed in Stata v . 13 . 1 ( Statacorp , USA ) .
During the 12-month follow-up period , 248 of the 300 children eligible to participate in the cohort study were enrolled; 46 children were lost to follow up and the remaining 202 children were successfully followed ( Fig 2 ) . As shown in Table 1 , most children included were between 5 and 10 years of age ( 53 . 5% ) , and a minority had experienced previous malaria episodes ( 25 . 2% ) . Hemoglobin below 11 . 9g/dL was observed in 110 children ( 54 . 5% ) and the most common helminth parasite infecting children was Ascaris lumbricoides ( 26 . 7% ) . During the follow-up period , 87 children ( 43 . 1% ) presented with at least one episode of malaria , and a total of 164 malaria episodes were observed . The remaining 115 children ( 56 . 9% ) did not develop malaria during the study period . Among the infected subjects , 46 children ( 52 . 9% ) had one malaria episode , 21 children ( 24 . 1% ) had two malaria episodes , and 20 children ( 23% ) had three or more malaria episodes . Regarding the Plasmodium species involved , 119 episodes ( 72 . 6% ) consisted of mono infections with P . vivax , 37 episodes ( 22 . 5% ) involved P . falciparum alone , and 8 episodes ( 4 . 9% ) were mixed P . vivax and P . falciparum infections . No episodes of severe malaria occurred amongst the cohort participants . Table 2 shows the nutritional status of the children at the baseline survey . In general , nutritional profile was similar amongst the three age groups . Children under 5 years of age showed a better nutritional status , considering WAZ and HAZ means scores . Frequencies of HAZ < -2 ranged from 7 . 7% in children under 5 to 10 . 9% in children between 10–14 years . Body mass index-z < -2 ranged from 1 . 9% in children 5–10 years of age to 7 . 7% in children under 5 . The multivariable analyses considering age , gender and helminth infection could not identify any factor associated with inappropriate scores at baseline ( S1 Table ) . There was an association between final assessment being underweight , low BMI and inadequate growth velocity in children 5–10 years of age who developed malaria during follow-up ( Table 3 ) . Table 3 summarizes the results of the univariable and multivariable analysis evaluating nutritional profile according to malaria status . For 5–10 years-old children who had malaria [aOR 4 . 0 ( 95% CI 1 . 4; 11 . 4 ) ; p = 0 . 008] and a time of 6–12 months from the last malaria episode to the second nutritional assessment [aOR 4 . 4 ( 95% CI 1 . 3; 15 . 3 ) ; p = 0 . 020] there was a significant association with increased odds of inadequate growth velocity . In this age group , although with imprecise aORs , there is a trend of association between presenting one malaria episode [aOR 9 . 1 ( 95% CI 0 . 9; 85 . 8 ) ; p = 0 . 053] and a time of 6–12 months from the last malaria episode to the second nutritional assessment [aOR 8 . 4 ( 95% CI 0 . 9; 79 . 2 ) ; p = 0 . 062] and WAZ<-2 , as there was for BMI-Z<-2 for having one malaria episode [aOR 7 . 5 ( 95% CI 0 . 8;73 . 0 ) ; p = 0 . 081] and a time of 6–12 months from the last malaria episode to the second nutritional assessment [aOR 6 . 9 ( 95% CI 0 . 7;65 . 9 ) ; p = 0 . 093] . No significant interaction was found between nutritional indicators at baseline and second measurements and malaria for children <5 year of age ( Table 4 ) . For the 5–10 years group , malaria was significantly associated with WAZ [adjusted β = -0 . 3 ( -0 . 5;-0 . 1 ) ; p = 0 . 025 ) ] and HAZ [adjusted β = -0 . 1 ( -0 . 3;0 . 0 ) ; p = 0 . 035 ) ]; one malaria episode was significantly associated with WAZ [adjusted β = -0 . 3 ( -0 . 6;0 . 0 ) ; p = 0 . 024 ) ] and BMI-Z [adjusted β = -0 . 6 ( -1 . 1;-0 . 2 ) ; p = 0 . 005 ) ]; two [adjusted β = -0 . 3 ( -0 . 5;-0 . 1 ) ; p = 0 . 012 ) ] or ≥3 [adjusted β = -0 . 3 ( -0 . 5;0 . 0 ) ; p = 0 . 023 ) ] malaria episodes were significantly associated with HAZ . A time of 6–12 months from the last malaria episode to the second nutritional assessment was significantly associated with WAZ [adjusted β = -0 . 4 ( -0 . 7;-0 . 1 ) ; p = 0 . 006 ) ] and BMI-Z [adjusted β = -0 . 6 ( -1 . 0;-0 . 2 ) ; p = 0 . 006 ) ] . For 10–14 years-old children , ≥3 malaria episodes was significantly associated with HAZ [adjusted β = -0 . 3 ( -0 . 6;-0 . 1 ) ; p = 0 . 014 ) ] . The results of the analysis on the risk of malaria episodes related to the baseline nutritional status was assessed only for the or the HAZ indicator as this reflects chronic malnutrition more accurately and are shown on Fig 3 . It was possible to observe a lower rate of malaria amongst children classified with low levels ( Z-score < -2 ) on both the log-rank test ( p = 0 . 057 ) and the multivariable Cox-proportional hazards regression ( Hazard Ratio = 0 . 31 , 95%CI [0 . 10; 0 . 99] p = 0 . 049 ) . We have additionally examined the relationship between baseline anemia and risk of malaria episodes , with no evidence of an association ( HR = 0 . 8; 95%CI = 0 . 5; 1 . 2; p = 0 . 239 ) .
Malaria is one of the main public health problems in several developing countries affecting especially children , a particularly vulnerable population with the highest morbidity and mortality burdens associated with this disease [1 , 34] . Malaria usually co-exists with other diseases and poor socioeconomic status , further impairing the development of the affected populations . Malnutrition is one of the most common and worrying conditions , impairing child development and the severity of other health conditions [35] . The concomitance of both conditions has been studied , but the mechanisms and clinical impacts of this association remain incompletely understood . In this study , we show that malaria episodes interferes with the nutritional status of children , with notable reduced linear growth velocity as well as impairment in other indices associated with chronic malnutrition , even after adjustment for other factors . Suffering from one or multiple episodes of malaria had a significant negative effect on the linear growth velocity of children , especially amongst children between 5 and 10 years of age ( aOR = 4 . 0; 95%CI: 1 . 4; 11 . 4 , p = 0 . 008 ) . This important finding suggests that even in areas where malaria transmission is low and P . vivax is the most prevalent species , the infection may influence the physical development of children . Lee et al . monitored a cohort of children in the Peruvian Amazon and observed that children with P . vivax malaria experienced delayed linear and weight growth [20] . Available reports indicating that malaria infection can trigger acute undernutrition [36 , 37] and the effect of reducing malaria transmission on the improvement of undernutrition among children [38–40] are in agreement with our findings . We did not find an association for many of the investigated relationships , in particular for children below the age of 5 . ( a particularly vulnerable group for physical and developmental changes ) due to limitations related to the low number of events during the study period . A trend of inadequate classification of WAZ and BMI-Z for children between 5 and 10 years of age presenting malaria was not observed , what could be related to specific developmental pathways being affected within this age range . Although we found an association between undernourishment measured by HAZ and the risk of developing malaria , one needs to be careful at implying causality as other unmeasured confounders could be associated , including , for example , improved malaria preventive measures for ill children . We also did not observe a protective effect between baseline anemia and the risk of developing malaria as has been suggested elsewhere [41] . Due to the absence of severe episodes , we were not able to investigate the association of this condition with the anthropometric measurements . Some studies investigating the pathophysiologic mechanisms suggest that in addition to the anorexia and vomiting caused by acute-phase malaria and the negative nitrogen balance during fever episodes [7] , a lack of micronutrients , such as vitamin A and zinc , is a mechanism that can explain the effect of malaria on the nutritional status of infected children , particularly children younger than five years of age [42 , 43] . After several malaria episodes , a delay in the physical development of these children may occur . Our findings provide further evidence from a specific endemic setting . There were some important limitations in our study . Overall there was a low rate of malaria infection , especially amongst children in the lowest age group , which resulted in low power to examine the association of the infection with nutritional status and may , at least partially , explain the lack of effect observed amongst this particular group . Combined with the possibility of non-differential measurement error of the anthropometric indices , we assume the risk of bias towards a null effect of the association . No assessment of dietary patterns was performed , and although there is evidence of homogeneous habits within the study population , there may be important individual variations that would have a strong impact on the nutritional status of children . The lack of stool examination on the end of follow-up after the decision to treat all patients at inclusion has impaired the capacity to evaluate and control for the important influence that helminths have on the nutritional status of children [21 , 22] , especially in rural areas , what must undoubtedly be done in future studies examining this association . There would have been beneficial to the understanding of the causal mechanisms involved if we had systematically examined the children for asymptomatic infection , which could have had an important effect on our assessments , although this is assumed to be a rare finding amongst children of moderate transmission settings . In summary , we found malaria to be associated with the impairment of nutritional status in children , with evidence pointing to specific effects for different age groups , as indicated by anthropometric indexes associated with chronic ( BMI , HAZ ) and acute ( linear growth velocity ) nutritional deficits . We were not able to establish a clear association between baseline anthropometric classification and the risk of developing malaria . Bigger cohorts may be necessary especially in areas of low malaria incidence , and future studies should focus on specific nutritional deficiencies that may influence Plasmodium infections . Undernutrition and malaria are important morbidities with relevance to public health; thus , coordinating the actions of malaria control programs and nutrition programs could substantially impact these morbidities among children , meriting further attention by researchers and policy makers . | Malaria is one of the most serious public health problems in the world , with 3 . 3 billion people at risk of contracting the disease and almost one million deaths annually , primarily in children younger than five years of age . Undernutrition is also a morbidity of importance to the public worldwide and primarily affects children in tropical regions . In the present study , we sought to analyze the relationship between malaria and undernutrition in children living in a rural Amazonian community where malaria is endemic . Children from 1 month to 14 years of age were followed-up for 12 months through passive case detection ( i . e . , the presence of malarial parasites in peripheral blood in case of fever ) . Anthropometric and hemoglobin measurements and active malaria case detection tests ( to detect asymptomatic parasitaemia ) were conducted at the beginning and end of the follow-up period of 12 months . Children who had at least one episode of malaria during this period presented lower mean anthropometric index scores compared to children who did not have malaria . We concluded that malaria had a negative impact on the nutritional status of children living in an endemic area at the Western Brazilian Amazon , where Plasmodium vivax predominates . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | The Association between Nutritional Status and Malaria in Children from a Rural Community in the Amazonian Region: A Longitudinal Study |
Contact tracing is one of the key response activities necessary for halting Ebola Virus Disease ( EVD ) transmission . Key elements of contact tracing include identification of persons who have been in contact with confirmed EVD cases and careful monitoring for EVD symptoms , but the details of implementation likely influence their effectiveness . In November 2015 , several months after a major Ebola outbreak was controlled in Liberia , three members of a family were confirmed positive for EVD in the Duport Road area of Monrovia . The cluster provided an opportunity to implement and evaluate modified approaches to contact tracing . The approaches employed for improved contact tracing included classification and risk-based management of identified contacts ( including facility based isolation of some high risk contacts , provision of support to persons being monitored , and school-based surveillance for some persons with potential exposure but not listed as contacts ) , use of phone records to help locate missing contacts , and modifications to data management tools . We recorded details about the implementation of these approaches , report the overall outcomes of the contact tracing efforts and the challenges encountered , and provide recommendations for management of future outbreaks . 165 contacts were identified ( with over 150 identified within 48 hours of confirmation of the EVD cases ) and all initially missing contacts were located . Contacts were closely monitored and promptly tested if symptomatic; no contacts developed disease . Encountered challenges related to knowledge gaps among contact tracing staff , data management , and coordination of contact tracing activities with efforts to offer Ebola vaccine . The Duport Road EVD cluster was promptly controlled . Missing contacts were effectively identified , and identified contacts were effectively monitored and rapidly tested . There is a persistent risk of EVD reemergence in Liberia; the experience controlling each cluster can help inform future Ebola control efforts in Liberia and elsewhere .
The largest outbreak of Ebola Virus Disease ( EVD ) on record began in Guinea in 2013 and spread to Liberia by March 2014 [1 , 2] . With the support of partner agencies and organizations , the Liberian Ministry of Health ( MOH ) directed the implementation of multiple interventions that led to control of the initial epidemic [3] and two subsequent EVD clusters [4 , 5] . After a previous declaration several months before [6] , Liberia was declared free of EVD transmission for a second time on September 3 , 2015 [7] . Contact tracing , a cornerstone intervention to halt transmission of infection , is the process of identifying , assessing , and monitoring people who may have been exposed to a disease to prevent onward transmission [8 , 9] . The recommended practice to control Ebola outbreaks is to identify contacts of confirmed EVD cases and systematically monitor them twice daily for 21 days from their most recent potential exposure to an infectious case [9] . This allows for the rapid identification of people who become symptomatic and facilitates early isolation and treatment to prevent further transmission [10 , 11] . While contact tracing was critical to control of the main Ebola outbreak and subsequent clusters , several challenges were encountered . These included difficulty locating contacts , difficulty with contacts completing 21 days of monitoring and unwillingness of symptomatic contacts to attend an Ebola Treatment Unit ( ETU ) be tested for Ebola among others . Following the second declaration of no active EVD transmission in Liberia , the country maintained heightened EVD surveillance [10] . On November 19 , 2015 , a 15-year-old boy with symptoms compatible with Ebola was seen at a health care facility in Monrovia , in Montserrado County , Liberia , resulting in an alert to public health authorities . He was isolated and subsequently confirmed to have EVD [12] . The Liberia Incident Management System ( IMS ) was immediately activated to respond . Since the index patient was from the Duport Road area of Monrovia , the cluster was referred to as the “Duport Road Cluster” . The confirmed case and family members residing in the same household were transferred to an ETU . Two of these family members were confirmed to have EVD . The response team immediately initiated identification and monitoring of contacts , incorporating adaptations to previous approaches that aimed to improve the completeness and effectiveness of these activities . As this was the first cluster response in Liberia to incorporate administration of Ebola vaccination to identified contacts and contacts of contacts , procedures for monitoring vaccinated contacts were developed . On March 29 2016 , WHO declared the Ebola Public Health Emergency of International Concern ( PHEIC ) over but recognized new clusters due to reemergence had occurred and are likely to continue to occur . Thus , countries must maintain the capacity and readiness to prevent , detect , and respond to any new cases or clusters [13] . We describe the approaches to contact tracing during the response to the Duport Road Cluster , and outcomes of these activities , to inform future Ebola control efforts .
Active case finding teams conducted interviews with families , neighbors , employers and co-workers of cases to locate missing contacts . For contacts that had left Montserrado County , the County Health Teams in the respective areas worked with the Montserrado County Team to assist in finding the missing contacts through the established surveillance system . Reluctance to provide locating information and unwillingness of some contacts to be monitored led the MOH to subpoena mobile phone companies in order to use phone records to track down missing contacts . The teams used the phone records to determine the communities and locations where previous calls were made or text messages sent and received , and conducted house-to-house searches around the various communities to find missing contacts . The MOH arranged return transportation for contacts who left Montserrado County before tracing efforts began . The ZSOs reviewed the contact monitoring forms and used them to complete a daily summary that was reported to the DSOs . The DSOs , in turn , used this information to compile a district summary form and report to the county data manager each evening . A contact tracing feedback session was held at 6pm daily for all DSOs , ZSOs , and contact tracers as needed . An aggregated summary table , including tracers , contacts , and contact status ( number monitored/not monitored , lost to follow up , or missing ) was created and updated during the nightly feedback session together with the contact tracing database . Daily descriptive analysis was conducted on contact tracing activities and these data , in conjunction with case data , were presented during the IMS meeting the following morning ( Fig 1 ) . A dynamic contact tracing dashboard was created to enable tracking of contacts over time by household , using data collected from the field . Contacts were grouped by household , with the head of household’s name listed on the dashboard . The total number of contacts , contacts by risk status , and overall risk status of the household were displayed . The dashboard provided the name of the contact tracer , supervising DSO , district , and zone and allowed for visualization of the 21-day follow-up period of all contacts . Color coding indicated the last date of possible exposure from which the 21-day follow up period started , the date contact tracing was initiated , successful daily follow-up , dates contacts were not seen , dates any contacts were symptomatic , and the last date of contact tracing . A moveable arrow bar indicated the date/day of follow-up ( Fig 2 ) . Lastly , a master list of all contacts was maintained in Microsoft Excel , and was updated and resaved with new corresponding data each evening .
At the start of the monitoring period , 29 contacts were missing; however , all were successfully located and classified using the methods described above . One of the missing contacts travelled to Rivercess County prior to initiation of contact tracing . MCHT informed Rivercess County Health Team of the contact . A general Community Health Volunteer ( gCHV ) found this missing contact when she gave birth at a local clinic . The contact and her newborn subsequently returned to Montserrado County and were monitored twice daily at home . All other missing contacts were located within Monrovia . The new dynamic contact tracing dashboard was developed mid-response so teams were unable to collect full household data for each contact , which would have been important had any contact become a case . Competing priorities created difficulties for effective data management . The county data team was maintaining day-to-day responsibilities as well as those required for outbreak response without increasing human resources or training . This led to an over-reliance on partners for data management support , as well as unsustainable working hours leading to fatigue of the team and potentially impacting the quality control of data . The provision of support to quarantined contacts required robust information sharing between the contact tracing teams and the partners providing this support . In the initial stages of the response , this information sharing was incomplete , partly due to a lack of clear terms of reference for each response pillar , resulting in incomplete delivery of food and other support items/services . The adapted data management procedures eventually supported good information sharing and effective support of persons in quarantine .
From experiences and lessons learned through the contact tracing activities for the Duport Road EVD outbreak , we recommend the following activities be implemented ( Table 3 ) . First , needed personnel ( regular and surge staff ) should be clearly identified and included in the county level epidemic preparedness and response ( EPR ) plan . All of these staff should receive regular refresher training , including knowing acceptable temperature ranges for contacts under monitoring , proper use of Thermoflash thermometers , correct daily recording of monitoring information , reporting flow , and protocols . Refresher training should include simulation exercises with mentorship from experienced contact tracers . These trainings should address any fears and concerns among contact tracers so they know how to protect themselves while conducting monitoring activities . The county health teams should maintain a register of trained contact tracers with the date of the most recent refresher training completed so they can be used for future responses to EVD or other epidemic-prone diseases . Appropriate training and supervision of case investigators is also required to ensure only true contacts are listed and monitored . Second , we recommend development of contact tracing procedures and job aides for field use and data management . These should be clear and concise documents that contact tracers and supervisors can take into the field with them for guidance . These procedures should include specific information about sharing of information between contact tracing teams and teams conducting vaccination activities . The data management procedures should explain best practices in contact tracing data management and detail how to use all the tools developed as a result of this outbreak , including the contact tracing data management dashboard . The third recommendation is the development of a package for contact tracers that consists of all the necessary supplies and can be provided at the beginning of an outbreak to reduce the delay in deployment . When combined with regular refresher trainings , and a comprehensive set of guidance documents , this will enable rapid deployment of contact tracers . The fourth recommendation is to ensure that procedures for data sharing are in place before responses are needed . These procedures should include development of clear data sharing agreements at county level and the national level , as well as between the county heath team and supporting partners . A pre-existing data sharing agreement will allow for immediate collaboration between those responding to an event of public health concern . We also recommend the development of a secure data sharing platform to protect confidential patient information . Procedures for sharing information with the press should also be developed . It may be useful to establish regulations and/or implement training about privacy for journalists . The final recommendation is to focus on the improvement of communication and coordination to ensure that all teams are aware of the needs on the ground so they can respond accordingly , and to mitigate logistical challenges as they arise . As well as daily IMS meetings there is a need for clear terms of reference for each response pillar together with an organization chart stating lead persons and their contact details for each sector , this would need to be developed at the beginning of a response . Past outbreaks have demonstrated the need for strong coordination of partners and the engagement of community leaders to end EVD transmission [17] . The Duport Road response offered an opportunity to improve contact tracing methods employed in response to Ebola clusters in Liberia . The likelihood of EVD reemergence is high [3 , 5 , 17] and EVD remains a threat to the region . Therefore , prompt identification and monitoring of contacts remains one of the key actions necessary for ending the transmission of EVD . | Contact tracing is one of the key response actions necessary for controlling spread of Ebola Virus Disease ( EVD ) . Contact tracing is comprised of several different activities: identification of persons who have been in contact with confirmed EVD cases , close monitoring contacts for EVD symptoms , and management of symptomatic persons . Closely monitoring contacts of confirmed EVD cases allows for the rapid identification of symptomatic individuals , which in turn facilitates early testing , medical intervention , and isolation of new cases . This reduces the possibility of the continued spread of the virus within communities . Delayed and ineffective contact tracing contributed to the extensive transmission of EVD during the 2014–2015 outbreak in West Africa . Clusters of EVD reemergence are likely to occur , therefore understanding and addressing the challenges of implementing and managing contact tracing remains essential to halting transmission and minimizing morbidity and mortality associated with EVD . This paper assessed the contact tracing activities in response to EVD reemergence to identify best practices for responses to future Ebola clusters . The work is also applicable to contact tracing for other infectious diseases . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"education",
"immunology",
"geographical",
"locations",
"sociology",
"social",
"sciences",
"quarantines",
"tropical",
"diseases",
"ebola",
"hemorrhagic",
"fever",
"pediatrics",
"preventive",
"medicine",
"data",
"management",
"negle... | 2017 | Ebola virus disease contact tracing activities, lessons learned and best practices during the Duport Road outbreak in Monrovia, Liberia, November 2015 |
Simple surgical intervention advocated by the World Health Organization can alleviate trachomatous trichiasis ( TT ) and prevent subsequent blindness . A large backlog of TT cases remain unidentified and untreated . To increase identification and referral of TT cases , a novel approach using standard screening questions , a card , and simple training for Community Treatment Assistants ( CTAs ) to use during Mass Drug Administration ( MDA ) was developed and evaluated in Kongwa District , a trachoma-endemic area of central Tanzania . A community randomized trial was conducted in 36 communities during MDA . CTAs in intervention villages received an additional half-day of training and a TT screening card in addition to the training received by CTAs in villages assigned to usual care . All MDA participants 15 years and older were screened for TT , and senior TT graders confirmed case status by evaluating all screened-positive cases . A random sample of those screened negative for TT and those who did not present at MDA were also evaluated by the master graders . Intervention CTAs identified 5 . 6 times as many cases ( n = 50 ) as those assigned to usual care ( n = 9 , p < 0 . 05 ) . While specificity was above 90% for both groups , the sensitivity for the novel screening tool was 31 . 2% compared to 5 . 6% for the usual care group ( p < 0 . 05 ) . CTAs appear to be viable resources for the identification of TT cases . Additional training and use of a TT screening card significantly increased the ability of CTAs to recognize and refer TT cases during MDA; however , further efforts are needed to improve case detection and reduce the number of false positive cases .
Trachoma , the leading cause of infectious blindness , affects an estimated 84 million individuals worldwide [1] . Although eliminated in Europe and the United States , trachoma persists in much of the developing world , disproportionately affecting the poorest and most vulnerable populations [2] . Chronic trachomatous inflammation typically begins during childhood , with years of repeated infections with Chlamydia trachomatis leading to the development of conjunctival scarring in a portion of individuals [3–5] . As scarring worsens , an inturning of the lid margin , or entropion , brings the lashes into contact with the cornea . This trachomatous trichiasis ( TT ) , as well as the scarred conjunctiva , leads to damage and opacification of the cornea , ultimately resulting in blindness [6] . To control trachoma and mitigate potentially blinding TT , the Alliance for Global Elimination of Trachoma by 2020 ( GET 2020 ) and World Health Organization ( WHO ) advocate the multifaceted SAFE strategy , an approach consisting several components: Surgery to prevent blindness from TT , Antibiotics to reduce the infectious burden in children , and Facial cleanliness along with Environmental improvement to reduce transmission of C . trachomatis [7] . As part of the “A” component , WHO recommends mass drug administration ( MDA ) using oral azithromycin with high coverage , at least 80% of the population in communities where the prevalence of follicular trachoma is >10% [8–10] . Surgical management of TT involves a tarsal rotation procedure , which can be performed by a trained surgical technician within the community [11 , 12] . However , despite the relative effectiveness of surgery , coverage and uptake are often low due to poor acceptance and awareness of available surgical intervention [13] . Village-based promotional efforts and awareness campaigns have shown success in improving surgical uptake , although identifying TT cases in the community and linking them to services still poses a major challenge to meeting the GET 2020 surgical goal of reducing untreated TT to less than 1 case per 1 , 000 population at the district level [14–16] . We considered that Community Treatment Assistants ( CTAs ) providing MDA have the potential to reach all members of a community and may provide a platform for the identification of TT cases . This community randomized trial sought to develop and test a novel community-based screening approach designed to enhance the ability of CTAs to recognize and report TT cases during MDA . We developed a set of standard screening questions , a TT screening and recognition card , along with additional training . In the randomly selected intervention villages , the approach was introduced immediately prior to MDA , during training of CTAs . Control communities had standard training only . To evaluate the new approach , we compared the number of cases found , as well as the sensitivity and specificity , between the communities with enhanced training group of CTAs and the communities with the control group of CTAs .
A controlled community randomized trial design was implemented , using two parallel arms to compare the novel TT screening tool ( intervention ) to usual care ( control ) during mass drug administration . The trial was conducted in 36 geographically distinct communities in Kongwa district , Tanzania , between January and November of 2013 . Communities were selected based on eligibility to receive antibiotic treatment through MDA . The community served as the unit of randomization , and enrolled villages were randomly allocated on a 1:1 basis to either intervention ( n = 18 ) or usual care ( n = 18 ) arms . This study was built on to the baseline study of another project , which enrolled 52 communities in Kongwa Tanzania , of whom 36 were to receive MDA . Communities were eligible for this study if they were eligible to receive MDA , and had the consent of village leadership . Within the 36 study villages , individual members of the community were eligible for TT screening if they were 15 years of age or older . In each community , 2–10 CTAs were chosen to distribute antibiotic treatment . CTAs were selected by local leadership based on standing in the community and basic reading skills . Supervisory staff conducted training sessions for CTAs prior to MDA , including how to measure heights of children for antibiotic dosage , to mix liquid doses , to observe treatment provided , to log and manage medication , and to ask about TT and refer self-identified cases of trichiasis for surgery at the local hospital . In communities assigned to the intervention arm , CTAs received an expanded training session on trichiasis case recognition in addition to the standard instruction on MDA protocol . This additional training lasted one afternoon and consisted of both didactic learning and role-play designed to teach a basic understanding of trichiasis and focusing on four key components: In addition to this training , CTAs received a TT screening card containing simple questions on symptoms in Swahili , instructions on performing a TT exam , and images depicting proper TT examination and examples of TT cases for reference ( Fig 1 ) . Interview questions were designed to identify potential TT cases by asking if an individual had TT ( “kope”in Swahili ) or had epilated their lashes or had common ocular symptoms of TT , including foreign body sensation and tearing . If a participant said they had TT or epilated , or responded positively to at least one of the three questions , the CTA was directed by the TT screening card to proceed with eyelid examination using a Maglite Solitaire torch as described on the card . Positive cases of TT were defined as individuals who either had one or more lashes touching the globe or epilated . Any suspected TT cases were noted on the treatment log for subsequent verification , at which time they would be provided an opportunity for surgery if indicated . In communities assigned to the usual care arm , CTAs received the normal half-day of training on MDA procedures and a half hour basic overview of trichiasis and TT case recognition . CTAs were directed to ask during MDA whether individuals self-identified as having TT , and to record suspected cases of TT on the treatment log for each person . CTAs in usual care communities were given Maglite Solitaire torches identical to those used by CTAs in intervention villages but no other instructions . Following completion of MDA in each village , a survey was conducted by an experienced trichiasis grader , the “gold standard . ” He was masked to TT grade assigned during MDA screening . We also attempted to mask him to the village intervention assignment , but once in the village it was difficult to maintain that masking . The follow-up survey was designed to: For the survey , all TT cases reported by CTAs were re-examined to confirm their status . In addition , treatment records were used to randomly select a sample of 100 adults per community initially screened as negative for TT to determine the rate of cases missed by the CTAs during MDA . A random sample of 50 adults per community who did not attend MDA was also selected to estimate the number of TT cases missed because of absence from the MDA . To conduct the follow-up surveys , invitations were sent to selected individuals following MDA . Invitations notified individuals of the time , date , and location within the community of the survey and requested their attendance . On the day of the survey , the master TT grader examined all presenting individuals for trichiasis . For those found to have confirmed TT , the option of surgical treatment was explained , and interested individuals were added to a surgical referral list for the upcoming surgical camps to be held after the MDA program had concluded . After the initial day of survey , the TT grader continued follow-up on an individual basis until at least 70% participation was achieved for those who had attended MDA . CTA screening results obtained during MDA were recorded on paper forms in the mass treatment books . These forms were entered into a customized Access database by staff with double data entry performed for fields of particular importance . All ocular examinations for TT were entered into Samsung Tab 2 . 0 tablets using a customized ODK electronic form . The ocular exam form included identification information , personal details , and TT status . All electronic data was uploaded into a customized Access database and transferred electronically to Johns Hopkins University for analysis . All statistical analysis was conducted using SAS , and Stata version 12 [17–19] . To assess comparability between study arms , differences in demographic and socioeconomic characteristics at the community level were evaluated using Wilcoxson rank sum tests . Additionally , characteristics of participants who were re-examined were compared to those lost to follow up within each arm , using chi-square tests or Fisher’s exact test as appropriate . Sensitivity , specificity , positive predictive value and negative predictive value were estimated for the intervention and usual care arms using a 2 × 2 table comparing the extrapolated results from the samples verified from the Master Grader with those obtained by the CTAs during MDA , as depicted below in Table 1 . Values used in these calculations as well as in the estimation of disease prevalence were obtained using a simple extrapolation method applying the findings from the random samples in the survey to the study population obtained from the full census of the community . This extrapolation was performed as follows: A simple extrapolation was deemed appropriate after review of the samples’ representativeness of the respective study population based on age and gender indicators . This method of extrapolation allowed for the comparison of confirmed cases to the estimated total missed cases in both arms , accounting for the rescreening of all individuals initially screened positive and only a sample of individuals initially screened negative . Oral informed consent was obtained for eligible adults prior to screening . All study protocols and procedures were approved by the Johns Hopkins Institutional Review Board and the National Institute for Medical Research in Tanzania . This randomized trial is registered in the ClinicalTrials . gov database under identifier NCT01783743 .
A total of 27 , 473 individuals in the 36 villages holding MDA were eligible for this study , of whom 19 , 607 ( 71 . 4% ) attended MDA and were screened for trichiasis . Fig 2 shows the randomization scheme of communities and the flow of participants through the study . Baseline characteristics of these villages were compared between arms ( Table 2 ) , with no significant differences found except a slightly higher proportion of household with bicycles in the usual care arm ( median 45% versus 43% in the intervention arm ) . MDA achieved similar overall coverage in both arms ( median 76% versus 72% , p = 0 . 56 ) . The master trichiasis grader re-examined at follow up 3 , 233 ( 81 . 6% ) of the 3 , 962 individuals who were initially screened during MDA and who were chosen for follow-up . Additionally , 1 , 695 individuals who did not attend MDA were randomly selected for follow-up to estimate the number of TT cases missed during MDA . Of these 1695 persons , 1 , 149 ( 67 . 8% ) were examined . Loss to follow-up was attributed primarily to refusal or being away from the home , as many residents travel long distances to farmland during the planting and harvest seasons . Table 3 compares characteristics of individuals who were examined by the master TT grader to those who were lost to follow up after MDA . Among the sample who were selected for follow up as screened negative for TT , older males were more likely to be lost to follow up compared to those who were examined . Otherwise , the groups appeared comparable . After conclusion of MDA , CTAs in the intervention arm reported a significantly higher number of suspected cases ( n = 409 , ( 4 . 3% ) ) than those assigned to usual care ( n = 46 , ( 0 . 5% ) , p < 0 . 01 ) ( Table 4 ) . The proportion of cases confirmed by the master grader as correctly identified was higher in the usual care group ( 28 . 1% ) compared to the intervention group ( 15 . 8% ) , although this difference did not reach statistical significance ( p = 0 . 070 ) . Among those that we could verify , CTAs in the intervention communities identified 5 . 9 times as many true cases as did those in the control arm ( n = 50 ( 0 . 525% ) and 9 ( 0 . 089% ) , respectively ) , but also had almost ten times as many false positives ( n = 316 vs . 32 ) . Within the random sample of individuals screened negative by CTAs , the proportions of missed TT cases were not significantly different between the two arms ( p = 0 . 269 ) , although fewer cases were missed in the intervention group ( 1 . 6% ) compared to usual care ( 2 . 1% ) . The sensitivity , specificity , and positive predictive value ( PPV ) for both groups were calculated using extrapolated values from the follow-up survey ( Table 5 ) . Sensitivity of the additional training and use of the TT screening card ( 31 . 2% ) was 5 . 6 times higher than without the extra training and card ( 5 . 6% , p < 0 . 05 ) , indicating that the intervention significantly increased the ability of CTAs to identify TT cases . PPV of this screening method did not statistically differ from that of usual care , although both were low ( 15 . 8% and 28 . 1% , respectively ) . The specificity in the usual care arm was higher ( 99 . 7% ) than that of the intervention ( 96 . 3% , p < 0 . 05 ) , although specificity for both groups was high . Based on case finding and the estimates derived from the random samples , the prevalence of TT in the study communities in those aged 15 and older was estimated to be 2 . 2% in both arms . Of the random sample of individuals who did not present to MDA , 14 ( 1 . 2% ) of the 1 , 149 individuals examined had TT . Six months after the follow-up survey , we re-contacted 19 of the 23 TT cases who were screened negative by the intervention CTAs to determine possible reasons for being incorrectly screened during MDA . Four persons were quite elderly and had no recollection of attending MDA or being screened despite mass treatment logs noting their attendance and completion of screening . Of the remaining 15 individuals , 5 ( 33% ) stated that they had not been asked screening questions , 5 ( 33% ) said they had been asked the questions but had not been examined , and 5 ( 33% ) said they had received both questions and examination . When these 15 individuals were asked again the questions using the TT screening card , those who had initially been screened out based on the interview questions ( n = 5 ) answered the questions negatively again despite having visible TT . However , those who had not been asked questions during MDA ( n = 5 ) and those who had been both interviewed and examined ( n = 5 ) answered positively to at least one interview question .
This study demonstrates that an expanded but simple training program and use of a TT screening card improves the ability of CTAs to identify cases of TT during MDA . Using this approach , CTAs identified over five times more TT cases than did those assigned to usual care . Additionally , training and use of the card required minimal additional resources , making this intervention an easily implementable approach to identify TT cases in MDA communities . Although this was a substantial improvement , sensitivity for CTA screening was still lower than expected . As the goal of TT screening is to find those individuals with TT who are currently in need of treatment , a higher sensitivity is necessary to reliably identify as many true positive cases as possible . While the scripted questions were designed to be specific for TT , many individuals answering yes did not have TT . In another study , 43% of false positive cases found by CTAs in fact had other eye pathology , including corneal disease and cataract [20] . From a primary eye care perspective , a high number of false positives may identify other eye conditions that require intervention . Therefore , having an eye care professional undertake further verification would not only help identify TT cases , but could also direct other patients into the eye care system . During a second follow-up of the false negative cases , one-third again reported not epilating or having symptoms of TT . This may indicate a reluctance of individuals to screen positive for fear of surgery , which has been commonly cited as an issue in TT surgical uptake [13 , 21–22] . It may also reflect absence of symptoms in some TT cases . Severity data was not collected , so perhaps these individuals had , for example , only 1–2 asymptomatic lashes . Another one-third of the false negatives remember being asked the TT screening questions and examined at MDA; however , they were not correctly identified as having TT by CTA examination . In these instances , the screening questions appear to have proven sufficiently sensitive , and at least three explanations are possible . Perhaps training was insufficient and CTAs failed to recognize TT . Alternatively , it was noted by the program coordinator during MDA that some older CTAs had difficulty seeing at close distances due to presbyopia , which made it difficult if not impossible for them to identify inturned lashes touching the globe . While it would be cost prohibitive to provide loupes for all the CTAs , choosing TT screeners with good near vision should improve detection . Finally , the TT screening card did not stress that epilation could also count as TT and may have led to the false negative cases . While the original intention was to refer patients who epilated or who had lashes touching , if there were no lashes touching the globe due to epilation , the CTAs may not have included the patient in the referral list . The final one-third of the missed TT cases claimed that they were not asked screening questions at all during MDA , suggesting a failure on the part of some CTAs to conduct the TT screening during treatment . All of these individuals responded positively to one or more of the TT screening card questions during their follow-up , raising the possibility that a lack of CTA adherence to the screening protocol may have posed an issue . During field visits , the program coordinator observed that when the MDA became particularly busy , CTAs did not ask questions properly . As CTAs typically work together in pairs during MDA , this could be improved by training CTAs to supervise one another . In fact , as MDA can continue for several years , missed TT cases could be detected in subsequent MDAs . Both the CTAs in the intervention group and in the control group demonstrated high specificity , which was expected in screening for a relatively rare disease in a population . Maintaining a high specificity is important for this type of screening as correctly identifying individuals without TT reduces the verification need , although as noted , many of these cases may have other eye diseases . The low PPV for TT can be largely explained by the low prevalence of TT . PPVs are typically low in the case of rare diseases for even the most sensitive and specific of tests , as false positives may outnumber true positives [23 , 24] . Even if expected , a low PPV has significant implications from a programmatic standpoint concerning the planning of surgical services for referred TT cases . Surgical camps , which are commonly implemented in rural and isolated areas , involve a great deal of preparation and cannot be planned for hundreds of cases if only 15% of them actually require surgery [25] . Using our approach , case verification would likely be needed before planning surgical services . CTAs represent a potential team of screeners already present within endemic communities where TT cases are likely located . Their extensive exposure to the village members during MDA presents a great opportunity for the identification of TT cases . Taking advantage of this exposure for screening comes at only a marginal cost to the MDA itself , requiring only a day of training and the provision of the TT identification card and a torch . As members of the community , CTAs also possess a deep understanding of the local culture and language , which can be beneficial , particularly when cases are to be referred for surgery . Additionally , charging local leadership with the task of determining recruitment criteria and selection of their own CTAs , as has been implemented in Kongwa , promotes a sense of community buy-in and partial ownership of the trachoma and trichiasis control efforts [26] . Such community-based approaches to disease detection and control also allow for the integration of trichiasis screening into other existing programs . Community-level education programs targeting the Guinea worm have already incorporated mass drug distribution for trachoma control in areas such as South Sudan[27] . Partnerships such as this present opportunity to greatly expand the reach of TT screening while pooling resources to further reduce implementation costs and improve health beyond TT and eye disease . While use of CTAs in TT screening during MDA offers clear advantages , this approach is limited to areas that merit MDA . This is problematic in areas that have reduced the prevalence of active trachoma to below 5% because even after cases of active trachoma have been reduced , trichiasis cases will likely still exist . Attrition of CTAs may also prove to be a limitation for this screening approach , as significant numbers of community-directed distributors ( CDD ) in national onchocerciasis programs in Nigeria and South Sudan left after their first year due largely to a lack of incentives offered [27 , 28] . In our study , both usual care and intervention CTAs participating in MDA were paid a nominal fee for their service , which likely influences retention of these trained assistants in ongoing MDA conducted in Kongwa district . However , this may not be practical for programs that cannot afford to pay CTAs . In summary , though surgical management of trichiasis has been shown to be cost-effective , the identification of cases needing surgery remains challenging and often resource-intensive . Our study evaluated an innovative approach to TT screening that sought to take advantage of the efforts already being made by trachoma control programs—the use of MDA—as a stage for TT screening . Additional training and use of a TT screening card significantly increased the ability of community treatment assistants to recognize and refer TT cases during MDA; however , further efforts are needed to improve case detection and reduce the number of false positive cases . | Surgical management of trachomatous trichiasis ( TT ) is recommended by the WHO as a cost-effective strategy to mitigate blinding trachoma . However , a large surgical backlog exists and many individuals suffering with TT remain unknown to the health system . To identify TT cases , we designed a standard set of screening questions , a card , and simple training for Community Treatment Assistants ( CTAs ) to identify trichiasis during community-wide Mass Drug Administrations ( MDA ) . To evaluate the sensitivity , specificity , and positive predictive value of this approach , we conducted a community randomized trial in 36 communities in trachoma-endemic Kongwa District , Tanzania . Additional training and the use of a TT screening card increased the sensitivity of TT identification and resulted in more cases identified compared to the usual training of CTAs . The positive predictive value was low , indicating a need for further verification of TT cases identified by the enhanced screening . MDA appears to be a good opportunity for TT screening by CTAs , but further training to improve screening sensitivity is suggested . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Assessment of a Novel Approach to Identify Trichiasis Cases Using Community Treatment Assistants in Tanzania |
Silky-feather has been selected and fixed in some breeds due to its unique appearance . This phenotype is caused by a single recessive gene ( hookless , h ) . Here we map the silky-feather locus to chromosome 3 by linkage analysis and subsequently fine-map it to an 18 . 9 kb interval using the identical by descent ( IBD ) method . Further analysis reveals that a C to G transversion located upstream of the prenyl ( decaprenyl ) diphosphate synthase , subunit 2 ( PDSS2 ) gene is causing silky-feather . All silky-feather birds are homozygous for the G allele . The silky-feather mutation significantly decreases the expression of PDSS2 during feather development in vivo . Consistent with the regulatory effect , the C to G transversion is shown to remarkably reduce PDSS2 promoter activity in vitro . We report a new example of feather structure variation associated with a spontaneous mutation and provide new insight into the PDSS2 function .
The feather is one of the most complex integumentary appendages due to the extensive diversity in shape , size , arrangement and pigmentation , and is therefore an excellent model for evolutionary and developmental biology as variations can occur at each step of development and differentiation . The development of feathers , beginning in the embryo and continuing through several cycles of regeneration after hatching , has been one of the most challenging subjects in the field of avian morphology since the early 19th century [1] . Feathers are clustered in different tracts on the skin and the chicken has about 20 tracts [1] . Feathers from different tracts vary widely and some of the differences are due to different origins of the mesenchyme . The three main types of feathers are contour feathers ( pennaceous ) , down feathers ( plumulaceous ) and filoplumes [1] . Contour feathers are divided into flight feathers known as remiges and rectrices , and the ordinary body contour feathers . A typical contour feather is composed of the calamus , rachis , barbs and afterfeathers . The barbs have two sets of barbules which are hooked together by the hooklets . Feathers can develop into different types with part or all of the branching structures . Several feather types exist in the domesticated birds and these variations have been selectively bred . Variations in the structure , distribution , length , arrangement and number of feathers are widespread in chickens and pigeons [2] . Factors controlling feather morphogenesis have been studied by poultry geneticists , zoologists and embryologists [1] , but only a few spontaneous mutations affecting feather morphogenesis are known . Well-studied natural variations in chicken include the BMP12 gene that changes the distribution of feathers on the neck ( Naked neck trait ) [3] , the HOXC8 gene that changes the length of the cranial feathers ( Crest trait ) [4] , and the KRT75 gene causing a characteristic curled feather rachis and barbs ( Frizzle feather trait ) [5] . Variations in contour feather structure usually occur in the distal barbules part of the vane and disturb the interlocking barbules . The structure variations in domesticated birds have been summarized [2] , including silky-feathers , frizzle feathers , hypoplasia of tail feathers and henny-feathering , hard and soft feather texture in the chicken; lace-feathering , curled feathers and powder downs , and fat quills in the pigeon; and spiraled feathers in the goose . The silky-feather phenotype is a primary characteristic of the Silkie breed of chicken which is named for this phenotype . The Silkie was first mentioned by Marco Polo in his Asian travelogues in 1298 as “chickens with hair like cats that lay the best of eggs” [6] . Darwin [7] noted that the progeny from matings between Silkie and wild-type chickens did not show the silky-feather phenotype , and a recessive mode of inheritance was subsequently confirmed by Dunn in 1927 [8] . All chickens have similar downy feathers at hatching . The first molting is initiated a few weeks after hatching at which time down feathers are replaced with juvenile pennaceous feathers . In silky-feather chickens beginning with the first pennaceous feathers a clear difference in feather structure is seen as compared to wild-type , with the Silkie chicken maintaining a more downy appearance in the body contour feathers ( Figure 1A and 1B ) . In the closed pennaceous feather portion , wild-type feathers have hooklets on the distal barbules and forms the vane ( Figure 1C and 1E ) while silky-feathers lack hooklets ( Figure 1D and 1G ) . In the afterfeather portion , both wild-type and silky-feather have the barbules structure but no hooklets ( Figure 1F and 1H ) . The flight feathers and some of the shank feathers can form hooklets in Silkie birds , but fewer than in wild-type chicken ( Figure 1B ) [9] . This phenotype has apparently been strongly favored during the development of the Silkie breed due to the beautiful fur-like feathers [9] . The Silkie chickens lose the flight function and do poorly in extreme temperatures because of the lack of closed pennaceous vane . A few variations of feather vane structure are related with hooklets in birds . One variation is the chicken frizzle feather which curves backward and mainly alters rachis structure . The mature frizzle feather structure also exhibits other modifications such as thickening of the barbs and barbules , alteration of the hooklets and other structural abnormalities , however without losing any morphological component [5] , [10] . The frizzle feather locus is inherited in an autosomal incomplete dominant manner and caused by the KRT75 gene deletion mutation [5] , [10] . Another variation is the silky plumage in pigeons which was also noted by Darwin although he didn't know its inheritance [7] . The structure of the silky plumage ( also named lace-feathering locus ) in domestic pigeons ( e . g . Silky Fantail ) and Ring-neck Doves is completely different with the silky-feather in chicken [11] . The silky pigeons have hooklets on the barbules , and the hooklets are abnormally thickened [12] . Moreover , the barbules are weak and their elasticity is poor . The lace-feathering locus is controlled by an autosomal gene with incomplete dominance [11] , [12] . Some signaling molecules involved in feather morphogenesis have been studied , and some feather-branching morphogenesis models are proposed involving the expression of sonic hedgehog ( SHH ) , bone morphogenetic protein ( BMP ) , Noggin , etc . [13]–[18] . SHH is found to mediate the interaction between the epithelium and mesenchyme during feather development [13] . In contrast , BMPs are found to inhibit feather formation [14] . Harris et al . [16] , [17] suggested that , ( i ) the activator-inhibitor models of SHH and BMP2 signaling to explain the barb formation , and ( ii ) an integrated model of feather morphogenesis and evolution to describe the feather branching structures . SHH and BMP2 signaling constitutes a functionally conserved developmental signaling pathway during epidermal appendage development , and the interaction between SHH and BMP2 signaling in feather epithelium has been demonstrated to control the formation of barb ridges and barb variation [16] . Additional inhibitory signal and signal gradient are required for the hierarchical branched structures and feather form ( e . g . plumulaceous and pennaceous structures ) . Variation regulating the SHH and BMP2 may be crucial for the evolution of feather-branching morphogenesis [16] , [17] . Yu et al . [18] found that the antagonistic interaction between Noggin and BMP4 mediated feather branching , and SHH was required for the formation of barbs . Noggin promotes branching , and BMPs promote rachis formation and inhibit the barb formation . The balance between BMPs and SHH signaling modulate the number and size of the barbs and barb ridges . The interaction of these signals determines their number and fate of marginal plate cells and barbule plate cells , which leads to the feather variants [18] . Previous studies have shown that many signaling molecules are involved in feather morphogenesis , but little is known about the molecular mechanisms of hooklet development and differentiation [1] , [19] , [20] . In the present study , we show that the silky-feather phenotype is associated with a single regulatory SNP modulating the expression of the PDSS2 gene .
Our initial linkage analysis to map the silky-feather locus was based on a genome-wide set of 125 microsatellite markers previously used for mapping growth trait QTLs in the CAU resource population ( CAURP ) comprised of a Silkie x White Plymouth Rock intercross [21] . The silky-feather locus was mapped to the SF01–SF06 ( GGA3:69 . 28–71 . 63 Mb ) interval on chromosome 3 and showed tight linkage with the SF03 marker ( 70 . 41 Mb ) with a LOD score of 30 . 1 , consistent with the results in another resource population [22] . Refined linkage mapping was done using the Illumina Chicken 60K SNP Beadchip which resulted in the assignment of the silky-feather locus to a 380 kb interval between positions 70 , 201 , 106–70 , 581 , 126 bp ( Table 1 ) . This interval contains three genes , Sex comb on midleg-like 4 ( SCML4 ) , Sine oculis-binding protein ( SOBP , also named Jxc1 ) and Prenyl ( decaprenyl ) diphosphate synthase , subunit 2 ( PDSS2 ) . Refined linkage mapping in another population [22] identified a similar albeit slightly larger location of 70 , 399 , 176–70 , 988 , 264 bp ( about 589 kb ) with an overlap of 70 , 399 , 176–70 , 581 , 126 bp ( about 182 kb ) from the two populations combined . We refined the location of the silky-feather locus in the CAURP using the identical-by-descent ( IBD ) mapping method [23] . In the F0 generation the White Plymouth Rock is fixed for the wild-type allele and the Silkie chicken is fixed for the silky-feather allele . The White Plymouth Rock is not suspected of carrying any silky-feather haplotype as no introgression between these breeds has occurred to the best of our knowledge . The 38 SNP markers genotyped on 12 White Plymouth Rock and 19 Silkie birds were used for the initial IBD mapping . Of the 38 SNP markers , 20 covered the 380-kb ( GGA3:70 , 201 , 106–70 , 581 , 126 ) interval without recombination in the linkage analysis ( Table 1 ) . Homozygosity for the silky-feather birds was limited to two short haplotype blocks defined by three SNPs respectively: the proximal 57 . 4-kb ( 70 , 384 , 172–70 , 441 , 580 ) and the distal 56 . 7-kb ( 70 , 447 , 648–70 , 504 , 365 ) ( Figure S1 ) . One White Plymouth Rock bird and some wild-type F2 individuals were homozygous for the proximal Silkie shared haplotype , whereas only the silky-feather birds in Silkie and the F2 generation were homozygous for the distal Silkie shared haplotype . Therefore , the proximal region was excluded and the silky-feather shared haplotype was narrowed to the distal 56 . 7-kb interval ( 70 , 447 , 648–70 , 504 , 365 ) ( Figure S1 ) . We sequenced one Silkie chicken BAC and one Red Jungle Fowl BAC to identify all sequence polymorphisms associated with the silky-feather allele and identified 886 polymorphisms in the 70 , 349 , 337–70 , 487 , 932 bp interval . An estimated 891 bp gap at 70 , 470 , 161–70 , 471 , 051 bp in the chicken galGal3 genome assembly was re-sequenced successfully in the Red Jungle Fowl BAC ( from the reference chicken genome individual ) and was determined to be 626 bp ( GenBank KC166241 ) . We genotyped 34 SNP markers within the 70 , 441 , 580–70 , 484 , 959 bp interval in 12 breeds for further IBD mapping . This breed panel consisted of 76 samples from Silkie , Kuaida Silky and Lanping Silky chickens all with silky-feather , and 95 samples from nine breeds that have the wild-type phenotype . The upstream IBD boundary ( 70 , 460 , 738 bp ) was established based on two heterozygous Silkie and one homozygous Lanping Silky chickens ( Table S2 ) . A downstream IBD boundary was not established using this panel . The silky-feather haplotype defined in the 70 , 461 , 033–70 , 484 , 959 bp interval was fixed in all 76 silky-feather chickens ( except one heterozygous Silkie bird at 70 , 466 , 750 bp ) and was not found in the 95 wild-type chickens , indicating that the causative mutation must be located in the 70 , 460 , 739–70 , 504 , 365 bp region . Based on this SNP screening , we selected seven silky-feather birds and nine wild-type birds for re-sequencing . Additionally , three F1 heterozygous birds and one F2 silky-feather bird were included . The F2 bird carried an intact Silkie chromosome and a recombined chromosome within the 70 , 467 , 968–70 , 581 , 126 bp interval that was of Silkie proximal descent and of White Plymouth Rock distal descent . Interestingly , among the nine wild-type birds , one White Leghorn and one White Plymouth Rock had the identical haplotype with silky-feather at 70 , 472 , 921–70 , 484 , 959 bp ( Table S2 ) . We re-sequenced the 49 kb interval ( 70 , 460 , 286–70 , 509 , 023 bp ) using the Sanger method . Four ∼1 . 2 kb fragments located within the 70 , 515 , 823–70 , 547 , 070 bp interval were also re-sequenced . Re-sequencing from the eight silky-feather birds revealed an exclusively silky-feather shared haplotype at 70 , 468 , 129–70 , 487 , 067 bp ( Figure 2 ) . The upstream and downstream boundaries were identified separately using two different Silkie birds . The 18 . 9 kb silky-feather haplotype was fixed in eight silky-feather birds , and absent in the wild-type birds . Whole genome sequencing was used to further investigate the location of IBD haplotypes in Silkie chickens in the USA . DNA was pooled from 15 Silkie chickens and was sequenced on the SOLiD platform as described previously [24] . Within the 182 kb region defined from the combined linkage mapping studies a single haplotype was detected as fixed in the Silkies ( Figure S2 ) . This haplotype was 21 . 7 kb ( 70 , 467 , 293–70 , 489 , 020 bp ) and overlapped perfectly with the 18 . 9 kb haplotype identified in the Silkie chickens from China . The causative mutation should be located in this 18 . 9 kb region , and no structural change in this region was identified in the present study via Sanger sequencing or in our previous study using array CGH [25] . The sequence of the assembly gap in this region ( GenBank KC166241 ) contained 83 . 5% G and C with numerous poly G , poly C and CpG sites . One SNP and a 9-bp deletion were found within the gap sequence but none was associated with the silky-feather phenotype . Of the other 85 variations , only a C to G transversion at 70 , 486 , 623 bp ( ss666793747 ) was perfectly associated with silky-feather ( Figure 2 ) . The remaining 84 polymorphisms were excluded because at least one wild-type ( H/H ) bird showed the same homozygous genotype as silky-feather ( h/h ) birds . We genotyped ss666793747 in 718 birds from 33 populations and found that all 337 silky-feather chickens were homozygous G/G , 341 wild-type chickens were homozygous C/C and 40 known heterozygous birds were G/C ( Table 2 ) . These results lead to the hypothesis that ss666793747 is the causal mutation responsible for the silky-feather phenotype . The candidate mutation ss666793747 is located between the SOBP and PDSS2 genes , which are two adjacent genes separated by 16 . 7 kb . The ss666793747 mutation is 103 bp upstream of the initiator codon ATG of PDSS2 and 16 . 6 kb upstream of SOBP . Hereafter , we refer to the ss666793747 mutation as PDSS2 ( -103C-G ) . Chicken SOBP contains six predicted exons and five introns according to the current annotation of the chicken genome assembly . The coding sequence is well conserved among chicken , human , mouse and other vertebrate species . The human and mouse have a non-coding exon 7 . The chicken exon 7 was confirmed by 3′ RACE analysis and the mRNA has been submitted to GenBank with number KC166240 . Chicken PDSS2 contains eight exons and the transcript has been submitted to GenBank with number JX982522 . To determine whether PDSS2 ( -103C-G ) disrupted the transcription start site ( TSS ) and the coding sequence of PDSS2 , we performed 5′ RACE analysis using dorsal skin from three wild-type and three silky-feather homozygous birds . The most common start site at position −90 ( 70 , 486 , 636 bp ) downstream of PDSS2 ( -103C-G ) was identical in all the wild-type and silky-feather birds ( Figure S3 ) . Meanwhile , the PDSS2 ( -103C-G ) was found to be present in the 5′ UTR in a few clones ( Figure S3 ) . Determining the exact TSS of PDSS2 with 5′ RACE is technically challenging due to the high GC content in this region . Another possible case is that multiple TSSs are located within the small region . The translation start site of PDSS2 was confirmed and showed no difference between the two genotypes . Exon re-sequencing revealed four synonymous SNPs in SOBP , eight synonymous SNPs and a nonsynonymous SNP in PDSS2 , but none of them was specifically associated with silky-feather ( Tables S3 and S4 ) . Expression analysis from postnatal ( P ) days 60 H/H and h/h homozygotes by RT-PCR revealed that PDSS2 and SOBP were expressed in all of the tissues for both genotypes ( Figure S4 ) . Transcript analysis of PDSS2 and SOBP revealed no splice difference between H/H and h/h genotypes ( Figures S5 and S6 ) . We investigated the relative expression levels of PDSS2 and SOBP mRNA in dorsal skin tissue at P10 , P60 , P130 and P200 from two homozygotes . Quantitative RT-PCR analysis showed that the expression of PDSS2 was significantly reduced in silky-feather birds at all stages , but SOBP expression was only decreased at P60 ( Figure 3A and 3B ) . To explore whether PDSS2 ( -103C-G ) is a tissue-specific regulatory mutation , the expression analysis was further performed in liver tissue at these four stages . A similar pattern with reduced PDSS2 expression in silky-feather birds at all stages and lower SOBP expression only at P60 was observed ( Figure 3C and 3D ) , showing that the mutation does not have a tissue-specific effect . Furthermore , the expression of PDSS2 was decreased significantly in silky-feather birds at embryonic stages , whereas SOBP expression was increased at embryonic ( E ) days 9 ( Figure 3E and 3F ) . We further compared PDSS2 and SOBP expression at P130 in all three genotypes . The expression of PDSS2 in H/h skin was higher than in h/h skin , but showed no difference with H/H skin ( Figure 3A ) . The expression of SOBP in H/h birds was similar to H/H and h/h birds ( Figure 3B ) . An allelic expression imbalance ( AEI ) analysis demonstrated significantly higher expression of the H allele for PDSS2 in P130 heterozygous H/h birds while no such allelic imbalance was observed for SOBP ( Figure 4 ) . The result implied that PDSS2 ( -103C-G ) is a cis-acting regulatory mutation , consistent with the expression differences in Figure 3 . We also analyzed the expression in dorsal skin tissue of the other nine genes located in the broad 70 , 177 , 177–70 , 816 , 125 bp interval . No SCML4 transcript was detected . Five genes ( BEND3 , C6orf203 , QRSL1 , RTN4IP1 and ENSGALT00000037260 ) were differentially expressed between silky-feather and wild-type birds at P10 and two genes ( OSTM1 and QRSL1 ) at P60 ( Figure S7 ) . The expression of QRSL1 was increased significantly in silky-feather birds at P130 and P200 , while the expression of OSTM1 showed an opposite trend at P200 comparing with P60 ( Figure S7 ) . These inconsistent expression patterns in skin suggested that these are more likely to be secondary effects or due to regulatory mutations in linkage disequilibrium with PDSS2 ( -103C-G ) rather than long range regulatory effects from PDSS2 ( -103C-G ) . The mRNA expression analysis suggested that silky-feather was likely due to reduced expression of PDSS2 in the skin . An anti-chicken PDSS2 antibody was custom made and its specificity was confirmed by western blotting and immunofluorescence analysis ( Figure S8 ) . The immunofluorescence assay showed that chicken PDSS2 protein is mainly located in the cytoplasm , which is identical with human PDSS2 protein that is localized to cytoplasm according to the Human Protein Atlas ( www . proteinatlas . org/ENSG00000164494/tissue ) and mitochondrion according to the UniProt database ( www . uniprot . org/uniprot/Q86YH6 ) . We performed immunohistochemical staining with the chicken PDSS2 antibody in both wild-type and silky-feather embryos at day E9 , E11 , E14 and E17 ( Figure 5 ) . PDSS2 expression was observed in feather short bud ( Figure 5A and 5E ) , long bud ( Figure 5B and 5F ) , follicle ( Figure 5C and 5G ) and complete feather ( Figure 5D and 5H ) . The staining revealed strong PDSS2 expression in feather epithelium and weak expression in distal mesenchyme at E9 ( Figure 5A and 5E ) . At E11 when feather buds elongated , PDSS2 showed a greater intensity in feather epithelium/shaft ( Figure 5B and 5F ) . At E14 when feather follicle formed , PDSS2 was strongly expressed in the collar region of the follicle and sheath ( Figure 5C and 5G ) . At E17 when complete feather formed , striking PDSS2 expression was observed in the proximal follicle and feather sheath ( Figure 5D and 5H ) . The expression of PDSS2 became gradually intense in the skin epithelium ( Figure 5 ) . Importantly , the expression of PDSS2 in silky-feather embryos was decreased compared with wild-type embryos at the same stage , consistent with the results of the mRNA expression ( Figure 3E ) . In addition , the strong expression of PDSS2 in the epidermal cells is consistent with abundant mitochondria in the epidermis of embryonic chick skin [26] . We did a BLAST search for the surrounding sequence of PDSS2 ( -103C-G ) ( 30 bp upstream and 30 bp downstream ) , however , we did not find any significant sequence conservation among birds , reptiles and mammals . Considering the position of PDSS2 ( -103C-G ) , we decided to test the promoter effect in “forward” ( toward PDSS2 ) orientation and enhancer effect in “reverse” ( toward SOBP ) orientation in the pGL3 luciferase vector ( Figure 6A ) . We transfected chicken fibroblast cells ( DF1 ) in which PDSS2 and SOBP were expressed ( Figure S4A ) , and measured luciferase activity after 24 hours . Two different length fragments were used separately . Compared to the promoterless vector , the wild-type vectors increased luciferase activity ∼46-fold ( long , H-LF ) and ∼53-fold ( short , H-SF ) whereas the silky-feather vectors increased the activity only ∼8-fold ( long , h-LF ) and ∼9-fold ( short , h-SF ) ( Figure 6B ) . Thus , the two constructs had promoter activity and most importantly , the silky-feather construct decreased promoter activity compared to wild-type consistent with the observed differences between alleles in vivo ( Figure 3A , 3C and 3E ) . No enhancer activity was detected in reverse direction in wild-type or silky-feather ( Figure 6C ) . The Abd-B factor was predicted to bind to wild-type promoter with score 86 . 6 ( transcriptional factor search , TF2 SEARCH ) but not the silky-feather allele [27] , which might be responsible for the difference in promoter activity .
Silky-feather is a recessive Mendelian trait found in only a few chicken breeds that dramatically alters the structure and appearance of the juvenile and adult chicken feather . We herein show that a single-base change upstream of PDSS2 is responsible for silky-feather in chicken based on high-resolution mapping and expression analysis . We mapped the silky-feather locus to a 182 kb region by linkage analysis using two separate mapping populations and then fine mapped it to an 18 . 9 kb region using IBD mapping . The identification of an overlapping 21 . 7 kb IBD region in a second population of Silkie chickens is further indication that this region harbours the causal mutation . Within the minimal haplotype region , SNP PDSS2 ( -103C-G ) was completely associated with silky-feather . Of the two flanking genes , PDSS2 showed differential expression in both skin and liver at all stages while SOBP only showed differential expression at P60 and E9 . Furthermore , consistent with this finding , the silky-feather allele was associated with reduced promoter activity in a luciferase assay . We conclude that PDSS2 ( -103C-G ) is a cis-acting regulatory mutation that causes altered expression of PDSS2 and is the primary cause for the development of the silky-feather phenotype . We should note that we cannot exclude the possibility that ( i ) this mutation in the 5′ UTR of PDSS2 can influence gene expression at the level of translation via other mechanisms such as translation efficiency and mRNA stability [28] , [29] , and ( ii ) differential expression of SOBP [30] and/or other genes at some specific stage may have some impact on feather ( or hooklet ) development . Among silky-feather chicken types , Kuaida Silky chicken was developed by crossing Silkie with Broiler , inheriting the silky-feather haplotype from the Silkie . To our knowledge , there is no gene introgression between Jinyang Silky or Lanping Silky and the Silkie breed , and both of these breeds share no other phenotypic characters with Silkie . However , both of them share the 18 . 9 kb IBD silky-feather haplotype with Silkie . Thus , the same ancestral silky-feather mutation most likely originated before the formation of these breeds . The interactions between epithelium and mesenchyme play a critical role in the formation and development of feathers [19] . The epithelium over feather tract has potential for forming the feather field and transforms into individual feather primordia through an activator-inhibitor mechanism [15] , [17] . The competition between signaling activators ( e . g . SHH ) and signaling inhibitors ( e . g . BMP ) results in the local expression of genes ( e . g . SHH ) in the feather primordium [13] , [14] , [17] . This inductive process resides in the mesenchyme and is also epithelium-dependent [19] . Then feather bud forms with an anterior-posterior orientation involving inductive signals , e . g . the expression of Wnt7a in the posterior bud epithelium [31] . In the late long-bud stage , the feather follicle begins to form in the dermis . Subsequently , the natal down feathers and adult alternative feathers with branches emerge from follicles . The immunohistochemistry showed that PDSS2 was persistently expressed in the feather epithelium , mesenchyme , follicle and sheath in both the wild-type and silky-feather chickens ( Figure 5 ) . The gradually enhanced expression of PDSS2 coinciding with feather development indicates that it plays an important role in the formation of the feather , perhaps together with other signals from the follicle . Feather hooklets are formed in the last differentiation stage of the life cycle of a pennaceous feather . The feather pulp is filled with mesenchymal tissues transiently , and the branch is formed through differential cell death [32] . The different parts of the feather involve different keratin proteins , the presence or absence of which is associated with several mutant feather phenotypes [33] , [34] . The formation of hooklets are primarily due to an increased number of barbule cells [35] , [36] , which can grow into hooklets after differentiation and keratinization [37] . Wedge cells , which are supportive cells of regenerating adult feathers , accumulate periderm granules and corneous material , and then degenerate gradually to produce hooklets that allow individual barbs to align and form a closed vane in mature pennaceous feathers [37] , [38] . The expansion of the feather β-keratins genes is thought to have contributed to the evolution and diversity of feathers [39] . An uncharacterized feather β-keratin gene , named barbule specific keratin 1 ( BLSK1 ) , is expressed specifically in feather follicles that generate pennaceous barbules but not in follicles that generated plumulaceous barbule [40] . Further expression experiments indicated that BLSK1 might be involved in the formation of pennaceous barbules and hooklets . In silky-feathers , no hooklets form in apical cells of barbules in plumulaceous or pennaceous feathers . The feather follicle forms at the embryonic stage while contour feathers form after the first molting . During development , silky-feathers lack hooklets in pennaceous feathers compared with the wild-type . The expression of PDSS2 was decreased in silky-feather dorsal skin and feathers during all stages ( Figures 3 and 5 ) . We hypothesize that the persistently decreased expression of PDSS2 changes barbule cell differentiation and spatial reorganization in the feather follicle and results in the lack of hooklets in silky-feathers . The theory of feather morphogenesis hypothesized a hierarchical series of stages characterized by successive evolution of tubularity , barbs , the rachis , barbules and hooklets resulting in the closed pennaceous vane [20] , [41] , [42] . Diversification of feather complexity and function is achieved by independence , covariation and interaction among plumage modules [42] . Barbule plate cells were hypothesized to mature earlier than the central ramus cells for the nutrient transport [43] . Microscopic and ultrastructural analyses of feathers have been focused on the barbs and barbules , but few on cell composition [1] , [42] , [44] . Electron microscopy has revealed some processes of cell differentiation within developing feathers , but the formation of barb ridges and the cell junctions between barb and barbule cells are nearly unknown and will be an essential area of feather biology research [36] , [37] . In contrast to the frizzle mutation that affects rachis and barb morphogenesis [5] , silky-feather is the first locus characterized at the molecular level that affects hooklet morphogenesis . Since the hooklets are uniquely absent in the pennaceous feathers of silky-feather chickens across all the birds , it is difficult to infer the exact role of the PDSS2 gene in the development and evolution of feathers from the molecular evidence presented in this study . Further studies will provide clues into whether the PDSS2 is involved in the formation of barbules and hooklets in different feather tracts of more species . PDSS2 encodes a prenyl diphosphate synthase that is essential for ubiquinone ( Coenzyme Q ) biosynthesis , which is required for mitochondrial respiratory electron transport . In human , missense mutations in PDSS2 were reported to cause Coenzyme Q10 ( CoQ10 ) deficiency with Leigh syndrome with nephropathy [45] . In the patient's skin fibroblasts , CoQ10 decreased to 12% of control cells and ATP synthesis decreased by 51% compared with the control [46] . Similar kidney disease was seen in mice carrying PDSS2 missense mutations and in glomerular podocytes conditional PDSS2 knockout mice , but not in renal tubular epithelium , monocytes , or hepatocytes conditional knockout mice [47] . In homozygous missense mutant PDSS2 mice , CoQ content in the kidney was significantly lower compared with that in wild-type mice [47] , [48] . Although these PDSS2 missense mutant mice showed mitochondrial respiratory chain deficiency in all organs studied ( brain , kidney , liver and muscle ) , another study showed that only the affected kidney organs showed increased reactive oxygen species ( ROS ) production and oxidative stress [49] . All of the results indicated that the tissue-specific PDSS2 dysfunction associated with CoQ deficiency might be responsible for the renal disease phenotype [47]–[49] . Cerebellum dysfunction is often associated with ubiquinone deficiency and consistent with that cerebellum hypoplasia and cerebellar ataxia were observed in conditional knockout mice due to increased ectopic apoptosis [50] . Despite PDSS2 having an important role in CoQ biosynthesis and abnormal PDSS2 expression initiating kidney and nervous system problems in human and mouse , little is known about the effect of PDSS2 mutations in the skin . In this study , we showed that lower expression of PDSS2 in silky-feather chicken skin affected feather development . Mice with keratinocyte-specific deficiency in mitochondrial transcription factor A had an abnormally thick epidermis , lacked hair and showed defects in differentiation [51] . Further analysis showed that the keratinocytes did not produce mitochondrial ROS , which resulted in the impaired Notch and β-catenin signaling pathways during skin development . The increased apoptosis in hair follicles and hair loss in these mice should be caused by the impaired Notch signaling in epidermal differentiation and β-catenin signaling in hair follicle growth . These findings revealed that the mitochondrial ROS influenced the keratinocyte differentiation and hair follicle development [51] . Except for the silky-feather phenotype , none of the QTLs controlling growth , body composition or body size traits is co-localized with the PDSS2 locus in the same pedigree , indicating that the silky-feather mutation has no pleiotropic effect on these quantitative traits [21] , [52]–[54] . Silkie skin transplantation experiments have indicated that the determining factors for hooklet development reside in the feather follicle [55] , indicating that the mechanism of the silky-feather mutation in the feather follicle is independent of the endocrine status . However , some similar but non-genetic silky-feather phenotypes are known . Similar long silky-feathers were first observed in thyroidectomized Brown Leghorn chickens in 1927 [56] , and later in hypothyroid White Leghorn chickens [57] . At the same time , hypothyroid chickens are usually small , obese with increased abdominal fat , and have a small , dry comb , which are different from Silkie fowl . Silky-like feathers can develop in normal chickens with excessive dosages of thyroid [58] . Furthermore , the silky-feather barbs in Silkie chickens can form barbules and hooklets after regional depluming and a subcutaneous injection of thyroxine [59] , [60] . The thyroid hormone , together with other hormones , are known to contribute to specific aspects of hair growth [61] . Those experiments revealed that the thyroid hormone played an important role in the feather development , which was also found to influence the rate of feather regeneration and the number of regenerated papillae in the Red-headed Buntings ( Emberiza bruniceps ) [62] . Thyroid hormone has been well reviewed for the effect on cellular respiration through changing expression of respiratory genes and modulation of inner membrane structure [63] . Some of the nuclear encoded respiratory genes mRNA appears to be induced by thyroid hormone . Considering that thyroidectomized and hypothyroid wild-type chicken lost the hooklet and compensatory thyroid hormone in Silkie birds formed hooklet , the hooklet morphogenesis might be dependent on normal mitochondrial respiratory function affected by the presence of thyroid hormone . In silky-feather chicken skin , lower expression of PDSS2 may decrease the CoQ10 synthesis , which influences the mitochondrial respiratory chain in the feather follicle . The evolutionary origin of feathers has been under debate for more than 150 years [41] , [43] , [64] . Numerous feather structures have been found in dinosaur fossils that have improved our understanding of the origin and evolution of this highly branched epidermal structure [41] , [43] , [64] . Pennaceous feathers found in dinosaurs demonstrated that modern feathers evolved in non-avian dinosaurs and their arrangement and distribution provided new insight into the dinosaurian hypothesis of bird origins [65] . The simple pennaceous feathers of Protarchaeopteryx showed that remiges and rectrices evolved earlier than flight in theropod dinosaurs [66] , with variation in feather size and possible color appearing later in Anchiornis [67] . By understanding the process by which feathers grow and develop , we can gain insight into how this highly branched epidermal structure may have evolved . Here we describe a mutation in the modern day pennaceous feather that affects a portion of the feather that is crucial to flight . Further studies on how PDSS2 contributes to normal hooklet morphogenesis may provide us with a better understanding of the origin and evolution of pennaceous feathers . The PDSS2 ( -103C-G ) mutation is presumed to alter the interaction with one or more transcription factors . Regulatory mutations are an important contributor to phenotypic diversity , but it is challenging to establish genotype-phenotype relationships for regulatory mutations and convincingly prove causality [68] , [69] . However , a number of cis-regulatory element ( CRE ) mutations have been found to underlie phenotypic variation in domestic animals and CREs can regulate cell-type-specific expression of genes [69] , [70] . For example , regulatory mutations with phenotypic effects in chicken include: regulatory mutations in BCDO2 associated with yellow skin [71] , copy number variation in intron 1 of SOX5 with Pea-comb [72] , a deletion of an enhancer element upstream of SOX10 underlies the Dark brown plumage phenotype [73] , a large insertion downstream of BMP12 causes Naked neck [3] , a complex rearrangement involving EDN3 causes fibromelanosis [24] , an inversion resulting in a genomic relocalization of MNR2 causes Rose-comb [74] and an EAV-HP insertion in SLCO1B3 causes Blue-eggshell [75] , [76] . In addition to these regulatory mutations representing structural changes , we now provide a single base nucleotide substitution in non-coding DNA underlying a monogenic trait in chicken .
All the chickens were fed and handled according to relevant national and international guidelines . The China Agricultural University Resource Population ( CAURP ) consisted of 31 F0 , 19 F1 and 229 F2 individuals that were used for linkage analysis . The F0 generation consisted of 12 White Plymouth Rock chickens and 19 Silkie chickens homozygous for the wild-type and silky-feather alleles , respectively . All 19 F1 birds showed the wild-type phenotype . The observed ratio of 162 wild-type birds and 67 silky-feather birds among the 229 F2 birds , did not deviate significantly from the expected 3∶1 ratio ( χ2 = 2 . 21 for 1 df , P>0 . 05 ) . The USA population used for linkage analysis was derived from a single New Hampshire breed male , homozygous wild-type , and a single Silkie female chicken , homozygous for the silky-feather allele . A single male F1 individual was mated to 10 Silkie females and 180 backcross progeny were used for linkage mapping as previously described [22] . The majority of the DNA samples used for IBD analysis and putative causal mutation genotyping were from Jiangsu Institute of Poultry Science . Additional Silkie samples were from Guangdong and Beijing , China; North Carolina and Wisconsin , USA . Genotyping of the CAURP was performed with 125 microsatellite markers and the Illumina Chicken 60K SNP Beadchip [21] , [53] , [77] . Eight microsatellite markers were genotyped using ABI 3100 DNA Genetic Analyzer ( Applied Biosystems ) ( See Table S5 for marker information ) . An additional 16 SNPs were genotyped using the SNPlex genotyping system ( See Table S1 for SNP information ) . Linkage analysis was performed using the CRIMAP software [78] . The USA population was genotyped for 30 markers within the region previously reported [22] using a custom GoldenGate BeadXpress SNP panel . Two BAC clones were resequenced . The RJF BAC CH261-103J18 spanning 70 , 324 , 632–70 , 520 , 054 bp on chromosome 3 was from BACPAC Resource Center ( BPRC ) , and the Silkie BAC 292C7 spanning 70 , 349 , 337–70 , 487 , 932 bp was identified by PCR from a Silkie chicken BAC library [79] . BAC DNA was sheared to target sizes from 1 . 5 kb to 3 kb and cloned into pUC118 vector . The sub-clones were sequenced using an Applied Biosystems 3730xl DNA Analyzer . The RJF genome sequence galGal3 , assembled by the Washington University Genome Sequencing Center ( WUGSC ) , was used as the reference for alignment . The RJF-BAC was accomplished with 7 . 4-fold genome coverage and the Silkie BAC with 8 . 5-fold coverage . IBD mapping was performed using a panel of 12 breeds ( Table S2 for details ) . This included 76 silky-feather individuals of Silkie , Kuaida Silky and Lanping Silky , all homozygous h/h , and 95 wild-type birds ( H/H ) representing nine breeds . The samples of Silkie chickens originated from Beijing , Yangzhou and Guangzhou , China . The Kuaida Silky population was generated by crossing Silkie and a commercial Broiler line , followed by selection for Silkie-like appearance and rapid growth for five generations . Lanping Silky originally came from Yunnan , China , and there was no genetic introgression with Silkie as far as we knew . SNPs were genotyped using the Sequenom MassARRAY platform ( see Table S1 for SNP information ) . Based on the SNP genotyping information , eight silky-feather birds , nine wild-type birds and three heterozygous birds were used for re-sequencing . Primers were designed for overlapping fragments ( primer sequence information in Table S5 ) . The PCR amplicons were purified and directly sequenced using ABI 3730xl DNA Analyzer ( Applied Biosystems ) . DNA sequences were analyzed using DNASTAR software ( DNASTAR ) . Genotyping of PDSS2 ( -103C-G ) ( ss666793747 ) and ss666793770 was performed using pyrosequencing . The pyrosequencing PCR assays contained 40 ng gDNA , 1× PCR buffer , 1× Q-solution ( Qiagen ) , 200 µM dNTP , 400 nM of each primer , 2 IU LongAmp Taq ( NEB ) , and H2O was added to achieve a final volume of 25 µl . A touchdown PCR protocol was used , including 94°C for 5 min , 10 cycles of 94°C , 63°C ( −1 . 0°C/cycle ) , and 65°C for 30 s each , followed by 30 cycles of 94°C , 53°C , and 65°C for 30 s each , and a final extension at 65°C for 5 min . The PCR product was analyzed by 2% agarose gel electrophoresis and used for pyrosequencing according to standard protocol of PyroMark ID ( Qiagen ) . Tissues used for expression analysis were from Silkie ( h/h ) birds at postnatal ( P ) days P10 ( n = 7 ) , P60 ( n = 7 ) , P130 ( n = 10 ) , P200 ( n = 10 ) and White Leghorn ( H/H ) birds at P10 ( n = 7 ) , P60 ( n = 7 ) , Beijing You ( H/H ) at P130 ( n = 10 ) , Huiyang Bearded ( H/H ) birds at P200 ( n = 10 ) . Heterozygous ( H/h ) birds at P130 ( n = 11 ) used for allelic expression imbalance originated from a cross between Silkie ( h/h ) and Youxi Partridge ( H/H ) chicken . Dorsal skin and liver were collected and stored in liquid nitrogen . Embryonic dorsal skin tissue was collected from Silkie ( h/h ) and White Leghorn ( H/H ) birds . Samples from five birds of each genotype were collected at embryonic ( E ) days 9 , 11 , 14 and 17 . Tissue was homogenized using TissueLyser LT ( Qiagen ) . Total RNA was extracted using TriZol ( Ambion ) . 1 µg total RNA was used in the first strand cDNA synthesis ( Promega ) with oligo ( dT ) 18 primers . The details of primer sequence used for RT-PCR were presented in Table S5 . The 5′ rapid amplification of cDNA ends ( RACE ) was performed with the 5′ RACE System for Rapid Amplification of cDNA Ends , Version 2 . 0 ( Invitrogen ) . RNA was extracted from three Beijing You ( H/H ) and three Silkie ( h/h ) chicken dorsal skin tissues at P100 using Trizol reagent ( Ambion ) . The RNA was treated with DNase I ( Qiagen ) and purified using RNeasy Mini Kit ( Qiagen ) . 3 µg RNA for each sample was used for reverse transcription in a 25 µl reaction with SuperScript III ( Invitrogen ) , which was incubated at 55°C for 50 min for the high GC content . The original and the nested PCR assays contained 3 µl dC-tailed cDNA ( or 1∶100 dilution of primary PCR product ) , 1× PCR buffer , , 1× Q-solution ( Qiagen ) , 5% DMSO , 400 µM dNTP , 400 nM of each primer , 2 . 5 IU LongAmp Taq ( NEB ) , H2O was added to give a final volume of 25 µl . A touchdown PCR protocol was used , including 95°C for 3 min , 15 cycles of 95°C for 20 s , 65°C ( −1 . 0°C/cycle ) for 30 s , and 65°C for 2 min , followed by 20 cycles of 95°C for 20 s , 50°C for 30 s , and 65°C for 2 min , and a final extension at 65°C for 5 min . The PCR fragments were purified and sub-cloned into the pMD19-T vector . At least 16 clones per sample were selected for sequencing . Primers used for 5′ RACE were presented in Table S5 . Quantitative real-time PCR were performed with triplicate on LightCycler 480 ( Roche ) . PDSS2 gene was detected using 40 ng cDNA , 1× Taqman Master mix ( Applied Biosystems ) , 750 nM of each primer and 250 nM probe in a total volume of 20 µl with a PCR condition of 15 min at 95°C and 40 cycles of 95°C for 15 s and 60°C for 1 min . For analysis of SOBP , OSTM1 , SEC63 , BEND3 , C3H6orf203 , QRSL1 , RTN4IP1 , AIM1 , and ENSGALT00000037260 , 40 ng cDNA , 1× SYBR green mix ( Applied Biosystems ) and 300 nM of each primer were used in a total volume of 20 µl with a PCR condition of 15 min at 95°C and 40 cycles of 95°C for 15 s and 60°C for 1 min , followed by a melting curve analysis . Expression level data was normalized using GAPDH as endogenous reference gene and calculated using the 2−ΔΔCt method . Student's T-test was used to compare different groups . Pyrosequencing was used to test allelic expression imbalance of PDSS2 and SOBP . Total RNA was purified with DNase I ( NEB ) and minus-RT ( not reverse transcribed ) PCR products were used as negative controls . Two SNPs in PDSS2 ( ss666793773 in exon 8 and ss189596174 in exon 6 ) and two SNPs in SOBP ( ss666793686 and ss666793687 in exon 6 ) were designed for PCR amplification . The pyrosequencing PCR assays contained 40 ng cDNA ( or gDNA ) , 1× PCR buffer , 200 µM dNTP , 400 nM of each primer , 2 IU LongAmp Taq ( NEB ) , H2O was added to give a final volume of 25 µl . A touchdown PCR protocol was used for the pyrosequencing SNP genotyping test including 94°C for 5 min , 10 cycles of 94°C , 63°C ( −1 . 0°C/cycle ) , and 65°C for 30 s each , followed by 30 cycles of 94°C , 53°C , and 65°C for 30 s each , and a final extension at 65°C for 5 min . The PCR product was analyzed by 2% agarose gel electrophoresis and used for pyrosequencing according to standard protocol of PyroMark ID ( Qiagen ) . The relative proportion of each allele was obtained using the AQ analysis mode ( allele quantification ) . The details of all primer sequences were given in Table S5 . DF1 ( chicken fibroblast cell line ) and 293T ( human embryonic kidney cell line ) cells were cultured at 37°C in a 5% CO2 atmosphere in Dulbecco's Modified Eagle's Medium ( DMEM ) containing 4 . 5 g/l of glucose ( Gibco ) and supplemented with 10% fetal bovine serum ( FBS ) . The enhanced green fluorescent protein ( eGFP ) was fused to the N-terminus of the chicken PDSS2 ( cPDSS2 ) gene by fusion PCR . The eGFP-cPDSS2 was cloned into pcDNA3 . 1 ( + ) using BamHI and XhoI sites and used for transfection . 293T and DF1 cells were transiently transfected using FuGENE HD Transfection Reagent ( Promega ) according to the technical manual . An anti-cPDSS2 monoclonal antibody was custom made in mouse using the immunizing peptide IGISTWKEQV-amide corresponding to amino acid residues 221–230 ( Abmart , Shanghai , China ) . Cells were collected at 48 h after transfection . Protein was extracted from cells with Cell Lysis Buffer ( Beyotime ) . The proteins ( 30 µg of total cell protein per lane ) were separated by 12% SDS-PAGE and transferred to polyvinylidene difluoride ( PVDF ) membranes according to standard protocols . The membranes were blocked and incubated with anti-cPDSS2 ( 1∶1000 ) , anti-GFP ( 1∶1000 , ab6556 , Abcam ) and anti-alpha tubulin ( 1∶1000 , sc-53646 , Santa Cruz ) antibodies overnight at 4°C . The Membranes were subsequently incubated with horseradish peroxidase ( HRP ) conjugated goat anti-mouse or goat anti-rabbit secondary antibodies ( 1∶10000 ) and visualized using SuperSignal West Dura Extended Duration Substrate ( Thermo Scientific ) . 293T and DF1 cells ( transient transfection and wild-type control ) were cultured on poly-lysine-coated coverslips for 36 h before staining . Cells were fixed in 4% paraformaldehyde for 15 min at room temperature , permeabilized with 0 . 3% Triton X-100 for 10 min , and blocked for 1 h in blocking solution ( 2% goat serum , 1% BSA , 0 . 1% Triton-X and 0 . 05% Tween 20 in PBS ) . Cells were incubated with cPDSS2 antibody ( 1∶50 ) overnight at 4°C . Cells were washed three times with PBS and incubated with Alexa Fluor 594 Goat Anti-Mouse IgG Antibody ( 1∶500 , Invitrogen ) for 1 h at room temperature . Cells were washed three times with PBS and treated with DAPI ( 1 µg/mL ) for 5 min . The slides were washed with PBS , mounted with Antifade mounting medium ( Beyotime ) and imaged with the Olympus Fluoview FV1000 confocal microscope . Images were formatted , resized , enhanced and arranged using FV10-ASW and Adobe Photoshop . The embryonic dorsal skin tissues were fixed in 4% paraformaldehyde in PBS for 3 h , followed by 15 min wash with PBS and incubated in 30% sucrose overnight at 4°C . Tissues were frozen in OCT ( Sakura ) and sectioned to obtain 10 µm thickness . The sections were rehydrated and blocked for 2 h in blocking solution ( 2% goat serum , 1% BSA , 0 . 1% Triton-X and 0 . 05% Tween 20 in PBS ) . cPDSS2 antibody was diluted 1∶100 in blocking solution and incubated overnight at 4°C . The sections were washed three times for 5 min in PBS and incubated with Alexa Fluor 488 Goat Anti-Mouse IgG Antibody ( 1∶400 , Invitrogen ) for 2 h at room temperature . The nuclei were stained with DAPI . Samples were analyzed using Olympus Fluoview FV1000 confocal microscope with the same parameters . Images were processed using FV10-ASW and Adobe Photoshop . Two fragments containing PDSS2 ( -103C-G ) were generated with PCR and cloned into pGL3 Basic vector ( Promega ) , a longer 643 bp fragment ( 70 , 486 , 018–70 , 486 , 660 bp ) and a shorter 405 bp fragment ( 70 , 486 , 256–70 , 486 , 660 bp ) in the “forward” ( toward PDSS2 ) orientation . A longer 752 bp fragment ( 70 , 486 , 339–70 , 487 , 090 bp ) and a shorter 399 bp fragment ( 70 , 486 , 444–70 , 486 , 842 bp ) in the “reverse” ( toward SOBP ) orientation were also amplified and cloned into pGL3 Promoter vector ( Promega ) . NheI and XhoI sites were selected to construct the vector . DF1 cells plated on 24 wells were transfected at 70–80% confluency with 720 ng of the pGL3 reporter plasmid and 80 ng of pRL-TK Renilla luciferase construct by 2 µl Lipofectamine 2000 ( Invitrogen ) for each well . The luciferase activity was measured 23–24 h after transfection using the Dual-Glo Luciferase Assay System ( Promega ) and an Infinite F200 Luminometer ( Tecan , Switzerland ) . Ratios of firefly luminescence/Renilla luminescence were calculated , and normalized to control samples ( Basic vector ) . For each test construct , one expression value was the average of three technical replicates in each plate and three separate operations were carried out to represent the final value . The pGL3 Basic vector , pGL3 Promoter vector and pGL3 Control vector were used as control . Information on the chicken genome sequence is available at http://www . genome . ucsc . edu ( May 2006 , WUGSC 2 . 1/galGal3 ) . The sequence data presented in this paper have been submitted to GenBank with accession numbers KC166240 , KC166241 , JX982522 and JX982523 . | The feather is an excellent model for evolution and development due to its complex structure and vast diversity . Some chickens have silky-feather because of a loss of hooklets in pennaceous feathers , while most chickens have the wild-type normal feather . Hooklets are formed in the last differentiation stage of the life cycle of a pennaceous feather . Chickens with silky-feather are homozygous for a recessive allele ( hookless , h ) . Silkie chicken from China is one of the breeds showing the fascinating silky-feather phenotype and the breed has been known for hundreds of years . In this study , we mapped the silky-feather locus to an 18 . 9 kb interval and identified a single nucleotide polymorphism ( SNP ) completely associated with silky-feather . The causative mutation is located 103 base pairs upstream of the coding sequence of prenyl ( decaprenyl ) diphosphate synthase , subunit 2 ( PDSS2 ) . The expression of the PDSS2 gene is decreased in silky-feather skin during feather development in vivo . The silky-feather allele also reduces the PDSS2 promoter activity in vitro . This is the first report of feather structure variation associated with PDSS2 and provides new insight into molecular signaling in the late development stage of feather morphogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"biology",
"and",
"life",
"sciences",
"animal",
"genetics"
] | 2014 | A cis-Regulatory Mutation of PDSS2 Causes Silky-Feather in Chickens |
Here we report the genetic analyses of histone lysine methyltransferase ( KMT ) genes in the phytopathogenic fungus Magnaporthe oryzae . Eight putative M . oryzae KMT genes were targeted for gene disruption by homologous recombination . Phenotypic assays revealed that the eight KMTs were involved in various infection processes at varying degrees . Moset1 disruptants ( Δmoset1 ) impaired in histone H3 lysine 4 methylation ( H3K4me ) showed the most severe defects in infection-related morphogenesis , including conidiation and appressorium formation . Consequently , Δmoset1 lost pathogenicity on wheat host plants , thus indicating that H3K4me is an important epigenetic mark for infection-related gene expression in M . oryzae . Interestingly , appressorium formation was greatly restored in the Δmoset1 mutants by exogenous addition of cAMP or of the cutin monomer , 16-hydroxypalmitic acid . The Δmoset1 mutants were still infectious on the super-susceptible barley cultivar Nigrate . These results suggested that MoSET1 plays roles in various aspects of infection , including signal perception and overcoming host-specific resistance . However , since Δmoset1 was also impaired in vegetative growth , the impact of MoSET1 on gene regulation was not infection specific . ChIP-seq analysis of H3K4 di- and tri-methylation ( H3K4me2/me3 ) and MoSET1 protein during infection-related morphogenesis , together with RNA-seq analysis of the Δmoset1 mutant , led to the following conclusions: 1 ) Approximately 5% of M . oryzae genes showed significant changes in H3K4-me2 or -me3 abundance during infection-related morphogenesis . 2 ) In general , H3K4-me2 and -me3 abundance was positively associated with active transcription . 3 ) Lack of MoSET1 methyltransferase , however , resulted in up-regulation of a significant portion of the M . oryzae genes in the vegetative mycelia ( 1 , 491 genes ) , and during infection-related morphogenesis ( 1 , 385 genes ) , indicating that MoSET1 has a role in gene repression either directly or more likely indirectly . 4 ) Among the 4 , 077 differentially expressed genes ( DEGs ) between mycelia and germination tubes , 1 , 201 and 882 genes were up- and down-regulated , respectively , in a Moset1-dependent manner . 5 ) The Moset1-dependent DEGs were enriched in several gene categories such as signal transduction , transport , RNA processing , and translation .
In eukaryotic cells , DNA-dependent processes can be regulated by covalent modifications of histones such as methylation , acetylation , phosphorylation , sumoylation , and ubiquitination [1] . Long amino-terminal tails of histones protruding from nucleosome cores are especially subject to post-translational modifications . The combination of histone modifications to regulate cellular processes is a dynamic language , and is referred to as the histone code [1] . Histone modifications serve as marks for the recruitment of various chromatin proteins or protein complexes to modulate diverse chromatin functions including gene expression , silencing , repair , and replication [2] . Numerous “writing” enzymes ( methylases , acetylases etc ) and “erasing” enzymes ( demethylases , deacetylases etc ) are involved in the histone code . Histone methyltransferases are a group of enzymes catalyzing the transfer of methyl groups from S-adenosyl methionine to histones . They can be divided into two groups based on their target amino acid residues: protein arginine methyltransferases ( RMTs ) and histone lysine methyltransferases ( KMTs ) [3–5] . A nomenclature system for the KMT family has recently been proposed , in which KMTs are classified into eight major subclasses , KMT1 to KMT8 , based on their phylogenetic relationships and domain structure/organization [6] . For example , KMT1 proteins , exemplified by Drosophila melanogaster Su ( Var ) 3-9 , Schizosaccharomyces pombe Clr4 , and Neurospora crassa DIM-5 , specifically methylate H3K9 , which leads to gene silencing and heterochromatin formation [7–9] . The KMT2 proteins , typified by Saccharomyces cerevisiae SET1 and D . melanogaster Trithorax , specifically catalyze methylation at H3K4 , a mark for gene activation [10 , 11] . Since all KMTs except the KMT4 class contain a SET domain , named after three Drosophila lysine methyltransferases: Su ( var ) 3-9 , Enhancer of zeste , and Trithorax , they are also often referred to as SET proteins . It is to be noted that there are also known possible KMT proteins that are not included in the nomenclature system such as SET3 and SET4 in S . cerevisiae . KMTs are conserved in a wide range of eukaryotes , playing roles in cellular signaling pathways related to the cell cycle , cell motility , transcription , apoptosis , and cancer [12 , 13] . In filamentous fungi , KMT-related gene regulation has been investigated mainly with regard to gene silencing and secondary metabolite ( SM ) production [9 , 14–21] . In N . crassa , H3K9me3 catalyzed by DIM-5 belonging to the KMT1 class directs DNA methylation and heterochromatin formation by recruiting a protein complex containing heterochromatin protein-1 ( HP1 ) and DIM-2 DNA methyltransferase through interaction of the chromo shadow domain of HP1 and PXVXL-like motifs in DIM-2 [9] . In Fusarium graminearum , H3K27me3 catalyzed by KMT6 was required for normal fungal development and contributed to regulating the “cryptic genome” including SM gene clusters [18] . Gene repression by H3K9 and H3K27 methylation was also recently shown to be involved in fungal symbiosis and pathogenicity through production of SM and effectors [19 , 20] . In Aspergillus nidulans , H3K4me2 and H3K4me3 , marks for gene activation play a role in chromatin-level regulation of SM gene clusters [21] . A loss-of-function mutation of the CclA gene , a member of the H3K4 methylating COMPASS ( Complex Proteins Associated with Set1 ) , resulted in a reduction of H3K4me2 and H3K4me3 at the SM gene clusters [22] . Surprisingly , cryptic SM gene clusters are activated in the ΔcclA mutant despite H3K4me2 and H3K4me3 being considered marks for gene activation [22] . While it is generally believed that H3K4 di- and tri-methylation are epigenetic marks for gene activation in higher eukaryotes , involvement of H3K4 methylation in gene repression is also reported in fungi and other organisms [21–25] . To date , it is not clearly known to what extent genes are up- or down-regulated in a H3K4 methylation-dependent manner , and what is the underlying mechanism for this apparent discrepancy . Rice blast caused by Magnaporthe oryzae ( Pyricularia oryzae ) is one of the most devastating worldwide rice ( Oryza spp . ) diseases . This fungal species consists of several host-specific pathotypes that cause blast disease on a wide range of gramineous hosts including wheat , oat , finger millet and etc . Owing to their economic importance and genetic tractability , rice and M . oryzae have emerged as a model system for studying fungi-plant interactions [26] . M . oryzae displays dramatic morphological changes during infection [27] . When a fungal spore lands on a plant’s surface it germinates and forms a melanized dome-shaped infection structure , called an appressorium , at the tip of the germ tube . The appressorium generates enormous turgor pressure and physical force to breach the host cuticle , and the fungus eventually develops invasive hyphae to colonize host cells . These morphological changes are accompanied with global transcriptional alterations [28–30] . However , the involvement of genome-wide histone modifications in infection-related transcriptional alterations is poorly understood in M . oryzae . We recently demonstrated that MoSET1 catalyzing H3K4 methylation was required for substrate-induced transcriptional activation of the MoCel7C cellulase gene in M . oryzae [31] . Here we report a reverse genetics study of the KMT gene family in the M . oryzae genome , and examine their roles in global gene regulation related to the formation of infection structures in M . oryzae , especially focusing on that of MoSET1 .
Eight putative KMT genes were identified in the M . oryzae genome based on sequence similarity and domain structure of known KMTs in the KEGG database ( http://www . genome . ad . jp/kegg ) . Moset1 ( MGG_15053 ) belonging to the KMT2 family , was previously named after S . cerevisiae Set1 [31] . The other seven genes were designated in this study as Mokmt1 ( MGG_06852 ) , Mokmt3 ( MGG_01661 ) , Mokmt4 ( MGG_05254 ) , Mokmt5 ( MGG_07393 ) , Mokmt6 ( MGG_00152 ) , Mokmt2h ( MGG_02937 ) , and Moset6 ( MGG_15522 ) ( Table 1 ) . Mokmt1 , Mokmt3 , Mokmt4 , Mokmt5 , and Mokmt6 likely belong to corresponding KMT families [6] . MoKMT2H and MoSET6 showed amino acid sequence similarity with N . crassa SET-3 and Schizosaccharomyces pombe SET6 , respectively . To determine biological functions of M . oryzae KMT genes , deletion mutants were constructed using the split-marker recombination method [32] ( S1–S7 Figs ) . Ectopic transformants , which had an insertion of a disruption construct somewhere in the genome other than the target locus , and gene complementation strains , which were deletion mutant-derived strains complemented by random insertion of a plasmid carrying the corresponding wild-type locus , were also created . These strains were used in further studies alongside the deletion mutants ( S1–S7 Figs and S1 Table ) . Deletion mutants and a complementation strain of Moset1 were previously made and used in this study [31] . Histone lysine methylation levels in the KMT deletion mutants were assessed using western blotting with specific antibodies ( Fig 1 ) . In the Δmoset1 mutant , the levels of H3K4me2 and H3K4me3 were strongly reduced , while H3K4me1 moderately decreased . Abundance of signals for H3K9me3 , H3K27me3 and H4K20me3 was also significantly reduced in theΔmokmt1 , Δmokmt6 , and Δmokmt5 mutants , respectively . These target sites for Magnaporthe KMTs were consistent with known target sites for corresponding KMT family proteins in other organisms . Western blotting was also used to test antibodies against H3K14me2 ( active motif #39350 ) , H3K36me3 ( active motif #61102 ) , and H3K79me2 ( active motif #39144 ) , however , no specific signal reduction in any KMT deletion mutant was detected ( S8 Fig ) . H3K36 and H3K79 methylation are known to be catalyzed by Set2 ( KMT3 family ) and Dot1 ( KMT4 family ) , respectively , in S . cerevisiae . Other KMT proteins might be involved in these histone marks in M . oryzae , or the specificity of the antibodies might not be strict enough to distinguish between marked and non-marked histones in M . oryzae . The decreased levels of histone methylation in the deletion mutants were completely recovered in gene complementation strains ( S9 Fig ) . These results indicated that MoKMT1 , MoSET1 , MoKMT5 , and MoKMT6 catalyze methylation of H3K9 , H3K4 , H4K20 , and H3K27 respectively , in M . oryzae . The rates of conidiation , germination , and appressorium formation were assessed for phenotypical characterization of the KMT mutants with regards to infection . In addition , their growth rates were examined on rich media . The assay used two deletion mutants and one ectopic transformant for each KMT gene , with a complement strain employed when phenotypic defects were observed . The growth rates of the KMT-mutants were generally lower than that of the wild-type strain ( Fig 2A ) . Especially , Δmoset1 exhibited the most severe reduction in vegetative growth , conidiation and appressorium formation but not in germination ( Fig 2B–2D ) . TheΔmokmt3 , andΔmokmt2h mutants showed moderate defects in all phenotypic traits investigated in Fig 2 . TheΔmokmt6 mutants also showed severe reduction in conidiation and slight defects in appressorium formation ( Fig 2B and 2D ) . Compared with the wild-type strain , the rates of conidiation and appressorium formation were reduced to less than 10% in the Δmoset1 mutants , and to 20–50% in theΔmokmt3 andΔmokmt2h mutants ( Fig 2B–2D ) . The rate of germination was decreased by 40–50% in theΔmokmt3 andΔmokmt2h mutants . Interestingly , while the Δmoset1 mutants germinated at levels comparable to the wild-type , the conidia of Δmoset1 mutants often appeared to be malformed . The wild-type strain produced three-celled , tear-drop-shaped spores . Conidia of the Δmoset1 mutants were also three-celled , but were more elongated than the wild-type spores . All phenotypic defects observed in theΔmoset1 , Δmokmt3 , Δmokmt6 , andΔmokmt2h mutants recovered to wild-type levels in the corresponding complement strains with an exception ( conidiation inΔmokmt6 ) , indicating that the KMT mutant phenotypes were caused by the corresponding KMT genes . Infection assays of the KMT mutants were performed using three wheat and two barley cultivars with different levels of resistance/susceptibility to the wild-type wheat-infecting M . oryzae strain ( Br48 ) used in this study . The order of susceptibility of the cultivars to Br48 was as follows: barley , Nigrate ( super susceptible ) > barley , Russian No . 74 ≈ wheat , Norin 4 ( susceptible ) > wheat , Chinese spring ( moderate susceptible ) > wheat , Thatcher ( moderate resistant ) [33] . Consistent with the rates of appressorium formation , pathogenicity to the wheat and barley cultivars was most severely impaired in the Δmoset1 mutants ( Fig 3 and Table 2 ) . The Δmoset1 mutants produced no visible symptoms on most host plants tested . Interestingly , the Δmoset1 mutants caused disease , albeit with fewer lesions , on the super susceptible barley cultivar Nigrate ( S10 Fig ) , indicating that mutants did not completely lose their ability to infect plants . That fewer lesions were produced by the Δmoset1 mutants could largely be attributed to the low rates of appressorium formation . Δmokmt1 , Δmokmt3 , Δmokmt6 andΔmokmt2h mutants showed significant reduction in pathogenicity on all tested plant cultivars except Nigrate ( Table 2 ) . Δmokmt2h mutants failed to cause compatible lesions on any plants other than Nigrate , indicating that mutants became non-pathogenic with certain host plants susceptible to the parent strain Br48 . Δmokmt3 mutants also become non-pathogenic with the wheat cultivar Chinese spring , which is moderately susceptible to Br48 . The other strains , including theΔmokmt4 , Δmokmt5 andΔmoset6 mutants , ectopic transformants , and complement strains were infectious to all plant cultivars at levels comparable to the wild-type strain . To further examine infection types of the KMT mutants , cytological analysis of inoculated leaves of the wheat cultivar Norin 4 was performed . At least 100 spores with an appressorium were assessed , and cytological interactions classified into four types: A , B , C , and D [34] . In Type A , no reaction of the host cells was observed . In Type B , inhibition of fungal growth was associated with papilla , a cell wall apposition at the penetration site . Type C represents the hypersensitive reaction of epidermal cells . Types A to C are resistance responses of host cells . Type D describes a susceptible response where infection hyphae were observed in infected cells . In leaves infected with the wild type strain , the incidence of susceptible response Type D was predominant ( 63 . 9% ) ( Table 3 ) . Similarly , Type D was predominant in leaves infected with theΔmokmt4 , Δmokmt5 , andΔmoset6 mutants at levels similar to the wild-type strain . In contrast , the rate of resistant responses ( Types A + B + C ) was the majority in leaves infected with theΔmokmt1 , Δmoset1 , Δmokmt3 , Δmokmt6 andΔmokmt2h mutants ( Table 3 ) . In leaves inoculated with theΔmoset1 mutant , Type A ( no reaction ) was predominant , ( 77 . 3% ) , suggesting that this mutant mostly failed to penetrate plant cuticle and/or cell walls . TheΔmokmt3 andΔmokmt2h mutants induced cytological responses at very similar rates in the host cells . Infection by the two mutants was mostly prevented by the HR ( ~60% ) , and partly blocked at the papilla ( ~20% ) . Only a small percentage of germlings successfully formed invasive hyphae in infected cells ( Table 3 ) . TheΔmokmt1 andΔmokmt6 mutant showed a slight reduction in compatible interaction rate ( Type D ) , and slight increases in incompatible interaction rates ( Type A–C ) ( Table 3 ) . The order of the degrees of reduction in KMT-mutant pathogenicity was as follows: Δmoset1 >Δmokmt2h >Δmokmt3 >Δmokmt1≈Δmokmt6 . The KMT mutantsΔmokmt4 , Δmokmt5 , andΔmoset6 showed no detectable differences from the wild-type strain in all infection assays in this study . Therefore , we concluded that MoSET1 played the most important role in infection-related morphogenesis in M . oryzae , and focused on MoSET1 for our further studies . Pharmacological examination was performed to gain an insight into which stage in the signaling pathway leading to appressorium formation was blocked in Δmoset1 mutants . Chemical and physical signals from host plants can trigger infection-related morphogenesis in M . oryzae . One such chemical signal is 1 , 16-hexadecanediol , a plant cutin monomer released from the plant cuticle by degradation enzymes produced by the fungus [35] . After perception of external signals , the secondary messenger cyclic AMP ( cAMP ) plays a crucial role in the signaling pathway leading to appressorium formation in M . oryzae [36 , 37] . Therefore , the effects of 1 , 16-hexadecanediol and cAMP on infection-related morphogenesis in the Δmoset1 mutant were examined . In the presence of 5 mM cAMP or 5 μM 1 , 16-hexadecanediol , the rates of appressorium formation were greatly restored ( to over 80% ) in the Δmoset1 mutant , though rates were still lower than seen in the wild-type strain ( Fig 4B and 4C ) . It is noteworthy that there was no significant difference in the rate of appressorium formation between treatments with cAMP and 1 , 16-hexadecanediol , suggesting that Δmoset1 mutants may have defects in the production and/or perception of external signals from host plants . Inoculation assays were performed to determine whether the addition of exogenous 1 , 16-hexadecanediol and cAMP restored the pathogenicity of Δmoset1 mutants . 5 mM cAMP or 5 μM 1 , 16-hexadecanediol was added to conidia suspension of Δmoset1 mutants , and then spotted on leaves of the susceptible wheat cultivar Norin 4 . No symptoms were observed on leaves inoculated with the chemical treatments ( S11A Fig ) , suggesting that defects in appressorium formation were not the only cause making the Δmoset1 mutant noninfectious in wheat . To further examine this finding , wound inoculation tests were performed on Norin 4 . The Δmoset1 mutant failed to cause disease on the wounded leaves ( S11B Fig ) , suggesting that the Δmoset1 mutant had some deficits in its ability to develop disease , even after entering into plant tissues . Overall , these results suggested that MoSET1 is involved in the regulation of genes required for external signaling perception and disease development in plant cells . Chromatin immunoprecipitation sequencing ( ChIP-seq ) and RNA sequencing ( RNA-seq ) analyses [38] , using chromatin and RNA samples extracted from vegetative mycelia and germination tubes of the wild-type andΔmoset1 strains , were performed to examine genome-wide H3K4 methylation and MoSET1 distribution patterns during infection-related morphogenesis , and to determine their relationships with gene expression ( S2 Table ) . Germination tubes were collected after 6 h incubation , when appressoria had begun to form; genes involved in appressorium formation were expected to be at their most active at this time point . To carry out ChIP-analysis of MoSET1 , N-terminal FLAG-tagged MoSET1 was constructed and introduced into theΔmoset1 mutant . The phenotypic defects of theΔmoset1 mutant recovered when FLAG-tagged MoSET1 was introduced in the mutant ( S12 Fig ) , indicating that FLAG-tagged MoSET1 was functional . ChIP- and RNA-seq data were visualized by showing a representative chromosomal region ( Fig 5 ) . DNA immunoprecipitated with H3K4me2 and H3K4me3 antibodies predominantly localized to coding regions in the M . oryzae genome . As a control to show overall H3 levels , ChIP-seq data with antibodies against a C-terminal peptide of histone H3 was also presented ( Fig 5 ) . H3K4me3 accumulated relatively more in the 5′ gene regions as reported in other organisms ( Figs 5 and S13 ) as reported in other organisms [39–41] . The number of genes showing differences in normalized mean coverage of H3K4me2 and H3K4me3 enrichment between vegetative mycelia and germination tubes is presented in Table 4 . Genes with altered ChIP enrichment ( p < 0 . 01 ) , and over-represented Gene Ontology ( GO ) categories in the gene sets , are listed in S3 and S4 Tables . Changes in the H3K4me3 ChIP coverage were more frequently detected than those in the H3K4me2 coverage , suggesting that H3K4me3 could be a dynamic mark for gene regulation during infection-related morphogenesis in M . oryzae . MoSET1 ChIP-seq reads were also largely mapped to gene regions . In contrast to H3K4 methylation that showed specific enrichment patterns among genes , MoSET1 appeared to be distributed rather ubiquitously to almost every gene , albeit at varying levels . In Table 4 , the number of genes showing differences in MoSET1 enrichment between vegetative mycelia and germination tubes is presented . In some cases such as MGG_11148 and MGG_11149 , relative enrichment of H3K4 methylation in mycelia than in germination tubes was associated with levels of MoSET1 enrichment at the loci ( Fig 5 ) . Consistently , in 91 of 133 ( 68 . 4% ) and 268 of 399 ( 67 . 2% ) genes showing significant enrichment of H3K4me2 and H3K4me3 , respectively ( Table 4 ) , normalized mean coverage of MoSET1 was concomitantly increased in germination tubes . However , it is not likely that different levels of H3K4 methylation among genes were simply attributed to levels of MoSET1 localization to their loci . For instance , while similar levels of MoSET1 coverage were observed at the MGG_01752 and MGG_01753 loci in germination tubes , much higher H3K4me2/me3 coverage was detected at the MGG_01752 locus than at the MGG_01753 locus ( Fig 5 ) . These results implied that MoSET1 could principally distribute throughout the genome but might not be always enzymatically active . To gain a global view of the relationship between H3K4 methylation and gene expression , the M . oryzae genes were categorized into groups of 100 genes based on their expression levels , and the levels of H3K4 methylation in the gene groups were analyzed . Levels of H3K4me2 and H3K4me3 decreased as RNA levels decreased in mycelia and germination tubes ( Fig 6A and 6B ) , indicating that H3K4 methylation was associated with active gene expression in M . oryzae in a similar way as reported in other organisms [39–41] . Next , we examined whether changes in the H3K4me2/me3 patterns were associated with gene activation or silencing during infection-related morphogenesis in M . oryzae at a global scale . RNA-seq analysis revealed that a total of 4 , 077 genes showed significant increases ( 1 , 936 genes ) or decreases ( 2 , 141 genes ) ( p < 0 . 001 ) in expression levels in germination tubes ( Table 4 ) . These were sorted into five up-regulated and five down-regulated gene groups , and changes in H3K4me2/me3 ChIP enrichment in the groups were plotted ( Fig 6C ) . The median H3K4me2 levels slightly decreased as the magnitude of transcript reductions increased . The H3K4me3 profile showed a more dynamic correlation with the transcript levels compared with the H3K4me2 levels , and the medians of the H3K4me3 levels were higher in the up-regulated gene groups and lower in the down-regulated gene groups . It is noteworthy that the transcriptional activity of a gene was not always associated with local enrichment of H3K4 methylation as reported previously [18] . Higher H3K4me2/me3 coverage in mycelia than in germination tubes was accompanied with higher gene expression in some cases such as MGG_11149 in Fig 5 . The up-regulation of the MGG_11149 gene in mycelia was significantly diminished in theΔmoset1 mutant , supporting the idea that H3K4me2/me3 contributes to gene activation . However , transcriptional activation of the neighbor genes ( MGG_01755 , MGG_01756 , and MGG_01757 ) in germination tubes did not concomitantly occur with apparent H3K4me2/me3 enrichment at their loci ( Fig 5 ) . In addition , with the MGG_11148 gene , H3K4me2/me3 enrichment in the wild-type strain appeared to be accompanied with gene activation but a similar change in gene expression also occurred in theΔmoset1 mutant ( Fig 5 ) . Such apparent discrepancies were observed in many other cases . For example , H3K4me3 were significantly enriched in 398 genes and depleted in 223 genes in germination tubes compared to in mycelium ( Table 4 ) . Increase and decrease in RNA-seq read coverage were not accompanied with the H3K4me3 enrichment and depletion in 70 ( 17 . 6% ) and 59 ( 26 . 5% ) genes , respectively . Thus , H3K4 methylation tends to be overall associated with transcriptionally active genes , but the mechanism of gene regulation by H3K4 methylation is fairly complex , and possibly affected by other histone modifications . The roles of MoSET1 in gene regulation were investigated by RNA-seq analysis of the Δmoset1 mutant during infection-related morphogenesis . A total of 2 , 572 genes were differentially expressed in Δmoset1 mycelia compared with the wild-type strain ( p < 0 . 01 ) , with 1 , 491 genes up-regulated and 1 , 081 down-regulated . Similarly , in germination tubes , 1 , 388 genes were up-regulated and 1 , 044 genes down-regulated in the Δmoset1 mutant . These results indicated that a significant amount of M . oryzae genes were affected by the Moset1 mutation . Interestingly , the number of genes up-regulated in the Δmoset1 mutant was comparable to , or even more than , the number of down-regulated genes , suggesting that MoSET1 directly or indirectly plays a role in gene repression , as well as in gene activation . To analyze the characteristics of differently expressed genes between the wild-type and Δmoset1 strains , we examined the frequency distribution of genes belonging to these gene groups in mycelia ( Fig 7A ) and germination tubes ( Fig 7B ) , based on the expression levels in the wild-type strain . In both mycelia and germination tubes , genes down-regulated in the Δmoset1 strain were highly biased to the high expression gene groups in the wild-type strain , while those up-regulated in the Δmoset1 strain were more frequently distributed in medium and low expression level gene groups ( Fig 7A and 7B ) . We next addressed how the moset1mutation affected gene regulation by comparing the fold change ( FC ) in gene expression during infection-related morphogenesis between the wild-type andΔmoset1 strains . A log2-scale scatter plot showed a positive linear correlation between the FC values in the wild-type andΔmoset1 strains ( p = 6 . 8E-06 ) ( Fig 7C ) . However , regression analysis gave the equation , y = 0 . 51x-0 . 04 with the correlation coefficient , r2 = 0 . 49 , indicating that the correlation was only moderate . The slope lower than one indicated that gene expression changes were generally more marked in the wild-type strain ( x-axis ) than in theΔmoset1 mutant ( y-axis ) for both up- and down-regulated genes ( Fig 7C ) . This supported the conclusion that MoSET1 contributed to bilateral and global gene regulation during infection-related morphogenesis in M . oryzae . To assess the effect of MoSET1 on gene induction and repression during infection-related morphogenesis , we focused on a subset of 4 , 077 genes that showed a significant change in expression levels between wild-type mycelia and germination tubes in the RNA-seq analysis ( Table 4 ) . There were less up-regulated genes in the subset ( 1 , 936 genes ) than down-regulated genes ( 2 , 141 genes ) . To understand the dependency of their gene expression on Moset1 , we defined the criterion of “Moset1-dependent genes” based on a comparison of FC values ( germination tubes/mycelia ) between wild-type and Δmoset1 strains . When the rate of FC increase or decrease of a gene in the Δmoset1 strain was less than 50% of the wild-type strain , the gene was categorized as a Moset1-dependent gene . Genes not meeting the criterion were classified as “Moset1-independent genes” . Based on these criteria , 1 , 201 and 735 genes were categorized as Moset1-dependent and -independent up-regulated genes , respectively; and 883 and 1 , 258 genes were grouped as Moset1-dependent and -independent down-regulated genes , respectively . Therefore , approximately half of the transcriptional changes during infection-related morphogenesis were directly or indirectly dependent on Moset1 in M . oryzae . Dependency on Moset1 was more evident with the up-regulated genes . Lists of the Moset1-dependent and -independent genes and over-represented GO categories in the gene sets were given in S5 and S6 Tables , respectively . Moset1-dependent and -independent genes were further classified using euKaryotic Orthologous Group ( KOG ) functional categories ( Fig 8 ) . The KOG category “signal transduction mechanisms” was highly over-represented in the MoSET1-dependent up-regulated gene set . Consistently , several GO categories related to “signal transduction mechanisms” were significantly over-represented in the gene set ( S6 Table ) . In this category , forty one kinases and thirteen GTPase regulators were detected , indicating that a large number of key signal mediators were transcriptionally regulated by MoSET1 , either directly or indirectly . Interestingly , twenty-four active transmembrane transporters , including MgAPT2 ( MGG_02767 ) , MgAPT3 ( MGG_04066 ) , and MgAPT4 ( MGG_04852 ) [42] were regulated by MoSET1 . The KOG categories over-represented in the Moset1-dependent down-regulated gene set were different from those in the upregulated gene sets , and included “translation , ribosomal structure and biogenesis” and “RNA processing and modification” . Interestingly , sixty-four structural constituents of the ribosome were found in this criterion ( S6 Table ) , indicating that MoSET1 was associated with down-regulation of a significant portion of ribosome-related genes . In addition , various nucleic acid binding proteins , especially those that bind RNA , were down-regulated in a Moset1-dependent manner . These included proteins homologous to nuclear ribonucleoprotein , RNA helicase , tRNA synthetase , rRNA biogenesis protein , poly ( A ) polymerase , and poly ( A ) -binding protein . Finally , we addressed whether Moset1-dependent gene regulation is directly related to H3K4 methylation . Levels of H3K4me2 and H3K4me3 enrichment in mycelia and germination tubes were plotted separately by the four gene criteria indicated in Fig 9 . In the up-regulated gene groups , levels of H3K4 methylation were generally higher in the Moset1-dependent genes than in the Moset1-independent genes . In addition , the Moset1-dependent genes showed stronger enrichment of H3K4me2 in germination tubes , where they were up-regulated , than did the Moset1-independent genes ( Fig 9 ) . Thus , in the up-regulated gene groups , changes in H3K4 methylation were more dynamic in the Moset1-dependent genes than in the Moset1-independent genes , suggesting direct contribution of H3K4 methylation to gene activation . In contrast , in the down-regulated gene groups , while both H3K4me2 and H3K4me3 levels decreased in germination tubes , where the genes in these criteria were down-regulated , only slight difference was observed in H3K4me2 and H3K4me3 enrichment patterns between the Moset1-dependent and -independent genes . This may suggest that the Moset1-dependency in the down-regulated genes was not directly resulted from H3K4 methylation .
The gene knockout studies of the eight KMT genes in M . oryzae revealed that MoKMT1 , MoSET1 , MoKMT3 , MoKMT6 , and MoKMT2H played significant roles in infection-related morphogenesis and/or pathogenicity to varying degrees , while MoKMT4 , MoKMT5 , and MoSET6 did not . Δmokmt1 mutants did not display detectable defects in infection-related morphogenesis , but showed a slight reduction in vegetative growth and of pathogenicity on host plants . MoKMT1 belongs to the KMT1 family responsible for methylation at H3K9 and is paralogous to N . crassa DIM-5 . In N . crassa , H3K9me3 catalyzed by DIM-5 is recognized by HP1 that forms a complex with DIM-2 DNA methyltransferase [9] . HP1 is a structural protein essential for heterochromatin formation , and leads to gene repression [43] . Since H3K9me3 is a conserved epigenetic mark for gene repression , MoKMT1 is likely involved in gene repression in M . oryzae . Interestingly , Δmokmt6 mutants showed a reduction in pathogenicity at levels similar toΔmokmt1 ( Tables 2 and 3 ) . The KMT6 family enzymes catalyze H3K27 methylation , a mark for gene repression . In F . graminearum , a KMT6 deletion mutant exhibited developmental defects and reduced pathogenicity as didΔmokmt6 mutants [18] . Similarly , in the plant pathogenic fungus , Leptosphaeria maculans , silencing of LmDIM5 belonging to the KMT1 family resulted in a reduction in pathogenicity [20] . Thus , gene repression itself or proper switching from gene repression to expression may be required for the full pathogenicity of the fungi , or these KMT genes may have functions other than gene repression . Δmokmt3 andΔmokmt2h mutants showed a significant reduction in vegetative growth , germination , appressorium formation , conidiation , and pathogenicity to host plants . Thus , these mutants had defects in every phenotypic assay performed in this study . The reduction in pathogenicity was more severe inΔmokmt2h mutants than inΔmokmt3 mutants . MoKMT3 is paralogous to N . crassa SET-2 , which belongs to the KMT3 family responsible for methylation at H3K36 . H3K36me3 levels peak within the body of active genes , and may be associated with transcription elongation through contributing to the maintenance of chromatin architecture [44 , 45] . In N . crassa , SET-2 loss-of-function mutants show various defects , including slow vegetative growth , low conidiation , and female sterility [16] . This is consistent with our results . Thus , MoKMT3 possibly affected the correct expression of a number of genes involved in phenotypic defects . While possible paralogs of MoKMT2H are widely conserved in ascomycete fungi , their biological roles have not been well-characterized . Based on phylogenetical analysis , the most closely related KMT to MoKMT2H in mammals is ASH1L , which is implicated in H3K4 and H3K36 methylation , and in transcriptional activation of certain genes , including Hox genes [46 , 47] . Therefore , if MoKMT2H is a functional homolog of ASH1L , it follows that MoKMT2H may also contribute to gene activation in M . oryzae . MoSET1 was among the most crucial KMTs for infection-related morphogenesis and symptom development on host plants in M . oryzae . Δmoset1 mutants showed severe defects in vegetative growth , appressorium formation , conidiation , and pathogenicity to host plants , but not in the rate of germination . MoSET1 catalyzes methylation at H3K4 , an evolutionary conserved epigenetic mark for gene activation . Our data suggested that this epigenetic mark could be the most important histone methylation for infection-related gene expression in M . oryzae . RNA-seq analysis of the Δmoset1 and wild-type strains suggested that approximately half of the genes induced or repressed during infection-related morphogenesis were dependent on MoSET1 . Moset1-dependent genes appeared to be involved in various infection-related processes . One such process is appressorium formation . cAMP signaling is crucial for appressorium formation in M . oryzae [36 , 37] . Several genes involved in the cAMP signaling pathway leading to appressorium formation , including MacI ( MGG_09898 ) , MacI-interacting protein ( MGG_05531 ) , Mck1 ( MGG_00883 ) , and Sum1 ( MGG_07335 ) were categorized as Moset1-dependent up-regulated genes [48] . The deficiency in transcriptional up-regulation of these genes in the germination tubes may be the cause of severe reduction in appressorium formation in the Δmoset1 mutants . This assumption is consistent with appressorium formation in the mutants being restored by exogenous cAMP . Pathogenicity of Δmoset1 mutants to host plants was not recovered by cAMP addition , indicating that Δmoset1 mutants had defects in pathogenicity other than appressorium formation . A large number of transporters were up-regulated in a Moset1-dependent manner in germination tubes ( Fig 8 ) . One such gene , MgApt2 is a P-type ATPase involved in exocytotic processes during plant infection [42] . Exocytotic mechanisms are involved in the delivery of proteins into plant cells to suppress plant defenses . Signal transducers were highly enriched in the Moset1-dependent gene set . MgATG1 ( MGG_06399 ) , a serine/threonine-protein kinase found in this group , is involved in autophagy and the generation of normal turgor pressure in the appressorium , and is thus essential for successful infection [49] . However , many important protein kinases for fungal pathogenesis , including Pmk1 ( MGG_09565 ) , CpkA ( MGG_06368 ) , Msp1 ( MGG_05344 ) , and MST7 ( MGG_00800 ) were not Moset1-dependent genes . Thus , not all signal transducers responding environmental stimuli required MoSET1 for their activation . Cell-wall degradation enzymes ( CWDEs ) are possible Moset1-dependent contributors to the full pathogenicity of the fungus . We previously reported that CWDEs , such as GH7 and GH8 cellulases and GH10 and GH11 xylanases , were greatly activated during infection [50 , 51] . Their gene activation was often induced by presence of relevant cellulose or xylan substrates . Substrate-dependent gene activation of the cellulases was severely compromised in the Δmoset1 [31] . Therefore , lack of CWDE activation during infection may be the cause of the severe reduction observed in the pathogenicity of the Δmoset1 mutants on the host plants . Gene activation of CWDEs was , however , barely detectable in the RNA-seq analysis; this is a consequence of RNA being extracted from germination tubes on slide glasses , thus having no available enzyme substrates . It should be note that , sinceΔmoset1 mutants were also impaired in vegetative growth , the role of MoSET1 in gene regulation was not infection specific . Thus , the down-regulation of such general genes could also contribute to the loss of pathogenicity inΔmoset1 mutants . In eukaryotes , H3K4 methylation is an epigenetic mark for gene activation . ChIP-seq or ChIP-chip analysis together with transcriptome analysis in human , Arabidopsis , and S . cerevisiae revealed a global positive correlation between H3K4me2/me3 and active transcription [39–41] . In M . oryzae , substrate-induced gene expression of GH6 and GH7 cellulases was associated with enrichment of H3K4me2 [31] . Interestingly , however , expression levels of GH6 and GH7 cellulases under non-inducing conditions increased in the Δmoset1 mutant , suggesting a possible role of H3K4 methylation in gene repression [31] . This is consistent with the observations in Aspergillus nidulans and A . fumigatus , where a deletion mutant of the CclA gene , which encodes a component of the COMPASS complex catalyzing H3K4 methylation , results in a reduction in H3K4me2 and H3K4me3 , and also causes increased gene expression of cryptic SM gene clusters [21 , 22] . Thus , CclA-mediated H3K4 methylation appears to contribute to gene silencing of SM clusters . It is noteworthy that SET1 in S . cerevisiae was originally identified as a gene required for transcriptional silencing of silent mating-type loci in the subtelomeric region [10] . Subsequently , SET1 was demonstrated to play a role in silencing rDNA and the retrotransposon Ty1 in S . cerevisiae [23 , 24] . Consistently , contribution of MoSET1 to the repression of ribosome-related genes was shown in this study ( Fig 8 ) . Recently , SET-1 was reported to have a role in DNA methylation of the frq promoter in N . crassa [52] . Thus , SET1 orthologs in fungi are involved in gene silencing in addition to gene activation . Our RNA-seq analysis revealed that significant numbers of M . oryzae genes were up- or down-regulated in the Δmoset1 mutant in comparison with the wild-type strain , supporting that H3K4 methylation is directly or indirectly involved in both gene activation and repression . Moset1-dependent gene up-regulation was largely detected in highly-expressed genes in the wild-type strain , whereas Moset1-dependent gene repression was more frequently observed in genes with middle or low expression levels ( Fig 7 ) . It is to be noted that the roles of Moset1-dependent gene repression in the pathogenicity of the fungus were so far unclear while Moset1-dependent gene activation , most likely , indeed contributed to infection-related morphogenesis in M . oryzae as discussed above . Apparent differences in the roles of KMT2 proteins among organisms can be attributed to the experimental approaches . In fact , this study demonstrates that results obtained by ChIP-related techniques in fungi are not much different from those obtained in higher eukaryotes . The role of SET1 orthologs in gene repression has mainly been revealed by gene knock-out approaches that are often used in fungi but seldom in higher eukaryotes . In higher eukaryotes , several SET1 homologs redundantly serve as catalytic enzymes for methylation at H3K4 . At least six Set1 homologs ( Set1A , Set1B , MLL1 , MLL2 , MLL3 , and MLL4 ) have been identified in mammalian cells , making gene knock-out strategies ineffective . Thus , it might be possible that the complete loss of SET1 homologs in the genome uncovered their additional functions that were hard to find by other approaches . However , it also should be noted that , in gene knockout approaches , it is difficult to distinguish direct and indirect effects of the loss of the target gene . Since SET1 homologs positively regulate global gene expression , their knockout mutants might fail to activate genes required for proper gene repression . Thus , de-repression of genes in theΔmoset1 mutant may not arise directly from the loss of the gene but may come from secondary causes , for example , insufficient expression of repressor genes in the mutant . Our results showed that changes in H3K4 methylation during germination tube formation were more dynamic in the Moset1-dependent gene set than in the Moset1-independent gene set among the up-regulated genes but not among the down-regulated genes ( Fig 9 ) , suggesting that the Moset1-dependency in the down-regulated genes could not be directly related to changes in H3K4 methylation . The data favors the hypothesis that the Moset1-dependent gene repression arose from indirect effects of the loss of MoSET1 even though other hypotheses are not completely eliminated . For example , recently , it has been reported that H3K4 monomethylation functioned as a mark for gene repression in several types of mammalian cells [25] . Thus , it might be possible that the lack or severe depletion of H3K4 monomethylation in the set1 or cclA deletion mutants resulted in activation of genes that were repressed under the wild-type background .
The wheat-infection M . oryzae isolate , Br48 [53] and its transformants constructed in this study ( S1 Table ) were kept on barley seeds media at 4°C for long-term storage [54] . For working culture , a barley grain from the stock culture was placed on a PDA ( potato dextrose agar ) slant media and cultured at 25°C . Fungal plugs were transferred to flasks containing complete medium ( 5% sucrose , 3% casamino acids , and 3% yeast extract ) and incubated in a shaker at 120 rpm at 25°C for 4 days . To prepare conidial suspension , fungal strains were cultured on oatmeal agar plates ( 40g of oatmeal , 17g of agar in 1000 ml water ) in the darkness at 25°C for 5 days . Then , aerial mycelia were removed by rubbing surface of mycelia with a sterile microtube , and further incubated under BLB light for 3 days at 25°C to induce conidiation . In this study , a split-marker gene disruption strategy [32] was used to obtain a gene knock-out mutant in M . oryzae ( see S1–S7 Figs ) . First , PCR products of the upstream and downstream of a targeted gene were cloned separately into the multiple cloning site of pSP72-hph that carries the Hygromycin resistance gene cassette [55] . Primers used in this study are given in S7 Table . PCR fragments amplified from the resulting 5’ and 3’ constructs were mixed and introduced into fungal spheroplasts by a polyethylene glycol ( PEG ) -mediated method as previously described [56] . For initial screening , colonies PCR were performed with appropriate sets of primers for each gene . The candidate strains were further examined by Southern blot analysis . Fungal genomic DNA was extracted using Plant Genomic DNA Extraction Miniprep System ( Viogene ) following the manufacturer’s instruction . Southern blot analysis was performed using the DIG DNA Labeling and Detection Kit ( Roche Applied science ) . Ten to twenty micrograms of genomic DNA were digested by appropriate restriction enzymes . The digests were separated by agarose gel electrophoresis and transferred to Hybond N+ ( Amersham biosciences ) . The hybridization procedures were carried out according to the manufacturer’s instructions . The positions of DIG-labeled probes used in Southern blot analysis are given in S1–S7 Figs . Genetic complementation of KMT deletion mutants was performed by introducing the corresponding native genomic fragment to them . Genomic DNA fragments containing KMT genes with their 5’flanking and 3’flanking regions were amplified with pairs of specific primers ( S7 Table ) using KOD FX Neo ( Toyobo ) , and cloned into pBluescript SK ( + ) . Each of the resulting plasmids was introduced into the corresponding KMT mutant with pII99 carrying the geneticin-resistance gene . Growth rates ( colony diameter ) of M . oryzae mutants on PDA media were measured up to 14 days with three replications . For conidiation assay , conidia were harvested 3 days after BLB induction by suspending them with 20ml sterile distilled water per plate . Spore concentration was estimated by microscopic observation of at least 20 visual fields using a hemocytometer . For conidial germination and appressorium formation assays , conidial suspension ( 105 spores per ml ) dropped on slide glasses was incubated in a humidity box at 25°C . The rates of conidia germination and appressorium formation were counted by microscopic observation of at least 200 spores after 5 and 24 h incubation at 25°C , respectively . Infection assay was performed as described previously [57] . Wheat and barley seedlings were grown in vermiculite supplied with liquid fertilizer in plastic pots ( 5 . 5cm×15cm×10cm ) at 22°C in a controlled-environment incubator with a 12h photoperiod for 8 days . The plant cultivars used were wheat cultivars “Norin 4” , “Chinese spring” and Thatcher , and barley cultivars “Russian No . 74” and “Nigrate” . Conidia suspension ( 1–2☓105 spores/ml ) containing 0 . 01% Tween 20 was sprayed to the primary leaves of 8-day-old wheat and barley seedlings . The inoculated seedlings were maintained under high humidity and dark conditions for 24 hours then moved to an incubator at 22°C with a 12h photoperiod for 5 days . Symptoms appeared were assessed based on the size and color of lesions to determine infection type . The size of a lesion was rated from 0 to 5: 0 , no visible evidence if infection; 1 , pinpoint spots; 2 , small lesion ( <1 . 5 mm ) ; 3 , lesion with an intermediate size ( <3 mm ) ; 4 , large and typical lesion; 5 , complete blighting of leaf blades . Green ( G ) and brown ( B ) lesions were regarded as susceptible and resistant responses , respectively [58] . For microscope observation of host cell response to M . oryzae , inoculated leaves were picked up at 48 h post inoculation ( hpi ) and deeply boiled in alcoholic lactophenol ( lactic acid/phenol/glycerol/distilled water/ethanol = 1:1:1:1 , v/v/v/v ) for 2 min as descried previously [57] . Samples were observed using an epifluorescence microscope under bright and fluorescent fields . Host response was classified into four types: no reaction , papilla formation , hypersensitive reaction ( HR ) , and hyphal growth . Fungal mycelia powder was suspended in 1×TBS buffer ( 50 mM Tris-HCl [pH 7 . 5] , 150 mM NaCl ) containing 1% Nonidet P-40 . The homogenates were centrifuged ( 12 , 000 rpm , 2 minutes ) , and supernatants were collected . Proteins in the supernatants were then heat-treated at 80°C for 10 min to precipitate contaminating proteins . The supernatant was recovered by centrifugation , and subjected to 15% SDS_polyacrylamide gel electrophoresis . After blotted to PVDF membrane , proteins were probed with the following primary antibodies; anti-H3K4me1 ( Active motif #39298 ) , anti-H3K4me2 ( Active motif #39141 ) , anti-H3K4me3 ( Active motif #39159 ) , anti-H3K9me3 ( Active motif #39161 ) , anti-H3K27me3 ( Active motif #39535 ) , anti-H4K20me3 ( Active motif #39181 ) and a C-terminal peptide of histone H3 ( Active motif #39163 ) . Washing was performed three times with TBS-T buffer containing a higher concentration of NaCl ( 50 mM Tris-HCl [pH 7 . 5] , 190 mM NaCl , 0 . 05% Tween 20 ) . Proteins reacting with the primary antibodies were visualized by appropriate peroxidase ( HRP ) -conjugated secondary antibodies and ECL plus western blotting detection regents . An N-terminal 2xFLAG-tagged MoSET1 construct was made by PCR amplification with the primers , FLAG-MoSET1-F and MoSET1-TGA-R ( S7 Table ) , and then fused to the native MoSET1 promoter and terminator sequences by In-Fusion HD cloning kit ( Clontech Laboratories ) at EcoRV site in pBluescript SK ( + ) with the primers , IF-PMoSET-F , IF-PMoSET-R , IF-TMoSET-F , and IF-TMoSET-R ( S7 Table ) . The resulting construct was introduced into the Δmoset1 mutant and used in chromatin immuno-precipitation ( ChIP ) analysis . ChIP experiments were performed with germinating conidia and vegetative mycelia using the ChIP-IT Express kit ( Active motif #53008 ) according to manufacturer's instructions using sonication as a method for chromatin shearing . In addition to the antibodies used in western blot analysis , Anti-DDDDK-tag mAb-magnetic beads ( Medical & Biological Laboratories , Japan ) was used in ChIP experiments . Briefly , samples ( 100mg ) were treated with 1% formaldehyde by shaking gently ( 100rpm ) for 30 minutes at room temperature . Chromatin was sheared on ice by sonication using a Bioruptor apparatus ( Diagenode ) for 3 cycles of 1 min on at high intensity ( 200 W ) and 30 sec off , followed by 4 cycles of 1 min on at medium intensity ( 160 W ) and 30 sec off . The size of the sheared chromatin was around 200 to 1 , 000 bp as determined by agarose gel electrophoresis . After immunoprecipitation with an appropriate antibody , DNA fragments were recovered by Proteinase K treatment . Indexed ChIP-seq libraries were prepared with the NEBNext ChIP-Seq Library Prep Master Mix Set for Illumina ( New England Biolabs ) according to the manufacturer’s instructions . Fragment size selection of ChIP-seq libraries was done using Agencourt AMPure XP beads ( Beckman Coulter ) . The products were purified and enriched with PCR to create the final double stranded cDNA library . The MiSeq system ( Illumina ) was used to sequence the cDNA library . RNA extraction and cDNA preparation were carried out as described previously with a few modifications [31] . Total RNA was isolated using Sepasol RNA I Super ( Nakalai Tesque ) , and used for cDNA synthesis using ReverTra Ace qPCR RT master mix with genomic DNA remover KIT ( Toyobo ) . Depletion of rRNA was performed using the Ribo-Zero rRNA removal kit for human/mouse/rat ( Epicentre ) . Indexed RNA-seq libraries were prepared with the NEBNext Ultra™ RNA Library Prep Kit for Illumina kit ( New England Biolabs ) or NEXTflex™ Directional RNA-Seq Kit ( BIOO Scientific Corp . ) according to the manufacturer’s instructions . Fragment size selection of RNA-seq libraries was done using Agencourt AMPure XP beads . The products were purified and enriched with PCR to create the final double stranded cDNA library . The MiSeq system ( Illumina ) was used to sequence the cDNA library . The RNA-seq and ChIP-seq reads ( 75–120 bp ) were mapped to the genome of the Magnaporthe oryzae strain 70–15 ( release 8 . 0 , http://www . broadinstitute . org/ ) using TopHat v2 . 0 . 10 [59] and bwa v0 . 6 . 2-r126 [60] , respectively . RNA-Seq and ChIP-seq data were visualized in the Integrative Genomics Viewer genome browser [61] . The edgeR package [62] for R v3 . 0 . 1 [63] was used for TMM normalization [64] , identification of differentially expressed genes ( DEGs ) from RNA-seq data , and detection of genes differentially enriched for histone modifications from ChIP-seq data with the corrected p-value [65] cutoffs of 0 . 01 ( ChIP-seq analysis ) or 0 . 001 ( RNA-seq analysis ) . The GOstats package [66] was used to identify statistically significant enriched Gene Ontology ( GO ) categories . | This paper provides two major contributions to the field of genetics . First , we systematically studied the biological roles of eight histone lysine methyltransferase ( KMT ) genes in the phytopathogenic fungus Magnaporthe oryzae . We investigated their roles , especially focusing on their involvement in infection-related morphogenesis and pathogenicity . The results showed that the eight KMTs were involved in various infection processes to varying degrees , and that MoSET1 , one of the KMTs catalyzing methylation at histone H3 lysine 4 ( H3K4 ) , had the largest impact on the pathogenicity of the fungus . Second , we focused on the role of MoSET1 in global gene regulation . H3K4 methylation is generally believed to be an epigenetic mark for gene activation in higher eukaryotes . However , in Saccharomyces cerevisiae , SET1 was originally characterized as being required for transcriptional silencing of silent mating-type loci . We addressed this apparent discrepancy by examining genome-wide gene expression and H3K4 methylation during infection-related morphogenesis in M . oryzae . RNA-seq analysis of a MoSET1 deletion mutant revealed that MoSET1 was indeed required for proper gene activation and repression . ChIP-seq analyses of H3K4 methylation and MoSET1 suggested that MoSET1 could directly play a role in gene activation while MoSET1-dependent gene repression may be caused by indirect effects . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | MoSET1 (Histone H3K4 Methyltransferase in Magnaporthe oryzae) Regulates Global Gene Expression during Infection-Related Morphogenesis |
Proteomics techniques can identify thousands of phosphorylation sites in a single experiment , the majority of which are new and lack precise information about function or molecular mechanism . Here we present a fast method to predict potential phosphorylation switches by mapping phosphorylation sites to protein-protein interactions of known structure and analysing the properties of the protein interface . We predict 1024 sites that could potentially enable or disable particular interactions . We tested a selection of these switches and showed that phosphomimetic mutations indeed affect interactions . We estimate that there are likely thousands of phosphorylation mediated switches yet to be uncovered , even among existing phosphorylation datasets . The results suggest that phosphorylation sites on globular , as distinct from disordered , parts of the proteome frequently function as switches , which might be one of the ancient roles for kinase phosphorylation .
Protein phosphorylation is important for many cellular processes , including signalling ( e . g . [1] ) , transcription ( e . g . [2] ) and metabolism ( e . g . [3] ) . Many phosphorylation sites act as switches to regulate inter-protein interactions ( e . g . [4] ) and there have been many studies into mechanisms , specificities and structures of kinases , phosphatases ( e . g . [5 , 6] ) and recognition domains ( SH2 , 14-3-3 , etc . ) that regulate or bind them ( e . g . [7 , 8] ) . Phosphosites also regulate enzymatic function ( e . g . [9] ) , target proteins for degradation ( e . g . [10] ) and play many other intriguing roles , e . g . in ultrasensitivity of Sic1/Cdc4 interactions [11] or in RNA polymerase II recognition during mRNA processing [12] . High-throughput efforts have identified thousands of phosphosites in many biological systems [13–16] . Few of them overlap with those identified in low-throughput studies ( e . g . [17] ) meaning that the molecular consequences of phosphorylation are not understood for most sites . Previous analyses have shown functional sites to be generally conserved [18] and over-represented in disordered regions [19 , 20] . Functional phosphosites have been proposed to have evolved from negatively charged amino acids , by making charge-mediated protein interactions tunable by kinases [21] . Functional coupling and/or co-evolution of sites has been suggested to be an important determinant of protein function [20 , 22] , with codes of post-translational modifications refining protein function , for example in transcription factors [23 , 24] . While many important proteins are known to be modified at multiple sites , the functional implications of these codes are understood for only a handful . There are now many thousands of three-dimensional ( 3D ) structures of protein-interactions [25–28] , providing an invaluable resource to study molecular mechanisms . These include structures of phosphorylated proteins and structures on which phosphosites from homologous proteins can be modelled . Phosphosites in known structures tend to be conserved when they occur at interfaces and only a minority of these alter binding affinity[29] . Mechanistic investigations show that certain phosphosites target interfaces , thus enabling predictions of function ( e . g . [30] ) . The now increased volume of both phosphoproteomic and 3D structure data provides an opportunity to study and predict the mechanistic impact of phosphosites on protein interfaces . Accordingly , we present here an approach to identify potential phosphosite switches , using structures of phosphorylated proteins and of their homologues , and to predict whether they turn interactions on or off . From a large phosphosite dataset we predict hundreds of new switches , a selection of which , via mutations to phosphomimics , we demonstrate are likely responsible for mediating protein-protein interactions .
To search for new potential switches we used a processed dataset of 223 , 971 phosphosites in 19 , 483 proteins from five organisms , defining the 1 . 6 million to date unphosphorylated Serine , Threonine and Tyrosine residues in the same proteins as background ( Fig 1A ) . The vast majority of known sites ( >90% ) come only from high-throughput studies , meaning their particular functions and consequences have not been studied in any detail . The majority ( 55% ) of the phosphosites are in disordered regions , as noted previously [19 , 31] , which is significantly higher than the background ( Fig 1B , 32% , P << 0 . 01 ) . 56 , 209 sites ( 25% ) , including 8341 ( 7% ) of those in disordered regions , could be matched to 3D structures , either of the protein itself or a homolog [32] . 8714 ( 16% ) phosphosites were within contacting distance of a small molecule ( more than background: 16% vs 13% P << 0 . 01 ) , including some known enzymatic switches ( e . g . [33] ) , though the majority have no known functional role . Whether these sites are regulatory or trapped phosphoenzyme intermediates requires additional investigation . Phosphosites are more likely to lie on protein surfaces ( 90% vs 87% , P << 0 . 01 , Figs 1B & S1 ) , to be at protein-protein interaction interfaces ( 10% vs 6% , P << 0 . 01 ) and , when at an interface , to be conserved or aligned to Aspartate or Glutamate in orthologues ( P << 0 . 01 , S2 Fig ) . A total of 34 sites at interfaces are aligned to at least 50% Aspartate/Glutamate residues , supporting the idea ( e . g . [21] ) that some sites have evolved from negative residues to modulate protein interactions . Only 1455 sites ( 0 . 7% ) are matched to phosphorylated residues visible in at least one 3D structure and only 122 of these are at interaction interfaces ( i . e . potential switches ) , emphasizing that few sites are understood in any mechanistic detail . We defined phosphosite-switches as Serine , Threonine and Tyrosine residues in protein interfaces that make interactions stronger ( enabling ) or weaker ( disabling ) through interplay between the physicochemical properties of the modification and the interface . To identify such sites we first computed a set of pair-potential scores that compare the frequency of pairs of contacting residues at interfaces to a random model ( Fig 1C , S5 Table ) , summed the differences in scores between phosphorylated and unmodified residues to give the Interaction Effect ( IE ) , and defined enabling as those where the IE increases upon phosphorylation ( i . e . a better interaction according to statistical preferences ) and disabling where it decreases [32] . Accuracy of interface structures is proportional to the sequence similarity between the protein of interest and the 3D template used to model it [28] , and our identified sites span the entire range of sequence identities . Similarly , the likelihood that a phosphosite is a true switch will increase with the degree to which it is conserved across orthologous sequences [20] . To account for both of these effects , we multiplied IE by the similarity between the protein and the 3D template ( fraction of identical residues , fID ) and the site conservation across orthologues ( fraction of residues that are either conserved or Aspartate or Glutamate , fCons ) to give an overall score Sswitch , where high positive/negative values indicate the best switch candidates . We benchmarked Sswitch using known phosphosite-switched interactions extracted from UniProt and PhosphoSitePlus [34] . These sets are biased towards enabling sites ( S1 Table ) since most sites are related to gain of interaction upon phosphorylation . Incorporation of the measures of structural match quality and residue conservation improves performance , though only marginally , perhaps reflecting the variability of sites and the relatively weak conservation of sites outside of closely related species ( Fig 1D ) . We also observed that absolute Sswitch is better able to find any effect , disabling or enabling , than are structural match quality and residue conservation by themselves ( S3 Fig ) , suggesting that conserved phosphosites seen directly in protein-protein interfaces may play roles other than switching . Values of Sswitch ≥ 1 . 7 or ≤ -1 . 7 give a false positive rate = 0 . 05 with reasonable sensitivity ( = 0 . 35 ) , positive predictive value ( > 0 . 78 ) and accuracy ( 0 . 74 ) , and a very low p-value ( << 1 x 10−6 ) ( Fig 1D , S4 Fig , S6 Table ) . Attempts to improve performance using logistic regression ( see Methods ) slightly reduced the sensitivity to 0 . 33 ( but with the same accuracy ) at our desired false positive rate ( 0 . 05; See S4 Fig and S6 Table ) . We believe this to be a function of the small benchmark rather than any issue with the regression approach; a larger benchmark would likely lead to an improved performance . To check for possible bias towards enabling sites from kinase-substrate interactions , we removed kinase interactors from the benchmark set ( see Methods ) and re-calculated the benchmark statistics , resulting in a slightly increased sensitivity ( 0 . 39 ) and the same accuracy ( for the desired false positive rate ( 0 . 05; See S7 Table ) at the cost of an increased Sswitch threshold . To separate the effect of using homologous structures from the prediction of effects on interactions , we re-calculated the benchmark statistics using only structures with a very high sequence identity ( > = 99% ) to the proteins in question . This gave a slightly higher sensitivity ( 0 . 42 ) but with lower accuracy ( 0 . 65 ) and p-value ( 0 . 0001 ) for the desired false positive rate ( 0 . 05; See S8 Table ) , which we believe to be a function of the reduced size of the benchmark . Finally , to allow for different thresholds for predicting enabling and disabling sites , we split the benchmark in to these two classes and analysed Sswitch separately . For our target false positive rate of < = 0 . 05 , enabling and disabling sites gave sensitivities of 0 . 37 and 0 . 24 respectively , accuracies of 0 . 76 and 0 . 67 respectively , and p-values of << 1 x 10−6 and 0 . 01 respectively ( See S9 and S10 Tables ) . These differences probably reflect the larger number of enabling sites in the benchmark . Here , for simplicity and the reasons given above , we used the simple Sswitch score with a threshold calculated from our combined benchmark . We did not use the optimised classifier , the kinase-deficient or homologue deficient benchmarks , or the separate disabling and enabling benchmarks . Hereafter , we only consider enabling or disabling sites above/below this threshold unless otherwise mentioned . The majority of significant sites have comparatively high sequence identities as might be expected by the nature of the score ( >70% have >90% sequence identity , S2 Table ) . There are other methods to calculate or predict the effect of mutations or modifications on protein interactions . Most of these use protein structures of interacting proteins to compute ΔΔG values ( i . e . the change of the interaction Gibbs free energy comparing wild-type and modified interactions ) . We compared our Sswitch score to ΔΔGs calculated by FoldX [35] on models we built with Modeller [36] using default parameters . These ΔΔGs were a poor predictor of effects on interactions ( True positive rate = 0 . 01 for a false positive rate 0 . 05; S6 Table , S4 Fig ) , highlighting the probable need for manual intervention to get the best results from modelling and energy calculations . For example , the Dynein Ser-88 phosphosite that we predict and is also known to disable homodimerisation ( see below ) is predicted by FoldX to have a negative ΔΔG ( i . e . a more favorable interaction ) . Inspection shows that the FoldX optimized structure has the two phosphate groups pointing away from each other and accommodated in the dimeric structure instead of pointing towards each other which would prevent dimerisation ( S5 Fig ) . It is possible that more careful consideration of each interface would give better results using FoldX , though this is not practical for the many thousands of Phosphosites considered here . Considering the 5690 phosphosites at protein-protein interfaces ( S2 Table ) , Sswitch predicts 827 ( 15% ) to be enabling and 255 ( 4% ) to be disabling , fractions significantly higher than background ( P << 0 . 01 , Fig 1B ) . Among these are several known enabling switches , such as the Syk Tyrosine kinase SH2 domain bound to an immunoreceptor activation motif [37] ( Fig 2A ) and Serotonin N-acetyltransferase bound to 14-3-3 zeta [38] . There are also known disabling sites , such as Dynein light chain Ser-88 , which is adjacent to a Glutamate and a copy of itself at the dimer interface [39] ( Fig 2B ) and where phosphorylation leads to inactive monomers [40] . Ser-429 in Mdm2 is also correctly predicted to disable oligomer formation [41] . Of the 123 sites matched to phosphorylated residues visible in at least one 3D interaction interface , 72 are enabling and only two are disabling ( the rest are neutral ) . Most predicted switches are unknown , including the weakly disabling PKC phosphosite in Glutamate receptor subunit zeta-1 , which lies in a negatively charged interface with its regulator Calmodulin ( Fig 2C ) . This has a high negative IE ( -6 . 74 ) but is poorly conserved ( fCons = 0 . 1 ) resulting in an Sswitch of -0 . 7 , below the threshold . Examination of the eggNOG group from which fCons was calculated shows that the majority of the 315 sequences to which this protein was aligned do not align at this point , giving a low fCons . Of those that do , 44% have Threonine at this position . Novel enabling sites are possibly more difficult to identify since phosphorylation might be required to determine a structure . However , many interactions of known structure are low affinity ( possibly half are > 1μM; one third are > 50μM [46] ) and high protein concentrations used in structure determination can produce structures without all features necessary for biological interactions . Analysis of our dataset supports this: of the 522 non-redundant phosphosites ( in all species ) at interfaces that are seen to be phosphorylated in a 3D structure , 16 are unphosphorylated in at least one homologous interface ( S3 Table ) . Thus there are also interesting candidate enabling switches , such as Tyr-65 in human Dynein light chain , predicted to strongly enable homodimer formation by interacting with lysine residues at the interface [39] ( Fig 2D ) . These predicted switches could also be more subtle changes to affinity than ( e . g . ) SH2 or 14-3-3 domain binding sites , perhaps enhancing or diminishing an interaction that would occur anyway . Of the 5690 non-redundant sites at protein-protein interfaces , 3225 ( 57% ) represent individual sites that are involved in interactions with multiple partner proteins and 55 represent individual sites that are enabling for one interaction and disabling for another ( with another six non-redundant sites being enabling in one protein and disabling in another ) , suggesting that phosphorylation selects interaction partners . For example , phosphorylation of Tyr-32 of the GTPase CDC42 appears to enable the ARHGAP1 interaction and disable that with the GEF MCF2L ( Fig 2E & 2F ) . Mutation of Tyr-32 in CDC42 is known to abolish exchange activity with GEFs [47] , though it is unclear how phosphorylation is involved in this process . As the set of known phosphosites is incomplete [20] , it is likely that many of the background sites are phosphorylated under conditions not yet tested . We thus searched for additional potential switches among these 1 . 6 million sites . Of these , 31 , 815 are at a protein-protein interface , of which just 2730 ( 9% ) would , if phosphorylated , be enabling , 780 ( 2% ) would be disabling and 78 ( 0 . 2% ) would enable some interactions and disable others in the same species . Among these is Ser-1055 in the Apoptosis-stimulating of p53 protein 1 , which lies in a long loop directly at the interface with TP53 and interacts with Arg-273 and Arg-248 ( Fig 2G ) . which are mutated in many human cancers [45] . This Serine , which is Aspartate in the closely related TP53BP2 , lies in a stretch of three to four Glutamate or Asparate residues in both proteins and is predicted to be a possible Casein kinase phosphorylation site [48 , 49] . We tested twenty sites with a range of Sswitch scores , including known or predicted switching by 13 phosphosites and seven background sites using the yeast two-hybrid system . Based on the few known disabling examples ( e . g . Dynein Ser-88 above ) , we selected five sites ( regardless of switch score ) for which phosphorylated residues were close to copies of themselves at a homodimer interface . Interestingly , the residue-residue parameters disfavour interactions between unphosphorylated residues ( particularly Serine & Threonine ) almost as much as between phosphorylated equivalents ( Fig 1C ) , suggesting that their adjacency alone would be insufficient to disable an interface ( and indeed at least one of these instances is weakly enabling , see SAT1 below ) . We compared the interactions of the natural sequence to those with mutations of the site to Glutamate ( commonly used as a phosphosite mimic ) or Alanine using the two-hybrid system . Nine of 20 interactions considered gave positive results when using the wild-type clones , a proportion that broadly agrees with the expected sensitivity of the two-hybrid system [50] . Of the sites tested by mutagenesis , four showed definite switching behaviour and five did not ( S4 Table ) . Perhaps highlighting the difficulties in predicting/identifying enabling switches ( see above ) , four of five instances where growth was seen ( suggesting an interaction ) , but no difference could be perceived between wild type and phosphomimic , were predicted enablers ( though this finding is not significant; p<0 . 3 by a hypergeometric distribution ) . Additionally , while the pair-potential for Glutamate-Glutamate interactions ( i . e . our phosphomimetic ) is similar to that for pairs of phosphorylated residues except phosphotyrosine ( S5 Table ) , it is also known that Glutamate is an imperfect mimic , particularly for tyrosine-phosphate [51] , but also for Serine or Threonine . Indeed , switching behavior for Thr-31 in AANAT/YWAZ ( S5 Table ) is known to be more apparent when using a chemical phosphomimetic instead of Glutamate [52] . For the known disabling Ser-88 in Dynein ( above ) both the wild-type and alanine mutants are able to interact , with the Glutamate mutant abolishing the interaction as known ( Fig 3A ) . High-throughput studies in human [53] and yeast [54] identify Ser-68 in yeast Adenine phosphoribosyltransferase from the purine nucleotide salvage pathway to be phosphorylated , and the assay confirms our prediction of a weak disabler ( Fig 3B ) . Another high-throughput site Ser-149 in human diamin acetyltransferase 1 ( SAT1 ) is also enabling as predicted ( Fig 3C ) , with the phosphomimic showing a stronger interaction than wild-type . We also predicted that phosphorylation of Thr-68 of DNA fragmentation factor A ( DffA ) would enable interactions with DffB . This site is not known to be phosphorylated ( i . e . it is a background site ) , though other sites in the same protein have been identified , including Tyr-75 [34] from the same interface loop . The site does appear to modulate the interface , but is surprisingly disabling ( Fig 3D ) . Inspection shows that the two lysines giving rise to the enabling score are oriented in a way that might preclude effective interactions with the phosphate group and that moreover might lead to steric clashes .
This study is the first large-scale investigation of phosphosites within interacting 3D structures , and has identified hundreds of potential interaction switches . These provide an immediate starting point for additional studies into proteins , interactions and processes affected by such modifications . The phosphoproteome has been estimated to be no more than 22% complete [58] . By this estimate there could be in excess of 4000 enabling or disabling switches across the species we investigated . New candidate switches will be a boost for efforts to unravel the complexity of PTM codes that are critical for fine tuning cellular processes [20] . The fact that so many phosphosites come from high-throughput studies makes structural/mechanistic tools like that presented here important to rank , filter and interpret these data as suggested previously [59] . As with many new technologies in the life sciences , interpretation increasingly lags behind data generation . Our method to predict the direction of the effect of phosphorylation on a protein-protein interface correctly identified several real enabling or disabling sites , though in some instances we saw no effect or switching in the direction opposite to our predictions . The simple metric does not yet consider the complexities of protein structures , such as conformational rearrangements and steric clashes , multi-faceted interfaces and complex regulation , nor coupling with other modified sites , which determine how phosphorylation might ultimately affect an interaction . It would also benefit from a larger benchmark set of phosphosites known to affect protein-protein interactions , phosphosites known not to affect protein-protein interactions , and phosphosites seen directly in protein 3D structures with which we can parameterise our pair-potential scores . The occurrence of many potential switches in ordered protein regions is surprising given the widely held view that phosphoregulation , particularly in eukaryotes , is predominantly a disordered phenomenon . Indeed , the observation of so many phosphorylation sites at the junction between globular proteins in Eukaryotes ( this study ) and Prokaryotes [60] and the apparent lack of phosphopeptide binding domains in the latter , suggests that regulation of globular interfaces could be an ancient role for Serine/Threonine kinases , which later diversified into the complex mechanisms—involving disorder and recognition modules—seen in Eukaryotes today .
We took phosphoproteins in five eukaryotes ( H . sapiens , M . musculus , D . melanogaster , C . elegans , S . cerevisae ) from a previous study [24] and identified 258 , 552 phosphosites in PhosphoSitePlus [61] , UniProt [62] ( those with experimental evidence only ) , dbPTM [63] and phospho . ELM [64] . We also extracted phosphorylated Serine , Threonine and Tyrosine residues within known 3D structures [65] which we mapped to UniProt sequences through MUSCLE [66] sequence alignments of SIFTS [67] pairs of PDB and UniProt sequences . For each phosphosite we defined high throughput sites as those seen only in publications reporting 100 or more phosphoproteins . We defined background sites as all 2 , 068 , 843 unphosphorylated Serines , Threonines and Tyrosines in the same set of proteins . To avoid over-counting because of redundancy from sites with equivalents in closely homologous proteins , we grouped all sites ( both phosphosites and background ) according to their positions in alignments of UniProt UniRef50 sequence groups [68] . We considered potential background sites that were aligned to real phosphosites to be ambiguous and ignored them in our counts and predictions . To avoid grouping poorly aligned sites , we did not group aligned sequences where the number of gaps divided by the sequence length was > = 0 . 09 ( a value deduced by inspection of several hundred phosphoprotein alignments ) . This gave 223 , 971 and 1 , 611 , 565 non-redundant phosphosites and background sites respectively . We mapped the sequences and sites described above to 3D structures , including interactions with proteins and small-molecules , using Mechismo [32] which uses a non-redundant set of 3D structures of interactions in PDB biological assemblies [69] , considers structures of homologues as well as the actual protein in question and transfers positional information via sequence alignments . We used the ‘low’ stringency setting , which identifies the best possible protein-interface for any pair of proteins that interact physically or for which an interaction is known for closely homologous proteins . This setting includes any possible interface of known structure as identified by sequence comparison . In practice , few low identity interfaces are used as the Sswitch score ( below ) down-weights switches arising from more remote homologues . As in Mechismo itself , we do not construct protein models , but transfer residue contacts from the template structure to a target sequence ( even if matched amino acids are different ) . In cases where multiple templates were available for a site at a particular interface ( as a result of different alignments between UniProt and the PDB , which can come from SIFTS or from BLASTP within Mechismo ) , we took the most significant score ( either enabling or disabling ) . 3D interaction structures with phosphorylated Serine , Threonine or Tyrosine ( PDB SEP , TPO and PTR ) residues seen directly in interfaces , from any species , were compared to similar interfaces ( at least 50% sequence identity across at least 50% of the sequence , and at least 50% interface residues in common after alignment ) to identify homologous interactions with unphosphorylated residues at the equivalent position . Multiple phosphorylated residues at the same position in the same interface group were counted only once . We defined intrinsically disordered residues as those where the mean IUPred long disorder [70] of the matching fragment residue over a sliding window of eleven residues was ≥ 0 . 5 . We defined residues as buried when the side-chain accessible surface area of the aligned residue in the structural template was < 5Å2 and exposed otherwise ( using NACCESS [71] ) . We defined the switch score as: Sswitch=IExfIDxfCons Where IE ( Interaction Effect ) is the sum of changes in residue pair-potentials upon phosphorylation [32] ( S5 Table ) , fID is the minimum of the fraction of identical residues in the alignment of either sequence with its structural template , and fCons is the fraction of sequences in the alignment of the animal or fungus ( i . e . opisthokont ) eggNOG 4 . 5 [72] orthologous group that have a residue of the same amino acid type or Aspartate or Glutamate aligned to the site . For homodimeric interactions , the site was assumed to be phosphorylated in both copies of the protein . For sites for which fCons was unavailable ( i . e . not aligned to any other sequence ) , we used the average fCons of all Serines , Threonines and Tyrosines in proteins of the same species . We defined the positive benchmark set by extracting all 1339 phosphosites from UniProt ‘MOD_RES’ records from the species studied here and where the annotated text gave indications of binding/interaction ( “bind*” or “interact*” ) and/or mentioned multimerisation or at least one additional protein by gene name . We then inspected these and marked relationships as enabling , disabling , phosphorylation/dephosphorylation or unknown which left 795 phosphosite-interaction pairs in 222 proteins . We also downloaded regulatory sites from PhosphoSitePlus [34] and extracted protein interaction pairs marked as being induced or disrupted by a phosphosite , given 5225 interaction pairs involving 3323 sites in 1588 proteins from 13 species . We defined the negative benchmark set by shuffling positions in this set , along with their interactors and the given effect , to a random position in the same protein and did this ten times for each site . In doing so we preserved the distribution of surface exposures of these sites as described previously [32] . This gave 41813 site-interaction pairs involving 28441 sites in the same set of proteins . We mapped the benchmark sites and their interactors to interaction structures and discarding unmapped pairs , leaving 122 unique positives and 224 negatives ( S1 Table ) . We then evaluated classifier performance using the R package 'ROCR' [73] . To account for possible bias towards enabling sites from kinase-substrate interactions , we classified all interactors as kinases when they matched to a protein kinase domain in Pfam [74] ( specifically , Pfam accession PF00069 ) and re-calculated the benchmark statistics using this reduced set . To optimise the combination of IE , fID and fCons , we applied logistic regression to our benchmark using R [75] . We balanced the benchmark data by randomly undersampling the negative set , ran five-fold cross-validation , repeated this 100 times , and took the means of the following summary statistics to evaluate the model: Area Under the Curve ( AUC ) , threshold that gave a False Positive Rate ( FPR ) of < = 0 . 05 , and the accuracy , True Positive Rate ( TPR ) , True Negative Rate ( TNR ) and Positive Predictive Value ( PPV ) at this threshold . We then applied logistic regression to the full benchmark set . For each phosphosite interaction in our benchmark , using the same template structure as for Sswitch , we used Modeller [36] to build a model of the unphosphorylated interaction and FoldX [35] to produce the phoshorylated version . We then used FoldX to calculate the ΔΔG between these two models . We calculated the significances of the differences of distributions ( accessible surface area , fCons ) of phosphosites and of background sites with Wilcoxon-Mann-Whitney rank sum tests . We used chi-square tests to calculate the significances of the differences in the fractions of phosphosites and of background sites under the various binary classifiers ( ordered , mapped to structure , exposed , in an interaction interface , and enabling or disabling ) . In all cases , P was << 0 . 01 . We calculated p-values for the selected score thresholds on the benchmark using a two-sided Fisher's exact test . A total of 70 open reading frames encoding putative phospo-switchable proteins and their interactors were obtained as sequence optimised synthetic clones flanked by attb-Gateway sites ( GeneArt/ Invitrogen ) . All clones were Gateway-cloned into the Donor vectors pDONR221 or if necessary into pDONR/Zeo by Gateway BP-reaction and subsequently by LR-reaction into the Y2H bait and prey vectors pDEST32 and pDEST22 respectively for the Yeast two Hybrid experiments . All constructs were sequence verified . All code is available from the Mechismo website , mechismo . russelllab . org/downloads . We performed two-hybrid assays following an altered “Testing specific Two-Hybrid interaction” protocol of the ProQuest™ Two-Hybrid System Handbook ( Invitrogen ) . Briefly , all interaction pairs ( wild-type , Glutamate- and Alanine-mutants ) were double-transformed into yeast strain MaV203 ( Invitrogen , MaV203 Competent Yeast Cells , Library Scale cat# 11281–011 ) . Colonies from each transformation were grown on 15-cm plates of synthetic complete media lacking leucine and tryptophan ( Sc-Leu-Trp ) . After 2–3 days 3 individual colonies of each transformation were picked and suspended in 100 μl autoclaved saline in a 96-well PCR plate . From here they were replicated by 96-needle replicator onto rectangular SC-Leu-Trp agar plates lacking histidine and containing three different concentrations ( 10 , 25 , 50 mM ) of 3-aminotriazol ( 3AT ) . 2–5 days after plating interaction phenotypes were assessed . For phosphotyrosine sites we also tested the Tyrosine to Alanine-Glutamate mutation which is proposed to be a better mimic of phosphotyrosine [51] . For homodimeric interactions , colonies where both copies of the protein contained the phosphomimetic were examined . | Most biological processes occur by molecules connecting to other molecules , and the precise details of these connections can often be seen in their three-dimensional structures or inferred from those of similar molecules . The ways in which molecules fit together are often affected and regulated by small chemical modifications to the structures of the molecules . Thousands of these modifications have been found in large-scale experiments , without knowing what connections they might affect or how . Some make molecules fit together better and some make the fit worse . We have combined 3D structures with data for a particular type of modification known as 'phosphorylation' to predict these effects and have found more than a thousand phosphorylations that may strengthen or weaken molecular connections , thereby allowing us to explain how certain biological processes are regulated . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"neurochemistry",
"protein",
"interactions",
"neuroscience",
"mathematics",
"forecasting",
"statistics",
"(mathematics)",
"protein",
"structure",
"prediction",
"protein",
"structure",
"neurotransmitters",
"research",
"and",
"analysis",
"methods",
"sequence",
... | 2017 | Systematic identification of phosphorylation-mediated protein interaction switches |
Positive-stranded RNA viruses replicate inside cells and depend on many co-opted cellular factors to complete their infection cycles . To combat viruses , the hosts use conserved restriction factors , such as DEAD-box RNA helicases , which can function as viral RNA sensors or as effectors by blocking RNA virus replication . In this paper , we have established that the plant DDX17-like RH30 DEAD-box helicase conducts strong inhibitory function on tombusvirus replication when expressed in plants and yeast surrogate host . The helicase function of RH30 was required for restriction of tomato bushy stunt virus ( TBSV ) replication . Knock-down of RH30 levels in Nicotiana benthamiana led to increased TBSV accumulation and RH30 knockout lines of Arabidopsis supported higher level accumulation of turnip crinkle virus . We show that RH30 DEAD-box helicase interacts with p33 and p92pol replication proteins of TBSV , which facilitates targeting of RH30 from the nucleus to the large TBSV replication compartment consisting of aggregated peroxisomes . Enrichment of RH30 in the nucleus via fusion with a nuclear retention signal at the expense of the cytosolic pool of RH30 prevented the re-localization of RH30 into the replication compartment and canceled out the antiviral effect of RH30 . In vitro replicase reconstitution assay was used to demonstrate that RH30 helicase blocks the assembly of viral replicase complex , the activation of the RNA-dependent RNA polymerase function of p92pol and binding of p33 replication protein to critical cis-acting element in the TBSV RNA . Altogether , these results firmly establish that the plant DDX17-like RH30 DEAD-box helicase is a potent , effector-type , restriction factor of tombusviruses and related viruses . The discovery of the antiviral role of RH30 DEAD-box helicase illustrates the likely ancient roles of RNA helicases in plant innate immunity .
Positive-stranded ( + ) RNA viruses replicate inside cells and depend on many co-opted cellular factors to complete their infection cycle . These viruses build elaborate membranous viral replication compartments , consisting of viral replication proteins , viral RNAs and recruited host factors , in the cytosol of the infected cells . The hijacked host factors participate in all steps of RNA virus replication , including the assembly of membrane-bound viral replicase complexes ( VRCs ) , viral RNA-dependent RNA polymerase ( RdRp ) activation and viral RNA synthesis . The growing list of co-opted host factors facilitating VRC assembly includes translation initiation and elongation factors , protein chaperones , RNA-modifying enzymes , SNARE and ESCRT proteins , actin network , and lipids [1–9] . Many ( + ) RNA viruses extensively rewire metabolic pathways , remodel subcellular membranes and take advantage of intracellular trafficking . The host utilizes cellular proteins to sense viral pathogenicity factors and block virus replication with the help of cell-intrinsic restriction factors ( CIRFs ) as an early line of defense [2 , 10–12] . These CIRFs can be part of the innate immune responses and used for antiviral defense as sensors or effectors [13–16] . The identification and characterization of the many CIRFs against different viruses is still in the early stages . Viral RNA replication is intensively studied with Tomato bushy stunt virus ( TBSV ) , a tombusvirus infecting plants , based on yeast ( Saccharomyces cerevisiae ) surrogate host [17–19] . Expression of the two TBSV replication proteins , termed p33 and p92pol , and a replicon ( rep ) RNA leads to efficient viral replication . p92pol is the RdRp [20 , 21] , whereas the more abundant p33 is an RNA chaperone . P33 functions in RNA template selection and recruitment and in the assembly of VRCs within the replication compartment [21–26] . TBSV , which does not code for its own helicase , usurps several yeast and plant ATP-dependent DEAD-box RNA helicases as host factors promoting TBSV RNA replication . The yeast DDX3-like Ded1p and the p68-like Dbp2p , and the plant DDX3-like RH20 , DDX5-like RH5 and the eIF4AIII-like RH2 DEAD-box proteins were shown as pro-viral factors , which affect plus- and minus-strand synthesis , maintenance of viral genome integrity and RNA recombination in TBSV [27–29] . DEAD-box helicases are the largest family of RNA helicases and are known to be involved in cellular metabolism [30–32] , and affect responses to abiotic stress and pathogen infections [33–35] . They function in unwinding of RNA duplexes , RNA folding , remodeling of RNA-protein complexes , and RNA clamping [36] . They have no unwinding polarity and can open up completely double-stranded RNA regions , however , unlike many other helicases , DEAD-box helicases do not unwind RNA duplexes based on translocation on the RNA strand . Instead , DEAD-box helicases directly load on duplexes and open up a limited number of base pairs . Strand separation within the duplexes is not coordinated with ATP hydrolysis , which is used for enzyme dissociation from the template . This unwinding mode is termed local strand separation [36 , 37] . DEAD-box helicases also affect RNA virus replication [38–41] , and viral translation [42 , 43] . In case of plant viruses , turnip mosaic virus and brome mosaic virus have been described to co-opt cellular DEAD-box helicases for proviral function in translation or replication [39 , 42 , 44] . Altogether , cellular helicases are important co-opted host factors for several viruses , playing critical roles in virus-host interactions . However , cellular RNA helicases also act as antiviral restriction factors , including functioning as viral RNA sensors ( e . g . , Dicer or RIG-I ) or directly inhibiting RNA virus replication as effectors [45–47] . For example , DDX17 restricts Rift Valley fever virus [48] , while DDX21 helicase inhibits influenza A virus and DDX3 blocks Dengue virus infections [49–52] . Thus , the emerging picture is that host helicases are important for the host to restrict RNA virus replication , but the mechanism of their activities or substrates are not well characterized . In this work , we find that the plant DDX17-like RH30 DEAD-box helicase plays a strong restriction factor function against tombusviruses and related plant viruses . RH30 DEAD-box helicase is expressed in all plant organs , but its cellular function is not known yet [53] . We find that RH30 is re-localized from the nucleus to the sites of tombusvirus replication via interacting with the TBSV p33 and p92pol replication proteins . Several in vitro assays provide evidence that RH30 inhibits tombusvirus replication through blocking several steps in the replication process , including VRC assembly , viral RdRp activation and the specific interaction between p33 replication protein and the viral ( + ) RNA . RH30 knockout lines of Arabidopsis supported increased accumulation level for the related turnip crinkle virus , confirming the restriction factor function of RH30 against a group of plant viruses . This is the first identification and characterization of a plant helicase with an effector type restriction factor function against plant viruses . Since plant genomes codes for over 100 RNA helicases , it is likely that additional helicases have CIRF function against plant viruses .
To test if the host RH30 RNA helicase could affect tombusvirus replication , we expressed the Arabidopsis RH30 using agroinfiltration in Nicotiana benthamiana plants . Interestingly , expression of AtRH30 blocked TBSV replication by ~90% in the inoculated leaves ( Fig 1A ) . The closely-related cucumber necrosis virus ( CNV ) , which also targets the peroxisomal membranes for VRC formation , was also inhibited by ~4-fold through the expression of AtRH30 ( Fig 1B ) . Replication of another tombusvirus , carnation Italian ringspot virus ( CIRV ) , which builds the replication compartment using the outer membranes of mitochondria , was inhibited by ~9-fold by the transient expression of AtRH30 in N . benthamiana ( Fig 1C ) . To test if RH30 was also effective against TBSV when expressed in yeast cells , we launched the TBSV repRNA replication assay in wt yeast by co-expressing the viral components with RH30 . After 24 h of incubation , TBSV repRNA analysis revealed strong inhibition of viral replication by RH30 expression ( Fig 1F ) , suggesting that RH30 is a highly active inhibitor against TBSV replication even in a surrogate host . To learn if the putative helicase function of RH30 is required for its cell intrinsic restriction factor ( CIRF ) function , we expressed a motif IV helicase core mutant of RH30 ( F416L ) in N . bentamiana via agroinfiltration . Mutation of the highly conserved F residue within the helicase core domain ( see S1 Fig ) has been shown to greatly decrease both ATP binding/hydrolysis and strand displacement activities in Ded1 and other DEAD-box helicases [54] . Northern blot analysis revealed the lack of inhibition of TBSV replication , and only partial inhibition of CIRV replication by RH30 ( F416L ) ( Fig 1D and 1E , lanes 9–12 ) . Thus , we suggest that the full helicase/ATPase function of RH30 is required for its CIRF function against tombusviruses . VIGS-based silencing of the endogenous RH30 in N . benthamiana led to ~5-fold , ~3-fold and ~11-fold increased accumulation of TBSV , CNV and CIRV , respectively , in the inoculated leaves ( Fig 2 ) . The leaves of virus-infected and VIGS-treated plants showed severe necrotic symptoms earlier and died earlier than the control plants ( i . e . , TRV-cGFP treatment ) in case of all three tombusvirus infections ( Fig 2 ) . On the other hand , the VIGS-treated plants became only slightly smaller than the TRV-cGFP treated control plants ( Fig 2 ) . Based on these and the RH30 over-expression data , RH30 DEAD-box helicase seems to act as a major restriction factor against tombusviruses in plants and yeast . To identify the cellular compartment where RH30 DEAD-box helicase performs its CIRF function , first we used co-localization studies in N benthamiana protoplasts co-expressing GFP-RH30 , p33-BFP ( to mark the site of viral replication ) and RFP-tagged H2B , which is a nuclear marker protein . We detected the re-localization of GFP-RH30 into the large p33 containing replication compartment from the nucleus during CNV replication ( Fig 3A , top panel versus second panel ) . Both the p33-BFP and RFP-SKL ( a peroxisomal matrix marker ) showed the re-localization of GFP-RH30 into the large TBSV replication compartment , which consists of aggregated peroxisomes . Part of the ER is also recruited to the p33 and RH30 containing replication compartment ( Fig 3A bottom panel ) , as shown previously [55 , 56] . Similar re-localization pattern of RH30 was observed in epidermal cells of whole plants infected with CNV ( Fig 3B , top panel versus second panel ) . The expression of only p33-BFP was satisfactory to recruit the RH30 into the replication compartment ( Fig 3B ) . RH30 was also re-targeted in CIRV-infected N . benthamiana cells into the p36 and p95pol containing replication compartment ( Fig 3B , bottom panel ) , which consists of aggregated mitochondria [57 , 58] . Based on these experiments , we propose that the mostly nuclear localized RH30 helicase is capable of entering the tombusvirus replication compartment via interaction with the replication proteins . However , the formation of large tombusvirus-induced replication compartments seemed to be normal in the presence of RH30 , indicating the lack of interference with the biogenesis of the replication compartment by RH30 . To test if the cytosolic localization of RH30 is required for its CIRF function , we fused RH30 with a nuclear retention signal ( NRS ) [59] to enrich RH30 in the nucleus at the expense of the cytosolic pool of RH30 . Interestingly , unlike WT RH30 , expression of NRS-RH30 did not result in inhibition of TBSV replication in N . benthamiana ( Fig 4A ) . Confocal microscopy experiments confirmed that NRS-RH30-GFP is localized exclusively in the nucleus ( Fig 4B ) . Infection of the N . benthamiana protoplasts with CNV did not result in the re-targeting of NRS-RH30-GFP from the nucleus to the replication compartment visualized via p33-BFP . The nuclear retention of NRS-RH30-GFP was also confirmed in N . benthamiana epidermal cells infected with CNV or mock inoculated ( Fig 4C ) . Altogether , these experiments demonstrated that re-localization of RH30 helicase from the nucleus to the replication compartment is critical for its CIRF function in plants . To learn about the tombusviral target of RH30 DEAD-box helicase , we co-expressed the His6-tagged RH30 with Flag-tagged p33 and Flag-p92pol replication proteins and the TBSV repRNA in yeast , followed by Flag-affinity purification of p33/p92pol from the detergent-solubilized membrane fraction of yeast , which is known to harbor the tombusvirus replicase [20 , 60] . Western blot analysis of the affinity-purified replicase revealed the effective co-purification of His6-RH30 ( Fig 5A , lane 3 ) , suggesting that RH30 targets the VRCs for its CIRF function . Interestingly , His6-RH30 was co-purified from yeast co-expressing either Flag-p33 or Flag-p92pol replication proteins ( Fig 5A , lanes 1–2 ) , suggesting that RH30 likely directly interacts with the tombusvirus replication proteins in a membranous compartment . To show direct interaction between RH30 DEAD-box helicase and the TBSV p33 replication protein , we performed pull-down assay with MBP-tagged RH30 and GST-tagged p33 proteins from E . coli . We found that MBP-RH30 captured GST-p33 protein on the maltose-column ( Fig 5B , lane 2 ) , indicating direct interaction between the host RH30 and the viral p33 protein . In the pull-down assay , we used truncated TBSV p33 replication protein missing its N-terminal region including the membrane-binding region to aid its solubility in E . coli [61] . Interestingly , the helicase core mutant RH30 ( F416L ) also bound to p33 replication protein as efficiently as the wt RH30 ( Fig 5B , lane 3 versus 2 ) . Altogether , these data suggest that the direct interaction between RH30 host protein and the replication protein of TBSV occurs within the viral protein C-terminal domain facing the cytosolic compartment . To provide additional evidence that RH30 helicase interacts with the tombusvirus replication protein , we have conducted bimolecular fluorescence complementation ( BiFC ) experiments in N . benthamiana leaves . The BiFC experiments revealed interaction between RH30 and the TBSV p33 replication protein within the viral replication compartment , marked by the peroxisomal matrix marker RFP-SKL ( Fig 5C ) . Altogether , these experiments revealed direct interaction between the cellular RH30 DEAD-box helicase and the TBSV p33 replication protein , which results in re-targeting of RH30 into the viral replication compartment . To gain insight into the mechanism of CIRF function of RH30 helicase , we affinity-purified the recombinant RH30 and tested its activity in vitro in a TBSV replicase reconstitution assay , which is based on yeast cell-free extract [26 , 62] . Addition of RH30 to the replicase reconstitution assay led to inhibition of TBSV repRNA replication by ~10-fold ( Fig 6A , lanes 9–10 ) . The in vitro production of double-stranded repRNA replication intermediate was also inhibited by ~10-fold by RH30 , indicating that RH30 likely inhibits an early step , such as the VRC assembly during TBSV replication . We then used a step-wise TBSV replicase reconstitution assay [26 , 29] , in which RH30 was added at different stages of VRC assembly ( schematically shown in Fig 6B ) . RH30 showed significant inhibitory activity when added at the beginning of the TBSV replicase reconstitution assay ( Fig 6B , lanes 3–4 versus 1–2 ) . On the contrary , RH30 was ineffective , when added to TBSV replicase reconstitution assay after the VRC assembly step and prior to RNA synthesis ( Fig 6B , lanes 7–8 ) . These in vitro data support the model that the inhibitory role of RH30 is performed during or prior to the VRC assembly step , but RH30 is ineffective at the latter stages of TBSV replication . We also utilized an in vitro RdRp activation assay based on the purified recombinant TBSV p92 RdRp , which is inactive and requires Hsp70 chaperone and the viral ( + ) RNA template to become an active polymerase [21] . Addition of the recombinant RH30 helicase strongly inhibited the polymerase activity of the p92 RdRp ( Fig 6C ) , suggesting that RH30 blocks the critical RdRp activation step during tombusvirus replication . Several RNA helicases are involved in regulation of cellular translation [63] . Therefore , we tested if RH30 affected the translation of tombusvirus genomic RNA , which is uncapped and lacks poly ( A ) tail [64] . CIRV genomic RNA was used in this in vitro assay based on wheat germ extract [65] . Addition of recombinant RH30 to the in vitro translation assay inhibited slightly the production of p36 replication protein from the gRNA when RH30 was used in high amount ( Fig 6D ) . The highest amount of RH30 also had minor inhibition on translation of the control Tdh2 mRNA ( Fig 6D ) . Thus , RH30 is unlikely to specifically affect the translation of the tombusvirus RNAs during infection . Since the canonical function of RNA helicases to bind RNA substrates and unwind base-paired structures [36] , we tested if RH30 DEAD-box helicase could perform these functions with the TBSV RNA in vitro . First , we used gel-mobility shift assay with purified recombinant RH30 , which showed that RH30 bound to both the ( + ) and ( - ) repRNA ( Fig 7A and 7B ) . Since each of the four regions in the TBSV repRNA contains well-defined cis-acting elements , we performed template competition assay with the four regions separately in the presence of recombinant RH30 helicase . This assay defined that the best competitors for binding to RH30 was RII ( + ) and RII ( - ) , whereas RI ( + ) , RIV ( + ) and RI ( - ) , RIV ( - ) also become competitive when added in high amounts ( Fig 7C ) . Because RII ( + ) contains a critical cis-acting stem-loop element , termed RII ( + ) SL , which is involved in p33-mediated recruitment of the TBSV ( + ) RNA template [24] , and the activation of the p92 RdRp [21] , we tested if the purified RH30 could bind to this stem-loop element in vitro . Interestingly , RH30 bound to RII ( + ) SL in the absence of added ATP ( Fig 7D ) . However , the presence of extra ATP enhanced the binding of RH30 to RII ( + ) SL , suggesting that RH30 binds to RNAs in an ATP-dependent fashion , similar to other DEAD-box helicases [36 , 54 , 66] . The control p33 ( an N-terminally-truncated , soluble version ) bound to RII ( + ) SL more efficiently and in an ATP-independent manner ( Fig 7D ) , as also shown previously [24] . This highlights the possibility that RH30 and p33 replication protein compete with each other in binding to this critical cis-acting element . To test the RNA helicase function of RH30 , we performed strand separation assays , where parts of the TBSV repRNA was double-stranded as shown schematically in Fig 7E and 7F . The RNA helicase activity of RH30 in the presence of ATP was found to efficiently separate the partial dsRNA templates , involving RI and RII sequences ( Fig 7E and 7F ) . RH30 was much less efficient to separate the partial dsRNA templates in the absence of ATP or when we added its helicase core mutant RH30 ( F416L ) ( Fig 7E , lanes 6–9; 7F , lanes 5–8 ) . It is possible that the residual strand-separation activity of RH30 ( F416L ) might come from its RNA binding and RNA chaperone activity with the TBSV RNA substrates . Additional biochemical assays will be needed to test if the partial activity of RH30 in the absence of added ATP is due to the possibly copurified residual ATP bound to RH30 . To test if RH30 ( F416L ) helicase core mutant still has antiviral activity , we performed a TBSV replicase reconstitution assay with yeast cell-free extract [26 , 62] . Addition of RH30 ( F416L ) to the replicase reconstitution assay led to minor inhibition of TBSV repRNA replication ( Fig 7G , lanes 1–2 ) . Thus , mutation within the helicase core region of RH30 affected its antiviral activity on TBSV replication in vitro . To further characterize the restriction function of RH30 during tombusvirus replication , we tested if RH30 helicase could inhibit the selective binding of p33 replication protein to the viral RNA template in vitro . To this end , we biotin-labeled RII ( + ) sequence of the TBSV RNA , which represents RII ( + ) -SL RNA recognition element required for template recruitment into replication by p33 replication protein [24] . Moreover , RII ( + ) -SL RNA is also essential part of an assembly platform for the replicase complex [67] . The biotin-labeled RII ( + ) RNA was then pre-incubated with purified RH30 ( Fig 8A ) . Then , purified p33C ( the soluble C-terminal region , including the RNA-binding and p33:p33/p92 interaction region of p33 replication protein ) was added , which can bind specifically to RII ( + ) -SL if the hairpin structure with the C•C mismatch in the internal loop was formed [24] . After a short incubation , the biotin-labeled RII ( + ) RNA was captured on streptavidin-coated magnetic beads . After thorough washing of the streptavidin beads , the proteins bound to the RNA were eluted . Western blot analysis with anti-p33 antibody revealed that RH30 in the presence of ATP inhibited the binding of p33C to RII ( + ) -SL by 50% ( Fig 8A , lane 2 versus lane 3 ) when compared with the control containing the MBP protein that does not bind to RII ( + ) -SL [24] . RH30 was less inhibitory of the p33C—RII ( + ) -SL interaction in the absence of ATP ( Fig 8A , lane 4 ) . We also performed the experiments when RH30 and p33C were incubated with biotin-labeled RII ( + ) RNA simultaneously . Western-blot analysis showed that RH30 was still inhibitory of p33C binding to RII ( + ) -SL ( Fig 8B ) , but less effectively than above when RH30 was pre-incubated with the RII ( + ) RNA . These in vitro results suggest that one of the mechanisms by which RH30 helicase inhibits tombusvirus replication is to inhibit the binding of p33 to the critical RII ( + ) -SL RNA recognition element required for template recruitment into replication . This inhibition is likely due to local unwinding RII ( + ) -SL , because the presence of ATP enhanced the inhibitory effect of RH30 . In another set of experiments , we first incubated biotin-labeled RII ( + ) RNA with p33C , followed by capturing the RNA-p33 complex with streptavidin-coated magnetic beads and then , the addition of RH30 helicase to the beads ( Fig 8C ) . Here we tested the released p33C from the beads in the eluted fraction by Western blotting . Interestingly , increasing the amounts of RH30 added in the presence of ATP led to the release of p33C from the RII ( + ) RNA ( Fig 8C , lane 3–4 ) , whereas RH30 was less efficient in replacing p33C in the absence of ATP ( lanes 1–2 ) . Based on these in vitro data , we suggest that RH30 helicase could replace the RNA-bound p33C by likely remodeling the RNA-p33 complex in an ATP-dependent manner . We also tested the localization of RH30 helicase in comparison with the viral repRNA in N . benthamiana . The TBSV repRNA carried six copies of the RFP-tagged coat protein recognition sequence from bacteriophage MS2 in either plus or minus polarity [68] . CNV served as a helper virus in these experiments . Interestingly , RH30 was co-localized with both ( - ) repRNA and ( + ) repRNA , which were present in the replication compartment decorated by the TBSV p33-BFP ( Fig 9A and 9B ) . The RFP signal within the replication compartment was usually weaker when RH30 helicase was expressed , likely due to the inhibitory effect of RH30 on tombusvirus replication . Similar outcome was observed when the viral dsRNA replication intermediate , detected via dsRNA probes [69] , was co-localized with RH30 helicase within the replication compartment ( Fig 10 ) . These data demonstrate that RH30 helicase relocates to the replication sites where tombusvirus RNA synthesis takes place . To learn if RH30 has restriction function against additional plant viruses , we tested the effect of RH30 expression on TCV carmovirus and red clover necrosis mosaic virus ( RCNMV ) dianthovirus , both of which belong to the Tombusviridae family . Expression of AtRH30 in N . benthamiana plants led to complete block of TCV gRNA accumulation and ~4-fold reduction in RCNMV RNA1 accumulation ( Fig 11A and 11B ) . On the contrary , two separate transgenic RH30 knock-out lines of Arabidopsis thaliana supported increased levels of TCV gRNA accumulation by up to 2-fold ( Fig 11C ) . The Arabidopsis-TCV system was also used to estimate if TCV infection could induce RH30 gene transcription . RT-PCR analysis revealed induction of RH30 mRNA transcription in TCV-infected versus mock-inoculated plants ( Fig 11D ) . All these data are in agreement that RH30 is a strong restriction factor against tombusviruses and related viruses in plants . To learn if RH30 also has restriction function against an unrelated plant virus , we over-expressed AtRH30 in N . benthamiana and measured the accumulation of the unrelated tobacco mosaic tobamovirus ( TMV ) . We observed a ~3-fold reduction in TMV RNA accumulation in N . benthamiana leaves expressing the WT RH30 , but not in those leaves expressing the helicase core mutant of RH30 ( F416L ) ( Fig 11E ) . Expression of WT RH30 , but not that of the RH30 ( F416L ) helicase core mutant , also inhibited the accumulation of the insect-infecting Nodamura virus ( NoV ) by ~3-fold in yeast ( S2A Fig ) . Interestingly , the accumulation of Flock House virus ( FHV ) , an alphanodavirus , which is related to NoV , was only slightly inhibited by the expression of WT RH30 in yeast ( S2B Fig ) . Based on these observations , we suggest that the plant RH30 DEAD-box helicase has a broad-range CIRF activity against several RNA viruses .
DEAD-box RNA helicases are the most numerous among RNA helicases [33 , 37] . They are involved in all facets of RNA processes in cells . RNA viruses and retroviruses also usurp several DEAD-box helicases to facilitate their replication and other viral processes during infection [70 , 71] . However , the host also deploys DEAD-box helicases to inhibit RNA virus replication [70 , 72] . Accordingly , in this work we present several pieces of evidence that the DDX17-like RH30 DEAD-box helicase restricts tombusvirus replication , including the peroxisomal replicating TBSV and CNV and the mitochondrial-replicating CIRV in yeast and plants , and the more distantly related TCV and RCNMV and the unrelated TMV in plants . On the contrary , knock-down of RH30 enhances the replication of these three tombusviruses in N . benthamiana or the related TCV in RH30 knock-out lines of Arabidopsis . On the other hand , the helicase core mutant RH30 can only partially inhibit tombusvirus replication in plants or in vitro , suggesting that the helicase function of RH30 is needed for its full antiviral activity . How can RH30 restrict TBSV replication ? We show that the antiviral RH30 helicase binds to p33 and p92pol replication proteins based on co-purification experiments of the viral replicase complex , a pull down assay , and BiFC in N . benthamiana . We propose that the interaction of RH30 helicase with the viral replication proteins might be important for the targeting of RH30 into the viral replication compartment ( Fig 12 ) . Accordingly , RH30 is recruited into the viral replication compartment from the cytosol and the nucleus based on live imaging in plant cells ( Fig 3 ) . The targeting of RH30 into the replication compartment is critical for its antiviral function , because fusion of a nuclear retention signal with RH30 , which leads to its enrichment in the nucleus at the expense of the cytosolic pool of RH30 , in turn , cancelled out the antiviral effect of RH30 . Yeast CFE-based replicase reconstitution assays showed that RH30 acts in the early steps of replication , since both ( - ) and ( + ) RNA synthesis was inhibited by RH30 ( Fig 6 ) . Moreover , the in vitro RdRp activation assay demonstrated that RH30 inhibited the TBSV RdRp activation step during the replication process as well ( Fig 6C ) . In contrast , the CFE-based TBSV replication was not inhibited by RH30 after replicase assembly was completed ( see step 2 , Fig 6B ) . These data suggest that RH30 DEAD-box helicase must act at the earliest steps in the replication process to inhibit TBSV replication . RH30 also binds to the viral RNA , including the 5’ UTR ( i . e . , RI ) and RII internal sequence present within the p92pol coding region ( Fig 7 ) . Using in vitro interaction and replication assays between RNA-p33 replication protein , we show that RH30 inhibits several steps in tombusvirus replication . These include the RH30-based inhibition of ( i ) the specific recognition of the critical RII ( + ) -SL cis-acting element in the viral ( + ) RNA by p33 replication protein , which is absolutely required for template recruitment into VRCs , ( ii ) the activation of the viral p92 RdRp , and ( iii ) the assembly of the VRCs [21 , 26 , 73] . Moreover , RH30 helicase could disassemble viral RNA-p33 complexes by likely remodeling the RNA structure in an ATP-dependent manner ( Fig 8 ) . However , RH30-mediated disassembly of viral RNA-p33 complexes is unlikely to occur after VRC assembly is completed , because RH30 helicase was not an effective restriction factor when added at a late step of TBSV replication ( step 2 , Fig 6B ) . We propose that the membrane-bound TBSV VRCs are protecting the viral RNA-p33 complexes by restricting accessibility of the VRC complex to RH30 DEAD-box helicase . Accordingly , we have shown before that the fully-assembled TBSV VRCs are resistant to cellular ribonucleases [74] . Therefore , RH30 helicase might only be able to disassemble viral RNA-p33 complexes before the vesicle-like spherule formation , which is the characteristic structure of the TBSV VRCs in yeast and plants [75] . Altogether , the in vitro assays provide plentiful data on the direct inhibitory effect of RH30 helicase on TBSV replication , indicating that RH30 functions as an effector-type , not signaling-type , DEAD-box helicase , which detect viral RNA and send signals to downstream components of the innate immunity network [72] . Future experiments will address if RH30 might have additional mechanisms to restrict tombusvirus replication . A recently emerging concept in innate immunity is the significant roles of DEAD-box helicases expressed by host cells that greatly reduce virus replication and facilitate combating viruses and making the induced and passive innate immune responses more potent . Many of the identified yeast DEAD-box helicases with restriction functions against TBSV are conserved in plants and mammals . Altogether , the genome-wide screens performed with animal viruses have shown that helicases are the largest group of host proteins affecting RNA virus replication . For example , in case of HIV , the involvement of several cellular helicases has been demonstrated , including DDX17 and DDX3 [71 , 76 , 77] . Yet , the functions of the cellular helicases during virus replication are currently understudied . The emerging pricture in plant-virus interactions , similar to animal-virus interactions , is the diverse roles of various host RNA helicases . Different plant viruses have been shown to co-opt plant RNA helicases for pro-viral functions . These include RH8 and RH9 for potyvirus replication and RH20 , RH2 and RH5 for TBSV replication [27–29 , 39 , 44 , 78] . However , this paper shows evidence that a plant DEAD-box helicase , RH30 , can also be utilized by host plants for antiviral functions . Thus , in addition to the previously identified Dicer-like RNA helicases [16 , 79–81] , additional plant RNA helicases might function as CIRFs by recognizing plant virus RNAs . The DDX17-like RH30 DEAD-box helicase characterized here opens up the possibility that among the more than 100 helicases of plants , there are additional ones with antiviral functions , serving as effector-type or sensor-like RNA helicases . The discovery of the antiviral role of RH30 helicase illustrates the likely ancient roles of RNA helicases in plant innate immunity . In summary , we have demonstrated that the plant DDX17-like RH30 DEAD-box helicase acts as a major restriction factor against tombusvirus replication when expressed in plants and yeast surrogate host . We show that RH30 DEAD-box helicase is targeted to the large TBSV replication compartment . In addition , we find that RH30 blocks the assembly of viral replicase complex , the activation of the RNA-dependent RNA polymerase function of p92pol and binding of p33 replication protein to critical cis-acting element in the TBSV RNA ( Fig 12 ) . Altogether , the plant DDX17-like RH30 DEAD-box helicase is a potent , effector-type , restriction factor of tombus- and related viruses .
Biotinylated RII RNA of DI-72 ( + ) was synthesized by in vitro T7 transcription in the presence of 7 . 5 μl of 10 mM ATP , CTP , GTP and 5 mM UTP as well as 0 . 35 μl of 10 mM biotin16-UTP ( Roche ) in a total of 50 μl reaction volume . The interaction assay was performed with 3 . 8 μM of recombinant MBP-RH30 and 1 . 9 μM of MBP-p33C along with 0 . 1 μg of biotinylated RNA , 0 . 1 μl of tRNA ( 1 mg/ml ) , 2 U RNase inhibitor , and 1 mM ATP in the presence of biotin-RNA binding buffer ( 100 mM Tris [pH 7 . 9] , 10% glycerol , 100 mM KCl , 5 mM MgCl2 , 0 . 1% NP-40 ) in a 10 μl reaction mixture . Non-biotinylated RII of DI-72 ( + ) RNA or absence of ATP was used as controls . Assay #1: Recombinant MBP-RH30 was incubated first with biontinylated RII ( + ) RNA at 25°C for 15 min . Then , the recombinant MBP-p33C was added to the reaction and incubated for another 15 min . Assay #2: Recombinant MBP-RH30 and MBP-p33C were co-incubated simultaneously with biontinylated RII ( + ) RNA at 25°C for 30 min . The reaction mixtures were incubated with 20 μl of Promega Streptavidin MagneSphere Paramagnetic Particles ( VWR ) at room temperature for 20 min . The particles were collected in a magnetic stand and washed with binding buffer for five times . The protein-RNA complexes were then eluted with 20 μl of SDS loading dye containing β-mercaptoethanol by boiling for 15 min . The eluted samples were analyzed by Western blot with anti-p33 antibody . Assay #3: For the detection of p33 released from protein-biotinylated RNA complex , 1 . 9 μM of recombinant MBP-p33C was incubated with 0 . 1 μg of biontinylated RII of DI-72 ( + ) RNA at 25°C for 15 min , followed by the addition of 20 μl of Promega Streptavidin MagneSphere Paramagnetic Particles for another 30 min incubation at room temperature . After collection of the beads and washing with biotin-RNA binding buffer for five times , the particles were incubated with either 0 . 95 or 3 . 8 μM of MBP-RH30 or MBP ( used as control ) in the presence of biotin-RNA binding buffer containing 1 mM ATP at 25°C for 15 min . The supernatant of the mixture was collected after collecting the particles in a magnetic stand and was analyzed by Western blot with anti-p33 antibody . The conditions for the EMSA experiments were described previously [24] . Briefly , the EMSA assay was performed with 0 . 1 pmol of 32P-labeled RNA probes along with different concentrations ( 0 . 4 , 1 . 9 , and 5 . 7 μM ) of purified recombinant MBP-fusion proteins or MBP in the presence of RNA binding buffer ( 10 mM HEPES [pH7 . 4] , 50 mM NaCl , 1 mM DTT , 1 mM EDTA , 5% Glycerol , 2 . 5 mM MgCl2 ) , 2 U of RNase inhibitor , as well as 0 . 1 μg of tRNA in a total of 10 μl reaction volume . Two different amounts ( 2 and 4 pmol ) of unlabeled RNAs together with 5 . 7 μM of either MBP-RH30 or MBP were used for template competition . To study if purified proteins could unwind partial dsRNA duplex , the dsRNA strand-separation assay was performed as described [28] . Firstly , the unlabeled single-stranded DI-72 ( - ) or DI-72 ( + ) RNAs were synthesized via T7 polymerase- based in vitro transcription . The 32P-labeled single-stranded RI ( - ) or RII ( + ) RNAs were synthesized by T7-based in vitro transcription using 32P-labeled UTP . To prepare partial dsRNA duplexes , consisting of either RI ( - ) /DI-72 ( + ) or RII ( + ) /DI-72 ( - ) ( see Fig 7E and 7F ) , 2 pmol of 32P -labeled RI ( - ) or RII ( + ) were annealed to 6 pmol of unlabeled DI-72 ( + ) or DI-72 ( - ) in STE buffer ( 10 mM TRIS [pH 8 . 0] , 1 mM EDTA , and 100 mM NaCl ) by slowly cooling down the samples ( in a total volume of 20 μl ) from 94°C to 25°C in 30 min . To test if the purified recombinant proteins could separate the partial dsRNA duplex , 1 . 9 and 5 . 7 μM purified MBP fusion proteins or MBP as a negative control were added separately to the partial dsRNA duplex in the RNA binding buffer ( 10 mM HEPES [pH7 . 4] , 50 mM NaCl , 1 mM DTT , 1 mM EDTA , 5% Glycerol , 2 . 5 mM MgCl2 ) along with 1mM ATP , followed by incubation at 25°C for 25 min . The reaction mixtures were then treated with Proteinase K ( 2 μg/per reaction ) at 37°C for 20 min , followed by loading onto 5% nondenaturing polyacrylamide gel with 200V for 1 h . Additional methods can be found in S1 Text and the primers used are listed in S1 Table . | Positive-stranded RNA viruses are important and emerging pathogens that greatly depend on the host during infection . The host uses conserved innate and cell-intrinsic restriction factors as a first line of defense to combat viruses . Among the most intriguing host restriction factors are the family of DEAD-box RNA helicases , which can function as viral RNA sensors or directly as effectors by inhibiting RNA virus replication . RNA helicases are involved in cellular metabolism and perform RNA duplex unwinding and remodeling of RNA-protein complexes in cells . The authors demonstrate that the plant DDX17-like RH30 DEAD-box helicase acts as a strong restriction factor of tombusviruses by blocking multiple steps in the viral replication process . Overall , the findings presented open up a new avenue based on DEAD-box RNA helicases to improve the resistance of plants against viral infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"plant",
"anatomy",
"nucleic",
"acid",
"synthesis",
"enzymes",
"microbiology",
"enzymology",
"fungi",
"plant",
"science",
"rna",
"helicases",
"rna",
"synthesis",
"dead-box",
"chemical",
"synthesis",
"research",
"and",
"analysis",
"methods",
"proteins",
"guide",
"rna",... | 2019 | Blocking tombusvirus replication through the antiviral functions of DDX17-like RH30 DEAD-box helicase |
The murine model of T . cruzi infection has provided compelling evidence that development of host resistance against intracellular protozoans critically depends on the activation of members of the Toll-like receptor ( TLR ) family via the MyD88 adaptor molecule . However , the possibility that TLR/MyD88 signaling pathways also control the induction of immunoprotective CD8+ T cell-mediated effector functions has not been investigated to date . We addressed this question by measuring the frequencies of IFN-γ secreting CD8+ T cells specific for H-2Kb-restricted immunodominant peptides as well as the in vivo Ag-specific cytotoxic response in infected animals that are deficient either in TLR2 , TLR4 , TLR9 or MyD88 signaling pathways . Strikingly , we found that T . cruzi-infected Tlr2−/− , Tlr4−/− , Tlr9−/− or Myd88−/− mice generated both specific cytotoxic responses and IFN-γ secreting CD8+ T cells at levels comparable to WT mice , although the frequency of IFN-γ+CD4+ cells was diminished in infected Myd88−/− mice . We also analyzed the efficiency of TLR4-driven immune responses against T . cruzi using TLR4-deficient mice on the C57BL genetic background ( B6 and B10 ) . Our studies demonstrated that TLR4 signaling is required for optimal production of IFN-γ , TNF-α and nitric oxide ( NO ) in the spleen of infected animals and , as a consequence , Tlr4−/− mice display higher parasitemia levels . Collectively , our results indicate that TLR4 , as well as previously shown for TLR2 , TLR9 and MyD88 , contributes to the innate immune response and , consequently , resistance in the acute phase of infection , although each of these pathways is not individually essential for the generation of class I-restricted responses against T . cruzi .
T . cruzi is an intracellular protozoan parasite that causes Chagas' disease , an endemic disorder affecting 16–20 million people which remains a health problem in Latin America . Although both innate and acquired immune responses are triggered during early infection and are critical for host survival , around 5% of individuals die due to myocarditis during the acute phase of the disease . In most cases , despite of the immune response , T . cruzi manages to subsist within the host and in approximately 30% of infected individuals it establishes a lifelong chronic illness presenting different clinical forms , including cardiomyopathy and megasyndrome in the gut [1] . Immunopathology due to parasite persistence is considered a key element in the development of chagasic cardiomyopathy , although a secondary role for autoimmunity is not completely excluded . Different members of the family of Toll-like receptors ( TLRs ) , by recognizing diverse pathogen-associated molecular patterns ( PAMPs ) of bacterial , viral , fungal , and protozoan origin trigger the activation of innate immunity and the subsequent development of Ag-specific adaptive immunity [2] . To date , TLR2 , TLR4 , and TLR9 have been implicated in recognition of different T . cruzi-derived PAMPs [3]–[6] . TLR2 recognizes GPI-anchors of mucin-like proteins and the T . cruzi-released protein Tc52 [3] , [4] , whereas TLR4 is responsible for recognition of free glycoinositolphospholipids [5] and TLR9 is involved in recognition of the CpG motif present in T . cruzi DNA [6] . Mice deficient in MyD88 , the adaptor molecule required for signaling events by most TLRs as well as IL-1R and IL-18R , show greatly enhanced susceptibility to infection with this protozoan parasite [7] . The susceptibility to infection of Tlr2−/ , Tlr9−/− and Tlr2−/Tlr9−/− double knockout mice ( all in the C57BL/6 background ) has also been analyzed [6] , [7] . Interestingly , although mice simultaneously lacking TLR2 and TLR9 are highly vulnerable to infection , their mortality rate is still less than that of Myd88−/− mice , pointing to the involvement of other TLRs and/or IL-1/IL-18 in the control of mortality . In addition to MyD88-dependent activation , another transduction pathway is involved in signaling through TLR3 and TLR4 . This pathway is mediated by the TIR domain-containing adaptor inducing IFN-γ ( TRIF ) . Interestingly , Myd88−/−Trif−/− and Myd88−/−Ifnar−/− double knock out mice were even more sensitive to in vivo infection with T . cruzi than Myd88−/− mice , indicating that in addition to MyD88-dependent induction of proinflammatory cytokines , the TRIF-dependent production of type I IFN also contributes resistance to T . cruzi infection [8] . In accord with this observation , we have previously demonstrated that the lack of expression of functional TLR4 in mice of C3H background caused higher parasitemia and accelerated mortality to T . cruzi infection [5] , although the mechanisms by which this occurs are not yet fully determined . However , since C3H WT mice are known to be more susceptible to T . cruzi infection when compared to mice of the C57BL strains , the direct comparison between the levels of susceptibility of C3H/HeJ ( TLR4-deficient ) mice and the other above mentioned Tlr−/− and Myd88−/− mice is difficult to interpret . Therefore , one of the aims of the present work was to analyze the role of TLR4 in the C57BL background in the innate response to T . cruzi . For this , host cell invasion , parasite survival and release from infected macrophages , as well as nitric oxide ( NO ) production were quantified in C57BL/6 ( WT ) and TLR4-deficient cell cultures . We also evaluated the contribution of TLR4 to the in vivo control of parasitemia levels and survival , as well as to IFN-γ and TNF-α production in the B6 and B10 backgrounds . Importantly , the participation of TLR2 , TLR4 , TLR9 and MyD88 in the induction of crucial effector mechanisms of the adaptive response against T . cruzi was also investigated , measured as the Ag-specific IFN-γ production and cytotoxic response mediated by CD8+ T cells in infected mice .
In order to compare the anti-T . cruzi microbicidal activity of WT and TLR4-deficient macrophages , it was first necessary to investigate whether the infection rate and parasite load were equivalent in both cases . Therefore we first compared the capacity of T . cruzi trypomastigotes to infect TLR4-deficient and WT macrophage ( MO ) cultures in three different genetic backgrounds: C3H , C57BL/10 and C57BL/6 . Strains C3H/HeJ and C57BL/10ScN are natural mutants in which the Tlr4 gene suffered mutations that result either in a residue substitution ( P712H ) , rendering the receptor non-functional , or a deletion , with non-expression of TLR4 , respectively [9] . Engineered Tlr4−/− in the B6 background was also previously described [10] . As shown in Figure 1A , no difference in the percentage of infected macrophages or in the number of parasites per macrophage after one hour of infection could be detected between cultures from the TLR4-deficient strains and their respective WT controls . However , when non-internalized parasites were extensively washed out after 1 h of interaction and the cultures were left to continue for three more hours , a significantly higher percentage of infected MO was found in TLR4-mutant cultures ( Fig . S1 ) . This result suggests the existence of an early microbicidal mechanism which is dependent on a functional TLR4 . In agreement with that , the number of trypomastigotes released in the supernatant after the parasite completes its intracellular cycle in long-term cultures is significantly higher in the TLR4-deficient MO cultures ( Fig . 1B–G ) . This is true for T . cruzi Y ( Fig . 1C–G ) and CL strains ( Fig . 1B ) and for both resident and elicited macrophages ( Fig . 1F and 1G ) . Together these results indicated that although cell invasion by the parasite is not affected by the absence of a functional TLR4 , T . cruzi growth is favored in TLR4-deficient MO , possibly due to a defective early anti-trypanosomacidal mechanism in TLR4-deficient MO . The expression of fluorescent TLR4 in cell lines allowed us to map TLR4 subcellular location , demonstrating its presence on the cellular surface and in the Golgi , similar to the TLR4 distribution observed in human monocytes [11] . It is also known that early after cell invasion the T . cruzi localizes in a host cell vacuole which fuses with peripheral lysosomes . HEK293 cells stably transfected with TLR4-yellow fluorescent protein and MD-2 ( HEK-TLR4YFP ) were infected with labeled T . cruzi trypomastigotes and 2 . 5 h later we analyzed parasite-TLR4 co-localization by confocal microscopy . Staining these infected cells with a lysosome probe also revealed that T . cruzi-TLR4 co-localization occurs in acidic compartments ( Figure 2 ) . Both NO and reactive oxygen species ( ROS ) have been shown to mediate T . cruzi killing [12]–[15] . Thus , we next analyzed the effects of adding NO and/or ROS inhibitors to the MO cultures during infection . Figure 3A shows that the addition of desferroxamine ( DFO ) , an iron chelator which can also act as a free radical scavenger [16] , causes a significant increase in the percentage of infected WT MO . This treatment abolishes the otherwise significant difference found in the percentage of infected MO between the non-treated WT and TLR4-deficient cultures . The same results are obtained when the inducible NO synthase ( iNOS ) inhibitor L-NMMA , or the combination of DFO and L-NMMA are added to these cultures . In order to further confirm the relevance of this early microbicidal mechanism , absent in TLR4-deficient MO , we tested the effect of inhibiting NO production in long term MO cultures , in which the number of parasites released by infected cells in the supernatant was evaluated several days after initial infection . As shown in Figure 3B , while iNOS inhibition had no effect in the number of parasites released by Tlr4−/− MO cultures , the addition of L-NMMA to the infected WT MO raised the number of trypomastigotes found in the supernatants to the levels observed in the Tlr4−/− MO cultures . In contrast , the addition of rIFN-γ an iNOS inducer to the MO cultures from the beginning of infection reduces the quantity of free trypomastigotes and results in equal numbers of released parasites from WT or TLR4-deficient MO ( Fig . S2 ) . Together , these results strongly suggest that the early trypanosomacidal mechanism absent in TLR4-deficient MO depends on ROS and NO induction . We next compared parasitemia and mortality between different pairs of WT and TLR4-deficient mice ( B10 or B6 versus B10/ScN or Tlr4 KO , respectively ) , after i . p . infection with 2×103 T . cruzi strain Y bloodstream trypomastigotes . Results in Figures 4A-C and 4D-F show that in both cases we found significantly higher parasitemia levels in TLR4-deficient mice , although the levels of parasites in the blood returned to very low or undetectable levels by day 11–12 post-infection and did not rise again , differently from what was previously described for C3H/HeJ mice , in which parasitemia levels were not controlled after day 15 pi [5] . We further monitored the mortality after infection and found that TLR4-deficiency in both B10 and B6 backgrounds results in higher lethality . Statistically significant differences , however , were consistently found only when comparing B10 and B10/ScN mice , while results with B6 and Tlr4−/− were more variable and did not reach statistical significance ( Figure 4G and H ) . Of note , these results were obtained in male mice of 6–7 weeks of age , while in older TLR4-deficient mice the higher susceptibility could not be observed ( data not shown ) . We have also performed experiments with lower ( 102 ) and higher ( 104 and 105 ) doses of infective T . cruzi forms/mice obtaining the same results ( data not shown ) . Therefore , mice lacking TLR4 expression in a C57BL genetic background are more sensitive than their WT controls to infection with T . cruzi , although these strains do not display the uncontrolled parasitemia and the remarkable earlier mortality previously observed in the TLR4-mutant C3H/HeJ mice [5] . We then analyzed whether a lower NO production by the infected TLR4-deficient mice could explain their higher susceptibility to infection as suggested by the results obtained in vitro . In accordance with that hypothesis , the production of NO ( inferred from nitrite levels in the supernatants ) by spleen cells from infected TLR4-deficient mice was significantly reduced when compared with NO released by spleen cells from infected WT mice at day 10 post-infection ( Figure 5A and B ) . Nitrite levels are also lower in the sera of TLR4-deficient mice , compared to WT animals , at this time point of infection ( data not shown ) . Furthermore , the in vivo blockade of NO production in T . cruzi infected animals , by injection of the inducible NO synthase ( iNOS ) inhibitor aminoguanidine ( AG ) in the early phase of acute infection , brought the parasitemia and mortality of treated WT mice to the same levels obtained in treated Tlr4−/− animals ( Figure 5C and D ) . In animals injected every other day with AG , following a previously reported protocol [17] , parasitemia kept rising until treatment was stopped on day 13 pi , attaining 3 and 7 fold higher levels of what was usually observed in Tlr4−/− and WT non-treated animals , respectively ( Figure 5C ) . This is due to the prevention of all NO generation , as for example in response to TLR2 and/or TLR9 signaling pathways , rather than exclusive inhibition of NO triggered by TLR4 engagement . Also , while non-treated infected animals usually die only after day 20 pi , earlier mortality was observed among AG-treated mice , with 50% mortality in both Tlr4−/− and WT AG-treated groups by day 12 pi ( Figure 5D ) . Hence , these results suggest that the lower NO production due to the absence of TLR4 expression during the early phase of infection with T . cruzi is responsible for the higher sensitivity observed in Tlr4−/− mice . As IFN-γ is thought to be the most important inducer of iNOS in macrophages and thus essential for mediation of NO-dependent parasite control during acute infection [18] , we quantified IFN-γ production by spleen cells from WT and TLR4-deficient infected mice . As shown in Figure 6A , higher IFN-γ levels are indeed secreted by WT infected splenocytes at day 10 pi . We also compared the secretion of another crucial cytokine for iNOS expression and host resistance to T . cruzi , TNF-α [14] As shown in Figure 6B , the levels of TNF-α secreted by splenocytes from infected WT mice are also significantly higher than from TLR4-deficient mice . Both IFN-γ and TNF-α can be produced after T . cruzi-induced triggering of the innate immune response ( mainly by NK/NKT cells and macrophages/DC respectively ) , as well as by CD8+ and CD4+ T lymphocytes later in the infection course , as part of the acquired response to the parasite . Since several previous studies have demonstrated the importance of IFN-γ secretion by CD8+ T cells in resistance to infection with T . cruzi [19] , we asked if the frequency of Ag-specific IFN-γ-secreting cells would be altered in the absence of TLR4 expression . To do so , a previously defined H-2Kb-restricted epitope ( PA8 ) derived from the amastigote surface protein-2 ( ASP2 ) , which is a member of the trans-sialidase family of surface proteins , was employed in ELISPOT assays [20] . However , as shown in Figure 6C no significant difference in the frequency Ag-specific IFN-γ secreting cells could be observed between WT and TLR4-deficient infected mice . Two other previously described trans-sialidase-derived peptides TSKB20 and TSKB18 [21] were alternatively employed in ELISPOT assays , giving the same results ( not shown ) . The frequency of IFN-γ secreting CD4+ and CD8+ T cells in the spleens of WT and TLR4-deficient mice at day 13 pi was also investigated by intracellular staining and results are shown in Figure 6D-G . These data show that the frequencies of CD4+ and CD8+ T cells secreting IFN-γ in response to T . cruzi-derived antigens are not reduced in Tlr4−/− mice . At this point , the present investigation was extended to compare the frequency of PA8/Kb-specific IFN-γ secreting lymphocytes between WT and mice which are deficient in other members of the TLR family , as well as Myd88−/− mice , whose susceptibility to T . cruzi infection was previously described [6] , [7] . To our surprise , the frequency of these important effector cells of the acquired response is not altered in the spleens of Tlr2−/− , Tlr9−/− or Myd88−/− mice , as with Tlr4-deficient mice ( Figure 7A ) . We also estimated the percentage of IFN-γ secreting CD8+ T cells in the spleen of infected Myd88−/− mice at day 10 pi by intracellular cytokine staining ( ICS ) following in vitro stimulation with PA8 peptide and obtained the same results , that is , no significant difference in the frequency of IFN-γ+CD8+cells between WT B6 and Myd88−/− mice ( Fig . 7B–D ) . In order to further evaluate the adaptive response to T . cruzi in Myd88−/− mice , the percentage of IFN-γ secreting T cells was measured by ICS in both CD4+ and CD8+ subsets at days 11 and 13 pi . As shown in Figure 8 , although the absence of MyD88 signaling strongly affects the percentage of CD4+ IFN-γ cells in these mice , MyD88 expression is not essential for the differentiation of IFN-γ producing CD8+ T cells specific against T . cruzi-derived epitopes . Finally , we tested whether expansion of specific CD8+ cytotoxic T cells was affected in any of the above Tlr−/− or Myd88−/− mice . For this purpose , we used a functional cytotoxic assay which measures the in vivo elimination of target cells ( total splenocytes ) coated either with PA8 ( Figs . 9A and C ) , TSKB20 or TSKB18 ( Fig . 9B ) peptides , as previously described [21] . The phenotype of effector cells mediating peptide-specific in vivo cell killing was established earlier as being CD8+ T cells [22] . The kinetics of Ag-specific cytotoxic CD8+ T cell development during infection with the Y strain of T . cruzi in mice was also previously determined , showing that the maximum cytotoxicity ( close to 100% specific lysis ) is attained around day 15 pi and continued at a high level in B6 mice , even until 100 days after challenge [22] . As shown in Fig . 9B , at day 20 post-infection , no difference in peptide-specific cytotoxicity could be detected between Tlr4−/− and WT mice for any of the tested peptides . The same was true for Myd88−/− mice , in which the in vivo cytotoxicity assay was performed at an earlier post-infection time point ( day 10 pi ) due to their earlier mortality [7] ( Fig . 9B ) . A summary of the cytotoxicity experiments is shown in Figure 9C , where the results of specific killing obtained with the PA8 immunodominant peptide in Tlr2−/− , Tlr4 −/− , Tlr9−/− or Myd88−/− mice are compared to B6 controls . No difference in the levels of specific cytotoxicity was observed in any of these deficient mice . Together , these results clearly indicate that deficiency in TLR2 , TLR4 , TLR9 or even MyD88 expression does not impair CD8+ T cell effector responses during infection with T . cruzi .
Different T . cruzi-derived molecules are able to induce host innate immune responses through the activation of different members of the TLR family , [3]–[6] , [23] , including glycoinositolphospholipids derived from the parasite membrane which induce a pro-inflammatory response through the TLR4 pathway [5] , [23] . Thus , the documented high susceptibility of Myd88−/− mice to infection with T . cruzi could not be attributed to a single TLR , suggesting that different members of the TLR family act in concert in determining resistance to the pathogen [6] . Bafica and collaborators have shown that doubly deficient Tlr2−/− Tlr9−/− mice , although more susceptible than the single TLR2- or TLR9-deficient mice , do not display the acute mortality exhibited by Myd88−/− mice , suggesting that additional TLR/IL-1R family members are involved in the protection against infection with T . cruzi in mice [6] . In this context , the contribution of TLR4 signaling to control of the parasite burden in the C57BL/6 background was not investigated until the present study , as the only previous work on the subject was performed in mice of a different genetic background [5] . Importantly , the present work is the first to study the contribution of the different TLR and MyD88 pathways to the development of anti-T . cruzi responses mediated by CD8+ effector T cells , a critical element of the acquired immune response to the parasite . The first question we addressed was to assess the role of TLR4 in T . cruzi internalization and triggering of very early microbicidal activity by macrophages . Infective T . cruzi trypomastigotes invade host cells using at least two different strategies; either by an active process recruiting host-cell lysosomes to the area of parasite cell contact or by an alternative pathway , in which the parasite infects phagocytic cells through conventional phagocytosis/endocytosis mechanism [24]–[26] . In both cases , the parasite may escape to the cytoplasm where it differentiates into the aflagellated amastigote form and begins intracellular replication . During cell invasion , T . cruzi interacts with different macrophage receptors to induce its own phagocytosis , but the nature of those receptors and the molecular mechanisms involved are still poorly understood . Although the general current view is that TLRs do not function directly as phagocytic receptors [27] , a recent report indicated that during the invasion of T . cruzi , the activation of the Rab5-dependent phagocytic pathway is regulated by signals emanated through the parasite interaction with TLR2 in macrophages [28] . Our present results with Tlr4-deficient macrophages from three different mouse strains show that internalization of T . cruzi by macrophages is not affected by the absence of functional TLR4 expression . Some studies on the other hand , have demonstrated that TLR signaling by means of MyD88 can enhance phagosome acidification and function , the so-called phagosome maturation , which is required for effective sterilization of its contents [29] . In accord with those results , we found that after 2 . 5 h of infection , TLR4 and parasite co-localize into acidic compartments . Also , 4 h after infection , the percentage of TLR4-deficient macrophages infected with T . cruzi is significantly higher when compared to WT cells . The addition of iNOS or ROS inhibitors abolished the difference in the frequency of infected macrophages between cultures from TLR4-deficient and WT origin , indicating that this early trypanosomicidal mechanism triggered by TLR4 depends on the production of reactive nitrogen intermediates ( RNI ) and ROS , which have been described to participate in the microbicidal activity against T . cruzi and other pathogens [12]–[15] . Moreover , the fact that the simultaneous usage of iNOS and ROS inhibitors did not increase further the percentage of infected macrophages , suggests that the peroxynitrite anion ( ONOO− ) , a strong oxidizing and against T . cruzi , formed by the reaction between nitric oxide ( NO ) and superoxide radical ( O−2 ) , may be the main species responsible for the elimination of T . cruzi , as described [30] , [31]cytotoxic effector molecule . Therefore , TLR4 signaling triggers an important early parasiticidal event against T . cruzi , which is dependent on the formation of NO and ROS . Significantly lower production of NO was also found in splenocyte cultures from Tlr4−/− mice at day 10 post infection . In conformity to these results , we demonstrated that Tlr4−/− splenocyte cultures produce lower levels of the main iNOS inducer cytokines , IFN-γ and TNF-α . As with the in vitro results obtained with macrophage cultures , the inhibition of NO production during in vivo infection made WT and Tlr4−/− mice equally susceptible , as measured by mortality and parasite levels in the blood . Our results are in agreement with previous studies demonstrating that mice deficient for inducible nitric oxide synthase ( iNOS ) are highly susceptible to T . cruzi [32] , and that the inhibition of iNOS from the beginning of infection lead to an increase in trypomastigotes in the blood and to high mortality [15] , [33] . Together , our results point to a significant contribution of the TLR4 pathway to the innate immune response against T . cruzi infection , with the production of NO playing a major role . We show that mice of either B10 or B6 genetic background with TLR4 deficiency presented significantly elevated parasite numbers in the blood compared to their WT controls after in vivo infection with T . cruzi . These results are in accordance with our previous work showing higher parasitemia levels of the Tlr4 mutant C3H/HeJ mice [5] , although this is more pronounced in the latter lineage . Concerning mortality , however , the absence of functional TLR4 expression in B10 or B6 mice do not lead to the acute mortality previously observed in C3H/HeJ mice [5] . Therefore , the effects of TLR4 deficiency on susceptibility to infection with T . cruzi are more evident in the C3H background . Inbred strains of mice may vary from highly resistant to highly susceptible , as reflected by parasitemia levels and survival time and , following these criteria , C3H strains have been classified as “susceptible” . Classical genetic studies previously established that the resistance to T . cruzi is governed by multiple genetic factors , including H-2-linked gene ( s ) [34] , [35] and the combination of different alleles in a group of loci confers resistance or susceptibility to infection . Therefore , analogous to the effects due to the absence of TLR2 , which only become perceptible in mice with the concomitant deficiency on TLR9 [6] , we have shown here that the susceptibility resulting from the absence of TLR4 is less pronounced in the resistant B6 and B10 backgrounds , compared to C3H strains . IFN-γ is an important mediator of resistance to T . cruzi . Besides iNOS , IFN-γ regulates the expression of a large number of genes , including chemokines and chemokine receptors , which were shown to play a role in IFN-γ-mediated protection in T . cruzi infection [18] , [32] , [36] . Early during infection , IFN-γ is secreted by NK cells and other cell types , as part of the innate response , and later on the infection course by activated CD4+ and CD8+ T cells . Since TLRs have been implicated in the modulation of acquired immunity against several pathogens , we have herein addressed the question of whether the frequencies of IFN-γ secreting CD8+ and CD4+ T cells are altered in Tlr4−/ mice . Our data showed no significant difference in the frequencies of IFN-γ producing CD8+ or CD4+ T cells in the spleens of Tlr4−/− and WT mice , indicating that TLR4 deficiency does not interfere with these important effectors of the acquired response against T . cruzi . Therefore , the higher level of IFN-γ detected in the supernatants of WT splenocyte cultures at day 10 pi is probably contributed by cells of the innate response . A number of different cells types may account for that and we are currently evaluating their phenotype . On the other hand , a significant reduction in the levels of the IFN-γ+CD4+ T cell population was observed in the spleens of Myd88−/− infected mice . This finding is in line with previous in vitro experiments [6] , but contrasts to the results of a recent paper , where IFN-γ production by CD4+ T cells is shown to be preserved in Myd88−/− mice infected with the Tulahuen strain of T . cruzi [37] . The reason for this apparent discrepancy is not clear and could be due to the different strain of parasite used for infection or to the method employed for CD4+ T cell re-stimulation in vitro . Therefore , our results demonstrate that although the CD4+ T cell-mediated response is substantially diminished , unaltered frequencies of CD8+ IFN-γ T cells specific for Kb-restricted T . cruzi-derived peptides are present in the spleens of Myd88−/− mice compared to WT controls at days 10 , 11 and 13 pi . CD8+ T cell mediated responses are a critical component of protective immunity in T . cruzi infection , since in their absence mice quickly succumb to the infection or develop a more severe chronic disease ( reviewed in [19] ) . Moreover , CD8+ T cells can be induced by vaccination to provide protection from lethal infection [38] . CD8+ T cells can control infection via a number of mechanisms: in addition to the already discussed secretion of IFN-γ inducing microbicidal activity in the host cell , the direct cytotoxic function against cells infected with T . cruzi is also a main effector response . Therefore , it is an important issue to define whether TLR-MyD88 mediated pathways can play a role in the priming and/or control of the cytotoxic T cell response against T . cruzi-infected targets . According to the present paradigm , this could be mainly achieved by the engagement of TLRs on antigen-presenting cells ( APCs ) such as dendritic cells ( DCs ) , promoting upregulation of co-stimulatory molecules , enhancement of antigen processing and presentation , as well as secretion of Th1 polarizing pro-inflammatory cytokines by the DCs [2] . We demonstrated here , however , that the cytotoxic response mediated by CD8+ T cells against H-2Kb-restricted immunodominant peptides in T . cruzi infected mice is not dependent on TLR2 , TLR4 nor TLR9 expression . Unexpectedly , the Ag-specific cytotoxic function was also preserved in Myd88−/− mice . As the cytotoxic response against the immunodominant PA8 T . cruzi epitope tested here was previously shown to be dependent on MHC class II restricted CD4+ T cells [39] , our results indicate that although diminished in frequency , the residual response of CD4+ activated T cells observed in infected Myd88−/− mice is sufficient for their licensing function , which results in the development of parasite-specific CD8+mediated cytotoxic response . A first possible interpretation of these results is that none of the tested TLR and MyD88 pathways are involved in the generation of cytotoxic CD8+ T cells during T . cruzi infection . In fact , other signaling molecules and innate recognition systems might contribute to adaptive immunity to T . cruzi as the members of the Nod-like receptor protein ( NLR ) family [40] . Other examples are: 1 ) the release of pro-inflammatory bradykinin peptide by the parasite proteases during infection and consequent DC maturation induced by bradykinin B2 receptors ( B2R ) [41] and 2 ) the recently described DC maturation induced by NFATc1 activation and consequent IFN-γ production in a TLR-independent pathway [37] . However , to date , it was not determined if these TLR-independent pathways can fully account for the preserved CD8+ T cell cytotoxic response against T . cruzi-infected targets observed in Myd88−/− mice . In our opinion , there is another plausible hypothesis to be considered for explaining the preserved CD8+ T cell cytotoxic response in Tlr2−/− , Tlr4−/− , Tlr9−/− or Myd88−/− mice: it is known that type I IFNs affect DC maturation [42] , [43] and can also stimulate survival , development of cytolytic function , and production of IFN-γ by CD8+ T cells [44] , [45] . Moreover , mice deprived of the type I IFN receptor , Ifnar−/− , develop higher parasitemia levels in comparison with control 129Sv mice [46] and doubly deficient Myd88−/−Ifnar−/− mice are highly susceptible to infection with T . cruzi [8] . Both TLR9 and TLR4 could induce type I IFN secretion through MyD88-dependent and -independent pathways , respectively . Therefore , TLR4 and TLR9 would be redundant concerning type I IFN production and might compensate for each other's absence in Tlr4−/− or Tlr9−/− mice . The TLR4-triggered TRIF pathway is also preserved in Myd88−/− mice and its activation would lead to type I IFN secretion and DC maturation , with the consequent normal adaptive responses against T . cruzi in these mice . Testing whether the CD8-mediated cytotoxicity against T . cruzi is affected in Tlr4−/−Tlr9−/− or Tlr4−/−Myd88−/− doubly deficient mice is one of our future goals . According to this hypothesis , cytotoxic CD8+ T cells would not be preserved in doubly deficient Myd88−/−Trif−/− mice , which is in agreement with the fact that these mice are even more susceptible to infection , as indicated by accelerated mortality when compared to single Myd88−/− mice [8] . Also , in opposition to Myd88−/− , the doubly deficient Myd88−/−Trif−/− mice are not able to control the levels of parasite in the blood [8] . The maintenance of CD8+ acquired responses against T . cruzi in Myd88−/− mice finds a parallel in studies of murine infection with Toxoplasma gondii , another intracellular protozoan parasite . As for T . cruzi , multiple TLR ligands were identified in T . gondii and Myd88−/− mice were shown to be highly susceptible to infection ( reviewed in [47] ) . Interestingly , a recent work demonstrated that a robust and protective IFN-γ response can be elicited in Myd88−/− mice infected with an avirulent T . gondii strain [48] . Therefore , the MyD88 pathway is required for innate immunity to control infection with Toxoplasma , even though adaptive immunity against the pathogen can be triggered without the need for this TLR adaptor molecule . The same picture emerges from our present results with T . cruzi . The absence of a role for the MyD88 pathway in the generation of CD8+ adaptive responses during T . cruzi infection is also in line with other previous reports analyzing the immune response to other pathogens [49] , [50] , including the described protective CD8+ T cell response against the intracellular bacteria Listeria [51] . The preservation of CD8+ T cell mediated effector mechanisms in MyD88-deficient mice is in agreement with the fact that despite their high mortality , these mice do succeed in controlling the number of parasites in the blood , in contrast to the even more susceptible Ifng−/− , IfngR−/− , iNos−/− or Myd88−/−Trif−/− mice [8] , [32] , [36] . At present we do not know why Myd88−/−mice succumb earlier than WT mice and display 100% mortality to T . cruzi infection , notwithstanding their capacity of controlling blood parasitemia [7] and their preserved CD8+ T cell-mediated responses shown here . We would like to consider four non-exclusive possibilities: First , Myd88−/−mice , whose IFN-γ levels in serum were shown to be significantly lower [7] , would also be affected by the fact that several genes , like iNOS and IP-10 , have been shown to be 5- to 100-fold less extensively induced by IFN-γ in macrophages lacking MyD88 expression [52] . Second , the higher susceptibility of Myd88−/− mice could be directly attributed to the defective activation of CD4+ T cells demonstrated here , as this cell population has also been demonstrated to be essential for resistance to infection [53] , probably through IFN-γ and TNF-α secretion . We do not know at the present what mechanism , absent in Myd88−/− mice , affects CD4+ T cell activation . Both attenuated DC maturation due to the absence of TLRs/MyD88 triggering and the absence of IL-1R/IL-18R signaling in CD4+ Myd88−/− T cells should be considered . Third , the lower levels of CD4+ helpers might also have indirect consequences as a defect in the B cell mediated response , which was also described to be necessary for resistance to the parasite [54] . In fact , although controversy exists , the requirement of TLR-MyD88 signaling for the generation of T-dependent antigen-specific antibody responses was proposed [55] , [56] and , interestingly , antibody responses against different virus are altered or completely lost in Myd88−/− mice [57]–[59] . Finally , another possible consequence of a deficiency in CD4+ cell activation could be the defective migration of CTLs into peripheral sites of infection distinct of the spleen ( liver and heart , for example ) , as recently demonstrated by Nakanishi et al . in a mouse model of herpes simplex virus ( HSV ) infection of the vagina [60] . In summary , the results obtained in the present study strongly argue in favor of a role for the TLR4 signaling in the innate immune response against T . cruzi displayed by B6 mice . Notably , we have also shown here that neither the absence of TLR2 , TLR4 or TLR9 individually , nor the ablation of all MyD88-mediated pathways affect the development of cytotoxic and IFN-γ-producing CD8+ T cells , which are crucial effector mechanisms against this parasite . Determining precisely how TLR-TRIF-MyD88 activation contributes to trigger protective immunity against T . cruzi will be of critical relevance for vaccine development against this important human parasite .
All animal experiments were approved by and conducted in accordance with guidelines of the Animal Care and Use Committee of the Federal University of Rio de Janeiro ( Comitê de Ética do Centro de Ciências da Saúde CEUA -CCS/UFRJ ) . Tlr2−/− , Tlr4−/− , Tlr9−/− and Myd88−/− mice were generated by and obtained from Dr . S . Akira ( Osaka University , Japan ) via Dr . R . T . Gazzinelli ( Federal University of Minas Gerais , Brazil ) . Tlr2−/− and Tlr4−/− mice were maintained along C57BL/6 mice at the Laboratório de Animais Transgênicos ( LAT , IBCCF° , UFRJ , RJ , Brazil ) . C3H/HeJ and C3H/HePas mice were from ICB , Universidade de São Paulo ( USP , SP , Brazil ) . C57BL/10 and C57BL/10ScN mice were maintained at the Biotério of the Department of Immunology ( IMPPG , UFRJ , RJ , Brazil ) . Tlr9−/− and Myd88−/− mice were maintained at the Centro de Pesquisas René Rachou ( FIOCRUZ , MG , Brazil ) . Mice used for experiments were sex- and age-matched , and housed with a 12-h light-dark cycle . Bloodstream trypomastigotes of the Y strain of T . cruzi [61] were obtained from BALB/c mice infected 7 days earlier . The concentration of parasites was estimated and each mouse ( at least 4 per group ) was inoculated intraperitoneally ( i . p . ) with 0 . 2 ml ( 2×103 trypomastigotes ) . Parasitemia was monitored by counting the number of bloodstream trypomastigotes in 5 µl of fresh blood collected from the tail vein . Mouse survival was followed daily . Resident or elicited macrophages ( obtained from the peritoneal cavity on day 4 after injection of 2 . 5 ml of 3% thioglycollate ) were plated in triplicates and infected with trypomastigotes at a 1∶10 ( macrophage:trypomastigote ) ratio . After 1 h of infection , the cells were washed four times with PBS to remove the extracellular parasites and cultured in DMEM supplemented with 10% FBS ( GIBCO , Invitrogen ) for the indicated time periods at 37°C in an atmosphere containing 5% CO2 . Trypomastigotes in the culture supernatants were counted microscopically in triplicates . Alternatively , extracellular parasites were removed by repeated washing after 1 h of infection and the cells were either washed , fixed and stained with Giemsa or cultured for a further 4 h in DMEM supplemented with 10% FBS before fixation and staining . In other experiments , macrophages were infected for 1 h in the presence of L-NMMA ( 1 mM ) and/or DFO ( 100 µM ) ; after washing , cells were cultured for further 4 h in DMEM supplemented with 10% FBS in the presence of L-NMMA ( 1 mM ) and/or DFO ( 100 µM ) and subsequently fixed and stained with Giemsa . The percentage of infected macrophages and the intracellular parasite numbers in 100 macrophages were counted under a light microscope . The stable cell line of HEK293 cells expressing the fluorescent protein TLR4YFP and MD-2 constructs were described previously [11] and kindly donated by Dr D . Golenbock ( University of Massachusetts Medical School , MA ) . Trypomastigotes were labeled with TO-PRO-3 ( 1 µl/ml , Molecular Probes , Invitrogen ) for 30 min at RT , washed twice and then cultured with HEK- TLR4YFP cells at 10∶1 ratio for 1 hour at 37°C , 5% CO2 . After repeated washing with PBS for extracellular parasite removal , cells were stained with LysoTracker Red probe ( 75 nM , Molecular Probes , Invitrogen ) for 1 hour . Cells were then washed in PBS containing 1 mM MgCl2 and 1 mM CaCl2 and fixed in 3 , 7% paraformaldehyde/PBS for 15 min . Confocal microscopy was performed with a Zeiss Axiovert 200-M inverted microscope equipped with an LSM 510 Meta laser-scanning unit . Image analysis was performed with LSM 510 software ( Zeiss ) . The Griess reaction was performed to quantitate nitrite concentrations in the supernatant of macrophage or spleen cell cultures , as previously described [62] . Briefly , 50 µl of sample plus 50 µl of Griess reagent were incubated for 10 min at RT , followed by detection at 550 nm in an automated ELISA plate reader . The results are expressed in units of micromolar , and were determined comparing the absorbance readings of the experimental samples to a sodium nitrate standard curve . Inhibition of iNOS in vivo was performed by injecting mice i . p . with 50 mg/kg of aminoguanidine/body weight , ( AG , Sigma-Aldrich , St . Louis ) , diluted in sterile phosphate-buffered saline ( PBS ) , every other day , as previously described [20] . Treatment started 4 h before infection with T . cruzi ( performed as described above ) , and animals were treated until day 13 pi . Control mice received the same volume ( 200 µl ) of PBS . A third group of mice received AG only . The ELISPOT assay was performed essentially as described earlier [63] . Briefly , the preparation of plates was done by coating 96-well nitrocellulose plates Multiscreen HA ( Millipore ) with 60 µl/well of sterile PBS containing 10 µg/ml of the anti-mouse IFN-γ mAb R4-6A2 ( BD Biosciences , San Jose , CA ) . After overnight incubation at RT , mAb solution was removed by sterile aspiration and the plates were washed three times with plain RPMI 1640 medium under sterile conditions . Plates were blocked by incubating wells with 100 µl RPMI medium containing 10% ( v/v ) FBS for at least 2 h at 37°C . Responder cells were obtained from spleens of B6 mice . Responder cells were ressuspended to a concentration of 106 viable cells per ml in RPMI medium ( GIBCO , Invitrogen ) supplemented with 10 mM HEPES , 2 mM L-glutamine , 5×10−5 M 2-β-mercaptoethanol , 1 mM sodium pyruvate , 100 U/ml of penicillin and streptomycin , 10% ( v/v ) FBS ( all purchased from GIBCO , Invitrogen ) . B6 spleen cells adjusted to a concentration of 4×106 viable cells per ml were used as antigen presenting cells after incubation or not with the synthetic peptide at a final concentration of 10 µM for 30 min at 37°C . One hundred microliters of suspension containing responders or antigen presenting cells were pipetted into each well . The plates were incubated for 24 h at 37°C in an atmosphere containing 5% CO2 . After incubation , the bulk of cultured cells was flicked out . To remove residual cells , plates were washed 3 times with PBS and 3 times with PBS-Tween . Each well received 75 µl of biotinylated anti-mouse IFN-γ mAb XMG1 . 2 ( BD Biosciences ) diluted in PBS-Tween to a final concentration of 2 µg/ml . Plates were incubated overnight at 4°C . Unbound antibodies were removed by washing the plates at least 6 times with PBS-Tween . Peroxidase-labeled streptavidin ( KPL ) was added at a 1∶800 dilution in PBS-Tween in a final volume of 100 µl/well . Plates were incubated for 1–2 h at RT and then washed three to five times with PBS-Tween and three times with PBS . Plates were developed by adding 100 µl/well of peroxidase substrate ( 50 mM Tris–HCl at pH 7 . 5 containing 1 mg/ml of DAB and 1 µl/ml of 30% hydrogen peroxide solution , both from Sigma ) . After incubation at RT for 15 min , the reaction was stopped by discarding the substrate solution and rinsing the plates under running tap water . Plates were dried at RT and spots were counted with the aid of a stereomicroscope ( Nikon ) or in the ImmunoSpot® Analyzer ( Cellular Technology Ltd . , Shaker Heights , OH , USA ) . Results of the ELISPOT assay are representative of two or more independent experiments . Tissue culture trypomastigotes of the Y strain of T . cruzi were transformed to amastigotes in acidic DMEM/10% FCS for 24 h at 37°C , as previously described [64] . Parasites were pelleted , washed in PBS , and subjected to more than five rounds of freeze-thawing followed by sonication . Cellular debris were removed by centrifugation at 12 , 000 rpm , and the soluble fraction was boiled for 5 min to denature the proteins . Protein concentrations were determined using a Bio-Rad protein assay . Splenocytes isolated from infected mice were cultured either with T . cruzi amastigote extract at 10 µg/ml ( see above ) or with PA8 peptide ( VNHRFTLV ) at 10 µM , or left unstimulated , for 5 h to 14 h at 37°C in the presence of brefeldin A ( Sigma-Aldrich ) . Cells were surface stained with anti-CD8-PerCP and anti-CD4-FITC ( BD Biosciences ) and fixed for 10 minutes with a solution containing PBS , 4% paraformaldehyde at RT . Then , cells were permeabilized for 15 minutes with PBS , 0 . 1% bovine serum albumine , 0 . 1% saponin ( Sigma-Aldrich ) . Intracellular cytokine staining was performed with anti- IFN-γ -PE ( BD Biosciences ) . At least 10 , 000 gated CD8+ lymphocyte events were acquired . Analytical flow cytometry was conducted with a FACSCalibur ( BD Biosciences ) and the data were processed with CellQuest software ( BD Biosciences ) . For the in vivo cytotoxicity assays , splenocytes of the different mouse strains were divided into two populations and labeled with the fluorogenic dye CFSE ( Molecular Probes , Invitrogen ) at a final concentration of 5 µM ( CFSEhigh ) or 0 . 5 µM ( CFSElow ) . CFSEhigh cells were pulsed for 40 min at 37°C with 1–2 . 5 µM of either H-2Kb -restricted ASP-2 peptide , also called PA8 , ( VNHRFTLV ) , H-2Kb- restricted TsKb-18 peptide ( ANYDFTLV ) or H-2Kb- restricted TsKb-20 peptide ( ANYKFTLV ) . CFSElow cells remained unpulsed . Subsequently , CFSEhigh cells were washed and mixed with equal numbers of CFSElow cells before injecting i . v . 15–20×106 total cells per mouse . Recipient animals were mice that had been infected or not with T . cruzi . Spleen cells of recipient mice were collected 20 h after transfer , fixed with 2 . 0% paraformaldehyde and analyzed by cytometry , using a FACSCalibur Cytometer ( BD Biosciences ) . Percentage of CFSElow ( M1 ) and CFSEhigh ( M2 ) cells were obtained using CellQuest software ( BD Biosciences ) . Percentage of specific lysis was determined using the formula: 1 - ( ( M2infected/M1infected ) / ( M2naïve/M1naive ) ) ×100% . Statistical analyses were performed using GraphPad Prism version 4 . 00 for Windows ( GraphPad Software , San Diego California USA , www . graphpad . com ) . Data were compared using a two-tailed Student's t test and are expressed as mean ± SEM . Data were considered statistically significant if p values were <0 . 05 . The LogRank test was used to compare the mouse survival rate after challenge with T . cruzi . The differences were considered significant when the P value was <0 . 05 . | Innate and acquired immune responses are triggered during infection with T . cruzi , the etiologic agent of Chagas' disease , and are critical for host survival . Parasite burden is usually controlled by the time the adaptive response becomes operational . Nevertheless , T . cruzi manages to subsist within intracellular niches and establishes a chronic infection , leading to the development of cardiomyopathy in approximately one-third of infected individuals . Recently , Toll-like receptors ( TLRs ) have been shown to recognize T . cruzi molecules and mice lacking MyD88 , the key adaptor for most TLRs , are extremely susceptible to infection . Although TLRs are known to link innate and adaptive responses , their role in the establishment of crucial effector mechanisms mediated by CD8+ T cells during T . cruzi infection has not yet been determined . We analyzed the induction of IFN-γ and cytotoxic activity in vivo in TLR2- , TLR4- , TLR9- or MyD88-deficient mice during infection , and found intact responses compared to WT mice . We also demonstrated that TLR4 is required for optimal production of inflammatory cytokines and nitric oxide and , consequently , for a better control of parasitemia levels . Understanding how TLR activation leads to resistance to infection might contribute to the development of better strategies to improve immune responses against this pathogen . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology",
"immunology/immune",
"response",
"immunology/immunity",
"to",
"infections",
"immunology/innate",
"immunity"
] | 2010 | Impaired Innate Immunity in Tlr4−/− Mice but Preserved CD8+ T Cell Responses against Trypanosoma cruzi in Tlr4-, Tlr2-, Tlr9- or Myd88-Deficient Mice |
The preparedness of health systems to detect , treat , and prevent onward transmission of Ebola virus disease ( EVD ) is central to mitigating future outbreaks . Early detection of outbreaks is critical to timely response , but estimating detection rates is difficult because unreported spillover events and outbreaks do not generate data . Using three independent datasets available on the distributions of secondary infections during EVD outbreaks across West Africa , in a single district ( Western Area ) of Sierra Leone , and in the city of Conakry , Guinea , we simulated realistic outbreak size distributions and compared them to reported outbreak sizes . These three empirical distributions lead to estimates for the proportion of detected spillover events and small outbreaks of 26% ( range 8–40% , based on the full outbreak data ) , 48% ( range 39–62% , based on the Sierra Leone data ) , and 17% ( range 11–24% , based on the Guinea data ) . We conclude that at least half of all spillover events have failed to be reported since EVD was first recognized . We also estimate the probability of detecting outbreaks of different sizes , which is likely less than 10% for single-case spillover events . Comparing models of the observation process also suggests the probability of detecting an outbreak is not simply the cumulative probability of independently detecting any one individual . Rather , we find that any individual’s probability of detection is highly dependent upon the size of the cluster of cases . These findings highlight the importance of primary health care and local case management to detect and contain undetected early stage outbreaks at source .
The last five years have seen an unprecedented number of cases of Ebola virus disease ( EVD ) , which has taken an enormous toll in terms of mortality , economic damage , disruption to other public health programs and infrastructure , and public fear and mistrust [1–3] . Months of delay and dozens of cases may occur before an outbreak is reported , as in the 2013–2016 West African outbreak [4] and the most recent outbreak in the Democratic Republic of the Congo , which is ongoing at the time of writing [5–6] . This delay raises questions about how often and how early EVD outbreaks are detected , particularly those that lead to fewer cases . We expect most EVD spillover events to be “dead ends” that do not transmit further , and most of these likely remain undetected . Spillover events from wildlife to people face myriad barriers to transmission and establishment—including host susceptibility , mobility , and onward contact with other susceptible individuals—often resulting in “stuttering chains” of transmission in people [7–8] . All diseases face the possibility of stochastic extinction upon introduction , and this is especially likely for a disease such as EVD that results in highly heterogeneous secondary transmission events [8–9] . Given a sufficiently skewed distribution of secondary infections , even a disease with a basic reproduction number ( R0 ) greater than 1 is more likely than not to die out after a single index case [10] . However , few single-generation spillover events of EVD have been documented [11] . It is difficult to assess the efficiency of current health systems in detecting , treating , and preventing onward transmission of EVD , as the number of unobserved outbreaks is by definition unknown . Here we use three distinct data sets from the 2013–2016 outbreak in West Africa , using the properties of person-to-person EVD transmission to estimate the likely true distribution of EVD outbreak sizes . We do this by simulating the early stages of outbreaks and using maximum likelihood estimation of size-dependent detection rates to link them to the reported distribution of outbreak sizes . We thereby provide estimates for 1 ) the probability of observing an EVD outbreak of a given size and 2 ) the number of small outbreaks and spillover events that are likely to have gone undetected since EVD was first reported in 1976 .
We performed a secondary data analysis of several sources of data already in the public domain in anonymized form [9 , 12–14] . We followed best practices guidelines for secondary data research issued by our institutional ethics committee , which indicated that no full ethical review was necessary . Because published estimates of the true secondary case distribution of EVD vary widely—with estimates of R0 alone ranging from subcritical ( i . e . , <1 ) to >3 [15–16]—we parameterised our simulations with three previously estimated distributions , each based on different assumptions and data from different geographic areas . Each of these previous analyses provided parameter estimates and credible intervals for a negative binomial distribution . Negative binomial distributions are commonly used to represent secondary infections [8] and can be parametrised by the disease’s basic reproduction number ( R0 ) and a dispersion parameter k measuring heterogeneity in secondary case numbers , with probability distribution f ( x ) =Γ ( x+k ) Γ ( k ) x ! ( kk+R0 ) k ( 1−kk+R0 ) x . One set of estimates was obtained from all reported exposures from over 19 , 000 cases from Guinea , Liberia , and Sierra Leone ( henceforth referred to as the full outbreak dataset ) [12] . The second estimates were derived from cases in a single district of Sierra Leone , Western Area [13] . The final set of estimates was based on chains of transmission from 152 cases in early 2014 in Conakry , Guinea [9 , 14] . A diagram of the full analysis is shown in Fig 1 , and more information about the datasets and their resulting secondary infection distributions is presented in S1 Text . For each dataset , we sampled 500 values each of R0 and the dispersion parameter k to approximate bounds supplied in the original papers ( see S1 Text ) . From each parameter set , we simulated the early stages of each of 104 outbreaks as stochastic branching processes for up to 50 generations or 57 cases ( whichever limit was reached first ) . As expected , most outbreaks died out within a few generations , representing the stuttering chains of interest to our study . The 57-case threshold was chosen based on the following rationale . To allow us to link these simulated outbreaks to observed ones , we set a cutoff value based on two assumptions: 1 ) outbreaks larger than this cutoff are always observed , and 2 ) outbreaks smaller than this cutoff die out for primarily stochastic reasons , while larger outbreaks may die out for other reasons ( such as interventions ) . Under the latter assumption , we fit our observation models according to the likelihood of any outbreak reaching a certain size rather than the ( infinitesimally small ) likelihood of large outbreaks dying out for purely stochastic reasons . After close inspection of the data , we set the initial cutoff to 57 cases to include the 1994 Gabon outbreak ( 52 cases ) , which was misidentified as a yellow fever outbreak while ongoing , and to exclude the 1996 Gabon outbreak ( 62 cases ) , which was subject to nosocomial control measures [17] . Our results are robust to the choice of cutoff value in a sensitivity analysis ( S2 Fig ) . We then modelled the probability Pr ( i ) of detecting an outbreak of size i using two possible linking functions: 1 ) the cumulative distribution function of the geometric distribution based on the probability p of detecting a single case , Pr ( i ) = 1− ( 1−p ) i ( see e . g . , [18] ) , and 2 ) a generalized logistic linking function ( Pr ( i ) = ( 1+e ( β−i ) ) −α ) . The observation function derived from the geometric distribution models assumes that the probability of detecting an outbreak is the cumulative probability of detecting at least one case , where each case has an equal and independent probability of being detected . The generalized logistic observation function allows that individual probability of detection to vary with outbreak size in a flexible way . We fit both distributions using the reported and expected ( as generated above ) distributions of outbreaks of 57 cases or smaller since 1976 using a coordinate descent method to iteratively select the expected numbers of outbreaks of each size and the parameters of the observation process ( see S2 Text for details ) [11] . We used corrected Akaike information criteria ( AICc ) to compare the two linking functions , then performed sensitivity analyses with the linking function with the lowest AICc values overall . We excluded from our distribution of reported outbreak sizes instances of laboratory infection and outbreaks of Reston virus , which follow extremely different spillover and transmission dynamics than African ebolaviruses ( see S2 Table for full list of included outbreaks ) . Finally , we performed several sensitivity and goodness-of-fit analyses . We assessed goodness of fit by simulating outbreaks and the observation process 104 times . For each simulation , we sampled outbreak sizes and the observation process until 13 reported outbreaks smaller than the cutoff were detected ( as per the data; see S2 Text ) . Our sensitivity analyses included variations in the cutoff value used to define outbreaks subject to stochastic extinction , a wider range of values of R0 and dispersion parameters , and the addition of a decreasing effective reproduction number with each generation of an emerging outbreak ( e . g . , due to control interventions ) .
Parameter estimates for the generalized logistic observation function support a sigmoidal effect of cluster size on EVD detection probability . The geometric model fit to the full outbreak data produced a median individual probability of detection of 8 . 4% per case , irrespective of cluster size . In contrast , the logistic model predicts that an isolated case has a median of 2 . 4% probability of detection , while outbreaks of sizes greater than 10 approach 100% detection ( Fig 2A ) . The generalized logistic model consistently outperformed the geometric model according to AICc ( Fig 2B ) ; all results below are from the generalized logistic model . For the full outbreak , Sierra Leona , and Guinea datasets , respectively , we estimate that medians of 67 ( range 35–283; Fig 3C ) , 26 ( range 15–37; Fig 3D ) , and 118 . 5 ( range 75–192; Fig 3E ) spillover events have gone undetected since EVD was first reported . These represent overall detection probabilities of 26 . 4% ( range 7 . 8–40 . 7% ) , 48 . 0% ( range 39 . 3–61 . 5% ) , and 16 . 8% ( range 11 . 1–24 . 2% ) , respectively . Our model predicts that most of these undetected outbreaks were dead-end zoonotic spillovers causing a single human case . We estimate medians of 56 such undetected spillovers from the full outbreak data ( range 28–263 , corresponding to detection probabilities of 0 . 1–6 . 7% ) , 22 from the Sierra Leone data ( range 14–31 , corresponding to detection probabilities of 6 . 0–12 . 5% ) , and 101 . 5 from the Guinea data ( range 64–161 , corresponding to detection probabilities of 1 . 2–3 . 0% ) . Simulations of outbreak sizes and the observation process produce predicted observation counts concordant with the data ( see S1 Fig ) . Our results were consistent across several sensitivity analyses . Varying the outbreak size cutoff value from 5 cases to 55 cases ( for the full outbreak data ) only resulted in a total difference in estimated detection probabilities of about 8% of the minimum value ( from 24 . 2% at a cutoff of 5 cases to 26 . 2% at 50 or 55 cases; see S2 Fig ) . We did not extend the range of considered values higher than our selected cutoff of 57 cases due to the sparseness of the data ( e . g . , there are no outbreaks between size 66 and size 124; see S2 Table ) and the documented control efforts used to limit these outbreaks . Repeating our analysis for the full outbreak data across a wider range of plausible dispersion parameters and R0 values still produces detection probabilities in a similar range . The range of median estimates for the proportion of small outbreaks and spillover events within common existing estimates of R0 ( 1–1 . 5 by intervals of 0 . 25 , inclusive ) and k ( 0 . 1–0 . 6 by intervals of 0 . 25 , inclusive ) for EVD is 13%-45% . To estimate greater than 50% detection requires that R0 be >2 , that k be >1 . 1 , or that both R0>1 . 5 and k>0 . 85; all of these values are greater than most existing estimates ( S3A Fig ) . Finally , we tested the effect on our results of allowing outbreaks to become slightly less infectious ( e . g . , due to successful attempts to control the outbreak , behavioural modifications , or pathogen evolution to become less virulent ) by having Reff decay 10 , 20 , or 30% per generation ( S3B Fig ) . At very high values of the dispersion parameter ( i . e . , with minimal superspreading ) , the addition of this decay decreases the median proportion of detections . Within the same parameter ranges of common estimates ( R0 1–1 . 5; k 0 . 1–0 . 6 , inclusive ) , median estimates for the detected proportion of small outbreaks and spillover events with Reff decay of 0 . 3 range from 8% to 34% .
Our median estimates from all three datasets suggest at least half of all EVD spillover events ( and possibly as many as 83% ) have gone undetected . Although most of these spillover events have been ‘dead-ends’ or very small outbreaks , our models suggest this could represent well over 100 cases ( from a minimum of 16 from the Sierra Leone data to a maximum of 317 from the full outbreak data ) . While the specific estimates of these missed small outbreaks are highly sensitive to assumptions about the underlying secondary case distribution , the central prediction is robust to different datasets , methodological choices such as the cutoff value for stochastic outbreaks , and a wide range of plausible R0 values and dispersion parameters . We found consistently lower AICc values for the logistic observation model , which assumes a dependence of individual detection probability on cluster size . This suggests that outbreak surveillance is not adequately modelled as the combination of independent individual detection probabilities . We predict instead that individual cases are more likely to be correctly detected when part of a cluster of cases; identifying EVD cases that do not exist within clusters ( or have not yet caused clusters ) may be even more difficult to detect than previously assumed . Across three datasets with very different estimates of the underlying offspring distribution ( e . g . , values of R0 from subcritical for the Guinea data to 2 . 4 for the Sierra Leone data ) , our analysis consistently predicts that most EVD spillover events and small outbreaks are not detected . That all three analyses from which we have drawn our estimates generate such different parameters highlights the difficulty of estimating them in a naïve , uncontrolled scenario when relevant data is collected in outbreak settings . However , the importance of superspreading is a consistent result of models applied to EVD , even at very early stages of outbreaks [19]; our sensitivity analysis ( S3 Fig ) predicts that at least 40% of spillover events are undetected as long as moderate superspreading occurs ( k<1 ) and R0<2 . The true value of R0 for a typical spillover case is most likely lower than those used here , due to the unusual epidemiology of the West African outbreak [20]; the very low human population densities at which spillover events often occur [21–22]; and potential asymptomatic cases and uncharacteristic low-transmitting cases , which may not be sufficiently accounted for in existing R0 estimates [23] . Additionally , although asymptomatic infections are uncommon among contacts of human EVD cases [24] , they may be a more common outcome of direct zoonotic spillover , e . g . , with an ebolavirus strain that is poorly adapted to human hosts , as appears to be the case with Reston virus [7 , 25] . Accounting for outbreaks that become less infectious over time ( due to , e . g . , control interventions , susceptible depletion , or host behavioural modification ) causes our estimated outbreak detection rates to drop . We therefore expect that models combining our analysis with more accurate dynamics of EVD and its control would estimate lower detection rates than we present here . It is possible that all these factors render our results underestimations of the true frequency of spillover . If our underestimation is particularly extreme , it is possible hundreds or thousands of EVD spillovers have gone undetected , potentially explaining high seroprevalence of Ebola virus antibodies in some populations [26] . Due to these and other assumptions with less clear consequences , we intend this analysis not as a precise quantification of rates of EVD detection , but rather as a demonstration of the high probability that many spillovers go undetected and that many large outbreaks are not detected early . The regions from which our data come may be unrepresentative in ways we have not considered; no EVD spillovers have been reported in Sierra Leone , so the generalizability of the Sierra Leone dataset to the typical spillover case is unknown . Finally , we assume that each documented outbreak had a single index case from spillover . While we know of no outbreaks with multiple index cases , the origins of some have not been fully traced , and outbreaks originating from a multiple-spillover event are less likely to die out stochastically [27] . There is a clear need to improve outbreak detection and rapid response , and investment in these areas is among the most efficient ways of reducing EVD mortality [28] . Paving the way , Uganda instated a viral haemorrhagic fever surveillance programme in 2010 that has increased the number of outbreaks detected while reducing their mean size and mean time to confirmation [29] . We note that Uganda is one of only two countries to have detected spillovers resulting in a single case ( S2 Table ) , which we expect to be the true most common outbreak size . This potential success does , however , highlight the lack of spatial resolution as a limitation of our study . The small number of total observed EVD spillover events means this analysis is uninformative on a country-by-country basis , obscuring potential differences in detection between countries . Additional work could consider more systematically spatial variation in contexts likely to lead to spillover and onward transmission , as well as barriers to treatment and reporting . Our estimates suggest that most spillover events and small outbreaks of EVD are not reported to international bodies but rather are handled locally , likely as fevers of unknown origin or mischaracterised as more common causes of fever ( e . g . , malaria ) [30] . Supporting core public health and sanitation infrastructure in the areas where spillover is likely to occur may prove vital to preventing the onward transmission of these unseen index cases . Furthermore , promoting the safe management of fever and enhancing local diagnostic capacity has the potential to improve facility-based national surveillance systems and ultimately increase the chance of early detection of EVD outbreaks , both large and small . | Emerging infectious diseases are often not investigated in rural Africa unless outbreaks involve a sizeable number of cases . A number of different Ebola virus disease ( EVD ) outbreaks have been reported in the literature and in surveillance reports since its discovery in 1976 . The majority of the reports are of large outbreaks . Given the low reported rate of transmission of Ebola , and the high frequency with which cases infect no one else , one might expect most outbreaks to be very small ( <5 people ) . This is the first study to the authors’ knowledge that quantitatively estimates the number of undetected EVD outbreaks or probabilities of EVD outbreak detection by outbreak size . Although the total amount of evidence in this area is still limited , this study’s main result—that at least half of EVD outbreaks go undetected—is consistent under many different sets of assumptions . This is the most thorough estimation of EVD outbreak detection to date and corroborates the majority of more qualitative work on EVD surveillance , suggesting greater investment in primary health care and local surveillance will be important to detect EVD outbreaks early and consistently . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"infectious",
"disease",
"epidemiology",
"pathogens",
"tropical",
"diseases",
"geographical",
"locations",
"microbiology",
"ebola",
"hemorrhagic",
"fever",
"viruses",
"filoviruses",
"pr... | 2019 | Estimating undetected Ebola spillovers |
Herpes Simplex Virus 1 ( HSV1 ) is amongst the most clinically advanced oncolytic virus platforms . However , efficient and sustained viral replication within tumours is limiting . Rapamycin can stimulate HSV1 replication in cancer cells , but active-site dual mTORC1 and mTORC2 ( mammalian target of rapamycin complex 1 and 2 ) inhibitors ( asTORi ) were shown to suppress the virus in normal cells . Surprisingly , using the infected cell protein 0 ( ICP0 ) -deleted HSV1 ( HSV1-dICP0 ) , we found that asTORi markedly augment infection in cancer cells and a mouse mammary cancer xenograft . Mechanistically , asTORi repressed mRNA translation in normal cells , resulting in defective antiviral response but also inhibition of HSV1-dICP0 replication . asTORi also reduced antiviral response in cancer cells , however in contrast to normal cells , transformed cells and cells transduced to elevate the expression of eukaryotic initiation factor 4E ( eIF4E ) or to silence the repressors eIF4E binding proteins ( 4E-BPs ) , selectively maintained HSV1-dICP0 protein synthesis during asTORi treatment , ultimately supporting increased viral replication . Our data show that altered eIF4E/4E-BPs expression can act to promote HSV1-dICP0 infection under prolonged mTOR inhibition . Thus , pharmacoviral combination of asTORi and HSV1 can target cancer cells displaying dysregulated eIF4E/4E-BPs axis .
Oncolytic viruses are promising immunotherapeutic agents for the treatment of cancer [1] . HSV1 application is amongst the most advanced and successful oncolytic platforms; with Amgen’s oncolytic HSV1 talimogene laherparepvec ( T-Vec , Imlygic ) being the first oncolytic virus to receive FDA and EMA ( Food and Drug Administration and European Medicines Agency ) approval in October 2015 [2] . Numerous groups are developing oncolytic HSV1 variants , including the pre-clinical development of an infected-cell-protein-0-deleted HSV1 ( HSV1-dICP0 ) [3] . Rapid viral clearance within tumour tissues constitutes a limitation for oncolytic viral therapies . Thus , potentiating viral replication could further increase efficacy [4] . We , and others reported that the drug rapamycin , an allosteric inhibitor of mTORC1 , and approved therapy for certain cancers , improves oncolytic viral replication within tumours through suppression of innate immunity and type-I IFN production [5–7] . The protein kinase mTOR , which integrates extra- and intracellular signals to affect cellular growth , proliferation , metabolism , and survival , exists in two complexes: mTORC1 , which is sensitive to rapamycin and regulates mRNA translation , and mTORC2 , which primarily controls actin cytoskeleton organization [8] . Evidence for the role of mTORC1 signaling in innate immunity emerged from the findings that rapamycin suppresses type-I IFN in plasmacytoid dendritic cells ( pDCs ) , which are the major producers of systemic type-I IFN [9] . Thereafter , it was shown that silencing or chemically inhibiting upstream activators of mTORC1 compromises innate immunity [5 , 9–11] , and that genetic deletion of the mTORC1 downstream targets ribosomal S6 kinases ( S6Ks ) results in impaired type-I IFN response [5 , 12] . Conversely , depletion of the upstream mTORC1 repressors TSC1/2 , or the absence of the translational repressors eIF4E-binding proteins 1 and 2 ( 4E-BP1/2 ) downstream of mTORC1 , leads to enhanced interferon-regulatory factor 7 ( Irf-7 ) mRNA translation and type I IFN production [13–16] . These studies have highlighted the mTOR signaling pathway as a critical component of innate immunity by controlling translation initiation of antiviral mRNAs [17] . Initiation of protein synthesis generally involves the recognition of the mRNA 5’-m7G-cap structure by the eIF4F complex; consisting of eIF4E , a cap-binding protein; eIF4A , an RNA helicase and eIF4G , a scaffolding protein that recruits eIF3 and the 40S ribosomal subunit to the mRNA [18] . eIF4E can be sequestered from eIF4F by 4E-BPs; in conditions of mTORC1 inhibition , hypophosphorylated ( activated ) 4E-BP1/2 , and inducible 4E-BP3 , bind strongly to eIF4E and repress translation , whereas hyperphosphorylated ( inactivated ) 4E-BPs do not , allowing formation of the eIF4F complex on the mRNA and initiation of protein synthesis [19 , 20] . In cancer , mRNA translation is frequently dysregulated because of increased eIF4E protein levels and/or elevated phosphorylation of 4E-BPs due to a hyperactivated mTORC1 pathway [21] . Targeting the aberrant mRNA translation in cancer represents an attractive strategy; while the allosteric mTORC1 inhibitor rapamycin is a poor activator of 4E-BPs , ATP-competitive active-site mTORC1 and 2 inhibitors ( asTORi ) , are superior to rapamycin in inhibiting mTORC1/2 , activating 4E-BPs , and providing potent anti-cancer effects [22] . Interestingly , the ratio of eIF4E to 4E-BPs was shown to determine the anti-proliferative efficacy of asTORi [23 , 24] . In contrast to rapamycin , which was previously reported to increase HSV1 infection in some cancer cells [7] , asTORi treatment strongly limits HSV1 replication in normal cells [25 , 26] . Strikingly , while suppressing viral replication in normal fibroblasts and epithelial cells , we observed that prolonged exposure to asTORi dramatically increases HSV1-dICP0 infection in cancer cells . asTORi treatment results in reduction of type-I IFN responses in both cancer and non-transformed cell lines , and strongly inhibits viral mRNA translation in normal cells . In contrast , viral protein synthesis persists in cancer cells and cells transduced to either increase eIF4E levels or deplete 4E-BP1/2 , ultimately resulting in enhanced viral replication and spread . Importantly , the combination of asTORi and HSV1-dICP0 reduces tumour size of an aggressive syngeneic breast cancer mouse model . Thus , our data reveal that cancer cells harboring altered mRNA translation via dysregulated eIF4E/4E-BPs axis can be targeted by the combination of asTORi and HSV1-dICP0 .
Rapamycin augments the oncolytic potential of several viruses [5–7 , 27–29] . asTORi are currently in clinical development [22 , 30] . Therefore , we sought to determine whether the asTORi PP242 and INK1341 could augment the infection of different oncolytic viruses in cancer cells . The human glioblastoma cell line U251N or the mouse mammary carcinoma cell line 4T1 were infected with several oncolytic viruses including vesicular stomatitis virus ( GFP-expressing VSVΔ51M ) , myxoma virus ( GFP-expressing MV ) , vaccinia virus ( GFP-expressing JX594 ) and HSV1 ( infected cell protein 0 ( ICP0 ) -defective oncolytic HSV1 expressing GFP ( HSV1-dICP0 ) ) in the presence or absence of asTORi . Our data showed that asTORi treatment limited VSVΔ51M , MV , and vaccinia virus infection , but unpredictably strongly increased HSV1-dICP0 infection and spread by 48 hours post-treatment ( S1A–S1D Fig ) . This was unexpected as it had been previously reported that asTORi strongly suppress wild type HSV1 infection of primary human fibroblasts and primary mouse embryonic fibroblasts ( MEFs ) [25 , 26] . To address this conundrum , we infected primary human foreskin fibroblasts ( HFF ) and MEFs with wild type HSV1 in the presence of rapamycin , or the asTORi PP242 . As reported , wild type HSV1 infection was repressed ( more than 10-fold ) by treatment with PP242 in normal cells ( Fig 1A ) . In stark contrast , this treatment markedly enhanced ( 5–10 fold ) wild type HSV1 infection in the human glioblastoma cell lines U251N and HTB-14 ( Fig 1B ) . Similar to asTORi , rapamycin enhanced HSV1 infection in U251N ( Fig 1B ) and 4T1 cells ( S1D Fig ) , however unlike asTORi , rapamycin had only a partial effect in suppressing virus expression in MEFs ( Fig 1A ) . We then examined the infection of various cancer cell lines with HSV1-dICP0 [3] . Supporting our data in U251N and 4T1 cells , asTORi ( PP242 , INK1341 , INK128 or Torin1 ) treatment increased HSV1-dICP0 infection and viral protein levels in multiple transformed human cell lines ( HEK293T , HCT116 , Huh7 ) and mouse mammary tumour cell lines ( 4T1 , NT2196 ) by 48–72 hours post-infection . In contrast , viral infection and protein synthesis were repressed in several cell lines derived from normal tissues that were treated with asTORi: fibroblasts ( MEF , HFF ) , the non-transformed neuroblast cell line SHEP [31 , 32] , and the mouse epithelial mammary cell line NMuMG ( Fig 1C and 1D , S1E and S1F Fig and S2A–S2D Fig ) . To better understand the kinetics of the enhanced viral infection in the presence of the asTORi PP242 , we used IncuCyte real-time imaging to monitor GFP-expressing HSV1-dICP0 in both the NMuMG cell line and the NeuT-transformed NMuMG cell line NT2196 [33] . We found that the increased GFP expression in the transformed NT2196 cells relative to DMSO-treated cells begins only at 24 hours post-infection , and that asTORi strongly suppressed the virus in the normal NMuMG cells ( S1E Fig ) . Importantly , this effect was not unique to HSV1-dICP0 , as g34 . 5-deleted HSV1-1716 , that has undergone clinical development [34] , also led to an increase in fluorescence upon treatment with PP242 as compared to DMSO control in the transformed NT2196 cell line ( S1E Fig ) . However , the increase with g34 . 5-deleted HSV1-1716 was not as dramatic as seen with HSV1-dICP0 . Furthermore , g34 . 5-deleted HSV1-1716 led to no detectable GFP fluorescence in the normal breast epithelial cell line NMuMG . Therefore , to characterize the differences between normal and transformed cells , we pursued experiments primarily using HSV1-dICP0 . By comparing 4T1 and NT2196 cells to NMuMG , we confirmed that combining asTORi with HSV1-dICP0 potentiates viral infection ( measured by Western blot , GFP expression , and titration ) , and oncolysis ( measured by crystal violet and trypan blue exclusion ) in the transformed mammary cells ( Fig 1D , Fig 2A–2C , S1E and S1F Fig and S2E Fig ) . Furthermore , intratumoural injection of luciferase-expressing HSV1-dICP0 in mice bearing late-stage 4T1 tumours showed detectable luciferase expression only in tumours of mice that had been administered asTORi ( Fig 2D ) . Correspondingly , the combinatorial pharmacoviral treatment resulted in significant reduced tumour growth ( by ~50% at day 18 post-implantation , p = 0 . 019 and p = 0 . 021 over PP242 or HSV1-dICP0 monotherapy respectively ) , and moderately prolonged animal survival bearing this aggressive breast tumour as compared to either pharmacological or viral single therapy ( median survival 15 days post-treatment for the combination compared to 8–10 days for PP242 or HSV1-dICP0 monotherapies , respectively , p = 0 . 045 ) ( Fig 2E and 2F ) . Collectively , these data demonstrate that mTOR inhibition by asTORi suppresses HSV1 replication in non-transformed fibroblasts and epithelial cells ( as previously reported [25 , 26] , but surprisingly the same treatment causes a robust , albeit delayed , enhancement of HSV1 replication in cancer and transformed cells . Inhibition of mTORC1 by rapamycin limits type-I IFN responses [5 , 6 , 9] . Additionally , mTORC2 has been implicated in the regulation of innate immunity [35] . Thus , asTORi that strongly impede the activation of both mTORC1 and 2 , and potently activate 4E-BPs , are expected to exhibit robust inhibition of type-I IFN production and antiviral responses in cells . To assess whether asTORi treatment and activation ( hypophosphorylation ) of 4E-BPs limits type-I IFN signaling , we transfected synthetic double stranded poly ( I:C ) RNA to mimic natural infection by viruses , or infected primary MEFs and the glioma U251N cells with wild type HSV1 in the presence of rapamycin or the asTORi PP242 . As previously reported [36 , 37] and shown above ( Fig 1A and 1B ) , asTORi treatment resulted in complete dephosphorylation of 4E-BP1 in both MEFs and U251N cells , whereas rapamycin exerted only a partial effect on 4E-BP1 phosphorylation ( Fig 3A and 3B , top panels ) . RT-PCR for Ifn-β after poly ( I:C ) RNA transfection ( Fig 3A and 3B , middle panels ) , or wild type HSV1 infection ( S3A Fig ) , revealed that cells treated with asTORi had limited induction of Ifn-β mRNA expression as compared to rapamycin or DMSO control . Similarly , the ~4–5 fold increase in interferon-stimulated response element ( ISRE ) promoter activity and type-I IFN production in presence of poly ( I:C ) RNA was blocked in PP242-treated cells while a partial reduction was detected in rapamycin-treated cells ( Fig 3A and 3B , bottom panels ) . Using conditioned media from MEFs or U251N cells treated with poly ( I:C ) RNA and DMSO ( control ) , rapamycin , or asTORi , a reduced protection from subsequent wild type HSV1 infection was primarily observed in asTORi treated cells ( Fig 3C and S3B Fig ) . Similarly , RT-PCR for HSV1 gC transcripts and ISRE reporter activity from the human glioblastoma HTB-14 cells infected with wild type HSV1 , confirmed that PP242 treatment limits the induction of ISRE reporter activity , with a corresponding increase in HSV1 gC transcript levels ( Fig 3D ) . Furthermore , the induction of Ifn-β mRNA levels measured by RT-qPCR was suppressed to a similar extent upon asTORi treatment and HSV1-dICP0 infection in 4T1 , NT2196 , and NMuMG cells ( Fig 3E ) . Additionally , the inhibitory effect of asTORi on type I IFN production was maintained in poly ( I:C ) -stimulated normal human foreskin fibroblasts HFF and human glioblastoma cell lines U343 and U373 ( S3C Fig ) . To further demonstrate that dual mTORC1/2 inhibitors impair the innate immune response , we performed polysome profiling and translation reporter assays of cellular and viral mRNAs . These experiments showed that Irf7 mRNA is excluded from polysomes in 4T1 and NT2196 cells infected with HSV1-dICP0 in the presence of the asTORi PP242 ( S3D–S3G Fig ) . In contrast , the interferon-induced gene Isg15 was rather weakly induced transcriptionally upon asTORi treatment as compared to control ( S3G Fig ) . Importantly , these experiments also demonstrated that HSV1-dICP0 viral mRNAs ( ICP4 , gC , TK ) are highly transcribed and more abundant in polysome fractions when 4T1 or NT2196 cancer cells were treated with asTORi as compared to DMSO control ( S3E–S3G Fig ) . To examine the translational repression of Irf7 mRNA versus HSV1 genes , we generated reporter constructs by merging the 5’ UTR of Irf7 , or those of HSV1-TK or ICP0 , to a luciferase reporter gene , and measured luciferase expression upon DMSO ( control ) versus asTORi treatment in 4T1 cells . TK and ICP0 were selected since they have well annotated 5’ UTR that could be readily cloned into our reporter system . As expected from previous data using the asTORi Torin1 or silencing of eIF4E [38] , these experiments showed that the reporter construct containing the 5’ UTR of Irf7 , which is highly structured [14] , was repressed translationally , in contrast to the reporter constructs containing the ICP0 or the TK 5’ UTR ( S3H Fig ) . These results demonstrate that mTORC1/2 inhibition by asTORi strongly limits innate antiviral responses and production of type-I IFN in both normal and cancer cells . Our results show that asTORi treatment suppresses type-I IFN signaling in both normal fibroblast/epithelial cells and cancer cell lines . To determine whether early infection of HSV1-dICP0 was accelerated by asTORi treatment , we assessed the mRNA levels of the immediate early viral gene ICP4 at 8 hours post-infection in 4T1 , NMuMG and NT2196 cells treated either with DMSO or the asTORi INK1341 . RT-qPCR data revealed no significant changes in ICP4 mRNA levels between the conditions used for all cell lines , suggesting that viral entry was generally not affected by asTORi treatment ( S4A Fig ) . Therefore , we reasoned that the specific augmentation of HSV1-dICP0 replication in cancer cells cannot be explained by either a more pronounced reduction in type-I IFN responses , or an increase in viral entry . Protein synthesis is frequently dysregulated in cancer and contributes to resistance to therapeutic drugs , particularly mTOR inhibitors , by allowing the continued translation of a subset of mRNAs involved in cell proliferation [23 , 24 , 37] . Interestingly , a recent report demonstrated that transient fasting , a condition that causes reduced mTORC1 activity , enhances the replication of oncolytic HSV1 in glioblastoma cells , which was proposed to be mediated by dysregulated protein synthesis in cancer cells [39] . We examined global protein synthesis by supplementing with [35S]-Met in the growth media of 4T1 , NT2196 and NMuMG cells infected with HSV1-dICP0 and treated with the asTORi PP242 . PP242 and HSV1-dICP0 treatment resulted in reduction of global protein synthesis , an effect that was more pronounced and sustained in NMuMG cells as compared to the transformed 4T1 and NT2196 cell lines ( ~50% [35S]-Met incorporation over DMSO control for NMuMG , versus ~70% and 85% for 4T1 and NT2196 , at 24 hours post-infection respectively ) ( Fig 4A ) . As previously observed for early time points post-infection in cancer cells ( S1A and S1E Fig ) , HSV1-dICP0 protein levels at 24 hours post-infection were comparable in control ( DMSO ) and PP242-treated 4T1 and NT2196 cancer cells , but were barely detected in the normal epithelial cell line treated with PP242 ( Fig 4A ) . In agreement with the [35S]-Met incorporation results , polysome profiling analyses showed that PP242 treatment engendered a stronger reduction in polysomal RNA in NMuMG cells than in 4T1 and NT2196 cell lines ( Fig 4B ) . PP242 treatment also reduced [35S]-Met incorporation to a higher extent in HFF as compared to U251N , and suppressed wild type HSV1 protein expression only in HFF ( S4B Fig ) . Interestingly , PP242 inhibited HSV1-dICP0 protein expression in the non-transformed NMuMG cells at low concentrations ( 0 . 4–0 . 6μM ) , while exerting the opposite effect on transformed NT2196 cells up 2μM . However , higher concentrations of asTORi ultimately reduced HSV1-dICP0 infection in certain cancer cell lines ( S4C and S4D Fig ) , suggesting that excessive mTOR suppression can limit HSV1-dICP0 protein synthesis and concomitant infection . These data indicate that asTORi treatment induces a stronger and more sustained translational repression in normal cells than in cancer cells , which ultimately dampens efficient HSV1 protein synthesis and replication . Transformed cells maintain higher levels of protein synthesis under asTORi treatment , recover more quickly , and sustain HSV1 mRNA translation thereby promoting the replication of the virus . The ratio of eIF4E/4E-BP protein expression determines the anti-proliferative efficacy of mTOR inhibitors [23 , 24 , 37 , 40] . Elevated eIF4E expression , in relation to 4E-BP1/2/3 , renders cells more resistant to the repression of proliferation and mRNA translation by asTORi . Interestingly , asTORi inhibition of wild type HSV1 in MEFs was previously shown to be dependent on 4E-BP1 expression [26] . Therefore , we hypothesized that dysregulated protein synthesis due to altered eIF4E/4E-BP ratio in cancer cells is responsible for the differences in HSV1-dICP0 infection between normal and cancer cells treated with asTORi . Initially , we sought to compare the ratio of eIF4E to 4E-BPs in normal and transformed mouse mammary cells lines . While eIF4E and 4E-BP1 protein expression appeared similar between these cells , the normal NMuMG cell line exhibited higher expression of 4E-BP2 ( and a slight increase of 4E-BP3 upon mTOR inhibition ) in comparison to the transformed 4T1 and NT2196 ( S4E Fig ) . To test the hypothesis that elevated eIF4E or reduced 4E-BPs protein expression contributes to HSV1-dICP0 viral propagation in the presence of asTORi , we transduced 4T1 , NT2196 , and NMuMG cells with constructs expressing V5-tagged eIF4E , 4E-BP1 or the control Blue Fluorescent Protein ( BFP ) . Strikingly , V5-4E-BP1 overexpression ( i . e . reduced eIF4E/4E-BP1 ratio ) prevented the asTORi-mediated increase of HSV1-dICP0 protein synthesis in transformed cells ( 4T1 and NT2196 ) , and resulted in greater repression of HSV1-dICP0 infection by asTORi in normal cells ( NMuMG ) . In contrast , eIF4E overexpression in the NMuMG cells led to sustained viral protein synthesis during asTORi treatment ( Fig 5A and 5B ) . To confirm the contribution of the eIF4E/4E-BP axis in potentiating HSV1-dICP0 infection during asTORi treatment , we knocked down eIF4E , 4E-BP1 or 4E-BP2 in 4T1 , NT2196 and NMuMG cells . Silencing eIF4E led to a strong repression of HSV1-dICP0 infection in cancer cells treated with asTORi , while silencing 4E-BP1 or 4E-BP2 had a reverse effect ( Fig 6A and 6B and S4F Fig ) . To examine this response in other cell models , we used NIH3T3 mouse fibroblasts that overexpress eIF4E [41] . PP242 treatment in non-transformed control NIH3T3 cells limited wild type HSV1 infection . However , transformed NIH3T3 cells with excess levels of eIF4E were resistant to PP242-mediated suppression of wild type HSV1 protein expression ( S5A Fig ) . Similar results were obtained using the non-transformed human neuroblast cell line SHEP [31 , 32] , where PP242 treatment decreased wild type HSV1 protein levels in cells treated with empty vector control , but not in cells overexpressing eIF4E ( S5B Fig ) . As expected , a stronger suppression of wild type HSV1 protein expression was also observed in SHEP cells overexpressing 4E-BP1 ( S5C Fig ) . Consistent with these findings , silencing eIF4E or increasing expression of 4E-BP1 in U251N cells diminished the elevated infection of HSV1-dICP0 in presence of asTORi , whereas eIF4E overexpression had the reverse effect ( S5D–S5H Fig ) . These data demonstrate that cells with overexpression of eIF4E or reduced expression of 4E-BPs , can sustain HSV1 protein synthesis during mTOR inhibition . The results further support the notion that in cancer cells with dysregulated eIF4E/4E-BP ratio , administration of asTORi could promote the replication of HSV1-dCIP0 .
In this study we investigated the pharmacoviral combination of asTORi and HSV1-dICP0 . asTORi treatment suppressed type-I IFN responses in normal and cancer cells . The treatment impaired the replication of HSV1-dICP0 in non-transformed cells , but infection was strikingly augmented in cancer cells . Furthermore , in an aggressive syngeneic mammary cancer mouse model , the combination of asTORi and HSV1-dICP0 reduced tumour size and prolonged survival compared to either monotherapy . Mechanistically our data show that asTORi treatment initially represses global mRNA translation , but HSV1-dICP0 protein expression persists in cancer cells and cells with dysregulated eIF4E/4E-BP ratio . Since cancer cells with high eIF4E expression demonstrate resistance to the anti-proliferative effects of asTORi [23] , combining dual mTORC1 and 2 inhibitors with HSV1-dICP0 could potentially provide the selective advantage of increasing viral oncolysis of mTOR-inhibitor resistant cancer cells . The eIF4E/4E-BP1 ratio is a key determinant of cell response to mTOR inhibitors as tumour cells with a high eIF4E/4E-BP1 ratio exhibit low sensitivity to asTORi-induced cell proliferation arrest [23 , 24 , 37] . Furthermore , in hepatocellular carcinoma cells or aggressive B-cell lymphomas , high eIF4E/4E-BP1 confers resistance to metformin-induced apoptosis or mTOR inhibitors , respectively [24 , 40] . In addition , the mTORC1 signaling pathway is often hyperactivated in cancer resulting in elevated phosphorylation of 4E-BPs [21] . mTORC1 activity was previously demonstrated to be required for wild type HSV1 infection and replication as primary human and mouse fibroblasts treated with asTORi were resistant to wild type HSV1 infection [25 , 26] . Interestingly , Moorman and Shenk showed that primary MEFs depleted of 4E-BP1 produced equal yields of HSV1 in the presence or absence of asTORi [26] . We also report here that asTORi treatment potently repressed wild type HSV1 and HSV1-dICP0 in several normal cell lines , but found that viral replication and spread is markedly augmented under prolonged asTORi treatment within cancer cells . We further show that varying the cellular expression levels of eIF4E , 4E-BP1 or 4E-BP2 results in differential HSV1-dICP0 infectivity in the presence of asTORi . Notably , forced expression of eIF4E was recently reported to strengthen the potency of oncolytic measles virus [42] . Since wild type HSV1 infection is known to stimulate mTORC1-dependent 4E-BP1 inactivation [25] , our results suggest that dysregulated eIF4E/4E-BP ratio and/or hyperactivated mTORC1 in cancer cells contribute to sustain , and ultimately favor viral protein synthesis in the presence of asTORi over antiviral mRNA translation . This model is consistent with previous data showing translation of antiviral Irf7 mRNA to be highly sensitive to 4E-BPs [14–16] , and to asTORi treatment as presented here . Nonetheless , it is noteworthy that overexpression or silencing of eIF4E in cells can induce a compensatory response vis-a-vis 4E-BP1/2 expression or 4E-BP1 phosphorylation [23 , 43] . We did observe changes in expression and phosphorylation of 4E-BP1 upon altering the levels of eIF4E in the transduced cells treated with asTORi . This compensatory mechanism may have impacted partly on the replication of HSV1-dICP0 . Recently , short-term fasting was shown to enhance oncolytic HSV1 replication in glioblastoma cell lines and tumours; critically , this was not observed in normal astrocytes [39] . 4E-BP1 phosphorylation was sustained during fasting in the glioblastoma cell lines , which was not the case in normal astrocytes [39] , highlighting a resistance to mTORC1 inhibition and translation repression phenotype in transformed cells , similar to that described here . High eIF4E versus 4E-BP expression is suggested as a cancer prognosis measure [44] , and we have shown that tumour samples of patients with hepatocellular carcinoma exhibit variable eIF4E/4E-BP1 ratios [24] . Thus , it is plausible that certain types of tumours could be selectively targeted by the pharmacoviral approach of dual mTORC1/2 inhibition and oncolytic HSV1 . In addition , our results suggest that alterations in the mRNA translation machinery , in the context of elevated eIF4E or decreased 4E-BP expression , can promote HSV1 infection in the presence of asTORi ( S6 Fig ) . In our study we used an ICP0-deleted HSV1 that has been investigated and developed as an oncolytic virus platform but that has not been translated to the clinic [3 , 45–47] . A limitation of our study is that the oncolytic HSV1 platforms that have progressed to clinical trials are attenuated through gamma34 . 5 ( g34 . 5 ) gene deletion . Using the clinically relevant g34 . 5-deleted HSV1-1716 [34] , we have observed a similar augmentation of viral infection by asTORi treatment in the transformed NT2196 cell line . Nonetheless , additional experiments with g34 . 5-deleted HSV1 are necessary to fully address the clinical implications of the findings presented here . Furthermore , oncolytic HSV1 research is now focused on improving immunotherapeutic anti-tumour responses , and the virus is administered together with checkpoint inhibitors in clinical trials [48] . Finding the right balance of immunosuppression during delivery of the oncolytic virus ( to minimize innate antiviral responses ) while retaining the humoral immunity required for an efficient and durable anti-tumour response will prove challenging . Indeed , “pre-conditioning” with the immunosuppressive drug cyclophosphamide during viral delivery has been shown to increase viral loads in several pre-clinical studies and early-phase trials [49] . However , this regimen may not elicit potent anti-tumour immune responses . Another concern is that asTORi treatment can reactivate latent HSV1 infections , the sequelae of which could complicate patient outcomes . This has been observed in a latently infected neuron culture model [50] as well as in organ transplant patients treated with immunosuppressive agents including mTOR inhibitors [51] . Despite these concerns , augmenting the initial phase of viral replication specifically within tumour tissues , either through transient systemic or localized intra-tumoural mTOR inhibitor therapy , could provide the added benefit of reducing cancer cell proliferation and inducing a subsequent increase in viral replication , which upon drug removal could elicit a more potent therapeutic immune response [52] . Such transient or tumour-targeted mTOR inhibition would need to be investigated in additional immune-competent cancer animal models to determine the correct dosing and its potential therapeutic efficacy in combination with oncolytic HSV1 . Finally , other complementary pathways ( e . g . MEK/ERK and MNK/eIF4E phosphorylation ) can become activated during prolonged mTOR inhibition or upon infection by viruses [21 , 53 , 54] . While we did not assess the contribution of these alternative signaling cascades here , additional studies are required to investigate their functions in cellular and HSV1 mRNA translational control during mTOR inhibition . In conclusion , we have shown that asTORi treatment induces a cellular state in transformed cell lines or cells exhibiting a dysregulated eIF4E/4E-BPs axis that HSV1 usurps for increased replication .
The human cell lines: foreskin fibroblast HFF , transformed embryonic kidney HEK293T , malignant glioblastomas U251N and HTB-14 , hepatocellular carcinoma Huh7 , the non-transformed neuroblasts SHEP , and colon carcinoma HCT116 were obtained from ATCC and cultured in Dulbecco's modified Eagle's medium ( DMEM-Wisent ) supplemented with 10% fetal bovine serum ( Wisent ) and 100 units/ml Penicillin-Streptomycin ( Wisent ) . The transformed murine mammary gland cell line 4T1 ( ATCC ) was propagated in Roswell Park Memorial Institute medium 1640 ( RPMI1640-Wisent ) supplemented with 10% fetal bovine serum and 100 units/ml Penicillin-Streptomycin . The normal murine mammary gland cells NMuMG and the NeuT transformed NMuMG , NT2196 , described in [33] were generously provided by Dr . William J . Muller . Both cell lines were cultured using the same DMEM medium as before but with the addition of 10mM HEPES pH7 . 5 ( Wisent ) and 10μg/ml insulin ( Wisent ) . NT2196 medium was also supplemented with 1μg/ml puromycin ( Bioshop ) . The mTOR inhibitors Rapamycin , PP242 , and Torin1 were purchased from Sigma , INK1341 and INK128 were obtained from Intellikine INC , as previously described in [23 , 55] . TRCN0000468751 ( pLX317 EIF4EBP1 ) and TRCN0000471416 ( pLX317 eIF4E ) lentiviral vectors from the MISSION TRC3 Human LentiORF Collection ( Sigma ) were used to generate cells stably over-expressing wild type human eIF4E or 4E-BP1 . 4E-BP1/2 and eIF4E shRNA vectors were previously described [23] . Wild type Herpes simplex virus 1 ( HSV1 ) KOS strain , GFP-expressing HSV1-dICP0 mutant and FLuc-expressing HSV1-dICP0 have been previously described [3] . The wild type HSV1 and HSV1-dICP0 were propagated and titered on Vero cells , following purification and concentration via sucrose cushion ultracentrifugation . The GFP-expressing HSV1-1716 was provided by Virttu Biologics , UK / Sorrento Therapeutics , San Diego , USA . Cells were washed with ice-cold PBS and lysed on ice using RIPA buffer ( 50mM Tris-HCl pH 8 . 0 , 150mM NaCl , 0 . 1% SDS , 0 . 5% Sodium deoxycholate , 1% Triton X100 , 10mM Sodium Fluoride , 1mM Sodium orthovanadate , 1mM DTT and protease inhibitor cocktail ( Roche ) ) . Cell debris were removed by centrifugation at 10 , 000g for 10min at 4°C . Protein concentration was determined using BioRad assay . Herpes Simplex Virus Type 1 polyclonal antibody was purchased from Dako ( Agilent Technologies ) , total 4E-BP1 , 4E-BP2 , phospho-S240–S244 rpS6 antibodies were from cell signaling , total rpS6 and alpha-tubulin antibodies were from Santa Cruz , eIF4E antibody was from BD Biosciences , β-actin antibody was from Sigma and V5-tag antibody was from Life Technologies . Anti-4E-BP3 number 1791 was previously described [20] . The plasmids encoding the ISRE promoter linked to firefly luciferase and Renilla luciferase ( transfection control ) were transfected using Lipofectamine and Plus reagent ( Invitrogen ) according to the manufacturer’s instruction . Cell extracts were prepared in 1X passive lysis buffer ( Promega ) 48 hours post-transfection and assayed for RLuc and FLuc activity in a Lumat LB95507 bio luminometer ( EG and G Bertold ) using a dual-luciferase reporter assay system ( Promega ) according to the manufacturer’s instructions . FLuc activity was normalized against RLuc activity . For the HEKBLUE assay , cells were transfected with 1μg/ml poly ( I:C ) ( InvivoGen ) using lipofectamine 2000 ( Life Technologies ) as per manufacturer’s instructions . Type-I IFN-containing supernatant was collected 24 hours post-transfection . HEKBLUE cells were seeded into a 96well plate ( 50 , 000 cells per well ) , mixed with 20μl of supernatant for a total volume of 200μl per well and incubated overnight to allow for the expression of Secreted Embryonic Alkaline Phosphatase ( SEAP ) . To quantify the levels of IFN-induced SEAP , 160 μl of Quanti-Blue ( InvivoGen ) were mixed with 40μl of supernatant and incubated at 37°C for 30min . The absorbance of the resulting reaction was measured at 650nm . NMuMG , NT2196 and 4T1 cells were infected at the indicated multiplicity of infection for 96 hours after which cells were fixed using 20% TCA solution for 5min and stained with crystal violet for 30 min followed by gentle wash with distilled water . Crystal violet stock solution was prepared by dissolving 1g of crystal violet in 100ml of 20% ethanol . Working solution prepared by mixing 20ml of the stock solution with 40ml of 100% ethanol and 140ml of distilled water . For quantification purposes , remaining viable cells from a separate experiment were manually counted using a hemacytometer and Trypan Blue exclusion . Live cell monitoring of virus spread was measured using the IncuCyte ZOOM system ( Essen BioScience , MI , US ) . NMuMG and NT2196 cells were seeded at 80–90% confluency and were pretreated with DMSO or PP242 prior to infection at the indicated MOI . Multiple images per well were taken every 2 hours until the predermined end point . Images were then analyzed using the IncuCyte ZOOM software ( Essen BioScience , MI , US ) to calculate for the GFP cluster integrated intensity ( Green calibrated unit x μm2 ) as a measurement of virus infection . In order to assess the in vivo efficacy of PP242/HSV1 combination , 3x105 4T1 cells were injected subcutaneously in the right flank of syngeneic mice . PP242 or vehicle was administered by gavage at 60mg/kg at 10 , 12 , 14 and 17 days post-implantation . FLuc expressing HSV1-dICP0 was intra-tumourally injected at 10 million pfu on days 11 , 13 and 18 . HSV1-dICP0 spread was assessed by measuring luciferase activity using IVIS at day 14 . Tumour size was monitored using electronic caliper . End point at which animals were sacrificed was set to tumour size 15mm X 15mm . Adherent cells were lysed using Trizol reagent ( Life Technologies ) immediately in the plate . RNA was isolated as per manufacturer’s instructions . RNA quality control and qPCR methodology was adopted from “the Minimum Information for Publication of Quantitative Real-Time PCR Experiments” ( MIQE ) guidelines [56 , 57] . RNA purity was verified by nanodrop and A260/A280 ratio for all samples used was between 1 . 8 and 2 . RNA integrity was verified by denaturing TBE/agarose gels . An equal amount of total RNA per condition for IFNβ level assessment following HSV1 stimulation were reverse transcribed into cDNA using SuperScript III Reverse Transcriptase ( Invitrogen ) following manufacturer’s instructions . Random Hexamer , RiboLock ribonuclease inhibitor and dNTPs were purchased from Thermo Scientific . Transcripts abundance was determined by qPCR ( eppendorf realplex 2 ) using iQ SYBR Green Supermix from Bio-Rad following manufacturer’s instructions . qPCR and sqPCR primers were designed using the NCBI primer blast software ( https://www . ncbi . nlm . nih . gov/tools/primer-blast/ ) following MIQE guidelines for optimal primer length , GC content and Tm . List of primers: HSV1 gC Fwd: 5’-GCCAGATCGACACGCAGACG-3’ HSV1 gC Rev: 5’-CGAAATGGGCAGGGTGGACC-3’ HSV1 ICP4 Fwd: 5’-CGACACGGATCCACGACCC-3’ HSV1 ICP4 Rev: 5’-ATCCCCCTCCCGCGCTTCGTCCG-3’ HSV1 TK qPCR Fwd: 5’- TCGGGGACACGTTATTTACCCTG-3’ HSV1 TK qPCR Rev: 5’- GCCCAGGCAAACACGTTATACAG-3’ hβ-Actin Fwd: 5’-GGACTCCTATGTGGGTGACGAGG-3’ hβ-Actin Rev: 5’ -GGGAGAGCATAGCCCTCGTAGAT-3’ mβ-Actin qPCR Fwd: 5’-GTACCACCATGTACCCAGGC-3’ mβ-Actin qPCR Rev: 5’-CGCAGCTCAGTAACAGTCCG-3’ hIFNβ Fwd: 5’-AATTGAATGGGAGGCTTGAA-3’ hIFNβ Rev: 5’-AGCCAGGAGGTTCTCAACAA-3’ mIFNβ qPCR Fwd: 5’-CTCCAGCACTGGGTGGAATG-3’ mIFNβ qPCR Rev: 5’-AGTGGAGAGCAGTTGAGGAC-3’ mISG15 qPCR Fwd: 5’- TGGTACAGAACTGCAGCGAG-3’ mISG15 qPCR Rev: 5’- AGCCAGAACTGGTCTTCGTG-3’ mIRF7 qPCR Fwd: 5’- GCACTTTCTTCCGAGAACTGGAGG-3’ mIRF7 qPCR Rev: 5’- GTCTTGCCCAAAACCCAGGTA-3’ Fluc qPCR Fwd: 5’- ATCCGGAAGCGACCAACGCC-3’ Fluc qPCR Rev: 5’- GTCGGGAAGACCTGCCACGC-3’ Cells cultured in complete growth media were incubated for 30 min at 37°C with complete culture medium containing 10 μCi/ml [35S]methionine . After 30 min incubation , cells were washed with PBS twice and lysed in Laemmli sample buffer . Incorporation was measured by TCA precipitation of proteins followed by scintillation counting . Radiolabeled proteins were also separated by SDS-PAGE followed by autoradiography . Sucrose density gradients ( 10 to 50% ) were prepared using Biocomp Gradient Master 108 as per manufacturer’s instructions and as previously published [58] . 10% and 50% sucrose solutions were prepared in a buffer containing 20 mM HEPES-KOH pH 7 . 6 , 100 mM KCl , 5 mM MgCl2 , 100 μg/ml cycloheximide , EDTA-free protease inhibitors mixture tablets ( Roche ) , and 200 units/ml ribonuclease inhibitor ( Ribolock from Thermo Fisher Scientific ) . A total of 7 . 5x106 cells for 4T1 and NMuMG and 15x106 for NT2196 were seeded per 15cm dish 24 hours prior to HSV1-dICP0 infection . Cells were pretreated for 30min with vehicle or PP242 followed by infection with GFP-expressing HSV1-dICP0 at 0 . 1 MOI for 24 hours . Prior to collection , cells were treated with 100μg/ml cycloheximide for 5 min at 37°C , washed twice with ice-cold PBS containing 100 μg/ml cycloheximide and collected by scraping . Cell pellet following centrifugation at 1000 rpm for 10min at 4°C were lysed in hypotonic lysis buffer ( 5 mM Tris pH 7 . 5 , 2 . 5 mM MgCl2 , 1 . 5 mM KCl , EDTA-free protease inhibitors cocktail tablets ( Roche ) , 100 μg/ml cycloheximide , 2 mM DTT , 200 units/ml Ribolock , 0 . 5% ( v/v ) Triton X-100 , and 0 . 5% ( w/v ) sodium deoxycholate . Cell debris were cleared by centrifugation at 21 , 000g for 5 min at 4°C . RNA concentration was determined using a Thermo Scientific Nanodrop 2000 spectrophotometer . A total of 500 μg total RNA per condition were loaded on each gradient , which were spun at 36 , 000 rpm ( 230501 X g ) for 2 hours at 4°C using a Beckman coulter SW40Ti rotor and Optima L80 XP ultracentrifuge . Gradients were fractionated using Teledyne ISCO fractionator with the pump set at 1 . 5ml per min . Optical density at 254nm was recorded at 10 measurements per second frequency . RNA from each fraction was isolated using TRIzol ( Invitrogen ) and treated with DNaseTurbo ( Ambion ) according to the manufacturer's instructions . Reverse transcription PCR ( RT-PCR ) and quantitative RT-PCR ( qRT-PCR ) reactions were carried out using SuperScript III First-Strand Synthesis System ( Invitrogen ) and iQ SYBR Green Supermix ( BIO-RAD ) according to the manufacturer's instructions . For qPCR experiments , each fraction was spiked with 5ng of polyA+ firefly luciferase mRNA ( Promega ) prior to extraction . Measurements were then normalized to luciferase abundance , and plotted as relative transcript amount over luciferase control . Luciferase reporter mRNAs were generated using MAXIscript T7 in Vitro Transcription kit ( Ambion ) according to the manufacturer’s protocol in the presence of the cap analog . PCR products encoding a T7 promoter followed by luciferase and a poly ( A ) sequence were used as templates for in vitro transcription . 4T1 cells were seeded on a 24-well-plate and cultured overnight . Cells were transfected with capped HSV-ICP0-5’UTR-Rluc-pA , HSV-TK-5’UTR-Rluc-pA , IRF7-5’UTR-Rluc-pA or Rluc-pA together with capped Fluc-pA as a transfection control using Lipofectamine ( Invitrogen ) . Cells were treated with 1μM PP242 for 24 hours and lysed 24 hours post transfection . Luciferase activities were determined using Dual-Luciferase Reporter Assay System ( Promega ) according to the manufacturer's instruction . Statistical tests were performed within Prism ( GraphPad ) or Excel . Error bars for data presented are standard deviation ( SD ) from the mean . P values were calculated as follow: For comparison between two independent groups , a two-tailed unpaired t test was used . For comparison between more than two independent groups , a one-way analysis of variance was performed with Dunnett's or Bonferroni's post hoc tests , and adjusted p values were reported . p values are reported as follows: * , p < 0 . 05; ** , p < 0 . 01; *** , p < 0 . 001; ns denotes non-significant p values . The animal experiment was performed in accordance with the guidelines for animal care at the University of Ottawa Animal Care and Veterinary Services under the approved protocol OGHRI-58 . The University of Ottawa is a registered research facility under the Animals for Research Act and is certified by the Canadian Council on Animal Care . | Dysregulated mRNA translation occurs frequently in tumours due to elevated eIF4E expression or a hyperactive mTOR complex 1 ( mTORC1 ) signaling pathway that results in the inactivation of the eIF4E binding proteins ( 4E-BPs ) . Targeting the mTORC1/4E-BPs/eIF4E axis is a promising strategy in cancer therapies and for preventing resistance to treatment . Enhanced mTORC1 activity also drives innate immune responses by modulating protein expression of antiviral genes . It was previously shown that the mTORC1 inhibitor rapamycin limits antiviral responses and promotes replication of oncolytic viruses within tumour tissues . Active-site dual mTORC1 and mTORC2 inhibitors ( asTORi ) have been developed for superior mTOR inhibition and anti-cancer potency but have not been studied in the context of oncolytic viral infection . We show here that prolonged treatment with asTORi strongly augments infection of HSV1-dICP0 in cancer cells , but not in normal cells , an effect modulated via eIF4E/4E-BP expression . Thus , cancer cells with dysregulated translation could be amenable to the pharmacoviral combination of HSV1 and asTORi treatment . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"luciferase",
"enzymes",
"cancer",
"treatment",
"messenger",
"rna",
"biological",
"cultures",
"enzymology",
"oncolytic",
"viruses",
"microbiology",
"transformed",
"cell",
"lines",
"oncology",
"protein",
"synthesis",
"chemical",
"... | 2018 | Active-site mTOR inhibitors augment HSV1-dICP0 infection in cancer cells via dysregulated eIF4E/4E-BP axis |
Visceral leishmaniasis ( VL ) is distinguished by a complex interplay of immune response and parasite multiplication inside host cells . However , the direct association between different immunological correlates and parasite numbers remains largely unknown . We examined the plasma levels of different disease promoting/protective as well as Th17 cytokines and found IL-10 , TGFβ and IL-17 to be significantly correlated with parasite load in VL patients ( r = 0 . 52 , 0 . 53 and 0 . 51 for IL-10 , TGFβ and IL-17 , respectively ) . We then extended our investigation to a more antigen-specific response and found leishmanial antigen stimulated levels of both IL-10 and TGFβ to be significantly associated with parasite load ( r = 0 . 71 and 0 . 72 for IL-10 and TGFβ respectively ) . In addition to cytokines we also looked for different cellular subtypes that could contribute to cytokine secretion and parasite persistence . Our observations manifested an association between different Treg cell markers and disease progression as absolute numbers of CD4+CD25+ ( r = 0 . 55 ) , CD4+CD25hi ( r = 0 . 61 ) as well as percentages of CD4+CD25+FoxP3+ T cells ( r = 0 . 68 ) all correlated with parasite load . Encouraged by these results , we investigated a link between these immunological components and interestingly found both CD4+CD25+ and CD4+CD25+FoxP3+ Treg cells to secrete significantly ( p<0 . 05 ) higher amounts of not only IL-10 but also TGFβ in comparison to corresponding CD25- T cells . Our findings shed some light on source ( s ) of TGFβ and suggest an association between these disease promoting cytokines and Treg cells with parasite load during active disease . Moreover , the direct evidence of CD4+CD25+FoxP3+ Treg cells as a source of IL-10 and TGFβ during active VL could open new avenues for immunotherapy towards cure of this potentially fatal disease .
Visceral leishmaniasis ( VL ) or kala-azar , a deadly infection affecting the organs of lymphoreticular system , is caused by the protozoan parasites of the Leishmania donovani complex . Human infections can be asymptomatic or oligosymptomatic with manifestations including persistent low grade fever , weight loss , hepatosplenomegaly , lymphadenopathy and cachexia . Other disease specific characteristics include pancytopenia , hypergammaglobulinemia , hypoalbuminemia and severe parasite infestations in visceral organs such as the liver and spleen , and in the bone marrow [1] . An estimate of 0 . 2–0 . 4 million global VL cases are reported each year with more than 90% of them occurring in India , Bangladesh , Sudan , South Sudan , Brazil and Ethiopia [2] . In India , the disease is mostly prevalent in the state of Bihar , parts of Eastern Uttar Pradesh and West Bengal [3] with an estimated annual incidence ( cases from 2004–2008 ) ranging from 146 , 700 to 282 , 000 and a mortality rate of 2 . 4% ( 853/34 , 918 ) [2] . The mechanisms underlying the multiplication and spread of parasites in VL are not well understood . Infection with Leishmania parasites does not always lead to illness and the majority of people residing in endemic regions never develop VL . How these individuals develop resistance against infection is an area of intense interest but remains largely unknown . In most tegumentary forms of leishmaniasis parasite growth is restricted and antileishmanial cell mediated immunity ( CMI ) is developed [4] . However , the absence of antigen specific CMI has been regarded as a signature of VL [5] . Development of a Th1 type immunity appears to be needed for limiting parasite growth and resistance against infection [6] . PBMCs from clinically asymptomatic subjects have been observed to respond to triggering with leishmanial antigen and secrete IL-2 , IFNγ and IL-12 [7] . However , a key immunological attribute of VL remained is the apparent incapability of its PBMCs to manifest an antigen specific immune response as suggested by their failed proliferative [8 , 9] or cytokine producing abilities [10 , 11] . Although active VL was initially characterized as presenting a predominant Th2 type response with elevated levels of serum IL-4 and IL-13 [12 , 13] , other studies reported increased levels of IFNγ and IL-12 along with IL-5 and IL-6 in plasma [14–17] as well as enhanced IFNγ mRNA expression [11 , 18 , 19] during disease , suggesting that its development is not due to a predominant Th2 response . Rather clinical studies implicate a strong role for IL-10 associated with the immunosuppressive nature of kala-azar [10] . VL patients have been noted for an augmented production/expression of IL-10 at both cellular [15–17 , 20] and molecular levels [18 , 19 , 21] . More recently IL-10 could be measured in leishmanial antigen-stimulated PBMC from culture supernatants [17 , 20 , 22] and whole blood cultures of VL patients [23 , 24] . Another cytokine important from the perspective of human VL pathology is TGFβ [25 , 26] . Its elevation has been reported in plasma as well as antigen-stimulated PBMCs of VL patients [20 , 22] . IL-10 regulated parasite progression and development of infection have been linked with accumulation of CD4+CD25+ regulatory T cells ( Treg ) in both experimental [27] and human cutaneous leishmaniasis ( CL ) [28] . Natural Treg cells were found to secrete large amounts of IL-10 and/or TGFβ in CL patients [28 , 29] and asymptomatic L . major infected individuals [30] , as well as in PBMCs of healthy subjects activated by live L . guyanensis parasites [31] . However , although CD4+CD25+ T cells have been related with kala-azar [20] and patients suffering from PKDL [20] , FoxP3-CD4+ T cells were reported to be the key source of enhanced IL-10 mRNA expression in spleen of VL patients [11] . Still other reports indicate that CD4+CD25+FoxP3+ cells could play a part in human VL as a vital source of IL-10 [32] . Th17 cells , an independently regulated and highly proinflammatory subset of CD4+ T cells , are also known to produce IL-17 and IL-22 [33] , effectors of innate immunity . Activation of Th17 cells is characterized by neutrophil recruitment and depending on their differentiation program they can either cause damaging effects or play a role in protection [34] . While Th17 cells have been associated with pathological roles and tissue damage in human mucocutaneous leishmaniasis [35] , a protective role has been attributed to them in the case of human VL [36] . CD8+ T cells , by virtue of their cytotoxic activity , lyse pathogen-infected cells and produce IFNγ thereby conferring protection to the host . However , they can also act as suppressor T cells , down-modulating the host immune response and thereby increasing the chances of survival of invading parasites inside their hosts [37] . Thus CD8 T cells have the potential to contribute to both resolution and pathogenesis of VL . Interestingly a low CD4 to CD8 ratio has been reported in PBMC from VL patients [38] . However , the mechanisms of evasion of CD8+ T cell-mediated host protective immunity by L . donovani in human VL are largely unknown . To understand how the immune response generated during VL in the host is impacted by the multiplication of the pathogen resulting in clinical pathogenesis and disease , we have studied the direct relation between different immunological correlates and the number of parasites in VL patients . Our study demonstrates , in addition to IL-10 , for the first time a positive correlation of TGFβ in plasma , as well as IL-10 and TGFβ in antigen-stimulated PBMCs from culture supernatants , with parasite loads of VL patients , supporting their roles in disease pathology . Additionally , our results also point to a positive correlation between absolute numbers of CD4+CD25+ Treg cells with parasite load , highlighting their involvement in L . donovani infection . On the basis of the initial results we further investigated whether these cellular subsets could be the sources of disease promoting cytokines ( IL-10 and TGFβ ) , and our observations revealed , along with IL-10 , for the first time TGFβ to be secreted from CD4+CD25+FoxP3+ Treg cells during active VL .
Twenty VL patients ( males and females , and HIV negative ) admitted to School of Tropical Medicine ( STM ) , Kolkata , West Bengal , and Rajendra Memorial Research Institute of Medical Sciences ( RMRIMS ) , Patna , Bihar were included in this study ( Table 1 ) . Patients were initially diagnosed by the usual clinical presentations such as prolonged fever , hepatosplenomegaly and pancytopenia . Confirmation criteria for the disease , however , was restricted to the rK39 strip test , as the risky and painful nature of aspiration ( both splenic and bone marrow ) techniques limit them to be employed for all patients studied . 4–5 ml of venous blood samples were collected from all patients . The patients were then subjected to a standard treatment protocol as selected by the study centers ( STM and RMRIMS ) and discharged after completion of therapy as clinically cured . The non-endemic healthy individuals ( NEC ) , both male and female , included in the study were from different areas of Kolkata with an age range of 23–35 years . Endemic healthy individuals ( EC ) [healthy family members living in the same endemic regions of Bihar] were chosen from among the persons attending patients , and included both males and females with an age range of 35–50 years . The study was approved by and carried out under the guidelines of the Ethical Committee of the Indian Institute of Chemical Biology , Kolkata . All participants including ECs , NECs and patients or responsible adults provided written informed consent for the collection of samples and subsequent analysis . Blood samples were collected in heparinized tubes . DNA extraction was performed using QIAamp DNA Blood mini kits ( Qiagen , Germany ) from whole blood according to the manufacturer’s instructions . DNA was isolated from 400 μl of sample and eluted in 25 μl elution buffer . A stock solution of L . donovani DNA was also obtained by extraction ( QIAamp DNA mini kit , Qiagen , Germany ) from 2 . 5×106 promastigotes . After eluting in 25 μl elution buffer , assuming the extraction was nearly 100% efficient , the DNA concentration corresponded to105 parasites/μl . All the DNAs were stored at -20°C until use . SYBR Green-based real-time PCR was applied for quantification of the Leishmania parasites in patient blood . For accurate sensitivity , L . donovani-specific kinetoplast DNA was chosen as the target region . The PCR was performed in a final volume of 20 μl containing 10 μl of SYBR Green master mixture ( 2× ) [Applied Biosystems ( ABI ) , Carlsbad , CA , USA] , 3 μl of MiliQ water , 5 μl of DNA template and 1 μl ( 50 pmol/μl ) of forward and reverse primers , 5′-CTTTTCTGGTCCTCCGGGTAGG-3′ and 5′-CCACCCGGCCCTATTTTACACCAA-3′ ( Integrated DNA Technologies , Coralville , IA , USA ) respectively . To construct the standard curve ten-fold serially diluted L . donovani DNA stocks corresponding to 105 , 104 , 103 , 102 , 10 , 1 parasite/μl were included in the assay . Amplification was conducted using a StepOnePlus Real-Time PCR system ( ABI ) . The thermal cycling conditions included an initial incubation at 50°C for 2 min , followed by a 10 min denaturation at 95°C and 40 cycles at 95°C for 15 s and 60°C for 1 min each . Each sample was tested in duplicate . The calculation of the melting temperature of each amplicon was done directly by the software provided ( StepOne Software ) . Negative controls with no template were included in each plate to deal with contamination issues . For all the real-time PCR reactions we used specific primers for the constitutively expressed beta actin gene ( forward: 5’-GGCCAACCGCGAGAAGAT-3’; reverse: 5’-CGTCACCGGAGTCCATCAC-3’ ) as a quality control to verify the integrity of the DNA samples . Threshold cycle values ( Ct ) were calculated for patient samples by determining the point at which the fluorescence exceeded the threshold limit plotted against known concentrations of parasite DNA , and the parasite load of patient samples was then determined using the standard curve [3 , 39] . L . donovani promastigote membrane antigens ( LAg ) were prepared as previously described [40] and stored at –70°C . The amount of protein thus obtained was determined by the method of Lowry et al . [41] . The endotoxin level in LAg was measured using the chromogenic Limulus Amebocyte Lysate ( LAL ) assay kit ( QCL-1000; Lonza , Walkersville , MD , USA ) according to the manufacturer's recommendations and was found to be virtually endotoxin free ( <10 endotoxin units per mg of protein ) . Plasma and PBMCs were isolated from the blood samples as described previously [17] . Plasma was then stored at -20°C until use and PBMCs were cultured with LAg ( 12 . 5 μg/ml ) . After 96 hrs , the supernatants of the cultures were collected and stored at -70°C until use . IFNγ , IL-12 , IL-10 ( BD OptEIA ELISA kit; BD Biosciences , San Jose , CA , USA ) , TGFβ , IL-17 ( eBioscience , Inc . , San Diego , CA , USA ) and IL-22 ( R&D Systems , ELISA DuoSet , McKinley Place , MN , USA ) were measured in plasma samples according to manufacturers’ instructions . IFNγ , IL-12 , IL-10 and TGFβ were also measured in culture supernatants following the same protocol . The color reaction was performed using avidin-HRP and tetramethylbenzidine , and read at OD450 [17 , 20] . For intracellular detection of cytokines , samples having a minimum cell count of 4 × 106 were chosen . PBMCs from such samples were stimulated with 5 μl of 10 ng/μl phorbol 12-myristate 13-acetate ( PMA ) and 1 μl of 1 mg/ml ionomycin ( Sigma , Saint Louis , MO , USA ) at 37°C and under 5% CO2 for 2 hrs . Brefeldin A ( 10 mg/ml; Sigma ) was added , and the samples were incubated for an additional 1 hr . Initially , stimulating abilities were compared between PMA/ionomycin , LAg and a combination of both PMA/ionomycin and LAg with all the alternatives triggering cytokines to similar levels ( S1 Fig ) . Our observations were consistent with those of Owens et al . [42] , who also failed to differentiate between antigenic and mitogenic stimulation when used with spleen cells from an experimental VL model . PMA/ionomycin was chosen as a standard triggering agent for its shortest stimulation time . The exact interval for the addition of Brefeldin A was standardized through a time kinetics study , in which negligible amounts of cytokines were found to be present in the culture supernatant 2 hr after the addition of PMA/ionomycin . The cells were stained with CD4-phycoerythrin-cyanine 7 conjugate ( PE-Cy7 , BD Biosciences ) and CD25-allophycocyanin-cyanine 7 conjugate ( APC-Cy7 , BD Biosciences ) to identify the CD4+CD25+ T cell population . In our conditions APC-Cy7 conjugated CD25 was able to produce a strong signal and a distinct demarcation between CD25l°w and CD25hi populations . After washing with FACS buffer [0 . 2 M PBS ( 17 . 8 g Na2HPO4 , 6 . 9 g NaH2PO4 , 9% NaCl , and distilled water to obtain 1 L , pH 7 . 2 ) and 1% FBS ( Gibco , Life Technologies , Grand Island , NY , USA ) ] , the samples were fixed and permeabilized by adding 100 μl of Cytofix/Cytoperm solution ( BD Biosciences ) in the dark for 20 min . After centrifugation , the supernatant was discarded , and the cells were intracellularly stained to detect FoxP3-fluorescein isothiocyanate ( FITC , BD Biosciences ) , IL-10- allophycocyanin ( APC , BD Biosciences ) , TGFβ-phycoerythrin ( PE , BD Biosciences ) and IFNγ-PE-Cy7 ( BD Biosciences ) for 30 mins in the dark . The samples were washed with FACS buffer supplemented with 0 . 1% Saponin ( Sigma ) . Cell acquisition and analysis were performed using FACS Diva Software ( BD ) on a FACSCanto flow cytometer ( BD ) . Prior to the addition of Cytofix/Cytoperm a small aliquot of the sample was analyzed for viability through PI staining and cells were found to be almost 95% viable . To further exclude any dead cells or debris we performed doublet discrimination through forward scatter-area/ forward scatter-width gating along with normal forward scatter-area/ side scatter-area gating . T cells were identified based on CD4-PE-Cy7 staining and different forward scatter and side scatter parameters . In total , 50 , 000 events were acquired . Regulatory T cells were gated as FoxP3-positive cells among CD4+CD25+ population , and the percentages of cells producing IL-10 or TGFβ were determined using thresholds based on unstimulated samples ( S2 Fig ) . Isotype-matched controls ( BD Biosciences ) for FITC , PE , APC and PE-Cy7 were used in all staining experiments [42 , 43] . Negative controls ( cells plus all reagents minus PMA/ionomycin ) were also used in the intracellular experiments as a background control to rule out the effect of any nonspecific stimulation [44] . For absolute counting , 50 μl of heparin-treated whole blood samples were taken in TruCount tubes ( BD Biosciences ) and incubated for 30 mins in the dark at room temperature with 5 μl each of CD45-PE-Cy7 , CD3-peridinin chlorophyll ( PerCP ) , CD4-APC , CD8-FITC and CD25-PE ( BD Biosciences ) . Erythrocytes were lysed with 450 μl lysing solution ( BD Biosciences ) according to the manufacturer’s protocol . Cell acquisition and analysis were performed using FACS Diva Software ( BD ) on a FACSCanto flow cytometer ( BD ) . For TruCount tubes , the analysis software calculated the percentage of lymphocytes that were positive for the subset of interest ( CD3 , CD4 , CD8 , CD4+CD25+ and CD4+CD25high ) by identifying lymphocytes as being positive for CD45 but with low side scatter ( S3 Fig ) [45] . Absolute numbers were calculated as ( number of cells of interest counted/number of beads counted ) × ( total number of beads in tube [from manufacturer]/50 μl [volume of blood tested] ) [46] . Correlation was evaluated using Spearman/Pearson correlation test . All data are presented as mean ± SD and a difference in mean values was considered significant when the P value was <0 . 05 . Statistical analysis was carried out using GraphPad Prism 5 ( GraphPad Software , Inc . , San Diego , CA , USA ) .
The real-time PCR assay was initially tested for its specificity of detecting L . donovani parasites . A sequential dilution of L . donovani culture with corresponding final concentrations ranging from 105 to 1 parasites was carried out and DNA was extracted from each dilution . Ct values obtained by subjecting those DNA samples to real-time PCR were used to construct a standard curve for determining parasite load in patient blood . The standard curve , containing 6 logarithmic parasitic DNA dilutions , was linear with a correlation coefficient ( r2 ) of 0 . 965 ( Fig 1 ) . The Ct values , as obtained using real-time PCR assay , were 1 . 91 , 2 . 46 , 2 . 84 , 2 . 93 , 3 . 06 and 3 . 12 for the six different concentrations . Melting temperature ( Tm ) of the samples , as calculated through melt curve analysis , was found to be 84 . 32 ± 0 . 09°C ( mean Tm values ± standard error ) and was comparable between all the samples ( S4 Fig ) . Percentage PCR efficiency was found to be 105 . 86% , using the following equation: %Efficiency = [10 ( -1/sl°pe ) -1] × 100% and slope of the standard curve ( -3 . 189 ) . Negative control ( water instead of template DNA ) was included in each PCR assay to control for any contamination issues . Determination of parasite loads in blood samples was performed in all the VL patients ( n = 20 ) . The average parasite load in circulation was 2549 parasites/ml , ranging from 487 . 5 parasites/ml to 5000 parasites/ml ( Table 2 ) . No parasites were detected in blood samples of the healthy controls ( both NECs and ECs ) . To evaluate the association of parasite load with disease pathology different clinical as well as hematological parameters were considered ( Table 2 ) . Significant positive correlations were found between parasite load and duration of illness [length of time the patients suffered from fever before being admitted to hospital] ( r = 0 . 57 , p = 0 . 008 ) ( Fig 2A ) as well as spleen size ( r = 0 . 59 , p = 0 . 006 ) ( Fig 2B ) , whereas the correlation was significantly negative with white blood cell counts ( r = -0 . 53 , p = 0 . 017 ) ( Fig 2C ) and albumin levels ( r = -0 . 57 , p = 0 . 009 ) ( Fig 2D ) . Expression levels of IFNγ , IL-12 , IL-10 , TGFβ , IL-17 and IL-22 were analyzed in plasma of VL patients ( n = 20 ) by ELISA ( S1 Table ) . Patients harboring higher number of parasites seemed to produce higher levels of plasma cytokines IL-10 , TGFβ and IL-17 . Indeed , their circulating levels were found to correlate strongly and significantly with parasite load ( r = 0 . 52; p = 0 . 02 for IL-10 , r = 0 . 53; p = 0 . 018 for TGFβ and r = 0 . 51; p = 0 . 02 for IL-17 ) . The correlation , however , was not so strong for the other three cytokines tested ( r = 0 . 23 for IFNγ , 0 . 25 for IL-12 and 0 . 25 for IL-22 ) . However , since VL is immunosuppressive and often involves co-infection , circulating cytokine levels do not always accurately represent the actual disease scenario . Thus investigation of the correlation between parasite load and antigen specific levels of cytokines in PBMCs of VL patients that were already found to be elevated in plasma was carried out . PBMCs were collected from VL patients ( n = 10 ) having a minimum cell count of 4 × 106 and cultured in vitro in the presence of LAg for 4 days as reported earlier [20 , 47 , 48] . LAg-stimulated culture supernatants contained significantly higher levels of IL-10 ( mean level of 57 . 8 pg/ml ) and TGFβ ( mean level of 53 pg/ml ) in comparison to unstimulated cultures ( mean levels of 28 . 5 pg/ml for IL-10 and 17 pg/ml for TGFβ ) ( Fig 3 ) . Moreover , when compared with ECs ( n = 5 ) and NECs ( n = 10 ) , VL patients were found to produce significantly higher levels of both LAg stimulated IL-10 ( p = 0 . 013 for EC and 0 . 0002 for NEC ) and TGFβ ( p = 0 . 0075 for EC and 0 . 0012 for NEC ) , suggesting the specificity of LAg towards infected individuals ( Fig 3 ) . Levels of IL-10 and TGFβ produced by unstimulated cells correlated positively ( r = 0 . 62 for IL-10 and r = 0 . 53 for TGFβ ) with parasite load ( Fig 4A and 4C ) , although the correlation was not statistically significant . However , the correlation was significant when IL-10 ( r = 0 . 71 , p = 0 . 02 ) and TGFβ ( r = 0 . 72 , p = 0 . 02 ) produced by LAg stimulated PBMCs were plotted against parasite load ( Fig 4B and 4D ) . Correlation between parasite load and unstimulated levels of IFNγ and IL-12 could not be performed due to the low titres of these cytokines and while the association between LAg stimulated levels of both these cytokines with parasite load was not so strong ( r = -0 . 4 for IFNγ and -0 . 06 for IL-12 ) , a negative slope in either cases signified their inhibitory role in parasite multiplication ( S5A and S5B Fig ) . Correlation of parasite load with antigen specific IL-17 and IL-22 , however , also could not be performed due to the low titre of these cytokines in both unstimulated and LAg stimulated PBMCs of VL patients . The increase in absolute number of circulating CD4+CD25+ Treg cells observed in VL patients suggested its probable role during L . donovani infection . To further elucidate if there is a direct association between these two variables , the absolute numbers of CD4+CD25+ Treg cells were compared with parasite load ( Fig 5D ) . The data clearly show that the number of CD4+CD25+ Treg cells is positively correlated with parasite load ( r = 0 . 55 , p = 0 . 01 ) in VL patients . Additionally , since high expression of CD25 has been reported as a more specific marker for Tregs [49] , we also assessed CD4+CD25hi Treg cells and found their numbers to be significantly correlated ( r = 0 . 61 , p = 0 . 004 ) with increasing levels of parasite load ( Fig 5E ) . The best characterized Treg cell is , however , defined by the expression of its transcriptional regulator , FoxP3 [11] . To evaluate the role of this subpopulation in active VL we investigated the association between frequency of CD4+CD25+FoxP3+ Treg cells with parasite load in patients ( n = 10 ) having at least 4 × 106 cells . Our data exhibited a significant positive correlation ( r = 0 . 68 , p = 0 . 03 ) between parasite load and percentage of CD4+CD25+FoxP3+ T cells ( Fig 5F ) . These observations confirm the involvement of CD4+CD25+FoxP3+ Treg cells in VL pathogenesis . Although no significant correlation with parasite load was observed for overall CD3 ( Fig 5A ) , CD4 ( Fig 5B ) and CD8 ( Fig 5C ) T cell subsets , the regression line showed a negative slope for CD3 ( Fig 5A ) and CD4 ( Fig 5B ) implying a decrease in their numbers with increasing parasite load whereas a positive slope for CD8 ( Fig 5C ) indicates an increase in CD8 count with increasing number of parasites . Significant positive correlation between parasite load and different cytokines as well as cellular subsets provided us the opportunity to investigate the link between these immune components . Since the cellular origins of IL-10 and TGFβ during active VL remains obscure , we examined CD4+CD25+ and CD4+CD25+FoxP3+ Treg cells for their potential role as sources of these disease promoting cytokines . To understand the scenario , we first analyzed the expression of PMA/ionomycin–activated , IL-10–producing CD25 in CD4 and CD4+FoxP3+ cells from patients with active VL ( n = 10 ) as well as from ECs ( n = 5 ) and NECs ( n = 5 ) . Although the percentage of IL-10 produced by CD25-positive and negative populations was comparable between healthy controls ( both ECs and NECs ) ( Fig 6A and 6B ) , both CD4+CD25+ and CD4+CD25+FoxP3+ cells of infected individuals were found to secrete significantly ( p = 0 . 013 for CD4+CD25+ and p = 0 . 049 for CD4+CD25+FoxP3+ ) higher levels of IL-10 in comparison to the corresponding CD25-negative populations ( Fig 6A and 6B ) . Fig 6C ( i ) –6C ( iv ) are representative of one healthy control and Fig 6D ( i ) –6D ( iv ) are representative of one active VL case , respectively . We then focused our study on probable source ( s ) of TGFβ in VL patients and there also found CD4+CD25+ and CD4+CD25+FoxP3+ cells to secrete significantly ( p<0 . 05 ) higher levels of this cytokine in comparison to CD25-negative cells ( Fig 7A and 7B ) . The percentage of TGFβ secreting CD4+CD25+FoxP3+ cells ( 32 . 06% ) ( Fig 7B ) , however , was significantly higher than that of CD4+CD25+ cells ( 6 . 04% ) ( Fig 7A ) . Moreover , the percentage of TGFβ producing CD4+CD25+FoxP3+ cells ( 32 . 06% ) was also found to be significantly ( p<0 . 05 ) higher than the similar population in ECs ( 12 . 9% ) as well as NECs ( 13 . 08% ) ( Fig 7B ) reaffirming our previous observations with circulating levels of TGFβ ( Fig 3 ) . Fig 7C ( i ) –7C ( iv ) are representative of one healthy control and Fig 7D ( i ) –7D ( iv ) are representative of one active VL case , respectively . Additionally , our observation with IFNγ yielded similar results , as both CD4+CD25+ and CD4+CD25+FoxP3+ cells were found to secrete significantly higher ( p<0 . 01 ) levels of this cytokine in comparison to corresponding CD25- cells ( S6A and S6B Fig ) . To the best of our knowledge this is the first report regarding the involvement of CD4+CD25+FoxP3+ Tregs as a cellular source of TGFβ in VL patients . Taken together , our results showed that among PBMCs from L . donovani infected patients with VL , CD4+CD25+ , or more specifically CD4+CD25+FoxP3+ Treg cells , are the most important sources of the disease promoting cytokines IL-10 and TGFβ .
The present study investigated the correlation between infective parasites and different immune components associated with VL and assessed the impact of their interplay in predicting the disease outcome . Interestingly , our results revealed a strong correlation of increasing parasite load with different clinical parameters implying disease manifestations to be linked with the number of infective parasites as well as symptoms like splenomegaly , leukocytopenia and hypoalbuminemia . Our data also demonstrated a positive association between parasite load and elevated levels of IL-10 and TGFβ along with absolute numbers of CD4+CD25+ T cells and percentages of CD4+CD25+FoxP3+ positive T cells suggesting the involvement of these cytokines and Treg cells in course of disease . Taking a cue from this parasitological evidence , we went further and were able to identify the significance of CD4+CD25+FoxP3+ Treg cells in secretion of these cytokines , in comparison to corresponding CD25- T cells . The course of VL depends on the interplay between the host and the infecting pathogen in which the survival of parasites is primarily regulated by the type of cytokine being secreted upon and their interaction with host immune cells . Therefore , we analyzed some important disease promoting/protective as well as Th17 cytokines in VL patients to understand the influence of the immune response on parasite load . IL-10 has been well known for its role in VL pathogenesis and a significant correlation was already reported between parasite load and plasma IL-10 levels in VL patients [3] . Our study with plasma IL-10 also supported this notion , and an extreme case having the highest level of IL-10 was presented with very high parasite load ( 4500 parasites/ml of blood with 350 pg/ml plasma IL-10 ) . However , since VL patients are often co-infected with other pathogens such as mycobacteria [50] , which could also result in elevated plasma lipopolysaccharide concentration and subsequent impairment in immune effector function [51] , or a systemic inflammatory response syndrome called sepsis [52] , total plasma cytokine profiles might not be a conclusive representation of infection . Therefore , we investigated the recall response in LAg stimulated PBMCs of some of these patients and found even a stronger correlation between parasite load and IL-10 levels in these patients . To the best of our knowledge , this is the first report of the correlation between number of parasites and antigen-specific production of IL-10 in the human VL scenario . Another cytokine which could contribute to the pathology of human VL progression is TGFβ [20] . It has been shown to impart down-regulatory effects on macrophages and its blockade could restrict parasite progression in these cells [53] . Since there has been no previous report on its association with parasite load in human VL , we examined the correlation between parasite load and circulating/antigen specific TGFβ levels in VL patients . The correlation was significant in plasma and appeared stronger in LAg stimulated culture supernatants of VL patients , confirming the role of TGFβ in parasite multiplication and evolution of disease . A correlation , although not statistically significant , was also recorded between parasite load and both these cytokines ( IL-10 and TGFβ ) produced by unstimulated cells , indicating that in vitro production of these cytokines partly resulted from stimuli received in vivo . IL-10 has also been reported to synergize with TGFβ to inhibit macrophage microbicidial activity thereby enhancing parasite growth and survival [54] . Our data supports this hypothesis as production of both IL-10 and TGFβ were clearly correlated with active disease . The role of Th17 cells and their cytokines in leishmaniasis remain obscure as both protective and pathogenic responses have been influenced by them [55] . In human VL , there are very few reports , and while in one case IL-17 levels were found to be higher in VL resistant individuals [36] , in other instances it was found to be elevated in VL susceptible patients [56 , 57] . We evaluated the association of parasite load with plasma IL-17 and found a positive correlation between them . However , as already explained , total circulating levels of any cytokine was not enough to depict the actual disease scenario . In fact , IL-17 level in plasma was found to be triggered in VL patients co-infected with malaria [57] and since disease resistance was associated with elevation in L . donovani antigen stimulated IL-17 levels [36] , a more precise antigen-specific response would be desirable to draw any conclusion in this respect . Although IL-10 and TGFβ have been reported to contribute to VL pathogenesis [58] , information regarding the cellular subtypes associated with the elevation of these cytokines during infection remains limited . In human CL , it has been observed that T cells expressing IL-10 and/or TGFβ are mostly natural Treg cells and thus can be considered as the source of these disease promoting cytokines [28 , 29] . However , in human VL the evidence is not conclusive . While some studies at flow cytometric as well as transcriptional levels failed to show any significant increase in CD4+CD25+FoxP3+ Treg cells during active VL [11 , 59] , other studies recorded increased percentages of CD4+CD25+ Treg cells in addition to upregulation in the levels of IL-10 and TGFβ in both VL and PKDL patients [20] . Furthermore mRNA levels of different Treg cell markers like CD25 and FoxP3 as well as IL-10 were also shown to increase in lesional tissues of PKDL patients , and this elevation exhibited a significant positive correlation with parasite load [60] . This evidence suggests CD4+CD25+FoxP3+ Treg cells to be one of the important sources of IL-10 in VL patients [32] and which may play significant role in favoring parasite survival . In the case of TGFβ , the cellular sources were reported to be mostly CD4+CD25+ T cells in L . guyanensis [31] and L . major [30] infected individuals . However , there is no direct evidence for the origin of TGFβ during active VL . Our experiments with absolute numbers of different T cell subsets revealed a significant positive correlation of parasite load with not only CD4+CD25+ T cells but also CD4+CD25high T cells . Higher percentages of CD4+CD25+FoxP3+ T cells were also found to be significantly correlated with increased parasite load . Since the surface marker and transcriptional regulators for Treg cells were well expressed in VL patients and correlated significantly with parasite load , it could be logical to infer that the heightened Treg activity may be the key determinant for parasite survival and multiplication inside the host . In fact , the abundance of Treg cells , along with IL-10 and TGFβ , in VL patients observed in the present study , may be the main reason for large numbers of parasites inflicting disease aggravation . Although these data did not allow us to predict the sources of disease promoting cytokines during infection , it encouraged us to take a more direct approach to characterize Treg cells as an integral part of disease pathology . Indeed , our data on different CD4+ T subsets and their cytokine producing ability conclusively identified CD4+CD25+ or specifically CD4+CD25+FoxP3+ Treg cells to be an important source of IL-10 , when compared to CD4+CD25- or CD4+CD25-FoxP3+ populations , respectively , in PBMCs of active VL patients . We were also able , for the first time , to describe CD4+CD25+ cells or more precisely CD4+CD25+FoxP3+ Treg cells as one of the sources of circulating TGFβ in L . donovani infected VL patients . Collectively our data suggest a possible role of Tregs , IL-10 and TGFβ in parasite establishment and course of disease progression among VL patients . The finding that CD4+CD25+FoxP3+ Treg cells could produce both IL-10 and TGFβ upon L . donovani infection is of prime importance in the knowledge of VL pathogenesis . Such findings could lead us to devise new vaccine strategies or immunotherapies against VL , which are important from the perception of eradication of parasites and cure of VL . | Visceral leishmaniasis ( VL ) is one of the most widespread parasitic diseases worldwide and is caused by kinetoplastid protozoa of the Leishmania donovani complex . The disease begins with internalization of L . donovani parasites and their multiplication within host macrophages followed subsequently by immune suppression . However , the immunological factors responsible for disease progression and their association with parasite dynamics are not completely understood . Herein , we investigated the correlation of different immune components ( cytokines and cellular subsets ) with parasite load and their involvement in the course of VL . Our study revealed a significant positive correlation between parasite load and plasma as well as antigen specific levels of IL-10 and TGFβ . In addition to cytokines , cellular subsets could also contribute to disease pathogenesis through their regulatory mechanisms . Our results indicate different Treg cell markers ( absolute numbers of CD4+CD25+ and CD4+CD25hi and percentages of CD4+CD25+FoxP3+ ) to be strongly correlated with parasite load . Exploring an association between these immunological correlates revealed Treg cells to be the source of these cytokines during VL . Therefore , this study points to a significant role of IL-10 , TGFβ and Treg cells in parasite load and active VL , providing evidence which could be helpful in devising new immunotherapeutic strategies against this disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"cytokines",
"immune",
"cells",
"body",
"fluids",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"developmental",
... | 2016 | Induction of IL-10 and TGFβ from CD4+CD25+FoxP3+ T Cells Correlates with Parasite Load in Indian Kala-azar Patients Infected with Leishmania donovani |
Hedgehog transduces signal by promoting cell surface expression of the seven-transmembrane protein Smoothened ( Smo ) in Drosophila , but the underlying mechanism remains unknown . Here we demonstrate that Smo is downregulated by ubiquitin-mediated endocytosis and degradation , and that Hh increases Smo cell surface expression by inhibiting its ubiquitination . We find that Smo is ubiquitinated at multiple Lysine residues including those in its autoinhibitory domain ( SAID ) , leading to endocytosis and degradation of Smo by both lysosome- and proteasome-dependent mechanisms . Hh inhibits Smo ubiquitination via PKA/CK1-mediated phosphorylation of SAID , leading to Smo cell surface accumulation . Inactivation of the ubiquitin activating enzyme Uba1 or perturbation of multiple components of the endocytic machinery leads to Smo accumulation and Hh pathway activation . In addition , we find that the non-visual β-arrestin Kurtz ( Krz ) interacts with Smo and acts in parallel with ubiquitination to downregulate Smo . Finally , we show that Smo ubiquitination is counteracted by the deubiquitinating enzyme UBPY/USP8 . Gain and loss of UBPY lead to reciprocal changes in Smo cell surface expression . Taken together , our results suggest that ubiquitination plays a key role in the downregulation of Smo to keep Hh pathway activity off in the absence of the ligand , and that Hh-induced phosphorylation promotes Smo cell surface accumulation by inhibiting its ubiquitination , which contributes to Hh pathway activation .
Hedgehog ( Hh ) signaling governs cell growth and patterning in species ranging from insects to human [1] , [2] . Because of its pivotal role in embryonic development and adult tissue homeostasis , misregulation of Hh signaling activity has been linked to many human disorders including birth defects and cancers [1] , [3] , [4] . Hh exerts its biological influence through a largely conserved signaling cascade that culminates at the activation of latent transcription factors Cubitus interruptus ( Ci ) /Gli [1] . The core Hh reception system consists of a 12-transmembrane protein Patched ( Ptc ) that acts as the Hh receptor and a seven-transmembrane protein Smo that acts as the Hh signal transducer [5] , [6] . Hh and Ptc reciprocally regulate the subcellular localization and active state of Smo [7]–[10] . In Drosophila , Hh stimulation or loss of Ptc leads to cell surface accumulation of Smo [7] , [11] . Increased cell surface expression and activation of Smo are regulated by Hh-induced and PKA/CK1-mediated phosphorylation of Smo carboxyl intracellular tail ( C-tail ) [12]–[14] . Several observations suggest that Smo cell surface expression is controlled by endocytic trafficking . A transmission electron microscopic study of Drosophila imaginal discs indicated that Smo is localized primarily in the lysosome of anterior compartment cells but is enriched on the plasma membrane of posterior compartment cells [15] . In Drosophila salivary gland cells , blocking endocytosis promotes Smo cell surface accumulation [11] . Using antibody uptake assay in S2 cells , we have shown that Smo reaches the cell surface but quickly internalizes in the absence of Hh and that Hh stimulation diminishes internalized Smo with a concomitant increase in cell surface Smo [13] . Taken together , these observations suggest that Hh signaling may regulate Smo cell surface expression by blocking its endocytosis and/or promoting its recycling back to the cell surface after internalization . The mechanisms by which Smo endocytic trafficking and cell surface expression are regulated have remained unknown . Smo intracellular regions lack recognizable endosomal-lysosomal sorting signals such as the NPXY and dileucine-based motifs [16] . However , many membrane receptors are internalized after covalently modified by ubiquitination , as has been demonstrated for receptor tyrosine kinases ( RTKs ) and G protein coupled receptors ( GPCRs ) [17] , [18] . The close relationship between Smo and GPCRs prompted us to investigate whether Smo cell surface expression is regulated by the ubiquitin pathway . Here we provide both genetic and biochemical evidence that Smo trafficking and degradation are regulated through multi-site ubiquitination of Smo C-tail and that Hh promotes Smo cell surface expression by inhibiting its ubiquitination . We also provide evidence that the non-visual β-arrestin Kurtz ( Krz ) acts in parallel with Smo ubiquitination to control Smo cell surface expression , and that the deubiquitinating enzyme UBPY promotes Smo cell surface expression by counteracting Smo ubiquitination .
In Drosophila wing discs , Smo cell surface level is low in anterior ( A ) compartment cells away from the A/P boundary but is elevated in response to Hh in A-compartment cells near the A/P boundary or in posterior ( P ) compartment cells ( Figure 1A ) [7] . To determine whether Smo is downregulated by the ubiquitin pathway , we generated mutant clones for Uba1 , which encodes the only ubiquitin-activating enzyme ( E1 ) in Drosophila [19] , [20] . We employed a temperature-sensitive allele of Uba1 , Uba1H33 , which behaves like a null allele at the restrictive temperature [19] . Uba1H33 clones were induced at second instar larval stage ( 48–72 h AEL ) by FRT/FLP mediated mitotic recombination . Larva carrying Uba1H33 clones were grown at permissive temperature ( 18°C ) for 3 d and then shifted to non-permissive temperature ( 30°C ) for 24 h before dissection for immunostaining . We found that anteriorly situated Uba1H33 clones accumulated high levels of Smo compared with neighboring wild type cells ( Figure 1A–B' ) , suggesting that Smo is downregulated via the ubiquitin pathway in the absence of Hh . Immunostaining with anti-Smo antibody before membrane permeabilization suggested that Smo was accumulated on the cell surface in anteriorly situated Uba1H33 clones ( Figure S1A–A' ) . A 12-h temperature shift resulted in a less robust Smo accumulation in Uba1H33 clones ( Figure S1B–B' ) , likely due to the perdurance of Uba1 activity . In general , Smo elevation coincided well with Uba1 mutant clones . Intriguingly , Uba1H33 mutant cells situated in the posterior compartment also exhibited slightly higher levels of Smo than neighboring wild type cells ( arrowhead in Figure 1B ) , suggesting that a fraction of Smo still undergoes ubiquitin-mediated degradation in the presence of Hh . To examine whether Smo is directly ubiquitinated and whether Uba1 is responsible for this activity , we carried out a cell-based ubiquitination assay ( see Materials and Methods ) [21] . We employed RNAi and/or pharmacological inhibitor to inactivate Uba1 . S2 cells stably expressing a Myc-tagged Smo ( Myc-Smo ) were treated with Uba1 or control double-stranded RNA ( dsRNA ) in the absence or presence of PYR-41 , a cell permeable E1 inhibitor [22] . The efficiency of Uba1 RNAi was confirmed by Western blot analysis of an exogenously expressed tagged Uba1 ( Figure 1C ) . Myc-Smo was ubiquitinated efficiently in the absence of Uba1 inhibition ( Figure 1D ) ; however , ubiquitination of Smo was attenuated by Uba1 RNAi and more significantly inhibited by PYR-41 ( Figure 1D ) . The incomplete blockage of Smo ubiquitination by Uba1 RNAi is likely due to partial inactivation of Uba1 by the RNAi approach . Indeed , a combined treatment with Uba1 RNAi and PYR-41 resulted in a more complete inhibition of Smo ubiquitination ( Figure 1D ) . We next applied a cell-based immunostaining assay to determine whether Uba1 regulates Smo cell surface expression [13] . Myc-Smo expressing cells were treated with control or Uba1 dsRNA in the absence or presence of PYR-41 . Cell surface and total Smo were visualized by immunostaining with an anti-SmoN antibody prior to and after cell membrane permeabilization , respectively . As shown in Figure 1E , inhibition of Uba1 either by RNAi or PYR-41 increased the levels of Smo cell surface expression and combined treatment resulted in more dramatic cell surface accumulation of Smo . Ubiquitinated membrane proteins are internalized through the endocytic pathway and targeted to lysosome for degradation [17] . We therefore examined the effect of inactivation of endocytic components on Smo accumulation in wing imaginal discs . We found that Smo was accumulated in intracellular puncta in mutant clones lacking the Drosophila homolog of HGF-regulated tyrosine kinase substrate ( Hrs ) ( Figure 2A–A' ) , a protein involved in sorting ubiquitinated membrane proteins into multivesicular bodies ( MVBs ) [23] . Of note , not all hrs mutant cells exhibited Smo puncta . This could be due to perdurance of Hrs activity and/or disc folding so that Smo puncta are present at different focal planes . RNAi of other endocytic components , including Tsg101 [24] , Avalanche ( Avl ) , a Drosophila syntaxin located in early endosomes [25] , and Rab5 , resulted in Smo accumulation in anterior compartment cells distant from the A/P boundary ( arrows in Figure 2B–E ) , as well as Hh pathway activation as indicated by Ci accumulation and ectopic expression of a Hh target gene decapentaplegic ( dpp ) ( Figure 2B–E ) . Taken together , these observations suggest that Smo is downregulated via the endocytic pathway in the absence of Hh . Consistent with Smo being downregulated through the endocytic pathway , treating Myc-Smo expressing S2 cells with a lysosome inhibitor , NH4Cl , stabilized Smo ( Figure 3A ) . Interestingly , treating cells with a proteasome inhibitor , MG132 , stabilized Smo more dramatically than treating cells with NH4Cl ( Figure 3A ) . Furthermore , combined treatment of cells with MG132 and NH4Cl had an additive effect on Smo stabilization ( Figure 3A ) , suggesting that Smo is downregulated by both lysosome- and proteasome-dependent mechanisms . However , unlike the case of Hh stimulation or Uba1 inaction where Smo was accumulated on the cell surface , proteasome inhibition stabilized Smo in intracellular vesicles ( Figure 3B ) . Double labeling with endosomal markers YFP-Rab5 ( for early endosomes ) or YFP-Rab7 ( for late endosomes ) revealed that Smo was stabilized in Rab7 positive late endosomes after MG132 treatment ( Figure 3C ) . Taken together , these observations suggest that a fraction of internalized Smo was degraded by proteasome in the endocytic pathway before reaching to lysosome . Hh induces Smo cell surface accumulation both in vitro and in vivo [7] , [11] , [13] . If Smo ubiquitination is responsible for its internalization , Hh may increase Smo cell surface expression by inhibiting its ubiquitination . Indeed , treating Myc-Smo stably expressing cells with Hh-conditioned medium markedly reduced but did not completely abolish Smo ubiquitination ( Figure 4A ) . Similarly , Ptc RNAi also reduced Smo ubiquitination ( Figure 4B ) . Our previous study demonstrated that Hh induced Smo cell surface accumulation through PKA/CK1-mediated phosphorylation of Smo C-tail [13] . We therefore determined whether Hh regulates Smo ubiquitination in a manner depending on Smo phosphorylation . We found that Hh stimulation failed to inhibit Smo ubiquitination in the presence of a PKA inhibitor H-89 ( Figure 4C ) . On the other hand , expressing a constitutively active PKA catalytic domain ( mC* ) inhibited Smo ubiquitination in the absence of Hh ( Figure 4D ) . To further determine whether Smo ubiquitination is regulated by PKA/CK1-mediated phosphorylation of its C-tail , S2 cells were transfected with Myc-tagged wild type Smo , a phosphorylation deficient form of Smo ( SmoSA ) with three PKA sites ( S667 , S687 , and S740 ) mutated to Ala , or a phospho-mimetic form of Smo ( SmoSD ) with three PKA/CK1 clusters mutated to Asp [13] , treated without or with Hh conditioned medium , and followed by the ubiquitination assay described above . As shown in Figure 4E , Hh inhibited the ubiquitination of Myc-Smo but did not significantly affect the ubiquitination of Myc-SmoSA . Furthermore , the phospho-mimetic Smo mutant , SmoSD , exhibited diminished ubiquitination and its residual ubiquitination was further reduced by Hh treatment ( Figure 4E–F ) . These results support the notion that Hh-induced phosphorylation by PKA/CK1 inhibits Smo ubiquitination , leading to its cell surface accumulation . Our previous study revealed that the Smo autoinhibitory domain ( SAID ) inhibits Smo activity in part by preventing Smo cell surface expression because a Smo variant lacking the SAID domain ( SmoΔ661–818 or SmoΔSAID ) accumulated on the cell surface in the absence of Hh stimulation [8] . To determine whether the SAID domain regulates Smo ubiquitination , we examined the ubiquitin status of a Myc-tagged SmoΔSAID ( Myc-SmoΔSAID ) . As shown in Figure 4G , deleting the SAID domain diminished Smo ubiquitination , and the residual ubiquitination of Myc-SmoΔSAID was further reduced by Hh treatment . To determine whether the SAID domain suffices to promote ubiquitination and internalization of a heterologous membrane protein , we fused it to the C-terminus of the Wingless ( Wg ) receptor Frizzle 2 ( Fz2 ) to construct Fz2-SAID chimeric protein ( FS ) . When expressed in S2 cells , CFP-tagged Fz2 ( CFP-Fz2 ) was largely accumulated on the cell surface with a small fraction internalized and colocalized with the endosomal marker Rab5 ( Figure 5A–A” ) . In contrast , CFP-FS was barely detectable on the cell surface but largely accumulated in Rab5-positive endosomes ( Figure 5B–B” ) , suggesting that SAID can promote endocytosis of Fz2 . Introducing the phosphorylation-mimetic mutation to the SAID domain of FS ( CFP-FS-SD ) reduced its endocytosis ( Figure 5C–C” ) , whereas the chimeric protein carrying a phosphorylation deficient form of SAID ( CFP-FS-SA ) was internalized as efficiently as CFP-FS ( Figure 5D–D” ) . In addition , we found that adding the phosphorylation-deficient form but not the phospho-mimetic form of SAID to Fz2 promotes the ubiquitination of the corresponding chimeric protein ( Figure 5E ) . Taken together , these observations suggest that the SAID domain suffices to promote ubiquitination and internalization of a membrane protein in a manner inhibited by phosphorylation . Combined with our earlier work [8] , it seems that the SAID domain autonomously regulates ubiquitination independent of the C-terminal negatively charged region . If Smo ubiquitination is responsible for its internalization and degradation , one would expect that ubiquitination-deficient Smo variants should be stabilized and accumulated on the cell surface . We therefore attempted to identify Lys residues responsible for Smo ubiquitination . In general , ubiquitin acceptor sites lack a strict consensus and target proteins can be ubiquitinated at multiple Lys residues . Smo C-tail and intracellular loops contain a total of 49 Lys residues , many of which may serve as ubiquitin acceptor sites , making it difficult to generate Smo variants devoid of ubiquitination . As deleting the SAID domain diminished Smo ubiquitination ( Figure 4G ) , we speculated that this region might contain Lys residues critical for Smo ubiquitination . There are a total of 13 Lys residues between aa 661 and aa 818 . We therefore constructed SmoK6R with K665 , K695 , K700 , K702 , K710 , and K733 mutated to Arg; SmoK7R with K752 , K753 , K762 , K772 , K773 , K782 , and K801 mutated to Arg; and SmoK13R with all the 13 Lys residues mutated to Arg . Using the cell-based ubiquitination assay described above , we found that both Myc-SmoK6R and Myc-SmoK7R exhibited reduced ubiquitination compared with Myc-Smo ( Figure 6A ) . The combined mutations ( K13R ) resulted in a more dramatic reduction in Smo ubiquitination ( Figure 6A ) , suggesting that Smo is ubiquitinated at multiple Lys residues between aa 661 and aa 818 . In addition , the residual ubiquitination of Myc-SmoK13R suggests that Smo is also ubiquitinated at one or more Lys residues outside the SAID domain . We next determined whether the K13R mutation affects Smo stability and cell surface expression . Myc-Smo and Myc-SmoK13R expression constructs were transfected into S2 cells together with a Myc-CFP expression construct as an internal control . The levels of Myc-Smo and Myc-SmoK13R were monitored at different time points after treatment with the protein synthesis inhibitor , cycloheximide ( CHX ) . As shown in Figure 6B , Myc-SmoK13R exhibited increased half-life compared with Myc-Smo , suggesting that inhibition of Smo ubiquitination leads to its stabilization . We also measured the steady state levels of Myc-Smo and Myc-SmoK13R in the absence or presence of MG132 and/or NH4Cl . While Myc-Smo was stabilized by both MG132 and NH4Cl , Myc-SmoK13R was stabilized by NH4Cl but insensitive to MG132 treatment ( Figure 6C ) , suggesting that inhibition of Smo ubiquitination blocks its degradation by proteasome . To determine whether inhibition of Smo ubiquitination leads to its cell surface accumulation , S2 cells were transfected with Myc-Smo or Myc-SmoK13R expression construct , followed by treatment with or without Hh-conditioned medium . Cell surface and total Smo were monitored by immunostaining with the anti-SmoN antibody before and after cell permeabilization , respectively . As shown in Figure 6D , Myc-SmoK13R exhibited higher basal level of cell surface expression than Myc-Smo; however , the level of cell surface Myc-SmoK13R in the absence of Hh was still lower than that of Myc-Smo or Myc-SmoK13R in the presence of Hh ( Figure 6D ) . Thus , although SmoK13R exhibits increased stability and cell surface expression , it is still internalized and degraded by lysosome and can be further stabilized by Hh . To determine whether the K13R mutation affects Smo stability in vivo , we generated transgenic flies expressing either UAS-Myc-Smo or UAS-Myc-SmoK13R from the same genetic locus using the phiC31 integration system to ensure similar expression level from different constructs [26] . We used the wing specific Gal4 driver MS1096 coupled with tub-Gal80ts to drive a pulse of UAS-Myc-Smo or UAS-Myc-SmoK13R expression by shifting late third instar larvae to the non-permissive temperature for 12 h . After chasing for different periods of time , wing discs were immunostained with an anti-Myc antibody . As shown in Figure S2 , after a 10 h chase , Myc-Smo was barely detectable in A-compartment cells distant from the A/P boundary , whereas Myc-SmoK13R persisted in these cells , suggesting that Myc-SmoK13R has a longer half-life than Myc-Smo . Internalization of SmoK13R is likely due to its residual ubiquitination at a Lys residue ( s ) outside the SAID domain . In addition , SmoK13R could also be internalized by Smo interacting proteins , as have been shown for other receptors [27] , [28] . It has been shown that the non-visual arrestin , β-arrestin 2 , can bind and internalize mammalian Smo [29] . The Drosophila non-visual arrestin is encoded by krz [30] . We therefore carried out both gain- and loss-of-function studies to determine whether Krz regulates Smo cell surface expression . We found that overexpression of Krz in wing imaginal discs using a dorsal compartment specific Gal4 driver , ap-Gal4 , blocked Smo accumulation in posterior-dorsal compartment cells ( compare Figure 7B with Figure 7A ) . However , we found that Smo was not accumulated in krz mutant clones located in the anterior compartment of wing discs ( Figure 7C ) . Similar observations were obtained by a recent study [31] . Using a coimmunoprecipitation assay , we found that Smo interacted with Krz through its C-tail as both Myc-Smo and Myc-SmoCT ( a Smo variant only containing its C-tail ) but not Myc-SmoΔCT ( a Smo variant with its C-tail deleted ) pulled down a C-terminally YFP-tagged Krz ( Krz-YFP ) when expressed in S2 cells ( Figure 7D ) . Furthermore , Krz-YFP could internalize SmoSD but not SmoΔCT in S2 cells ( Figure 7F ) , suggesting that Krz internalizes Smo by binding to its C-tail . The association between Smo and Krz was attenuated by Hh stimulation because Myc-Smo pulled down less Krz-YFP in the presence of Hh conditioned medium ( Figure 7E ) . In addition , Myc-SmoSD pulled down less Krz-YFP than Myc-SmoSA ( Figure 7E ) , suggesting that Smo/Krz interaction is inhibited by Hh and PKA/CK1-mediated phosphorylation . The observations that overexpression of Krz promoted Smo internalization but its loss of function did not lead to Smo cell surface accumulation suggest that a redundant mechanism ( s ) may act in parallel with Krz to internalize Smo . For example , in the absence of Krz , ubiquitination of Smo might be sufficient to promote its internalization and degradation . On the other hand , Krz could internalize Smo when Smo ubiquitination is compromised . This may explain , at least in part , why SmoK13R is still internalized and degraded by lysosome . To test this model , we examined the effect of Krz inactivation on the cell surface expression of Myc-Smo and Myc-SmoK13R in S2 cells . Consistent with the finding that loss-of-Krz has no effect on the cell surface expression of endogenous Smo in wing discs ( Figure 7C ) , Krz RNAi did not significantly affect the cell surface expression of Myc-Smo in S2 cells ( Figure 7G ) . In contrast , Krz RNAi increased the cell surface expression of Myc-SmoK13R ( Figure 7G ) , suggesting that SmoK13R is , at least in part , internalized by Krz . Similarly , Krz RNAi enhanced the cell surface accumulation of Myc-Smo induced by Uba1 RNAi or PYR41 ( Figure S3 ) , suggesting that Krz acts in parallel with ubiquitination to internalize Smo . On the other hand , overexpression of Krz-YFP blocked the cell surface accumulation of Myc-SmoK13R and this blockage was alleviated by Hh treatment ( Figure 7I ) , suggesting that Hh inhibits Krz-mediated Smo internalization . Ubiquitination is a reversible process and ubiquitin attached to target proteins can be removed by deubiquitinating enzymes/DUBs [32] . Compared with the large number of E3 ubiquitin ligases that catalyze ubiquitination of targeted proteins , each genome encodes a much smaller number of DUBs . For example , the Drosophila genome encodes over 200 annotated E3s but less than 30 annotated DUBs ( Flybase; Table S1 ) . To determine whether Smo ubiquitination is regulated by DUBs , we systematically knocked down individual DUBs by RNAi and examined the effect on Smo ubiquitination in S2 cells stably expressing Myc-Smo . From this screen , we found that RNAi of the Drosophila UBPY/USP8 significantly increased the basal levels of Smo ubiquitination ( Figure S4 ) . The effect of UBPY RNAi on Smo ubiquitination was confirmed by an independent dsRNA for UBPY ( Figure 8A ) . We also found that inactivation of UBPY by RNAi increased Smo ubiquitination in the presence of Hh ( Figure 8A ) , suggesting that UBPY counteracts Smo ubiquitination in both Hh signaling “off” and “on” states . Consistent with UBPY being able to counteract Smo ubiquitination independent of Hh signaling states , overexpression of UBPY reduced Smo ubiquitination in S2 cells both in the absence and presence of Hh ( Figure 8B ) . We then carried out coimmunoprecipitation assays to determine whether UBPY physically interacts with Smo . As shown in Figure 8C , Myc-Smo and Myc-SmoCT but not Myc-SmoΔCT pulled down a flag-tagged UBPY ( Fg-UBPY ) when expressed in S2 cells , suggesting that UBPY interacts with Smo through its C-tail . The association between UBPY and Myc-Smo was not significantly affected by Hh stimulation ( Figure 8D ) . Furthermore , UBPY appears to interact equally well with Myc-Smo , Myc-SmoSA , and Myc-SmoSD , suggesting that the bulk of Smo/UBPY association is not regulated by Hh signaling . We next examined the effect of loss- or gain-of-UBPY on Smo cell surface expression . In wing discs carrying UBPY mutant clones , Smo cell surface accumulation was attenuated in P-compartment situated UBPY mutant cells ( Figure 8E–E” ) . On the contrary , expression of UAS-UBPY using the wing specific Gal4 driver MS1096 resulted in Smo accumulation in anterior compartment cells away from the A/P boundary ( Figure 8G , J ) . Similarly , overexpression of UBPY in S2 cells markedly increased the cell surface expression of Myc-Smo ( Figure 8K ) . Overexpression of UBPY in wing discs stabilized full-length Ci ( Figure 8G' , J' ) and induced ectopic expression of dpp-lacZ in anterior dorsal compartment cells where MS1096 was expressed at high levels ( Figure 8G” ) . Smo RNAi suppressed the ectopic dpp-lacZ expression induced by UBPY overexpression as well as the endogenous dpp-lacZ expression near the A/P boundary ( Figure 8H–H” ) . However , overexpression of UBPY induced little if any ectopic expression of ptc-lacZ ( Figure 8J” ) , which is normally induced by higher levels of Hh signaling than dpp-lacZ . Taken together , these results suggest that UBPY can reverse Smo ubiquitination to promote its cell surface accumulation and induce low but not high levels of Hh pathway activation . This is in line with our previous finding that overexpression of wild type Smo only induced low levels of Hh pathway activation and full activation of Smo requires additional steps , including a phosphorylation-mediated conformational switch in Smo C-tail [7]–[10] , [13] . It is generally thought that monoubiquitination or multiubiquitination ( monoubiquitination at multiple sites ) is responsible for receptor internalization and degradation by lysosome , whereas Lys 48-linked polyubiquitination targets proteins for proteasome-mediated degradation . The observation that Smo is degraded by both lysosome and proteasome dependent mechanisms implied that Smo might undergo both types of modification . To determine if Smo could be monoubiquitinated , Myc-Smo or its KR variants was coexpressed with a HA-tagged mutant form of Ub with all Lys residues mutated to Arg ( HA-UbK0 ) in S2 cells . In this case , addition of HA-UbK0 prevents the formation of polyubiquitination chains , generating modified proteins with one or more sites monoubiquitinated . We found that Myc-Smo was effectively modified by HA-UbK0 ( Figure 9A ) . HA-UbK0 was also incorporated into Myc-SmoK6R , Myc-SmoK7R , and Myc-SmoK13R , albeit with reduced efficiency compared with Myc-Smo ( Figure 9A ) , suggesting that Smo can be monoubiquitinated at multiple sites . In the absence of proteasome inhibitor , HA-UbK0 and wild type HA-Ub were incorporated into Myc-Smo at similar levels ( Figure 9B ) , suggesting that the ubiquitinated Smo species modified by HA-UbK0 or HA-Ub detected under these conditions were mostly mono- or multi-ubiquitinated . Furthermore , Hh stimulation inhibited Smo ubiquitination under these conditions ( Figure 9B ) . However , after MG132 treatment , more HA-Ub conjugated Smo was detected than HA-UbK0 modified Smo ( Figure 9B ) , suggesting that a fraction of Myc-Smo underwent polyubiquitination that was normally degraded by proteasome . The proteasome inhibitor also increased the level of HA-UbK0 conjugated Smo ( Figure 9B ) , suggesting that a fraction of HA-UbK0 conjugated Smo might undergo polyubiquitination via endogenous Ub . To confirm that Smo could be modified by Lys 48-linked polyubiquitination , we probed Smo immunopurified from S2 cells stably expressing Myc-Smo with a Lys 48-linkage specific polyubiquitin antibody ( K48 , Cell Signaling ) . As shown in Figure 9C , immunoprecipitated Myc-Smo was recognized by the K48 antibody and the signal was markedly increased by MG132 treatment , suggesting that Smo can also be modified by Lys 48-linked polyubiquitination that targets it for proteasome-mediated degradation .
Regulation of Smo cell surface expression is a key step in Hh signal transduction [7] , [11] , [13] , but the underlying mechanism has remained unknown . In this study , we provide the first evidence that Smo is ubiquitinated in a manner regulated by Hh signaling and PKA/CK1-mediated Smo phosphorylation . We provide both genetic and biochemical evidence that Smo ubiquitination regulates its endocytic trafficking and cell surface expression . In addition , we provide evidence that the non-visual β-arrestin Krz acts in parallel with Smo ubiquitination to promote its internalization and that Smo ubiquitination is antagonized by the deubiquitinating enzyme UBPY . Several lines of evidence suggest that the ubiquitin pathway regulates Smo endocytic trafficking and degradation: ( 1 ) Smo was accumulated in mutant clones lacking the ubiquitin-activating enzyme Uba1 in wing imaginal discs , and inactivation of Uba1 in S2 cells inhibited Smo ubiquitination and promoted its cell surface accumulation; ( 2 ) Smo was accumulated when the activity of several endocytic components or lysosome was inhibited; ( 3 ) Hh and PKA/CK1-mediated Smo phosphorylation inhibited Smo ubiquitination and increased Smo cell surface expression; ( 4 ) the Smo autoinhibitory domain ( SAID ) promoted receptor ubiquitination and internalization; ( 5 ) Smo was ubiquitinated at multiple sites both inside and outside the SAID domain and mutating the ubiquitin acceptor sites in SAID increased Smo half-life and cell surface expression; and ( 6 ) Smo cell surface expression was promoted by the deubiquitinating enzyme UBPY that binds Smo and counteracts Smo ubiquitination . Early studies with yeast membrane receptors provided evidence that monoubiquitination of GPCRs mediates their agonist-induced internalization [33] , [34] . Later studies with mammalian GPCRs and other receptors suggested that both mono- and polyubiquitination could be involved in receptor endocytosis and degradation [18] . However , it has been shown that “polyubiquitination” of some receptors is due to monoubiquitination at multiple sites ( multiubiquitination ) instead of forming a polyubiquitination chain at a single site [35] , [36] . Here we provide evidence that Smo is both mono- and polyubiquitinated . It is possible that mono- or multiubiquitination may lead to Smo internalization and that internalized Smo could be further ubiquitinated in the endocytic pathway , leading to the formation of Lys 48-linked polyubiquitin chain that targets Smo for proteasome-mediated degradation ( Figure 10 ) . Thus , multiple ubiquitination events provide a robust mechanism for Smo downregulation to prevent aberrant Smo activity in the absence of Hh . Regulation of Smo trafficking and cell surface expression provides a new paradigm for how the ubiquitin pathway controls the activity of a membrane receptor . Unlike all the other cases whereby receptor ubiquitination is triggered by ligand or agonist stimulation and serves as a mechanism to control the duration of cell signaling , Smo ubiquitination occurs in the absence of ligand stimulation and serves as a mechanism to keep the basal pathway activity in check . Smo ubiquitination is inhibited upon ligand stimulation; as a consequence , Smo is accumulated on the cell surface where it becomes activated . Thus , the regulation of Smo ubiquitination by the upstream signal is in the opposite direction compared with other receptors . How does Hh block Smo ubiquitination ? Smo intracellular regions such as SAID could recruit one or more E3 ubiquitin ligases to catalyze Smo ubiquitination and E3 recruitment could be inhibited by Hh stimulation and PKA/CK1-mediated Smo phosphorylation . An alternative but not mutually exclusive mechanism is that Hh and Smo phosphorylation could promote Smo deubiquitination by regulating the binding and/or activity of one or more DUBs . In a systematic RNAi-based screen , we identified UBPY as a Smo DUB . UBPY binds Smo C-tail and antagonizes Smo ubiquitination . UBPY may modulate Smo cell surface expression by attenuating Smo endocytosis and/or promoting Smo recycling ( Figure 10 ) . However , we found that UBPY decreases Smo ubiquitination regardless of the Hh signaling states and that the association between UBPY and Smo is not significantly affected by either Hh stimulation or Smo phosphorylation , suggesting that Smo deubiquitination by UBPY is unlikely to be a major mechanism by which Hh inhibits Smo ubiquitination , although we cannot rule out the possibility that Hh regulates UBPY binding to Smo in a subtle way that escaped the detection by our coimmunoprecipitation assay . The mechanism underlying the regulation of Smo ubiquitination might be analogous to those regulating the phosphorylation of many proteins in which kinases instead of phosphatases are usually regulated by upstream signals . Thus , identifying the E3 ligase ( s ) involved in Smo ubiquitination may shed important light on the mechanism by which Smo ubiquitination is regulated . We have also obtained evidence that the non-visual β-arrestin Krz can promote Smo internalization by binding to its C-tail and this activity is inhibited by Hh . However , while Krz overexpression effectively internalized Smo , loss-of-Krz-function did not lead to a significant change in Smo cell surface expression ( Figure 7C , G ) [31] . Our results suggest that Smo ubiquitination can act independently of Krz to internalize Smo , leading to its degradation by both proteasome and lysosome so that the requirement of Krz in internalizing Smo can only be revealed when Smo ubiquitination is compromised ( Figure 7G–I ) . It is possible that Smo ubiquitination plays a major role while Krz only plays a minor role in the regulation of Smo trafficking and cell surface expression . The mechanisms that regulate Smo trafficking and cell surface expression exhibit interesting similarities to as well as important differences from those regulating GPCRs . For example , it has been shown that agonist-induced downregulation of β2-Adrenergic Receptor ( β2AR ) is mediated by both β-arrestin and receptor ubiquitination [27] . In addition , β2AR internalization and degradation is regulated by both proteasome- and lysosome-dependent mechanisms [27] , [37] . However , β2AR ubiquitination is induced by agonist and serves as a mechanism for desensitization [27] , [37] , whereas Smo ubiquitination is inhibited by Hh and serves as a mechanism for keeping pathway activity off in the absence of the ligand . β-arrestin binding to β2AR is induced by agonists and requires GRK2-mediated phosphorylation of the activated receptor [27] , whereas Krz binding to Smo is attenuated by Hh and Smo phosphorylation ( Figure 7 ) . Although GPRK2/GRK2 also regulates Smo in Drosophila , its function appears to be uncoupled from that of Krz because loss of GPRK2 exhibits a phenotype distinct from that exhibited by loss of Krz [31] , [38]–[40] . Furthermore , Krz can internalize Smo in the absence of GPRK2 [31] . β-arrestin is required for β2AR ubiquitination [27] , [37] , whereas Krz inactivation does not significantly affect Smo ubiquitination ( unpublished observations ) . Finally , while the proteasome inhibitor MG132 blocks agonist-induced β2AR internalization [27] , it does not prevent Smo internalization but instead inhibits Smo degradation after internalization ( Figure 3 ) . It is also interesting to note that β-arrestin has been implicated in the regulation of Smo trafficking and Shh signaling in vertebrates [29] , [41] , [42] . Furthermore , β-arrestin binds to mammalian Smo ( mSmo ) in a manner promoted by Shh and GRK2-mediated phosphorylation of mSmo C-tail [42] , [43] , which is analogous to agonist-induced β-arrestin binding to GPCRs . However , instead of internalizing mSmo for degradation , β-arrestin appears to promote mSmo ciliary accumulation [42] , which correlates with its positive role in Shh signaling . Both Drosophila and vertebrate Smo proteins can activate trimeric G-proteins [44]–[46] , suggesting that they are not only structurally but also functionally related to GPCRs . It is conceivable that Smo proteins may employ multiple mechanisms utilized by GPCRs to control their intracellular trafficking and activity . Thus , it will be interesting to determine whether vertebrate Smo is also regulated by the ubiquitin pathway .
Mutations used in this study are Uba1H33 [19] , l ( 2 ) 23AdD28/hrs [23] , krz1 [47] , and UBPYKO [48] . Mutant clones were generated by FLP/FRT-mediated mitotic recombination as previously described [49] . The genotypes for making clones are as follows: Uba1 clones: yw 122; FRT42 Uba1H33 /FRT42 hs-Myc-GFP; hrs clones: yw 122; l ( 2 ) 23AdD28 FRT40/ hs-Myc-GFP FRT40; krz or UBPY clones: yw 122; FRT82 krz1 or UBPYKO /FRT82 hs-Myc-GFP . Transgenic RNAi lines used are UAS-Tsg101-RNAi ( VDRC# 23944 ) , UAS-Avl-RNAi ( VDRC# 5413 ) , and UAS-Rab5-RNAi ( VDRC# 34096 ) . UAS-Krz and UAS-UBPY are previously described [47] , [48] . Constructs for various tagged forms of wild type Smo , SmoΔCT , SmoCT , SmoΔ661–818 , SmoSA , and SmoSD are previously described [8] , [13] , [50] . CFP-tagged Fz2 is described [8] . To construct Fz2/Smo chimeric proteins , the coding sequence for the wild type and mutant forms of SAID ( aa 661–818 ) was amplified by PCR and inserted at a Kpn I site between the coding sequence for Fz2 and CFP . To construct Krz-YFP , the coding sequence of Krz was amplified by PCR and inserted between Not I/ Kpn I digestion sites of pUAST vector , and YFP was inserted in frame to the C-terminus of Krz between Kpn I/ XbaI digestion sites . SmoK6R , SmoK7R , and SmoK13R were generated using PCR-based site-directed mutagenesis to introduce K to R mutations in corresponding Lys residues . Drosophila S2 cells were cultured in Drosophila SFM ( Invitrogen ) with 10% fetal bovine serum , 100 U/ml of penicillin , and 100 mg/ml of streptomycin at 23°C . Transfection was carried out by Calcium Phosphate Transfection Kit ( Specialty Media ) according to the manufacturer's instructions . Hh-conditioned medium treatment was carried out as described [51] . Cells were treated with 50 µM MG132 ( Calbiochem ) for 4 h to inhibit proteasome or 20 mM NH4Cl ( Sigma ) for 18 h to inhibit lysosome . Immunoprecipitation and Western blot analysis were carried out using standard protocols as previously described [52] . For Smo cell surface staining assay , S2 cells were harvested and washed with PBS , fixed with 4% formaldehyde at room temperature for 20 min , and incubated with the mouse anti-SmoN antibody in PBS at room temperature for 90 min . Cells were washed 3 times by PBS followed by secondary antibody staining . Immunostaining of imaginal discs was carried out as described [13] , [49] . Quantification of immunostaining and autoradiography densitometric analysis was performed using ImageJ software . Antibodies used in this study were: mouse anti-SmoN ( DSHB ) , rat anti-Ci 2A1 [53] , rabbit and mouse anti-Flag ( Sigma ) , mouse anti-Myc ( Santa Cruz ) , mouse anti-HA ( Santa Cruz ) , mouse anti-GFP ( Millipore ) , rabbit anti-GFP ( Santa Cruz ) , rabbit anti-LacZ ( ICN Pharmaceuticals , Inc . ) , anti-Ub ( P4D1 ) ( Santa Cruz ) , and anti-Poly-UbK48 ( Cell signaling ) . Ubiquitination assays were carried out based on the protocol described previously [21] . Briefly , Myc-Smo stably expressing S2 cells or S2 cells transfected with Smo variants with or without HA-Ub ( wild type or mutants ) were treated with MG132 or NH4Cl before harvesting . Cells were lysed in 100 µl of denaturing buffer ( 1% SDS/50 mM Tris , pH 7 . 5/0 . 5 mM EDTA/1 mM DTT ) . After incubation for 5 min at 100°C , the lysates were diluted 10-fold with lysis buffer and then subjected to immunoprecipitation and Western blot analysis . dsRNA was generated by MEGAscript High Yield Transcription Kit ( Ambion: #AM1334 ) according to the manufacturer's instruction . DNA templates targeting Uba1 ( aa 1–172 ) , Krz ( aa 191–365 ) , UBPY ( aa 25–191 and aa 124–290 ) or other DBUs ( Table S1 ) were generated by PCR and used for generating dsRNA . Ptc RNAi was carried out as previously described [51] . dsRNA targeting the Fire Fly Luciferase coding sequence was used as a control . For RNAi knockdown experiments , S2 cells were cultured in serum free medium containing indicated dsRNA at 23°C for 8 h . After adding fetal bovine serum to a final concentration of 10% , dsRNA treated cells were cultured overnight before transfection . 48 h after transfection , cells were harvested for further analysis . | The Hedgehog ( Hh ) family of secreted proteins governs cell growth and patterning in diverse species ranging from Drosophila to human . Hh signals across the cell surface membrane by regulating the subcellular location and conformation of a membrane protein called Smoothened ( Smo ) . In Drosophila , Smo accumulates on the cell surface in response to Hh , whereas in the absence of Hh it is internalized and degraded . The molecular mechanisms that control this intracellular trafficking and degradation of Smo were unknown , but here we show that Smo is modified by attachment of several molecules of a small protein called ubiquitin , which tags it for internalization and degradation within the cell . Hh inhibits this ubiquitination of Smo by inducing another modification , phosphorylation , of its intracellular tail by two types of protein kinase enzymes . This loss of ubiquitination and gain of phosphorylation causes the accumulation of Smo at the cell surface . What's more , we find that another protein called Kurtz interacts with Smo and acts in parallel with the ubiquitination process to promote internalization of Smo , and that the deubiquitinating enzyme UBPY/USP8 counteracts ubiquitination of Smo to promote its cell surface accumulation . Our study demonstrates that reversible ubiquitination plays a key role in regulating Smo trafficking to and from the cell surface and thus it provides novel insights into the mechanism of Hh signaling from the outside to the inside of the cell . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2012 | Hedgehog-Regulated Ubiquitination Controls Smoothened Trafficking and Cell Surface Expression in Drosophila |
The Gene Ontology ( GO ) provides biologists with a controlled terminology that describes how genes are associated with functions and how functional terms are related to one another . These term-term relationships encode how scientists conceive the organization of biological functions , and they take the form of a directed acyclic graph ( DAG ) . Here , we propose that the network structure of gene-term annotations made using GO can be employed to establish an alternative approach for grouping functional terms that captures intrinsic functional relationships that are not evident in the hierarchical structure established in the GO DAG . Instead of relying on an externally defined organization for biological functions , our approach connects biological functions together if they are performed by the same genes , as indicated in a compendium of gene annotation data from numerous different sources . We show that grouping terms by this alternate scheme provides a new framework with which to describe and predict the functions of experimentally identified sets of genes .
The Gene Ontology ( GO ) [1][2] has been around for over a decade , during which time it has been widely used both to validate and to predict the results of biological experiments ( see , for example [3–9] ) . The structure of the ontology , in which different functional “categories” or terms are related to each other in a hierarchical fashion , provides a well-established format with which to classify and subclassify all biological functions and processes . This classification approach is well-structured and well-characterized . However , we seek to determine if there is an alternate method for organizing biological functions that may in some instances be more biologically relevant or lead to important new insights . We focus on two main questions . First , can we use the information encoded in gene annotations ( which report the relationships between individual genes and functional terms , and are derived from various sources of evidence ) to identify an alternate organization for biological functions ? Secondly , if such an alternate classification exists , how can it be used to interpret biological data ? In order to answer our first question , we link functional terms together if they are performed by many of the same genes , creating a complex network of term-term relationships . We point out that although many researchers have investigated relationships between GO terms , previous studies have focused on quantifying the similarity between biological functions using the distance between terms in the ontology [10] and/or their semantic similarity as derived from ancestor terms [11 , 12] , using functional relationships for improving gene set analysis [13] or for protein function prediction [14 , 15] , discovering and incorporating links between functional terms that were not previously in the Gene Ontology [5 , 16 , 17] and even building data-driven ontologies by combining annotations made using GO with empirical data on gene interactions [18] . By contrast , in this paper we focus on the network structure of the term-term relationships that result solely from shared annotations . In doing so , our method identifies an alternate organization of biological terms that is largely distinct from the ontological organization of GO . In recent years , complex networks tools have been used alongside traditional bioinformatics techniques to study many different kinds of biological networks [19] , including , but not limited to , gene regulatory networks [20 , 21] , protein-protein interaction networks [22 , 23] , and metabolic networks [24 , 25] . Developments in network theory provide the computational tools needed to calculate the global properties of these networks , lending insights into the behavior of the systems they represent . For example , many networks exhibit community structure , meaning that there are clusters of nodes in the network within which edges are relatively dense [26] . Within the field of complex networks , many recent papers [27–33] have focused on methods to detect such modules in various types of networks in a computationally efficient and accurate manner . In this study , we leverage the community structure in gene annotation networks to develop an endogenous organization of biological functions . Our complex networks approach to organizing biological functions using annotations made to the Gene Ontology is outlined in Fig 1 . We begin by considering term relationships defined by the GO hierarchy . We then add in gene-term annotation information collected from different evidence sources and encapsulate these connections in the form of a bipartite network . Next , we use this bipartite network of gene-term relationships to construct another network describing connections between functional terms based on shared gene annotations . We apply community structure finding algorithms to partition this annotation-driven network into communities of terms and compare these communities to branches ( ontological groupings of terms ) from the GO hierarchy . We show that , although there are some similarities , there are also very strong differences between the two ways of organizing terms . Finally , we test the applicability of the community-derived classification , using functional analysis techniques to evaluate the enrichment of cancer signatures ( sets of genes associated with cancer ) in both term communities and GO branches . We find that certain signatures are enriched primarily in our term communities and not GO branches . Therefore , we suggest that by linking functional terms based on shared genes , we can create an alternate , biologically meaningful , network-derived organization of terms that is both distinct from the GO DAG and can also be used to investigate biological systems . We emphasize that our goal is not to supplant the traditional use of GO but rather to offer an alternate organization for biological functions that may in some cases provide important additional insights into the functional enrichment of experimentally derived gene sets . The annotation files and code needed to reproduce the analysis and figures in this manuscript are included in the Supplemental Material ( S1 Code ) . This information , as well as all intermediate and output data-files , can also be downloaded from [34] .
The Gene Ontology describes the relationships between different biological concepts or functions [1] . It breaks these concepts into three distinct ontologies , or primary domains: “Biological Process” ( BP ) , describing sets of molecular events , “Molecular Function” ( MF ) , describing the activities of gene products , and “Cellular Component” ( CC ) , describing parts of a cell or its external environment . Each of the three primary domains in GO takes the form of a directed acyclic graph ( DAG ) , in which “child” functional categories , or “terms” , are subclassified under one or more “parent” terms . Terms in the GO hierarchy can then be grouped into multiple overlapping sets called “branches , ” with each individual branch corresponding to a parent term and all of its descendants . Using GO , genes are annotated to individual terms representing their particular role in a cell , and these annotations are transitive up the relationships in the DAG such that each “parent” term takes on all the gene annotations associated with any of its progeny [35] . In the following analysis we explore if there exists an alternate , annotation-driven way to classify terms that is distinct from this ontology structure . To begin , we use term-term ontology relationships and gene-term annotation information for human genes downloaded from the GO website ( geneontology . org; access date: May 28 , 2015 ) to construct a gene-term bipartite network . We choose to represent this network in the form of an nG × nT adjacency matrix , where nG is the total number of genes and nT is the total number of terms . In this matrix a value of one indicates a known connection between the corresponding gene and term , and a value of zero indicates that the gene is not associated with that term . Thus , B p i = {1 if gene p is annotated to term i 0 if gene p is not annotated to term i . ( 1 ) Because annotations are transitive , edges in B will not only extend from a gene to its annotated term , but also from that gene to all the term’s ancestors ( parents , parents of parents , etc . ) in the GO DAG . The bipartite network described by B represents a summary of the relationships between 19329 human genes and 19403 functional terms , derived from many different types of biological evidence and contributed to by multiple laboratories [36] . We note that GO is divided into three primary domains and gene-annotations are made to the ontology for many species . However , for simplicity in the following analysis we combine information from all three domains and use annotation information only that pertains to human genes . Domain-specific and comparative species analysis is provided in the Supplemental Material ( S1 Text ) . Next , we used gene-term annotations to construct a network representing term-term relationships . Using the bipartite network Eq ( 1 ) one could create a term network by simply joining together any pair of terms that share common genes; however , the number of genes annotated to each term has a heavy-tailed distribution [37 , 38] , thus this approach would lose a large amount of information as connections between pairs of terms with many genes annotated to them would be given the same weight as connections between pairs of terms that only have few gene annotations . We correct for the skewed term degree distribution by constructing a diagonal weighting matrix , w , and then projecting a term network T , whose edges are modified by this weighting matrix: w i j = δ ( i , j ) ∑ q = 1 n G B q i , T = w ′ B ′ B w , ( 2 ) where δ ( i , j ) is the Kronecker delta function and takes a value of one when i = j and zero otherwise . The values of Tij take a maximum value of one when terms i and j each only have the same single gene annotation and a minimum value of zero when none of the genes annotated to term i are annotated to term j . We note that because every parent term takes on all of the annotations of each its children , Tij will necessarily be nonzero for every parent-child pair of terms . However , the weight of these relationship can be very low ( this is especially likely when the parent has a large number of annotations ) . In other words , the use of the weighting matrix serves to accentuate relationships between low degree terms . Since these terms represent biological functions performed by only a handful of genes , we believe this weighting is more likely to capture highly-specific shared biological information . We also note that because we are using gene annotations to terms in all three primary domains of GO , edges in T have the potential to link terms in different primary domains . The connections between terms in different domains has been investigated by others [39] and there are also documented “cross-domain” relationships in GO that are not subject to the DAG structure described above . In Supplemental Material , we explore these documented “cross-domain” relationships and show that they are enriched and have relatively higher edge-weights in the term-term network described by T ( Figure A ( a ) in S1 Text ) . We next sought to explore the community structure in annotation-driven term-term relationships , by identifying term communities , i . e . clusters of terms within which there are many or high-weight relationships in our projected network Eq ( 2 ) , but between which there are only few or low-weight relationships . In order to quantify the strength of community structure we use a quantity known as modularity [27] . Modularity ( Q ) can be defined as: Q = 1 2 m ∑ i j [A i j - ( 1 + r ⟨ k ⟩ ) k i k j 2 m] δ ( x i , x j ) ( 3 ) where δ is the Kronecker delta function , xi is the community of node i , ki is the degree of node i , A is the adjacency matrix , a matrix with values representing the weight between nodes i and j , and m is the total weight of the edges in the network [40 , 41] . Traditionally , in order to divide a network into communities , the resolution parameter , r in Eq ( 3 ) , is set equal to zero and a heuristic is employed to identify a partition of the network that maximizes the modularity . Varying this value allows one to look for alternate divisions of a network into communities at different scales , or resolutions , with r > 0 uncovering sub-structures in the network [41] . We used a weighted version of the Fast Greedy Community Structure algorithm [28] to investigate the community structure of our term network , and found 51 communities at maximum modularity . We then implemented a modified version of the Fast Greedy that maximizes modularity for non-zero values of the resolution parameter in order to find many different viable partitions . We varied the resolution parameter several orders of magnitude , choosing values that resulted in communities whose sizes are roughly similar to those defined by the branches of the GO DAG ( see Figure A ( b ) in S1 Text ) . This process identified 14013 different communities ( see Table A in S1 Text ) . Like GO branches , which represent overlapping sets of functional categories rather than one discreet partition of terms , communities found at different resolutions are highly overlapping and represent functional structure at many different levels of specificity . We give our communities numeric identities that vary from TC:0000001 to TC:0014013 and will refer to them as such in the following analysis . A file including these communities and their term members can be found in the Supplemental Material ( S1 Data ) .
To better understand the relationships between the communities found at different resolutions , we visualized the term communities with ten or more members for the six lowest values of resolution used ( Fig 2 ) . In this visualization each community is represented by a single circle , whose radius scales as the log of the number of terms belonging to that community and whose color corresponds to the percentage of members from each primary domain that belong to that community . Between the communities found at adjacent resolutions , we draw a line from a community at a higher resolution to a community at a lower resolution if at least 10% of the members of the community from the higher resolution also belong to the community at the lower resolution . The thickness of the line is indicative of the overlap between the two communities . For more details on the creation of this figure see the Supplemental Material ( S1 Text ) . The structure of annotation-driven term relationships is distinct from the structure of those relationships as defined by GO branches . This is evidenced clearly by the fact that , although each GO branch can only belong to one primary ontology , and thus would be pure yellow , cyan or magenta in this type of visualization , communities , even smaller ones and those found at higher resolutions , generally contain members from multiple ontologies , resulting in a rainbow of colors . We also observe that communities at higher resolutions do not merely represent the “splitting apart” of communities at lower resolutions ( represented by a child community only connecting to a single parent ) , but instead each resolution often brings about a new way of partitioning the network . An analogous visualization of GO branches reveals a similar complex partitioning , albeit segregated by primary domain ( see Figure C in S1 Text ) . Next we directly compared the membership of the term communities with that of branches in the GO DAG . In order to quantify the similarity between each community and branch , we calculated the Jaccard similarity , which takes the value J ( x , y ) = |x∩y|/|x∪y| . Then , for each community ( x ) , we determined the corresponding branch ( y ) that has the highest overlap in membership by this measure: Jm ( x ) = max{J ( x , y ) :y ∈ Y} , and vice versa . Because the exact value of the Jaccard similarity is highly sensitive to incremental changes in set membership when comparing sets with only a few members , we limit all the following analysis to communities and branches that contain ten or more terms in order to focus on the most robust results . Fig 3 ( a ) shows the distribution of Jm comparing these 2929 communities and 2439 branches . Although a handful of communities and branches are quite similar to each other , the majority of communities are dissimilar to the GO Branches and vice versa . We have repeated this analysis constructing the term network and corresponding partitions three more times , using annotations specific to each of the three primary domains , and observe similar results ( see Figure B and Table B in S1 Text ) . To better interpret these values , we selected several communities to inspect more closely . First we selected a community with a high Jm value to inspect ( Fig 3 ( b ) ) . TC:0003876 is most similar to GO:0015298 ( “solute:cation antiporter activity” ) with Jm = 0 . 5 . Overall , we observe the terms found in the community but not the branch are consistent with known biology , indicating that these connections may lead to important insights into the relationships between these functions . For example , members of the community that are not in GO:0015298 include “antiporter activity” and “potassium ion antiporter activity” . It is also interesting that in addition to members from the MF domain , TC:0003876 also includes two members from the BP domain , “calcium ion export from cell” and “calcium ion export” . One of the primary mechanisms for calcium export from the cell is through an antiporter , or exchanger [42 , 43] . Next we selected TC:0011556 , which is most similar ( Jm = 0 . 1 ) to GO:0090559 ( Fig 3 ( c ) ) . We note that the dissimilarity found between this community and branch cannot be attributed to community membership from multiple primary domains , as all of TC:0011556’s members belong to the “Biological Process” primary domain . Interestingly , the branch defined by GO:0090559 has members that belong to six different communities , demonstrating that not only are communities often distinct from branches , within the branches themselves the annotation-driven classification is often very distinct from the defined ontological relationships . The branch/community pair shown in Fig 3 ( c ) is a representative example of the maximal shared information that is typically found between a community and branches , therefore we conclude that although there is occasional similarity between our found communities and GO branches , the communities are not simply a recapitulation of the DAG . We remind the reader that because every parent term takes on the annotations of its children , in T a parent term is connected to all of its children , and vice versa . However , these relationships are differentially-weighted based on the specificity of shared annotation information ( Eq ( 2 ) ) . Therefore , what this analysis is telling us is that the specificity of shared annotations is often not the highest between a parent and a child term , but between two terms that reside in different branches of the GO . One advantage of the hierarchical organization of GO is that the collection of terms that make up a GO branch can be easily summarized by considering only the parent node of that branch . At this point we have identified strongly connected groups of terms that are organized differently from the GO DAG , but we lack a way to probe the biological information captured in these communities . We know that on a mathematical level they represent sets of biological functions that are generally performed by the same collection of genes . However identifying and understanding the biological meaning behind these communities is vital if they are to have wide-range applications similar to the GO branches . As a step toward interpreting the contents of our term communities , we visualize the names of member terms in the form of word clouds . To create a word cloud , we first make a list of all the member terms in the community , recording the primary domain of each . For each different word that appears in the list , we color it according to the percentage of its occurrences that come from terms in the different domains . For example , a coloring of yellow indicates that , within the specified community of terms , the word appeared only in term names from the domain “Biological Process . ” Similarly , cyan indicates words derived solely from “Molecular Function” terms , and magenta denotes words derived solely from “Cellular Process” terms . In this scheme , words are colored black if they have an equal ( normalized ) percentage of occurrences from all three primary domains . We also count the number of times a word appears across all member terms in a community and compare that to the word’s frequency across all terms . We then set the size of the word proportional to its statistical enrichment in the community , calculated using the hypergeometric probability . Thus the size of a word does not simply reflect its number of occurrences in the list of terms that make up a community . Rather , it reflects the statistical enrichment of its frequency in the term community compared to its frequency across all terms . Additional details about the construction of word clouds can be found in the Supplemental Material ( S1 Text ) . Following the word cloud construction technique described above and further detailed in Supplemental Material , we illustrate the biological content of two communities in Fig 4 ( a ) and 4 ( b ) . These word clouds display the richness of the biological information contained in their corresponding term communities . For example , although Community TC:0000228 ( Fig 4 ( a ) ) contains 945 members harking from all three primary domains , the word cloud presentation easily summarizes this information . We observe that this community includes biological concepts related to the cell-cycle and DNA repair , such as “mitotic” , “meiotic” , “checkpoint” , “repair” , “nucleotide-excision” , “recombination” , “replication” and more . Interestingly , the individual words are often contained in terms associated with multiple domains , resulting in a complex coloration . Our second example , TC:0000227 contains words such as “integrin” , “insulin” , “adherens” , “adhesion” and “junction” ( Fig 4 ( b ) ) . Neither community is very similar to any particular branch in GO , although they represent similar biological information . TC:0000228 is most similar ( Jm = 0 . 12 ) to GO:0022402 or “cell cycle process” , and TC:0000227 has the highest similarity ( Jm = 0 . 046 ) with GO:0016773 , or “phosphotransferase activity , alcohol group as acceptor” . We point out that one can also represent the biological information contained in branches in the form of word clouds , although , because the members of each branch can only belong to one of the three primary domains , all the words in the cloud will be the same color . Word Clouds for two branches are illustrated for comparison in Fig 4 ( c ) and 4 ( d ) . The first , GO:0050896 , or “response to stimulus” contains 905 member terms , but the corresponding word-cloud is dominated by a handful of words , including “cellular” , “stimulus” , “response” and “detection” . However , several of the smaller words , such as “stress” , “defense” , “damage” and “bacterial” , are also indicative of the types of functions encapsulated in this branch . Similarly , the cloud for GO:0002376 , whose parent term name is “immune system process” contains words pertaining to the immune system . In contrast to GO:0050896 , the richness of word-information in this cloud is more similar to that represented in the term community clouds . Finally , we wanted to test how our communities might be used in one common application of the Gene Ontology: functional enrichment analysis . The goal of functional enrichment analysis is to determine the biological functions associated with experimentally determined gene sets . Traditional methods for using the GO database to determine the functional enrichment of gene sets are designed to estimate the statistical significance of the overlap between two groups of genes: ( a ) gene set of interest and ( b ) the set of genes annotated to a particular GO term [44] . Because all genes annotated to the progeny of a given term are also annotated to that term itself , calculating the enrichment of a gene set for a specific functional term can be thought of as determining the functional enrichment of the set with respect to the group of terms represented by the term’s GO branch ( i . e . , the term itself and all of it’s descendants ) . In a similar way , we seek to determine the functional enrichment of experimentally derived gene sets with respect to the groups of terms defined by our term communities . We note that the aforementioned gene-set overlap statistics for determining functional enrichment do not account for the high level of heterogeneity in the number of functions associated with individual genes or the number of genes annotated to individual functions . Because of these limitations , we instead use Annotation Enrichment Analysis , which has been shown to address these biases by properly accounting for the heterogeneities in the null model used to determine statistical significance [38] . In practice , the appropriate treatment of these heterogeneities is particularly important when evaluating the connection between an experimentally derived gene set and a group of terms that has many associated genes . In the Supplemental Material , we provide some comparisons between AEA and a traditional method using Fisher’s Exact Test to determine enrichment , illustrating that the traditional approach erroneously identifies gene signatures as being statistically enriched with randomly constructed groups of functional terms ( Figure E in S1 Text ) . Despite the advantage of AEA in this context , we acknowledge that there are likely other biases in annotation data that it does not properly account for , and which may affect our functional enrichment results . For our analysis , we downloaded a collection of experimentally derived genes sets from the Gene Signatures Database ( GeneSigDB ) [45] . This database is a manual curation of previously published gene expression signatures , focusing primarily on cancer and stem cell signatures [46] , and includes 497 human signatures that contain 100–1000 genes annotated in the Gene Ontology . We then used Annotation Enrichment Analysis ( AEA; [38] ) with 10 , 000 randomizations to determine functional enrichment in both term communities and GO branches . For simplicity we focus on term communities and branches that have ten or more members , and exclude those with more than one thousand members . To evaluate whether term communities reflect important biological information , we determined , for each gene signature , the percentage of term communities that signature was enriched in at the p < 0 . 01 significance according to AEA . Similarly , we determined percentage of GO branches each signature was enriched in at the p < 0 . 01 significance . We then compared these values ( Fig 5A ) . We observe that not only are cancer signatures enriched in GO branches ( as might be expected ) , there is also a large level of enrichment in term communities . More interestingly , we observe a number of gene signatures that are enriched in at least one community at the p < 0 . 01 cutoff , but in no branches at this same cutoff . More specifically , there are 34 cancer signatures which are only enriched in communities at the p < 0 . 01 significance level , while only 2 signatures are only enriched in branches at the p < 0 . 01 significance level . Although the GO branches contain important biological information , this analysis demonstrates that the term communities can capture this information and additional , potentially important , functional associations . Knowing that our communities are statistically associated with experimental gene signatures , we next sought to determine in what context our term communities captured biological information from these signatures that was missed by the branches , or vice versa . Along these lines , we selected signatures that are enriched in at least one community at p < 0 . 001 but no branches at p < 0 . 01 . Fourteen signatures met this criteria . We also identified signatures that are enriched in at least one GO branch at p < 0 . 001 but no term communities at p < 0 . 01 . Only one signature met this criteria . Fig 5 ( b ) shows a heat map of the significance values representing the association of these fifteen signatures across any community or branch statistically enriched in at least one of those signatures . Additional information about these signatures can be found in the Supplemental Material ( S2 Data ) . The only signature enriched in at least one GO branch but no communities is a Leukemia signature ( bottom signature in Fig 5 ( b ) ) , which is significantly associated with GO:0050896 , or “response to stimulus” . The word cloud for the branch defined by this term and all its progeny is shown in Fig 4 ( c ) . Curiously , this signature includes all genes localized on chromosome 8 in a copy-number variation experiment exploring trisomy 8 in Acute Myeloid Leukemia ( AML ) . As noted in the original publication , the median expression of these genes was 1 . 27-fold higher in trisomy-8 cases compared to AML patients with a normal karyotype . Based on this , one could hypothesize that one effect of trisomy 8 in AML is a differential response to stimuli , something that has been observed for AML in other contexts [47] . Among the fourteen signatures that are enriched in at least one community , but no GO branches , we find that TC:0000228 , illustrated in Fig 4 ( a ) , plays an important role . This community contains terms that are related to both cell proliferation ( with words such as “cell-cycle” and “mitotic” ) and DNA repair ( with words such as “break” , “damage” , “mismatch” , and “DNA-integrity” ) . It makes sense that the cellular activities described in this community would be important across a range of cancer signatures , especially given the high rate of cell proliferation [48] and the importance of mutations in many cancers [49 , 50] . We hypothesize that one reason that this community is highlighted in our functional enrichment results may be due to the fact that genes often perform multiple functions; for example they could be simultaneously involved in both cell-cycle processes and DNA-repair . However , if the number of overall cell-cycle genes or DNA-repair genes is relatively low in a given signature , the signature will not be enriched in the corresponding GO branches . By combining these concepts rather than evaluating them separately , we believe we are highlighting important information about the biological processes important for these genes . In addition to TC:0000228 , there are also several term communities in Fig 5 ( b ) that are only enriched in a small number of our selected signatures . One example , TC:0000227 , illustrated in Fig 4 ( b ) , is most enriched in a colon cancer signature . This signature includes genes that are differentially-expressed between responders and non-responders to preoperative radiotherapy . TC:0000227 included words such as “insulin” , which is known to be associated with colon cancer risk [51] and important for mediating tumor growth [52] . TC:0000227 also includes “adherens” , “integrin” and “adhesion” . In colon cancer cells , cadherin-17 has been found to interact with α2β1-integrin to regulate cell proliferation [53] . Overall , we find interesting functional information in this term community that is highly relevant to the biology of the associated gene signature . In this section we have only discussed a subset of the term communities shown in Fig 5 ( b ) , which are themselves a small portion of all the term communities that are enriched in these cancer signatures . We note that in investigating these enrichment results , we found many other interesting biological features , which are too numerous to explore in sufficient detail here . However , our desire is that these term communities will be a resource that will help provide many future biological insights .
The network structure of gene annotations made to GO terms has not previously been exploited in a manner that reveals an organization of biological function unique from the published hierarchical classification of the Gene Ontology DAG . By analyzing the relationships between genes and functional terms reported in the GO database , we were able to construct an alternate , annotation-driven , and biologically-relevant way in which to categorize cellular functions . This categorization is structurally and conceptually distinct from the GO DAG and allows us to uncover multiple , strong connections between terms that do not share a parent/child relationship . It takes advantage of a large amount of data from a variety of sources and creates a classification scheme that is driven primarily by the data reported . Our aim is for this new organization of biological functions to be used alongside the one captured by the Gene Ontology to evaluate the functional properties of experimentally derived gene sets . The term communities defined in this work represent an integration of information across all three primary domains in GO that , to the authors’ knowledge , has not previously been investigated in this manner . However , we do not suggest that the communities we define here are the only endogenous way to group functional terms outside of the ontology structure . A different construction of T or the application of other community structure methods , such as those published in [54–57] , would likely lead different sets of functional communities . Such other alternate classifications represent a generalization of our approach and we hope to see such explorations in the future . Because annotations are continually being improved and added to the GO database ( reflecting the substantial efforts of many curators [58] ) , the organization of functional terms uncovered by our approach can change over time . This is both a drawback and a benefit of our method . It’s a drawback because researchers might reasonably desire that the results of functional enrichment calculations be independent of the state of the database at the time of the calculation . It’s a benefit because newly discovered connections between genes and functions can reveal previously missed relationships between functional terms . Here we have reported our results and analysis for a version of the GO database downloaded on May 28 , 2015 . We note that we obtained very similar results with older versions of the database . Consequently , we expect the results of our approach to be relatively stable over time , with a few exceptions that may reflect newly discovered biological phenomena . In this study we investigated if an alternative classification of GO terms exists and whether this different organization of biological functions could be used to help interpret experimental data . We believe that our functional enrichment analysis demonstrates that the term communities we define are more than a mathematical artifact and have a high potential to be applied to better interpret biological data . | Investigating how a set of genes might collectively work together to perform various cellular processes has become a routine part of many biological analyses . In such analyses , genes of interest are compared to sets of genes annotated to various biological functions ( or pathways ) defined within carefully curated databases . One of the most comprehensive and widely used resources of this type is the Gene Ontology ( GO ) database . The Gene Ontology database is comprised of two important elements: ( 1 ) the ontology itself , which provides a controlled vocabulary of terms describing genetic function and also specifies how these functional terms are related to one another via a hierarchical structure; and ( 2 ) the set of annotations made using GO that connect individual genes to different functional terms . In our paper we investigate a method for organizing functional terms that results from connecting terms based on shared gene annotations . We find that this alternate classification has an organization that is highly distinct from the Gene Ontology hierarchy , challenging the way we think about the relationships between different biological functions . Finally , we show that these alternate collections of terms are highly associated with published cancer gene signatures , demonstrating that this alternative organization of biological functions can highlight important relationships between cellular processes and has the potential to lead to new insights and discoveries . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Finding New Order in Biological Functions from the Network Structure of Gene Annotations |
Fly development amazes us by the precision and reproducibility of gene expression , especially since the initial expression patterns are established during very short nuclear cycles . Recent live imaging of hunchback promoter dynamics shows a stable steep binary expression pattern established within the three minute interphase of nuclear cycle 11 . Considering expression models of different complexity , we explore the trade-off between the ability of a regulatory system to produce a steep boundary and minimize expression variability between different nuclei . We show how a limited readout time imposed by short developmental cycles affects the gene’s ability to read positional information along the embryo’s anterior posterior axis and express reliably . Comparing our theoretical results to real-time monitoring of the hunchback transcription dynamics in live flies , we discuss possible regulatory strategies , suggesting an important role for additional binding sites , gradients or non-equilibrium binding and modified transcription factor search strategies .
During development reproducible cell identity is determined by expressing specific genes at the correct time and correct location in space in all individuals . How is this reproducible expression pattern encoded in the noisy expression of genes [1 , 2] , and read out in a short amount of time ? We study this question in one of the simplest and the best understood developmental examples—the Bicoid-hunchback system in Drosophila melanogaster . In the fly embryo , the exponentially decaying Bicoid ( Bcd ) gradient [3–5] acts as a maternal source of positional information along the embryo’s Anterior-Posterior ( AP ) axis [6] . The hunchback ( hb ) gene extracts this positional information from the local Bicoid concentration and forms a steep binary-like expression pattern , observed as early as in nuclear cycle ( nc ) 10 ( see Fig 1A ) [3 , 7–9] . This Hb pattern later becomes a source of positional information for the formation of other gap gene patterns [10 , 11] , forming the first step in the differentiation of cecullar phenotypes . From real-time monitoring of the hb transcription dynamics [12 , 13] using the MS2-MCP RNA-tagging system [14] , we observed that from nc11 to nc13 the positional readout process of the hb gene is interrupted by mitosis , leaving a window of 5-10 minutes for gene expression in each cycle . Once the pattern stabilizes 2-3 minutes after mitosis , as we describe in detail in a companion experimental paper [13] , the boundary between regions of high and low hb transcription is already steeper than even the Hb protein concentration profile in nc14 [13 , 15 , 16] . Several studies have proposed that the steep boundary between regions of high and low hb expression , given the smooth Bcd transcription factor ( TF ) gradient , is due to the cooperativity between the TF binding sites ( Fig 1B ) [7 , 15 , 17–20] . This cooperativity diversifies gene expression levels given small changes in the input [21–23] . Conventionally , the pattern steepness is represented by the Hill coefficient H . We define the hb gene readout as the hb gene transcription state of one locus in a single nucleus averaged over a given transcription window , fP , ( Fig 1C ) . We can evaluate this quantity as a function of the TF concentration [TF] ( Fig 1D ) , and thanks to the exponential nature of the Bcd gradient [15] , uniquely associate a position along the AP axis to a Bcd concentration [TF] . The Hill coefficient is then estimated by fitting the mean readout value averaged over all nuclei at a specific position along the AP axis , 〈fP〉 , to a sigmoidal function: ⟨ f P ⟩ ∼ [ T F ] H T F 0 H + [ T F ] H , ( 1 ) where TF0 is the Bcd concentration that results in half-maximal hb expression , 〈 f P | [ T F ] = T F 0 〉 = 0 . 5 ( Fig 1D ) . TF0 defines the middle of the boundary , which we will call the mid-boundary point , that separates the highly expressing “ON” nuclei in the anterior region and the minimally expressing “OFF” nuclei in the posterior region of the embryo . Within a simple model where hb expression depends only on the binding and unbinding of Bcd to the hb promoter , the maximal steepness of the hb expression pattern was shown to depend on the number of operator binding sites in the promoter region of the gene N . Depending on whether this process conserves detailed balance or not , the maximal Hill coefficient is 2N − 1 or N , respectively [18] . These studies did not address whether such a steep boundary is achievable within the limited time window of 3 to 15 minutes in nuclear cycles 11-13 , in which the Bcd concentration is read . The effects of the time constrained readout are further aggravated by the fact that transcription is stopped before and during each mitosis [13 , 24] , suggesting that the hb expression pattern needs to be re-established in each nuclear cycle . In addition , the intrinsic noise in chemical processes leads to inherent errors in the Bcd concentration readout [16 , 25] . This noise results in a lower bound for the Bcd concentration readout error , defined as the standard deviation of the concentration of the measured molecule divided by its mean ( Fig 1E ) , that depends on the readout integration time and the diffusion constant of ligand molecules [5 , 26–29] . Extending the original work that considered a single or an array of non-interacting receptors [5 , 26] , other work pointed out that cooperativity from receptor arrays increases the readout noise [30 , 31] . Given these effects , it is unclear how the readout precision of the Bcd concentration ( or nuclei position ) changes quantitatively given the highly cooperative readout process by the promoter observed as the steep hb expression pattern [9 , 15 , 16] ( Fig 1F ) and what are the consequences for the ability of neighboring nuclei to take on different cell fates . In this work , we investigate how the constraints coming from short cell cycles affect the steepness and errors in the hb expression pattern .
In the early stage of development , the hb transcription pattern is steep , despite relying mostly on the exponential Bcd gradient as the source of positional information [8] . It was hypothesized that Bcd molecules can bind cooperatively to the many Bcd binding sites on the hb promoter , enabling the gene to have diverse expression levels in response to gradual changes in the Bcd concentration [7 , 18] . We use a simple model of gene expression regulation by binding of Bcd transcription factors ( TF ) to the operator sites ( OS ) of the target promoter [18] ( Fig 1B , Fig A in S1 Text ) . The promoter activity depends on the occupancy state of the operator sites and we consider different activation schemes , which we specify below . The binding rates are functions of the position-dependent TF concentration and we further assume their value is bounded by the promoter search time of individual TFs ( see S1A Text ) . The promoter readout , fP , is defined as the mean of the promoter activity level n ( t ) , calculated from the temporal average of the promoter state n ( t ) over the steady state expression interval T of a given nuclear cycle interphase ( see Fig 1C ) . We first focus on a simplified version of the general model of gene regulation for binding of Bcd TF to the N OS [18] , where all the binding sites of the target promoter are identical . This assumption gives a Markov model of TF binding/unbinding to the many identical OS of the target promoter: P 0 ⇌ k − 1 k 1 [ T F ] P 1 ⇌ k − 2 k 2 [ T F ] P 2 … ⇌ k − N k N [ T F ] P N , ( 2 ) where Pi denotes the promoter state with i bound OS and N − i free OS . [TF] is the relative Bcd TF concentration with respect to that at the mid-boundary position . Since Bcd concentration decays exponentially along the embryo AP axis , we estimate the relative nuclei position X measured in terms of the gradient decay length from the TF concentration ( X = ln ( [TF] ) , such that at mid-boundary X = 0 and [TF] = 1 . The binding and unbinding of TF to the promoter occur with rate constants ki and k−i . If all the rates are non-zero , all reactions are reversible and Eq 2 defines an equilibrium model . Throughout the paper , we randomize the binding and unbinding rates to explore the behavior of the model ( see Methods for details ) . When comparing models with different parameters we rescale unbinding rate values k−i in order to keep the binding rate at the mid-boundary position constant . In order to best align to experimental observations , we estimate this fixed binding rate at −5% embryo length ( EL ) ( −50% EL and 50% EL are the embryo’s anterior and posterior poles ) , which is the typical boundary position in the analyzed wild type embryos [13] ( see S1B Text ) . We first consider the “all-or-nothing” case , i . e . the promoter is active when the OS are fully bound by TF ( PN ≡ Pactive ) , although the qualitative conclusions remain the same for the “K-or-more” scenario [18 , 30] , where the promoter is active if at least K sites are occupied ( see the next section ) . At steady state , we find the probability that the promoter is in the active state given the nucleus position X ( see S1A Text ) : P ( P active , X ) = K ˜ N e N · X ∑ i = 0 N K ˜ i e i X , ( 3 ) where for convenience of notation we define the effective equilibrium constant K ˜ i = ∏ j = 1 i k j / ∏ j = i i k - j and K ˜ 0 = 1 . We assume the target gene transcription rate at steady-state is proportional to the probability of the promoter to be in the active state ( Eq 3 ) . The steepness of the expression pattern is quantified by the Hill coefficient H [32] , calculated as the slope of the expression pattern at the mid-boundary position ( see S1C Text ) H = N - ( ∑ i = 1 N - 1 i · K ˜ i ) / K ˜ N . H is bounded from below by 1 , and from above by N—the OS number , confirming previous results [18] . Maximum steepness ( H = N ) is achieved when the system spends most of the time in the fully free ( P0 ) or fully bound states ( PN ) while H = 1 when the system spends most of the time in highly occupied states PN−1 and PN ( see S1D Text ) . Lastly , we also consider a full non-equilibrium binding model ( defined in Fig A in S1 and S1G Text ) , in which not all binding reactions are reversible , and reversible equilibrium models with two different types of TF factors ( defined in S1H Text ) . To explore the properties of all of these models , we solve the time dependent equations of motion for the stochastic binding models numerically and , when possible analytically in steady state , considering different expression schemes ( “all-or-nothing” and “K-or-more” ) , different numbers of TF binding sites and randomizing binding and unbinding parameters ( see Methods ) . The hb expression pattern in the early phase of development is always formed under rigorous time constraints: the total time of transcription during an interphase of duration Tfull varies in nc 10-13 from ∼100 seconds to ∼520 seconds ( Fig 2A ) . During mitosis , Bcd molecules leave the nuclei and only re-enter at the beginning of the interphase [5] . The steep expression pattern takes time to reestablish . Assuming that at the beginning of the interphase all OS of the hb promoter in all nuclei are free , the mean probability μP ( t , X ) for the promoter to be active at position X at time t following the entering of the TF to the nuclei is initially large only in the anterior of the embryo ( see Fig C in S1 Text ) . By propagating the time dependent equations of motion for the stochastic equilibrium binding models ( see S1E Text ) in time , we see that with time μP ( t , X ) increases also in other regions of the embryo , to reach its steady-state form ( P ( Pactive , X ) ) with a border between low and high expressing nuclei that defines the mid-boundary position . Whether the interphase duration Tfull in a given nuclear cycle is long enough for the system to reach steady state depends on the parameters of binding and unbinding of the TF to the operator sites . However , the binding and unbinding rates also determine the expression pattern steepness , leading to constraints between expression pattern steepness and formation time . Considering the “all-or-nothing” equilibrium model , when the promoter is active ( PN ≡ Pactive ) , any unbinding from the promoter inactivates the promoter . We find that a steep expression pattern requires the promoter to stay in the active state for a long time or to be regulated by binding of TF to many OS ( see S1D Text ) : H ≤ N - ( N - 1 ) τ bind τ active , ( 4 ) where τactive = ( k−N ) −1 is the mean time for the promoter to switch from the active state to the inactive state and τbind = ( kN ) −1 is the mean time for a TF to bind to the last unoccupied OS . The maximum value of H is reached when the promoter spends most of the time in the P0 , PN−1 and PN states ( see S1D Text ) . This limit corresponds to very slow promoter switching , τactive ≥ τbind , similarly to conclusions obtained for cell surface receptors [30] . With typically considered parameters for the Bcd-hb system [9 , 15 , 33] , τbind ∼ 4s and the H ∼ N limit in Eq 4 corresponds to τactive ≫ 4 s ( see S1D Text ) . Even if the currently available estimates for the value τbind prove inaccurate , the qualitative conclusion about slow promoter switching will remain unchanged . Given the limited interphase duration in nc 11 , τfull ∼ 270s ( Fig 2A ) , randomizing parameters of the equilibrium model ( Eq 2 ) shows that a steep steady-state expression pattern cannot be established during the interphase: the upper bound for the mean promoter activity level μP ( Tfull , 0 ) at the mid-boundary position ( X = 0 ) at the end of the interphase of duration Tfull is less than the steady state value of 0 . 5 for kinetic parameters giving large Hill coeffcients H ( Fig 2B ) . For long interphases ( Tfull ≥ 100 s ) , all patterns but those close to the maximum allowed steepness of H ≈ N reach steady state . For H ≈ N , Eq 4 imposes large τactive , which means there are not enough binding and unbinding events to achieve the steady state expression pattern with μP ( Tfull , 0 ) ∼ 0 . 5 at the boundary . Generalizing the model to allow for non-equilibrium binding ( Fig A in S1 Text ) increases the possible Hill coefficients above H > N = 6 , but does not alleviate their inaccessibility within the considered nuclear cycles 11-13 ( Fig L in S1 Text ) . Given the observed steep boundary H ∼ 7 in nuclear cycles 11-13 [13 , 16] and the relatively short interphase duration ( Tfull ∼ 520s in nc 13 , see Fig 2A ) , it seems unlikely that the steep steady state boundary is reached in early fly development with only the N = 6 known Bicoid operator sites of the proximal hb promoter [7 , 19] . Nevertheless the steady state results give a best case scenario for readout error estimates so we focus on an equilibrium steady state system in the next section . We then extend the arguments to out-of-equilibrium binding . Even when the mean promoter dynamics over the nuclear population has reached steady-state , each individual gap gene in each nucleus must independently read the positional information and express mRNA in a way to ensure the transcription pattern’s reproducibility . The promoter in each nucleus switches between an active and an inactive state n ( t ) = 0 , 1 ( Fig 1C ) . The reproducibility of the transcriptional readout f P = 1 T ∫ t = 0 T n ( t ) d t at the mid-boundary position in steady state is described by the nuclei-to-nuclei readout error of the mean activity of the nuclei CVP = δfP/〈fP〉 , where the average 〈〉 is over nuclei at the same position X = 0 calculated during the steady state expression window T in a given nuclear cycle ( Fig 1C and 1E , see S1E Text ) . Randomizing binding parameters in the equilibrium model ( Eq 2 ) we see that the lower bound for the nuclei-to-nuclei readout error , CVP , increases with increasing Hill coefficient H and decreases with the nc duration ( Fig 2C ) . A steep pattern requires slower promoter switching dynamics ( Eq 4 ) , which results in less independent measurements that take part in the single locus readout during each interphase . Therefore , the steeper the pattern , the larger the nuclei-to-nuclei readout error in the expression pattern due to the increased variability in the readouts , fP , between different nuclei [34] . When the steepness H approaches its upper bound limited by the maximum number of binding sites N , due to a small number of switching events during the interphase , the distribution of readout fP approaches a Bernoulli distribution with p = 0 . 5 with the relative error always equal to ( 1 - p ) / p = 1 , regardless of T and N . The decrease in readout error at small steepness depends on the length of the nuclear cycle ( Fig 2C ) . For very short cycles ( i . e . T < 10 s ) , only non steep patterns ( H ≤ 2 ) are able to significantly reduce the readout errors . For long interphases ( T > 100 s ) , significant reduction in readout errors can be achieved with steep patterns ( H ∼ 5 ) , and further decreasing H yields little improvement in reducing the readout error ( Fig D in S1 Text ) . In our models , τbind is the only external time scale in the problem . We assume it is set by diffusion ( S1A Text ) and all other timescales ( e . g . the time to establish the steady state expression pattern , the value of τactive that will minimize the time to establish the steady state profile ) depend on it . If our estimate of τbind ∼ 4s is inaccurate and differs by orders of magnitude , then the conclusions about not being able to establish the steep steady state expression pattern may not hold . However the point of the analysis presented in this section remains valid—steep expression profiles result in large nuclei-to-nuclei variability . The above analysis uncovers a trade-off between the readout error and steepness of the expression pattern at the boundary: the steeper the boundary , the larger the minimal nuclei-to-nuclei variability , quantified as the readout error ( Fig 2C ) . Additionally , while long nuclear cycles seem desirable both to obtain the observed steep expression patterns and decrease nuclei-to-nuclei variability , the nuclear cycles 11-13 during which these steep patterns are experimentally observed [13] are very short ( Fig 2A ) . In light of the experimental facts , steep expression patterns seem like an obstacle to reducing readout errors . The trade-off between the expression pattern steepness and the nuclei-to-nuclei variability suggests that neither of these features alone can be used as the sole criterion for a reproducible pattern . This observation is not surprising given that these features emerged from looking at the embryo from two different perspectives ( Fig 1A ) : the expression pattern steepness is perceived from an external observer’s perspective when looking at the whole embryo at a fixed time ( Fig 1D ) , while the readout error is calculated by comparing nuclei at a similar position along the AP axis averaged over time ( Fig 1E ) . These features are likely to be unobtainable to individual nuclei ( Fig 1F ) , in which the decisions about transcription are made , since they require averaging or comparing the readout of different nuclei . In order to better understand the readout of reproducible cell fates from the perspective of an individual nucleus in the fly embryo , we use the positional resolution of the expression pattern , ΔX [9 , 35] , defined as the minimum distance between two nuclei located symmetrically on the two sides of the mid-border position X = 0 that have distinct readout levels in steady state ( Fig 3A ) . Specifically , if F+ and F− are the distributions of mRNA concentrations in two nuclei at positions +ΔX/2 and −ΔX/2 ( see S1F Text ) , we define the positional resolution ΔX such that the probability of a false positive readout is small , P ( F+ ≤ F− ) ≤ 0 . 05 . Positional resolution is a distance measure that we report in length units of % egg length ( EL ) or nuclei widths , where one nucleus width corresponds to 2% EL . The width of one nucleus ( 2% EL ) sets a natural resolution scale for the problem—the embryo cannot achieve a better resolution than that of one nucleus . While positional resolution tells us how well a nucleus can distinguish its position from that of other nuclei , it is not a measure of information between the position along the AP axis and Bicoid concentration , such as the previously proposed positional information [37 , 38] . The term positional resolution is borrowed from optics , and the higher the resolution the better , since it corresponds to a smaller minimal distance between nuclei that make distinct readouts . To avoid confusion , in the text we refer to the minimal value of the positional resolution ΔX as the best case scenario when nuclei separated by a small distance make discernable readouts . The trade-off between the pattern steepness and the readout error translates into constraints on the positional resolution . For a flat expression pattern ( low H , Fig 3A , panel ( i ) ) , F+ and F− have a small difference in their mean value , which makes it hard to differentiate the mRNA concentration in closely positioned nuclei , but the variance around their mean is also small . On the other hand , with a very steep pattern ( Fig 3A , panel ( iii ) ) , F+ and F− have a big difference in their mean mRNA expression but also an increased variance , due to the increased readout errors in particular nuclei . An intermediate Hill coefficient offers the best positional resolution ( Fig 3A , panel ( ii ) ) . To evaluate the positional resolution for a given pattern steepness H and steady state expression interval T in a given nuclear cycle we randomize all the binding/unbinding parameters for a promoter with N = 6 OS—a number inspired by the number of Bicoid binding sites found on the hb promoter [7 , 19] . We identify the parameters that give the smallest CVP to ensure the smallest ΔX . CVP and ΔX are tightly correlated ( Fig I in S1 Text ) but CVP is faster to evaluate . For short nuclear cycles ( small T ) , there are hardly any promoter switching events during the readout time window and the readout error CVP ∼ 1 for all values of H ( Fig 2C ) . In this case , the positional resolution is mainly governed by the increase in the difference between F+ and F− , with increasing Hill coefficients H , which leads to a decrease in ΔX ( Fig 4A ) . As T lengthens , the value of the positional resolution ΔX for small Hill coefficients decreases with increasing H , due to the reduced readout error from averaging promoter switching events , until a certain value , ΔXmin ( T ) . As H approaches N , the readout error increases drastically since CVP → 1 ( Fig 2C ) . As a result , the value of the positional resolution ΔX increases and converges to a fixed value ΔXN ≈ 24% EL independently of T ( see S1F Text ) . We asked what values of Hill coefficients give the best ability for close-by nuclei to distinguish their position along the AP axis , and whether these values change with the duration of the nuclear cycle . To this end for each steady state transcription period T , we read-off the minimal value of the positional resolution ΔX predicted by our model , ΔXmin ( T ) , from Fig 4A to produce the orange line in Fig 4B . We also plot the optimal Hill coefficients corresponding to the minimal value of the positional resolution , H* = H ( ΔXmin ) as a function of T—the dashed blue line in Fig 4B . We found that the Hill coefficients H* that guarantee the best positional resolution decrease with the nuclear cycle duration . Since the embryo need not be performing an optimal positional readout , we found the range of Hill coefficients that allow for a margin of error of about one nucleus ( 2% of the embryo’s length ) . The choice of 2% of the embryo’s length is arbitrary , yet motivated by the observation that close-by nuclei do make different readout and this assumption allows us to explore the properties of the model . The solid blue lines in Fig 4B denote a confidence interval of H that results in a positional resolution within 2% of the embryo’s length of the optimal value . We see that for short nuclear cycles ( up to nc 11 ) , the embryo can best discriminate readouts when producing a very steep pattern ( intersect of dashed blue and dashed gray nc 11 line in Fig 4B ) . For longer nuclear cycles ( 12 and 13 ) , a narrow range of moderately steep profiles ( H* between 2 and 5 ) result in the smallest values of positional resolution ( intersect of dashed blue and dashed gray nc 12 and nc 13 line in Fig 4B ) . As the steady state transcription period T increases , ΔX becomes very small for expression profiles with a wide range of H and the constraint on H* is relaxed ( blue solid lines for large T in Fig 4B ) . In this case a discernible readout owing to small values of positional resolution can be reached even for very flat expression profiles , since time averaging alone can result in reproducible readouts . To compare the model predictions to experimental data , in Fig 4B–4D we plot the Hill coefficient ( blue dot ) and positional resolution ( orange cross ) obtained from the analysis of MS2-MCP imaging of fly embryos in nc 12 and 13 [13] . To avoid variability in the Bcd concentration between embryos , the analysis was performed by aligning 8 embryos in nc 12 and 4 embryos in nc 13 at the point of their half-maximal value of the integral fluorescence intensity . The Hill coefficients are calculated by fitting a sigmoidal curve to the mean normalized fluorescence intensity averaged over nuclei at similar positions as a function of the AP axis from data combined from multiple embryos ( see S1I Text for details ) . To calculate the positional resolution we take the normalized fluorescence intensity as the readout of each nucleus within a 5% EL bin around X = 0 and follow the procedure described above and in S1I Text . The errors bars in Fig 4B–4D for both observables represent the 95% confidence intervals . The experimental positional resolution is ΔXdata ∼ 14% EL ( confidence interval from 11% to 20% ) in nc 12 and ΔXdata ∼ 12% EL ( confidence interval from 8% EL to 18% EL ) in nc 13 . The experimental Hill coefficient value is Hdata = 6 . 9 ( confidence interval [5 . 80 , 8 . 64] , p < 0 . 05 ) in nc 12 and Hdata = 7 . 1 ( confidence interval [6 . 20 , 8 . 32] , p < 0 . 05 ) in nc 13 . The experimental positional resolution in these early nuclear cycles is well predicted by an equilibrium model with N = 6 binding sites ( orange dashed line and orange dots in Fig 4B ) , but the experimental Hill coefficient is larger than the model prediction ( blue dashed line and blue dots in Fig 4B ) . To explore the effect of multiple gene copies on positional resolution we generalize the model with M = 1 that describes the readout from a heterogenous gene to M = 2 , which describes a homogenous gene readout made independently in one nucleus ( Fig 3B ) . Although the density of nuclei does increase as nuclear cycles progress , assuming that each nuclei is making an independent measurement of the Bcd concentration ( M = 1 for a heterogenous gene or M = 2 for a homogenous gene ) , the minimal distance between nuclei that make a distinct readout measured in units of length , will not change . However , spatial averaging of the readout concentration changes the positional resolution ΔX . In our model we account for spatial averaging of mRNA in the cytoplasmic space coming from different nuclei [35] in an effective way by assuming that the readout in a given nucleus is the average of more than two genes ( M > 2 , Fig 3B ) . The results for the mRNA readout in a nucleus coming from a single expressing gene copy ( M = 1—a heterozygous fluorescent marker such as in recent MS2-MCP experiments [12] ) hold for a readout coming from more gene copies ( M > 1 , Fig 3B , Fig H in S1 Text ) . As expected , averaging over many gene copies further reduces the readout noise and slightly decreases the minimal value of positional resolution ( Fig H in S1 Text ) . We opt for an effective treatment of spatial averaging at the mRNA level , since the scale of the phenomenon has not yet been quantified in experiments in nc 11-13 and a more detailed model would require making arbitrary assumptions . In general , the strength of the averaging effect is likely to increase with time , as the nuclei density increases and the nuclear cycles get longer . Our model does not capture these time dependent effects because the role of averaging is likely to be limited during the very short time of ∼2 minutes when the steep expression pattern is established [13] . Comparing the results of the equilibrium binding site model to experimental observations , we note that the steepness values obtained in experiment ( Hdata ∼ 7 ) cannot be reached by an equilibrium model with the identified N = 6 Bcd binding sites on the proximal hb promoter . Estrada et al . [18] noted that this threshold of H = N can be overcome with a non-equilibrium binding model . We considered a full non-equilibrium model for N = 3 ( Fig A in S1 Text ) and a hybrid model for N = 6 due to the computational complexity of performing a parameter scan of a full N = 6 non-equilibrium model . In the hybrid model , the promoter has 3 OS whose interactions with TF are in equilibrium and 3 OS whose interactions with TF are out-of-equilibrium ( see S1G Text ) . The boundary steepness within these models can be larger than the number of operator sites ( H ≤ 5 for the N = 3 case , Fig K in S1 Text , and H ≤ 8 for the hybrid N = 6 case , Fig L in S1 Text ) . However , the conclusions drawn from the equilibrium model are still valid even for H > N . Large Hill coefficients result in larger readout errors ( Figs K and L in S1 Text ) . For the N = 6 hybrid model , the value of the positional resolution is minimal for large H only for very short interphase durations , and for longer interphase durations lower Hill coefficients give smaller ΔX ( Fig M in S1 Text ) . For interphase durations found in the fly embryo , intermediate Hill coefficient values , 2 ≤ H ≤ 5 , provide the best positional resolution of ∼ 6 to 10% EL or 6 to 7 nuclei lengths ( Fig 4C ) , smaller than the observed experimental values of ∼14% EL for nc 12 and 12% EL for nc 13 [13] ( orange crosses with error bars in Fig 4C , see S1I Text ) . Until now we assumed that the gene is read out only if all the binding sites are occupied . We relax this assumptions and consider the equilibrium “K-or-more” model ( Pactive ≡ Pi≥K , 1 < K < N ) , where the gene is transcribed if at least K binding sites are occupied , assuming for simplicity that transcription occurs at the same rate regardless of the promoter state . As in the “all-or-nothing” model , the attainable pattern steepness is also bounded by the number of OS ( H ≤ N − τbind/τactive ) , but to achieve a specific steepness H , the τactive in the “K-or-more” model is N−1 times smaller than that of “all-or-nothing” model . However , since the deactivation process now involves several reversible steps , τactive is also noisier . As a result , the “K-or-more” model has only a slightly faster pattern formation time and slightly lower readout error than the “all-or-nothing” case ( Fig N in S1 Text ) . In general , the ‘K-or-more” setup does not change the conclusions about the parameter regimes where the minimal value of the positional resolution ΔX can be obtained ( Fig O in S1 Text ) . We also investigated whether two mirrored transcription factor gradients , one anterior activator TF and one posterior repressor TF’ , could lower the predicted pattern steepness , at the same time keeping low values of positional resolution . While there is no direct evidence for additional regulatory gradients acting in the early nuclear cycles , the idea of an inverse gradient , possibly indirectly due to Caudal , has been suggested [39] . We assume N = 6 binding sites for the Anterior-Posterior decreasing gradient ( TF ) and L = 6 binding sites for Posterior-Anterior decreasing ( TF’ ) gradient . Transcription is allowed only when the promoter is fully bound by TF and free of TF’ and we assume that the interactions of TF and TF’ with the promoter are independent ( see S1H Text ) . In the equilibrium model , the pattern can achieve a maximum steepness of H* ∼ 7 given the total of 12 binding sites ( Fig 4D ) . The quantitative conclusions are the same as for the previously considered models ( Figs P and Q in S1 Text ) but the minimal value of the positional resolution ( ΔX ∼ 10% EL in nc 12 and nc 13 ) is smaller than that achieved with a single TF gradient , and smaller than observed experimentally . Lastly , we investigated the pattern formation when an additional repressor is concentrated in the mid-embryo region ( see S1H Text ) . This scenario is motivated by the known pattern of the Capicua ( Cic ) protein and its potential effect on transcription . In the hb promoter sequence there is one known binding motif for the Cic protein [40] . Since the Cic concentration is relatively constant at the hb pattern boundary ( ∼−5% EL from mid-embryo ) , Cic does not affect the pattern steepness . We also find that the Cic gradient contributes little to the positional resolution of the hb pattern ( Fig R in S1 Text ) . Since the interphase duration varies during the early development phase but the molecular encoding of regulation is unlikely to change , we can use the results of the simplest equilibrium model Fig 4B to define a value of a Hill coefficient , Hrobust , that gives the minimal value of the positional resolution in the widest range of steady state transcription periods T ( see Fig 5 inset ) as a function of the number of operator sites ( N ) for different numbers of expressing gene copies ( M ) . For M = 1 , Hrobust is slightly greater than N/2 , resulting in not so steep boundaries ( Fig 5 ) . Hrobust increases with M but is always smaller than its highest possible value of N allowed by the equilibrium model , even for very large numbers of expressing genes . The optimal value of the Hill coefficients in nc 12 and 13 for all the considered models , as well as the Hrobust values , are all between H ∼ 2 − 4 . These values are in very good agreement with in vitro experiments that measured the cooperativity of 6 Bcd binding sites on the hb promoter [20 , 41] ( Hdata ∼ 3 ) . Comparing the model predictions to the experimental data [13] , one can construct an equilibrium model that correctly captures the experimentally observed positional resolution , but it is much harder to achieve the readout steepness observed from the endogenous promoter given the currently identified number of binding sites . As has been shown before [18] , non-equilibrium models allow for steeper expression profiles . However , increasing the Hill coefficients to Hdata ∼ 7 [13 , 16] also increases the minimal obtainable value of the positional resolution within a hybrid non-equilibrium model to ΔX ∼ 20% EL ( ∼10 nuclei widths ) , slightly above the the experimentally observed value of ΔX ∼ 12% EL in nc 13 ( ∼6 nuclei widths ) ( Fig 4C ) . Unfortunately , from the experimental data it is hard to reliably extract Hill coefficients for nc 11 . Steep boundaries are only possible if the promoter spends most of its time in the fully occupied or fully bound states , which sets boundaries on the switching parameters [30 , 42] ( Fig F in S1 Text ) . We looked for the kinetic parameter set that yields the smallest positional resolution ΔX and , although the values vary with the interphase duration , we find that a parameter set that results in the experimentally observed ΔXdata ∼ 12% EL in nc 12 does not change over multiple nuclear cycles of varying duration . This stability throughout the nuclear cycles is consistent with experimental observations that the Bcd interactions with the hb promoter are likely independent of other TF , which suggests the binding rate constant coefficients are independent of the nuclei’s positions along the AP axes [43] . Varying the only parameter of the model τbind , which is set by the 3D diffusion assumption , rescales the steady state transcription period T ( see Fig S in S1 Text ) . However , this rescaling does not quantitatively change the conclusions of our analysis for the equilibrium models , since only the non-equilibrium model with N = 6 binding sites is able to produce boundaries as steep as those observed in the experiments ( Fig 4C ) . Within a non-equilibrium model longer binding timescales ( τbind = 40s ) than currently estimated within the diffusion approximation ( τbind = 4s ) result in a model that reproduces the observed steepness in nc 11-13 ( Fig S in S1 Text ) but , as discussed above , also results in a much higher minimal value of the positional resolution . Conversely , short binding timescales ( τbind = 0 . 4s ) allow the model to reach very low values of positional resolution in models with much smaller corresponding Hill coefficients than Hdata ( Fig S in S1 Text ) . We also asked what value of the binding rate τbind in a non-equilibrium model results in both Hill coefficients and positional resolution that is consistent with experimentally observed values . For this , we calculate the positional resolution as a function of τbind and randomized the remaining set of binding and unbinding parameters to achieve the experimentally observed Hdata ≈ 7 and the lowest value of the positional resolution given the fixed Hdata constraint ( see Fig V in S1 Text ) . The difference with the analysis in Fig S in S1 Text is that now we add an additional constraint on H = Hdata , so the minimal value of positional resolution is greater than in the results in Fig S in S1 Text . We find that for small values of τbind ∼ 0 . 01s , the mean values of the experimentally observed positional resolution ( ΔXdata ≈ 14% EL in nc12 and ΔXdata ≈ 12% EL in nc13 ) are close to the minimal value calculated in the model ( Fig V in S1 Text ) . Taking into account the confidence interval of the experimentally measured positional resolution , the experimental values are very close to the minimal predicted values of positional resolution even for τbind ∼ 0 . 1s . We conclude that a hybrid non-equilibrium model with N = 6 binding sites can reproduce both the experimentally observed Hill steepness and positional resolution , if the binding timescales are smaller than currently estimated . Achieving small τbind ∼ 0 . 1–0 . 4s requires a diffusion coefficient of D ∼ 100μm2/s , which seems an order of magnitude larger than the current estimates ( D ∼ 7 . 4μm2/s ) [9 , 33] . Misestimates in τbind = 1/Dac[TF] ) coming from the binding site size a and Bcd concentration [TF] separately are unlikely to be at the origin of such a large difference . Even considering a combined effect of a misestimate in the binding site size , Bcd concentration and the diffusion coefficient , the diffusion coefficient would need to be an order of magnitude larger . However , a different diffusion model , such as a combination of a 1D and 3D TF search for the operator site [44] could help lower the binding timescale . As a result , a non-equilibrium model with a slight modification ( additional binding site , additional regulation ) and a smaller binding rate does seem a likely candidate for explaining the experimental data . We can also compare the readout error δmRNA/〈mRNA〉 calculated directly from the MS2-MCP experiments in nc 12 and nc 13 ( Fig U in S1 Text ) . The experimental readout error in nc 12 is δmRNA/〈mRNA〉 = 0 . 82 and in nc 13 is δmRNA/〈mRNA〉 = 0 . 69 , which are lower than expected from the equilibrium model CVP ∼ 1 for the maximum allowed Hill coefficient of N = 6 , but higher than the CVP ∼ 0 . 45 in nc 12 and CVP ∼ 0 . 25 in nc 13 for the non-equilibrium hybrid model that yields the minimal value of the positional resolution . The higher experimentally observed readout error may be due to the the fact that the living embryo does not saturate the lower bound of positional resolution , as well as additional sources of noise in the experiments that are not considered in this model . These sources of noise include the random arrival times of RNA polymerases [45] , non-uniform progression of the polymerases along the DNA [46] or additional modes of regulation that manifest themselves in bursty expression even in the anterior region where Bcd binding should be saturated [25 , 47] , and possibly experimental noise . To focus on the regulatory architecture , following previous work [48–52] , we assumed the mean expression and noise at the promoter level is correlated with the mRNA readout . Exploring the role of these different sources of noise that lead to the observed readout error in conjunction with binding models of different complexity remains a future direction . The δmRNA/〈mRNA〉 values reported above are also less than the previously reported δmRNA/〈mRNA〉∼1 . 5 [25] for nuclei in a 10% EL strip centered at mid-embryo for the same 4 nuclei in nc 13 . In the previous analysis the embryos where aligned in the middle of the embryo ( 0% EL ) , which is close to the half-maximal expression point based on the mean probability of the nuclei to transcribe the gene at any point during the interphase . In the current analysis , based on the discussion in the experimental companion paper [13] , we align the embryos at their half-maximal expression point of the integral fluorescence intensity , which is typically positioned anterior to the middle of the embryo at ∼−5% EL . These results suggest that either fluctuations in the Bicoid concentration between embryos influence δmRNA/〈mRNA〉 , or that nuclei that are positioned to the posterior of the mid-boundary point ( X = 0 ) contribute more to the readout error , which is likely due to their lower expression probability . We confirm the latter hypothesis by finding for the 4 embryos aligned at X = 0 in nc 13 the CVP in a 5% strip around 0% EL , CVP = 1 . 78 , which is larger than the CVP = 0 . 69 in the strip centered at −5% EL . However , without experiments that simultaneously measure Bcd concentrations and hb expression , we cannot rule out that fluctuations in the Bicoid concentration also play a role . To explore the role of a transcriptional repressor in these trade-offs , we also considered the possibility of binding sites for an inversely directed gradient . The choice of a gradient repressor was arbitrary , since the only known mirror gradient in early fly development , Caudal , has no known binding sites in the hb promoter , no known repressor function in fly development and its maternal component has been shown to be non-essential in early fly development [53 , 54] . Nevertheless it provided for a simple choice of parameters and was motivated by earlier theoretical ideas [39] , and known activator-repressor pairs in other systems [55] . Since we are only looking at a small part of the embryo the precise form of the gradient will not strongly influence our qualitative conclusions , so we opted for the mirror image for simplicity . This two gradient model , even in its equilibrium version , does decrease the positional resolution in short cell cycles while increasing the steepness of the expression profile ( Fig 4D ) . Again , the exact results of the model do not position the experimental results for the endogenous promoter within the predictions of the model , but for the two TF gradient the minimal value of the positional resolution observed at nc 12 is obtained at earlier nc with H* ∼ 7 , ΔX ∼ 16% EL is not far from the experimentally measured value of ΔXdata ∼ 12% EL in nc 13 ( Fig 4D ) . Together these results suggest that a repressor gradient working together with Bcd in a non-equilibrium setting , possibly with additional Bcd or Hb binding sites , could explain all of the experimentally observed results . Following the above results for different binding timescales ( Fig V in S1 Text ) , an equilibrium repressor gradient model with a smaller τbind is another way to agree the model and the data . There are also other repressor candidates in the fly development , such as Capicua , which is a known repressor gradient albeit with a different profile [56 , 57] . For simplicity , motivated by Capicua , we studied a model with a constant additional repressor gradient in the middle of the embryo . Not surprisingly , due to its symmetry around the boundary , this type of gradient neither increases steepness nor severely modifies the readout error .
In order to better understand the trade-off between short cell cycles , steepness , readout error and positional resolution we studied a family of models where transcription is controlled by the binding and unbinding of the Bcd TF to multiple operator sites on the hb promoter: equilibrium binding models with different expression rules , non-equilibrium models and equilibrium models with two TF gradients . One possible way to reconcile steep profiles with small values of positional resolution are additional unidentified binding sites in the promoter . Currently the minimal hb promoter used in the experiments we are analyzing [13] is known to consist of 6 Bcd binding sites , one proximal and one distal Hb binding site . Of course , it could also include unidentified binding sites . Since we were interested in nc 11-13—the early cell cycles when the profile is already steep—we did not include the Hb binding sites in our analysis . At that stage of development the zygotic Hb gradient is weak , although there exists a maternal step-like Hb profile with a smaller amplitude than the final zygotic profile [8] . Since these Hunchback gradients have the same direction as Bcd , Hb binding sites would most likely have the same effect as additional Bcd sites so we did not add them to the model promoter . However due to the step-like shape with a boundary in the middle of the emrbyo , maternal hunchback may play a role in establishing the steep profile . The usually characterized minimal hb promoter also includes one to two Zelda binding sites but they either do not change or they decrease the pattern steepness [13] . Nevertheless additional unknown Bcd binding sites would certainly increase steepness , as could Hb binding sites . The disagreement between the model and the data is not manifested by the fact that the experimental points do not precisely fall on the theoretical predictions . The fly embryo does not need to function close to the optimal parameter regime and probably it does not . The disagreement arises because the experimentally measured values of these two observables , the Hill coefficient and the positional resolution , cannot be simultaneously obtained within the current regulatory model with the experimentally estimated diffusion limited binding time . In general , within the current models , steep boundaries increase the minimal obtainable value of the positional resolution . Specifically , the results of the model tell us that in the case of the observed steep profiles the best positional resolution that can be achieved has a much larger value than is experimentally measured . Since this is the minimal value of the positional resolution , the experimentally observed value of the positional resolution must be larger . Yet , in experiments we observe much smaller values of the positional resolution . This suggests different modes of regulation , such as described above , or smaller binding timescales than currently estimated ( Fig V in S1 Text ) . Yet if this process is fast , and early fly development is very fast , the undiscovered modes of regulation have to be simple [43] . Another explanation to consider for the discrepancy between the experimental observations and our current discussion of the model is that the assumption we made about the positional error being minimized in the developing embryo is not valid . However , even if we relax this assumption , the general conclusions do not change: the Bcd-only equilibrium N = 6 model is not compatible with the experimentally observed Hill coefficient regardless of this assumption and in the hybrid non-equilibirum model , the predicted positional regulation for the experimentally observed Hill coefficient values within the current regulatory model is larger than the observed positional regulation ( Fig M in S1 Text ) . Relaxing this assumption does , however , make it even more likely for models with different binding timescales or additional regulators or binding sites can explain both the observables simultaneously . The observed steep boundaries minimize the positional resolution only for very short cell cycles . Another possible regulatory strategy involves setting up an imprecise boundary with low positional resolution at nc 11 using a steep expression profile . This boundary would further be refined during the following cell cycles , using additional regulatory mechanisms , such as Hb regulation or epigenetic modifications encoding memory in the translational state [9] , leading to lower positional regulation . We also demonstrated that if the system starts from an out-of-steady-state condition after mitosis , the interphase duration may not be long enough for steep steady state expression patterns to establish ( Fig C in S1 Text ) . This may lead the pattern to shift along the AP axis from nuclear cycle to nuclear cycle , as observed in fly development [9] . The “all or nothing” model is clearly a simplifying assumption but we have shown that a “K-or-more” model does not change the quantitative conclusions . In the “K-or-more” model , we further , incorrectly , assume that the transcription rate is the same for all of the promoter states that enable transcription . However , given the generality of our conclusions , introducing intermediate transcription rates would change the precise numerical values of the achievable positional resolution but not the general constraints on steepness and the positional resolution . As has been pointed out in the context of maximizing information flow between the Bcd gradient and Hb output [38] , very steep boundaries decrease the ability of the nuclei to discriminate between similar Bcd concentrations . The optimal expression profiles for minimizing positional resolution are always relatively steep H > 1 , since large input fluctuations in the posterior end of the embryo coming from small Bcd concentrations limit extremely flat expression profiles . In general , we give a real biological example of the previously identified phenomenon that utlrasensitive systems require extremely slow receptor switching dynamics , which results in increased errors at the single-cell readout level [58] . Other trade-offs imposed by a need for a precise or informative readout have also been explored , including the energy—speed—accuracy constraint that shows that these three quantitates cannot be simultaneously optimized [59] or the cost of optimal information transmission in a finite time [60] . The variability in the expression states of different nuclei in the considered models comes from the binding and unbinding noise of TF to OS . The binding rates are assumed to be diffusion limited , which we implement using the Berg-Purcell bound [26] . In order to concentrate on the trade-off between steepness and positional resolution and simplify the parameter space exploration , we make the simplifying assumption that the binding and unbinding dynamics are uncoupled from diffusion . This approximation means that after an unbinding event the TF diffuses far enough from the OS so that it does not have an increased probability of binding compared to other TF molecules and its rebinding can be considered as an independent event [29] . For the equilibrium model , where all binding sites are the same , allowing for fast rebinding renormalizes the binding rates depending on the number of available free binding sites [29] . This renormalization would rescale the time axes to shorter times ( or shift the time axis to the left on the log scale ) , but would not qualitatively change the discussed results ( see Fig S in S1 Text ) . The effects of the full model of coupled binding and diffusion in the non-equilibrium model remain to be investigated in detail . Coupling the search process to the non-equilibrium process is also interesting in light of recent experimental evidence of two Bcd populations , one that spends a long time bound ( ∼ 1s ) ( < 0 . 1s ) to the DNA , and the other that spends a short time bound ( < 0 . 1s ) [61] , which could be a manifestation of specific or non-specific rebinding . We compared the experimentally measured positional resolution and steepness in the MS2-MCP experiments [12 , 13 , 24] to the M = 1 “all or nothing” model , since these experiments look at heterozygous constructs . The developing fly embryo is homozygous and has M = 2 genes , and the total resolution of the gene readout that matters for downstream genes should be determined at the protein level . Therefore the overall resolution at the protein level is different than measured by the MS2-MCP system [15] . At the protein level Gregor et al . [15] measured a Hill coefficient of H data protein = 5 in nc 14 and concluded that within the equilibrium limit of H ≤ N the known six binding sites are sufficient to achieve this steepness . In this work we consider the steepness of the mRNA readout in nc 13 and earlier , which is steeper ( Hdata ∼ 7 ) than the protein boundary at later cycles [13] . Our results therefore do not contradict previous observations [15] . The protein boundary is likely to benefit from averaging of protein concentrations between nuclei [15 , 35] . The fast timescale of about 2 minutes for achieving the steep mRNA boundary [13] suggests that the readout mechanism initially produces a steeper boundary , which is then made less steep with time , possibly due to diffusion [35] . While spatial averaging is clearly important for Hb proteins [15 , 35] , given that the steep expression profile is established in ∼2 minutes [13] , spatial averaging of hb mRNA in nc 11-12 probably plays a smaller role . Inspired by the experiments of Lucas et al . [13] we focused on nc 11-13 . The hb gene is also expressed during later stages of development [62–64] . In nc 14 , additionally to the proximal promoter active in nc 11-13 , expression of hb is also controlled by distal and shadow enhancers [1 , 2 , 43] . However they are unlikely to play a major role in nc 11-13 . Recent studies have also used an optogenetically modifiable Bcd protein [65] that makes it possible to modify the transcription of Bcd target genes . Combining all these experimental approaches with the knowledge gained both about hb mRNA [12 , 13 , 24] and Hb protein regulation [36] is a much needed future direction . In summary , we show how trade-offs between steep expression profiles and positional resolution influence the possible regulatory modes of hb expression in the short early cell cycles of fly development . We propose a number of possible solutions from non-equilibrium binding , additional regulatory gradients and binding sites , faster binding rates to epigenetic regulation . Additional experiments are needed to discriminate between the proposed scenarios . For example , testing whether the binding of TF to the promoter is equilibrium or non-equilibrium requires analysis of experiments that track TF bound to fluorescent probes that follow their binding and unbinding . Equilibrium dynamics results in time reversible traces—a property that can be evaluated based on such tagged TF data collected using high resolution microscopy .
The general model of transcription regulation through transcription factor ( TF ) binding/unbinding to the operator sites ( OS ) is based on the graph-based framework of biochemical systems [18 , 66] . In short , for a promoter with N TF binding sites the model considers all the possible 2N promoter occupancy states and all transitions between these states that involve the binding and unbinding of one TF . In most treatments of the model we randomize parameters to explore its behavior . The full non-equilibrium model is described in S1A Text and solved numerically . Assuming the binding sites are indistinguishable results in the one dimensional equilibrium model in Eq 2 . The kinetic rate constants are randomized in R + space . Assuming binding is diffusion limited by the Berg-Purcell limit [26] , the binding rate constants ki have an upper bound depending on the OS search time τbind . Based on measured and typically taken parameters for diffusion , concentration and operator size the we estimate τbind = 4s ( S1A Text ) . However , since we randomize the parameters , our quantitative conclusions do not depend on the exact values taken for these parameters . For the non-equilibrium model , i ranges from 1 to 2N−1 and there are no further constraints on the binding rates . For the equilibrium model , a reaction from Pi−1 to Pi is the binding of a TF to one of the remaining N − i + 1 free OS , so the rate constants k+i are bound by N − i + 1 ) /τbind . There are no bounds on the unbinding rate constants k−i , but their values are rescaled a posteriori so that the boundary is located in the middle of the embryo ( P ( Pactive , X = 0 ) = 0 . 5 , see S1B Text ) . The values of the rate constants are sampled to be uniformly distributed on the logarithmic scale , from 10−20 s−1 to 1020 s−1 . The number of randomized configurations tested is on the order of 105 . To find the value of ΔX for a specific kinetic parameter set , we test the condition P ( F+ ≤ F− ) ≤ 0 . 05 with increasing nuclei distance ΔW . The distribution of F+ and F− is taken as the marginal distributions of the gene readout from 500 stochastic simulation runs ( SSA ) [67 , 68] implemented in the SGNS2 simulator [69] . F− and F+ are not well-fit by Gaussian distributions , especially for short interphase durations . ΔX is the smallest value of ΔW yielding a tolerable error of P ( F+ ≤ F− ) ≤ 0 . 05 . ΔX and ΔW and nuclei position X can be expressed in units of length relative to the decay length of the TF gradient λ ≈ 100μm [5] , which corresponds to ∼20% of the embryo length ( EL ) . The data on the dynamics of hb pattern are taken from Lucas et al . 2018 [13] . In this work , hb transcription in nuclear cycle 11 to 13 is monitored using the MS2-MCP RNA tagging system [12 , 14] . From the total amount of mRNA produced per nuclei at any given position , we extracted the pattern steepness ( Hdata ) and positional resolution ( ΔXdata ) ( see S1I Text ) . | Despite very limited time , organisms develop in reproducible ways . In the early stages of fly development the information about maternal signals is read out in a few minutes to produce steep and precise gene expression patterns . Motivated by recent live imaging experiments in fly embryos , we explore the consequences of the trade-off between a rushed but reproducible readout and a steep expression pattern on the regulatory modules of gene expression . We show that the current view of one anterior gradient morphogen binding to six binding sites is quantitatively inconsistent with the experimental data given the short readout time , suggesting other regulatory features . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"chemical",
"characterization",
"gene",
"regulation",
"cell",
"cycle",
"and",
"cell",
"division",
"regulatory",
"proteins",
"messenger",
"rna",
"cell",
"processes",
"dna-binding",
"proteins",
"dna",
"transcription",
"mitosis",
"developmental",
"biology",
"transcription",
... | 2018 | Precision in a rush: Trade-offs between reproducibility and steepness of the hunchback expression pattern |
The hallmarks of Alzheimer’s disease ( AD ) are characterized by cognitive decline and behavioral changes . The most prominent brain region affected by the progression of AD is the hippocampal formation . The pathogenesis involves a successive loss of hippocampal neurons accompanied by a decline in learning and memory consolidation mainly attributed to an accumulation of senile plaques . The amyloid precursor protein ( APP ) has been identified as precursor of Aβ-peptides , the main constituents of senile plaques . Until now , little is known about the physiological function of APP within the central nervous system . The allocation of APP to the proteome of the highly dynamic presynaptic active zone ( PAZ ) highlights APP as a yet unknown player in neuronal communication and signaling . In this study , we analyze the impact of APP deletion on the hippocampal PAZ proteome . The native hippocampal PAZ derived from APP mouse mutants ( APP-KOs and NexCreAPP/APLP2-cDKOs ) was isolated by subcellular fractionation and immunopurification . Subsequently , an isobaric labeling was performed using TMT6 for protein identification and quantification by high-resolution mass spectrometry . We combine bioinformatics tools and biochemical approaches to address the proteomics dataset and to understand the role of individual proteins . The impact of APP deletion on the hippocampal PAZ proteome was visualized by creating protein-protein interaction ( PPI ) networks that incorporated APP into the synaptic vesicle cycle , cytoskeletal organization , and calcium-homeostasis . The combination of subcellular fractionation , immunopurification , proteomic analysis , and bioinformatics allowed us to identify APP as structural and functional regulator in a context-sensitive manner within the hippocampal active zone network .
Alzheimer’s disease ( AD ) , characterized by a massive loss of synapses , cognitive decline and behavioral changes , is mainly associated with an accumulation of neurofibrillary tangles and senile plaques [1–3] . The most prominent brain region affected by the progression of AD is the hippocampal formation . The pathogenesis involves a successive loss of hippocampal neurons accompanied by a decline in learning and memory consolidation . More than 20 years ago , the amyloid precursor protein ( APP ) was cloned and identified as precursor of Aβ-peptides , the main constituents of senile plaques [4 , 5] . Within the last decades much effort has gone into understanding the pathogenesis of AD . However , little is known about the physiological role of APP within the central nervous system ( CNS ) . Currently , a variety of functions have been proposed , including neurite outgrowth , synaptogenesis , and synaptic plasticity , but the underlying molecular mechanism by which APP executes its functions in neurons is still elusive [6–10] . Allocating APP to the proteome of the presynaptic active zone ( PAZ ) , a highly dynamic substructure of the presynapse , identifies APP as an yet unknown player within the neuronal communication and signaling network [11] . The presynaptic active zone is the central setting , where synaptic vesicles release their neurotransmitter into the synaptic cleft , after the arrival of an action potential and the calcium-triggered docking and fusion process [12 , 13] . Neuronal communication and signal transduction is highly dependent on the concerted action of individual proteins within the PAZ [14] . The multitude of individual proteins identified by proteomic analysis at hippocampal neurotransmitter release sites [14] emphasizes the need for identification and characterization of functional protein clusters involved in the regulation of neurotransmitter release . Within this study we combined state-of-the-art proteomics with bioinformatics to analyze the impact of APP deletion on the proteome of the hippocampal presynaptic active zone . We compared the PAZ proteomes of wild type controls , single APP-KO mice , and conditional APP/APLP2 double knockout mice for creating a common core proteome of hippocampal neurotransmitter release sites . Lack of APP induces only a mild phenotype in young animals ( e . g . reduced body weight , reduced grip strength ) [10] . This condition is changed in the elderly mice , when impairments in spatial and working memory become obvious [7 , 9 , 10] . Compared to the APP-KO mice , deletion of both APP and APLP2 provokes a lethal phenotype [15] . To circumvent postnatal lethality a conditional APP/APLP2 double knockout has recently been designed making use of the NexCre mouse line [7] . Within excitatory forebrain neurons Cre recombinase becomes activated under the control of NEX at embryonic day E11 . 5 allowing deletion of floxed proteins in distinct brain regions such as the hippocampus [16] . The NexCre APP/APLP2 cDKO ( NexCre-cDKO ) turned out to be viable but displayed severe impairments in memory related tasks associated with reduced long term potentiation ( LTP ) already in young animals [7] . The effect of APP deletion on hippocampus-specific function ( LTP ) points to an essential role of APP for learning and memory consolidation . Furthermore , the impact of APP deletion and APP/APLP2 deletion within the CNS point to general physiological role in synaptic plasticity and synaptic maintenances [6–10] . In this context , we addressed the question of the cellular function of APP at synaptic release sites by comparing wild type and knockout mice and creating protein-protein interaction networks . Former observations point to an as yet unknown link that incorporates APP into the physiological network of the presynaptic active zone . Our novel experimental design for the isolation and purification of the native hippocampal PAZ [14] along with quantitative biochemical and proteomic approaches allowed us to compare APP-mutants to their controls and to unravel the proximate impact of APP deletion . Apart from this methodological approach we visualized the impact of APP on the hippocampal PAZ proteome by creating protein-protein interaction ( PPI ) networks that incorporate APP into protein networks of the synaptic vesicle cycle , cytoskeletal organization and calcium-homeostasis . Or results derived from the combination of multidisciplinary biological techniques suggest that APP is as structural and functional regulator in a context-sensitive manner within the hippocampal active zone network .
Animal treatment was performed under veterinary supervision in accordance with animal welfare regulations of the German animal protection law ( Regierungspräsidium Darmstadt and Karlsruhe , Germany ) . All mice were maintained in an animal facility , and the experimental procedures were approved by the Animal Welfare office of the Regierungspräsidium Darmstadt and Karlsruhe . APP-KO ( APP-/- ) compared to wildtype ( C57BL/6N ) , and APPflox/floxAPLP2−/−NexCre+/T ( NexCre-cDKO ) compared to APPflox/floxAPLP2-/- ( APLP2-/- ) , mice of both sexes and age-matched ( 12 or 16 weeks respectively ) were kept under 12 h light and dark cycle with food and water ad libitum . The generation of APP mutant mice has previously been described [7 , 17] . Antibodies were directed against APLP1 ( CT-11 , rabbit polyclonal , 1:2000 , Calbiochem—Merck Millipore , Darmstadt , Germany ) APP ( Y188 , rabbit monoclonal , 1:1000 , Abcam , Cambridge , UK ) , Munc18 ( rabbit polyclonal , 1:1000 Synaptic Systems , Göttingen , Germany ) , NCAM ( rabbit polyclonal , 1:1000 , Abcam , Cambridge , UK ) , SNAP25 ( mouse monoclonal 1:1000 Synaptic Systems , Göttingen , Germany ) , synaptophysin ( G63 , rabbit polyclonal , 1:1000 , kindly donated by Dr . R . Jahn , Göttingen , Germany ) , synaptotagmin-1 ( rabbit polyclonal , 1:1000 , Synaptic System , Göttingen , Germany ) , syntaxin-1 ( mouse monoclonal , 1:1000 Synaptic Systems , Göttingen , Germany ) , SV2 ( the clone CKK 10H4 producing the monoclonal anti-SV2 antibody , kindly donated by Dr . Regis B . Kelly , San Francisco , CA , USA; was cultured in-house ) , SV2A ( rabbit polyclonal , 1:1000 kindly donated by Dr . S . Bajjalieh , Seattle , USA ) , and VAMP2 ( cl . 69 . 1 , mouse monoclonal , 1:1000 Synaptic Systems , Göttingen , Germany ) . Dynabeads M-280 conjugated with monoclonal sheep anti-mouse IgGs ( cat . No . 112 . 02D ) were purchased from Invitrogen , Darmstadt , Germany . The hippocampus was dissected from native mouse brain prior to subcellular fractionation . Synaptic vesicles were isolated from synaptosomes according to the protocol guidelines of Whittaker [18] . The protocol has previously been adapted to the fractionation of individual mouse brains [19] and downscaled for individual mouse brain regions [14] . The following modifications were applied: Individual hippocampi was homogenized in 0 . 4 mL of preparation buffer ( 5 mM Tris-HCl , 320 mM sucrose , pH 7 . 4 ) containing the protease inhibitors antipain , leupeptin , chymostatin ( 2 μg/mL each ) , pepstatin ( 1 μg/mL ) and benzamidine ( 1 mM ) . Unless otherwise mentioned the material was kept at 4°C during the entire preparation . The hippocampal homogenate was centrifuged using a Beckman TLX Optima 120 and rotor TLA 120 . 2 by acceleration ( mode 4 ) up to 2800gav for 2 min . The resulting pellet was discarded and the supernatant was further fractionated by discontinuous Percoll gradient centrifugation . The Percoll gradient was prepared by layering 1 . 0 mL supernatant solution onto three layers of 1 . 0 mL Percoll solution [3% , 10% , 23% ( v/v ) in preparation buffer] . After centrifugation using the TLA 100 . 4 rotor for 7 min at 35 , 000gav , fractions containing synaptosomes were collected and diluted twofold in preparation buffer and centrifuged using TLA 100 . 4 rotor for 35 min at 50 , 000gav . For hypoosmotic lysis of synaptosomes the resulting pellet was triturated in four volumes of lysis buffer ( 5 mM Tris-HCl , pH 7 . 4 ) at room temperature . The suspension was centrifuged using the TLA 100 . 4 rotor for 60 min at 250 , 000gav . The pellet was resuspended and homogenized in 300 μL sucrose buffer ( 10 mM HEPES-NaOH , 0 . 5 mM EGTA , 0 . 1 mM MgCl2 , 200 mM sucrose , pH 7 . 4 ) . This microsomal solution was layered onto 900 μL of a discontinuous sucrose gradient ( 0 . 3 M , 0 . 75 M , and 1 . 2 M; containing 10 mM HEPES , 0 . 5 mM EGTA , adjusted to pH 7 . 4 ) and centrifuged using a WX Ultra 90 Sorvall centrifuge and the TST 55 . 5 rotor for 2 h at 65 , 000gav . Thirty-six fractions ( 35 μL each ) were collected from top to bottom of the gradient . The pooled lower fractions ( LF ) 16 to 30 corresponding to sucrose concentrations of 0 . 5 to 1 . 1 M were further analyzed . Immunopurification of the hippocampal presynaptic active zone via docked synaptic vesicles The immunopurification protocol for the presynaptic active zone ( PAZ ) via docked synaptic vesicles was as described recently for individual mouse brain regions [14] . In brief , 100 μL magnetic beads pre-coupled with an anti-mouse monoclonal antibody were washed with Tris-buffered saline ( TBS , pH 7 . 4 ) and incubated with TBS containing 1% glycine , 1% lysine and 0 . 5% saponin followed by three washing steps in TBS . Magnetic beads were then incubated for 1 h with the anti-SV2 antibody ( 3 μg of antibody per 107 magnetic beads to gain representative SV2 population ) . Crosslinking of the antibodies was performed with 0 . 1% glutardialdehyde in TBS for 5 min and stopped by adding TBS containing 1% glycine and 1% lysine . Finally the beads were incubated over night at 4°C with the pooled lower sucrose gradient fractions ( LF , 16–30 ) . Beads containing the bound material were three times washed with TBS and incubated with ice-cold acidified acetone ( acetone containing 125 mM HCl ) for 30 min at 20°C . Elution was performed with different elution agents for 30 min . For Western blot analysis proteins were eluted with sample buffer containing 2% SDS . For MS analysis proteins were eluted with 100 mM triethylammonium bicarbonate ( TEAB ) . The elution of PAZ proteins was supported by applying short ultrasonic pulses . For quantification of protein contents the BCA-assay kit ( #23225; Pierce , Rockford , IL , USA ) was applied . Immunopurified material was eluted from the beads with 2% SDS , 62 . 5 mM Tris , pH 6 . 8 , prior to protein determination . The BCA kit tolerates up to 5% SDS , and 2% SDS are recommended to eliminate interference by lipids . Subsequently proteins were dissolved with sample buffer containing 2% SDS , 62 . 5 mM Tris , pH 6 . 8 , 10% glycerol , and 0 . 01% bromophenol blue . Equal amounts of protein ( 100 ng ) were resolved on a 15% Tris-glycine SDS-PAGE [20] and transferred onto nitrocellulose membrane ( GE Healthcare ) using semi-dry blotting techniques ( BioRad ) . Membranes were blocked with 5% skimmed milk powder in PBS/T ( 123 mM NaCl , 7 . 4 mM Na2HPO4 , 4 . 3 mM KH2PO4 , 0 . 1% Tween20 ) for 1 h . Incubation with the respective primary antibody was performed over night at 4°C followed by a second blocking step with 5% skimmed milk powder ( five times , 10 min each ) , subsequent incubation with the respective HRP-conjugated secondary antibody ( GE Healthcare ) and a final washing step in PBS/T ( five times , 10 min each ) . Immunoblots were incubated with Western Lightning ECL substrate and visualized using ImageQuant LAS 4000 ( both GE Healthcare ) . Quantification of immunosignals was performed with samples obtained under identical experimental conditions ( n = 3 ) and run in one gel . Pixel intensities of non-saturated bands ( ±SEM , standard error of the mean ) from the same blot , were measured in voxels using ImageQuant TL software . Data were statistically processed employing unpaired Student’s t-test . The immunopurified presynaptic active zone ( PAZ ) derived from mouse hippocampus was subjected to enzymatic digestion using the well-established serine protease trypsin . The amount of trypsin ( Proteomics Grade , Sigma Aldrich , St . Louis , MO ) was adjusted to an enzyme-to-substrate ratio of 1:50 for each sample according to the protein concentrations determined by BCA Protein Assay ( Pierce , Thermo Scientific ) . The digestion was performed at 37°C for 18 h and stopped by adding 3 μL of formic acid ( FA ) . Samples were dried down and solubilized in solvent A ( 5% MeCN , 0 . 1% FA ) to obtain a final concentration of 1 μg peptide mixture per μL . Peptide labeling with TMT sixplex was carried out according to the manufacturer’s protocol ( TMT6 , Thermo Scientific ) . The immunopurified hippocampal PAZ derived from the respective wild type mice was labeled with TMT-126 , TMT-127 and TMT-128 , whereas the immunopurified hippocampal PAZ from APP mutants was labeled with TMT-129 , TMT-130 , and TMT-131 . The peptide mixtures were combined , purified and desalted by employing Pierce C18 spin columns ( Thermo Scientific ) and finally dried down . Prior to separation the sample was solubilized in solvent A ( 2% MeCN , 0 . 1% formic acid ) to a final concentration of 1 μg/μl . Chromatographic separation of the mixture was performed on a Dionex Ultimate 3000 RSLC nano system ( Thermo Scientific ) . Peptides were trapped on an Acclaim PepMap 100 μCartridge Column C18 , 300 μm x 0 . 5 cm , 5 μm , 100 Å ( backflush mode ) trapping column prior to separation on an Acclaim PepMap 100 C18 ( 2 μm , 100 Å , 75 μm i . d . x 50 cm ) EASY-Spray nano column at a flow rate of 250 nL/min . A gradient of 150 minutes with increasing amounts of solvent B ( 80% MeCN , 0 . 08% FA ) was applied . The LC system was coupled online to a Q-Exactive Plus hybrid quadrupole-Orbitrap instrument ( Thermo Scientific ) and operated in a data-dependent acquisition mode selecting the top 12 most intense peaks from a survey scan for fragmentation . The survey scan was performed using the following parameters: scan range between 300–1500 m/z at a resolution of 70000 at m/z 200 . The AGC target value was set to 3e6 . MS/MS scans were acquired at a resolution of 17 , 500 at m/z 200 with an AGC target of 1e5 . The precursors were isolated with an isolation width of 1 . 6 Da and the normalized collision energy ( NCE ) was 35 . Dynamic exclusion was set to 20s , peptide recognition mode was enabled while singly and charged state above 8 and unassigned precursor ions were disabled . The immunopurified hippocampal PAZs were analyzed in triplicate . Data processing , database searches for protein identification and relative quantification were performed using Proteome Discoverer ( V1 . 4 . 0 . 288 , Thermo Scientific ) . Spectra were deconvoluted by the integrated software module and a signal-to-noise filter of 1 . 5 was applied . The database search was performed on an in-house Mascot server . Precursor mass tolerance was set to 10 ppm and fragment mass tolerance to 0 . 02 Da . Oxidation of methionine and deamidation of asparagine and glutamine were allowed as variable modifications and TMT was set as fixed modification for lysine residues and peptide N-termini . Spectra were searched against a database of murine proteins ( SwissProt , released on 2014-02-19 ) and a two-level decoy search was performed , with a target FDR value of ≤ 1% . For quantification , the mass window for peak integration tolerance was set to 20 ppm and the most confident centroid peak detection was applied . Only unique peptides were considered and the results were normalized on protein median by the built-in software module . Only proteins that were identified in each experiment and present in the hippocampal PAZ of all animals ( n = 12 ) were considered for further analysis . Cytoscape 3 . 2 . 1 [30] , an open source software platform for visualizing , integrating , and analyzing networks , was used to create the actual PPI network . All data files were imported into Cytoscape and the networks were merged into a single network using Cytoscape’s network merging tool . Filters were applied to retain only physical interactions which were experimentally detected . We used a selection of proteins that are constituents of the hippocampal PAZ . Singletons and small connected components which had no connection to the giant connected component were excluded . Additional node information was loaded from UniProt and the relative abundance was attributed to the respective node . Localization of proteins was derived from UniProt and mapped manually onto each node . The Cytoscape 3 session file of or data set is provided as S1 Cytoscape 3 File . The Cytoscape NetworkAnalyzer tool was used to analyze the topology of the network and to compute centrality values for each node . The importance of nodes in the network was ranked in accordance with their centrality values . Centralities are numerical values assigned to nodes based on statistical properties . We used the clustering coefficient , degree , and shortest path betweenness as implemented in the Cytoscape NetworkAnalyzer tool . The cluster coefficient describes the densities of connections between neighbors with a range from 0 to 1 the latter with the highest density ( 100% ) , degree is the number of interactions , and betweenness represents the ratio of all shortest paths between two nodes to the number of shortest paths a given node ( e . g . APP ) is participating in . Community detection was performed for unweighted modularity [31 , 32] using Radatools 3 . 2 [33] that outputs the best partition found in form of a text file which contains information about the number of elements in each partition , a list of elements , the size of each community , and the indices of each element . The Radatools software implements several algorithms for the optimization of modularity . Here , a combination of tabu search [33] , extremal optimization [34] , fast heuristics [35] and spectral optimization [36] was used . To visualize the distributions of up- and downregulated proteins in APP-KO and NexCre-cDKO the relative abundances of all proteins were mapped on the network . We used Cytoscape 3 . 2 . 1 for all network layouts and visualizations . All network visualizations are optimized for color vision deficiency . Changes in protein abundance of more than ±10% are reflected by increasing sizes of the nodes .
To further characterize the impact of APP deletion the abundance of hippocampal constituents was visualized according to both their subcellular localization ( Fig 3A , left ) and their functional allocation ( Fig 3B , right ) . The final network layout illustrates proteins as nodes and interactions ( experimentally validated physical interactions ) between proteins as edges ( Figs 3–8 ) . The resulting 615 proteins revealed a high degree of networking , confirming their direct association with the hippocampal PAZ core proteome . The separation into localization and function allowed us to gain further insights into the topological connectivity of single proteins and the organization of individual proteins within the proteomic network ( Fig 3 ) . Localization , embodied by six groups according to the pie chart layout , highlights all structural core constituents of the PAZ network ( color , protein number ) : mitochondria ( red , 188 ) , plasma membrane ( blue , 108 ) , synaptic vesicle ( turquoise , 81 ) , signaling cascade ( green , 113 ) , cytoskeleton ( yellow , 79 ) , and metabolic enzymes ( orange , 46 ) ( Fig 3A ) . The term function in a network is associated with topological features that allow insight into the connectivity and relationship not only of single proteins but also insight into the functional organization of a proteome ( community structures ) . A community is a densely connected subnetwork within the PPI network of the PAZ and represents a functional module within the entire community structure . To find a linkage between proteins and their functions within the network we applied a community detection algorithm resulting in seven communities ( Fig 3B ) . In the next step , we mapped the relative abundance of each protein within each layout to visualize the impact of APP deletion on the PPI-network ( Fig 3C–3F , APP-KO Fig 3C and 3D , NexCre-cDKO Fig 3E and 3F ) . Proteins affected by the deletion were highlighted in magenta ( upregulation ) or green ( downregulation ) , whereas slightly affected proteins display a color-gradient and non-affected proteins are in yellow ( Fig 3C–3F ) . Within the localization and community structure network , APP deletion displays a wide distribution of up- and downregulated proteins ( Fig 3C and 3D ) whereas deletion of APP in NexCre-cDKO revealed a defined pattern allocating significantly dysregulated proteins to individual groups ( e . g . mitochondria , Fig 3E and 3F ) . The community structure provided by Radatools does not resolve the functional structure unambiguously . Within the community structures individual protein groups ( functional clusters ) are represented , e . g . synaptic vesicle exo- and endocytosis , Ca2+-homeostasis and cytoskeleton . Therefore , each community structure consists of a collection of proteins that belongs to individual functional cluster . This collection of proteins can be summarized under the term heterogeneity to emphasize this complex and diverse layout of the community structures . For better visualization of this “heterogenic nature” , we created a subcommunity layout representing all individual functional clusters ( Fig 4 ) . Proteins of the respective functional cluster are listed in S1 Table ( Worksheet SubCom—Subnet ) In addition , we mapped the relative abundance of each protein within the subcommunity structure layout to visualize the impact of APP deletion on the PPI network ( Fig 5A , APP-KO left , Fig 5B , NexCre-cDKO right ) . The visualization of functional units within this network allows distinguishing differently affected clusters such as vesicle organization ( 2 . 6 in Fig 4 ) and energy metabolism ( 3 . 2 in Fig 4 ) . Several members involved in energy metabolism are downregulated at the hippocampal PAZ in APP-KO mice; in contrast the majority is upregulated in NexCre-cDKO . The analysis of subcommunity structure was a prerequisite to define the impact of APP deletion on distinct functions including synaptic vesicle cycle , cytoskeletal organization and calcium homeostasis within the hippocampal PAZ proteome . For this purpose subnetworks were created containing APP and its family members APLP1 and APLP2 ( S1–S3 Figs ) . Remarkably APP appears as a highly connected node with 70 interactions within this network with a clustering coefficient of 0 . 0402 and a betweenness of 0 . 1198 indicating a central regulating function . In the synaptic vesicle cycle APP represents a hub that interacts with 22 proteins , in cytoskeleton organization with 18 proteins , and in calcium homeostasis with 20 proteins . The cluster coefficient that describes the densities of connections between neighbors in the synaptic vesicle cycle is 0 . 1125 with a betweenness of 0 . 4172 indicating a regulating role of APP within the network . Values for cytoskeleton organization were 0 . 0909 with a betweenness 0 . 2529 and for calcium homeostasis 0 . 0684 with a betweenness 0 . 7049 indicating that APP is not part of the cluster but a linker monitoring information flow . Subsequently , the impact of APP deletion was mapped onto these subnetworks ( APP-KO left , NexCre-cDKO right ) . The thickness of the connections represents the importance of the respective edges for information flow within the network ( edge betweenness ) . In the synaptic vesicle cycle ( Fig 6A ) –comprising members of the cytomatrix of the active zone , SNARE complex , G-proteins , glycolysis , endocytosis , and vATPase—several proteins were downregulated ( given as ratio APP-KO/control±variability in percent ) . These include α-synuclein ( 83 . 6±4 . 2% , Snca ) , syntaxin-1B ( 86 . 5±0 . 7% , Stx1b ) , ERC protein 2 ( 85 . 9±2 . 4% , Erc2 ) , guanine nucleotide-binding protein G ( I ) /G ( S ) /G ( O ) subunit gamma-2 ( 37 . 0±5 . 8% , Gng2 ) , clathrin light chain A ( 69 . 9±11 . 4% , Clta ) and B ( 80 . 4±2 . 3% , Cltb ) , AP-2 complex subunit sigma ( 84 . 9±19 . 1% , Ap2s1 ) , and v-type proton ATPase subunit d 1 ( 62 . 9±6 . 4% , Atp6v0d1 ) and only a few were upregulated ( given as ratio APP-KO/control±variability in percent ) such as putative tyrosine-protein phosphatase auxilin ( 111 . 7±4 . 2% , Dnajc6 ) and synaptophysin ( 124 . 7±2 . 5% , Syp ) while the majority remained unaltered in APP-KO mice ( Fig 6A ) . In contrast , hippocampal PAZ proteins derived from NexCre-cDKO were mainly downregulated ( given as ratio NexCre-cDKO/control±variability in percent ) ( Fig 6B ) . The mediator of SNARE complex formation α-synuclein ( 52 . 8±3 . 7% , Snca ) is severely downregulated whereas members of the SNARE complex the vesicle-associated membrane protein 2 VAMP2 ( 85 . 3±2 . 1% , Vamp2 ) , the synaptosomal-associated protein 25 SNAP25 ( 99 . 7±0 . 8% , Snap25 ) , syntaxin-1A ( 104 . 3±1 . 3% , Stx1a ) , syntaxin-1B ( 98 . 7±1 . 2% , Stx1b ) are less or not affected . In addition , downregulated proteins comprise guanine nucleotide-binding protein G ( I ) /G ( S ) /G ( O ) subunit gamma-2 ( 53 . 6±11 . 0% , Gng2 ) , guanine nucleotide-binding protein G ( I ) /G ( S ) /G ( T ) subunit beta-1 ( 87 . 0±5 . 5% , Gnb1 ) , members of the glycolysis cycle the L-lactate dehydrogenase A chain ( 87 . 6±2 . 2% , Ldha ) , fructose-bisphosphate aldolase A ( 87 . 5±1 . 6% , Aldoa ) , phosphoglycerate mutase 1 ( 87 . 8±2 . 5% , Pgam1 ) , clathrin light chain A ( 89 . 3±3 . 6% , Clta ) and B ( 81 . 0±1 . 4% , Cltb ) , the v-type proton ATPase 16 kDa proteolipid subunit ( 68 . 3±3 . 1% , Atp6v0c ) , and v-type proton ATPase subunit d 1 ( 81 . 3±2 . 5% , Atp5v0d1 ) ( Fig 6B ) . APP is strongly interconnected with proteins governing cytoskeletal organization ( S2 Fig ) . Important connections for the information flow within this network are via actin and the microtubule motor dynein . Several organizers of the cytoskeleton are affected by APP deletion ( Fig 7A ) such as coactosin-like protein ( 86 . 3±4 . 4% , Cotl1 ) , cofilin-1 ( 83 . 1±4 . 7% , Cfl1 ) , tropomyosin alpha-1 chain ( 80 . 6±2 . 5% , Tpm1 ) , and vimentin ( 113 . 9±2 . 6% , Vim ) . Especially in the NexCre-cDKO ( Fig 7B ) the MARCKS-related protein ( 74 . 3±7 . 1% , Marcksl1 ) that interacts with calmodulin , and actin is downregulated . Other affected proteins are cofilin-1 ( 83 . 5±2 . 2% , Cfl1 ) , tropomyosin alpha-1 chain ( 86 . 7±2 . 8% , Tpm1 ) , actin-related protein 2/3 complex subunit 4 ( 87 . 1±0 . 4% , Arpc4 ) and subunit 5 ( 86 . 1±1 . 0% , Arpc5 ) , and the src substrate cortactin ( 87 . 4±4 . 4% , Cttn ) . The protein network of calcium homeostasis reveals that APP is a central regulator ( S3 Fig ) that is interconnected with e . g . bassoon , neural cell adhesion molecule 1 , neural cell adhesion molecule L1 , neuromodulin , SNAP-25 , and neuroplastin . In the hippocampal PAZ derived from APP-KO mice neuromodulin ( 79 . 0±1 . 0% , Gap43 ) , the guanine nucleotide-binding protein G ( I ) /G ( S ) /G ( O ) subunit gamma-2 ( 37 . 0±5 . 8% , Gng2 ) , α-synuclein ( 83 . 6±4 . 2% , Snca ) , brain acid soluble protein 1 ( 80 . 1±4 . 3% , Basp1 ) , and calmodulin ( 30 . 9±17 . 8% , Calm ) are downregulated ( Fig 8A ) . Mapping the change in protein abundance derived from NexCre-cDKO demonstrates the downregulation of calmodulin ( 54 . 2±14 . 3% , Calm ) with its modulator neuromodulin ( 72 . 0±2 . 6% , Gap43 ) , brain acid soluble protein 1 ( 63 . 6±1 . 1% , Basp1 ) , myristoylated alanine-rich C-kinase substrate MARCKS ( 61 . 3±8 . 0% , Marcks ) , α-synuclein ( 52 . 8±3 . 7% , Snca ) , vesicle-associated membrane protein 2 VAMP2 ( 85 . 3±2 . 1% , Vamp2 ) , neuronal membrane glycoprotein M6-a ( 83 . 4±8 . 5% , Gpm6a ) , guanine nucleotide-binding protein G ( I ) /G ( S ) /G ( T ) subunit beta-1 ( 87 . 0±5 . 5% , Gnb1 ) , and the guanine nucleotide-binding protein G ( I ) /G ( S ) /G ( O ) subunit gamma-2 ( 53 . 6±11 . 0% , Gng2 ) . Neural cell adhesion molecule L1 ( 114 . 0±3 . 1% , L1cam ) and protein kinase C gamma type ( 112 . 6±1 . 1% , Prkcg ) reveal an increase in protein abundance ( Fig 8B ) .
The application of network analysis by e . g . centrality and path length , visualization techniques , and functional analysis by subcommunity structures allowed insights into the changes of protein abundance following APP deletion . APP is highly interconnected in all networks analyzed and serves as a hub , implicating that it is important for the structure and function of the entire network . The high value for betweenness centrality indicates a bridging role of APP between the functional modules . The low values for the clustering coefficient supports the linking function of APP in the interaction network of the hippocampal PAZ . APP and its interactors are central in the information flow as indicated by the high value for edge betweenness . For example , APP interacts with bassoon physically [38] and is thereby embedded into the cytomatrix of the active zone ( CAZ ) . Interactors of bassoon are in turn the ERC protein 2 , piccolo and regulating synaptic membrane exocytosis protein 1 . Deletion of APP leads to a decomposition and rearrangement of the entire network structure . This network is based on current database knowledge about physical interactions between proteins . Therefore APP deletion results in loss of network components . Interestingly , this is not reflected by a loss of biological functions . For example following APP deletion bassoon and its interactors are still part of the synaptic vesicle cycle . However they are no longer connected in cytoskeleton organization and calcium homeostasis subnetworks . In addition reorganization of the network results in changes in path lengths . For example in the synaptic vesicle cycle the vesicular proton pump vATPase and bassoon are linked via APP . Deletion of APP results in an increase in the shortest path length by three steps ( interactions ) . Similarly , in calcium homeostasis loss of APP extends the shortest path length from extracellular cell adhesion ( e . g . NCAM1 ) to trimeric G-proteins involved in intracellular signaling by three interactions . In contrast deletion of APP adds three additional interactions ( 4→7 ) from the homophilic cell-cell adhesion molecule 1 , that acts in a calcium independent manner , to the actin-related protein complex . Based on our data , we hypothesize that APP functions as a regulator involved in all functional activities at the release sites . Depending on its respective microenvironment it may exert varying functions . This is illustrated in a scheme ( Fig 9 ) which depicts the physical interaction between APP ( regulator , R ) , mediators ( M ) and central players ( C ) . Deletion of APP results in significant changes in the abundance of mediators but not of central players . For example the mediator of synaptic vesicle exocytosis α-synuclein [39] is downregulated in the mutants , however the central players the SNARE proteins VAMP2 , SNAP25 , and syntaxin1 [13] are not affected . In this context it is noteworthy to mention that deletion of SV2A does not affect the abundance of SNARE proteins but results in reduced SNARE complex formation [40] . Similarly , the mediators in calcium homeostasis calmodulin and neuromodulin are downregulated whereas the central player in learning and memory CaMKII as well as calcium-channels remain unaltered . Our data support the role of APP as a context-sensitive regulator of the hippocampal PAZ . This is in line with the current knowledge of the physiological function of APP in the CNS . Recently , APP has been suggested as a G-protein coupled receptor ( GPCR ) mediating presynaptic signaling and neurotransmitter release by activation of calcium-channels [41] . A selective interaction of APP with the neural cell adhesion molecule NCAM has been reported [42] . NCAM-induced activation of calcium-channels is a downstream effect of signaling through heterotrimeric G-proteins [43] . Our PPI network displays APP interaction via Munc18 with calcium/calmodulin activated Ser-Thr kinase CASK that is recruited to the cytosolic tail of neurexin-1 [44] . CASK may recruit calcium-channels to the presynapse [45] providing strong support that the presynaptic function of APP is mediated through similar mechanisms [46] . The APP-Mint1-Cask complex has also been implicated in neurexin-mediated signaling in presynaptic organization [47 , 48] . This complex is thought to play a regulatory role in synaptic vesicle release , and may be involved in development and maintenance of synaptic architecture [49] . The physical interaction of APP with bassoon [38] regulates the recruitment of ERC protein 2 ( ELKS ) , RIM , and Munc13 that are involved in the recruitment of Ca2+-channels [13] . The regulatory impact of APP on the abundance of Ca2+-channels at the presynaptic plasma membrane [50] points to a functional integration of APP within this conserved network of the active zone key players . Furthermore , our data suggest , that APP is physically implemented into synaptic vesicles exocytosis . The link is provided by α-synuclein an essential mediator of SNARE-complex formation [39] . How APP and α-synuclein are connected to each other is still elusive and needs to be further investigated . However , embedding APP as a context-sensitive regulator into the active zone network supports the notion of a physiological role in synaptic transmission and plasticity . Taken together , our findings provide a new perspective of the functional integration of APP into the hippocampal PAZ proteome . Furthermore , our network analysis incorporates APP into the evolutionary conserved active zone protein complex , comprising ELKS , CASK , as well as bassoon , Rim and Munc18 . The coordinated arrangement of these proteins mediating synaptic vesicle docking and priming is essential for neuronal signaling . Interestingly , this subnetwork persists after APP deletion maintaining functional neurotransmission . We could demonstrate that deletion of APP differently affects individual PAZ subcommunities . Our results suggest , APP acts as a context-sensitive regulator within the presynaptic proteome linked to neuronal communication . Future studies have to reveal how APP deletion affects these subcommunities in a spatiotemporal way that eventually lead to impairments in learning and memory . These alterations may provide a molecular link to the pathogenesis of Alzheimer’s disease and open new strategies for therapeutic approaches . | More than 20 years ago , the amyloid precursor protein ( APP ) was identified as the precursor protein of the Aβ peptide , the main component of senile plaques in brains affected by Alzheimer’s disease . However , little is known about the physiological function of amyloid precursor protein . Allocating APP to the proteome of the structurally and functionally dynamic presynaptic active zone highlights APP as a hitherto unknown player within the presynaptic network . The hippocampus is the most prominent brain region for learning and memory consolidation , and a vulnerable target for neurodegenerative disease , e . g . Alzheimer’s disease . Therefore , our experimental design is focused on the hippocampal neurotransmitter release site . Currently , the underlying mechanism of how APP acts within presynaptic networks is still elusive . Within the scope of this research article , we constructed a network of APP within the presynaptic active zone and how deletion of APP affects these individual networks . We combine bioinformatics tools and biochemical approaches to address the dataset provided by proteomics . Furthermore , we could unravel that APP executes regulatory functions within the synaptic vesicle cycle , cytoskeletal rearrangements and Ca2+-homeostasis . Taken together , our findings offer a new perspective on the physiological function of APP in the central nervous system and may provide a molecular link to the pathogenesis of Alzheimer’s disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"vesicles",
"nervous",
"system",
"neurodegenerative",
"diseases",
"protein",
"interaction",
"networks",
"electrophysiology",
"neuroscience",
"membrane",
"proteins",
"synaptic",
"vesicles",
"network",
"analysis",
"alzheimer",
"disease... | 2016 | APP Is a Context-Sensitive Regulator of the Hippocampal Presynaptic Active Zone |
Assembly and disassembly of viral capsids are essential steps in the viral life cycle . Studies on their kinetics are mostly performed in vitro , allowing application of biochemical , biophysical and visualizing techniques . In vivo kinetics are poorly understood and the transferability of the in vitro models to the cellular environment remains speculative . We analyzed capsid disassembly of the hepatitis B virus in digitonin-permeabilized cells which support nuclear capsid entry and subsequent genome release . Using gradient centrifugation , size exclusion chromatography and immune fluorescence microscopy of digitonin-permeabilized cells , we showed that capsids open and close reversibly . In the absence of RNA , capsid re-assembly slows down; the capsids remain disintegrated and enter the nucleus as protein dimers or irregular polymers . Upon the presence of cellular RNA , capsids re-assemble in the nucleus . We conclude that reversible genome release from hepatitis B virus capsids is a unique strategy different from that of other viruses , which employs irreversible capsid destruction for genome release . The results allowed us to propose a model of HBV genome release in which the unique environment of the nuclear pore favors HBV capsid disassembly reaction , while both cytoplasm and nucleus favor capsid assembly .
Viral capsids facilitate multiple functions in the viral life cycle . Outside the cell , they protect the enclosed viral genome against nucleases , and in case of non-enveloped viruses they mediate attachment and entry . For both enveloped and non-enveloped viruses , they carry the viral genome to the site of replication where they have to release the genome in order to allow access of transcription and/or replication factors . After replication new capsids have to be assembled for encapsidation of the progeny genomes and subsequent release of mature virions . Capsids are assigned to be metastable: early in infection they have to open , later they have to assemble and close . Most data on stability of capsids and kinetics of their formation and dissociation are obtained in vitro allowing analysis by biophysical or electron microscopical techniques ( e . g . [1]–[5] ) In vivo data on capsid disintegration are rare despite of their importance for genome release and viability of infection . Two different examples are Adenovirus 5 ( Ad-5 ) capsids and the capsid of Herpes simplex virus-1 ( HSV-1 ) . Ad-5 capsids disintegrate to penton and hexon subunits after modification upon endocytosis [6] , [7] for genome release . Capsids of HSV-1 in contrast remain stable after release of one penton [8] which allows the injection of the viral DNA from that opening [9] . Intensive studies on capsid assembly were performed in vitro on capsids of the medically important Hepatitis B virus ( HBV ) [1] , [5] , [10] , [11] . HBV infection is endemic in large parts of the world and ∼350 Mio people are chronically infected , accounting for 1 million deaths per year . HBV is an enveloped virus with an icosahedral capsid that is composed of 240 or 180 copies of a single protein species called core protein [12] . Within the oxidizing environment outside the cell the two core protein subunits of a dimer become linked by three disulfide bonds ( Cys 48 , 61 and 183 [13] , [14] ) . The capsid encloses the relaxed circular viral DNA ( rcDNA ) [15] , which is covalently attached to the viral polymerase [16] . HBV cell entry is the limiting step that prevents infection of most cell cultures but it can be by-passed by lipofection of hepatoma cells with virion-derived capsids . Using this artificial mode of capsid entry , HBV production reaches in vivo-like efficiency [17] . This suggests that the capsids either do not become modified during natural entry or that lipofection changes their structure in the same way . Complex interactions with the cellular transport machinery mediate capsid translocation to the nuclear periphery , passage through the NPC and liberation of the viral DNA [17]–[19] . Transport and genome release must be highly efficient and well-coordinated , because ∼80% of virions are infection-competent in vivo [20] . Within the nucleus , viral DNA is converted by cellular repair mechanisms to a covalently closed circular form ( cccDNA ) , which is the template for viral mRNA synthesis including the RNA pregenome . Interaction of the RNA pregenome with the viral polymerase facilitates encapsidation into the viral capsid [21] . The polymerase retrotranscribes RNA pregenome into the rcDNA , but this occurs only within the capsids . This genome maturation requires multiple phosphorylation steps within the C terminus of the core protein [22] , [23] . Mature capsids ( Mat-C ) can either be enveloped by the viral surface proteins in order to form virions or they can be transported through the NPC causing amplification of nuclear viral DNA . Liver histology of HBV-infected individuals shows massive intranuclear capsid accumulation . However , the number of intranuclear viral genomes is low . Thus , it is believed that unassembled core protein pass the NPC and assembles intranuclear [24] . HBV capsids exhibit predominantly a T = 4 symmetry with a diameter of ∼36 nm . Core protein assembly is independent of eukaryotic host proteins as it occurs also upon core protein expression in E . coli resulting in “recombinant” capsids ( rC ) . In contrast to natural capsids , rC are unphosphorylated and contain E . coli RNA [25] . In addition , they exhibit one unnatural disulfide bond linking the C terminal Cys ( C183 ) of one core protein with a C183 of a neighboring dimer [26] . Structure of the first 143 aa of core proteins in rC has been obtained by X ray crystallography with a resolution of 3 . 3 Å [27] . The core monomer comprises two long α-helices with a hairpin structure . Association of two hairpins from two monomers forms a spike which protrudes from the particle surface . The connecting loop is exposed on the spike tip and comprises the immune dominant c/e1 B cell epitope , so that most antibodies are conformation-dependent . While the first 143 aa of the core protein are well structured [12] , the C terminus is flexible: whereas the C terminus is localized within capsid lumen in E . coli-expressed capsids [28] , it is exposed to the exterior in viral mature capsids [19] . The latter observation may however indicate that capsids dynamics increase with genome maturation . In vitro association kinetics , performed on E . coli-expressed , C terminally truncated core proteins ( aa 149 ) , showed that capsid formation starts with core protein dimers . It is thought that the dimers trimerize and the resulting hexamer nucleates capsid assembly without accumulating further distinct populations of capsid subassemblies [10] , [29] . According to the laws of thermodynamics , disassembly could just be the inverse reaction because the attractive forces between the subunits are weak , allowing them to transiently dissociate and re-associate ( capsid breathing ) [1] , [5] in a way similar to that observed for polio- , flockhouse- and rhinoviruses [2]–[4] . In fact , recent in vitro evaluations showed that chaotropic agents as urea cause disassembly down to core protein dimers without distinguishable capsid subassemblies [30] . Several differences of these in vitro studies with the in vivo situation deserve attention: the C terminus , which interacts with encapsidated nucleic acids [31] and comprises the phosphorylation sites , was deleted in these studies; the capsids contained neither RNA nor DNA nor the polymerase . Moreover , host factors explaining the highly time- and site-coordinated HBV genome release were not present . Accounting for the poor knowledge on in vivo disassembly and the medical importance of HBV , we evaluated the fate of HBV capsids within the cell . As no efficient infection system exists , we used digitonin-permeabilized cells that promote genome liberation into the nucleus . In order to distinguish reliably between input capsids and products of disassembly and re-assembly , we determined antibody binding properties , density , and size of authentic and recombinant capsids , and compared these particles with the products of nuclear import .
We characterized 32P labelled mature DNA-filled capsids from cell culture and RNA-filled capsids from E . coli as reagents . These capsids exhibited different densities upon Nycodenz gradient centrifugation . DNA-filled capsids undergo a transition to a density that resembled the RNA-filled state that correlates with nuclear entry . We demonstrated that E . coli expressed capsids dissociate and reassociate . Using the assembly/disassembly states we characterized two anti core protein antibodies and found different specificities for assembly states . These characteristics allowed us to analyse the intracellular fates of DNA-filled capsids by immune fluorescence microscopy . Phosphorylation of both HBV virion-derived ( Mat-C ) and E . coli-derived ( rC ) capsid species resulted in specific labelling of the core proteins ( Fig . 1A ) . In order to exclude that contaminant proteins accounted for the radiolabelled 21 . 5 kDa band we performed an immune precipitation using a polyclonal anti HBV capsid antibody ( DAKO Ab ) . This antibody does not react with denatured core protein or core protein aggregates generated by acidification [18] . Fig . 1B showed that the DAKO Ab completely precipitated the 21 . 5 kDa band . This finding confirmed the identity of the band as core protein and showed that no other 32P-labelled protein co-migrated . It further indicated that all core proteins exhibited their proper conformation after phosphorylation . To analyze whether the core proteins were assembled to capsids we performed a native agarose gel electrophoresis of ( Fig . 1C ) . No disintegrated capsids or aggregated core proteins exhibiting slower and diffuse migration were observed [18] . Phosphorylated rC ( P-rC ) caused a minor , slower migrating band . The additional band is characteristic for E . coli-expressed capsids . It is assumed that the minor band represent two capsids linked by RNA , as RNase A-treatment reduces the slower migrating band . The presence of this band thus indicates that the trapped RNA was not degraded upon in vitro phosphorylation . Both capsid species reacted with the DAKO Ab confirming their exposure of core protein epitopes . We next analyzed possible contaminations with nucleases in the capsid preparations . We used 32P labelled dsDNA , ssDNA and RNA and determined the hydrolysis after 2 h incubation at 37°C ( Fig . 1D ) . We observed reductions of TCA-precipitable 32P up to 10 . 2% . As the pipetting error in this assay was determined to be 4% we assume that the nuclease activities in the capsid preparations were not significant . The 32P capsid preparations were analyzed by Nycodenz gradient centrifugation . Nycodenz was chosen as it conserves protein interactions better than any other gradient media , allowing the recovery of functionally active protein complexes [32] . In our hands , Nycodenz allowed the recovery of 95% of E . coli-expressed capsids loaded , while CsCl , which is known to allow separation of genome-containing and empty capsids - caused significant capsid disintegration to 10% ( data not shown ) . As Nycodenz has properties similar to sucrose that allows a separation based on the sedimentation coefficient . In contrast to sucrose however , Nycodenz allows capsids to reach their equilibrium . In order to determine sedimentation of unassembled core proteins , we analyzed urea disintegrated capsids by centrifugation . Due to the unphysiological disulfide bonds of the C terminal Cys upon expression in E . coli , the assay was performed using the C183S P-rC mutant . Sedimenting the capsids on Nycodenz gradient resulted in a peak of P-rC from 1 . 257 to 1 . 283 g/ml , with a maximum at 1 . 283 g/ml ( Fig . 2A , peak ( 1 ) ) . The three major fractions contained 91% of the radioactivity . P-Mat-C banded at a slightly lower density between 1 . 242 and 1 . 284 g/ml with a maximum at 1 . 263 g/ml ( Fig . 2B , peak ( 2 ) ) . The four major fractions contained 98% of the radioactivity indicating that capsids had not undergone significant degradation during centrifugation . Urea dissociated core proteins remained close to the top of the gradient exhibiting a peak at 1 . 156 g/ml ( Fig . 2C , peak ( 3 ) ) . Likely , these core proteins represent dimers as it was reported upon urea disintegration by others [30] . Despite the fact that rcDNA has a ca . 1 . 8 fold higher molecular mass than RNA , the density of P-Mat-C was slightly lower . In Nycodenz however , RNA has a higher density ( 1 . 18 g/ml ) than dsDNA ( 1 . 13 g/ml ) while proteins have 1 . 31 g/ml in Nycodenz [32] ) . Therefore , the different density distribution of Mat-C and rC implies that the gradient medium entered capsid lumen . To test this hypothesis we searched for Nycodenz entry by electron microscopy . Nycodenz , an electron-rich solute , can be seen without addition stain . Fig . 2D depicts that Nycodenz caused 20–24 nm large dots which were absent in the negative control . The size of the stain was similar to the lumen of P-rC stained by phospho tungstic acid , which exhibited the external diameter of 35 nm known for HBV capsids . We used rC capsids in order to separate capsids from core protein dimers and capsid subassemblies . Separation was performed by Superdex 75 size exclusion chromatography using C183S rC capsids . Separating a stock solution of C183S-rC revealed a single peak in fractions A4 ( 0 . 85 ml ) ( Fig . 3A ) . Its appearance in the exclusion volume of the column implies that practically all core proteins were assembled to capsids . However , when a 20 fold lower core protein amount was applied three peaks occurred ( Fig . 3B ) : the strongest peak ( peak A; 51% of total protein ) appeared in fraction A3/4 ( 0 . 85 ml ) , which is within the exclusion limit of the column . A small peak ( B , ∼3% ) appeared in fraction B9 ( 1 . 73 ml ) and another stronger peak ( peak C , 46% ) was observed in fraction C1/2 ( 1 . 92 ml ) . Re-application of peak A after 1 h at RT resulted in the re-appearance of peaks A , B and C ( 69% peak A , 10% peak B and 23% peak C; Fig . 3C ) . Re-injection of peak C revealed the same peaks but with a different distribution ( A 11% , B 10% , C 79% ) ( Fig . 3D ) . This finding confirms that rC undergoes dissociation , as described previously [1] , [5] . For analyzing the immune reactivity of the fractions , we used two anti capsid antibodies: ( 1 ) polyclonal DAKO Ab reacts with entire wt capsids but only weakly with denaturated core proteins ( <1% [18] ) , ( 2 ) monoclonal FAb3105 which was shown previously to bind to an epitope on core protein dimers involving the immune dominant loop ( aa 77–80 and 83–84 [33] ) . Determination of the immune reactivity of peaks A , B and C by dot blot is shown on Fig . 4A . The graph shows the reactivity , normalized to the protein amounts . FAb3105 reacted similarly well with all fractions giving evidence that they contained core proteins which were at least assembled to dimers . The polyclonal DAKO Ab reacted best with fraction B and somewhat less with fraction A containing the capsid . It must be considered however that the capsid peak contains the encapsidated heterogeneous E . coli RNA , which leads to an overestimation of the core protein amount . Reactivity with fraction C was faint . Given that peak C contained core proteins already assembled to dimers we conclude that the DAKO Ab recognizes preferably epitopes that are formed on capsid subassemblies with a higher complexity . The interference of spike insertions with DAKO Ab binding implies binding at or close to the spikes , similar to FAb3105 . We thus analyzed the antibody competition for their epitopes on rC . Fig . 4B shows that preincubation with DAKO Ab completely inhibited FAb3105 binding , suggesting overlapping binding sites . Saturation with FAb3105 prior to DAKO Ab incubation reduced DAKO Ab binding but not entirely , potentially due to its polyclonality . In order to analyze the fate of Mat-C upon nuclear entry we analyzed the sedimentation in Nycodenz after subjecting P-Mat-C to nuclear import . Import reaction was performed using digitonin-permeabilized cells , which is a well-established system for analysis of nucleo-cytoplasmic traveling and HBV genome release at the nuclear envelope [19] , [34] . A control reaction was performed by addition of WGA , which blocks active nuclear import by nuclear transport receptors [35] . Following import reaction , nuclei were lysed by the same non-ionic detergents used for capsid purification of Mat-C from secreted virus . A similar protocol was chosen to exclude that nuclear lysis has an impact on the subsequent capsid analysis on Nycodenz gradients . Upon WGA inhibition , 97% of P-core proteins of the P-Mat-C migrated as purified P-Mat-C without cell exposure , exhibiting a peak between 1 . 283 and 1 , 252 g/ml ( Fig . 5A , peak ( 2 ) ) with a maximum at 1 . 267 g/ml . It has been shown that P-Mat-C from transport reaction is attached to the nuclear import receptors Importin α and β [18] . Thus , one might expect a higher density . However , precipitating with the DAKO Ab capsids that were preincubated with the cytosolic extract showed , by immune blot , that the number of coprecipitated import receptors is rather low ( 4 molecules per capsid , not shown ) . This accounts for an undetectable increase of protein mass of <13% only . Core proteins derived from P-Mat-C following nuclear import peaked between 1 . 304 and 1 . 250 g/ml with a maximum at 1 . 285 g/ml ( Fig . 5B , peak ( 1 ) ) . This maximum was identical to the one of the purified RNA-containing P-rC ( 1 . 283 g/ml ) . As 99% of 32P core proteins were found within these fractions we concluded that all core proteins were assembled into particles without exhibiting significant amounts of dimers or subassemblies . A summary of all capsid sedimentation profiles is given in Fig . 5C . The requirement for an active nuclear import suggested however that the transition from “light” to “heavy” capsids occurred inside the nucleus . The changed density of the DNA-filled P-Mat-C to the density of RNA-filled E . coli-expressed capsids upon nuclear transport implied that the DNA genome in Mat-C was replaced by RNA . Such a replacement likely involves at least partial capsid disintegration , although transitory capsid subassemblies could not be detected in our assay . Reduction of temperature or modification of salt concentration - successfully used in biophysical assays [1] - could not be applied as physiological nuclear import is temperature and salt dependent . We used an alternative approach based on the observation that HBV capsid assembly rate is enhanced by core protein RNA-interaction [36] . We treated digitonin-permeabilized cells by RNase A , which is a small 13 . 7 kDa protein that is far below the threshold of diffusion into the nucleus ( up to 68 kDa [37] ) . RNA degradation was monitored by ethidium bromide staining after agarose gel electrophoresis of the cell lysate ( Fig . 6A ) . Specificity of degradation was shown by control digestions using DNase I ( 37 kDa ) . DNase I treatment however causes collapse of the nuclear structure with diffuse distribution of the nuclear pores ( not shown ) . RNase A-treated cells were used for nuclear import of P-Mat-C . Applying the nuclear lysate onto Nycodenz gradient showed two peaks , one with densities from 1 . 241 to 1 . 300 g/ml , with a maximum at 1 . 285 g/ml ( Fig . 6B , peak ( 1a ) ) and a second with densities from 1 . 141 to 1 . 185 g/ml , with a maximum at 1 . 156 g/ml ( Fig . 6B , peak ( 1b ) ) . Quantification showed that 47% of the 32P core protein migrated in the first peak while 53% were found in the lighter fractions . The heavier peak showed the same peak density of P-Mat-C after nuclear import but with a broader distribution . The lighter one showed the same distribution as urea treated core proteins . A control was performed by adding WGA during the import reaction . Here , core proteins migrated with the density of P-Mat-C , ranging from 1 . 242 to 1 . 282 g/ml ( maximum 1 . 266 g/ml; Fig . 4B , peak ( 2 ) ) , showing that RNase A has no impact on P-Mat-C density prior to nuclear import . To confirm the presence of capsids , dimers and capsid subassemblies , we analyzed capsids derived from nuclear import by size exclusion chromatography . Fig . 7A ( upper panel ) shows the elution profile at OD280 . The peaks were derived from cellular proteins , as the same profiles were obtained from permeabilized cells without P-Mat-C and from RNase A-treated cells . We analyzed the presence of 32P capsids by native agarose gel electrophoresis ( Fig . 7A , lower panels ) . In lysate from cells not treated with RNase A , 32P signals were obtained in fractions A3–5 migrating as integer rC capsids . 32P-core proteins from RNase A-treated cells , in contrast , were observed in fractions A5–7 and exhibited a migration slower and more diffuse than rC . No core proteins were observed in fractions B9 and C2 in which dimers and more complex capsid subassemblies were found . It must be considered however , that in native agarose gel electrophoresis unassembled core proteins migrate diffusely . Given that the signals were already at the detection limit , diffuse bands could have caused false negative results . We thus analyzed the fractions derived from the RNase A-treated cells by dot blot ( Fig . 7B , upper row ) , revealing 32P label in fractions C1/2 by phosphoimaging . Immune reaction of the dot blot by FAb3105 confirmed the presence of core proteins in all the 32P-positive fractions ( Fig . 7B , middle panels ) . These fractions however were not reactive with DAKO Ab , which showed a faint signal only in fraction A8 . As it overlaps with the peaks of cellular proteins , we assume that it results from unspecific interactions , which is in accordance with the absence of any FAb3105 or 32P label in this fraction . To confirm the immune reactivity of the capsids , dimers and other capsid subassemblies , we investigated the light and the heavy fractions derived from Nycodenz gradient centrifugation by immune precipitation . Heavy fractions ( 2–5 ) from P-Mat-C exposed to untreated cells ( 1 . 250–1 . 304 g/ml ) and heavy fractions 3–5 , derived from RNase A-treated cells in which nuclear import was blocked by WGA ( 1 . 242 to 1 . 282 g/ml ) served as controls . From the RNase A-treated sample , we analyzed the heavy fractions 2–5 ( 1 . 241 to 1 . 300 g/ml ) and the light fractions of the same gradient ( 1 . 141–1 . 185 g/ml ) . Fig . 8 depicts that FAb3105 precipitated the 32P-core proteins in all heavy and light fractions from all gradients with similar efficiency . In contrast , DAKO Ab precipitated 32P-core proteins only from heavy fractions of P-Mat-C from untreated cells and from RNase A-treated cells in which nuclear import was blocked by WGA . No precipitate was found in the heavy and light fractions from RNase A-treated cells in which nuclear import occurred . Gradient analyses and size exclusion chromatography indicated that P-Mat-C is subject to an RNA-dependent intranuclear reassembly process upon nuclear import . For confirmation we analyzed the intranuclear appearance of capsids and their subassemblies in individual cells . Due to the lack of a suitable infection system we used Digitonin-permeabilized cells , which are a wide-spread technique for studying nuclear import including HBV capsids [18] , [19] , [34] . Digitonin permeabilizes the plasma membrane leaving the nuclear and ER membrane integer . After addition of exogenous cargos and the addition of nuclear import receptors the fate of the cargo can be analyzed using microscopy . We added Mat-C and rabbit reticulocyte lysate , which is common source of transport receptors [34] . The localization of capsids and capsid subassemblies was determined by confocal laser scan microscopy using indirect immune fluorescence . A control stain was performed with propidium iodide ( PI ) in order to depict cell nuclei . Mat-C were added to permeabilized cells that were either untreated or RNase A-treated . Controls were performed by inhibiting nuclear import using WGA and by using cells to which no capsids were added . Fig . 9 shows that both DAKO Ab and FAb3105 exhibited nuclear fluorescence after addition of Mat-C to permeabilized cells which had not been treated with RNase A . This staining pattern is in accordance with data reported previously in permeabilized and integer cells [17] , [19] . It is also in agreement with the typical findings in liver histology [38] . The signals of both antibodies were specific , as no fluorescences were observed in cells to which no Mat-C was added , or in which nuclear import was blocked by WGA [18] . In RNase A-treated cells however , no immune fluorescence could be observed using DAKO Ab , but strong signals were obtained using FAb3105 . These findings are in accordance with the observation that capsid subassemblies present in the lysates from RNase A-treated cells were not recognized by DAKO Ab but recognized by FAb3105 . FAb3105 staining in RNase A-treated cells was significantly stronger than in cells untreated with RNase A . As both antibodies were added together this observation is in agreement with the competition for binding to capsids .
Several studies describe the in vitro assembly of viral capsids , but there are only few investigations targeting their disassembly or their intracellular fate . In vivo investigations can hardly yield capsids in amounts suitable for biochemical analysis . We thus used two anti capsid antibodies and characterized at first their binding specificity for capsid subassemblies . Both antibodies , raised against entire capsids , are conformation-dependent . We used the well-studied FAb3105 in order to comparatively characterize the DAKO Ab for its reactivity against different capsid subassemblies , obtained by separation on size exclusion columns . Comparison was required , as the calibration of the column with globular proteins ( see Material and Methods ) showed that the migration of the subassemblies did not correspond to the MW of a single core protein ( 21 . 5 kDa ) neither to a core protein multimer , so that a form-dependent retardation of the core proteins or unspecific interactions with the matrix were assumed . Both subassemblies ( B , C ) reacted with the FAb3105 and thus represented dimers or larger multimeric association of core proteins . Based on the known kinetics of HBV capsid assembly , we assume that peak C exhibits core protein dimers and peak B corresponds to a larger assembly state . Our experiments do not allow drawing a conclusion on how many dimers form this complex . According to the literature trimers of the dimers could be present in peak B , as this assembly state was shown to be the only distinct capsid subassembly population apart of dimers and capsids [10] . The DAKO Ab reacted with the capsid subassembly in peak B but poorly with the dimer fraction ( C ) . In fact , the faint signal obtained after blotting of this fraction can be well explained by the limited formation of capsid subassemblies larger than dimers occurring between harvest from the column and blotting . We assume that the DAKO Ab either reacts with epitopes at the interface of dimers or , alternatively , the larger subassemblies may exhibit conformational differences compared to free dimers , as it was suggested previously [1] . The antibodies competed for their substrate implying that they both attach at , or near , the capsid spikes . This was not surprising , as the loop comprises the immune dominant c/e1 epitope of the capsids . In summary , the antibody characterizations indicate that DAKO Ab requires core protein assemblies larger than dimers , and that both antibodies exhibited sterical interference . Considering the kinetics of in vitro assembly [10] , it must be assumed that the DAKO Ab reacts with core protein hexamers and fully assembled capsids . Anti capsid antibodies do not detect small amounts of core protein monomers . For following the migration of all states of capsids assembly on gradient we depended on sensitive and quantitative detection . We took advantage of capsid phosphorylation , which allows labelling of the core proteins by radioactivity . Phosphorylation of the capsids did not significantly affect capsid structure , as indicated by their unaltered migration on native agarose gel electrophoresis and their unchanged reactivity with the DAKO Ab ( not shown ) . This result was expected as the number of transferred phosphates was low . In addition , data from others have shown that E . coli-expressed capsids are identical to liver-expressed capsids within a resolution limit of 30 Å [39] . Recently these data were confirmed with better resolution of 16 Å further showing that no gross structural changes are linked with genome maturation and envelopment [40] . Better resolution with 10 Å however showed that a hydrophobic pocket is present on DNA-containing capsids [41] . As these differences do not affect capsids size , the different densities of Mat-C and rC in Nycodenz gradients suggest that the nucleic acid content caused the difference . This assumption is in accordance with the reported densities of RNA and dsDNA in Nycodenz [32] . Entry of Nycodenz in the capsid lumen was proven by electron microscopy , and could have occurred either via the 1 . 5 nm-measuring holes in the capsid shell [12] , or by capsid breathing . For analyses of the intracellular fate of HBV capsids , we investigated the dissociation reaction of HBV capsids using digitonin-permeabilized cells . Addition of P-Mat-C to permeabilized cells caused a change of migration towards the density of RNA containing capsids , but only upon active nuclear import . Although the differences appear to be minor , the superposition of the different gradients allows a clear differentiation of the sedimentation profiles ( Fig . 3B ) . Strikingly , P-Mat-C showed a single density peak after nuclear import , implying that virtually all capsids were converted , and that no significant subpopulations failed to deliver their encapsidated DNA into the nucleus . This high efficiency in our system suggests that it may reflect the situation in infected individuals where up to 80% of all HBV particles are infectious [20] . Genome release and subsequent replacement of the genome by RNA could hardly be explained by passage through holes in the threefold or quasi-threefold axis of the capsid shell [12] . It can be concluded from structural data on HBV capsids that even dissociation of one core protein hexamer upon capsid breathing would cause holes of only 4 . 3 – 5 . 7 nm . Further dissociation of the capsid is probably necessary for genome release . The required incubation period of the transport assay for obtaining detectable nuclear import ( 15 min ) was however much longer than the short concentration-dependent association times of >50 s observed in vitro [29] . To observe the disassembly of capsids into subassemblies within nuclei we decelerated reassociation by removal of RNA . Both sedimentation and size exclusion chromatography showed two reaction products . The smaller product corresponded in size to urea treated C183S P-rC by sedimentation , and to peak C of the chromatography , thus likely representing core protein dimers . This assumption was confirmed by their reactivity with FAb3105 and their failure to bind to the DAKO Ab following both methods of separation . The second product migrated similarly to RNA filled capsids , but exhibited a slightly broader peak . Size chromatography confirmed a similar but nevertheless different migration than capsids , as the peak was shifted by two fractions . Further confirmation was obtained by native agarose gel electrophoresis , after chromatography , showing a diffuse and retarded migration compared to capsids . Such a migration was reported for P-rC upon acid denaturation [18] causing protein aggregation . Immune blotting confirmed the absence of DAKO Ab reactivity arguing against capsid or core protein hexamer formation similar to that observed after acid denaturation . The remaining reactivity with FAb3105 indicated however core protein dimer formation , so that we conclude that the oligomeric reaction product consists of core protein dimers which were assembled in a misfolded manner . In fact , misdirected folding under preservation of intact dimer formation was observed recently upon addition of the HBV capsid assembly inhibitor HAP1 [42] . To confirm the intranuclear localization of the dissociation and reassociation events , we analyzed capsids and their subassemblies by indirect immune fluorescence in permeabilized cells . In accordance with gradient centrifugation and size exclusion chromatography , we observed that RNase A-treatment generated intranuclear subassemblies , which were reacting with FAb3105 but not with the DAKO Ab . Despite the absence of capsids , the nuclear presence of core protein dimers indicates that RNase A treatment did not interfere with nuclear import of core protein . This supports the conclusion that HBV capsids become imported into the nuclear basket as entire particles [43] but disassemble in the nuclear environment . In the presence of RNA , in contrast , DAKO Ab and FAb3105-reactive assembly forms of the core protein appeared . Considering the absence of detectable capsid subassemblies between complete capsids and core protein dimers upon in vitro disassembly [30] , this finding implies that the capsids disintegrate to core protein dimers followed by a rapid RNA-dependent reassociation to capsids , which is misdirected in the absence of RNA . As we used unphosphorylated Mat-C in this assay , these results further exclude that phosphorylation of Mat-C has had a significant impact on formation of capsids and on the capsid subassemblies . In the present paper we showed that the dissociation of the HBV capsid follows the in vitro association reaction in inverse direction . Our observation that core proteins reassemble to capsids inside the nucleus implies that both compartments - the cytoplasm in which initial capsid formation occurs - and the nucleus , support the assembly reaction . The environment in which disassembly occurs should be consequently before capsid entry into the free karyoplasm . A potential candidate compartment would be the nuclear basket at the karyoplasmic side of the NPC , as it provides the unique proteins of the NPC . We thus hypothesize that the capsids assemble from core proteins via dimer and hexamer formation , as it was proposed recently [10] , [29] ( Fig . 10 ) . The time at which the polymerase-RNA pregenome complex interacts with the capsid subassemblies has to be left open but may enhance the assembly reaction . During genome maturation and transport , the capsid is subject to capsid breathing and remains stable [17] , [19] , [43] . Apparently , the basket of the nuclear pore comprises the environment promoting capsid disintegration and genome liberation , which occurs after arrest of the capsid [19] , [43] . We assume that core protein dimers derived from disassembled capsids are however able to diffuse deeper into the nucleus . When the threshold concentration of nuclear core protein dimers [36] is reached , capsid formation would occur , which is probably enhanced by cellular RNA . As cytoplasmic capsids , they undergo breathing but remain stable , explaining the capsid accumulation observed in liver biopsies .
Mat-C were prepared and purified from virions of the permanently Hepatitis B virion-expressing hepatoma cell line HepG2 . 2 . 15 [44] accordingly to Rabe et al [19] . The capsids of these virions , which were shown to be infectious in chimpanzees [45] , contain viral DNA in a relaxed circular form . [44] . E . coli-derived capsids ( rC ) and a mutant in which the C terminal Cys was replaced by Ser ( C183S rC ) were expressed and prepared as described previously [12] . Electron microscopy did not show any difference between wt capsids and this mutant ( not shown ) . While Mat-C showed strong contamination of proteins of cell culture medium ( approx . 30 fold ) both E . coli-expressed capsid preparations exhibited high purity of >90% with respect to the total protein: SYPRO staining after SDS PAGE showed a single band of 21 . 5 kDa when 400 ng were loaded . Disintegration of capsids was achieved by treating C183S rC by 4 M urea for 10 min at 42°C . All capsid preparations were analyzed for nuclease contaminations by incubating 50 ng capsids with 32P labelled nucleic acids in 50 µl transport buffer for 2 hours at 37°C in siliconized Eppendorf tubes and determining the amount of radioactive TCA precipitable material at 0 and 120 min . Five µl were removed and spotted in duplicates on Whatman 3 M filters , dried , then immersed in a beaker containing 300 ml 5% TCA and 1% PPi for 15 min on ice . Filters were rinsed 3 times 5 minutes with 300 ml 5% TCA 1% PPi , then immerged 1 minute in 70% ethanol , dried and finally counted in a liquid scintillation counter . Radiolabelled DNA was obtained by random priming of a 700 nt PCR product , whereas radiolabelled RNA was obtained by T7 transcription of a linearized plasmid containing a cloned gene downstream a T7 promoter . Radiolabelled DNA and RNA were separated from free unincorporated nucleotide either by spin column or by ethanol precipitation in the presence of ammonium acetate . Single stranded DNA was obtained by 5 min heat denaturation of DNA at 100°C followed by rapid chilling in ice . Mat-C was labelled by addition of [γ32P]ATP using the activity of the in vivo incorporated cellular protein kinase ( P-Mat-C ) as described previously [19] and resulted in the transfer of 2–4 P/capsid ( T4 ) which corresponds to 0 . 008–0 . 017 P/core protein P . rC and C183S rC were labelled in vitro by PKC according to Kann et al . [26] ( P-rC and P-C183S ) . In brief this phosphorylation requires partial disintegration of capsid structure by low salt treatment , followed by phosphorylation . Capsids from the stock solution were at first diluted 1∶20 in water and incubated for 15 min at 40°C . 5 µg of the diluted capsids were preincubated in PKC buffer ( 20 mM HEPES pH 7 . 4; 10 mM MgCl2; 1 . 7 mM CaCl2; 1 mM DTT ) together with 3 µg phosphatidylserine for 10 min at 42°C . Then 15 ng protein kinase C ( Promega ) and 10 µCi [γ32P]ATP ( 3000 Ci/mmol ) were added ( 5 µl final volume ) and the kinase reaction let to proceed for 30 min at 37°C , after which 0 . 5 µl PBS 10X was added . In contrast to the previously reported protocol we did not performed an RNase A digest , resulting in a 50fold reduced labelling of about 0 . 01 P/core protein . After phosphorylation the capsids were reconstituted by adjusting the salt concentration to physiological concentrations . As we have shown previously by EM these capsids exhibit the same structure than original capsids [26] . For separation by native agarose gel electrophoresis , capsids were loaded on the gel using sample buffer without SDS . Electrophoresis was performed on 1% agarose/TAE using a TAE buffer . Blot onto PVDF membranes was performed according to Southern [46] . Quantification of radioactivity was performed by phosphoimaging using either Typhoon 9200 ( Amersham Biosciences ) or Pharos FX ( BioRad ) . Quantification of immune blots was performed by ECL ( PerkinElmer ) using ChemiDoc XRS ( BioRad ) . For analysis of capsid integrity during transport on Nycodenz gradients , or on Superdex 75 PC 3 . 2/30 columns , 4×106 HuH-7 cells were permeabilized by digitonin as described previously [19] , with the modification that the permeabilized cells were harvested after the washing steps prior to the transport reaction . The transport was performed as described , but using a volume of 100 µl comprising 50 ng P-Mat-C and rabbit reticulocyte lysate . If required , 100 µg/ml of wheat germ agglutinin ( WGA ) was added during the washing steps . After incubation , cells were lysed using 1% NP-40/PBS/5 mM MgCl2 for 1 h at 37°C . After lysis , Triton X-100 was added to a final concentration of 0 . 2% , followed by 10 min incubation at 37°C and subsequent sonification ( 6×15 sec ) . The sample of 200 µl was then subjected to analysis . RNase A-treatment was performed for 15 min at 37°C at a final concentration of 4 µg/µl . Transport assays for immune detection , immune staining and subsequent confocal laser scan microscopy were done accordingly to Rabe et al . [19] . Immune staining was performed with the polyclonal rabbit anti capsid antibody ( 1∶200 , DAKO Ab ) and with the monoclonal mouse anti capsid protein antibody ( 1∶200 , Fab3105 , Institute of Immunology CO . , LTD , Tokyo , Japan ) . Cy5-labelled anti rabbit antibody ( 1∶400 , Dianova ) and FITC-labelled anti mouse antibody ( 1∶200 , Dianova ) served as secondary antibodies . Nuclei were stained with propidium iodide ( 1∶5000 ) . Two hundred µl of samples were added onto 3 . 6 ml continuous Nycodenz/TN gradient ( 1 . 11–1 . 32 g/ml ) in a SW60 rotor . Centrifugation was done for 22 h at 10°C and 36 , 000 rpm . Fractions of 220 µl were harvested . Density d was determined by refractometry ( σ ) using the formula: d = ( σ×3 , 287 ) – 3 , 383 . The fractions were vortexed prior to density determination . Capsid preparation was centrifuged for 5 min at 17 000 g before being processed through the size-exclusion column . The apparent molecular size of the proteins was analyzed by chromatography on a Superdex 75 PC 3 . 2/30 column ( GE Healthcare ) , which has a fractionation range of 3 to 70 kDa . The column was equilibrated with transport buffer ( 20 mM HEPES pH 7 . 3 , 2 mM Magnesium acetate , 110 mM Potassium acetate , 5 mM Sodium acetate , 1 mM EGTA and 1 mM DTT ) . Proteins ( 50 µl ) were eluted with a flow rate of 40 µl/min and recorded by continuously monitoring the absorbance at 280 nm . The column was calibrated with the following standard proteins: ovalbumin ( 43 kDa; 1 . 13 ml ) ) , chymotrypsinogen A ( 25 kDa; 1 . 26 ml ) , RNase A ( 13 . 7 kDa; 1 . 36 ml ) and the void volume was determined with dextran blue ( >2000 kDa ) . Quantification of the capsids were done by immune blot accordingly to Rabe et al . ( 11 ) using either the polyclonal rabbit anti capsid antibody ( 1∶5000 , DAKO Ab ) or the monoclonal mouse anti capsid protein antibody ( 1∶2500 , Fab3105 ) . As secondary antibody , a horse radish peroxidase anti rabbit or anti mouse antibody ( 1∶10000 , Dianova ) were used . Detection was performed by ECL ( PerkinElmer ) using ChemiDoc XRS ( BioRad ) . For immune precipitations 3 . 5×106 sheep anti rabbit- or anti mouse-conjugated biomagnetic beads ( Dynal ) were added to 22 µg DAKO Ab or mouse anti capsid protein antibody ( Fab3105 ) and incubated in 0 . 1% BSA/PBS overnight at 4° C on a rotating wheel . Afterwards unbound antibodies were removed by washing the beads three times with 0 . 1% BSA/PBS . The antibody saturated beads were subjected to light and heavy capsid fractions from Nycodenz gradients in the presence of 0 . 1% BSA and incubated overnight at 37°C on a rotating wheel . The precipitate was washed three times in 0 . 1% BSA/PBS , one time in 0 . 1% Nonidet P-40/PBS , transferred into a new cup and again washed three times with PBS . The samples were loaded onto a 4–12% Bis-Tris gradient gel SDS-PAGE ( NuPAGE ) and transferred onto a PVDF ( VWR International ) by electro blotting . Precipitated capsid proteins were detected by their intrinsic radioactive signals using phosphoimaging . | Viral capsids facilitate protection of the enclosed viral genome and participate in the intracellular transport of the genome . At the site of replication capsids have to release the genome , but after replication new capsids have to be assembled for encapsidation of the progeny genomes . Detailed data on stability of capsids and kinetics of their formation and dissociation are obtained for several viruses in vitro , allowing biophysical or electron microscopical techniques . These approaches , however , do not consider the impact of cellular interaction partners . Using digitonin-permeabilized cells which support hepadnaviral genome release actively , we analyzed the disassembly kinetic of the hepatitis B virus ( HBV ) capsid . Using different analytical methods we found that HBV capsids disintegrate to protein dimers which reassemble to capsids inside the nucleus . The study provides a link between in vitro and in vivo data showing that HBV uses a unique strategy . We propose a model in which the unique environment of the nuclear pore favors the disassembly reaction , while both cytoplasm and nucleus favor assembly . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"virology/virion",
"structure,",
"assembly,",
"and",
"egress",
"virology/viral",
"replication",
"and",
"gene",
"regulation",
"virology/host",
"invasion",
"and",
"cell",
"entry"
] | 2009 | Nuclear Entry of Hepatitis B Virus Capsids Involves Disintegration to Protein Dimers followed by Nuclear Reassociation to Capsids |
Low-oxygen tolerance is supported by an adaptive response that includes a coordinate shift in metabolism and the activation of a transcriptional program that is driven by the hypoxia-inducible factor ( HIF ) pathway . The precise contribution of HIF-1a in the adaptive response , however , has not been determined . Here , we investigate how HIF influences hypoxic adaptation throughout Drosophila melanogaster development . We find that hypoxic-induced transcriptional changes are comprised of HIF-dependent and HIF-independent pathways that are distinct and separable . We show that normoxic set-points of carbohydrate metabolites are significantly altered in sima mutants and that these animals are unable to mobilize glycogen in hypoxia . Furthermore , we find that the estrogen-related receptor ( dERR ) , which is a global regulator of aerobic glycolysis in larvae , is required for a competent hypoxic response . dERR binds to dHIFa and participates in the HIF-dependent transcriptional program in hypoxia . In addition , dERR acts in the absence of dHIFa in hypoxia and a significant portion of HIF-independent transcriptional responses can be attributed to dERR actions , including upregulation of glycolytic transcripts . These results indicate that competent hypoxic responses arise from complex interactions between HIF-dependent and -independent mechanisms , and that dERR plays a central role in both of these programs .
The ability to adapt to limiting oxygen requires metabolic versatility , with cells transitioning toward glycolytic lactate production for energy production . Complementing this strategic change of metabolism are complex shifts in the transcriptome , which add durability to the initial hypoxic response . At the vanguard of the transcriptional reply to hypoxia is the HIF transcriptional complex , which is comprised of the oxygen-labile hypoxia-inducible factor-1a ( HIF-1a ) and its stable partner HIF-1b . This ancient pathway is central to the hypoxic response and is highly conserved from worms to human [1] . The actions of the HIF complex exert considerable influence in the etiologies of many diseases , including cancers and heart disease [2]–[4]; these conditions have a hypoxic component – and therefore an altered metabolic component – that is critical to disease progression . In normoxia ( N ) , HIF-1a is marked by a set of 2-oxoglutarate-utililizing prolyl hydroxylases ( PHDs ) that recognize specific proline residues within the oxygen-dependent degradation ( ODD ) domain [5] , [6] . Prolyl modification of HIF-1a allows it to associate with the von Hippel-Lindau ( VHL ) tumor-suppressor and it is subsequently degraded [7]–[10] . Hypoxia disrupts the degradative cascade , allowing HIF-1a accumulation and activation of the HIF transcription pathway [11] , [12] . The number of transcripts impacted by HIF-1a is large and ontologically diverse . Despite this , a few affected pathways generally characterize HIF-mediated adaptation responses , including upregulation of angiogenic [13] , [14] , erythropoietic [15] , and glycolytic transcripts [16] , [17] . The total hypoxic response , however , is not entirely dependent on the HIF pathway . For example , Shen et al . found 110 hypoxia response genes in C . elegans , 47 of which were induced in the absence of HIF [18] . Although HIF-independent hypoxia-induced activities have also been identified in other organisms , these pathways remain poorly understood , though even in mammalian cells , HIF-1a is dispensable for hypoxic upregulation of a host of transcripts [19]–[21] . These results suggest that HIF-independent hypoxic signaling mechanisms may act in concert with , or even supplant , the HIF response pathway in a context-dependent manner . Drosophila melanogaster deal with no/low oxygen conditions well when compared to mammals , and can survive anoxic challenge for hours at a time [22] , [23] . Flies maintain the three fundamental components of the HIF pathway: 1 ) the HIF prolyl hydroxylase ( Fatiga ) ; 2 ) dVHL; and 3 ) both components the HIF complex – dHIFa ( encoded by sima ) and Tango ( dHIFb ) [24] . As in mammals , dHIFa has an ODD domain that is sufficient to direct oxygen-sensitive degradation when hydroxylated [25] . While previous studies have examined hypoxic responses in adult flies [26] , [27] , the precise input that dHIFa has in this process has not been examined . In contrast , detailed studies have shown that dHIFa plays a vital role in directing hypoxia-driven terminal branching of the tracheal system during development [28] , [29] . The Drosophila tracheal network serves as the fly respiratory system , and it is noteworthy that its developmental branching bears a striking resemblance to processes controlling mammalian angiogenesis [30] . In addition , similar hypoxia-induced metabolic transitions have been reported in flies and mammals [31] , although these remain poorly defined . The highly conserved dERR nuclear receptor directs a developmentally-regulated transcriptional switch towards glycolytic metabolism that supports developmental growth [32] . This function is akin to that described for ERRa in vertebrates , which is associated with glycolytic metabolism and breast cancer [33]–[38] . Importantly , mammalian ERRs are also active participants in HIF-mediated hypoxic responses . They are directly recruited by HIF-1a to HREs and are required for a complete transcriptional response at specific promoters [39] , suggesting that ERRs play a critical role in hypoxic responses . We set out to interrogate hypoxic responses in Drosophila and wanted to assess the influence of dHIFa on transcriptional and metabolic adaptation . We report here that the hypoxic transcriptional response segregates into distinct HIF-dependent and HIF-independent pathways . These pathways are differentially sensitive to hypoxic challenge in a temporal fashion during development , but both pathways are most sensitive prior to metamorphic onset and least active in the immediate hours following pupariation . Contrary to expectations , we find that upregulation of glycolytic transcripts is non-HIF-dependent . Our metabolic analysis suggests that loss of dHIFa has a profound and wide-ranging affect on all aspects of carbohydrate catabolism when unchallenged in normoxia . In hypoxia , however , sima mutants remain unable to mobilize glycogen , which is preferentially depleted under hypoxic conditions . We additionally show that dERR is required during hypoxia , in that it controls a unique set of hypoxia-regulated dERR-dependent transcripts that include HIF-independent glycolytic genes . Altogether , our studies raise important questions regarding the breadth of HIF involvement in hypoxic transitions and identify dERR as an essential factor that complements HIF-dependent and -independent responses .
To better understand the contribution of in the hypoxic adaptation response , we wanted to determine the developmental time point when dHIFa was most active . To start , we examined the wild-type expression of two known hypoxia-responsive transcripts in Drosophila , lactate dehydrogenase ( LDH , known also as: ImpL3 , CG10160 ) , and the HIF prolyl hydroxylase , fatiga ( CG31543 ) [40] , [41] . We also examined the rate-limiting enzyme of glycolysis , phosphofructokinase , encoded by Pfk ( CG4001 ) as a potential hypoxia-responsive gene . We surveyed three times points , late embryo , mid-second instar ( mid-L2 ) larvae , and late-L3 larvae by qRT-PCR to examine transcriptional responses of whole animals that were allowed to develop in normoxia and then challenged with a 4% O2 treatment for 6 hours – hereafter referred to as H-treatment ( Figure 1A ) . This level of oxygen , and this time course , has previously been shown to mobilize the fly HIF pathway [42] . As seen in Figure 1D , the late-L3 time point of wandering larvae [−10 to −4 hours relative to the onset of pupariation ( RTP ) ] is a period where each of the three genes is significantly induced by H-treatment . This expression profile is different from responses observed in embryos 18–24 hr after egg laying ( AEL ) ( Figure 1B ) and mid-L2 larvae ( Figure 1C ) , when LDH and Pfk were unresponsive to treatment , indicating that hypoxic responses are developmentally tempered . To establish the identity of the full complement of H-regulated transcripts , RNA samples were prepared from N- and H-treated pools of control w1118 animals and sima mutants at the late-L3 time . The sima mutant line ( sima07607 ) contains a lethal P-element insertion in the first intron in the sima locus , which eliminates detectable expression of the transcript , rendering the animals incapable of directing expression of an oxygen-sensitive murine LDH-reporter and , importantly , unable to respond competently to hypoxic challenge [42] , [43] . As expected , H-treatment resulted in a pronounced change in the transcriptome . Using the microarray scheme outlined in Figure S1 , we extracted a series of significantly altered gene sets ( Table S1 ) . We were primarily concerned with identifying two mutually exclusive H-regulated gene sets – HIF-independent ( HI ) and HIF-dependent ( HD ) . Transcripts that did not exhibit at least a 1 . 5-fold change in expression , and which did not have a false discovery rate ( FDR , q-value ) of less than 1% , were not included in any set . This high stringency means that we have likely excluded genuinely H-regulated transcripts from our final sets , be they HD or HI . Despite this , we classified 254 transcripts as HI and 171 as HD . It is important to note that the HI and HD categorizations reflect the hypoxic responsiveness of individual transcripts at the late-L3 time alone . The top 20 affected transcripts from the HIF-dependent and -independent categories are listed in Table 1 . Gene ontology ( GO ) analysis [44] was performed on the hypoxia genes sets ( Figure 2A and 2B ) . Notably , the HI genes set , and not the HD genes set , contain glycolytic transcripts that are upregulated in hypoxia , which was the single most statistically-impacted process in either the HD or the HI sets ( Figure 2A ) . Instead of glycolytic genes , significant GO categories were identified for HD genes involved in translational control and RNA processing ( Figure 2B ) . However , among the HD H-regulated transcripts are fatiga and dVHL . This suggests that dHIFa participates in a feedback regulatory loop that attenuates its own activity . Ontology-focused heat maps were generated to compare hypoxic transcriptional responsiveness . In addition to glycolytic genes , comparisons were made for other metabolic categories where GO significance was identified , including oxidoreductase activity , lipid metabolism , vitamin binding , and amino acid metabolism ( Figure 2C ) . With the exception of lipid metabolic genes , when hypoxic responsiveness is seen ( up- or downregulated ) in the control background , the majority of genes also respond in-kind in the sima background , and usually with a similar fold-change increase . These results suggest that HIF-independent , H-sensitive mechanisms account for a large percentage of the hypoxic response . The unexpected breadth of contribution of the HI pathway in the hypoxic response led us to reconsider our initial observations made in Figure 1 , where fatiga displayed a similar response profile across each of the times assayed , and LDH and Pfk displayed a hypoxic response at only the late-larval time . Indeed , fatiga is a HD gene , whereas LDH and Pfk are HI genes ( Table S1 ) . Were LDH and Pfk unresponsive at earlier developmental times because the HI pathway was not active until just prior to metamorphic onset ? To address this question , we collected RNA from control animals and sima mutants staged at times that spanned development . In all , twelve samples were gathered: 4 embryonic times ( 0–6 hrs AEL [w1118 background only] , 6–12 hrs AEL , 12–18 hrs AEL , and 18–24 hrs AEL ) ; 4 larval times ( mid-L1 , mid-L2 , mid-L3 , and −4 hr RTP ) ; 3 metamorphic times ( 0 hr RTP , +12 hr RTP , +72 hr RTP ) ; and 1 adult time ( 1 day-old males ) . qRT-PCR was used to assess H responses of 13 select genes that displayed varying levels of H-sensitive expression . Of those genes analyzed: five were classified as HD genes – fatiga , spermine oxidase ( CG7737 ) , sequoia ( CG17724 ) , branchless ( CG4608 ) , and Peroxiredoxin 2540-2 ( Prx2540 , CG11765 ) ; seven were classified as HI genes – LDH , Pfk , NMNAT ( CG13645 ) , Alas ( CG3017 ) , Cyp9b1 ( CG4485 ) , Cyp6a17 ( CG10241 ) , and Cyp6a22 ( CG10240 ) ; and one was highly affected by H , but did not meet the stringency requirements for H set inclusion – CG31769 , which had a largely HIF-dependent expression profile in late-L3 . Several patterns emerged from the developmental analysis . First , hypoxic transcriptional induction is most evident at the late-L3 time for each gene assayed . Second , without exception , HD and HI genes display marked drops in H responsiveness just after metamorphic onset . In most cases , H responsiveness is eliminated during the hours surrounding head eversion , which is the initiation of the pupal phase . Among the transcripts examined , HD genes were not induced in hypoxia in a sima background at any developmental time , with the notable exception of a single point in mid-L3 for branchless ( Figure 3D ) . Finally , in the absence of dHIFa , HI genes tend to be hyper-responsive to H challenge throughout development – this was true for all genes examined except LDH ( Figure 4A ) . The LDH profile was unique amongst those assayed , in that late-L3 expression was HI , while pupal expression appears to be dominated by HD expression . The super-activation of LDH during the pupal phase in the w1118 background ( vs . sima ) suggests that both the HD and HI pathways are capable of converging simultaneously at the same locus to contribute to its overall expression . Collectively , these developmental expression data indicate that hypoxic responses are comprised of a patchwork of HD and HI activities throughout life-stage progression . The observation that glycolytic transcripts are effectively upregulated in sima mutants challenged with hypoxia raised the question of how metabolism was affected under these conditions . As before , we concentrated on the late-L3 time because of its particularly robust transcriptional response to H-treatment . We found that glycogen was significantly depleted by control w1118 animals in H , in addition to a near 50% reduction in the level of ATP ( Figure 5A and 5B ) . We tested for additional HIF-dependent metabolic defects in carbohydrate catabolism using mass spectrometry tied to gas and/or liquid chromatography ( GC/MS , LC/MS ) . Extracts were prepared from animals subjected to N- and H-treatments and 32 carbohydrate metabolites were measured ( Table S2 ) . The metabolites correspond to four broad categories: 1 ) aminosugar metabolism; 2 ) fructose , mannose , galactose , starch , and sucrose metabolism; 3 ) glycolysis , gluconeogenesis , and pyruvate metabolism; and 4 ) nucleotide sugars and pentose metabolism ( Figure 5C ) . We found that the control response to hypoxia is characterized by a remarkable level of metabolic stability for carbohydrate catabolites ( third column in Figure 5C ) . Among those compounds that do display significant H-induced depletions are oligomeric forms of glucose ( maltose , maltotriose , maltotetraose , and maltopentaose ) , which are catabolic products from glycogen and starch breakdown ( Figure 5C and 5E ) . These sugars feed into the glycolytic cascade by replenishing glucose . They are successively depleted in H the larger they are , and their reductions are consistent with a depletion of total glycogen seen in the w1118 response ( Figure 5A ) , as well as the HIF-dependent upregulation of amylase in H in the same background ( Figure 5D ) . In contrast to the effects that H-treatment has on w1118 animals , sima mutants cannot deplete glycogen in H ( Figure 5A ) . Instead , they adopt a profile for the maltose oligomers in normoxia that resembles the hypoxia-mobilized profile in control animals ( Figure 5C and 5E ) . This is likely a combination of two factors – the sima mutant's inability to effectively upregulate amylase in H and its constitutively elevated expression profile for amylase in normoxia that is greater than w1118 expression in hypoxia ( Figure 5D , Table S1 ) . Curiously , despite the clear transcriptional switch toward glycolytic energy production at late-L3 , lactate levels remained unchanged for either genotype in H ( Figure 5C ) . This failure to generate lactate in hypoxia is a stage-specific block . We independently performed lactate measurements by an enzymatic assay to confirm the late-L3 findings made by GC/MS . Indeed , we find that mid-L1 larvae and young adults from either the w1118 or sima backgrounds produce lactate in hypoxia , but not late-L3 larvae ( Figure S5A–S5C ) . Notably , early larval and young adult sima mutants exhibit an exacerbated hyperlactatemic phenotype when subject to hypoxia ( Figure S5A , S5C ) – this does not happen to late-L3 animals . Additionally , though the transcriptional H response profile was largely normal for glycolytic genes in the sima mutant , profound depletions were still observed for glucose-6-phosphate and fructose-6-phosphate in H ( Figure 5C and 5E ) . This is because the normoxic levels for these compounds , rather than H-induced changes , dominate their metabolism . We also observed a HIF-dependent increase for pyruvate in N and H . This is consistent with findings in HIF-1a−/− MEFs , which maintain higher levels of ATP in hypoxia than WT MEFs do in normoxia [45] . Finally , the elevated level of Ru5P:Xu5P and ribulose , coupled with the depleted levels of S7P , reveal that sima mutants display a clear split in the oxidative ( NADPH-generating ) and non-oxidative phases of the pentose phosphate pathway in normoxia , which is exacerbated by H-treatment . The only factor known to transcriptionally regulate glycolytic transcripts in Drosophila is dERR [32] . Our lab identified this orphan nuclear receptor as a potential factor that may participate in hypoxic signaling when the dERR ligand-binding domain ( LBD ) was used to repeatedly isolate sima clones in a large-scale yeast two-hybrid screen . Of the 20 positive clones recovered in the screen , seven encoded different C-terminal fragments of dHIFa . These findings are consistent with a previous report that demonstrated HIF/ERR interactions between the Drosophila proteins and their mammalian homologs [39]; however , there are two important aspects about HIF/ERR complexes that we note differentiate the fly and mammalian complexes . First , we find that the dERR DBD is dispensable for interaction with dHIFa , whereas the Ao report showed that interaction occurs between the mammalian ERR DBD and the HIF-1a/b heterodimer . Second , unlike in mammals [39] , HIF-1b ( tango ) is not required for dERR association with dHIFa in Drosophila – tango was not present in the screen . On this aspect , our findings are consistent with the findings made by Ao et al . Their two-hybrid screen of Drosophila components also did not have a HIF-1b [39] . We validated our two-hybrid screen findings by performing a GST-pulldown with GST-fused dERR LBD protein with full-length dHIFa , which confirmed a robust interaction ( Figure 6A ) . The C-terminal AF-2 helix of nuclear receptors often mediates interaction with transcriptional coregulators through an LXXLL motif that is found on the interacting protein [46] . A single such sequence resides within dHIFa , at amino acids 1289–1293 ( LKNLL ) . When the last two leucines of this sequence were mutated to alanine and/or when the last 12 amino acids of the dERR LBD were deleted , spanning the AF-2 helix ( 479–491 ) , the interaction between the proteins was severely reduced , but not eliminated ( Figure 6A ) . These data indicate that the dERR AF-2 helix mediates a docking point with the dHIFa LXXLL motif , but that at least one additional point of contact is maintained between dERR and dHIFa . We have recently shown that the orphan nuclear receptor dERR is essential for triggering the pro-growth glycolytic program during Drosophila development [32] . Without the dERR-initiated metabolic switch , development cannot successfully proceed . Many of the same metabolic genes that exhibit H-sensitive regulation are also misregulated in the dERR mutant . If dERR is important in the hypoxic response , as suggested by its association with dHIFa , then the mutants should be sensitive to H-treatment . To test this , we challenged dERR mutants and compared their H-sensitivity with sima mutants and control animals . Indeed , 24–30 hr AEL L1 larvae challenged with constant hypoxia resulted in sima mutant lethality ( Figure 6B ) . The dERR mutants were also H-sensitive , but not to the same extent as sima embryos . Nevertheless , dERR is critical , less than 25% of animals survived as compared to 97% survival for the w1118 background . These data indicate that dERR is essential for hypoxic adaptation . Using the same analytic framework that was used to assess sima involvement in the late-L3 larval hypoxic response , we collected RNA samples from dERR mutants and dERR , sima double-mutants for microarray analysis to determine how loss of ERR alone or ERR and dHIFa together would affect hypoxic responses ( Figure S2 ) . Through these analyses , we identified 282 dERR-dependent ( ED ) transcripts and 207 double-mutant-dependent ( DM ) transcripts whose expression changed in hypoxia ( Figure 7A , Table S1 ) . The ED and DM H-genes sets encompass a variety of highly significant GO categories , including H-induced kinases and transferases that specifically require dERR , and a host of nucleolar and RNA processing transcripts that are coordinately upregulated in hypoxia due to the lack of both dERR and dHIFa ( Figure S6A , S6B , Table S1 ) . Venn analysis was used to assess the similarity of the independently derived H-gene sets ( HI , HD , ED , DM ) . The overlapping pattern of the ED genes set with the mutually exclusive HI and HD sets demonstrates that dERR significantly affects both HIF-dependent H-genes ( 71 transcripts ) and HIF-independent genes ( 54 transcripts ) ( Figure 7B ) . Interestingly , among the overlap between the HI and ED genes sets are all the glycolytic transcripts that are upregulated in hypoxia . These data reinforce our earlier findings that demonstrate that at metamorphic onset , dHIF is not part of the hypoxic-induced glycolytic shift . They also suggest that a portion of the HI response it attributable to dERR . Given that dERR can interact with dHIFa , and that it can impact hypoxic transcription independent of HIF , we anticipated that the DM H-genes set would significantly overlap the ED and HD genes sets . Indeed , this is the case – as shown by Venn analysis , the DM set has more overlap with the HD and ED sets than not ( Figure 7C ) . A listing of the top hypoxia-sensitive transcripts in the various Venn overlapping regions can be found in Figure 7D . To verify that loss of dERR and/or dHIFa selectively eliminates/diminishes hypoxic induction , RNA samples were independently collected from control animals , sima mutants , dERR mutants , and double-mutants ( Figure 7D ) . Six genes were chosen for further analysis by qRT-PCR . The results demonstrate that the factor-dependent classification we employed for hypoxic responsiveness is accurate . For example , Pfk is classified as HI , ED , and DM , indicating that hypoxic regulation should be affected in the double-mutant and the dERR mutant backgrounds , but unaffected in the sima mutant – this is the pattern that is observed ( Figure 7E ) . Similar trends also held true for fatiga and spermine oxidase , which were expected , respectively , to only respond in the dERR mutant background , or not in any of the three mutant lines . With the exception of a modest H-induction in the dERR mutant for spermine oxidase the responses were true . Hypoxic responses for NMNAT , LDH , and ALAS were all expected to display the same pattern; which is that only in the dERR background will H-responsiveness be significantly reduced/eliminated . Responses were , by-and-large , as expected , except for the significant H-induction of LDH in the in dERR mutants . These data indicate that dERR and dHIFa have a different activity profile when in the presence of the other , than either protein has by itself , and suggest that promoter-specific actions of different HIF and/or ERR complexes drive a large percent of hypoxic responses at metamorphic onset . In certain cases , loss of one factor does not influence the other's response , as with loss of dHIFa for the dERR-mediated Pfk response ( Figure 7E ) . In other cases , loss of either dHIFa or dERR renders the H-response incomplete , such as occurs with spermine oxidase . And , still in other cases , loss of one factor is more detrimental for H-induction than is loss of both , as with ALAS . Responses of this type appear to suggest that , at certain loci , dHIFa acts as a negative regulator of hypoxic transcription in the absence of dERR but not in its presence .
Our results underscore the complexities of adaptive responses in hypoxia , which are life-stage specific and controlled by multiple H-sensitive pathways . Although our data confirm that HIF is a major transcriptional driver of hypoxic responses , we also define distinct HIF-independent responses . These data raise new questions about dHIFa collaboration and challenge the notion that the HIF complex has little or no normoxic role . In addition , we show that a significant fraction of HIF-independent pathways can be attributed to the ERR nuclear receptor . Among the HIF-independent genes were numerous glycolytic transcripts that are well-known responders to hypoxia [26] , [31] , [40] . The fact that these genes are as effectively upregulated in sima mutants as they are in a control response was surprising , particularly considering the known role of HIF-1a in this process [47] . We find that dERR is the overriding factor that mediates hypoxic upregulation of glycolytic genes ( Pgi , Pfk , GAPDH2 , enolase ) just prior to metamorphic onset . Our findings , however , do not exclude dHIFa contribution in hypoxic expression of HI genes at other developmental times . The super-induction of LDH during metamorphosis in w1118 animals versus sima mutants is consistent with this scenario ( Figure 4A ) . These temporal- and context-specific differences may explain the wide variability in hypoxic responses that have been seen between cell-types [16] , [48] , despite the ubiquitous presence of the HIF pathway . Furthermore , they may account for discrepancies between our data collected on Drosophila and reports on mammalian systems . For example , LDH is a HIF-independent hypoxia-regulated gene in late-L3 animals . However , loss of dERR has a greater effect on the diminution of hypoxic induction at this developmental time than does loss of dHIFa ( Figure 7C ) . But , this effect is short-lived , because just hours later , when the larva transitions into a pupa , dHIFa appears to work in combination with a non-HIF pathway to elicit hypoxic responsiveness ( Figure 4A ) . This combinatorial response during Drosophila metamorphosis is consistent with vertebrate studies that show LDH expression is the product of HIF-1 action that also requires the presence a cAMP response element for full hypoxic induction [47] , [49] . Thus , in addition to different pools of potential coregulatory molecules that may significantly alter HIF-dependent transcription , entirely different transcriptional pathways , with their own triggers of hypoxic induction , refine the H response . Given the right spatiotemporal setting , HIF-independent pathways may displace ( or substitute for ) the HIF pathway altogether , a result that is consistent with our data . Further support of this idea is evident in the expression of Alas2 , the rate-limiting enzyme for heme production . Alas2 has been identified as a HIF-dependent and a HIF-independent hypoxia-regulated gene in mammals [50]–[52] . In our hands , ALAS is H-responsive , and displays HIF-independent and ERR-dependent upregulation , which may be subject to dHIFa negative regulation in dERR's absence ( Figure 7E ) . The dynamic patterns of temporal expression of HI and HD genes raise the fundamental question of how hypoxic responses are regulated through development and into the adult . Low-oxygen responses are not one-size-fits-all programs that mitigate oxidative damage and metabolic imbalance; they must be coordinated with developmental progression and metabolic state . In particular , late-L3 wandering larvae exhibit a hypersensitive transcriptional response to hypoxia for HIF-independent/ERR-dependent glycolytic genes . This includes a robust LDH induction ( Figure 4A ) . Paradoxically , however , late-L3 larvae do not produce lactate in the 6-hr hypoxic challenge ( Figure 5C , 5E , and Figure S5 ) . In contrast , at other developmental times ( L1 , adult ) , animals correspondingly produce lactate in hypoxia , even though they remain transcriptionally incompetent to induce LDH transcript ( Figure S5 and Figure 4A ) . We speculate that the atypical transcriptional and metabolic hypoxic profiles of the late-L3 larva are a product of its developmentally programmed energetic state , which at this time is transitioning from low to high efficiency ( see the dramatic decrease of Pfk expression in late larvae in Figure S4 ) . Just prior to the wandering L3 time , larvae are prolifically growing , and in a state of metabolism that is fueled by aerobic glycolysis – this metabolic program is ERR-dependent [32] . Just after this developmental time , larvae initiate metamorphosis , which will impose 5 days of developmentally forced starvation . During this lipid-driven phase [53] , metabolism is characterized by high efficiency OXPHOS . In contrast to the switch-like hypoxic expression of HIF-independent glycolytic transcripts , the HIF prolyl hydroxylase fatiga displays relatively uniform expression throughout development ( Figure 3A and Figure S3 ) , suggesting that regulation of the HIF pathway , by HIF itself , is equally important at all times for the animal . Such disparities in induction are only understood in context . While our studies here provide a framework with which to view H responses , they indicate that further developmental analysis is needed to more fully appreciate hypoxic response pathways and the mechanisms that specifically support their activities . Although we have emphasized the transcriptional and metabolic impacts of hypoxia on carbohydrate catabolism , the breadth of our data sets indicate that many important hypoxia-induced changes are thus far unappreciated and await further investigation . What is the significance , for instance , of the greater than 10-fold increase of HIF-dependent expression of dDPH-1 ( CG11652 ) in hypoxia ( Table 1 ) ? DPH-1 is a tumor suppressor that is responsible for the first step of the unique protein modification that occurs on elongation factor 2 ( eEF2 ) , which converts a histidine residue to diphthamide . This residue is the target of diphtheria toxin that can shut down protein synthesis through ADP-ribosylation . Although diphthamide formation is conserved from archaea to human , its significance on cellular function is not clear , as it is dispensable for protein elongation [54] . However , it has been implicated in translational fidelity [55] and is likely an asset under stress [56] . GO analysis performed on HD H-regulated genes indicate that dHIFa is important in replenishing select protein translation/RNA processing transcripts . From this perspective , DPH-1 induction by dHIFa may be indicative of a regulatory role of hypoxic translation for HIFs . Such a role would be consistent with a recent report from mammals that demonstrates a HIF-2a-dependent association with ribosomal/translational control proteins and the selective hypoxic translation of transcripts containing an RNA hypoxic response element in the 3′UTR via a mechanism involving eIF4E2 [57] . Our analysis of carbohydrate catabolism identifies amylase-mediated breakdown of glycogen as the fuel of first resort in hypoxia ( Figure 5 ) . This catabolic pathway feeds into glycolysis and supplies needed glucose for increased glycolytic flux , obviating the need to draw on circulating sugar in the form of trehalose , which did not change in the 6-hr challenge . The strategy of glycogen mobilization allows animals to maintain a remarkably stable profile for a wide variety of carbohydrate catabolites . Trehalose levels are substantially elevated in sima mutants , regardless of oxygen status ( Figure 5C and 5E ) . These data may indicate a role for dHIFa in the insulin receptor pathway . Numerous studies demonstrate that trehalose levels are altered by genetic disruptions of the insulin-signaling components [58]–[61] . Alternatively , elevated trehalose levels may be the result of constitutively high expression of amylase ( Figure 5D ) . Although the increased amylase expression does not translate into a depleted level of glycogen in the sima mutant ( Figure 5A ) , it is conceivable that increased glycogen deposition compensates for increased glycogenolysis . It is important to note that post-transcriptional control mechanisms are well known to impact glycolytic enzymes . Although we did not document them , we consider such influences on hypoxic glycolytic flux likely to have genotype-specific effects . sima mutants do not mobilize glycogen in hypoxia , but they are able to initiate H-induced changes for other carbohydrates . This is the case for the glycolytic intermediate DHAP , which more than doubles in a control hypoxic response and significantly accumulates in mutants ( Figure 5C , 5E ) . These findings are consistent with appropriate transcriptional responses we noted for glycolytic transcripts in sima animals , which are upregulated in hypoxia by dERR , not dHIFa . The results for glycogen notwithstanding , it is the widespread derangement of normoxic set points for metabolites that characterizes the metabolic incompetency of the sima mutant . Our data indicate that dHIFa has it greatest impact on metabolism in the unchallenged normoxic state , rather than in hypoxia . The mechanism whereby dERR participates in hypoxic responses needs to be explored further . We identified dERR as a potential player in hypoxic responses through its association with dHIFa ( Figure 6A ) , suggesting that it acts in a collaborative role with the HIF complex through direct recruitment to HREs . This model was favored by the Ao et al . report for ERR participation in hypoxic responses in vertebrates [39] . Additionally , dERR may recruit dHIFa to ERR-specific response elements to facilitate H responses . Another possibility is that dERR actively regulates hypoxic transcription without dHIFa at all; or , in parallel to the actions of dHIFa , which may occur independently , yet simultaneously . Each of these scenarios is consistent with hypoxic expression analysis that we performed to generate HD , HI , ED , and DM gene sets . Moreover , in the presence of dERR , dHIFa may act as a negative regulator of hypoxic responses at select hypoxia-regulated sites ( Figure 7E , NMNAT , ALAS ) . Of further interest also , will be the identification of the triggers for ERR participation in hypoxic-induced responses . Apart from dERR and dHIFa , our data indicate that at least one more hypoxic-sensitive pathway is active and important for mediating hypoxic adaptation , as we found many H-sensitive transcripts that fall outside the regulation of either factor . The nature of the alternate pathway ( s ) is unknown . The results shown here suggest that identifying the sensors and effectors that regulate these HIF- & ERR-independent hypoxic response pathways will have profound impacts on our understanding of hypoxic signaling , and will undoubtedly provide new avenues with which to approach the complex problem of metabolic transition .
Flies were maintained on regular cornmeal-molasses-yeast media at 25°C . sima mutants ( sima07607 ) [43] were obtained from Bloomington Stock Center . w1118 animals were treated as controls . dERR mutants ( dERR1/dERR2 ) are described elsewhere [32] . dERR , sima double-mutants were generated by recombination of the sima07607 allele with each of the individual dERR1 and dERR2 mutations . Embryos were collected at 25°C for 14 hrs onto egg caps ( molasses-agar media in 35 mm×10 mm dishes ) with yeast paste . Mid-L2 larvae were transferred to a fresh egg cap with blue yeast paste ( 0 . 3% bromophenol blue ) , and allowed to develop until achieving the partial clear-gut L3 stage ( −10 to −4 hrs RTP ) . Staged animals were moved to fresh agar plates and allowed to age an additional 6 hours at 25°C ( normoxic treatment ) ; or , animals were placed in an airtight Modular Incubator Chamber ( Billups-Rothenberg , Inc . , Del Mar , CA ) for 6 hours at 25°C after a gas mixture containing 4% oxygen balanced with nitrogen was flashed into the chamber ( hypoxic treatment ) . The sima07607 chromosome was carried over a TM3 , twi-GFP ( green fluorescent protein ) balancer chromosome . Homozygous mutant larvae were sorted for the absence of GFP expression using a Zeiss Discovery V . 8 dissecting stereoscope with fluorescence at mid-L2 . For lethal phase analysis in Figure 6B , 0–4 hr post-hatch L1 larvae were sorted for fluorescence to assign genotype . Larvae were placed in vials containing fresh yeast paste and were then exposed to 21% ( normal air ) or 4% oxygen for 48 hrs and scored for lethality or completion of L1 . Microarray analyses were performed on at least three biological replicates of w1118 animals , sima mutants , dERR mutants , and sima , dERR double-mutants at the partial clear-gut L3 stage and treated for 6 hrs in normoxia or 4% O2 . For each biological replicate , at least 10 larvae were collected and washed with 1×PBS before homogenization in TRIzol ( Invitrogen , Carlsbad , CA ) using a VWR disposable pellet mixer . Total RNA was isolated using a TRIzol/RQ1 DNase hybrid extraction protocol ( Promega , Madison , WI ) . Template labeling was done using the GeneChip 3′ IVT Express Kit according to the manufacturer's specifications ( Affymetrix , Santa Clara , CA ) . Hybridizations to Affymetrix GeneChip Drosophila Genome 2 . 0 arrays were performed using the manufacturers recommendations . Every chip was scanned at a high resolution by the Affymetrix GeneChip Scanner 3000 according to the GeneChip Expression Analysis Technical Manual procedures ( Affymetrix , Santa Clara , CA ) . Raw data were normalized with RMA [62] and analyzed with the significance analysis of microarray ( SAM ) program [63] . No changes below 1 . 5-fold were considered significant . Additionally , the following false discovery rate percentages were imposed: 0 . 733% for w1118 normoxia vs . w1118 hypoxia; 0 . 414% for sima normoxia vs . sima hypoxia; 0 . 721% for w1118 hypoxia vs . sima hypoxia; 7 . 84% dERR normoxia vs . dERR hypoxia; 0 . 619% for w1118 hypoxia vs . dERR hypoxia; 0 . 662% dERR , sima double-mutant normoxia vs . dERR , sima double-mutant hypoxia; 0 . 703% for w1118 hypoxia vs . dERR , sima double-mutant hypoxia . Microsoft Access was used to compare data sets . Microarray data from this study can be accessed at the Omnibus website ( http://www . ncbi . nlm . nih . gov/geo ) with the accession number GSE33100 . Total RNA samples were isolated as described above . RNA was reverse transcribed with the High Capacity cDNA Reverse Transcription Kit ( Applied Biosystems , Carlsbad , CA ) using the manufacturer's specifications . For real-time PCR , premixed primer-probe sets were purchased from Applied Biosystems , with the exception of the primer set used for amylase . For amylase , a standard SYBR Green ( Bioline , Taunton , MA ) protocol was used with the primer sets: 5′ AACTACAACGACGCCAACGAG 3′ and 5′ TGGTCGGTGTTCAGGTTCTTG 3′ . All amplifications were carried out on a CFX96 real-time PCR system ( Bio-Rad , Hercules , CA ) . Experimental values were normalized to values obtained for the Rp49 probe set . Data are reported as the mean±SEM . All values reported represent experiments performed on at least three biological replicates . Analyses were performed on partial clear-gut L3 larvae treated for 6 hours in normoxia or 4% O2 . After treatment , animals were washed twice in PBS pH 8 . 0 and immediately frozen at −80°C . For glycogen measurements , 45 animals were split into three pools and the assay was performed essentially as described [64] . Color intensity was measured using a Bio-Tek Elx800 absorbance micro-plate reader at 540 nm . Glucose and glucose+glycogen amounts were determined using a standard curve . The amount of glycogen was determined by subtracting the glucose from the glucose+glycogen total . Glycogen amounts were normalized to protein content in each homogenate using a Bradford assay ( Bio-Rad ) . For ATP measurements , larvae were homogenized in 300 µl of 6M guanidine-HCl extraction buffer ( 100 mM Tris and 4 mM EDTA , pH 7 . 5 ) . The homogenate was heated at 70°C for 5 min and centrifuged in at 3000×g for 1 min . The supernatant was diluted 1∶750 in dilution buffer ( 25 mM Tris and 100 µM EDTA , pH 7 . 5 ) and spun at 14000×g for 3 min , after which 10 ll supernatant was transferred to a 96-well white opaque plate and mixed with 100 ll of luminescent solution ( Invitrogen , Molecular probes ) . Luminescence was immediately measured by a Bio-Tek Synergy 2 SL luminometer . The amount of ATP was determined using a standard curve . Amounts were normalized to total protein . For lactate measurements , 300 first instar larva , 60 third instar larva or 30 1-day-old males were split into three pools and measured as described Monserrate et al . ( 2012 ) using Lactate Assay Kit ( Biovision Milpitas , CA , ) [65] . For metabolomics , analyses were performed by Metabolon , Inc . ( Durham , NC ) . Replicates were normalized by protein content ( Bradford analysis ) . Recovery standards were added to samples prior to extraction using a proprietary series of organic and aqueous solutions . Extracts were divided into two fractions , one for GC and one for LC . Organic solvent was removed using a TurboVap ( Zymark ) . Briefly , for LC/MS , split samples were dried and reconstituted in acidic or basic LC-compatible solvents containing standards . Positive and negative ion-optimized sample conditions were analyzed in separate injections . For acidic reconstitutions a gradient of water and methanol containing 0 . 1% formic acid was used , and for basic extracts a water/methanol gradient with 6 . 5 mM NH4HCO3 . Analysis was performed on a Thermo-Finnigan LTQ mass spectrometer with an electrospray ionization source and linear ion-trap mass analyzer . For GC , samples were re-dried under vacuum prior to derivatization under nitrogen using bistrimethyl-silyl-trifluoroacetamide . The column was 5% phenyl with a temperature ramp of 40° to 300°C over 16 minutes . Samples were analyzed using a Thermo-Finnigan Trace DSQ fast-scanning single-quadrapole mass spectrometer with electron impact ionization . Refer to Table S2 for normalized data of each replicate and p- and q-values . Extensive quality control care was applied to minimize variability between days . The Metabolon platform has been described elsewhere [66] , [67] . Data values were imputed in the following way when values fell below the threshold level of detection: when all six replicates were undetectable , each was assigned the minimum detectable value of across all compounds tested; when five or less replicates were undetectable , sample values were assigned the minimum value obtained among those that were detected for a given compound . A yeast two-hybrid screen was conducted using the Invitrogen ProQuest Two-Hybrid System . For this purpose , three cDNA prey libraries were simultaneously prepared using the CloneMiner cDNA Library Construction Kit ( Invitrogen ) . All the libraries ( a , b , c ) were made from poly-A-selected RNA that was extracted from w1118 animals at −4 , +0 , or +4 RTP , which was reverse transcribed and pooled in equal proportions before library construction . Each library differs by only a single base pair in the adapter sequence to facilitate expression of clones in all three frames . Extensive procedures , provided by the manufacturer , were followed to capture clones into the pDONR222 vector . Clones in the donor vector were subsequently recombined into the pDEST22 vector . Libraries were titered ( a = 7 . 18E6 CFU , b = 4 . 44E6 CFU , c = 14 . 28E6 CFU ) and sampled for average insert size ( a = 1 . 64 kb , b = 1 . 25 kb , c = 1 . 6 kb ) before transformation into ElectoMax cells ( Invitrogen ) . Transformed cells for each library were pooled ( total of 6 . 4E6 CFU ) and grown for 22 hrs at 30°C for preparation of library DNA by standard techniques . 22lg of library DNA was transformed into the yeast bait strain containing the LBD of dERR ( L193-R496 ) that had been recombined into the pDEST32 vector . A total of 5 . 28E5 clones were screened by auxotrophic selection . All positive hits were sequenced . GST-pulldown experiments and the expression of GST-fused ERR constructs in pGEX-4T1 were performed as described [68] . A one-way ANOVA F-test was applied to test for the differences in glycogen levels , followed by Tukey's HSD method . For developmental qRT-PCR analysis , delta CT values were used to perform statistical analysis , whereby a two-tailed unpaired student's t-test was applied for the differences in gene expression using a Bonferroni correction . Following log transformation and imputation , a one-way ANVOA with contrasts was used to identify significance for metabolites in the mass spec analysis ( See Table S2 ) . Cumulative hypergeometric probability was used to determine significance between overlapping gene sets . | When oxygen levels fall below normal , cells are said to be in a hypoxic state . Once in hypoxia , dramatic changes are induced that allow for adaptation . In particular , energetic metabolism and transcription are highly affected . HIF ( hypoxia inducible factor ) is a highly conserved factor that is the driving force behind many hypoxia-induced changes—it is inactive in normal conditions and becomes active in hypoxia . Using the fruit fly as a model system , we show that hypoxic responses consist of HIF and non-HIF-dependent pathways . These response programs counteract the impacts of low oxygen by broadly influencing different cellular processes such as the breakdown of sugars , but only at appropriate developmental times . We provide evidence that HIF- and non-HIF-dependent pathways are complemented by the actions of the steroid hormone receptor estrogen-related receptor ( ERR ) , which we show is also essential in hypoxia . Our results place new emphasis on the actions of HIF and suggest that alternative HIF-independent pathways play a more prominent role than previously thought . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"models",
"biochemistry",
"drosophila",
"melanogaster",
"model",
"organisms",
"molecular",
"cell",
"biology",
"metabolic",
"pathways",
"gene",
"expression",
"biology",
"metabolism",
"molecular",
"biology",
"dna",
"transcription"
] | 2013 | HIF- and Non-HIF-Regulated Hypoxic Responses Require the Estrogen-Related Receptor in Drosophila melanogaster |
Epstein–Barr virus ( EBV ) is a ubiquitous oncogenic virus that induces many cancers . N6-Methyladenosine ( m6A ) modification regulates many cellular processes . We explored the role of m6A in EBV gene regulation and associated cancers . We have comprehensively defined m6A modification of EBV latent and lytic transcripts . Furthermore , m6A modification demonstrated a functional role in regulation of the stability of viral transcripts . The methyltransferase METTL14 was induced at the transcript and protein levels , and knock-down of METTL14 led to decreased expression of latent EBV transcripts . METTL14 was also significantly induced in EBV-positive tumors , promoted growth of EBV-transformed cells and tumors in Xenograft animal models . Mechanistically , the viral-encoded latent oncoprotein EBNA3C activated transcription of METTL14 , and directly interacted with METTL14 to promote its stability . This demonstrated that EBV hijacks METTL14 to drive EBV-mediated tumorigenesis . METTL14 is now a new target for development of therapeutics for treatment of EBV-associated cancers .
Epstein-Barr virus ( EBV or HHV4 ) , was the first human oncogenic virus discovered and isolated from a Burkitt’s lymphoma patient in 1964 [1] . EBV contributes to approximately 2% of all cancers [2] . EBV exhibits two distinct phases , lytic infection and latent infection . Latent infection is referred to as latency 0 , I , II and III [3] . Viral protein expression during latency III can potently promote indefinite proliferation of primary B-cells ultimately mediating immortalization of infected cells [4] . N6 methylation of adenosine ( m6A ) is the most abundant RNA modification and it is distributed at translation start sites , CDS regions , and 3’ UTRs [5–7] . Deposition of m6A on RNA is controlled by the cellular m6A machinery , which is the methyltransferase complex ( METTL3 , METTL14 and WTAP ) and the demethylase enzymes ( FTO and ALKBH5 ) [8] . Studies have shown that m6A modification has important roles during infection of cells with HIV-1 [9] , HCV [10] , Zika virus [11] , KSHV [12] and SV40 virus [13] . However , the role of m6A modification and the associated enzymes on infection by the ubiquitous human tumor virus EBV , and the effects on its potent oncogenic activities have not been explored . We investigated the role of m6A modification on EBV-mediated cell transformation , and further examined its contribution to the tumorigenic activities of the virus . We demonstrated that m6A modification of the EBV epitranscriptome facilitated enhanced expression of its latent genes and repressed lytic gene expression . Furthermore , we showed that the m6A methyltransferase METTL14 is a critical factor required for EBV-induced oncogenesis . We also showed that METTL14 was dramatically increased in EBV latently infected cells and down-regulated during EBV lytic infection . Knock-down of METTL14 led to decreased latent gene expression and an increase in lytic gene expression . EBV latent antigen EBNA3C , which is critical for viral-mediated transformation of cells , was up-regulated by METTL14-mediated m6A modification , and its expression led to a feedback loop by which METTL14 transcription was also induced , as well as its protein stability . We now provide for the first time a comprehensive understanding of the EBV epitranscriptome , and the host cell RNA processing machinery critical for regulation of viral transcripts to maintain latent infection required for its oncogenic properties .
To determine the role of m6A modifications in the EBV lifecycle , we first examined the viral epitranscriptome in EBV transformed lymphoblastoid cell lines ( LcLs ) and EBV-positive Burkitt's lymphoma , Akata cells , using methylated RNA immunoprecipitation followed by sequencing ( MeRIP-seq ) . Biological replicates of ribo-RNA deleted mRNA of each cell type were prepared for MeRIP-seq followed by peak calling using the MeTPeak package on both strands of the genome . We monitored expression of some lytic genes in LcLs through RNA-seq and RT-qPCR ( S1A and S1D Fig ) . However , we found that there was a much lower level of lytic transcripts compared to latent gene transcripts based on the results of our realtime PCR assays . In addition , the results of biological replicates were also highly consistent . The m6A conserved peaks among LcLs and Akata cells were found in transcripts of both latent and lytic genes . The latent genes with m6A modifications include EBNA1 , EBNA3A , EBNA3B , EBNA3C and LMP1 ( Fig 1A and 1B , S1 and S2 Tables ) . M6A modification was also detected on the lytic transcripts including BRLF1 , gp350 ( BLLF1 ) , DNA polymerase accessory subunit ( BMRF1 ) , as well as other lytic genes ( Fig 1A and 1B , S1 and S2 Tables ) . The m6A peaks on EBNA1 , EBNA3C , LMP1 , BRLF1 , gp350 , and BMRF1 transcripts were confirmed by methylated RNA immunoprecipitation , reverse transcription , and quantitative real-time PCR ( Fig 1C and 1F ) . The regions which encompassed the m6A peaks all displayed strong m6A enrichment compared to the IgG control . No m6A modification was detected on LMP2A transcripts from LcLs or Akata cells based on our qPCR analyses ( Fig 1C ) . These results suggested that m6A modification was specifically distributed on both EBV latent and lytic transcripts . We also investigated the levels at which the viral transcripts were modified during reactivation and induction of lytic replication . We detected m6A modification of EBV transcripts during lytic infection . TPA and Butyric acid were used to induce lytic reactivation . We observed a dramatic reduction of m6A modification of EBV transcripts both in reactivated LcLs and Akata cells ( Fig 1D and 1E and S3 and S4 Tables ) . However , m6A peaks were still detected on specific transcripts encoded by latent genes , including EBNA3A , EBNA3B , EBNA3C , EBNA1 , as well as BRLF1 , gp350 , and BMRF1 transcripts in lytically induced LcLs ( Fig 1F ) . Notably , the fold enrichment in lytic infection was lower than that seen during latent infection and there was a typically 5-20-fold difference ( Fig 1A , 1D , 1C and 1F ) . Additionally , the m6A modification signal was much lower in lytic Akata cells compared to lytic LcLs which ranged between 2–5 fold based on the MeRIP-seq data analysis and MeRIP-qPCR ( Fig 1D and 1F ) . The comparatively high enrichment of m6A modification of latent viral transcripts and low enrichment of m6A modification during lytic infection suggested that m6A modification may play a role in latent gene expression and maintenance of EBV latent infection . We identified an enrichment of m6A modifications of viral transcripts during EBV latent infection , and conversely a decrease in m6A modification of viral transcripts during lytic infection . Therefore , we hypothesized that m6A modification was important for latent antigen expression and stability in infected cells . One of the most important functions of m6A modification is its effects on mRNA stability . Therefore , we examined the role of m6A in regulating the mRNA stability expressed from several essential EBV genes . METTL14 is a major component of the methyltransferase complex required for m6A modification , and so we first determined binding of METTL14 to the transcripts modified with m6A . The target genes included 4 latent genes , EBNA1 , EBNA3C , LMP1 , LMP2A , and 4 lytic genes , BRLF1 , gp350 , BMRF1 and BGLF5 . For input samples , similar results were observed that there were relatively higher levels of latent transcripts compared to lytic gene transcripts according to the results of our real-time PCR assays in latently infected LcLs . We showed the RNA levels of transcripts in LcL input samples in Fig 2A and 2C ( S1B and S1C Fig ) , and also examined the corresponding DNA levels in IP samples and we did not find any DNA amplification before reverse transcription ( S1D Fig ) . We found that there was a substantial increase in enrichment of m6A at EBNA1 , EBNA3C , LMP1 , BRLF1 , gp350 , and BMRF1 , but not at LMP2A and BGLF5 with METTL14 immunoprecipitation ( Fig 2A and 2B ) . These results were consistent with the MeRIP-seq and m6A-qPCR data described above . To directly demonstrate the role of METTL14 in regulating these viral genes , we knocked down METTL14 followed by m6A RIP-qPCR . The results showed that m6A modification of EBNA1 , EBNA3C , LMP1 , BRLF1 , gp350 and BMRF1 in the absence of METTL14 was dramatically decreased in both LcLs and Akata cells ( Fig 2C–2F ) . We further investigated the viral mRNA stability after METTL14 knock down . The results showed that there were differences between the latent and lytic genes when METTL14 was knocked down . Consequently , the stability of the latent genes , including EBNA1 , EBNA3C , and LMP1 , was decreased in the absence of METTL14 ( Fig 2G ) . However , the stability of lytic genes , including BRLF1 , gp350 , and BMRF1 , was substantially increased in the absence of METTL14 as seen by the relative mRNA levels of the transcripts ( Fig 2H ) . and the stability of LMP2A and BGLF5 were not affected ( Fig 2G and 2H ) . The expression of METTL14 in the absence or presence of actinomycin D was monitored as a control . We found that the RNA level of METTL14 was decreased when the LcLs were treated with actinomycin D in both METTL14 knock-down cells and control cells ( S2 Fig ) . These results demonstrated that METTL14 played an important role in m6A modification of viral transcripts and that m6A modification played a different role for latent and lytic genes in terms of their stability . To investigate the role of specific m6A modification sites , we introduced 4 mutations to disrupt the DRACH motif ( D = A/G/U , R = A/G , and H = A/C/U ) [14] in the EBNA3C open reading frame ( S12A Fig ) . The sites were selected according to the MeRIP-seq ( Fig 1A and S1 Table ) and qPCR data ( Fig 1C ) . These residues are located at the amino-terminal region between G649 and C701 nucleotides where the highest enrichment of m6A modification within the EBNA3C transcript was identified . The mutations were designed to disrupt the DRACH motif where the adenosine is modified but would not change the underlying coding potential of the EBNA3C gene . The EBNA3Cwt and EBNA3Cmut constructs were transfected into Saos-2 cells and the expression of EBNA3C was detected using reverse transcription qPCR . We found that the levels of the EBNA3Cmut transcripts were lower than that of the wild-type EBNA3C ( Fig 2I ) . The m6A IP-qPCR experiment showed that there was less m6A modification of the EBNA3C mutant transcript of greater than 2-fold compared to that of the wild-type EBNA3C ( Fig 2J ) . To determine the effect of the mutations in EBNA3C , we used Actinomycin D to inhibit transcription for 6 and 12 hours . The RNA levels of EBNA3C was then measured . We found that there was a clear decrease in EBNA3C transcripts which contained the mutations ( Fig 2K ) . These results demonstrated that m6A modification of the EBNA3C transcripts played an important role in EBNA3C mRNA expression levels and stability . We identified a large number of m6A modifications in EBV transcripts , and more specifically a different profile of m6A modification during latent and lytic infection . Therefore , we wanted to determine whether m6A related enzymes were regulated on EBV infection . We first performed RT-qPCR to detect the expression of m6A modification enzymes including METTL3 , METTL14 , YTHDF1 , YTHDF2 , YTHDF3 , FTO and ALKBH5 in B cells , LcLs , Akata EBV negative and EBV-positive Akata cells . The RT-qPCR results showed that METTL14 was significantly up-regulated in EBV-positive LcLs and Akata cells ( S3 Fig ) . YTHDF2 expression showed a significant decrease in LcLs and a modest decrease in Akata cells while ALKBH5 expression was slightly increased in LcLs and Akata cells ( S3 Fig ) . We further examined the protein levels of METTL14 , YTHDF2 , and ALKBH5 since there were differences in their RNA expression . EBV negative B cells , Akata EBV- , BL41 and HEK293T cells , and latently infected LcLs , BL41-EBV and HEK293T-BAC-EBV cells were used for analysis of the protein levels by Western blot . The expression of the demethylase ALKBH5 was significantly decreased while the m6A methyltransferase METTL14 was dramatically increased in EBV-positive LcLs , BL41-B95 . 8 and there is a modest increase in Akata and 293T-BAC-EBV cells ( Fig 3A–3D , lane 1–2 and S4A–S4D Fig ) . These results suggested that the m6A methyltransferase and m6A modification are required for EBV latent infection . Further , the expression of METTL14 , YTHDF2 and ALKBH5 were also detected during lytic reactivation of infected cells . The data clearly showed that expression of METTL14 , YTHDF2 and ALKBH5 was down-regulated when EBV-positive cells were induced to lytic reactivation ( Fig 3A–3D , lane 3 , and S4A–S4D Fig ) . The expression of METTL14 , YTHDF2 and ALKBH5 was also monitored in several other EBV-infected cells . Latent Sav I , Sav III , Mutu I , and Mutu III as well as reactivated Sav I , Sav III , Mutu I , and Mutu III cells were used for analysis of the protein levels . BZLF1 was detected by western blot to confirm reactivation of EBV . We showed that METTL14 expression was increased in latency EBV III cells ( Mutu III and Sav III ) , compared to latency I cells ( Mutu I and Sav I ) ( Fig 3E and 3F , lane 1 and 3 ) . In addition , the expression levels of reader YTHDF2 and the demethylase ALKBH5 were decreased in latency III EBV cells compared to latency I EBV cells ( Fig 3E and 3F , lane 1 and 3 ) . The expression levels of METTL14 , YTHDF2 and ALKBH5 in lytically reactivated EBV-positive Sav I , Sav III , Mutu I and Mutu III cells were consistently down-regulated ( Fig 3E and 3F , lane 2 and 4 ) . This result suggests that essential viral antigens expressed during latency III may promote the expression of METTL14 and that its expression is reversed during lytic reactivation . METTL14-mediated m6A modification regulated the stability of EBV mRNA . To further examine the role of METTL14 , YTHDF2 and ALKBH5 in regulating the expression of EBV proteins , METTL14 , YTHDF2 and ALKBH5 were knocked down with lentivirus carrying shRNA in EBV-positive Burkitt lymphoma cells and LcLs . Knock-down of METTL14 led to a decrease in latent viral gene expression , including EBNA1 , EBNA3C and LMP1 ( Fig 3G and 3J , S4G , and S4J Fig ) . Expression levels of two lytic genes , BZLF1 ( immediately early ) and GP350 ( late ) were modestly induced by knock-down of METTL14 in EBV-positive cells ( Fig 3G and 3J ) , and EBNA1 , EBNA3C and LMP1 expression levels did not show any noticeable change in the YTHDF2 and ALKBH5 knock-down cells ( Fig 3H–3L , S4H–S4L Fig ) . However , BZLF1 expression was increased in YTHDF2 knock-down cells ( Fig 3H and 3K ) . EBV reactivation was induced by TPA and Butyric acid , and the expression of viral genes was monitored . In contrast to latently infected cells , expression of the viral proteins was seen to be significantly increased in reactivated cells ( Fig 3G–3L ) . Notably , latent gene expression was lower in METTL14 knock-down cells , but the lytic genes BZLF1 and GP350 were increased in METTL14 knock-down cells by at least 2-fold ( Fig 3G and 3J ) . These two lytic genes were also increased in the YTHDF2 knock-down cells ( Fig 3H and 3K ) . However , there were little or no obvious effects of ALKBH5 knock-down on viral antigen expression ( Fig 3I and 3L ) . We investigated the function of m6A modification in the virus lytic cycle progression and the production of infectious virions through knocking down METTL14 in LcLs . We found that knock-down of METTL14 alone will not lead to robust reactivation . However , knock-down of METTL14 will promote virion production when the virus entered the lytic cycle ( S5A Fig ) . These results demonstrated that METTL14 played a role in promoting EBV antigen expression during latent infection as well as lytic infection . Since METTL3 and METTL14 work as a methyltransferase complex . We further monitored the binding of METTL3 on the m6A sites of viral transcripts in LcLs and the results showed that METTL3 also bound to the m6A sites ( S5B Fig ) . We then investigated the effects of METTL3 on viral gene expression and we found that METTL3 knock-down showed a similar consequence for EBV protein expression as the METTL14 knock-down ( S5C Fig ) . Meclofenamic acid is reported to be a specific demethylation inhibitor [15] . To rule out any possible effects due to ALKBH5 shRNA artifacts , we treated LcLs with meclofenamic acid to support the results of the ALKBH5 knock-down . We found that the demethylation inhibitor showed similar effects as the ALKBH5 knock-down on viral gene expression ( S5D Fig ) . We further investigated which antigen of EBV was responsible for up-regulation of METTL14 . We transfected Myc-tagged EBNA2 , EBNA3C , LMP1 , LMP2A and LMP2B in Saos-2 cells ( S6A–S6G Fig ) . EBNA1 was not included as we observed a dramatic increase in METTL14 levels in type III EBV-positive latent cell lines compared to that of type I EBV-positive latent cell lines ( Fig 3E and 3F ) . 48 hours later , cells were collected and lysed . METTL14 and the transfected viral proteins were detected by western blot analyses . We found that EBNA3C was able to up-regulate METTL14 expression by approximately 2 . 5-fold ( S6G Fig ) . We repeatedly found that METTL14 was up-regulated in LcLs compared to EBV negative B cells . In addition , LMP1 upregulated METTL14 expression by approximately 24% at the protein level ( S6B Fig ) . We also showed that METTL14 expression levels were decreased by about 2-fold in EBNA3C knock-down LcLs ( S6H Fig ) , and further confirmed these results in EBNA3C stably expressed B cell lines BJAB7 and BJAB10 cells . METTL14 expression levels were increased in the presence of EBNA3C , when compared to EBNA3C negative B cells and was based on the levels of EBNA3C expressed ( S6I Fig ) . We also monitored the expression of METTL14 at the transcription level in EBV-positive and EBNA3C expressing cells . EBNA3C expression can up-regulate METTL14 transcription in BJAB7 and BJAB10 cells which stably expressed EBNA3C ( S6J Fig ) . We found that the RNA levels of METTL14 were also increased in EBV-positive cells and knock-down of EBNA3C in LcLs led to decreased expression of METTL14 ( S6K Fig ) . These results demonstrated that EBV infection , and specifically , EBNA3C expression can up-regulate METTL14 transcription . Further , we generated the METTL14 promoter-driven luciferase reporter system to determine whether EBNA3C was able to up-regulate METTL14 expression through activation of the METTL14 promoter . The reporter construct containing the METTL14 promoter and a dose-dependent increase of the Myc-EBNA3C expression vector was transfected into HEK293 or Saos-2 cells . Meanwhile , the thymidine kinase promoter-Renilla luciferase reporter plasmid ( pRL-TK ) was additionally transfected and used as a control for transfection efficiency . The results of the luciferase assay suggested that the METTL14 promoter activity was dramatically increased by EBNA3C in a dose-dependent manner ( S6L–S6M Fig ) . These results demonstrated that the essential EBV latent protein EBNA3C can up-regulate the methyltransferase METTL14 transcription through activation of the METTL14 promoter . We have shown that METTL14 transcripts and protein levels were significantly up-regulated by EBNA3C . To demonstrate whether EBNA3C can regulate METTL14 at the protein level , we initially performed protein stability assays by expressing METTL14 with or without EBNA3C in Saos-2 and HEK293 cells . Cells were treated with cycloheximide ( CHX ) for up to 12 hours . The protein level of METTL14 was decreased dramatically in the absence of EBNA3C ( Fig 4A and 4B ) . The results demonstrated that METTL14 was stabilized on expression of EBNA3C ( Fig 4A and 4B ) . We further validated the results by treating BJAB , BJAB7 , LcL-shCr and LcL-shE3C cells with CHX . The results showed that the stability of METTL14 was significantly enhanced in the presence of EBNA3C compared to a loss of up to 63% of the METTL14 signal at 12 hours after cycloheximide treatment in EBNA3C-negative cells ( y4C and 4D ) . We further examined whether EBNA3C can interact directly with METTL14 . Co-immunoprecipitation ( Co-IP ) assays were performed in B-cells with physiologic expression of EBNA3C . To determine the association in B-cell lines , BJAB , EBNA3C stably expressing BJAB ( BJAB7 ) and EBV-transformed ( LcL ) cells were used for the Co-IP experiment . We performed IP using an EBNA3C antibody or METTL14 specific antibody to immunoprecipitate the target proteins followed by western blot analyses . The Co-IP results showed that endogenous EBNA3C can physically associate with METTL14 and more importantly in EBV-transformed lymphoblastoid cell lines ( Fig 4E and 4F ) . To determine the domain of EBNA3C which can specifically interact with METTL14 , Myc-tagged full length and specific regions of EBNA3C ( 1-365aa , 366-620aa or 621-992aa ) were co-transfected into HEK293 cells with Flag-tagged METTL14 ( Fig 4I ) . The targeted protein was immunoprecipitated with anti-Myc or anti-Flag antibody , respectively . The results demonstrated that METTL14 interacted with an amino terminal region of EBNA3C ( 1-365aa ) as well as with the full-length EBNA3C protein ( 1-992aa ) ( Fig 4G and 4H ) . These results showed that EBNA3C amino acid residues 1-365aa were responsible for the interaction between EBNA3C and METTL14 ( Fig 4G and 4H ) . Plasmids with EBNA3C truncations were also transfected in HEK293 cells without Flag-METTL14 plasmid . IP was done and the results showed that there is no non-specific binding to Flag protein ( S7 Fig ) . To further determine the exact domain of EBNA3C which can specifically interact with METTL14 , Myc-tagged specific regions of EBNA3C ( 1-100aa , 100-200aa , or 200-300aa ) which delineated specific motifs in the amino terminal region were co-transfected into HEK293 cells with Flag-tagged METTL14 . Co-immunoprecipitation was performed with Myc antibodies . The results suggested that the region located within residues 100-200aa of EBNA3C containing the interaction domain for the transcription repressor RBP-JK was responsible for its interaction with METTL14 ( Fig 4J ) . To determine whether the EBNA3C residues EBNA3C100-200 or EBNA3CΔ100–200 ( EBNA3C without 100-200aa residues ) affect the protein expression , protein stabilities and the oncogene function of METTL14 , we expressed EBNA3C100-200 , EBNA3CΔ100–200 , or EBNA3C in HEK293 cells . We found that only the full-length EBNA3C can upregulate the expression of METTL14 . Neither the EBNA3C100-200 alone nor EBNA3C lacking the 100–200 residues up-regulated METTL14 expression ( S8A Fig ) . We also monitored the effects of EBNA3C100-200 and EBNA3CΔ100–200 on METTL14 protein stability . We found that neither the EBNA3C100-200 alone nor EBNA3C lacking the 100–200 residues promoted the stability of METTL14 ( S8B and S8C Fig ) . We have shown that EBNA3C binds to METTL14 , so it is expected that these two proteins can coexist in the same cellular compartment and colocalize with each other in cells . To determine the co-localization of EBNA3C and METTL14 , Saos-2 and HEK293 cells were initially transfected with plasmids expressing Myc-tagged EBNA3C , and the cellular localization of EBNA3C and METTL14 was monitored by fluorescence microscopy . In the absence of EBNA3C , the METTL14 proteins were distributed around the nucleus as seen with some punctate dots ( S9A Fig ) . In cells transfected with Myc-EBNA3C , the merged yellow fluorescence demonstrated that EBNA3C co-localized with METTL14 in Saos-2 and HEK293 cells ( S9B–S9D Fig ) . To further determine the co-localization of EBNA3C and METTL14 proteins in EBV-positive B-cells , immune-fluorescence assays were performed using specific antibodies against EBNA3C and METTL14 in order to examine the endogenous expression in different B-cell lines . The results corroborated the above results that EBNA3C co-localized with METTL14 in BJAB cells that can stably express EBNA3C , as well as in EBV-transformed LcLs ( Fig 5A–5D ) . This was consistent with the above results and further suggested that EBNA3C associated with METTL14 in nuclear complexes . We further investigated the expression and localization of METTL14 in EBV-negative and positive human tumor tissues . Immunohistochemistry for METTL14 expression showed that it was clearly up-regulated in EBV-positive Post-transplant lymphoproliferative disorders ( PTLDs ) tumor tissues and that EBNA3C signals were also clearly detected in the EBV-positive tissues compared to EBV negative tissues ( Fig 5 , compare panels 5E and 5F ) . We supported our co-localization data for METTL14 and EBNA3C in the PTLD tumor tissues with immunofluorescence assays . The results showed that EBNA3C co-localized with METTL14 in PTLD tumor tissues ( Fig 5 , compare panel 5G and 5H ) . To examine the effects of METTL14 on cell proliferation , Saos-2 and HEK293 cells were transfected with expression of EBNA3C and METTL14 and selected with G418 for two weeks . The expression of EBNA3C and METTL14 was confirmed by western blot ( Fig 6A and 6B ) . We monitored cell proliferation by expressing EBNA3C or METTL14 using CFSE staining assays . The results showed that both EBNA3C and METTL14 expression can promote enhanced cell growth ( Fig 6C ) . We further investigated the growth of colonies in monolayer cultures . Cells were seeded into 6-well plates and allowed to grow for 5 days to monitor formation of colonies . A significant increase in colony numbers was observed when EBNA3C and METTL14 were co-transfected compared to those transfected with only EBNA3C or METTL14 in both cell types ( Fig 6D and 6E ) . We performed overexpression of METTL14 , EBNA3C , and co-overexpression of EBNA3C and METTL14 in BJAB cells . And again , we found similar phenotypes of overexpression of METTL14 and EBNA3C in BJAB cells ( S10 Fig ) . These results demonstrated that METTL14 can promote growth of cells and that this ability was further enhanced by EBNA3C . We further monitored the role of METTL14 in cell growth in the context of EBV transformed LcLs . We performed CFSE staining and soft agar assays with LcL shCr , LcL shEBNA3C , LcL shMETTL14 and LcL shEBNA3C plus shMETTL14 . Knock-down of EBNA3C and METTL14 was confirmed by western blot ( Fig 6F ) . The results showed that knock-down of EBNA3C or METTL14 inhibited the ability of LcLs to form colonies in the soft agar assay ( Fig 6H and 6I ) . Furthermore , we observed the most prominent inhibition of colonies in LcLs with both EBNA3C and METTL14 knock-down ( Fig 6H and 6I ) . In support of these results , we observed a similar pattern of cell growth in Burkitt lymphoma BL41 cells with these genes knocked down ( Fig 6G and 6J ) . The results demonstrated that METTL14 enhanced cell proliferation and growth of colonies in EBV transformed cells by cooperating with EBNA3C . To investigate the functional role of METTL14 in EBV-associated tumorigenesis , we probed the effects of its knock-down in a xeno-transplant model in vivo . LcL shCr , LcL shMETTL14 , LcL shEBNA3C and LcL shMETTL14 plus shEBNA3C were subcutaneously injected into NOD-SCID mice to assess the effect of knock-down METTL14 or EBNA3C on tumor growth . The results demonstrated that knock-down of METTL14 or EBNA3C led to a dramatic slowing of tumor growth compared to the control group ( Fig 6K–6M and S11 Fig ) . Importantly , knock-down of both METTL14 and EBNA3C together led to the most dramatic decrease in tumor weight and volume ( Fig 6K–6M ) . Western blots to show levels of METTL14 and EBNA3C showed that METTL14 was knocked down greater than 50% and EBNA3C greater than 75% of the endogenous signals in the LcLs ( Fig 6M ) . We also investigated the functional role of specific m6A modification sites on EBNA3C for cell growth . We transfected the wild-type or EBNA3C mutant at the m6A sites into BJAB ( EBV negative cells ) and Raji cells ( EBV-positive but without EBNA3C expression ) [16] . Cell growth was monitored by CFSE staining and Soft agar assays . The results showed that mutant EBNA3C had a compromised capacity to promote cell growth when compared to the wild-type EBNA3C in both BJAB and Raji cells ( S12B–S12E Fig ) . These results confirmed a critical role for m6A modification of EBNA3C through regulation of METTL14 in driving the oncogenic phenotype . Finally , we investigated the function of EBNA3C100-200 and EBNA3CΔ100–200 in the oncogene function of METTL14 though Soft agar and CFSE staining assays . We found that neither the EBNA3C100-200 alone nor EBNA3C lacking the 100–200 residues promoted METTL14-mediated cell growth compared to full-length EBNA3C ( S8D–S8G Fig ) .
Modification of RNA by m6A functions extensively in cellular processes linked to RNA metabolism . These include mRNA stability , translation , splicing and RNA transport [17] . This modification is also involved in the self-renewal of cancer stem cells , promotion of cancer cell proliferation , and resistance to radiotherapy or chemotherapy [17] . Recent studies have provided clues as to understanding the roles of m6A modification during virus infection , however , the function of m6A modification in tumor virus-mediated oncogenesis is completely unknown and not previously explored . We have now mapped the m6A modification of viral transcripts during latent and lytic infection of EBV . We showed that m6A modification can play a major role in promoting latent gene expression and also repression of lytic gene expression . We also show that the level of methyltransferase METTL14 was up-regulated by the essential EBV latent antigen EBNA3C . Furthermore , EBNA3C promoted cell growth and proliferation by co-operating with METTL14 . We now establish a novel link between regulation of the EBV epitranscriptome and EBV-mediated oncogenesis ( Fig 6N ) . EBV latent infection allows the virus to persist in a mostly dormant state for the lifetime of the host and is associated with numerous cancers [18] . EBV protein expression in latency III potently drives B-cells to immortalization through regulation of cell growth , and promotion of cell survival . Increased expression of c-Myc can also be induced by EBV latent antigens , such as EBNA2 [19] , LMP2A [20] , and EBNA3C [21] . The ubiquitin ligase SCFSKP2 , cyclin D1 , cyclin A , c-Myc , MDM2 , p53 , CHK2 , E2F1 , and E2F6 are all directly regulated by EBNA3C [22–25] . Additionally , EBV latent proteins can play an essential role in epigenetic deregulation during B-cell lymphomagenesis [26 , 27] . EBV latent antigens are the major contributors to EBV-associated malignancies . Therefore , decreased expression of EBV latent antigens will result in attenuation of EBV-mediated tumorigenesis . Modification of RNA by m6A is catalyzed by a methyltransferase complex composed of METTL3 , METTL14 , and WTAP [17] , and mRNAs containing m6A modification can be recognized by YTH domain family members , YTHDF1 , YTHDF2 and YTHDF3 [28] . The consequence of m6A modification differs and depends on being recognized by the readers , and the position of the m6A modification [29] . YTHDF1 is also known to enhance translation of mRNA [30] . YTHDF2 is now known to induce mRNA degradation [31] , and YTHDF3 co-operates with either YTHDF1 or YTHDF2 to promote translation or degradation of transcripts , respectively [32] . More recently another family of m6A readers , insulin-like growth factor 2 mRNA-binding proteins ( IGF2BPs; including IGF2BP1/2/3 ) have been shown to up-regulate mRNA stability [33] . IGF2BPs belong to a conserved family of single-stranded RNA binding proteins [33] , and can recognize m6A modification to enhance mRNA stability and translation [33] . IGF2BPs and YTHDF2 show a different distribution on the transcripts and recognize m6A modification with different motif sequences [33] . Nevertheless , new m6A readers are emerging and there must be more unrevealed readers to be explored . We found that m6A modifications are distributed on both latent and lytic gene transcripts . M6A modification at the viral latent and lytic gene transcripts resulted in different outcomes . Modification of latent transcripts promoted stability while it had a negative role on stability of lytic transcripts . This difference may be due to the different readers responsible for recognition of the modification . Additionally , the enhanced stability of the latent transcripts , such as EBNA3C , had a direct effect on promoting cell growth and proliferation . METTL3-METTL14-WTAP-mediated m6A modification is associated with cancer [5] and the relationship between m6A and cancer has not been fully explored . Silencing of METTL3 led to P53 signaling pathway enrichment in HepG2 cells [5] , and knock-down of METTL3 or METTL14 promoted glioblastoma stem cell growth and tumorigenesis [34] . METTL14 is highly expressed in normal hematopoietic stem/progenitor cells and acute myeloid leukemia cells [35] . METTL14 is required for myeloid leukemia cell survival and proliferation [35] . We found that METTL14 expression was dramatically up-regulated in EBV-infected cells , and EBNA3C positive cells . We also showed that METTL14 promoted growth and proliferation of these cells . Importantly , we showed upregulation of METTL14 in EBV-positive PTLDs and knock-down of METTL14 resulted in a decreased tumorigenic activity of EBV-transformed cells in the xenograft animal model systems . Studies in KSHV revealed that blocking m6A inhibits splicing of the pre-mRNA encoding replication transcription activator ( RTA ) , a key KSHV lytic switch protein , and halts viral lytic replication [36] . Knockdown of YTHDF1 , YTHDC1 or YTHDC2 had no significant or consistent effect on viral lytic replication [12] . Knock-down of YTHDF2 led to a four-fold increase in KSHV production and an increase in RTA , ORF57 , ORF-K8 and ORF65 lytic transcripts through an increase of the half-life of KSHV transcripts . Another study found that depletion of the m6A machinery had differential pro- and anti-viral impacts on viral gene expression depending on the cell-type analyzed [37] . Here we focused on METTL14 , a component of methyltransferase complex and found that METTL14 can promote increased levels of the latent oncogene EBNA3C . METTL3-METTL14 mediated m6A modification can negatively regulate EBV lytic gene expression , but knock-down METTL14 cannot drive EBV into full-blown lytic reactivation from latency . Although we focused on the relationship between m6A modification and EBV-mediated tumorigenesis , the effects of METTL3-METTL14 mediated m6A modification on lytic gene expression is an interesting topic for further study . METTL14-mediated m6A modification can negatively regulate mRNA stability of some lytic genes according to our initial results . Correspondingly , METTL14 expression is dramatically downregulated during the EBV lytic infection cycle . These results suggest that m6A modification is likely not absolutely a necessary requirement for EBV lytic reactivation . The detailed mechanism related to EBV reactivation and m6A modification will need to be further explored . In this study , we have now shown for the first time that the host RNA m6A modification machinery can be targeted and manipulated specifically by a viral-encoded oncoprotein . This modulated latent and lytic viral gene expression enhancing the pro-oncogenic potential of EBV . The study now provides a deeper understanding of specific interactions between a potent oncogenic virus and host cells at the level of the viral epitranscriptome , and further suggests a critical role for viral mRNA modification in driving the oncogenic process . Furthermore , targeting METTL14 may be a critical strategy for controlling EBV-associated cancers .
The University of Pennsylvania Institutional Animal Care and Use Committee reviewed and approved all the animal experiments ( protocol 804549 ) . All the experiments were carried out in accordance with the Laboratory Animal Welfare guidelines for the Care and Use provided by National Institutes of Health . The University of Pennsylvania CFAR Immunology Core provided B-cells from deidentified humans . The CFAR Immunology Core maintains an Institutional Review Board ( IRB ) -approved protocol in which Declaration of Helsinki protocols are strictly followed . Every donor gave written and informed consent at the CFAR Immunology Core . All patients provided written informed consent . The plasmid pA3M-EBNA3C and the LcLs , Akata , and BJAB cell lines , have been described previously [4] . LcLs were generated in our laboratory [4] . BJAB , BL41 , Akata , Akata-EBV , and BL41-B95 . 8 cell lines were kindly provided by Elliott Kieff ( Harvard Medical School , Boston , MA ) . Saos-2 ( human osteosarcoma cell line ) and HEK-293 ( human embryonic kidney cell line ) were provided by Jon Aster ( Brigham and Woman's Hospital , Boston , MA ) . Sav I and Sav III cell lines , Mutu I and Mutu III cell lines were kindly provided by Dr . Paul M . Lieberman ( The Wistar Institute , Philadelphia , PA ) . Antibodies against Myc ( 9E10 ) and EBNA3C antibody ( A10 ) was generated from hybridomas . A rabbit anti-Flag monoclonal antibody was purchased from Sigma-Aldrich Corp . ( St Louis , MO ) . Rabbit anti-METTL14 , YTHDF2 and ALKBH5 antibodies were purchased from Proteintech Group Inc . METTL14 or EBNA3C knocked down cells were transduced with lentivirus containing specific shRNAs and shRNAs against luciferase were used as a control . The knocked down cell lines were then selected with 1ug/ml puromycin . Isolation of m6A-containing fragments was performed as previously described with minor modifications [38] . Briefly , total RNA was extracted from cells using TRIzol Reagent ( Invitrogen , Inc . , Carlsbad , CA ) . The total RNA was fragmented in a buffer containing 100 mM Tris-HCl at pH 8 . 0 and 100 mM ZnCl2 followed by incubation at 94°C for 8 min . Successful fragmentation of mRNA with sizes close to 120 nucleotides was validated using a BioRad Geldoc Station ( Bio-Rad , Hercules , CA ) . Before immunoprecipitation , 10 μg of anti-m6A antibody was incubated with 30 μl slurry of Pierce Protein A Agarose beads ( Thermo Fisher Scientific , Waltham , MA ) in 250 μl PBS with 0 . 5% BSA at 4°C for 2 h . The beads were washed three times in cold PBS with 0 . 5% BSA . To isolate the m6A-containing fragments , 900 μg of fragmented total RNA was added to the antibody-bound beads in 250 μl IP buffer supplemented with RNasin Plus RNase inhibitor ( Promega Inc , Madison , WI ) , and the mixture was mixed at 4°C for 2h . The beads were washed four times with 1 ml IP buffer before elution with 100 μl IP buffer supplemented with 6 . 67 mM of m6A salt ( M2780 , Sigma-Aldrich , St . Louis , MO ) . The mixture was incubated for 1 h at 4°C with continuous shaking and the eluate was collected . A second elution was carried out and the eluates were pooled together before purification with 2 . 5-fold ethanol and 1/10 volume of NaAC pH5 . 2 . RNA was washed with 75% ethanol for twice and dissolved in RNase free H2O . Purified eluate and Input samples were used for the preparation of libraries and sequencing at the Genome Sequencing Facility of the GATC core at the Washington University St . Louis . Approximately 300 ng of mRNA was used for RNA sequencing ( RNA-Seq ) . The library was prepared using the TruSeq stranded mRNA kit ( Illumina , San Diego , CA ) according to the manufacturer’s protocol . First , the elute-frag-prime stage was done at 80°C for 2 min to allow annealing without causing fragmentation [12] . RNA was reverse transcribed into first strand cDNA using reverse transcriptase and random primers . This was followed by the second strand cDNA synthesis using DNA Polymerase I and RNase H . The cDNA fragments were used for end repair process with the addition of a single ‘A’ base followed by ligation of adapters . The products were then purified and enriched by PCR amplification for 12 cycles to generate the final RNA-seq library . cDNA libraries were quantified and pooled and subsequent sequencing on an Illumina HiSeq3000 platform 50 bp single read sequencing module . After sequencing , the first step is quality and adapter trimming . Trimming was conducted by Trim Galore ( https://www . bioinformatics . babraham . ac . uk/projects/trim_galore/ ) to remove adapter sequences and low-quality bases ( bases with < 20 quality score will be removed ) . After trimming , the MeRIP-seq reads were aligned to EBV reference genome by the aligner HISAT2 [39] with the annotation of the splice sites ( the EBV reference genome and annotation were downloaded from https://www . ncbi . nlm . nih . gov/nuccore/NC_007605 . 1 ) . The m6A peaks were called by a graphical model-based peak calling method–MeTPeak [40] , with the parameters of window width = 50 , sliding step = 50 and read length = 50 . The peaks were visualized in IGV ( http://www . broadinstitute . org/igv ) . RIP was performed as previously described [33] . Briefly , 107 LcLs or Akata cells were collected and lysed with RIP buffer ( 10 mM Tris-HCl pH 7 . 4 , 150 mM NaCl , 5mM EDTA , 0 . 5mM DTT , 0 . 5% NP40 , 100 U/ml RNAase inhibitor SUPERase•in , 1×protease inhibitor ( 539131 , Millipore , Burlington , MA ) . 3μg of METTL14 ( Proteintech Group Inc , Rosemont , IL ) or IgG ( NI01 , Millipore , Burlington , MA ) was conjugated to protein A/G agarose beads ( Thermo Fisher Scientific , Waltham , MA ) by incubation for overnight at 4°C in RIP buffer at 4°C overnight . Beads were washed with RIP buffer for three times , followed by DNA digestion at 37°C for 30 min and incubation with 50μg of proteinase K ( Thermo Fisher Scientific , Waltham , MA ) at 55°C for 30min . Input and co-immunoprecipitated RNAs were recovered by TRIzol , extraction and analyzed by qPCR . Saos-2 and HEK293 cells were transfected with jetPRIME ( Polyplus Transfection , Illkirch , France ) according to the manufacturer’s instructions . BJAB cells were transfected by electroporation ( 220V , 950 μF ) with Bio-Rad Gene Pulser II electroporator in 400 ul of serum-free medium . The cells were then transferred to complete RPMI 1640 media , which was preincubated to 37°C . Immunoprecipitation ( IP ) and Western blotting were performed as described previously [41] . Briefly , cells were collected and were lysed in lysis buffer ( 10 mM Tris , 1% NP-40 , 2 mM EDTA , 150 mM NaCl [pH 7 . 5] ) with protease inhibitors . For IP , lysates were incubated with the antibodies indicated in the figures and 30 μl of a 1:1 mixture of protein A/G Sepharose beads at 4°C overnight . The beads were washed with RIPA buffer for 3 times , boiled and were subjected to SDS-PAGE for Western blotting . NOD . CB17-PrkdcSCID ( NOD/SCID ) male mice were purchased from Jackson Laboratory ( Jackson Labs , Bar harbor , ME ) at 6 weeks of age and 6 million cells were subcutaneously injected . 3 weeks later , mice were sacrificed and the weight and volume of the tumor were measured . Immunohistochemistry was performed as described previously [42] . Tissues were embedded in paraffin and sectioned . The slides were treated at 60°C for 60 min and then deparaffinized in xylene and a gradient concentration of ethanol . Antigen was retrieved in boiling Tris buffer ( pH 9 . 0 ) for 18 min . The slides were immersed in 3% H2O2 for 10 min , and permeabilized with 0 . 5% Triton X-100 at RT for 10 min . The slides were blocked in 5% bovine serum albumin ( BSA ) at RT for 30 min and incubated overnight with diluted primary antibody at 4°C , and then diluted HRP-labeled secondary antibody was added at room temperature for 60 min . After development with diaminobenzidine for 3 min , the slides were washed in water and counterstained with hematoxylin . For immunofluorescence , cells were seeded on glass coverslips in 24-well plates before transfection . After treatment , cells were fixed with 4% paraformaldehyde at 4°C for 60min and permeated with 0 . 2% Triton X-100 in PBS for 10min . Nuclei were visualized by staining with DAPI for 2 min . Images were acquired using a Fluoview FV300 confocal microscope and Fluoview software was used for image analysis . HEK293T cells were co-transfected with pGL4 . 1 plasmid with METTL14 promoter , pRL-TK ( Promega , Madison , WI , USA ) , Myc-tagged EBNA3C , or empty plasmids . Cells were harvested and the dual-luciferase reporter assay was performed according to the manufacturer's instructions ( Promega , Madison , WI , USA ) 48 hours after transfection . Luciferase value was read using a Cytation 5 ( BioTek , Winooski , VT , USA ) . The total RNA extraction was performed using Trizol reagent ( Invitrogen , Inc . , Carlsbad , CA ) and treated with DNase I ( Invitrogen , Inc . , Carlsbad , CA ) . cDNA was prepared with Superscript II reverse transcriptase kit ( Invitrogen , Inc . , Carlsbad , CA ) according to the manufacturer’s protocol . Quantitative real-time PCR analysis was performed by using SYBR green real-time master mix ( MJ Research Inc . , Waltham , MA ) . The primers used are listed in S5 Table . Cells were collected and suspended in 1XPBS at a concentration of 1 million cells/ml . The CFSE solution was added to make a final concentration of 5μM . An equal volume of 1XPBS containing 5% FBS was added after 10 min incubation at room temperature . Cells were washed three times with 1XPBS containing 5% FBS and equally divided into several plates for incubation . 3 days later , cells were harvested , washed with ice-cold 1XPBS and resuspended in 5ml 1XPBS , then run on FACS Calibur cytometer ( Becton-Dickinson Inc . , San Jose , CA ) followed by analysis with FlowJo software ( Treestar , Inc . , San Carlos , CA ) . HEK293 or Saos-2 cells were transfected with control vector , Myc-EBNA3C or Flag-METTL14 and allowed to grow in DMEM supplemented with 1mg/ml G418 ( Sigma-Aldrich , St . Louis , MO , USA ) . After two weeks of selection , 105 cells were seeded in 6-well plates and allowed to grow for 5 days . Cells were stained with 0 . 005% Crystal Violet overnight and scanned by PhosphorImager ( Molecular Dynamics , Piscataway , NJ ) and the area of the colonies measured by using Image J software ( Adobe Inc . , San Jose , CA ) . Three independent experiments were performed . The soft agar assays were performed using LcLs and Burkitt lymphoma cells . Briefly , 1 ml of 0 . 5% agar in RPMI media was poured into 6-well plates and set aside to solidify . 0 . 5 ml 0 . 3% agar/medium containing 2×105 cells were added to the plates as the middle layer . Then cells were covered with a top layer of another 1ml 0 . 5% agar/medium . After two weeks , colonies were stained with 0 . 005% crystal violet for 1 hour and scanned using a Licor Odyssey system ( LiCor Inc . , Lincoln , NE ) . The number of colonies was counted using ImageJ software ( Adobe Inc . , San Jose , CA ) . Each experiment was repeated at least three times . The mean scores were examined by using Student’s t-test . All statistical tests were performed using Microsoft Office Excel . A P-value of 0 . 05 was considered to be a statistically significant difference . A P-value of 0 . 01 was considered to indicate highly statistical significance . | Epstein–Barr virus ( EBV ) is a ubiquitous oncogenic virus that contributes to approximately 2% of all cancers by modulating a myriad of host cell activities . M6A is the most abundant RNA modification which is important for infection with HIV-1 , HCV , Zika virus , KSHV and SV40 virus . The role of m6A modification and the associated enzymes in EBV infection and related oncogenic activities has not been explored . We investigated the role of m6A modification on EBV-infection and examined its effects on the proliferative activities of the virus which contributes to cancer . We showed that the m6A methyltransferase METTL14 is an important factor in EBV-induced oncogenesis . METTL14 was dramatically increased in EBV latently infected cells . Knock-down of METTL14 led to decreased latent viral gene expression . The essential EBV latent antigen EBNA3C was up-regulated by METTL14-mediated m6A modification , and its expression led to a feedback loop in up-regulation of METTL14 transcription , and the stability of its protein . We now provide a more comprehensive understanding of EBV and the host cell RNA processing machinery critical for regulation of EBV activities to maintain latent infection and its oncogenic properties . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"pathogens",
"messenger",
"rna",
"immunology",
"microbiology",
"cell",
"processes",
"enzymology",
"viruses",
"oncology",
"immunoprecipi... | 2019 | EBV epitranscriptome reprogramming by METTL14 is critical for viral-associated tumorigenesis |
Human monkeypox ( MPX ) occurs at appreciable rates in the Democratic Republic of Congo ( DRC ) . Infection with varicella zoster virus ( VZV ) has a similar presentation to that of MPX , and in areas where MPX is endemic these two illnesses are commonly mistaken . This study evaluated the diagnostic utility of two surveillance case definitions for MPX and specific clinical characteristics associated with laboratory-confirmed MPX cases . Data from a cohort of suspect MPX cases ( identified by surveillance over the course of a 42 month period during 2009–2014 ) from DRC were used; real-time PCR diagnostic test results were used to establish MPX and VZV diagnoses . A total of 333 laboratory-confirmed MPX cases , 383 laboratory-confirmed VZV cases , and 36 cases that were determined to not be either MPX or VZV were included in the analyses . Significant ( p<0 . 05 ) differences between laboratory-confirmed MPX and VZV cases were noted for several signs/symptoms including key rash characteristics . Both surveillance case definitions had high sensitivity and low specificities for individuals that had suspected MPX virus infections . Using 12 signs/symptoms with high sensitivity and/or specificity values , a receiver operator characteristic analysis showed that models for MPX cases that had the presence of ‘fever before rash’ plus at least 7 or 8 of the 12 signs/symptoms demonstrated a more balanced performance between sensitivity and specificity . Laboratory-confirmed MPX and VZV cases presented with many of the same signs and symptoms , and the analysis here emphasized the utility of including 12 specific signs/symptoms when investigating MPX cases . In order to document and detect endemic human MPX cases , a surveillance case definition with more specificity is needed for accurate case detection . In the absence of a more specific case definition , continued emphasis on confirmatory laboratory-based diagnostics is warranted .
Since the global eradication of smallpox , the most important Orthopoxvirus infection in humans in terms of ongoing numbers of cases , morbidity , and mortality has been human monkeypox ( MPX ) [1 , 2] . Monkeypox virus ( MPXV ) is maintained by an enzootic cycle , with zoonotic introductions to humans often being followed by more limited human-to-human transmission [3 , 4] . The animal reservoir for MPXV remains unknown , but the virus has been isolated in the wild from a squirrel ( Funisciurus anerythrus ) and a sooty mangabey ( Cercocebus atys ) [5 , 6] . Infection with MPXV can lead to a smallpox-like illness characterized by a febrile prodrome , lasting 1–4 days , followed by a slowly progressing rash . The rash proceeds from macules to papules to vesicles to pustules to crusts and finally to desquamation . This occurs over a period of two to four weeks . At any given site on the body , the rash is generally in the same stage of development ( e . g . , all vesicles ) , and the lesions are typically circumscribed , umbilicated , deep-seated and firm . The rash has a centrifugal distribution , with a concentration of lesions on the extremities and face . As with smallpox , MPX lesions often appear on the palms of the hands and soles of the feet . Lymphadenopathy ( inguinal , axillary , and/or cervical ) is common in MPX patients and can occur prior to or at the onset of rash . Ocular infection with MPXV can lead to permanent corneal scarring and blindness [7–9] . Infection with varicella zoster virus ( VZV ) has a similar presentation to that of MPXV infection , and in areas where MPXV is endemic these two illnesses are commonly mistaken [10 , 11] . However , there are several features of illness that typically set one infection apart from the other . For example , VZV patients typically exhibit a short , mild period of febrile prodrome , or none at all , followed by a quickly evolving ( 1–2 day ) pleomorphic rash ( i . e . , a rash for which neighboring lesions may be in different stages of development ) . VZV lesions also often have irregular borders , and are superficial on the surface of the skin ( relative to those of MPX ) . In addition , varicella lesions often appear in a centripetal distribution on the body [9 , 12] . Although noted in rare occurrences , lesions on the palms of the hands and soles of the feet are not hallmarks of VZV infections [13] . VZV patients do not typically have pronounced lymphadenopathy , and , thus , the presence of lymphadenopathy is one distinguishing characteristic that can differentiate MPX from both smallpox and varicella . Additional illnesses that can be mistaken for MPX are other herpetic infections ( in addition to VZV ) , drug eruptions , syphilis , yaws , scabies , and rickettsialpox [9] . Specimen collection followed by laboratory testing can be difficult to accomplish for all suspected cases in MPX endemic areas . A clinical case definition capable of enhancing the distinction between MPX and other illnesses would be useful to enable more accurate and expedient case detection , collection of higher-quality surveillance data , and improved patient management . MPXV is enzootic in western and central Africa , with the overwhelming majority of human infections reported each year from the forested areas of the Congo Basin of Democratic Republic of Congo ( DRC ) [2] . MPX is a nationally reportable disease in DRC and has been identified as one of the country’s priority diseases of epidemic potential . On a bi-weekly basis , notifications of suspected MPX cases from each of the country’s Health Zones are submitted to national public health authorities; few of the suspected cases are formally investigated ( i . e . , case investigation forms completed and diagnostic specimens collected ) . This study evaluates the diagnostic utility of two surveillance case definitions for MPX . Both definitions were herein applied to a cohort of suspected MPX cases that were identified over the course of a 42 month period via surveillance in one Province in DRC . This cohort is unique in that the dataset contained an in-depth list of signs/symptoms . The accuracy of case classification was determined using laboratory findings . We assessed clinical features of illness in patients with confirmed MPXV infection to identify characteristics distinctly associated with disease presentation and suggest modifications to the MPX surveillance case definition to improve specificity .
Data from suspect MPX cases were obtained by investigation in accordance with national guidelines . Patients were identified as suspect MPX cases if they had a vesicular or pustular eruption with deep-seated , firm pustules and at least one of the following symptoms: fever preceding the eruption , lymphadenopathy ( inguinal , axillary , or cervical ) , and/or pustules or crusts on the palms of the hands or soles of the feet . For suspect cases , a MPX-specific case investigation form was completed and , in most instances , two or more diagnostic specimens were collected from each suspected MPX case . The specimens were sent to the Institut National de Recherche Biomédicale ( INRB ) in Kinshasa for diagnostic testing . One specimen from each individual was tested at INRB for the presence of Orthopoxvirus DNA signatures using a real-time PCR assay [14] . If the initial PCR was negative for Orthopoxvirus , a second real-time PCR assay specific for VZV-specific DNA signatures was conducted ( reagents provided by the United States Army Medical Research Institute of Infectious Diseases ) . DNA extracted at INRB and additional independent specimens , if available , were shipped to the Poxvirus Laboratory at the U . S . Centers for Disease Control and Prevention ( CDC ) . At the CDC , DNA was extracted from original specimens and all specimens were tested with MPXV and VZV-specific real-time PCR assays [15 , 16] . An individual was classified as a laboratory-confirmed MPX case if at least one specimen was 1 ) positive with the Orthopoxvirus-specific assay , and/or 2 ) positive by MPX-specific real-time PCR . An individual was independently classified as a laboratory-confirmed VZV case if a crust specimen tested positive for VZV DNA signatures at INRB or if an original vesicular swab or crust specimen tested positive for VZV at CDC . These activities were determined to not be research by a CDC human subjects advisor . Suspect MPX cases in Tshuapa Province , DRC , with rash onset occurring between September 2009 and February 2014 were included in the analysis ( N = 1025 , Fig 1 ) . These individuals were all assessed by a surveillance officer , who determined the individual met the surveillance case definition for a suspect MPX case . Individuals whose laboratory test results were suggestive of a coinfection with both MPXV and VZV , and those with incomplete or inconsistent laboratory results were excluded from analyses ( n = 273 ) . A total of 752 cases were included for further analyses; 333 laboratory-confirmed MPX cases , 383 laboratory-confirmed VZV cases , and 36 cases that were determined to not be either MPX or VZV ( laboratory diagnosis undetermined ) . Clinical signs and symptoms were recorded as checkboxes—‘yes’ , ‘no’ , and ‘do not know’—on the case investigation form . If an individual case investigation form had a response ( ‘yes or ‘no’ ) for any signs/symptoms , but the absence of a response for another specific sign/symptom the variable was coded as “missing” for the specific sign/symptom without a response . Febrile prodrome status was determined by either a ) selection of ‘yes/no’ for the febrile prodrome variable on the case investigation form , or , if that information was missing , by , b ) presence of prodrome was ascertained by determining the time interval between onset of fever and that of rash , when the information was available . The presence of lymphadenopathy was determined using the individual lymphadenopathy-type fields ( inguinal , axillary , cervical ) . If at least one category of lymphadenopathy was reported as present , then the individual was coded “yes” for lymphadenopathy; if all fields were “no” , the individual was coded “no” for lymphadenopathy . If one or more of the fields were missing and others were “no” , the individual was coded “missing” for lymphadenopathy . Two surveillance case definitions were evaluated in this study . According to Case Definition A , a suspect case is an individual with fever followed by a vesicular or pustular rash with the following symptoms: rash on palms , soles , and face; or the presence of 5 variola-like scars . Case Definition A has been recommended for use in endemic areas . Case Definition B was developed as a discriminatory case definition with the inclusion of several criteria that distinguish MPX from VZV . Case Definition B encompasses individuals who have a vesicular or pustular eruption with deep-seated , firm pustules and at least one of the following symptoms: fever preceding the eruption , lymphadenopathy ( inguinal , axillary , or cervical ) , and/or pustules or crusts on the palms of the hands or soles of the feet . To evaluate the diagnostic accuracy of each case definition , the dataset was restricted to individuals who had all the information needed for classification using both case definitions ( i . e . , one dataset was used for the independent analyses of both definitions ) . A total of 645 of the 752 cases ( 85 . 8% ) had sufficient information to be included in the analyses of both case definitions; 314 laboratory-confirmed MPX cases , 305 laboratory-confirmed VZV cases , and 26 cases that were determined to be neither MPX nor VZV ( Fig 1 ) . Individuals met the criteria for Case Definition A if they had a febrile prodrome and either the presence of a ) rash/scars on the face , palms , and soles , or b ) more than 5 scars . Individuals met the criteria for Case Definition B if they had rash with deep-seated , firm lesions and either a ) febrile prodrome , b ) lymphadenopathy , or c ) lesions on the palms of the hands or soles of the feet . This dataset was also used for the receiver operating characteristic analysis . The frequencies of each sign/symptom were calculated for all cases included in the dataset , and individually for laboratory-confirmed MPX and VZV cases . Associations between reported signs/symptoms and laboratory determined diagnoses were calculated using chi-squared and Fisher exact tests . Signs/symptoms that occurred with significantly different frequency ( p<0 . 05 ) between laboratory-confirmed MPX and VZV cases were further assessed for their individual sensitivity , specificity , PPV and NPV in relation to the confirmed diagnosis of MPX . For the analysis of the two case definitions , real-time PCR diagnostic test results ( for MPX and VZV ) were used as the ‘gold standard’ to establish MPX and VZV diagnoses . The sensitivity , specificity , positive predictive values ( PPV ) , and negative predictive values ( NPV ) were computed for the two case definitions . A receiver operating characteristic ( ROC ) analysis was completed using ‘fever before rash’ and various summed frequencies of the 12 signs/symptoms that were identified as having high ( >80% ) sensitivities or specificities for laboratory-confirmed MPX cases . For example , individuals with ‘fever plus rash’ and one of these 12 signs/symptoms were categorized as having ‘1 criteria’; individuals with ‘fever plus rash’ and 12 signs/symptoms were ranked as ‘12 criteria’ . Each of these 12 signs/symptoms counted equally to the sum . For example , if an individual reported two symptoms of ‘nausea’ and ‘cough’ , they were categorized the same as an individual who reported two symptoms of ‘fatigue’ and ‘conjunctivitis’ . ‘Fever before rash’ was identified as a mandatory sign/symptom because this has been consistently noted in the literature for MPX patients and also was observed at a frequency of 98 . 1% for all suspect cases and 99 . 1% for laboratory-confirmed MPX patients in the present analysis . The sensitivity , specificity , PPV and NPV for each model ( with increasing number of signs and symptoms from 1 to 12 ) were assessed . All data analysis was performed using SAS version 9 . 3 .
Characteristics of the Population: A total of 752 suspect cases were included in the analysis . Approximately 53% of suspect cases were male and 46% were female; this proportion remained consistent after laboratory case classification . The mean age of suspect cases was 17 years ( median 13 , range 0 . 01–86 ) . Of these suspect cases , 333 ( 44 . 3% ) individuals were classified as laboratory-confirmed MPX cases and 419 ( 55 . 7% ) were classified as laboratory-confirmed VZV ( 383 ) or undiagnosed ( 36 ) cases ( Table 1 ) . Performance of specific clinical characteristics: With the objective of improving the specificity and PPV of the case definitions under examination , associations between the reported clinical signs/symptoms and a laboratory-confirmed diagnosis of MPX ( versus VZV ) were investigated . Significant ( p<0 . 05 ) differences between laboratory-confirmed MPX and VZV cases were noted for the signs/symptoms of nausea , cough , lymphadenopathy ( overall and each site ) , mouth ulcers , sore throat , malaise , fatigue , conjunctivitis , sensitivity to light , and bedridden ( Table 2 ) . Rash characteristics that were significantly different included same size , deep-seated , firm lesions , and the presence of lesions on the arms , legs , palms of the hands , soles of the feet , and genitals . Each of the significant signs/symptoms and rash characteristics occurred more frequently in laboratory-confirmed MPX cases than in laboratory-confirmed VZV cases . The majority of significant signs/symptoms ( 15/20 ) occurred in more than 50% in laboratory-confirmed MPX cases . Variables with high sensitivity for MPX ( ≥80% ) were lymphadenopathy , fatigue , and the following rash characteristics: same size , deep-seated firm lesions , presence on the arms , legs , palms of the hands , and soles of the feet ( Table 3 ) . Nausea , conjunctivitis , bedridden , and lesions present on the genitals were signs/symptoms with a high specificity ( ≥80% ) , but none of these , individually , were found in laboratory-confirmed MPX cases at a frequency > 32% . Analysis of the case definitions: Two-by-two tables and diagnostic values for the two case definitions are presented in Tables 4 and 5 . Two hundred ninety-one ( 92 . 6% ) laboratory-confirmed MPX cases satisfied Case Definition A; and 306 ( 97 . 5% ) laboratory-confirmed MPX cases satisfied Case Definition B; 245 ( 74% ) and 303 ( 91 . 5% ) non-MPX cases [laboratory-confirmed VZV cases and undiagnosed ( MPX and VZV negative ) cases] satisfied Case Definitions A and B , respectively . The sensitivity of both case definitions was high , with the value for the Case Definition B a bit higher than that of Case Definition A ( 97 . 45% vs . 92 . 86% , respectively ) . Similarly , the specificity of both case definitions was low and the Case Definition A had a higher specificity ( 25 . 98% ) than Case Definition B ( 8 . 46% ) . The PPVs were similar for both definitions ( 50 . 25% for Case Definition B and 54 . 92% for Case Definition A ) , as were the NPVs ( 77 . 78% for Case Definition B and 78 . 90% for Case Definition A ) . The NPV was higher than the PPV for both definitions . Receiver operating characteristic analysis: Using the 12 identified signs/symptoms with high sensitivity and/or specificity values , the ROC analysis tested the performance and accuracy of 12 models ( with increasing numbers of signs and symptoms ) ( Table 6 ) . In general , models with a greater number of signs/symptoms ( >8 but <12 ) demonstrated excellent specificity ( >90% ) but low sensitivity ( <40% ) . In contrast , models with a lower number of signs/symptoms ( <7 ) had excellent sensitivity ( >90% ) but low specificity ( <40% ) . The models for MPX cases that had the presence of ‘fever before rash’ plus at least 7 or 8 of the 12 signs/symptoms demonstrated a more balanced performance between sensitivity and specificity . There was greatly improved specificity for the models that included 7 ( 50 . 76% ) or 8 ( 70 . 69% ) signs/symptoms when compared to either Case Definition A ( 25 . 98% ) or B ( 8 . 46% ) . The area under the curve for the model using these summed symptom counts was 0 . 74 ( Fig 2 ) .
The choice and utility of a case definition will be guided by its intended use . Case Definition A was designed to detect a single case of MPX illness , followed by an immediate outbreak response and control efforts . Case Definition B was designed to be a discriminatory definition used in the context of surveillance for disease , to better understand the extent and burden of disease in an endemic area . Both case definitions were characterized by high sensitivities but very low specificities . The high values for sensitivity were expected since the dataset represented patients who were diagnosed with suspected MPX virus infection ( prior to laboratory confirmation ) . These characteristics indicate that both definitions are useful for screening purposes and are well-designed for outbreak detection . Given that MPX is an endemic , regularly occurring , reportable disease in Tshuapa Province , the case definitions should be sufficient to capture true MPX cases , such that local or national officials may want to launch an outbreak response if they observe an aberration or threshold in the number or rate of reported infections . For surveillance purposes , however , especially in resource-limited countries such as DRC , it is necessary for a definition to capture all true cases and at the same time minimize the number of false positives . Attributes of Case Definition B , including a low specificity and moderate PPV , are not optimal for the objectives of disease surveillance . Observations of a moderate PPV and high NPV for both case definitions is consistent with a disease with a low prevalence in the population . Although prominent and regularly occurring , MPX does have a relatively low incidence in Tshuapa Province with a cumulative incidence rate of 4 . 8/10 , 000 over a four year period ( data available upon request ) . Several signs/symptoms had a high sensitivity ( lymphadenopathy , fatigue , and the following rash characteristics: same size , deep-seated firm lesions , presence on the arms , legs , palms of the hands , and soles of the feet ) , which indicates that these signs/symptoms may be useful in ruling out other febrile rash illnesses that may be circulating in an area of active MPXV transmission . Signs/symptoms with a high specificity ( nausea , conjunctivitis , bedridden , lesions present on the genitals ) , on the other hand , may be useful in identifying true MPX cases . These four signs/symptoms and , also , sensitivity to light were characteristics that occurred more frequently in laboratory-confirmed MPX cases than VZV cases . However , none of these discriminatory signs/symptoms were found at a high frequency in MPX cases ( <50% ) . Thus , these signs/symptoms may be eligible components of a case definition to identify MPX cases , however , they cannot be a mandatory component . Case definitions for MPX include characteristics of disease presentation specific to the rash itself . This analysis reinforced the inclusion of lesion size and surface presentation ( deep-seated vs superficial ) . Notably , the “characteristic” lesion locations of palms of the hands and soles of the feet , although present in higher frequency in MPX cases than VZV cases , were prevalent in all suspect cases ( 91 . 2% and 81 . 3% , respectively ) and were not helpful in increasing the specificity for MPX cases . Lesions on the palms and soles have been previously noted in VZV cases in central Africa [13] , and the findings here indicate that this presentation may be more common than recognized before . The benefit of an ROC analysis is that one is able to evaluate the effect of increasing number of signs and symptoms on the sensitivity and specificity of case identification . Both laboratory-confirmed MPX and VZV cases presented with many of the same signs and symptoms . Instead of limiting the case definition to an all-or-nothing analysis , we chose to limit the ROC analysis to a subset of highly sensitive and specific signs/symptoms . As such , the analysis indicated that the combination of 7 or 8 signs/symptoms was the most optimal model to accurately predict laboratory-confirmed MPX cases . This analysis emphasized the utility of including the minimum 12 signs/symptoms ( nausea , conjunctivitis , bedridden , lesions on genitals , lymphadenopathy , fatigue , lesions of the same size , deep-seated firm lesions , lesions present on the arms , legs , palms of the hands , or soles of the feet ) on a MPX-specific case investigation , followed by the classification of a patient as a suspect MPX case if they possessed a combination of any 7 or 8 of these specific signs/symptoms plus ‘fever before rash’ . Suspected MPX patients are rarely followed over the course of their infection . Patients are often only seen once , the data on clinical signs/symptoms is assessed and captured at that single time point , and no further follow-up is conducted . A definition such as the one suggested by the ROC analysis may allow for greater flexibility and utility in detecting true MPX cases at any given point during the course of the infection , since it allows for the presence of 7 or 8 signs/symptoms ( versus 12 ) . Additional evaluations to discern a specific suite of signs/symptoms that can be easily identified by healthcare personnel in endemic areas are warranted . This could be followed by modification of the surveillance investigation tool to incorporate the 12 signs/symptoms and evaluation of the utility of a new , modified case definition that accounts for the presence of 7 or 8 signs/symptoms when determining if a patient is a suspected MPX case . The population used in this analysis was a population of suspect MPX cases identified through surveillance for human MPX illness in DRC . A total of 137 cases were excluded from the final analysis due to incomplete data from case report forms ( missing signs/symptoms ) . The majority of excluded cases were laboratory-confirmed VZV cases ( 78 or 72 . 9% of those excluded ) , which may lead to a slight bias in the dataset that contains relatively more laboratory-confirmed MPX cases than were identified in the surveillance dataset . However , a similar proportion of laboratory-confirmed MPX and VZV cases were used for the analyses . Further , data from the suspect MPX cases was collected at one point in time during their illness . The dataset is unique in that it captured many signs/symptoms present for patient; similar datasets with a large number of patients/cases are not available for independent comparison . This data does not represent the spectrum of signs/symptoms that a patient may experience over the course of their illness , and , the result may be a limitation in the frequency of signs/symptoms for suspect cases . Nevertheless , this data reflected the range of presentations of signs/symptoms recognized in suspected MPX patients in an area with endemic disease . In order to document and detect endemic human MPX cases , a surveillance case definition with more specificity in accurate case detection is needed . In rural DRC there are increasingly limited resources , competing health priorities , and a lack of regional testing capacity , which emphasizes the need to easily and efficiently deploy a case definition to accurately identify true MPX patients and limit false positives . A single MPX case or the decision to launch an outbreak response requires considerable resources . According to national guidelines , once a MPX case is identified , the case should be isolated and contacts should be followed for 21 days . Strict recommendations regarding hygiene and infection control are instituted for the entire period of illness , which can last for four weeks . In the absence of a more specific case definition , continued emphasis on laboratory-based diagnostics is warranted . More rapid and efficient methods of diagnosing suspect MPX patients , via a regional surveillance laboratory or a clinical laboratory , are needed to better identify and care for patients followed by appropriate control measures . Disclaimer: The findings and conclusions in this report are those of the author ( s ) and do not necessarily represent the views of the Centers for Disease Control and Prevention . | Human monkeypox is the most significant Orthopoxvirus infection since the eradication of smallpox . The disease is endemic in Africa and the majority of cases occur in the Congo Basin . Correct identification of patients is critical to deployment of efficient control measures to prevent further transmission and appropriate care for the patient . An evaluation of two surveillance case definitions revealed that the definitions had high sensitivities but low specificities for correct case identification , which results in the identification of patients with other rash illnesses . Several signs and symptoms of laboratory-confirmed monkeypox cases were identified that could be used to aid in the development of surveillance case definitions to correctly identify cases . Laboratory diagnostics continue to be an important part of correct patient identification in order to control the disease and provide adequate care . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"dermatology",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"limbs",
"(anatomy)",
"signs",
"and",
"symptoms",
"infectious",
"disease",
"control",
"infectious",
"diseases",
"musculoskeletal",
"system",
"zoonoses",
"hands",
"epide... | 2017 | Enhancing case definitions for surveillance of human monkeypox in the Democratic Republic of Congo |
The genome of Trypanosoma brucei is unusual in being regulated almost entirely at the post-transcriptional level . In terms of regulation , the best-studied genes are procyclins , which encode a family of major surface GPI-anchored glycoproteins ( EP1 , EP2 , EP3 , GPEET ) that show differential expression in the parasite's tsetse-fly vector . Although procyclin mRNA cis-regulatory sequences have provided the paradigm for post-transcriptional control in kinetoplastid parasites , trans-acting regulators of procyclin mRNAs are unidentified , despite intensive effort over 15 years . Here we identify the developmental regulator , TbZFP3 , a CCCH-class predicted RNA binding protein , as an isoform-specific regulator of Procyclin surface coat expression in trypanosomes . We demonstrate ( i ) that endogenous TbZFP3 shows sequence-specific co-precipitation of EP1 and GPEET , but not EP2 and EP3 , procyclin mRNA isoforms , ( ii ) that ectopic overexpression of TbZFP3 does not perturb the mRNA abundance of procyclin transcripts , but rather that ( iii ) their protein expression is regulated in an isoform-specific manner , as evidenced by mass spectrometric analysis of the Procyclin expression signature in the transgenic cell lines . The TbZFP3 mRNA–protein complex ( TbZFP3mRNP ) is identified as a trans-regulator of differential surface protein expression in trypanosomes . Moreover , its sequence-specific interactions with procyclin mRNAs are compatible with long-established predictions for Procyclin regulation . Combined with the known association of TbZFP3 with the translational apparatus , this study provides a long-sought missing link between surface protein cis-regulatory signals and the gene expression machinery in trypanosomes .
The pathway of mRNA control in eukaryotes involves regulatory steps at multiple stages . This is reflected by the large investment of eukaryotic genomes in RNA binding proteins , with hundreds of genes in yeasts and mammals being devoted to functions requiring RNA interaction . As the diverse roles of these proteins , and their interactions with specific subsets of mRNAs , are investigated , it is becoming clear that post-transcriptional control represents a regulatory network of a complexity and importance likely greater than transcriptional control [1] . Perhaps the most extreme example of a group of organisms with emphasis on post-transcriptional control is the kinetoplastid parasites . These organisms are of medical , veterinary and economic importance because they are responsible for an enormous burden of disease within the tropics , including a variety of cutaneous and visceral diseases ( caused by Leishmania spp . ) , Chagas disease ( caused by Trypanosoma cruzi ) and African sleeping sickness ( caused by Trypanosoma brucei ) . Here , an absence of detectable RNA II polymerase promoters for protein coding genes and the general organisation of transcription units into polycistronic arrays necessitates almost complete reliance on post-transcriptional control for regulated gene expression [2] . Supporting this , the genome of these parasites reveals a complexity and composition of encoded RNA binding proteins exceeding , and distinct from , that found in the crown group of eukaryotic organisms [3] , [4] . Gene regulation is particularly important in kinetoplastid parasites because their life cycle is complex , involving passage through a mammalian host and within distinct compartments of an arthropod vector [5] . One of the best-characterised life-cycle differentiation events involves exchange of the major surface antigens as African trypanosomes passage from mammalian blood to the midgut of their haematophagous vector , the tsetse fly [6] . In the bloodstream trypanosomes stay ahead of the immune response by expressing , sequentially and hierarchically , thousands of different antigenic surface coats comprised of variant surface glycoprotein ( VSG ) [7] . However , upon differentiation in the tsetse , the VSG coat is replaced by a family of glycophosphatidyl inositol ( GPI ) -anchored proteins known as Procyclins . There are two types of Procyclin proteins , which mainly differ by the type of amino acid repeats they contain at their C-termini . One set of proteins , the EP isoforms ( encoded by the EP1-1 , EP1-2 , EP2 and EP3 genes ) contain 22–30 [E-P] internal repeat peptides whereas GPEET Procyclins ( encoded by one copy of GPEET ) contain 6 [G-P-E-E-T] repeats , which can be phosphorylated at the Thr residues . In the tsetse , Procyclins follow a programmed expression and their C-terminal repeat peptides , together with their complex GPI anchors , may provide protection for the parasite from the action of tsetse gut hydrolases [8]–[11] . Sequence-dependent signals in the 3′ untranslated region ( 3′ UTR ) of each procyclin mRNA govern their expression and have been the subject of intense investigation , providing the paradigm for gene expression control in kinetoplastid parasites [2] . Although the 3′UTRs of EP1 , 2 and 3 and GPEET procyclin mRNAs are highly similar , the genes are differentially regulated in distinct phases of tsetse infection or in vitro . For example , the GPEET procyclin 3′UTR contains an element , absent in the closely related EP1 3′UTR , that differentially regulates its expression in response to glycerol and the activity of mitochondrial enzyme activities [12] , [13] . However , whilst the cis-acting control sequences for procyclin mRNAs are very well characterised [14]–[17] , protein factors that recognise these regulatory domains have remained unidentified , despite considerable effort . Here we establish the specific association and regulation of procyclin mRNA isoforms by a kinetoplastid-specific protein factor that associates with polyribosomes , providing the first example in these organisms of surface protein regulation by an mRNA-associated regulatory factor .
Previous immunoprecipitation experiments using an antibody specific for a small CCCH-protein implicated in developmental control , TbZFP3 ( Tb927 . 3 . 720 ) , demonstrated co-precipitation of mRNA for the procyclic-form specific surface proteins , Procyclins [18] . Since different Procyclin isoforms exhibit distinct profiles of mRNA and protein expression in the tsetse fly [8] , [11] , [19] we investigated whether each isoform mRNA was co-precipitated with equivalent efficiency by TbZFP3 . Figure 1A shows a typical experiment where immunoprecipitation from cell extracts resulted in a selection for TbZFP3 ( lane 1 ) , this being blocked in the presence of the peptide immunogen used to raise the TbZFP3-specific antibody ( lane 2 ) . The resulting co-selected mRNAs were then reverse transcribed and subjected to quantitative real-time ( qRT ) PCR using primers specific for each procyclin transcript isoform [11] ( Figure 1B ) . In parallel reactions , total RNA from the starting cultures was also analysed with each primer set , allowing us to compare the relative level of each procyclin isoform mRNA in unselected and TbZFP3-immunoprecipitated material . In total mRNA of the cell extracts , both EP2 and EP3 were present at 66% of EP1 levels ( EP1 is normalised to 100% in Figure 1B ) , approximating to their observed relative abundance in culture and in the tsetse midgut [11] . As expected in this parasite strain [12] , [20] , [21] , GPEET mRNA was also abundant ( 191% with respect to EP1 ) in the unselected material . In contrast to unselected cDNA , TbZFP3-immunoprecipitated material showed a strikingly differential abundance of the isoforms , such that EP1 and GPEET were the dominant selected transcripts , with EP2 and EP3 selected at much lower level ( 3 . 3% and 6 . 2% of immunoprecipitated EP1 , respectively ) . Importantly , use of the peptide block prevented the immunoprecipitation of each procyclin mRNA isoform , demonstrating specificity of the selection . Supporting this qRT-PCR data , non-selective amplification and cloning of procyclin cDNAs derived from TbZFP3-immunoprecipitated material isolated 21/27 ( 78% ) EP1 sequences and 6/27 ( 22% ) GPEET sequences , with no clones containing EP2 or EP3 derived sequences . We conclude that although EP1 , EP2 , EP3 and GPEET procyclin mRNAs are each abundant in the unselected mRNA pool , TbZFP3 is preferentially associated with EP1 mRNA , but also GPEET procyclin mRNAs . To determine whether the co-immunoprecipitation of procyclin mRNAs with TbZFP3 was dependent on its predicted RNA-binding domain we examined a cell line expressing a mutant form of TbZFP3 lacking the CCCH zinc finger domain ( TbZFP3 ΔCCCH; [18] ) . This mutant incorporated a C-terminal Ty1-epitope tag to allow it to be specifically immunoprecipitated in the context of endogenous TbZFP3 using the BB2 antibody which detects the Ty1 epitope [22] . As a control , wild type TbZFP3 was also expressed with a C-terminal Ty1 tag with the relative expression of each ectopically expressed protein being examined by Western blotting using either the antibody against TbZFP3 ( this detecting endogenous and ectopically expressed TbZFP3 ) or BB2 ( detecting only the ectopically expressed protein ) . Figure 2A shows the relative expression of each ectopically expressed protein in each cell line , confirming their approximately equivalent abundance . Thereafter , cell extracts from each line were used in immunoprecipitation experiments to select the ectopic TbZFP3 using the BB2 antibody , and the co-selection of procyclin EP1 mRNA assayed . This demonstrated selection of EP1 mRNA with TbZFP3-Ty , as expected , whereas deletion of the CCCH domain prevented co-precipitation of EP1 mRNA ( Figure 2B ) . Thus , the integrity of the predicted RNA binding domain in TbZFP3 is necessary for co-immunoprecipitation of EP1 mRNA . The sequences which regulate procyclin gene expression have been very well characterised in transgenic parasites by use of reporter genes linked to wild type or mutant forms of the EP1 mRNA 3′ UTR . This has identified a number of regulatory regions that act to either positively or negatively control expression [14] , [23]–[26] . Minimally , three domains contribute to EP procyclin regulation: a positive control element in the first 40 nt after the stop codon ( “Loop I” ) , a negative element contained within 101–173 nt ( ‘Loop II’ ) and a further positive element comprising a highly conserved 16 nt stem loop structure ( “Loop III” ) . To determine whether TbZFP3 RNA–immunoprecipitation generated sequence-specific selection of EP1 procyclin mRNA , we generated a series of cell lines transfected with previously characterised reporter constructs ( kindly provided by Professor I . Roditi , University of Bern ) . These comprised a GARP coding region reporter [27] linked to either the wild type EP1 procyclin 3′UTR or mutants lacking each regulatory domain ( Δ40 , ΔLII or Δ16mer ) ( Figure 3A ) . Initially the anticipated effects on reporter gene expression for each construct were confirmed by analysing the GARP mRNA and protein levels in the resulting transfected cell lines ( Figure 3B ) . Matching previous analyses of these deletions [23] , the mRNA abundance of GARP was reduced in the Δ40 ( 62% of wild type levels ) and Δ16mer cell lines ( 26% of wild type levels ) , but significantly elevated in the ΔLII cell line ( 210% of wild type levels ) . Similarly , Western blotting of protein extracts from these cell lines with a GARP antiserum [28] confirmed that the levels of GARP protein translated from the expression constructs matched previous observations , with abundant GARP generated in the ΔLII cell line and little detectable protein when the 16mer element was deleted . Having generated cell lines stably transfected with each reporter construct , extracts from each were subjected to TbZFP3-immunoprecipitation , either in the presence or absence of blocking peptide and analysed for the selection of TbZFP3 , ( Figure 4A , “TbZFP3 IP” ) or of the reporter GARP mRNA ( Figure 4A; “GARP-RT-qPCR” ) . Importantly , in each case the relative selection of GARP mRNA was compared with , and normalised to , the selection of endogenous EP1 procyclin , ensuring the efficiency of immunoprecipitation from each extract was equivalent ( Figure 4A and 4B ) . In the cell line containing GARP linked to wild type EP procyclin 3′UTR , efficient selection of the reporter mRNA was observed with this being abolished in the presence of the blocking peptide ( 80% and 1% respectively , normalised to the relative immunoprecipitation of endogenous EP1 mRNA ) . When the Δ40 cell line was examined efficient selection of GARP transcripts was also observed ( 81% of endogenous EP1 , with 3% of endogenous EP1 in the presence of the peptide block ) . However , when either the negative control element contained in Loop II of the EP1 procyclin 3′UTR , or the 16mer stem-loop structure were deleted , selection with TbZFP3 was reduced to only 1 . 5% or 9% of endogenous EP1 mRNA , respectively . This did not represent inefficient immunoprecipitation since endogenous EP1 procyclin mRNA was selected at an equivalent level in all cell lines ( Figure 4B and data not shown ) . Moreover , it was not simply dependent on target mRNA abundance because the ΔLII–derived GARP mRNA was highly expressed ( Figure 3B ) . This demonstrated that TbZFP3 immunoprecipitation showed sequence-specific selection of the EP procyclin 3′UTR , this being individually dependent upon integrity of the Loop II and the 16mer regulatory domains . Having demonstrated that EP1 procyclin mRNA co-selects with TbZFP3 via known regulatory domains we determined if TbZFP3 could specifically regulate EP procyclin mRNA abundance . Initially , we made use of transgenic procyclic and bloodstream form lines that ectopically over-express TbZFP3 under tetracycline control . Figure 5A ( lanes 1–4 ) shows endogenous and ectopically expressed TbZFP3 mRNA in each cell line , whereas lanes 5–8 shows hybridisation to the same RNAs of a generic EP procyclin riboprobe . This revealed no evidence for a specific enrichment of any EP mRNA in response to TbZFP3 induction in procyclic forms nor appearance of EP mRNA in bloodstream forms ( where Procyclin is not normally expressed ) . Furthermore , quantitative RT-PCR specific for EP1 , EP2 , EP3 procyclin revealed no specific change of EP1 mRNA with respect to EP2 or EP3 mRNAs , although the expression of all mRNAs increased slightly ( ∼20%; Figure 5B ) . Similarly , RNAi directed to TbZFP3 ( resulting in 60% reduction of protein expression; Figure 6A ) resulted in no specific regulation of any procyclin mRNA isoform , although all procyclin mRNAs as well as several housekeeping genes showed an overall reduction of mRNA abundance . This suggests non-specific or indirect effects or , potentially , a more widescale consequence of TbZFP3 knock-down on mRNA abundance ( Figure 6B and data not shown ) . Nonetheless , the analysis demonstrated that there was no differential change in the abundance of procyclin isoform mRNAs caused by enhanced or reduced TbZFP3 expression . To monitor the relative protein expression of individual Procyclin isoforms we made use of an established mass spectrometry approach to detect Procyclins . Thus , cell lines induced to ectopically express TbZFP3 for 48 h , 72 h or 1 week were subject to delipidation and butanol extraction followed by aqueous HF treatment to release the Procyclin proteins from their GPI-anchors . The released full length Procyclins were then further subject to mild acid treatment [29] , a procedure that cleaves the EP isoforms at the Asp-Pro bonds and partially cleaves GPEET between Asp-Gly . The resulting extracts were analysed by negative ion MALDI-TOF-MS to detect the presence and abundance ( based on their comparable ionisations ) of the [M-H]− pseudomolecular ions representing C-terminal fragments of GPEET and EP1-1 , EP1-2 , EP2 and EP3 proteins . Consistent with expectation for this parasite strain , the dominant surface protein was GPEET Procyclin [20] , [21] , with lower expression of the EP Procyclin isoforms ( Figure 7A–7C ) . However , ectopic expression of TbZFP3 ( generating 1 . 5 and 2 . 3 fold overexpression at 48 h and 72 h , respectively ) , progressively elevated EP1-1 and EP1-2 Procyclins ( Figure 7D–7F ) above either uninduced controls ( Figure 7A–7C ) or the parental cell line grown in the presence of tetracycline ( Figure S1 ) . Furthermore , expression of GPEET was progressively reduced ( ∼5 fold ) from being the dominant surface molecule to being a minor component with respect to EP1 after 7 days of induction ( Figure 7F ) . In contrast to these two proteins , both allelic variants of EP3 procyclin ( EP3-1 and EP3-5 ) remained relatively unchanged , whereas EP2 was not detected in any cell population matching previous studies . Consistent with the specific regulation of Procyclin expression by TbZFP3 , examination of the Procyclin protein signature of the cell line expressing the ΔCCCH mutant of TbZFP3 , which does not co-select procyclin mRNAs , did not reveal any change in the profile of expressed proteins regardless of whether the ectopic expression of the mutant protein was induced or not ( Figure S2 ) . To examine the basis of the altered expression of EP1 and GPEET Procyclins after TbZFP3 ectopic expression , we assayed the relative co-immunoprecipitation with TbZFP3 of each procyclin mRNA isoform in the TbZFP3-uninduced population or after 72 h induction . Figure 8 shows a representative semi-quantitative analysis of the relative selection of EP1 and GPEET mRNA in each cell population . This reveals that the relative selection of EP1 mRNA increased upon TbZFP3 ectopic expression , whereas the efficiency of GPEET mRNA selection was diminished . We conclude that moderate elevation of TbZFP3 levels alters the relative association with EP1 and GPEET mRNAs , this resulting in a change of trypanosome surface antigen expression , inducing a change from GPEET to EP1 Procyclin as the dominant surface protein on procyclic forms .
The experiments in this paper identify the developmental regulator , TbZFP3 , as an isoform-specific regulator of Procyclin surface coat expression in trypanosomes . Specifically , we demonstrate ( i ) that endogenous TbZFP3 shows sequence-specific co-association with distinct procyclin mRNA isoforms , ( ii ) that ectopic overexpression of TbZFP3 does not enhance the mRNA abundance of selected transcripts , but rather that ( iii ) their protein expression is regulated in an isoform-specific manner , as evidenced by mass spectrometric analysis of the Procyclin expression signature in transgenic cell lines . Unlike the wild type TbZFP3 protein , a mutant form of TbZFP3 lacking its C×8C×5C×3H predicted RNA-interaction motif and which cannot co-associate with procyclin mRNAs does not alter Procyclin expression . We have already demonstrated that TbZFP3 promotes differentiation when associated with the translational machinery ( this being dependent upon its predicted RNA and protein interaction motifs ) and that this occurs only in the parasite life cycle stage at which Procyclin proteins are expressed [18] . Hence , our work provides a long-sought ‘missing link’ between the intensely studied cis-regulatory signals for the procyclin gene family and the general gene expression machinery , this being the translational apparatus . TbZFP3 shows specific co-association in vivo with EP1 and GPEET procyclin mRNA , whereas the distinctly regulated transcripts EP2 and EP3 are not co-immunoprecipitated . Although deletion of the predicted RNA-binding domain in TbZFP3 prevents the co-association with EP1 mRNA , our studies do not formally distinguish between direct intermolecular contact between TbZFP3 and target mRNAs and indirect contact dependent on other protein factors . Hence , we use the term TbZFP3mRNP to define the composition of the immunoprecipitated material comprising TbZFP3 , procyclin mRNAs and , possibly , other identified ( e . g . [18]; see below ) and unidentified co-operating factors . Nonetheless , by using an immunoprecipitation approach employing an antibody directed to the endogenous TbZFP3 protein in wild type parasites we demonstrate that the observed co-association with procyclin mRNAs is physiological , and directed by TbZFP3 in its normal cellular context . Interestingly , the differential selection of different procyclin mRNA isoforms by the TbZFP3mRNP matches their overall sequence similarity , with EP1 and GPEET 3′UTR sequences being significantly more closely related than EP2 and EP3 ( Figure S3 ) . Nonetheless , EP1 and GPEET are differentially regulated in vivo , with GPEET expression being repressed as EP1 is upregulated during differentiation to late procyclic forms in vitro and in the tsetse fly [8] , [12] . Significantly , this matches the observed effects of TbZFP3 ectopic overexpression , whereby EP1 expression is elevated to become the dominant surface protein and GPEET expression is repressed , this correlating with enhanced association of the TbZFP3mRNP with EP1 mRNA and diminished association with GPEET mRNA . Although copy number control of Procyclins on the parasite surface could accentuate this switch , it is significant that the inverse control of these surface molecules is regulated by only subtle changes in the abundance of TbZFP3 ( ∼1 . 5–2 . 5 fold ) . This suggests exquisitely regulated control of Procyclin isoform expression in response to TbZFP3 levels . In addition to isoform-specific selection , the TbZFP3mRNP exhibits sequence-specific association with the EP1 mRNA 3′UTR , this being dependent upon the integrity of two well-characterised regulatory regions - the ‘Loop II’ and the 16mer stem loop region . Previous analyses have demonstrated that these sequences provide negative and positive control elements for EP1 procyclin expression , respectively . The Loop II region acts as a translational repressor and mRNA destabilisation element in procyclic forms , whereas the 16mer is a translational enhancer , which suppresses the action of the Loop II region [23] , [24] . In insect stages , it was predicted that a macromolecular complex would associate with both elements and so shield the ‘Loop II’ element from recognition by a negative regulator , thereby promoting gene expression [23] . Our findings are compatible with this , invoking a model ( Figure 9 ) in which TbZFP3 competes with a negative regulator binding ‘Loop II’ , such that TbZFP3 over-expression promotes EP1 Procyclin expression ( at the expense of GPEET ) , whereas RNAi mediated removal of TbZFP3 results in reduced procyclin mRNA abundance . Interestingly , expression of the Loop II deletion reporter construct revealed that TbZFP3mRNP-binding is not necessary for efficient mRNA or protein expression ( Figure 3B ) , suggesting that the TbZFP3mRNP acts primarily as an anti-repressor ( Figure 9 ) , matching earlier predictions for the procyclin regulatory machinery [2] , [17] , [23] , [30] . This is analogous to the regulation of nanos RNA during Drosophila embryogenesis , whereby overexpression of Oskar displaces the translational repressor Smaug bound to the nanos 3′UTR [31] . CCCH proteins in eukaryotes are involved in all levels of gene expression through RNA recognition , usually this being dependent upon the integrity of at least two juxtaposed CCCH fingers [32] . TbZFP3 , however , has only a single CCCH motif and is a member of a family of small CCCH zinc finger proteins ( TbZFP1 , TbZFP2 and TbZFP3 ) , unique to kinetoplastids , and each <140 amino acids long . These proteins interact in yeast 2-hybrid assays and co-immunoprecipitate in vivo , suggesting that they provide a modular function in order to confer specificity of binding [33] . Whether this is part of a single mRNP complex , or distinct complexes with differential specificities for different genes or in different life cycle stages , remains to be determined . This complexity of interactions may define the specificity of different mRNA classes selected by the TbZFP3mRNP , or moderate their differential efficiency of selection and hence regulation , as observed with EP1 and GPEET mRNA regulation under conditions of TbZFP3 ectopic expression . To be understood in depth , such interactions will need to be analysed on a case-by-case basis for individual RNAs as has been done here for procyclin mRNAs . Nonetheless , analogous combinatorial interaction between RNA binding proteins in kinetoplastid parasites to confer target specificity and regulation has previously been proposed for the small RNA binding proteins TbUBP1 and TbUBP2 [34] , homologues of the regulators of mucin gene expression in T . cruzi [35] . TbZFP2 and TbZFP3 are constitutively expressed and associate with the translation apparatus in procyclic forms but not in bloodstream forms [18] . Our demonstration here that the TbZFP3mRNP co-associates with procyclin mRNA regulatory elements that control translation , thus promoting EP1 surface protein expression without enhancing EP1 mRNA abundance , suggests a role for TbZFP3 in translational control . Supporting this , a mutant TbZFP3 lacking the predicted RNA-interaction domain neither co-associates with procyclin mRNAs ( Figure 2 ) nor the translational apparatus [18] and induces no consistent change in Procyclin expression ( Figure S2 ) . Temporally-regulated translational control is a key aspect of cell-type development in the Plasmodium parasite , whereby translational repression via the DDX6 RNA helicase family member DOZI regulates gametocyte mRNA expression and life-cycle differentiation [36] . Interestingly , these transcripts share a 47 nt U-rich control element [37] , similar to the regulatory U-rich 26mer elements enriched in procyclic form-specific transcripts [38] and comprising part of the Loop II region of EP1 procyclin recognised by TbZFP3 . This points to common mechanisms of developmental control among widely divergent eukaryotic protozoan pathogens . Translational control is believed to be a major mechanism of gene regulation in trypanososmatid parasites [39] , [40] . Although the general mRNA degradation and translational machineries are broadly conserved in these evolutionarily ancient eukaryotic organisms [3] , it is the kinetoplastid-specific trans-acting regulators that provide the key to understanding their extreme emphasis on post-transcriptional control . Moreover , targeting unique components of the translational machinery in pathogens is a major strategy in antimicrobial therapies . Thus , discovering novel regulators interacting with this apparatus provides both new understanding of gene expression and new possibilities to intervene in the virulence and spread of these devastating parasites .
Procyclic form or bloodstream form T . brucei Lister 427 trypanosomes were used throughout . Cell lines engineered for TbZFP3 ectopic expression in procyclic or bloodstream forms have been described previously and were cultured in SDM-79 or HMI-9 , respectively [18] . Immunoprecipitation using TbZFP3-specific antisera , RNA extraction and reverse transcription have been described previously [18] . Blocking peptides used were N-DSSQMQQVGHDVPPMA-C for TbZFP3 and N-EVHTNQDPLD-C for Ty1 , each being titrated prior to use . SYBR green qRTPCR reactions were performed using Roche reagents as per specifications for the LightCycler system . The 5′ primers for actin , ep and gpeet and 3′ primers specific to EP1 , EP2 , EP3 , GPEET , or the Anchor sequence were described previously [11] , [18] . cDNA was amplified as follows: 10 min , 95°C; 30×[8 s , 95°C; 9 s , 55°C; 12 s , 72°C] with fluorescence acquired at 82°C . The amplification was followed by a melting temperature analysis that measured PCR product fluorescence during a temperature increase from 65°C to 95°C at 0 . 1°C/s to determine product melting temperature and confirm specificity . Product identities were further verified by gel electrophoresis and DNA sequencing . In all cases , serial dilutions of input cDNAs confirmed the quantitative efficiency of the reactions and “no reverse transcriptase” controls confirmed the absence of contaminating genomic DNA in the RNA preparations . Northern blotting involved resolution of 3–5 µg of total trypanosome mRNA on formaldehyde agarose gels resolved in MOPS buffer . Hybridization of blots used digoxigenin labelled riboprobes , detected using anti-DIG alkaline phosphatase-conjugated antibody and visualised using CDP-star as a reaction substrate ( Roche ) . Western blots were detected and quantitated using a Li-COR Odyssey system , using alpha-tubulin as an internal standard . Mass spectrometry was carried out according to the methodology described in [8] , [19] , [29] . Briefly , parasite pellets were freeze-dried and then extracted twice with 200 µl of Chloroform/Methanol/Water , 10∶10∶3 ( V/V/V ) , under sonication ( 10 min ) . After centrifugation , the delipidated pellets were then extracted 3 times with 150 µl of 9% butanol ( ButOH ) , also under sonication . The ButOH fractions contain the Procyclins . All ButOH fractions were then freeze-dried and submitted to dephosphorylation using 50 µl of 48% aqueous hydrofluoric acid ( aq . HF ) , at 0°C for 24 h . After aq . HF incubation the samples were freeze-dried again and washed twice with water . The samples were then dried and further incubated with 200 µl of 40 mM TFA , 20 min at 100°C ( mild acid conditions ) , in order to assist visualization of the Procyclin C-termini and the identification of each isoform . Under this condition , the Asp-Pro bonds of most of the EP isoforms are cleaved whereas GPEET partially releases 13 amino acids at its N-terminus . Equivalent amounts of each sample were mixed with α-cyano ( matrix ) and analysed by negative-ion MALDI-TOF-MS using a Voyager-DE STR instrument . | Trypanosomes , the tropical parasites that cause African sleeping sickness , show a number of biological peculiarities that distinguish them from other eukaryotes . One is the unusual way in which they regulate gene expression . Unlike most eukaryotes , trypanosomes do not regulate gene expression by controlling the rate of messenger RNA synthesis , but , instead , control the stability of messenger mRNAs ( and , hence , their abundance ) and also their rate of translation into protein . The best-studied model for this “post-transcriptional” gene expression control in trypanosomes is the procyclin mRNAs , which encode the major surface proteins of the parasite in the tsetse fly . In this study we demonstrate that a small kinetoplastid-specific protein ( TbZFP3 ) co-associates with the mRNAs for some procyclin isoforms ( EP1 , GPEET procyclin ) but not others ( EP2 , EP3 procyclin ) . Furthermore , we show that this is dependent upon sequences in the EP1 procyclin 3′untranslated region known to govern its mRNA turnover and protein synthesis . Finally , we demonstrate that limited over-expression of TbZFP3 causes a change in the surface protein expression profile on cultured parasites from GPEET to EP1 Procyclin . Our data identify TbZFP3 as an important post-transcriptional regulator of Procyclin expression , the first such protein factor identified . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"biology/rna-protein",
"interactions",
"molecular",
"biology/translational",
"regulation",
"microbiology/parasitology",
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/tropical",
"and",
"travel-associated",
"diseases",
"cell",
"biology/gene",
... | 2009 | Differential Trypanosome Surface Coat Regulation by a CCCH Protein That Co-Associates with procyclin mRNA cis-Elements |
Critical to human innate immunity against African trypanosomes is a minor subclass of human high-density lipoproteins , termed Trypanosome Lytic Factor-1 ( TLF-1 ) . This primate-specific molecule binds to a haptoglobin-hemoglobin receptor ( HpHbR ) on the surface of susceptible trypanosomes , initiating a lytic pathway . Group 1 Trypanosoma brucei gambiense causes human African Trypanosomiasis ( HAT ) , escaping TLF-1 killing due to reduced uptake . Previously , we found that group 1 T . b . gambiense HpHbR ( TbgHpHbR ) mRNA levels were greatly reduced and the gene contained substitutions within the open reading frame . Here we show that a single , highly conserved amino acid in the TbgHpHbR ablates high affinity TLF-1 binding and subsequent endocytosis , thus evading TLF-1 killing . In addition , we show that over-expression of TbgHpHbR failed to rescue TLF-1 susceptibility . These findings suggest that the single substitution present in the TbgHpHbR directly contributes to the reduced uptake and resistance to TLF-1 seen in these important human pathogens .
Primate specific innate immunity plays a decisive role in defining the host range of African trypanosomes . Trypanosoma brucei brucei , Trypanosoma congolense and Trypanosoma vivax infect both domesticated and wild mammals but are unable to infect most primates , including humans , because of their susceptibility to two primate specific innate immune complexes , Trypanosome Lytic Factor-1 ( TLF-1 ) and TLF-2 [1]–[3] . TLF-1 and TLF-2 , isolated from humans , have similar protein compositions . Both complexes contain apolipoprotein A-1 ( apoA-1 ) , a characteristic protein of high-density lipoproteins ( HDLs ) , and two primate-specific proteins , apolipoprotein L-1 ( apoL-1 ) and haptoglobin related protein ( Hpr ) [3]–[8] . Despite similarities in protein composition , the two complexes differ significantly , TLF-2 containing IgM and having little associated lipid , while TLF-1 is a minor sub-class of HDL ( ρ = 1 . 21–1 . 26 g/ml ) which is ∼40% lipid by mass [9] . TLF-1 killing of T . b . brucei requires high affinity binding within the flagellar pocket , a specialized region of the trypanosome cell surface , followed by endocytosis and lysosomal localization [10] . Within the acidic lysosome , TLF-1 is activated leading to disruption of the lysosome and cell lysis [5] , [10]–[13] . Critical to initiating the lytic pathway is the binding of TLF-1 to the T . b . brucei haptoglobin-hemoglobin receptor ( TbbHpHbR ) [11] , [14] , [15] . Haptoglobin ( Hp ) is an acute phase protein produced at high levels in all mammals , which binds and detoxifies free hemoglobin ( Hb ) by facilitating its clearance from the circulation [15] . Since African trypanosomes are heme auxotrophs , TbbHpHbR has been proposed to function as a nutrient receptor providing heme to these parasites [14] . Unlike the mammalian HpHb scavenger receptor ( CD163 ) the TbbHpHbR also binds Hpr present in TLF-1 when complexed with Hb [16] . Two mechanisms of trypanosome resistance to TLF-1 , and therefore human infectivity , have been described [17] . Trypanosoma brucei rhodesiense , the cause of acute human African trypanosomiasis ( HAT ) , has evolved the human serum resistance associated protein ( SRA ) , which binds and neutralizes TLF-1 killing [16]–[19] . A member of the variant specific glycoprotein ( VSG ) family , SRA , is a glycophosphatidylinositol-anchored protein that is synthesized in the endoplasmic reticulum and transiently presented on the surface of the trypanosome within the flagellar pocket . However , its steady state distribution suggests it is rapidly endocytosed and localizes predominately to endosomes in T . b . rhodesiense [20]–[22] . SRA tightly binds the apoL-1 component of TLF-1 , providing complete protection against TLF-1 killing [5] . It is assumed that SRA also binds apoL-1 in TLF-2 and inhibits its activity . Trypanosoma brucei gambiense , the causative agent of chronic HAT , lacks SRA . We recently reported that expression of the T . b . gambiense HpHbR ( TbgHpHbR ) was reduced in the group 1 subtype of T . b . gambiense , suggesting that decreased expression of the receptor contributed to TLF-1 resistance and human infectivity [23] . We also observed that the TbgHpHbR gene , from four distinct geographic isolates of group 1 T . b . gambiense , contained four non-synonymous amino acid substitutions within the coding sequence for the mature protein [23] . A more extensive analysis of a large number of isolates further revealed a single leucine ( L ) to serine ( S ) substitution , at amino acid 210 of TbbHpHbR which was conserved in all group 1 T . b . gambiense isolates examined [24] . This led to the suggestion that this substitution might reduce the affinity of TbgHpHbR for TLF-1 [23]–[25] . Recently , the crystal structure of the HpHbR was deduced allowing new structure-function analysis . A domain of HpHbR , protruding beyond the VSG coat , possesses a hydrophobic core in which ligand binding is predicted to occur . The L210S substitution present in T . b . gambiense was predicted to disrupt the core of the head structure of HpHbR thus eliminating ligand binding [26] . Other mechanisms of resistance to TLF-1 and human serum must exist since group 2 T . b . gambiense lacks SRA yet expresses the functional HpHbR and takes up TLF-1 [24] . The mechanism of group 2 T . b . gambiense resistance to TLF-1 remains unknown . In the studies reported here , we developed a T . b . brucei double knockout line ( TbbHpHbR−/− ) to directly test the function of each of the four non-synonymous amino acid substitutions in the TbgHpHbR . Expression of TbgHpHbR in the TbbHpHbR−/− knockouts did not restore TLF-1 binding or killing . However , the substitution of serine for leucine , at position 210 , restored high affinity TLF-1 binding and susceptibility . Changes to the other three substituted amino acids in the TbgHpHbR had no effect on TLF-1 binding , uptake or killing . These results , together with our previous finding , indicate that TLF-1 resistance has exerted strong selective pressure on group 1 T . b . gambiense , resulting both in decreased expression levels and loss of function substitutions in the TbgHpHbR .
Bloodstream form T . b . brucei Lister 427 ( MiTat 1 . 2 ) were grown at 37°C under 5% CO2 in HMI-9 medium supplemented with 10% fetal bovine serum ( Sigma-Aldrich ) and 10% Serum-Plus ( Sigma-Aldrich ) . HpHbR KO constructs were generated after cloning HpHbR flanking sequences onto blasticidin and hygromycin resistance genes [23] . All primers used in the studies reported here are listed in Table S1 . 3×107/ml trypanosomes were transfected with 5 µg of NotI digested DNA using the Amaxa nucleofection system according to the manufacturer's instructions . ( Human T Cell Nucleofactor Kit , program X-001 ) . Transfected cells were then allowed to recover for 24 hours before addition of blasticidin ( 2 . 5 µg/ml ) or hygromycin ( 2 . 5 µg/ml ) . Cell lines were clonally selected prior to a second round of transfection . To obtain HpHbR double knockout cell line , we transfected the second HpHbR drug resistance construct into the single allele TbbHpHbR+/− lines . To examine the effects of amino acid substitution on HpHbR function , stable cell lines expressing ectopic copies of the TbbHpHbR , TbgHpHbR or the individual TbgHpHbR substitutions were prepared by targeting to the tubulin locus and selection with phleomycin ( 2 . 5 µg/ml ) [23] . To determine growth rates , cells were grown to mid-log phase and diluted to 1×104/ml . Cell counts , determined by hemocytometer , were carried out until stationary phase . Growth curve data is in triplicate . An HA-epitope tag was cloned into the TbgHpHbR construct via a three-step PCR method . The HA-tag was added downstream of the signal peptide . Once completed , the construct was sequenced and digested with NotI and ApaI ( 5 µg total ) prior to transfection . Transfections and cloning were carried out as described above . The construct used to generate the TbgHpHbR cell line [23] was subjected to site-directed mutagenesis to generate the four TbgHpHbR substitutions of S210L , V293A and GA369-370EG . Mutagenesis was carried out according to manufactures instructions ( Agilent Technologies ) . TbbHpHbR−/− cells were transfected independently with the mutagenized constructs . Transfections and cloning were carried out as described above . Mutant TbgHpHbR constructs were sequenced with HpHbR sequence primers ( sense and antisense ) . To prepare an HpHbR over-expressing cell line , PCR products were generated with Platinum High Fidelity Taq Polymerase ( Invitrogen ) , gel purified , digested with EcoRI and cloned into the pURAN over-expression constructs [27] . Prior to transfection into TbbHpHbR−/− cells , pURAN HpHbR constructs were linearized with BstXI . Both strands were sequenced with HpHbR sequence primers ( sense and antisense ) . TLF-1 purification , labeling and survival assays were performed as previously described [28] . Briefly , for the survival assays , trypanosomes were harvested from mid-log phase cultures , washed and re-suspended at a final concentration of 1×106/ml in complete HMI-9 media . Susceptibility to hemoglobin ( Hb ) bound TLF-1 was determined over a range of TLF-1 concentrations following incubation at 37°C for 16 hours . The number of surviving cells was determined by hemocytometer count with phase contrast microscopy . All survival assays were done in triplicate . For Southern analysis , 5 µg genomic DNA was digested with EcoRI . DNA was fractionated on a 0 . 6% agarose gel and transferred to a nitrocellulose membrane ( Amersham Hybond-N+ ) . Pre-hybridization was with a solution containing 40% formamide ( Sigma ) , 3× SSC , 10× Denhardt's , 20 µg/ml salmon sperm DNA , 5% dextran sulfate and 2% SDS at 42°C for three hours . The P32 labeled probe , specific for the region upstream of the 5′-flanking regions was added to the hybridization mix and incubated at 42°C overnight . The probe sequence is listed in Table S1 . Blots were then washed in a solution containing 3× SSC/0 . 1% SDS at 55°C for 30 minutes then a final stringency of 0 . 3× SSC/0 . 1% SDS at 65°C for 20 minutes . Blots were exposed to a storage phosphor screen ( Molecular Dynamics ) and analyzed on a STORM-860 PhosphorImager ( GE Healthcare ) . All TLF-1 binding and uptake studies were carried-out with Alexa-Fluor 488 TLF-1 that was labeled according to manufacture instructions ( Invitrogen ) . Alexa-488 TLF-1 was incubated with an excess of Hb for 10 minutes on ice prior to analysis of binding . The binding and uptake of Alexa-488 TLF-1 was examined by either fluorescence microscopy or FAC analysis . To measure the amount of binding by fluorescence microscopy , the fluorescence intensity values from AxioVision v4 . 6 software ( www . zeiss . com ) was plotted versus TLF-1 concentrations . Imaging was carried out using a Zeiss Axio Observer inverted microscope . Quantification of Alexa-488 TLF-1 was done on compressed images . To measure TLF-1 uptake by FAC analysis , cells were grown to mid-log phase , collected , washed and resuspended ( 1×107/ml ) in HMI-9 supplemented with 1% bovine serum albumin ( BSA ) , 1% glucose . Alexa-488 TLF-1 , and excess Hb , were added to the cells followed by incubation at 37°C for 30 minutes . Uptake was stopped by placing the tubes on ice followed by two washes with ice-cold phosphate buffered saline buffer ( PBS ) ( 10 mM NaPi , 137 mM NaCl , pH 7 . 4 ) . The amount of TLF-1 uptake was determined using Cyan cytometer ( DAKO ) and FlowJo software . Uptake was also measured by fluorescence microscopy . Following incubation , cells were washed two times with ice cold PBS . Following the washes , cells were spread onto glass slides , methanol-fixed for 5 min , at −20°C , and analyzed by fluorescence microscopy . Images were captured with the same exposure and were contrasted to the same extent . To analyze only binding in the flagellar pocket , pre-chilled Alexa-488 TLF-1/Hb complexes were added to cells in ice-cold HMI-9 supplemented with 1% BSA , 1% glucose and further incubated at 3°C for two hours . Cells were washed two times with ice-cold PBS , kept on ice and subjected to FAC analysis . All binding and uptake experiments analyzed by FAC analysis were done in triplicate . For fluorescence microscopy , approximately 100 cells per data point were analyzed in triplicate . All binding data collected were analyzed using Graphpad Prism software , version 4 . 0a . To better localize the distribution of TLF-1 binding to the flagellar pocket , PFA fixed and methanol treated trypanosomes were incubated with a mouse anti-paraflagellar rod ( PFR ) antibody ( generously provided by Dr . Diane McMahon-Pratt , New Haven ) at a dilution of 1∶1 , 000 followed by a secondary antibody staining with Alexa Fluor 594 mouse IgG ( Invitrogen ) . Serial image z-stacks were acquired through oil immersion optics with exposure times kept constant for each experiment . Imaging was carried out using a Zeiss Axio Observer inverted microscope equipped with an AxioCam HSm camera and analyzed with the AxioVision v4 . 6 software ( Zeiss ) . A single stack is shown for each experiment , with individual channels contrasted to the same extent for each image set . Specificity of TLF-1 binding to trypanosomes was analyzed using competition-binding studies with the unlabelled TLF-1 and Hp 1-1 . Cells were collected , washed and resuspended ( 1×107/ml ) in ice-cold HMI-9 supplemented with 1% BSA , 1% glucose then transferred to 3°C for at least 10 minutes . Alexa-488 conjugated TLF-1 ( 3 nM constant ) was complexed with hemoglobin ( 50 nM ) at 4°C for 10 minutes . Increasing concentrations of unlabeled competitor were incubated with Hb ( 50 nM ) for 10 minutes at 4°C . Competing ligands were then mixed with the Alexa-488 conjugated TLF-1/Hb , added to cells at 3°C and allowed to incubate for two hours . Cells were then transferred to ice , washed with ice-cold 1× PBS and analyzed by Cyan cytometer and FlowJo software . For studies without Hb , competitors were added to Alexa-488 TLF-1/Hb ( 6 nM ) in the same increasing molar concentrations and taken through the same protocol as previously described . All competition studies were done in triplicate . Total RNA was isolated with Tripure Isolation Reagent ( Roche ) . cDNA was generated in a Reverse Transcription ( RT ) reaction ( Promega ) . Control reactions were performed with enolase , as well as reactions without added RT . Real time PCR was performed with and iCycler ( iQ5 multicolor real-time PCR detection system; Bio-Rad ) using cDNA from an equivalent of 20 ng of total RNA , 6 pmol sense primer , 6 pmol antisense primer , 10 µl SYBR green PCR master mix ( Fermentas ) in a final volume of 20 µl . Real time PCR conditions were: one cycle of 95°C for 3 min , followed by 40 cycles of 95°C for 15 s , 60°C for 30 s and 72°C for 30 s . The relative amounts of specific cDNA between samples were calculated using CT values calculated with the iQ5 optical detection system software . All qRT-PCR reactions were carried out with a splice leader RNA sense primer and gene specific anti-sense primers . Triplicate analyses were performed for each parasite line . All primers were designed using Integrated DNA Technologies software . For western analysis of HA-tagged HpHbR , total cellular protein was prepared from TbbHpHbR−/− , Rab5aHA and TbgHpHbRHA cell lines and and analyzed , based on cell equivalents , as previously described [22] . Rat monoclonal anti-HA–biotin ( Roche Diagnostics , Indianapolis , IN ) was used at a dilution of 1∶1000 , with streptavidin-HRP conjugate ( Invitrogen , Camarillo , CA ) used for secondary detection at 1∶5000 .
Previously , we described the isolation of a TLF-1 resistant line of T . b . brucei following in vitro selection for growth in the presence of human HDLs [23] , [29] . The resistance phenotype correlated with reduced expression of the TbbHpHbR , susceptibility being restored by ectopic expression of the TbbHpHbR from a different chromosomal locus . We also observed that ectopic expression of the TbgHpHbR failed to restore TLF-1 uptake or susceptibility , suggesting that substitutions to the TbgHpHbR might contribute to TLF-1 resistance in this important human pathogen [23] . Initial sequence analysis of four T . b . gambiense isolates revealed five non-synonymous amino acid substitutions , four in the coding sequences of the mature protein , when compared to TbbHpHbR [23] . To test whether these substitutions lead to loss of TLF-1 binding , we generated a T . b . brucei HpHbR−/− knockout cell line and then systematically tested the ability of each of the four substitutions to restore TLF-1 binding to the TbgHpHbR . The HpHbR knockout cell lines were prepared in T . b . brucei 427-221 ( Lister 427 ) cells by sequentially replacing the complete coding sequence for each TbbHpHbR allele with the coding sequences for hygromycin and/or blasticidin ( Figure 1A ) . Replacement of the coding sequence for TbbHpHbR , in both single TbbHpHbR+/− ( sKO ) and double TbbHpHbR−/− ( KO ) knockouts , was verified by Southern blot hybridization of genomic DNA digested with EcoRI . The expected size restriction fragments were detected when blots were hybridized with a probe complementary to the 5′ flanking sequence of TbbHpHbR . A 7 . 0 kb fragment was detected in untransfected TbbHpHbR while EcoRI sites in both the blasticidin and hygromycin gene constructs gave rise to smaller fragments ( 3 . 9 kb and 4 . 2 kb respectively ) ( Figure 1A ) . PCR analysis of genomic DNA from WT and TbbHpHbR−/− cells , with oligonucleotide probes complementary to coding sequences in the TbbHpHbR , showed that the TbbHpHbR gene had been deleted in the TbbHpHbR−/− cells ( Figure S1 ) . Furthermore , ( q ) RT-PCR with total RNA from WT T . b . brucei and TbbHpHbR−/− showed that double knockout cells do not express TbbHpHbR ( Figure 1B , Table S2 ) . The generation of a stable TbbHpHbR−/− cell line showed that this gene was non-essential in the bloodstream stage of T . b . brucei used in these studies . In addition , only a very slight reduction in growth rate was observed ( Figure 1C ) . We examined whether the TbbHpHbR−/− cells were deficient in TLF-1 uptake . When WT T . b . brucei was incubated at 37°C with Alexa-488 conjugated TLF-1 ( 10 nM ) , cells rapidly accumulate TLF-1 in endosomes and lysosomes , whereas no detectable Alexa-488 TLF-1 internalization was observed in TbbHpHbR−/− cells ( KO ) ( Figure 1D ) . As previously reported , uptake of TLF-1 was dependent on the addition of Hb prior to incubation with trypanosomes [28] . Additionally , it was shown that at 3°C , Alexa-488 TLF-1 localized specifically to the flagellar pocket [22] . To determine whether TLF-1 uptake at this low concentration was dependent on flagellar pocket binding , WT and TbbHpHbR−/− cells were incubated at 3°C with Alexa-488 TLF-1 ( Figure 1E ) . At concentrations as low as 0 . 6 nM , TLF-1 binding to the flagellar pocket was detectable and was concentration dependent up to 66 nM . No TLF-1 binding to TbbHpHbR−/− cells was observed at concentrations up to 66 nM ( Figure 1E ) . In addition , TbbHpHbR−/− cells were highly refractory to TLF-1 killing at concentrations of 0 . 1 nM ( Figure 1F ) . These studies showed that the TbbHpHbR was required for high affinity TLF-1 binding and further supports the role of this receptor in trypanosome killing . In order to determine whether the TbgHpHbR was functional in TLF-1 binding and subsequent killing , stable cell lines , ectopically expressing the TbbHpHbR and TbgHpHbR genes , were prepared in the TbbHpHbR−/− background . In addition , to verify HpHbR expression , an HA-epitope tagged variant of the TbgHpHbR ( TbgHpHbRHA ) , in the TbbHpHbR−/− background , was also prepared ( Figure 2A ) . Expression of TbbHpHbR , TbgHpHbR and TbgHpHbRHA was determined by nested RT-PCR allowing detection of both endogenous and HA-tagged HpHbR mRNAs ( Figure 2A ) . The level of HpHbR mRNA was comparable in all cell lines ( Table S2 ) . The expression of TbgHpHbRHA was also evaluated by western blot with antibodies specific to the HA-tagged HpHbR ( Figure 2B ) . A single band , migrating around 80 kDa , was visible in TbgHpHbRHA cells , but not in the TbbHpHbR−/− cell line ( Figure 2B ) . Specificity of the anti-HA antibody was verified with a cell line expressing a HA-tagged Rab5A ( Figure 2B ) [22] . To determine whether the TbgHpHbRHA was functional in TLF-1 binding and uptake , cells were examined for TLF-1 binding and subsequent killing ( Figures 2C and 2D ) . TbbHpHbR , but not TbgHpHbR or TbgHpHbRHA , restored both TLF-1 binding and killing in the TbbHpHbR−/− background ( Figures 2C and 2D ) . We also prepared a HA-epitope tagged variant of the TbbHpHbR ( TbbHpHbRHA ) , in the TbbHpHbR−/− background . However , perhaps due to the positioning of the HA-epitope within the highly structured cytosolic domain of the receptor , no detectable signal was seen on western blots ( unpublished data ) . It is unlikely that the TbbHpHbRHA was not expressed since mRNA levels were comparable to the other receptor knock-in lines and TLF-1 binding and susceptibility were restored to wild type levels in these cells . Together these results indicate that TbbHpHbR retains the ability to bind TLF-1 while TbgHpHbR and TbgHpHbRHA , while expressed at comparable levels , do not function in TLF-1 binding , uptake or trypanosome killing . Sequence analysis of a small number of isolates of group 1 and 2 T . b . gambiense , T . b . rhodesiense and T . b . brucei led to the initial hypothesis that a limited number of amino acid substitutions may contribute to reduced uptake of TLF-1 by cells expressing the TbgHpHbR [23] ( Figure 3A ) . In a more comprehensive geographic and taxonomic analysis of HpHbR sequences , a single substitution replacing a leucine with a serine at position 210 was observed in all group 1 T . b . gambiense and was not observed in TLF-1 susceptible T . b . brucei or in T . b . rhodesiense [24] . To test whether any of the substitutions in the TbgHpHbR could individually restore TLF-1 binding and killing , each of the T . b . gambiense specific amino acid substitutions were changed back to the amino acid found in the TbbHpHbR ( Figure 3A ) . The steady state levels of TbbHpHbR , TbgHpHbR , TbgHpHbRS210L , TbgHpHbRV293A and TbgHpHbRGA369-370EG mRNA s were evaluated by ( q ) RT-PCR ( Figure 3B , Table S2 ) . The levels of expression of these ectopically expressed genes were comparable in all five analyzed cell lines . In order to determine whether amino acid changes in the TbgHpHbR affected TLF-1 uptake , each cell line was examined by fluorescence microscopy and flow cytometry following incubation with Alexa-488 TLF-1 for 30 minutes at 37°C ( Figure 3C and 3D ) . Fluorescence microscopy showed that TbbHpHbR expressing cells endocytosed TLF-1 and that most was localized to the posterior region of the cells between the kinetoplast and nucleus consistent with lysosomal trafficking ( Figure 3C ) . Flow cytometry indicated that the amount of TLF-1 taken up by the TbbHpHbR cells was similar to that seen in WT T . b . brucei ( Figure 3D ) . In contrast , TbgHpHbR cells showed no detectable uptake of TLF-1 either by fluorescence microscopy or flow cytometry analysis ( Figure 3C and 3D ) . Similarly , TbgHpHbRV293A and TbgHpHbRGA369-370EG did not take-up TLF-1 and appeared identical to TbgHpHbR cells . However , the single amino acid change at position 210 of the TbgHpHbR , from serine to leucine , restored TLF-1 uptake and localization to levels seen in the TbbHpHbR cells ( Figure 3C and 3D ) . The specificity of TLF-1 binding in TbbHpHbR and TbgHpHbRS210L was examined by competition binding studies with unlabeled TLF-1 or Hp1-1 in the presence or absence of Hb ( Figure 3E and 3F , respectively ) . When complexed with Hb both unlabeled Hp 1-1 and TLF-1 effectively competed for TLF-1 binding . These results indicated that the HpHbR mediated all TLF-1 binding in these cells . To determine whether susceptibility to TLF-1 killing was also influenced by the changes to the HpHbR , cell lines expressing TbbHpHbR , TbgHpHbR , TbgHpHbRS210L , TbgHpHbRV293A and TbgHpHbRGA369-370EG were incubated with increasing concentrations of TLF-1 and the percentage of cells surviving after 16 hours was determined ( Figure 3G and 3H ) . As expected , based on uptake studies , cells expressing TbbHpHbR and TbgHpHbRS210L were fully susceptible to TLF-1 . TbgHpHbR , TbgHpHbRV293A and TbgHpHbR GA369-370EG were resistant to TLF-1 killing . Together these studies show that the single amino acid change of serine to leucine at position 210 of TbgHpHbR is sufficient to restore both TLF-1 uptake and killing to levels seen in cells expressing the TbbHpHbR . This finding is consistent with the substitution to the HpHbR in T . b . gambiense playing a critical role in human infectivity . To evaluate the effect of the amino acid substitutions in the TbgHpHbR on the binding affinity for TLF-1 , a live cell-binding assay was developed with Alexa-488 TLF-1 . Cells were incubated at 3°C for 2 hours with varying concentrations of Alexa-488 TLF-1 . Unbound TLF-1 was removed by washing in ice-cold 1× PBS and the amount and location of TLF-1 binding evaluated by FAC analysis and fluorescence microscopy ( Figure 4 , Figure S3 ) . Alexa-488 TLF-1 localized exclusively to the flagellar pocket ( Figure 4D ) and cell associated fluorescence was concentration dependent in WT T . b . brucei , TbbHpHbR and TbgHpHbRS210L cell lines ( Figure 4A–C ) . No detectable TLF-1 binding was seen in TbbHpHbR−/− , TbgHpHbR , TbgHpHbRV293A , and TbgHpHbRGA369-370EG cell lines ( Figure 4C and D ) . To determine whether the affinity for TLF-1 differed in the TbbHpHbR−/− cell lines expressing TbbHpHbR , TbgHpHbR , TbgHpHbRS210L , TbgHpHbR V293A and TbgHpHbR GA369-370EG we performed saturation binding studies with Alexa-488 TLF-1 ( Figure 4A–C ) . The binding affinity was estimated based on half-maximal binding . Both WT T . b . brucei and TbbHpHbR had high affinity for TLF-1 ( 3 . 96±0 . 31 nM and 4 . 12±0 . 25 nM , respectively ) . The TbgHpHbRS210L cells also bound TLF-1 with similar affinity ( 3 . 96±0 . 35 nM ) . Consistent with the results obtained by microscopic analysis , TLF-1 binding was not observed in the TbgHpHbRV293A and TbgHpHbR GA369-370EG cell lines . Based on these results , the highly conserved amino acid substitution in the TbgHpHbR at position 210 is responsible for decreased binding affinity for TLF-1 . In flow cytometry studies , a small amount of TLF-1 binding to the TbgHpHbR was detected ( Figure 4B ) . To determine whether this represented binding to the TbgHpHbR or a low level of background binding we over-expressed the TbgHpHbR and the TbbHpHbR through ectopically expressing the HpHbR , driven by a ribosomal promoter , in the TbbHpHbR−/− cells . This resulted in a 15-fold increase in expression of TbgHpHbR and the TbbHpHbR mRNAs as measured by qRT-PCR ( Table 1 ) . Overexpression of the TbbHpHbR also resulted in a large increase in TLF-1 binding ( 16-fold ) and sensitivity to TLF-1 killing ( 18-fold ) . However , over-expression of TbgHpHbR , to similar levels , had no effect on TLF-1 binding or susceptibility ( Table 1 ) . Together these results indicate that the TbgHpHbR is unable to bind TLF-1 and that a single amino acid change in the TbgHpHbR is sufficient to spare T . b . gambiense from TLF-1 killing .
Previous studies have shown that the level of HpHbR expression can influence the susceptibility of African trypanosomes to TLF-1 and human serum [4] , [23] . Analysis of mRNA levels in five field isolates of group 1 T . b . gambiense showed that TbgHpHbR expression was reduced 20-fold relative to T . b . brucei [23] . In addition to reduced mRNA levels , four non-synonymous substitutions present in the TbgHpHbR and not in TbbHpHbR were identified [23] . A more extensive analysis of HpHbR gene sequences from 67 isolates of T . b . brucei , T . b . gambiense group 1 and group 2 and T . b . rhodesiense supported these findings and further narrowed conserved substitutions in TbgHpHbR . This led to the suggestion that substitution of leucine with serine at position 210 might abolish TLF-1 binding [23]–[25] . To directly test the consequence of amino acid substitutions within the TbgHpHbR , on TLF-1 binding , uptake and trypanolytic activity we established a TbbHpHpR−/− cell line by replacement of both alleles with drug resistance markers ( Figure 1A ) . Using this stable cell line , we tested each amino acid substitution in the TbgHpHbR individually by ectopic expression ( Figures 3B ) . By systematically changing each of the amino acid substitutions in the TbgHpHbR to the most common sequence in TbbHpHbR , we showed that the S210L change restores high affinity TLF-1 binding , uptake and trypanosome killing ( Figure 3 , 4 ) . Based on these new findings and our previous results we propose that group 1 T . b . gambiense has evolved two mechanisms to avoid uptake of TLF-1 . First , the abundance of HpHbR mRNA was reduced 20-fold in all group 1 T . b . gambiense isolates tested [23] . Secondly , as shown in the studies presented here , the TbgHpHbR had reduced affinity for TLF-1 due to an amino acid substitution that was highly conserved in all members of this subgroup . It is likely that both reduced HpHbR expression and TLF-1 affinity contribute to the overall resistance of group 1 T . b . gambiense to TLF-1 . Recent crystallographic studies of the T . congolense HpHbR have allowed a detailed structural analysis of the trypanosome HpHbR [26] . These studies revealed a hydrophobic core head domain predicted to be important in receptor-ligand interaction and further predicted that the hydrophobic core of the ligand-binding domain would be disrupted by the S210L substitution described in Figure 3 . We found that addition of a HA-epitope within the disrupted head domain of the TbgHpHbR was accessible for antibody detection ( Figure 2B ) . In contrast , the HA-epitope , in the stabilized head domain of the TbbHpHbR , was inaccessible to antibody binding yet retained TLF-1 binding and facilitated killing ( unpublished data ) . The in vivo binding results presented in Figures 3 and 4 , were also consistent with SPR binding assays with recombinant HpHbR , which showed that the leucine to serine substitution significantly reduced TLF-1 and HpHb binding to the HpHbR [26] . A potentially important difference in TLF-1 binding was revealed in the in vitro binding studies with recombinant HpHbR [26] and the in vivo studies reported here ( Figure 3 ) . The SPR binding results showed a striking difference in the affinity for TLF-1 and HpHb for the TbbHpHbR ( 5–10 µM and 4 . 5 nM respectively ) [26] . This is inconsistent with our findings showing that TLF-1 , when saturated with bound Hb , binds with a similar affinity as HpHb to the TbbHpHbR ( Figure 3 ) . The relatively low affinity binding of TLF-1 may result from sub-saturating levels of Hb in the TLF-1 samples used in their studies . Alternatively , the higher affinity measured in vivo may reflect a role for secondary binding proteins on the trypanosome surface that increase the binding affinity of the heterogeneous TLF-1 particles [30] . Other mechanisms of resistance contribute to human infectivity by group 1 T . b . gambiense since TLF-2 kills HpHbR deficient T . b . brucei lines , albeit at concentrations approximately 200-fold higher than needed to kill WT T . b . brucei [4] . It is possible that group 1 and group 2 T . b . gambiense share common , HpHbR independent , mechanisms of resistance . Unlike group 1 T . b . gambiense , the sequence of the HpHbR gene in subgroup 2 more closely resembles that of T . b . brucei . Critically , the group 2 T . b . gambiense HpHbR has a leucine at position 210 [24] , [25] . TLF-1 binding , uptake and trafficking to the lysosome of group 2 T . b . gambiense also more closely resemble the TbbHpHbR [24] . The HpHbR has been proposed to be an essential nutrient receptor in African trypanosomes functioning in hemoglobin scavenging in these heme auxotrophs [14] . The near wild type growth rate of T . b . brucei HpHbR−/− cell line showed that the receptor was not essential for survival in vitro ( Figure 1 ) . Furthermore , this suggests that heme scavenging , by the HpHbR , may not be necessary in bloodstream African trypanosomes . An attractive alternative is that the T . b . brucei HpHbR−/− cell lines have other mechanisms for heme uptake that can compensate for the loss of the HpHbR under in vitro growth conditions . Recently , a heme transporter has been described in Leishmania that is partially localized to the plasma membrane suggesting that heme may be transported into kinetoplastids in the absence of the HpHbR [31] . It is not surprising that group 1 T . b . gambiense has evolved diverse mechanisms for protection against TLF-1 and 2 . These parasites have a long and intimate involvement with the human host . Largely lacking wild game or domesticated animal reservoirs , these parasites have had ample opportunities to develop both redundant and augmenting mechanisms of resistance . It is likely that the observed reduced expression and loss of function substitution to the HpHbR gene in group 1 T . b . gambiense , though seemly redundant processes , heightens the collective resistance of these cells to the more complex assault by the human innate immune systems . Group 2 T . b . gambiense is genetically more diverse than group 1 and has evolved a novel HpHbR independent mechanism for inhibition of TLF-1 killing [24] . Since group 2 T . b . gambiense express a functional HpHbR , resistance requires inhibition of TLF-1 killing . It is appealing to speculate that group 1 T . b . gambiense may share this mechanism but its effect on TLF-1 killing is largely masked by reduced TLF-1 uptake . | African trypanosomes are parasites that are able to infect a wide variety of mammals; however , only two sub-species , Trypanosoma brucei rhodesiense and Trypanosoma brucei gambiense , are able to infect humans . A human innate immune molecule , trypanosome lytic factor-1 ( TLF-1 ) , is responsible for this selective protection . TLF-1 killing requires high affinity binding to the trypanosome haptoglobin-hemoglobin receptor ( HpHbR ) , which initiates endocytosis and lysosomal localization of the toxin . T . b . gambiense infects humans because it does not bind TLF-1 , and several amino acid changes in the HpHbR are conserved in group 1 T . b . gambiense . To better understand the mechanism of resistance in these parasites , we analyzed TLF-1 binding to trypanosomes expressing the T . b . gambiense HpHbR ( TbgHpHbR ) and variants in which single amino acids were changed . Our studies showed that a single , highly conserved , amino acid substitution in the TbgHpHbR was sufficient to ablate high affinity TLF-1 binding and contributed to TLF-1 resistance . This likely plays a key role in human infectivity by group 1 T . b . gambiense . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biomacromolecule-ligand",
"interactions",
"mutagenesis",
"biochemistry",
"flow",
"cytometry",
"genetic",
"mutation",
"genetics",
"biology",
"cytometry",
"molecular",
"cell",
"biology"
] | 2013 | A Single Amino Acid Substitution in the Group 1 Trypanosoma brucei gambiense Haptoglobin-Hemoglobin Receptor Abolishes TLF-1 Binding |
Malignant melanoma is a cancer of the skin arising in the melanocytes . We present a mathematical model of melanoma invasion into healthy tissue with an immune response . We use this model as a framework with which to investigate primary tumor invasion and treatment by surgical excision . We observe that the presence of immune cells can destroy tumors , hold them to minimal expansion , or , through the production of angiogenic factors , induce tumorigenic expansion . We also find that the tumor–immune system dynamic is critically important in determining the likelihood and extent of tumor regrowth following resection . We find that small metastatic lesions distal to the primary tumor mass can be held to a minimal size via the immune interaction with the larger primary tumor . Numerical experiments further suggest that metastatic disease is optimally suppressed by immune activation when the primary tumor is moderately , rather than minimally , metastatic . Furthermore , satellite lesions can become aggressively tumorigenic upon removal of the primary tumor and its associated immune tissue . This can lead to recurrence where total cancer mass increases more quickly than in primary tumor invasion , representing a clinically more dangerous disease state . These results are in line with clinical case studies involving resection of a primary melanoma followed by recurrence in local metastases .
Melanoma , the most dangerous form of skin cancer , arises in the melanocytes and progresses through two well-defined clinical stages . Following a period of radial growth in the epidermis , melanomas may switch to malignant , vertical growth melanoma ( VGM ) [1] . This switch generally occurs following the onset of angiogenesis and penetration of the basement membrane ( BM ) separating the dermis and epidermis [2] . These processes are tightly coupled in melanomas [2] , [3] . Angiogenesis is induced primarily by the release of angiogenic factors by melanoma cells and associated stromal cells and through the restructuring of the extracellular matrix ( ECM ) that occurs in concert with invasion [4] . Angiogenesis of lymphatic vessels ( lymphangiogenesis ) also occurs in melanomas and plays a role in lymphatic metastasis [5] . Melanomas are known to produce a number of angiogenic cytokines , the most prominent being vascular endothelial growth factor ( VEGF ) and basic fibroblast growth factor ( bFGF ) [2] . VEGF is overexpressed both constitutively and in response to hypoxia [6] . Although some angiogenesis may occur before penetration of the basement membrane , intense angiogenesis requires ECM remodeling , which in turn requires the cooperation of stromal cells . Even if the tumor is releasing large amounts of VEGF , most of it is sequestered in the ECM [7] . Fibroblast recruitment is essential for the angiogenic switch . Once recruited by platelet derived growth factor ( PDGF ) and other cytokines , fibroblasts begin producing ECM degrading matrix metalloproteases ( MMPs ) [8] . Matrix degradation releases large amounts of sequestered angiogenic growth factors , including VEGF , and eases the penetration of new capillaries [7] , [8] . MMPs also aid in recruiting stromal cells involved in angiogenesis . MMP-9 induces proliferation and motility of endothelial precursor cells ( EPCs ) in the bone marrow [8] , and MMP-2 recruits VEGF expressing macrophages and leukocytes [9] . Endothelial cells are the predominant cell type in the formation of new vasculature . Endothelial cell migration into the tumoral region is essential for angiogenesis and is facilitated by MMP mediated matrix remodeling [9] and migration up chemotactic and haptotactic gradients [10] . The tumoral vasculature incorporates epithelial cells from the existing vasculature as well as circulating EPCs . The latter are likely important in sustained , intense angiogenesis as they provide a much larger pool of endothelial cells than the native tissue alone can provide [11] , [12] . Chemotaxis occurs in response to a number of factors , the most important being VEGF . Growing epithelial sprouts extend filopodia , indicating that they , like epithelial cells , can respond to a VEGF gradient [10] . The immune system is able to effectively mobilize against tumor invasion . This occurs mainly through direct tumoricidal action by natural killer ( NK ) cells and phagocytes , such as macrophages , and through the T cell response [7] . Unfortunately , the tumor microenvironment is strongly immunosuppressive [13] , and while the immune response can hinder invasion , tumor-associated lymphocytes and macrophages have been observed secreting growth and angiogenic factors that may aid in tumor invasion [7] , [13] , [14] . Melanoma has a strong tendency to metastasize , with most metastases occurring in the skin , lymph nodes , and lungs [5] . Following resection of the primary lesion , cancer can recur locally . This is often due to growth in local satellite metastases and is not caused by incomplete resection [15] . Removal of the primary tumor can stimulate growth in previously dormant metastases [7] , [16] , [17] . Metastatic melanoma has a very poor clinical prognosis and is largely unresponsive to existing therapies [1] . In this paper we develop a spatially explicit model using partial differential equations ( PDEs ) to capture the dynamics of melanoma invasion in the skin . We first present a basic model that does not consider an immune response and examine tumor invasion in a cylindrical section of skin . Then we extend this model to include a cellular immune response and carry out a number of numerical experiments using this extended model . In these , we examine the possible dynamics of tumor invasion under different levels of immune response . We develop a method to realistically model the stochastic process of local metastatic spread , and surgical treatment is simulated in both locally metastasizing and non-metastasizing melanomas . These simulations are motivated by a clinical case where , following resection of a primary melanoma , widespread recurrence was seen in local satellite metastases [17] . Based on our simulations we make several observations and predictions . First we observe that angiogenesis is strongly tumorigenic , which is very well known from previous experimental and theoretical work . In accordance with the biological observations in [6] , we find that the production of angiogenic factors by melanoma cells in response to hypoxia is insufficient to induce angiogenesis and vertical growth; constitutive production of such factors is needed . These results validate the basic construction of the complex model . We show that the immune response can either aid or inhibit tumor progression; the outcome depends on the balance between angiogenic factors released by immune cells and the growth-inhibitory and cytotoxic effects of the same immune cells . We also make several predictions concerning treatment and metastasis . We observe that a relatively small safety margin is necessary to remove all primary tumor material and prevent local recurrence due to “local persistence” ( as defined in [15] ) . In the case of a tumor spawning satellite metastases , we predict that the immune response directed against a primary tumor can suppress these metastases “in passing . ” The removal of the primary tumor also removes most of this immune response , allowing the cancer to recur through the growth of these previously suppressed metastases . Thus , local recurrence in previously dormant satellite metastases seen in clinical cases [17] , [15] can be explained as a consequence of immune disruption .
We propose a spatially explicit model of melanoma invasion in the skin , formulated in terms of cell densities . We first present the full system with a detailed derivation of the governing equation for each variable . We then describe the model geometry and boundary conditions . The basic model considers seven variables: The basic assumptions we use to derive the model follow The mathematical model is formulated as follows: ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) ( 7 ) Where H ( x ) is the Heaviside step function: As our model is intended to capture the process of primary melanoma invasion in the skin we must build a reasonable approximation to this geometry . Therefore , we generally consider a three-dimensional domain using cylindrical coordinates – . We furthermore assume radial symmetry , eliminating dependence on and reducing our consideration to only two independent spatial variables - . Considering a cylindrical section of skin , let and , where the z-axis is directed downward from the surface of the skin . We assume that at ( i . e . the base of the skin segment ) there exists a vascular bed . Oxygen is supplied by this vasculature and at the skin surface . Dirichlet boundary conditions are used to hold oxygen concentration constant at the skin surface ( ) and at the vasculature ( ) . Also , the vascular bed is the source for all endothelial cells entering the skin section . In response to TAF , existing vasculature is destabilized and endothelial cells begin migrating up the TAF gradient . Circulating endothelial precursor cells can also be recruited by TAF and enter the skin from the vascular bed . Therefore , we use a Neumann boundary condition to represent an influx of endothelial cells in response to TAF from both these sources . Finally , the basement membrane separates the epidermis and dermis , serving as a barrier to migration for most cell types . The basement membrane is modeled using initial conditions in simulations - a thin strip of membrane is thought to exist at a depth of approximately 0 . 15 mm . A simple schematic of the modeled geometry is shown in Figure 1 . All boundary conditions are no-flux with several exceptions for oxygen and endothelial cells . The Dirichlet boundary conditions for oxygen are 150 mmHG for the skin surface [18] and 90 mmHg for the vasculature [30]: We also provide for an influx of ECs from the vascular bed and circulation through a Neumann boundary condition at as follows: The maximum EC influx is , and is the TAF concentration at which half-maximal EC stimulation occurs . While this model contains a large number of parameters , we can make at least a reasonable order of magnitude estimate for all of them from empirical biological data . A wide range of population doubling times has been observed for different melanoma lines , with more advanced tumors typically having higher growth rates . Doubling times range between approximately 1 and 4 days [31] , [32] . We always assume that the maximum death rate is roughly the same as the maximum growth rate . The maximum density of melanoma lesions also varies widely . A range from 1 . 0×105–5 . 0×105 TC mm−3 for several tumor strains is given in [33] . The linear diffusion coefficient for human keratinocytes was measured between . 002 and . 07 mm2 day−1 in [34] . Assuming that melanoma cells achieve a similar rate of diffusion at maximum density yields in the range 7 . 0×10−7–4 . 0×10−9 , using the range of densities reported in [33] . In [35] , a threshold for hypoxia of 10 mmHg is given , with 5–7 . 5 mmHg considered moderate hypoxia and considered severe . We estimate our parameters governing the sensitivity of growth and death to using these values as guidelines . Moreover , we generally assume that healthy cells are more sensitive to hypoxia than cancerous cells . We also assume that production of TAF is proportional to growth inhibition due to hypoxia , i . e . the same parameter , , is used in the respective Hill functions . In a tissue scaffold model the diffusion coefficient for VEGF has been measured as and the degradation rate as [36] . However , using the relation given in [37] with a molecular weight of 45 kDa for VEGF gives , about an order of magnitude lower . Maximum rates of constitutive and hypoxic VEGF expression for several melanoma lines are given in [20] , [6] . Using data from [38] , the growth rate for bovine endothelial cells is calculated as . Each endothelial cell corresponds to approximately 5 µm of vessel [22] . Mean vessel diameter for melanoma xenografts varies between 9 . 5 and 14 . 6 µm in [39] . This gives with a mean value of 1 . 893×10−4 . Measured values for the capillary permeability coefficient for oxygen are reported to range between 8 . 64 mm day−1 and 6 . 74×104 mm day−1 , depending on the tissue type [40] . The baseline oxygen consumption rate , , is given in [18] . A number of values for for different cells are given in [18]; we determine a maximum . All parameters and values with references ( if applicable ) are given in Table 1 . We extend the model to include a class of immune cells that are directly cytotoxic to cancer cells . The most important of these cells are macrophages , dendritic cells ( DCs ) , natural killer cells ( NKs ) , and cytotoxic T cells ( ) . We do not consider the humoral ( i . e . antibody generating ) immune response . Melanomas secrete inflammatory cytokines that attract circulating monocytes to the site of invasion . These monocytes can differentiate into macrophages or dendritic cells [13] . Macrophages and DCs are both important phagocytes , capable of engulfing tumor cells as well as fragments of necrotic debris . Macrophages are frequently observed both infiltrating the tumor mass and in the peritumoral region of melanomas . Interestingly , these macrophages can be either harmful or helpful to tumor development . Although macrophages have a cytotoxic effect on tumor cells , tumor-associated macrophages release a number of angiogenic factors , including VEGF , bFGF , IL-8 , and that can aid in tumor progression [41] . In a mouse model by Nesbit et al . [14] , modest expression of monocyte chemoattractant protein-1 ( MCP-1 ) in melanomas attracted macrophages that proceeded to aid the tumor through release of angiogenic factors . However , high expression led to massive macrophage infiltration that destroyed the tumor within days . Natural killers do not need any education to recognize and destroy neoplastic cells . On encountering aberrant cells NKs can initiate a large scale immune response through the release of cytokines that recruit other effector immune cells , the most important being tumor necrosis factor ( ) [7] . A strong response by effector immune cells is probably more harmful to the tumor than helpful . The presence of tumor-infiltrating lymphocytes ( TILs ) has been associated with a good clinical prognosis in a number of cancers . Patient survival is 1 . 5 to 3 times longer in melanoma patients with high numbers of TILs compared to patients with few TILs [7] . Here we present an extension to the basic model that considers effector cell-tumor interactions . Two new variables are introduced: Similar to the TAF formulation , is a generic aggregation of all chemokines that attract macrophages and other immune cells which we refer to as immune attracting factor ( IAF ) . These factors are primarily MCP-1 and . is an aggregation of all cytotoxic effector cells . This aggregate type most closely resembles macrophages and natural killers , as it is cytotoxic to tumor cells and clears necrotic debris . Like natural killer cells , it also has the ability to initiate a larger immune response through secretion of IAF . We assume that ICs are attracted by IAF and are activated by contact with either tumor cells or necrotic debris . Tumor cells express IAF at a constant constitutive level . ICs express TAF and IAF and have a cytotoxic effect on tumor cells . Contact with tumor and dead cells is assumed to activate ICs . This activation causes immune cells to express both TAF and IAF and increases anti-tumor and debris cytotoxicity . Given that the tumor microenvironment is often immunosuppressive and even directly toxic to immune cells [7] , [13] , we also assume that tumor cells have a cytotoxic effect on ICs . Additionally , we assume that ICs can devote energy either to debris cleanup or tumor cell destruction . A cartoon of the modeled immune response is shown in Figure 2 . The equations governing ICs and IAF , as well as the modifications to the equations for tumor cells , necrotic debris , and TAF , follow ( “…” represents the original terms of the equation ) : ( 8 ) ( 9 ) is the level of activation:IC mobility is achieved through simple diffusion and a chemotactic response to IAF . represents the level of activation in response to contact with cancer cells and necrotic debris . and weigh the relative ability of cancer cells and dead cells to activate ICs , and is the ( weighted ) cell density at which half-maximal activation occurs . is normal IC turnover . Each cancer cell kills ICs at maximum rate . Cancer cells generally do not actually kill immune cells , but if we assume the hostility of the microenvironment is directly proportional to cancer cell density then this formulation is a reasonable approximation . The term causes toxicity to saturate with sufficient IC density . Generally , is be taken to be quite small , so that toxicity is at nearly the maximum except at very low IC densities . ICs are imagined to expend energy at some per capita rate which can be devoted either to debris cleanup or killing cancer cells . The amount of energy devoted to either task is proportional to the presence of debris and tumor cells as well as some inherent bias toward one task or another . Energy devotion to debris cleanup is measured as: is a parameter measuring the bias towards debris cleanup . We suppose that can be scaled in such a way as to measure maximum rate of tumor cell destruction - , or dead cell destruction - . Modifying these further , and are the respective rates of destruction for a non-activated immune cell . and are the the rates of destruction beyond and for a fully activated immune cell . Thus , we arrive at the additional death terms for cancer cells: . The terms cause toxicity to be proportional to cell density , and is taken to be very small . The additional death term for necrotic debris is similar . We assume activated ICs produce TAF at a rate proportional to the level activation , with a maximum of . This assumption is justified because , as already noted , tumor-associated macrophages have been observed producing angiogenic factors , and it may be that macrophages that have ingested tumor-derived antigens produce very high levels of VEGF [7] . IAF diffuses by simple diffusion , is produced by tumor cells at the constant rate , and degrades at rate . ICs also produce IAF at a rate proportional to the level of activation . Boundary conditions are no-flux except for IAF . When IAF concentration is high enough we expect an influx of ICs into the domain , representing macrophage , NKs , and migrating from the circulation . The Neumann boundary condition used follows: The IAF density at which half-maximal IC migration occurs is given by , and is the maximum influx from the circulation . We have derived a baseline set of values for most parameters from empirical data . However , a wide range of values is allowed in simulations for those parameters representing cytotoxicity , as these are assumed to be quite variable among strains of melanoma . We have been unable to find any estimates for the diffusion coefficient for macrophages . However , given that measurements for keratinocytes and endothelial cells fall within the same rough range , . 002– . 07 [34] and . 000864– . 070848 , respectively [42] , [43] , we can reasonably assume a similar range for immune cells . The molecular weight of MCP-1 was measured to be about 4 kDa in [44] , although other authors have reported values of 8–10 kDa [45] . Using the relation given in [37] yields . We have been unable to find any estimates on the half-life of MCP-1 either in tissue or in plasma , but we assume it is on the order of several hours . This gives for a half-life of 1–6 hours . Data on MCP-1 production by melanoma cells is given in [14] , from which we calculate . We assume maximum expression by immune cells is similar . We also assume that maximum expression of TAF by immune cells is similar to the constitutive expression by melanoma cells ( see [20] , [6] ) . The turnover rate for human natural killer cells has been measured to be about 2 weeks [46] , giving a maximum . However , since the overall natural killer population remains static , the “effective” turnover rate is probably closer to 0 . To determine the rate at which immune cells kill tumor cells we use data on phagocytosis in macrophages given in [47] . We employ a simple ODE system considering macrophages ( ) , macrophages that have ingested an antigen ( ) , and a tracking variable counting the total number of antigens ingested and processed ( ) . Assuming the antigen level remains constant , this yields the ODE system ( 10 ) ( 11 ) ( 12 ) Solving numerically allows the slope of the kill tracking line ( ) to be calculated , yielding the rate at which macrophages destroy antigen without the need to explicitly model antigen complex populations in the full model . From [47] , we estimate and , yielding a maximum kill rate of 2 . 027 day−1 , presumably for fully activated macrophages . Using and , we estimate the maximum rate of destruction for inactivated macrophages to be about an order of magnitude lower , at 0 . 218 day−1 . The parameters , , and are scaling parameters , included in the formulation for model flexibility . However , for a first approximation we may simply set them all equal to 1 . In the absence of other data , and given that macrophage production of MCP-1 is similar to melanoma production of VEGF , we assume that . A tumor cell strain is considered immunoevasive when parameter values are used that give low anti-tumor cytotoxicity and/or a high cell density for activation . A cell strain is considered immunosuppressive if it is highly cytotoxic to immune cells . All immune extension parameters with values and references are given in Table 2 .
All simulations have been run using a finite difference method on the symmetric cylindrical domain described previously . We run several simulations using the base model without the immune response to characterize the basic dynamics . For biologically realistic parameter values , the model produces realistic patterns of invasion . Before the onset of angiogenesis , growth is restricted to the epidermis . TAF expression by melanoma cells causes an influx of endothelial cells into the domain , which leads to penetration of the basement membrane and vascularization of the tumor within several months . Following this angiogenic switch , density increases and invasion spreads throughout the domain . Live cancer cell density is highest at the skin surface and the vascular bed . Between these boundaries the effects of oxygen consumption by proliferating cancer cells can be seen . Hypoxia is most pronounced near the invasive edge , where oxygen demand is greatest . Necrotic debris is initially concentrated in a roughly spherical core . As the tumor continues to invade , this core expands as an annulus following the invasive edge . Thus , the model predicts that in the absence of an immune response a solid invasive tumor with a necrotic core will form . Angiogenesis is predicted to be strongly tumorigenic . To thoroughly demonstrate the model results , a 3-D isosurface of the evolution of three key variables , cancer cells , basement membrane , and endothelial cells , over several months of invasion is shown in Figure 3 . A 2-D projection of the same simulation is shown in Figure 4 . Finally , a 1-D projection of variable densities at the inner radius of the domain is shown in Figure 5 . A second set of simulations is performed in which the cancer cell line does not produce TAF constitutively . In these simulations , TAF production in response to hypoxia is not sufficient to induce angiogenesis , and tumor growth is restricted to the epidermis . Even when TAF production in response to hypoxia is set to the highest value reported in [6] and maximum tumor cell density is used , angiogenesis does not occur and growth remains radial . If the same simulations are run with no hypoxic TAF production but with the lowest rate of constitutive TAF production reported in [6] then angiogenesis occurs , leading to an aggressive , vertically invasive tumor . Therefore , we predict that the constitutive production of TAF is much more important in melanoma of the skin . This is in line with the conclusions made by Danielsen and Rofstad in [6] . However , we note that constitutive production of TAF by melanoma cells themselves may still be be unnecessary , as stromal cells play a very important role in producing the angiogenic factors and MMPs necessary for invasion , as discussed previously . We can therefore conclude that TAF production in response to hypoxia is insufficient to induce angiogenesis - TAF must be produced by melanoma cells constitutively or supplied by cooperating stromal cells . In simulations of primary tumor invasion with an immune response , tumor growth is generally slowed significantly . There appear to be three possible eventual outcomes: All three outcomes occur in biologically reasonable parameter space . The pattern of immune cell infiltration differs between tumors that can be characterized as immunoevasive versus immunosuppressive . In immunoevasive tumors , immune cell levels are high in the core , while suppressive tumors result in a high peritumoral concentration with little core presence . We have also examined the effect of TAF in expression by activated immune cells . This can cause a two-phase pattern of growth , where the immune response initially holds the tumor to a steady state within the epidermis . After a period of apparent quiescence , immune-induced angiogenesis leads to a second phase of more aggressive vertical growth . Surprisingly , the tumor can completely invade the domain in this second phase even if the immune response was sufficient to hold it to a steady state in the first phase . The results of such a simulation are shown in Figure 6 . To simulate treatment , the tumor is allowed to grow for some specified amount of time , after which surgical excision is performed . To simulate excision the value of all variables is set to zero within a prescribed region . Then the simulation is allowed to continue for several years . The results of a surgical excision using the basic model without an immune response are simply characterized . Immediately following excision the wound is quickly filled by healthy cells , and in the following absence of TAF what remains of the tumoral vasculature quickly dies off . However , within a year any surviving tumor cells begin invading again , and soon the tumor recovers to pre-surgery mass . In simulations without an immune response the tumor border is typically sharply defined with little spread beyond the visible border . A margin of several millimeters beyond the visible edge of the tumor is generally sufficient to ensure no tumor cells survive . To simulate a primary tumor seeding metastases in the nearby tissue a normal simulation is first run for some initial amount of time ( typically 6 months ) . Then a very small metastasis is introduced some distance from the primary tumor at the level of the vasculature . Note that this metastasis is necessarily a “metastasis ring , ” due to the symmetry of the domain . This is not an unreasonable approximation , as metastasis spread is be expected to be roughly symmetric , and a complete ring can be thought to represent a worst-case scenario . We assume that metastasis seeding can be modeled as a Poisson process , i . e . the probability of a metastasis being created within a given time period is an exponential random variable . We furthermore assume that the distance from the primary tumor at which the metastasis is seeded is exponentially distributed , with the greatest probability next to the tumor edge . Thus , a biologically reasonable approximation of metastatic spread into the surrounding skin tissue can be described by two exponential rate parameters , and . To explicitly differentiate between primary tumor and metastasis populations an additional cancer cell variable is introduced . A number of minor modifications must be made to the model; they are straightforward and we do not present the details . The immune response is included in these simulations . If no treatment is performed the primary tumor invades normally while the metastases , if sufficiently close to the primary tumor , are generally destroyed or held to an extremely low density . We have found that if , then all metastases are typically suppressed . If the primary tumor is removed , any surviving metastases are often able to begin invading . The immune response eventually reacts to the metastases , and the same asymptotic behavior occurs ( a steady-state is reached or the domain is invaded ) . However , the metastases often increase in total mass much more quickly than did the primary tumor . Such a course of treatment followed by metastatic recurrence is shown in Figures 7 and 8 . Resection can lead to aggressive metastatic recurrence , but in some cases where the immune response is very strong , resection can lead to a state where metastases persist but remain held to a small size . This state of persistence can last indefinitely , and an example is shown in Figure 9 . Numerical investigation has yielded a somewhat nonintuitive result concerning the rate at which metastases are seeded . In a sensitivity analysis of , we have found that for low values of a single secondary metastasis usually forms . Upon increasing to fairly large values , significant metastatic spread occurs , but it is eliminated by the immune response . Very large values of cause metastatic disease to overwhelm the immune response and invade widely . We note that the initial secondary metastasis generally occurs at a predictable distance from the primary tumor , just beyond the range in which the immune response directed against the primary tumor can incidentally suppress metastases . Figure 10 shows the results under several values of . These results suggest that moderately aggressive primary tumors that seed many metastases can induce a widespread local immune response that is sufficient to keep these metastases in check . Furthermore , those metastases that do manage to grow to significant sizes are quickly eliminated by the immune response , even though this response cannot eliminate the primary tumor .
We have examined the macroscopic dynamics of melanoma tumor growth using a reaction-diffusion framework . This diffusion framework models cell populations as a continuous density field , and has the effect of washing out local inhomogeneities . However , tumors have an irregular and heterogeneous architecture , and angiogenesis in particular is a spatially complex phenomenon , with the tumor vasculature marked by irregular construction and heterogeneity in blood flow [19] . Local perturbations and substructures in the vasculature likely have a minimal effect on the tumor's overall growth and architecture . Of more concern is the variability in nutrient supply , principally oxygen , caused by a regionally heterogeneous vasculature , and its potential effect on overall tumor growth . However , our model suggests that melanoma is rather unique among solid tumors , with significant hypoxia only occurring at the leading edge of invasion . This is due in large part to the significant amounts of oxygen diffusing into the tumor core from the skin surface . Thus , the unique geometry in which melanoma invades likely dampens the importance of vascular irregularity . Therefore , we argue that the diffusion approximation can reasonably be employed in examining melanoma tumor invasion at the macroscopic level . Using this model as a framework for early investigation , we have observed a wide range of interesting and biologically reasonable patterns of tumor invasion . Angiogenesis is strongly tumorigenic in this model . In simulations using the basic model , following the onset of angiogenesis the tumor spreads throughout the dermis and a significant necrotic core forms . Hypoxia is always most severe at the leading edge of radial invasion , and the necrotic core expands as an annulus in sustained invasion . The constitutive production of TAF ( particularly VEGF ) is more important than production in response to hypoxia in inducing angiogenesis in melanomas . In simulations , even the most aggressive cancer cell strains are unable to induce angiogenesis without at least a low level of constitutive TAF production . Therefore , TAF must be produced constitutively by melanomas or by cooperating stromal cells . When an immune response is considered , it usually inhibits tumor growth , often destroying invasive tumors or holding them at a steady state for many years . These outcomes are observed in biologically reasonable parameter space , implying the immune response often plays a clinically meaningful role in the control of cancer growth . However , immune cells expressing TAF can also aid melanoma invasion by inducing tumorigenic angiogenesis . This can lead to a qualitative change in tumor behavior as non-invasive melanoma tumors restricted to the epidermis become aggressively invasive following immune induced angiogenesis . We have investigated a primary tumor seeding micro-metastases into the local skin tissue . This line investigation is motivated by the case study reported by De Giorgi et al . in [17] , where a patient presented with a large polypoid melanoma lesion that had reportedly been growing slowly for about three years . Following resection there was rapid recurrence in a number of previously dormant satellite metastases within 5–7 cm of the original lesion . This case was previously examined in a mathematical model of tumor dormancy by Boushaba et al . [16] . This is not an isolated case , as many cases of local melanoma recurrence are actually caused by local micro-metastases and not by incomplete resection of the primary tumor [15] . We propose that local metastatic spread can be reasonably modeled using two exponential rate parameters . The first , , gives the rate at which micrometastases shed from the primary tumor into the circulation . The second , , determines the distance from the primary tumor at which extravasation occurs . Under this model , simulations have suggested that moderately metastatic melanomas may induce local immune activation that is optimal for the suppression of metastatic disease . Melanomas that seed metastases only occasionally do not induce sufficient immune activity to destroy metastases beyond a threshold distance , while extremely metastatic melanomas overwhelm the immune response . This framework also suggests an explanation for the phenomenon of aggressive metastasis growth following surgical excision of a primary tumor . The immune response directed against a primary tumor can suppress nearby metastases . Following surgical excision most of the immune cells attacking the primary tumor are removed , as is the major source of cytokines attracting other immune cells to the sight . In the absence of this immune response , previously checked metastases can begin growing aggressively . The total mass and growth rate of these metastases can exceed that of the primary lesion , making this recurrence potentially more clinically dangerous . This phenomenon was also studied in a mathematical model by Boushaba et al . [16] who found that the release of growth inhibiting cytokines produced by the primary tumor is a possible mechanism . Anti-angiogenic factors released by the primary tumor , such as angiostatin and endostatin , are another major explanation [17] . Metastasis suppression by these factors and the immune response are not mutually exclusive hypotheses , and all may play a role . There is clearly a threshold distance beyond which local immune activation is insufficient to suppress metastases . This distance may be highest in moderately metastatic tumors . It is possible that immune activation plays the dominant role in suppression near the primary tumor , where most metastases are expected to extravasate . Further away , growth inhibition due to soluble factors may become dominant . However , our model has not taken into account circulating lymphocytes or antibodies that may play an important role . Despite its limitations , the overall implication of this work is that therapy targeting a primary tumor can perturb the host immune response in a way that allows increased growth in disseminated disease without altering any of the underlying parameters describing the system . The full model framework presented here can be translated into more focused systems aimed at addressing specific questions concerning melanoma invasion and treatment . With our model we have demonstrated that disruption of the immune response caused by surgical excision of a primary tumor is a possible mechanism for increased metastasis growth . Following surgery the wound healing response and associated inflammatory response probably also plays a role in cancer recurrence . A more detailed examination of tumor excision with a wound healing cascade could give insight into the importance of immune disruption . The effects of additional treatments such as chemotherapy or radiation therapy on the immune response and metastasis growth should also be investigated . Mathematical modeling may be particularly suited to examining the effect of different treatment schedules . At the least , predictions could be made concerning the efficacy of pre- versus post-operative adjuvant therapy . Finally , this framework provides an opportunity to investigate the nature and power of the natural selective forces at work driving the evolution of aggressive melanoma tumors . By incorporating multiple cell strains with differential parameter values , we can study the spatial and temporal requirements for successful mutant strain invasion of a pre-existing tumor and how allopathic intervention alters the balance of selective forces in and around the primary tumor . | Melanoma is a deadly skin cancer that invades into the dermis and metastasizes into the surrounding tissue . In clinical cases , surgical excision of the primary tumor has led to widespread and accelerated growth in metastases . We develop a mathematical model describing the basic process of melanoma invasion , metastatic spread , and the anti-tumor immune response . This model is formulated using partial differential equations that describe the spatial and temporal evolution of a number of different cellular populations , and it uses a realistic skin geometry . Using simulations , we examine the importance of the immune response when a primary tumor is spawning satellite metastases . We find that local metastases can be suppressed by the immune response directed against the primary tumor , but grow aggressively following surgical treatment . We also find that moderately metastatic tumors optimally activate the local immune response against disseminated disease , and in this case tumor excision may have profound effects on metastatic growth . We conclude that surgical perturbation of the immune response controlling local metastases is one mechanism by which cancer can recur . This could have implications as to the appropriate clinical management of melanomas and other solid tumors . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"oncology",
"mathematics",
"dermatology/skin",
"cancers,",
"including",
"melanoma",
"and",
"lymphoma",
"immunology/immune",
"response",
"oncology/skin",
"cancers",
"computational",
"biology",
"computational",
"biology/systems",
"biology"
] | 2009 | Tumor-Immune Interaction, Surgical Treatment, and Cancer Recurrence in a Mathematical Model of Melanoma |
Cytomegalovirus ( CMV ) is the most common cause of congenital infection , and is a major cause of sensorineural hearing loss and neurological disabilities . Evaluating the risk for a CMV infected fetus to develop severe clinical symptoms after birth is crucial to provide appropriate guidance to pregnant women who might have to consider termination of pregnancy or experimental prenatal medical therapies . However , establishing the prognosis before birth remains a challenge . This evaluation is currently based upon fetal imaging and fetal biological parameters , but the positive and negative predictive values of these parameters are not optimal , leaving room for the development of new prognostic factors . Here , we compared the amniotic fluid peptidome between asymptomatic fetuses who were born as asymptomatic neonates and symptomatic fetuses who were either terminated in view of severe cerebral lesions or born as severely symptomatic neonates . This comparison allowed us to identify a 34-peptide classifier in a discovery cohort of 13 symptomatic and 13 asymptomatic neonates . This classifier further yielded 89% sensitivity , 75% specificity and an area under the curve of 0 . 90 to segregate 9 severely symptomatic from 12 asymptomatic neonates in a validation cohort , showing an overall better performance than that of classical fetal laboratory parameters . Pathway analysis of the 34 peptides underlined the role of viral entry in fetuses with severe brain disease as well as the potential importance of both beta-2-microglobulin and adiponectin to protect the injured fetal brain infected with CMV . The results also suggested the mechanistic implication of the T calcium channel alpha-1G ( CACNA1G ) protein in the development of seizures in severely CMV infected children . These results open a new field for potential therapeutic options . In conclusion , this study demonstrates that amniotic fluid peptidome analysis can effectively predict the severity of congenital CMV infection . This peptidomic classifier may therefore be used in clinical settings during pregnancy to improve prenatal counseling .
Cytomegalovirus ( CMV ) is the most common cause of congenital infection with an incidence of 0 . 7% at birth [1] . Congenital CMV infection is the leading cause of non-genetic hearing loss and the most frequent viral cause of neurodevelopmental delay . Primary maternal CMV infection in pregnancy carries a 30% to 40% risk of vertical transplacental transmission , and 10% of those infected fetuses will be born as infected infants with clinical symptoms and long-term disabilities including sensorineural hearing loss and cognitive deficits such as mental retardation , cerebral palsy or seizures [2] . In addition , 5 to 10% of asymptomatic infants will develop milder forms of sensorineural hearing loss and of psychomotor delay later in life [2] . When maternal primary infection is documented , it is important to evaluate the risk to the fetus of being infected and/ symptomatically affected by CMV in order to provide appropriate counseling to these pregnant women [3] . Fetal CMV infection is diagnosed by amplification of the viral DNA in the amniotic fluid obtained by amniocentesis . However , a diagnosis of fetal CMV infection does not equate to a symptomatic neonate since 80% to 90% of fetuses with congenital CMV infection are asymptomatic at birth . Prenatal assessment of prognosis is mainly based upon ultrasound or MRI imaging of the fetal brain or fetal blood biochemistry . When severe ultrasound features are present at the time of diagnosis in the second trimester of pregnancy , the prognosis is known to be poor [4][5] . However , when the infected fetus shows either no or only non-severe ultrasound anomalies , the prognosis is difficult to establish . Indeed , although the negative predictive value of fetal imaging has been reported to be as high as 90% [6 , 7] , damages to the fetal brain may be delayed or masked to ultrasound , and even to magnetic resonance imaging ( MRI ) , up until late in the third trimester or at birth . Fetal blood sampling by cordocentesis under ultrasound guidance has been advocated for the additional prognostic value of fetal platelet count , CMV DNA levels and beta-2-microglobulin serum levels [8 , 9] . However , this procedure is invasive and is associated with a 1–3% risk of fetal loss . Finally , the prognostic value of the levels of CMV DNA loads in amniotic fluid is controversial [10–13] . Altogether , the prenatal prognosis of fetal CMV infection remains difficult to establish . In the past , the only option discussed when infected fetuses were suspected of severe disabilities was elective termination of pregnancy ( TOP ) . Alternative medical strategies such as CMV hyperimmune globulins and other antiviral drugs are being evaluated with recent studies showing conflicting results on the incidence of sequelae among treated infected fetuses [14][15] . Therefore , the identification of new effective prognostic markers is critical to help counseling pregnant women carrying an infected fetus through the dilemma of continuing or not with their pregnancy but also to decide upon starting medical therapy in utero . Amniotic fluid has been subjected to proteome analysis ( i . e . analysis of the global protein content of a sample ) for the identification of biomarkers of several prenatal conditions [16] . Specific amniotic fluid proteins have been related to various conditions affecting the fetus including Down-syndrome [17][18] , intrauterine inflammation leading to preterm birth [19][20] , preterm premature rupture of the membranes [21] and intrauterine growth restriction [22] . Moreover , we and others have shown that capillary electrophoresis coupled to mass spectrometry ( CE-MS ) analysis of body fluids can help identifying peptide-based biomarkers of disease which can be clinically relevant [23–25] . In this study , our aim was to analyze the amniotic fluid peptidome of CMV infected fetuses at mid-gestation and to evaluate the presence of peptide biomarkers that could contribute to improve the prognostic evaluation of this condition .
CE-MS-based peptidome analysis using 150 μL of amniotic fluid from 47 samples from CMV infected fetuses allowed the detection of a total of 4076 different peptides in all samples ( Fig 1A ) . The samples were further divided in a discovery and validation cohorts ( Table 1 ) . The discovery cohort consisted of 26 pregnancies with primary CMV infection including 13 severely symptomatic cases ( Table 1 ) and 13 asymptomatic cases . All amniotic fluid samples were collected between 17 and 29 weeks of gestation with a mean of 22 weeks of gestation in the asymptomatic group and of 23 weeks in the severely symptomatic group ( p = 0 . 43 , not significant ) ( Table 1 and S1 Table ) . Follow-up of asymptomatic infants was obtained up to 12 to 53 ( mean 24 ) months of age . In the severely symptomatic group , 12 fetuses underwent TOP for severe brain lesions and all but one were confirmed at necropsy ( Table 2 ) . One child was born alive but developed as severely handicapped ( Table 2 ) . Comparison of the amniotic fluid peptides content of the discovery cohort led to the identification of 76 peptides that were consistently differentially excreted between severely symptomatic and asymptomatic cases between these two groups ( Fig 1B , S1 and S2 Datasets ) . These 76 amniotic fluid peptides are identified by a unique tag composed of specific mass and migration time . In the next step , using a combination of LC-MS/MS and CE-MS/MS , sequence information could be obtained in 37 of these 76 peptides . Most of these peptides were fragments of various collagens . However , a number of other non collagen-related peptides ( 13 out of 37 ) were also observed , including a fragment of hemoglobin subunit delta and of beta-2-microglobulin displaying a strong increase in abundance in amniotic fluid of symptomatic patients ( >8-fold change ) ( Table 3 ) . The 37 sequenced peptides were reduced to a support vector machine ( SVM ) classifier with 34 peptides , called “CMV34” , by a leave-one-out procedure whereby the classifier was assessed for all peptide candidates minus one . Peptides that did not influence the accuracy of the classifier in the total cross-validation of the training data were left out of the final classifier . Scoring the patients from the discovery cohort with the resulting CMV34 classifier clearly separated the majority of the symptomatic from asymptomatic patients ( Fig 1C ) . In the next step the CMV34 classifier was validated in a separate blinded cohort using 21 amniotic fluid samples of primary CMV infection pregnancies . Amniocentesis was performed between 18 and 30 weeks of gestation with means of 25 and 23 weeks in the asymptomatic and in the severely symptomatic groups respectively ( p = 0 . 49 , not significant ) ( Table 1 and S1 Table ) . The cohort consisted of 9 severely symptomatic and 12 asymptomatic fetuses that were not used in the discovery phase ( Table 1 ) . In the severely symptomatic group , all cases underwent TOP for severe brain lesions ( Table 4 ) . Six out of the 9 symptomatic fetuses had an autopsy and all confirmed the severity of the lesions . In the other 3 cases , although autopsy was not performed , prenatal imaging described indisputably severe cerebral lesions on ultrasound that correlate well with all other cases terminated on the same basis and confirmed by postmortem examination . We have therefore no reason to believe that such a rich cerebral semeiology could be a false positive interpretation . The mean postnatal follow-up of the asymptomatic group was 8 months . Amniotic fluid samples were analyzed by CE-MS ( S3 and S4 Datasets ) , scored using the CMV34 classifier and then compared with the clinical data . We first verified that the CMV34 classifier score was independent of gestational age ( Spearman r = -0 . 3992 , not significant ) ( Fig 2A ) . The CMV34 classifier predicted a symptomatic/asymptomatic outcome with 89% sensitivity and 75% specificity with an area under the curve ( AUC ) of 0 . 90 [95% CI: 0 . 68 to 0 . 98] ( Fig 2B and Table 5 ) . The distribution of CMV34 scores for the validation cohort also showed highly significant separation of the 2 populations ( p = 0 . 0025 , Fig 2C ) . Among the 12 asymptomatic fetuses , 3 were wrongly classified as symptomatic by the CMV34 classifier when considering a score above/below 0 as threshold ( VC_AS2 , 1 . 523; VC_AS6 , 0 . 122; VC_AS8 , 0 . 004 ( S1 Table ) . Imaging data showed that for the first misclassified asymptomatic fetus , the only ultrasound symptom was hyperechogenic bowel . MRI done at 32 weeks of gestation did not show any abnormality . The child was asymptomatic at 24 months of age . For the second misclassified fetus , no abnormalities were seen at antenatal ultrasound nor at MRI . The child was still asymptomatic at 8 months of age . The third fetus showed no abnormalities were seen at antenatal ultrasound nor at MRI . The child was asymptomatic at birth and then lost for follow-up . Among the 9 fetuses considered as severely symptomatic , 2 were classified as asymptomatic by the classifier with a score of -0 . 136 ( VC_S16 ) and of -1 . 164 ( VC_S20 ) ( Table 4 and S1 Table ) . We next compared the performance of the CMV34 classifier with other frequently used , but controversial [8 , 9] [10–13] , parameters used for the assessment of the severity of the CMV infection including CMV DNA levels in amniotic fluid and fetal blood and the fetal platelet count . Since the number of missing values for these parameters was high , mostly in the symptomatic fetuses ( S1 Table ) we combined the discovery and validation cohorts . The classifier displayed a higher AUC than the clinical parameters ( Fig 3 and Table 5 ) . However , the sensitivity of CMV DNA levels in fetal blood was slightly higher , 92 versus 89% for CMV34 ( Table 5 ) . Based on a prevalence of 13% risk of being symptomatic at birth among infected fetuses [1] , we also assessed the positive ( PPV ) and negative ( NPV ) predictive values . All four parameters displayed high NPV , while the classifier and CMV DNA levels in amniotic fluid showed highest and comparable positive predictive value ( Table 5 ) . A small group of 9 cases with moderate postnatal symptoms ( i . e . hearing loss and/or vestibular dysfunction ) was scored separately using the CMV34 classifier ( Table 1 and S2 Table ) . Amniocentesis was carried out between 19 and 28 ( mean 23 ) weeks . Postnatal follow-up was of 12 to 84 ( mean 34 ) months . All 4 children suffering from vestibular dysfunction and 2 of 5 with isolated hearing loss were classified as symptomatic prenatally by the CMV34 peptide classifier ( S2 Table ) . The comparison of the combined CMV34 scores of these 2 sub-groups ( hearing loss and vestibular dysfunction , HL+VD versus hearing loss only , HL ) showed a trend towards higher scores for children with vestibular dysfunction ( p = 0 . 06 ) ( Fig 4A ) . Interestingly such trend could not be observed for CMV DNA levels in amniotic fluid and fetal blood and the fetal platelet count ( Fig 4B–4D and S2 Table ) . An advantage of non-targeted approaches leading to panels of peptides is that such data can be subjected to pathway analysis aiming at a deeper understanding of the underlying pathophysiology of the disease . The 13 non-collagen peptides ( considering peptides as proteins ) were submitted to Ingenuity Pathway Analysis software . The second top canonical pathway ( ignoring the first pathway related to HER-2 signaling in breast cancer ) was “Virus Entry via Endocytic Pathways” confirming the relation of 3 ( beta-2-microglobulin ( B2M ) ; Phosphatidylinositol 4;5-bisphosphate 3-kinase ( PIK3CB ) and protein kinase and Serine/threonine-protein kinase D1 ( PRKD1 ) ) out of the 13 peptide markers with the disease under study . When focusing on enrichment of pathways related to diseases , “Organismal Injury and Abnormalities” was among the top pathways suggesting a general connection to the potential lesions observed in symptomatic CMV . Another major enrichment of the peptide biomarkers was found to be in “Neurological Disease” with 7 out of 13 peptides significantly enriched ( Fig 5 ) : beta-2-microglobulin ( B2M ) , hemoglobin delta subunit ( HBD ) , adiponectin ( ADIPOQ ) , protein scribble homolog ( SCRIB ) , T calcium channel alpha-1G ( CACNA1G ) , DNA replication licensing factor ( MCM6 ) and tuberin ( TSC2 ) . Neurological lesions are the main prognostic lesions among the manifestations observed in symptomatic CMV patients including epilepsy .
The management of fetal CMV infection remains controversial . One difficulty is to establish the prognosis timely and accurately before birth . Prognostic factors are mainly derived from prenatal ultrasound or MRI imaging of the fetal brain and these can either appear late in the pregnancy and/or be objectified only after birth . Laboratory markers of the severity of infection including thrombocytopenia and high serum levels of beta-2-microglobulin in fetal blood have been suggested to precede the development of brain lesions . These can be obtained following cordocentesis , another invasive procedure performed under ultrasound guidance and separate from the amniocentesis performed for diagnosis [8 , 9 , 26] . We therefore aimed at identifying reliable protein-based prognostic factors of fetal CMV infection in the amniotic fluid at the time of prenatal diagnosis by amniocentesis . Amniotic fluid of 26 fetuses infected with CMV contained 76 peptides showing a clearly distinct abundance between those leading to either severely symptomatic or asymptomatic children by the age of 12 months . Thirty-four of the 37 sequenced peptides were combined in a prognostic classifier called CMV34 . The classifier was independent of gestational age and reached 89% sensitivity , 75% specificity and an area under the curve of 0 . 90 to discriminate severely symptomatic from asymptomatic neonates in a validation cohort . The CMV34 classifier showed better AUC and NPV than the other biomarkers of severity of CMV fetal disease including fetal platelets , CMV DNA levels in fetal blood and in amniotic fluid . High NPV , leading to unambiguous identification of asymptomatic neonates , is crucial for clinical management . PPV values were low and comparable for all four biomarkers , which could be expected due to the low prevalence of symptomatic neonates among CMV infected fetuses . Hence , a score suggesting a symptomatic fetus should always to be confirmed by targeted imaging . These data support that the CMV34 classifier may be of substantial value in the clinical context of intrauterine CMV infection . The selected peptides included 21 collagen fragments suggestive of tissue remodeling differences between symptomatic and asymptomatic patients . The other 13 peptides that were significantly more abundant in amniotic fluid of severely symptomatic fetuses included fragments of beta-2-microglobulin , hemoglobin subunit delta and adiponectin . The beta-2-microglobulin and the hemoglobin subunit delta fragments were highly up-regulated ( >8 fold ) suggesting an important role of these peptides in congenital CMV disease . Beta-2-microglobulin is the light chain of class I major histocompatibility complex and is present in the majority of T- and B-lymphocytes . Fetal blood levels of beta-2-microglobulin have previously been reported as a prognostic marker of fetal CMV infection [27] and beta-2-microglobulin has also been described as an important antibacterial protein in amniotic fluid [28] . Increased serum levels of beta-2-microglobulin is also a predictive factor of CMV infection in adult renal transplant recipients [29] . In the context of this study , high levels of circulating beta-2-microglobulin might be suggestive of lymphoid cell stimulation . One other hypothesis could be that increased beta-2-microglobulin reflects intense viral replication since it has been shown that CMV binds to beta-2-microglobulin [30] . However , in our study the abundance of the beta-2-microglobulin fragment ( ID 44679 ) was neither correlated to CMV DNA levels in amniotic fluid nor to CMV DNA levels in fetal blood ( S1 Fig ) . In association with hemoglobin subunit alpha , hemoglobin subunit delta constitutes hemoglobin A2 , which represents the adult hemoglobin . The hemoglobin subunit delta peptide found in amniotic fluid is therefore of maternal origin . The high specificity and the large differences between symptomatic and asymptomatic groups ( 8 . 6-fold change ) led us to hypothesize that those findings are not related to contamination during amniocentesis . Vaibuch et al . have demonstrated that total hemoglobin can be detected in the amniotic fluid of all pregnant women , increasing in concentration with gestational age . They also demonstrated in the same study [31] that the total hemoglobin level in amniotic fluid was significantly higher in women with intra-amniotic infection and/or inflammation . We suggest that this finding of higher hemoglobin subunit delta in amniotic fluid from symptomatic fetuses could be due to a more severe inflammation related to CMV infection . There is emerging evidence indicating that the study of biological networks can provide insight into the pathobiology of disease and improve biomarker discovery . We used the Ingenuity Pathway Analysis software to identify networks related to the peptides identified within this classifier . The second top canonical pathway was related to virus entry via endocytic pathway . Mechanisms of CMV entry into the cells are not completely understood . There is evidence that CMV enters into fibroblasts using a pathway involving membrane fusion , whereas it enters epithelial and endothelial cells via an endocytic pathway involving macropinocytosis [32] . Up-regulation of the canonical pathways “virus entry” in the more severe cases probably reflects intensive viral multiplication making the case for prenatal antiviral treatment . Moreover , one of the top disease pathways was identified as “neurological” and included 7 proteins all up-regulated in the more severe cases: beta-2-microglobulin ( B2M ) , hemoglobin delta subunit ( HBD ) , adiponectin ( ADIPOQ ) , protein scribble homolog ( SCRIB ) , T calcium channel alpha-1G ( CACNA1G ) , DNA replication licensing factor ( MCM6 ) and tuberin ( TSC2 ) . A previous publication reported elevated level of adiponectin in the amniotic fluid of women with intra-amniotic infection compared to women without infection [33] . Adiponectin appears to be intimately involved in several neurodegenerative disorders that are associated with CMV brain disease , such as epilepsy and ischaemic stroke [34] Therefore , adiponectin up-regulation in severe cases probably underlines the potential importance of inflammatory response on fetal brain lesions; this inflammatory response could be a target for future antenatal therapeutic intervention . The protein scribble homolog ( SCRIB ) is involved in planar cell polarity and Scrib knockout animal models present with neural tube defects [35] . Up-regulation of SCRIB in fetuses with severely affected brains could be a counter-mechanism of the putative effect of CMV on the differentiation and the migration of neuronal stem cells . Tuberin is the product of gene TSC2 and is highly expressed in fetal neural tissue [36] . Mutation of TSC2 is responsible for the development of Tuberous Sclerosis whereas activation of TSC2 has been reported to increase autophagy of the neurons in a model of Parkinson‘s disease [37] . Finally , T calcium channel alpha-1G ( CACNA1G ) protein belongs to the neuronal T-type calcium channels that are critical contributors to membrane excitability in neuronal cells . These T-type channels have a role in ischemic neuronal cell death in brain undergoing oxygen-glucose deprivation [38] as could be the case in CMV infection . Up-regulation of T calcium channel alpha-1G protein might thus predict an activation of seizures , which are frequent consequences of severe CMV brain disease . Molecules that counteract the effect of T calcium channel alpha-1G could therefore also be an interesting field of development to help treating the symptoms of children with severe CMV brain disease . In this study , we did not seek for diagnostic markers of fetal CMV infection but we were investigating prognostic markers in established CMV infected cases . However , it would be interesting to further study whether these markers are specific for CMV infection or could represent , at least partially , global perturbations related to in utero infection . Therefore , the CMV34 classifier could be tested on amniotic fluid samples from uninfected fetuses , from other pregnancy-related conditions such as pre-eclampsia , or from other infectious diseases , such as chorioamnionitis . In conclusion , the analysis of the peptidome in the amniotic fluids infected by CMV led to the identification of a prognostic classifier based on 34 peptides that yielded a better performance than the currently used fetal biomarkers to predict the asymptomatic or symptomatic status at birth . This classifier highlights the importance of the intensity of viral replication . A subset of those highly up-regulated proteins is also linked within a neurological disease network . This study serves as basis for future investigations to assess the prospective prognosis value of the CMV34 in larger cohorts . Moreover , these findings provide new insight on the pathobiology of CMV-induced fetal brain lesions and they open the perspective of new therapeutic options .
Forty-seven amniotic fluid samples collected in the second trimester of pregnancy with CMV infected fetuses were extracted from the Necker Virology laboratory database and analyzed retrospectively . Prenatal data were reviewed , including fetal ultrasound and MRI examination and follow-up . Outcome was assessed by targeted neonatal examination or postmortem examination following termination of pregnancy ( TOP ) . All infected fetuses were followed-up in the Fetal Medicine Unit at Necker Hospital using the same management . Amniocentesis was performed either because of suggestive fetal ultrasound symptoms or in the context of documented maternal primary infection . Fetal blood sample by cordocentesis under continuous ultrasound guidance is part of the management proposed in all infected cases . Fetal blood viral DNA-load and fetal platelet count are measured to be used as part of the prognostic assessment of all infected fetuses in the Fetal Medicine Unit at Necker Hospital . Cerebral MRI was performed after 30–32 weeks of pregnancy in all cases that were not terminated . Amniotic fluid samples from women carrying a CMV infected fetus were divided into two cohorts . The discovery cohort was composed of 26 amniotic fluid samples obtained from infected fetuses that led to either termination of pregnancy ( TOP ) for severe brain lesions or symptomatic neonates ( n = 13 ) or to asymptomatic neonates ( n = 13 ) . A second blinded cohort was used for validation and was composed of 21 amniotic fluid samples obtained from infected fetuses that led to TOP or to symptomatic neonates ( n = 9 ) or from asymptomatic neonates ( n = 12 ) . All women undergoing amniocentesis in the Fetal Medicine Unit at Necker Hospital gave written informed consent for CMV prenatal diagnosis to be performed on amniotic fluid and for the amniotic fluid sample left-over to be used for research purposes . According to French laws , an ethics statement from an Institutional Review board was not required for this work . Moreover , all women gave written informed consent for the use of clinical data as they were included in different clinical studies on congenital CMV in the Fetal Medicine Unit at Necker Hospital . These studies were approved by the Institutional Review Board of University Paris Ouest ( IRB N°2011-001610-34 and 2013-A213-42 ) . Fetuses were classified as severely symptomatic when severe brain anomalies were identified by ultrasound , including ventriculomegaly measured ≥ 15mm , periventricular hyperechogenicity , hydrocephaly , microcephaly < -2DS , increased cisterna magna ≥ 8mm , vermian hypoplasia , porencephaly , lissencephaly , periventricular cysts or agenesis of the corpus callosum . In cases of TOP , severity was confirmed if postmortem examination showed microcephaly < 4SD , ventriculomegaly , white matter necrosis , associated with diffuse lesions of vasculitis and of encephalitis . Indeed , it has been documented in the literature that 77% of infected newborns with severe abnormalities on neonatal CT scan ( intracranial calcifications , ventricular dilatation , white-matter abnormalities , cortical atrophy and migration abnormalities ) develop at least one psychomotor sequela [39] . In another study 100% of newborns with microcephaly had severe mental retardation [40] . Therefore , severely abnormal brain imaging both on ultrasound and on MRI in utero is likely to be found mainly in cases destined to become neurologically symptomatic neonates . Symptomatic or asymptomatic status at birth was evaluated by clinical examination , biological assessment ( platelet count , liver enzymes , and bilirubin serum levels ) and audiometric assessment by automated auditory brainstem response ( AABR ) . Neonates with normal clinical examination , normal biological assessment and normal hearing were considered asymptomatic . Neonates with abnormal clinical examination and/or abnormal biological assessment , including petechiae , microcephaly , seizures , lethargy/hypotonia , poor suck , hepatosplenomegaly , thrombocytopenia and hearing loss were considered as symptomatic . Infected children are routinely followed-up with serial clinical examination and audiometric evaluation ( AABR ) performed at 4 , 12 , 18 , 24 and 36 months of age . For the purpose of the study , cases were classified into 3 groups . Cases with severe brain anomalies identified either at prenatal ultrasound or MRI , at necropsy or at birth were classified as severely symptomatic . Cases with isolated unilateral hearing loss > 40 decibels and/or vestibular syndrome identified at birth or at last follow-up visit were classified as moderately symptomatic . Asymptomatic cases at birth or at last available follow-up visit were classified as asymptomatic . CMV DNA was extracted from 200 μl of amniotic fluid with the MagNaPure LC using the Total Nucleic Acid extraction kit ( Roche Diagnostic , Meylan , France ) and from 200μl of fetal whole blood using the QiaAmp DNA mini blood kit ( Qiagen , Courtaboeuf , France ) . CMV DNA was amplified by CMV-R Gene ( Argene BioMerieux , France ) a real time commercial quantitative CMV PCR assay . We have recently developed a sample preparation protocol for the analysis of the fetal urinary peptidome by CE-MS starting from 150 μl of fetal urine [41] . This protocol was also used for amniotic fluid . Briefly , immediately before preparation , amniotic fluid aliquots kept at –80°C were thawed and 150 μl aliquots were diluted with the same volume of 2 M urea , 10 mM NH4OH containing 0 . 2% SDS . Subsequently , samples were passed over a Centristat 20-kDa cut-off centrifugal filter device ( Sartorius ) in order to eliminate high molecular weight compounds . The filtrate was desalted using a NAP-5 gel filtration column ( GE Healthcare ) to remove urea and electrolytes . Lyophilisation of the sample was performed using a Savant speedvac SVC100H connected to a Virtis 3L Sentry freeze dryer ( Fisher Scientific ) and stored at 4°C until use . Shortly before CE-MS analysis , the samples were re-suspended in 10μL of HPLC grade H2O . CE-MS analyses were performed as previously described [41–43] using a Beckman Coulter Proteome Lab PA800 capillary electrophoresis system ( Beckman Coulter ) on-line coupled to a micrOTOF II MS ( Bruker Daltonic ) . The electro-ionization sprayer ( Agilent Technologies ) was grounded , and the ion spray interface potential was set between –4 and –4 . 5 kV . Data and MS acquisition methods were automatically controlled by the CE via contact-close-relays . Spectra were accumulated every 3 s , over a range of m/z 350 to 3000 . Mass spectral ion peaks representing identical molecules at different charge states were deconvoluted into single masses using MosaiquesVisu software [44] . The software automatically examined all mass spectra from a CE-MS analysis for signals with a signal-to-noise ratio of at least 4 present in 3 consecutive spectra . Furthermore , the isotopic distribution was assessed , and charge was assigned based on the isotopic distribution , as well as conjugated masses , using a probabilistic clustering algorithm . This operation resulted in a list wherein all signals that could be interpreted are defined by mass/charge , charge , migration time , and signal intensity ( ion counts ) . Time-of-flight mass spectrometry ( TOF-MS ) data were calibrated utilizing Fourier transform ion cyclotron resonance mass spectrometry ( FT-ICR-MS ) data as reference masses applying linear regression . Both CE-migration time and ion signal intensity ( amplitude ) showed high variability , mostly due to different amounts of salt and peptides in the sample . Normalization of the amplitude of the amniotic fluid peptides was based on sequenced endogenous “housekeeping” peptides ( Table 6 ) that varied little among the samples . Based on the sequences most of the housekeeping peptides found in amniotic fluid were similar to those observed in urine , therefore allowing the application of a CE-MS procedure for sample analysis and data processing similar to the procedure used for urine [42 , 45] . Normalized levels of amniotic fluid peptides were compared between symptomatic and asymptomatic patients’ groups for the identification of amniotic fluid proteome biomarkers . Only peptides with a P-value<0 . 05 corrected for multiple testing ( Benjamini- Hochberg ) were considered significant [46] . The number of peptides with differential abundance was reduced to a support vector machine ( SVM ) classifier with 34 peptides ( CMV34 ) by a take-one-out procedure that had similar performance for the classification of the patients in the discovery CMV patient cohort . Sensitivity and specificity of the previously defined biomarker classifiers , and 95% confidence intervals ( 95% CI ) were calculated using receiver operating characteristic ( ROC ) plots ( MedCalc version 14 . 8 . 1 , MedCalc Software , Belgium ) . Candidate biomarkers and other native peptides from fetal urine and amniotic fluid were sequenced using LC-MS/MS and CE-MS/MS analysis [47] . LC-MS/MS analysis experiments were performed on a Dionex Ultimate 3000 RSLC nano flow system ( Dionex , Camberly UK ) . For CE-MS/MS , the samples were injected under constant flow and pressure conditions at a pH of 2 . 2 to ensure that all peptides are positively charged . Both , CE and LC , were directly interfaced with an LTQ-Orbitrap XL ( Thermo Finnigan , Bremen , Germany ) , using data-dependent HCD MS/MS sequencing of a maximum of the top 20 ions . All resultant MS/MS data were analysed using Proteome Discoverer 1 . 2 ( activation type: HCD; min-max precursor mass: 790–6000; precursor mass tolerance: 10 ppm; fragment mass tolerance: 0 . 05 Da; S/N threshold: 1 ) and were searched against the Uniprot human non-redundant database without enzyme specificity . No fixed modifications were selected , oxidation of methionine and proline and deamidation of aspartic acid and glutamine were selected as variable modifications . The peptide data were extracted using high confidence peptides , defined by an Xcorr≥1 . 9 , a delta mass between experimental and theoretical mass ± 5 ppm , absence of cysteine in the sequence as without reduction and alkylation it is forms disulphide bonds , absence of oxidized proline in protein precursors other than collagens or elastin , and top one peptide rank filters . For further validation of peptide identification , the strict correlation between peptide charge at pH 2 and CE-migration time was utilized to minimize false-positive identification rates [48] . Calculated CE-migration time of the sequence candidate based on its peptide sequence ( number of basic amino acids ) was compared to the experimental migration time . Peptides were accepted only if they had a mass deviation below ±50 ppm and a CE-migration time deviations below ±2 min . Ingenuity Pathway Analysis software ( version 24390178 , last release 18/06/2015 , www . ingenuity . com ) was performed using the parent proteins peptides with differential abundance in the symptomatic CMV infections . As urine , amniotic fluid is very rich in different collagen fragments . Therefore , collagens were omitted from the Ingenuity pathway analysis to avoid a bias in the analysis towards fibrosis and conjunctive tissue remodeling . All Ingenuity output related to cancer was omitted . | CMV is the most common cause of congenital infection , and can result in significant neonatal morbidity and neurological disabilities . The birth prevalence of congenital CMV is estimated at 0 . 7% worldwide , and 10 to 20% of these neonates develop severe symptoms . In such cases the outcome is generally poor . Therefore , identification of additional prognostic markers is crucial for prenatal counseling in cases with an infected fetus . This may influence the decision of continuing with the pregnancy or requesting its termination , but also the decision of starting experimental antiviral therapy . The pathophysiology of CMV brain injury is not completely understood , and the identification of new biomarkers of CMV infection might also pave the way towards the development of new therapeutic alternatives . Here , we apply a recently developed and modern non-targeted peptidomics approach to amniotic fluid obtained from symptomatic and asymptomatic CMV-infected fetuses/neonates , followed by network analysis of the peptides of interest in the context of fetal infection and in relation with outcome . Our study identified 34 amniotic fluid peptides that form new prognostic biomarkers that could be used in clinical settings to improve prenatal counseling . In addition , this study provides novel mechanistic insight into the pathobiology of CMV congenital disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Material",
"and",
"Methods"
] | [] | 2016 | Identification of Symptomatic Fetuses Infected with Cytomegalovirus Using Amniotic Fluid Peptide Biomarkers |
The identification of recessive disease-causing genes by homozygosity mapping is often restricted by lack of suitable consanguineous families . To overcome these limitations , we apply homozygosity mapping to single affected individuals from outbred populations . In 72 individuals of 54 kindred ascertained worldwide with known homozygous mutations in 13 different recessive disease genes , we performed total genome homozygosity mapping using 250 , 000 SNP arrays . Likelihood ratio Z-scores ( ZLR ) were plotted across the genome to detect ZLR peaks that reflect segments of homozygosity by descent , which may harbor the mutated gene . In 93% of cases , the causative gene was positioned within a consistent ZLR peak of homozygosity . The number of peaks reflected the degree of inbreeding . We demonstrate that disease-causing homozygous mutations can be detected in single cases from outbred populations within a single ZLR peak of homozygosity as short as 2 Mb , containing an average of only 16 candidate genes . As many specialty clinics have access to cohorts of individuals from outbred populations , and as our approach will result in smaller genetic candidate regions , the new strategy of homozygosity mapping in single outbred individuals will strongly accelerate the discovery of novel recessive disease genes .
The primary causes of most pediatric diseases remain elusive . However , it is becoming apparent that many pediatric diseases may represent recessive single-gene ( monogenic ) disorders . As an example , it was recently recognized that up to 25% of all cases with steroid resistant nephrotic syndrome ( SRNS ) in childhood are caused by recessive mutations of the NPHS2 gene , although SRNS had not previously been viewed as a genetic disorder [1]–[4] . Progress has been made in unraveling the primary causes ( etiologies ) of pediatric disorders by identifying recessive disease genes using positional cloning by homozygosity mapping [5] . Homozygosity mapping is performed by total genome scan with polymorphic markers in individuals with a recessive disease whose parents are related . It tests the assumption that a homozygous mutation in a recessive disease gene is “identical by descent ( IBD ) ” by segregating twice to the affected child from a common ancestor through both the maternal line and the paternal line ( Figure 1 ) . In addition , homozygosity mapping supposes that a short chromosomal segment surrounding the homozygous mutation has not been recombined by crossing over and that SNP ( single nucleotide polymorphism ) markers in a segment surrounding the mutation will therefore also be homozygous by descent ( Figure 1 ) [5] . This short segment of homozygosity by descent can then be detected by multipoint homozygosity mapping as a “ZLR peak” ( Figure 2 ) and will harbor the sought-after disease gene . Gene identification by positional cloning holds multiple promises: i ) It directly identifies the primary cause of a disease , thereby providing a secure starting point for the elucidation of the related pathogenesis . ii ) Positional cloning often leads to the discovery of novel genes , revealing new mechanisms of cell physiology and development . iii ) Since recessive single-gene disorders exhibit the strongest possible relationship between disease cause and disease phenotype , leading to a phenotype in virtually all individuals with mutations , they provide the most useful objective for molecular genetic diagnostics [6] . The number of causative genes for autosomal recessive disorders that are yet to be discovered is most likely in the thousands [6] . However , more rapid discovery of recessive disease genes is hampered by the lack of suitable consanguineous families with multiple affected children . In addition , mapping has resulted in too many putative loci that contain too many candidate genes for efficient gene identification . Considering these impediments to gene discovery , we tested the hypothesis that SNP arrays may systematically allow gene identification by homozygosity mapping in single individuals from non-consanguineous ( “outbred” ) populations . If this were possible , it would offer the following important advantages: i ) For any given disease phenotype , pediatric specialty clinics have access to far more patients who are from outbred populations and represent single cases than to patients who are from consanguineous background or represent sibling cases . ii ) Homozygosity mapping in individuals who are from outbred populations but bear homozygous disease gene mutations by descent from an unknown distant ancestor may provide a single genomic candidate region small enough to allow successful gene identification by mutation analysis .
We obtained blood samples and clinical and pedigree data following informed consent from individuals with two different groups of autosomal recessive kidney diseases , nephronophthisis ( NPHP ) or steroid resistant nephrotic syndrome ( SRNS ) . Human subjects research was approved by the University of Michigan Institutional Review Board . The diagnoses were made by nephrologists based on standardized clinical criteria [7] , [8] . Renal biopsies were evaluated by renal pathologists . Clinical data were obtained using standardized questionnaires ( www . renalgenes . org ) . In 1 , 069 families with a NPHP-like phenotype , who were referred to us for mutation analysis from worldwide sources over the course of 16 years , we performed mutation analysis by direct exon sequencing in the genes NPHP1 through NPHP9 , AHI1 , and MKS3 . Additionally , in 404 families with steroid-resistant nephrotic syndrome ( SRNS ) , who were referred to us worldwide for mutation analysis over the course of 8 years we performed mutation analysis by direct exon sequencing in the genes NPHS1 ( nephrin ) and NPHS2 ( podocin ) . We performed genome-wide scans for linkage in 72 individuals from 54 families , who had a homozygous mutation in any of the 13 genes that cause NPHP or SRNS ( NPHP1 through NPHP9 , AHI1 , MKS3 , NPHS1 or NPHS2 ) . We used single nucleotide polymorphism ( SNP ) arrays ( GeneChip ) from Affymetrix , Inc . Most individuals ( n = 40 ) were examined at 250 K resolution ( Human Mapping 250 K StyI Array ) , 9 families at 50 K resolution ( 50 K Hind Array ) , and 5 families at 10 K resolution ( 50 K Xba Array ) . Samples were processed , hybridized , and scanned using the manufacturer's' standard methods at the University of Michigan Core Facility ( www . michiganmicroarray . com ) . As non-parametric likelihood ratio z-scores ( ZLR scores ) provide a good estimate of excess allele sharing [9] , and since the GENEHUNTER [10] NPL score is conservative [11] , we used the ZLR score implemented in ALLEGRO [12] to detect narrow segments of homozygosity . Parameters were set for a disease allele frequency of 0 . 001 , and marker allele frequencies for Caucasians as specified by Affymetrix . We calculated ZLR scores under the hypothesis of consanguinity , using a standard pedigree structure first cousin marriage for parents of affected individuals . For single affecteds a non-existent sibling was included to enable non-parametric ALLEGRO runs . ZLR scores were plotted over genetic distance across the entire human genome using the GNUPLOT software ( http://www . gnuplot . info/ ) . In this way , maxima of ZLR scores , “ZLR peaks” , are expected to reflect segments of homozygosity by descent , which may harbor the homozygous mutation of the recessive disease gene . In the following we will refer to the ZLR plots as “homozygosity profiles” as described in Figure S1 . Using ALOHOMORA [13] , ZLR scores were calculated using one marker every 100 , 000 nucleotides of human genomic sequence under three different conditions , i . e . for minor allele frequencies of >0 . 2 , >0 . 3 , and >0 . 4 , respectively ( Figure S1 ) . In a small cohort of patients with known homozygous disease-causing mutations in NPHS2 we established that ZLR peaks , which exceed the value of 2 . 0 under two out of the three conditions ( minor allele frequencies of >0 . 2 , >0 . 3 , and >0 . 4 ) , did in fact contain the known homozygous disease-causing mutation in this ZLR peak and exhibited continuous segments of homozygosity upon haplotype inspection . We therefore refer to peaks that fulfill these criteria as “consistent ZLR peaks” ( “cZLR” peaks ) ( Figure S1 ) . All ZLR peaks of homozygosity considered “consistent” under these criteria were also present when using the criterion of 0 . 1 as cut off for minor allele frequency . We later confirmed that in 67 of 72 cases ( 93% ) the homozygous mutations in the causative gene were positioned within a cZLR peak ( see Results ) . Thus , cZLR peaks can be utilized to indicate the position of homozygous mutations in recessive disease genes with high likelihood . We demonstrate below that cZLR peaks very rarely occurred in negative control individuals without homozygous mutations ( see Results ) . Physical distance for SNP markers: genome browser , human , May 2004 at http://genome . ucsc . edu
Homozygosity mapping can only be applied to individuals that bear homozygous mutations of recessive disease genes . Therefore , we first evaluated whether in a significant fraction of patients with mutations in 13 different recessive disease genes the disease was caused by homozygous rather than compound heterozygous mutations . We chose 9 different genes ( NPHP1-9 ) [14] that cause the recessive kidney disorder nephronophthisis ( NPHP ) , 2 genes ( AHI1 and MKS3 ) causing NPHP-like phenotypes , and 2 genes ( NPHS1 and 2 ) that cause steroid resistant nephrotic syndrome ( SRNS ) [2] , as we had access to DNA from worldwide cohorts of patients with these disorders . NPHP is the most frequent genetic cause of terminal kidney failure in children and young adults . Positional cloning of nine responsible genes ( NPHP1-9 ) has revealed NPHP as a “ciliopathy” , relating its pathogenesis to dysfunction of primary cilia signaling [14] . In a worldwide cohort of 1 , 069 families with NPHP we detected the disease-causing mutation in 320 ( 30% ) . We excluded from the analysis 224 families with large deletions of NPHP1 , as they are due to a unique mechanism of genomic DNA rearrangement [15] . In the remaining 96 families , the fraction of homozygous mutations for each gene was as follows: NPHP1 ( 2/2; 100% ) , NPHP2 ( 6/11; 55% ) , NPHP3 ( 4/6; 66% ) , NPHP4 ( 15/20; 75% ) , NPHP5 ( 22/26; 85% ) , NPHP6 ( 3/10; 30% ) , NPHP7 ( 1/1; 100% ) , NPHP8 ( 6/6; 100% ) , NPHP9 ( 1/1; 100% ) , AHI1 ( 2/5; 40% ) , and MKS3 ( 5/8; 63% ) . Thus , the fraction of homozygous mutations for the 11 different NPHP-causing genes together was 70% ( 67/96 ) and ranged from 30–100% . The other patients had compound heterozygous mutations in these genes . In patients with NPHP there was some bias towards selection of consanguineous kindred . However , we have also ascertained a worldwide cohort of 404 families with SRNS for mutation analysis in the genes NPHS1 and NPHS2 without selection for consanguineous kindred . In 73/404 ( 18% ) of these families we detected both mutated alleles of NPHS2 . Thirty-three of the 73 ( 45% ) had homozygous mutations [4] . NPHS1 mutations were detected in 36 of 86 ( 42% ) families with congenital nephrotic syndrome . Twenty-four of the 36 ( 67% ) had homozygous mutations ( Heeringa , unpublished ) . We thus demonstrate that for 13 different recessive disease genes a very high fraction ( 124/205; 60% ) had homozygous mutations , although ascertainment was mostly by self-recruitment for mutation analysis and not specifically directed at patients from a consanguineous background . We conclude that in a large proportion of patients from worldwide cohorts with mutations in various rare recessive disease genes homozygosity mapping is potentially feasible . We next tested the assumption that homozygous mutations of recessive disease genes are virtually always localized in segments of contiguous homozygous SNP markers derived from a ( distant ) common ancestor and can therefore be located by homozygosity mapping . The rationale of homozygosity mapping is summarized in Figure 1 . In addition , we wanted to determine the shortest detectable size of homozygous segments as short segments will reduce the number of candidate genes . From our worldwide cohort of 124 families with known homozygous mutations in 13 different recessive genes ( NPHP1-9 , AHI1 , MKS3 , NPHS1 and NPHS2 ) we selected for total genomes scan 72 patients from 54 different families that had sufficient DNA available . The selected families represented the entire spectrum of consanguinity from first degree cousin relations to outbred populations . Families , patients , and their homozygous mutations are listed in Table S1 . Following total genome search by homozygosity mapping primarily using 250 K SNP arrays we plotted for each individual “homozygosity profiles” across the genome as defined in Figure S1 . Examples are given in Figure 2 for three individuals with homozygous mutations in NPHP3 , NPHP2 , and NPHP5 , respectively . Homozygosity plots of all 72 cases are available from the authors . Generation of homozygosity profiles suggested that the number of cZLR peaks per genome reflected the degree of consanguinity of an individual ( Figure 2 ) . For example , individual A14-2 who is from a consanguineous Venezuelan kindred with 17 documented consanguinity loops , exhibited 28 cZLR peaks ( Figure 2A ) , whereas individual A8-1 , who is not aware of consanguinity but originates from a geographic region with frequent consanguinity ( Turkey ) exhibited 12 cZLR peaks ( Figure 2B ) . Only one of the 12 cZLR peaks colocalized with one of the 9 different NPHP loci . This was the NPHP2 locus , in which the patient has a homozygous disease causing mutation ( Figure 2B ) . This demonstrates that in diseases with multiple responsible loci the generation of a homozygosity profile can help select the locus relevant for mutation analysis from a high number of alternative disease loci . Thus , homozygosity profiles will reduce the number of exons to be examined in molecular genetic diagnostics . In the example given it was reduced from 220 exons for NPHP1 through NPHP9 to 16 exons for NPHP2 only . In contrast , individual F399-1 , who originates from a geographic region where consanguinity is rare ( Germany ) exhibited only one single cZLR peak of 8 . 2 Mb width , which harbored the disease-causing homozygous mutation in NPHP5 ( Figure 2C ) . These data gave a first indication that the number of ZLR peaks per genome reflect the degree of consanguinity and that homozygous mutations in outbred populations can be localized to a single cZLR peak that is unique to an individual's genome . The strategy of homozygosity mapping carries the theoretical risk that the mutation in the recessive disease gene may not be detected in a homozygosity peak . In order to test whether homozygous disease-causing mutations were practically always positioned in cZLR peaks , we generated homozygosity plots as defined in Figure S1 ( see also Figure 2 ) of all 72 individuals from 54 families with homozygous mutations in one of the 13 different recessive genes tested . We found that in 67 of 72 cases ( 93% ) the homozygous mutation of the causative gene was in fact positioned within a cZLR peak ( data available from the authors ) . In the five individuals with absence of a cZLR peak the mutation was embedded in a homozygous haplotype too short for detection ( see below ) . In the 13 rare recessive disease genes studied , we thus confirmed the hypothesis that most homozygous mutations ( 93% ) can be mapped to a cZLR peak . To examine the relationship between the number of cZLR peaks and degree of consanguinity we plotted the number of cZLR peaks for each of the 72 individuals from 54 families with homozygous disease-causing mutations ( Figure 3 ) . The number of cZLR peaks ranged from 60 to 0 ( Figure 3 ) . When we sorted families from left to right from highest to lowest number of cZLR peaks , the number of cZLR peaks reflected the degree of inbreeding in each family as follows ( Figure 3 ) : i ) Families with known consanguinity ( orange highlighting in Figure 3 ) clustered in the left third of the plot with the number of cZLR peaks ranging from 60 to 8 . Fourteen of fifteen consanguineous families ( orange ) had 8 or more cZLR peaks . The highest number of cZLR peaks was seen in the following closely inbred families: A1337 ( 1st degree cousins , Turkey ) , F617 ( Native American ) , A1175 ( 1st degree cousins , Turkey ) , A1125 ( known consanguinity , Turkey ) , A364 ( known consanguinity ) , A14 ( 17 consanguinity loops , Venezuela ) , and F3 ( 1st degree cousins ) [16] . ii ) Families without known consanguinity , but originating from geographic regions where consanguinity is frequent ( yellow highlighting in Figure 3 ) , clustered in the middle third of the plot with , 8 of 12 families having 8 to 4 cZLR peaks . These families originated mostly from Turkey and Arab countries ( Table S1 ) . iii ) Outbred families , i . e . families without known consanguinity and originating from geographic regions where consanguinity is rare ( no highlighting in Figure 3 ) , clustered in the right third of the plot with cZLR peaks . Twenty-three of the 27 outbred families had 4 to 0 cZLR peaks ( 0 denoting that no cZLR peak was detected ) . These families originated mostly from Western Europe and the US ( Table S1 ) . No family from a consanguineous population had 3 or less cZLR peaks . We thereby demonstrated through graphical evaluation of total genome homozygosity profile data , that the number of cZLR peaks of individuals reflected their degree of inbreeding and thereby the extent of homozygosity by descent . Specifically , the number of homozygosity peaks was 8–60 in families with known consanguinity , 4–8 in families with unknown consanguinity but originating geographic background “at risk” for consanguinity , and it was 0–4 in families from outbred populations . Homozygosity profiles can therefore be viewed as “consanguinograms” that reflect the degree of inbreeding of an individual . The data set generated ( Figure 3 ) gave us the opportunity to examine the limit of resolution for detecting short segments of homozygosity in single individuals with homozygous disease-causing mutations from outbred populations . Until now , successful gene identification by homozygosity mapping has been mostly based on consanguineous kindred that have multiple affected individuals . Close consanguinity of such kindred generate homozygous segments broad enough to be detected by SNP marker sets of low density ( e . g . , 50 K marker set ) . The presence of multiple siblings or cousins helps refine the candidate region to fewer cZLR peaks that overlap between affected siblings or cousins . For example , the number of 9 , 12 and 10 cZLR peaks in siblings A1538-1 , -2 , and -3 , respectively is reduced to 3 , 4 , and 5 cZLR peaks if each pair of siblings is evaluated together and to 2 cZLR peaks if all three sibs are evaluated together ( Figure 3 ) . However , a severe limitation to successful positional cloning of recessive disease genes by homozygosity mapping lies in the fact that sufficient numbers of consanguineous kindred with multiple affecteds and mutation of the same gene are very hard to ascertain . In our 16-year experience in worldwide recruitment of 1 , 069 patients with NPHP , non-consanguineous single cases ( n = 918 ) were 6-times as frequent as consanguineous cases ( n = 151 ) . In 595 patients with SRNS and mutations in recessive disease-causing genes the ascertainment of non-consanguineous single cases ( n = 512 ) was also about 6-times as high as consanguineous cases ( n = 83 ) . Therefore , the ability to employ individuals from outbred populations in homozygosity mapping should greatly accelerate gene discovery . Another strong limitation to gene identification is imposed by the experience that studies of consanguineous kindred result in segments of homozygosity that are too numerous and too large and contain too many positional candidates for efficient gene identification by mutation analysis . For instance , in our worldwide cohort individuals from consanguineous background had 4–60 cZLR peaks ( Figure 3 ) . These homozygous segments contained hundreds of positional candidate genes . Because of both limitations we wanted to evaluate whether gene identification by homozygosity mapping is possible in single individuals from outbred populations . Using a medium resolution SNP marker set ( 250 K ) we were able to detect homozygous segments that contain the disease-causing mutations in single individuals from outbred populations as shown in Figure 3: With the exception of 4 families ( A1218 , A646 , F1158 , and A825 ) , all other 23 outbred families exhibited 4 or less cZLR peaks ( Figure 3 ) . In 2 of the 3 outbred families with 4 cZLR peaks , calculation of both affected individuals together reduced the number of cZLR peak from 4 and 3 cZLR peaks to 2 cZLR peaks ( A237 ) and from 4 and 2 peaks to 1 cZLR peak ( F60-61 ) . Four outbred families had 3 cZLR peaks ( F138 , F456 , F50 , and F1183 ) . Seven outbred individuals exhibited only 2 cZLR peaks ( A762 , F53 , A1298 , A567 , A7 , F54 , and A1338 ) ( Figure 3 ) . In these individuals the cZLR peak that contained the homozygous gene mutation had a median size of 8 . 3 Mb ( range 2 . 4 - 40 . 1 Mb ) ( Figure 3 , insert ) . Most interestingly , 4 outbred families exhibited only 1 cZLR peak ( F399 , A1730 , F30_10G , and F408 ) ( Figure 3 ) . These peaks , which contained the homozygous gene mutation , had a median size of only 2 . 7 Mb ( range 2 . 1–8 . 25 ) . A genomic segment of 2 . 1 Mb is equivalent to an average of 16 candidate genes , a number of genes that permits gene identification by exon sequencing in candidate genes . For comparison , we have so far successfully identified disease genes by mutation analysis of all exons within intervals of 8 . 7 Mb ( 57 genes ) in NPHP5 [17] , 1 . 5 Mb ( 9 genes ) in NPHP6 [18] , 0 . 94 Mb ( 7 genes ) in NPHP3 [19] , 0 . 90 Mb ( 11 genes ) in BSND [20] , 0 . 70 Mb ( 6 genes ) in NPHP4 [21] , and 4 . 0 Mb ( 43 genes ) in PLCE1 [22] . We thereby demonstrated that we were able to detect a cZLR signal of “homozygosity by descent from the disease allele” above the background of “homozygosity by descent from haplotype blocks common to the outbred population” . To be detectable , the former would have to be more recent and therefore reside on a longer segment of homozygosity than the latter . Finally , in 5 families ( F409 , A159 , A1686 , A1685 , and A887 ) , in whom the known homozygous mutation was positioned in homozygous segments of less than 2 . 1 Mb , we did not obtain any cZLR peak ( see Table 1 and inset in Figure 3 ) . Therefore , we conclude that the limit for detecting a homozygous segment in single outbred individuals with homozygous disease-causing mutations using this technique was 2 . 1 Mb . Three observations further substantiate this detection limit: i ) We demonstrated for the first time that the R138Q mutation of NPHS2 does in fact represent a founder mutation as it occurs on identical haplotypes ( Figure S2 ) . It was detected within a homozygous interval of 2 . 7 Mb ( 212 SNPs ) in both individuals of family A1730 ( Table 1 and Figure S2 ) . It was also detected within 2 . 3 Mb of homozygosity in families A237 , and A825 , who had additional cZLR peaks on other chromosomes ( Figure S2 ) . However , the R138Q mutation was not detected within homozygous intervals of 1 . 2 Mb , 1 . 2 Mb , and 0 . 70 Mb , in families A159 , A1686 , and A887 , respectively ( Table 1 and inset in Figure 3 ) . ii ) In family F30 a single cZLR peak of 2 . 1 Mb width bearing a homozygous NPHP4 mutation was detected in individual F30-1 , but not within the sibling's ( F30-2 ) interval of 1 . 14 Mb , which is shorter due to a recombination in the mother's meiosis ( Table 1 ) . We documented that the parents of F30 are related 10 generations ago . This gives an example of the “remoteness” of inbreeding that leads to homozygous intervals in the order of ∼1–2 Mb [23] . iii ) Families F408 and F409 bear the same homozygous mutation ( F142fsX14 ) in NPHP5 and originate from the same geographic region in Switzerland . We demonstrate that this mutation is embedded in a shared haplotype ( Figure S3 ) . Whereas we detected the 5 . 18 Mb homozygous interval as a single cZLR peak in F408 , the shorter haplotype of 0 . 65 Mb in F409 was not detected as a cZLR peak ( Table 1 and Figure S3 ) . In 28 unrelated individuals with recessive disease from outbred background with known homozygous mutations in recessive genes we failed to detect a “ZLR peak of homozygosity” in 6 ( 21% ) . In all of these patients the homozygous mutation resided in a homozygous segment of less than 2 . 1 Mb ( Figure 3 and Table 1 ) . Thus , the limit of detection of a homozygous mutation using our homozygosity mapping strategy was at 2 . 1 Mb ( Table 1 , Figures S2 and S3 ) . It is remarkable however , that even if homozygous mutations were not detected as a cZLR peak they were always embedded in a homozygous haplotype of >0 . 65 Mb . This indicates that it may be possible to further improve the sensitivity of the approach . The sensitivity for detecting a cZLR peak in an individual with a homozygous mutation was 93% ( 67/72 ) for all 72 patients ( see above ) . It was 76% ( 16/21 ) if only the 21 outbred individuals with 4 or less cZRL peaks were evaluated . In order to assess the specificity of the method , i . e . whether or not cZLR peaks occur frequently as “false positives” in individuals without homozygous mutations in outbred populations , we examined 20 parents of affected individuals with a homozygous mutation , as parents are obligate heterogygotes . Thirteen of 20 such parents had no cZLR peak , equivalent to a specificity of 65% . Five parents had one cZLR peak , one parent had 2 peaks , and one father had 7 peaks , being most likely from unknown consanguineous background himself ( Figure 3 ) . This specificity of 65% should not pose a problem for gene identification , since the likelihood that a false posititive cZLR peak of a few Mb size will colocalize by chance for two individuals , thereby pointing to a false candidate locus , is in the order of only 1∶1 , 000 . In order to assess how likely it is to find a disease-causing mutation in outbred single affecteds with mutations in a recessive disease gene using the approach described here , we prospectively performed 250 k SNP DNA microarray analysis in 24 unrelated outbred patients with infantile nephrotic syndrome , which is known to be caused by NPHS1 mutations in 22 . 5% of cases ( 18/80 ) [3] and by PLCE1 mutations in 28% of cases ( 10/35 ) [24] . We obtained the following distribution for the number of ZLR peaks of homozygosity per patient: 0 , 1 , 2 , 4 , and 6 peaks in 9 , 4 , 8 , 1 , and 2 patients , respectively . In all 24 patients mutation analysis was performed by exon PCR of all NPHS1 and PLCE1 exons . We identified both causative mutations in 21% ( 5/24 ) of these patients ( 4 patients had NPHS1 mutations and 1 patient had a PLCE1 mutation ) . In the 15 of 24 outbred patients that had ZLR peaks of homozygosity , we identified homozygous mutations in 27% ( 4/15 ) of these patients ( 3 patients with a homozygous NPHS1 mutation and 1 patient with a homozygous PLCE1 mutation ) . As expected , these 4 patients with homozygous mutations had a ZLR peak of homozygosity at the locus of their mutation . However 1 additional patient had a compound heterozygous mutation of NPHS1 although we observed 2 ZLR peaks of homozygosity at different loci . This patient represents an example of a “false positive result” in the homozygosity mapping approach . We conclude that the fraction of homozygous mutations detected in NPHS1 and PLCE1 in patients with infantile nephrotic syndrome ( 27% ) using the homozygosity mapping approach in outbred individuals prospectively was similar to the one reported by diagnostic sequencing ( 22 . 5% for NPHS1 and 28% for PLCE1 ) . We conclude that homozygosity mapping of recessive disease genes is possible in single individuals from outbred populations . The current resolution is at ∼2 . 1 Mb , containing an average of only 16 candidate genes .
These data are , to our knowledge , the first to quantify for a large cohort of patients with known homozygous mutations in a high number of different recessive disease-causing genes the extent of homozygosity by descent and its detection limit in outbred populations . Our findings have the potential to strongly facilitate the identification of recessive disease-causing genes by overcoming some of its major impediments: i ) First , large cohorts of consanguineous pedigrees are very difficult to ascertain , whereas most pediatric specialty clinics have direct access to sufficient numbers of single individuals of rare pediatric diseases . ii ) Second , homozygosity mapping in consanguineous kindred leads to large candidate regions that usually contain too many positional candidate genes to allow for gene identification by mutation analysis . Here we demonstrate that homozygosity mapping can be performed in single affecteds from outbred populations yielding a single cZLR peak per genome . We were able to map the disease-causing gene to single peaks representing segments of homozygosity that were as short as 2 . 1 Mb , which would contain 16 genes on average . For positional cloning of novel recessive disease genes this strategy offers a reduction of complexity of three orders of magnitude , i . e . from 3 , 300 Mb to ∼3 . 3 Mb or from 25 , 000 genes to ∼25 genes . This reduction makes gene discovery by mutation analysis feasible , either by exon sequencing or large-scale sequencing using emerging techniques . Additionally , in regions of homozygosity affected individuals from different families that originate from the same geographic region may , within the homozygous haplotype , be identical by descent from an unknown common ancestor . Candidate gene analysis can then be focused on this short interval under the hypothesis that both families might share the same disease allele , as we show for the NPHS2 mutation R138Q ( Figure S2 ) and for an NPHP5 founder mutation ( Figure S3 ) . This approach has also been useful in the identification of the NPHP6/CEP290 gene [18] . The strategy described here is especially useful in situations of pronounced genetic locus heterogeneity , in which similar autosomal recessive phenotypes may be caused by a high number of different genes . In this situation a first run of homozygosity mapping will show whether the individual most likely carries a homozygous disease mutation by presence of one or more cZLR peaks . Mutation analysis by exon sequencing can then be restricted to genes that are positioned in a cZLR peak of the individual . An example for NPHP is given in Figure 2B . Even if in the near future rapid and cost-effective sequencing of a person's genome will be possible , which has been referred to as the “$1 , 000 genome” , homozygosity mapping in outbred populations will be a valuable first step of molecular genetic diagnostics in recessive diseases for the following reason: It generates the hypothesis that a disease-causing homozygous mutation will be found in a homozygosity peak and nowhere else , thereby reducing the high number of difficult-to-interpret sequence changes found outside cZLR peaks . Our data show that the approach of homozygosity mapping in outbred individuals at a marker density of 250 K has not reached its theoretical limit of refinement yet , as we found some homozygous disease alleles embedded in haplotypes as short as 0 . 65 Mb equivalent to only 48 SNP of the 250 K array ( Table 1 , Figures S2 and S3 ) . The following measures may increase the likelihood of detecting even shorter unique homozygous segments that contain the disease-causing mutation in single individuals from outbred populations: i ) When increasing the density of markers evaluated in multipoint calculations to more than 1 marker per 100 , 000 nucleotides , additional homozygosity peaks appear as “background” . This background is most likely due to non-informative alleles ( identically homozygous in both parents ) . Linkage calculation together with parental genotypes will allow exclusion of these markers , thereby reducing the number of false positive cZLR peaks . This will permit increasing the number of markers evaluated , allowing detection of even smaller homozygous loci . iii ) SNP arrays with higher density are now available ( e . g . 1 Mill . SNPs ) . They contain HapMap-derived SNPs that are potentially more informative . However , as we observed a relatively high rate ( <4% ) of false heterozygous allele calls in the 250 K SNP arrays ( see Methods ) , the utility of higher density SNPs for homozygosity mapping in outbred individuals will need to be tested first . The individuals tested here , in whom we showed that homozygous mutations are embedded in a short homozygous haplotype by descent , provide a perfect resource for testing the limitations of higher resolution SNP arrays for homozygosity mapping in outbred individuals . As inclusion of parents of an affected individual in the total genome haplotype analysis establishes the origin of the paternal and maternal haplotypes , a data base of disease-allele specific haplotypes would permit even detection of known mutations by their haplotype in individuals with compound heterozygous mutations in the future . The new strategy of homozygosity mapping in single outbred individuals described here has a strong potential of accelerating the discovery of novel recessive disease genes as a critical step towards elucidating the pathogenesis of a wide variety of pediatric disorders independent of organ system involved . | Many childhood diseases are caused by single-gene mutations of recessive genes , in which a child has inherited one mutated gene copy from each parent causing disease in the child , but not in the parents who are healthy heterozygous carriers . As the two mutations represent the disease cause , gene mapping helped understand disease mechanisms . “Homozygosity mapping” has been particularly useful . It assumes that the parents are related and that a disease-causing mutation together with a chromosomal segment of identical markers ( i . e . , homozygous markers ) is transmitted to the affected child through the paternal and the maternal line from an ancestor common to both parents . Homozygosity mapping seeks out those homozygous regions to map the disease-causing gene . Homozygosity mapping requires families , in which the parents are knowingly related , and have multiple affected children . To overcome these limitations , we applied homozygosity mapping to single affected individuals from outbred populations . In 72 individuals with known homozygous mutations in 13 different recessive disease genes , we performed homozygosity mapping . In 93% we detected the causative gene in a segment of homozygosity . We demonstrate that disease-causing homozygous mutations can be detected in single cases from outbred populations . This will strongly accelerate the discovery of novel recessive disease genes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"nephrology/tubulointerstitial",
"diseases",
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"nephrology/chronic",
"kidney",
"disease",
"nephrology/hereditary,",
"genetic,",
"and",
"development",
"nephrology",
"genetics"... | 2009 | A Systematic Approach to Mapping Recessive Disease Genes in Individuals from Outbred Populations |
Colonies of the opportunistic pathogen Proteus mirabilis can distinguish self from non-self: in swarming colonies of two different strains , one strain excludes the other from the expanding colony edge . Predominant models characterize bacterial kin discrimination as immediate antagonism towards non-kin cells , typically through delivery of toxin effector molecules from one cell into its neighbor . Upon effector delivery , receiving cells must either neutralize it by presenting a cognate anti-toxin as would a clonal sibling , or suffer cell death or irreversible growth inhibition as would a non-kin cell . Here we expand this paradigm to explain the non-lethal Ids self-recognition system , which stops access to a social behavior in P . mirabilis by selectively and transiently inducing non-self cells into a growth-arrested lifestyle incompatible with cooperative swarming . This state is characterized by reduced expression of genes associated with protein synthesis , virulence , and motility , and also causes non-self cells to tolerate previously lethal concentrations of antibiotics . We show that temporary activation of the stringent response is necessary for entry into this state , ultimately resulting in the iterative exclusion of non-self cells as a swarm colony migrates outwards . These data clarify the intricate connection between non-lethal recognition and the lifecycle of P . mirabilis swarm colonies .
Organisms rarely live in complete isolation . Living in a community can provide benefits to each individual . However , there is a constant balance between the interests of individuals and the maintenance of community-wide advantages . A stable evolutionary strategy is for individuals to preferentially direct advantages to close kin [1–3] . This behavior , known as kin discrimination , has been the subject of focused study . Several examples of kin discrimination in bacteria have been elegantly described , including those mediated by Type IV [4] , Type VI [5 , 6] , and Type VII [7] secretion system based effector exchange , contact-dependent inhibition ( CDI ) [8 , 9] , and the MafB toxins of the Neisseria [10] . One common thread between these systems is that they characterize discrimination as immediate antagonism towards cells or strains that are non-kin , typically through delivery of lethal toxin effector molecules . Upon effector delivery , receiving cells must either neutralize it by presenting a cognate anti-toxin or suffer immediate negative consequences , typically cell death [5] or permanent inhibition of growth [8] . Here we describe an expansion of these mechanisms: the Ids self-recognition system mediates kin discrimination in Proteus mirabilis by selectively inducing non-self cells into a growth-arrested lifestyle incompatible with social behavior , thereby controlling access to that behavior . P . mirabilis , a major cause of recurrent complicated urinary tract infections [11] , engages in several sophisticated social behaviors such as swarming on rigid surfaces . Swarms are formed by many elongated ( ~ 10–80 μm ) “swarmer” cells moving cooperatively , allowing for colony expansion over centimeter-scale distances . Rounds of swarming are interspersed with periods of non-expansion termed “consolidation” . The oscillation between swarming and consolidation leads to a characteristic pattern of concentric rings on higher percentage agar plates [12] . Effective P . mirabilis swarming relies on the ability of swarmer cells to form large rafts that together move much more quickly than isolated individuals [13] . Rafts are fluid , transient collectives that cells frequently enter and exit . As such , an individual cell interacts with many different neighbors through the lifetime of a swarm . During swarming , P . mirabilis cells can communicate with each other by exchanging proteins through contact-dependent secretion systems [14 , 15] . These signals in turn cause emergent changes in swarm behavior [16 , 17] . P . mirabilis swarms exhibit the ability to recognize self in several ways . The oldest known example is Dienes line formation: two swarms of the same strain merge into a single swarm upon meeting , while two swarms of different strains instead form a human-visible boundary [18 , 19] . More recently , the phenomenon of territorial exclusion was described: in a mixed swarm comprising two different strains , one strain is prevented from swarming outwards by the other [15] . Clonal swarms of P . mirabilis have a coherent self identity , minimally mediated by the Ids system encoded by six genes idsA-F . Deletion of the ids locus results in the mutant strain no longer recognizing its wild-type parent as self [20] . Several of the molecular mechanisms governing Ids-mediated self recognition have been described in detail . However , how the Ids system functions in local behaviors has remained elusive . Briefly , two proteins , IdsD and IdsE , govern self identity . IdsD is transferred between cells in a Type VI secretion system ( T6SS ) -dependent fashion; disruption of the T6SS prevents all Ids signal transfer [17 , 21] . A cell in a swarm is considered to be self if it produces an IdsE protein that can bind IdsD proteins sent from neighboring cells . Disruption of these IdsD-IdsE interactions , either through deletion of idsE or through production of non-self IdsD or IdsE variants , result in strains that display extreme attenuation in swarm expansion without loss of viability [16 , 17] . Fig 1A shows cartoon representations of Ids-mediated self recognition , where endogenous IdsD and IdsE are represented by capital letters “D” and “E” , and a non-self variant of IdsD is represented by lowercase “d . ” The IdsD protein must be incoming: endogenous IdsD and IdsE proteins produced within a single cell do not impact self-recognition behaviors [21] . We describe conditions that lead to non-self recognition as “Ids mismatch . ” Crucially , territorial exclusion by Ids does not affect viability . Excluded cells grow and divide at a comparable rate to non-excluded cells when isolated from swarms [17] . While Ids has been described as a toxin-antitoxin system [22 , 23] , this characterization is inconsistent with experimental data and is likely due to the reliance on the T6SS for transport of IdsD [17 , 21] . The T6SS is often characterized as a lethal toxin delivery mechanism [24] . A mechanistic description of Ids-mediated recognition is needed to reconcile the data and would provide a model for other non-lethal mechanisms that might be attuned for surface-dwelling swarm migration . Here we show that even though an Ids recognition signal transfer happens while cells are actively migrating as a swarm , the recognition response is delayed until cells have stopped moving . We show that recognition of non-self is at least partially mediated by ppGpp levels within the cell , and this contributes to a concerted shift of cells into a distinctive , antibiotic-tolerant state that is incompatible with continuation of swarming . We found that this Ids-induced state is short-acting; induction requires continuous cell-cell interactions . In the context of a swarm , the collective consequence is an iterative winnowing of the non-self cells from the swarm fronts during periods of no active migration . We posit that the cell-cell communication of these non-lethal factors therefore acts as a control system during swarm expansion by diverting non-self from developing into swarm-compatible cells and thus preventing non-self cells from taking part in cooperative swarm behavior .
We initially sought to determine the method through which Ids caused non-self cells to be territorially excluded from swarms . Given the lack of lethality and the stark attenuation of swarm colony expansion observed during Ids-mediated territorial exclusion [15 , 17] , we hypothesized that an Ids mismatch caused broad changes in gene expression of the recipient cell . Ids mismatch is defined here as transcellular communication of IdsD to a recipient cell lacking a cognate IdsE . Therefore , we performed RNA-Seq differential expression analysis using conditions that would produce either self or non-self interactions , all within genetically equivalent backgrounds . We isolated total RNA from cells undergoing consolidation , because consolidation is the swarm development stage most tightly connected to major transcriptional changes [25] . As a baseline , we compared transcriptional profiles between cells from clonal swarms of wildtype and independently , of a derived mutant strain lacking the ids genes ( BB2000::idsΩCm [20] , herein referred to as “Δids” ) . The swarm colonies of both strains expand equivalently and have no notable morphological differences [20] . A complete list of genes with differential transcript abundance is found in S1 Table . Overall , few differences were apparent between clonal swarms of wildtype and the Δids strain . 10 genes were found to be significant ( fold-change > log2 1 . 5 , p < 0 . 05 ) , six of which were the ids genes deleted in the construction of the Δids strain ( Fig 1B ) . We next considered differences in strains experiencing Ids mismatch . We examined three different conditions: a clonal swarm in which every cell lacked the idsE gene ( CCS06 ) , a clonal swarm in which every cell is a chimera containing an IdsE protein unable to bind transferred IdsD proteins ( CCS02 ) , and cells of a Δids-derived strain constitutively producing Green Fluorescent Protein GFPmut2 ( Δids-GFP ) that were isolated from a 1:1 co-swarm with wildtype through fluorescent-activated cell sorting ( FACS ) , henceforth termed “co-swarmed Δids" . Strains CCS06 and CCS02 are the Δids background containing a plasmid expressing the ids operon under its native promoter; mutations to the ids genes were made in the plasmid-based allele . All strains have previously been verified and characterized [16 , 17 , 20 , 21] . As the two clonal swarm colonies have attenuated expansion , we were only able to harvest whole colonies as visible consolidation phases were less distinct . RNA-Seq differential expression analysis was performed on cells from each condition as compared to appropriate control samples: co-swarmed Δids were compared with clonal Δids , and CCS02 and CCS06 swarms were compared to a clonal Δids-derived swarm in which cells expressed plasmid-encoded ids genes ( CCS01 , Δids pidsBB ) . Large-scale changes to relative transcript abundances were apparent for each condition: 231 genes in CCS06 , 457 genes in CCS02 , and 836 genes in co-swarmed Δids , which represents approximately 6% , 13% , and 23% of total genes , respectively ( Fig 1 , S2 Table , S3 Table , S4 Table ) . General trends were apparent even though differences in relative abundance of transcripts were present in a diverse and widespread range of genes . We observed a concerted decrease in transcripts for class I , II , and III genes for flagellar synthesis such as flhDC , filA , and fliC . We also observed a decrease in transcripts for many genes associated with protein synthesis , such as the 50S ribosomal protein rplT and 30S ribosomal protein rpsP , along with ribosomal-associated elongation factors such as EF-Tu . Several genes involved in respiration , including those of the FoF1 ATP synthase , also had significantly fewer transcripts as did those transcripts for different virulence-associated protein families [26–31] , such as hpmA , umoA , and zapD . Overall , fewer genes had an increased relative abundance as compared to the control strains: 109 genes in CCS06 , 56 genes in CCS02 , and 332 genes in co-swarmed Δids , respectively . Within these genes , several endogenous toxin-antitoxin systems displayed an increased relative abundances of transcripts , including genes that encode homologous proteins to Escherichia coli YfiA ( also known as RaiA ) [32] and to toxin/antitoxin pairs ParE/CC2985 [33] and Phd/Doc [34] . There was also an increased relative abundance for several fimbriae families , including genes encoding MR/P and P-like fimbria . A large proportion of differentially regulated genes were proteins of unknown function . A subset of differentially abundant genes was shared among all three datasets; these were representative of families found in each Ids mismatch strain . Three genes had increased relative abundance in transcripts as compared to wildtype; 32 genes had decreased ( Fig 1D ) . The 32 genes with decreased relative transcripts included those involved in motility , chemotaxis , ribosomal proteins , and metabolism ( S5 Table ) . Of the three genes with increased relative transcripts , two encode the Phd/Doc endogenous toxin-antitoxin system . The third gene is rob , which has been associated with the induction of low-metabolism states in E . coli [35] . Thus , cells under the influence of incoming IdsD , and without a cognate IdsE protein present , enter a distinctive transcriptional state from either wild-type or Δids cells participating in a normal swarm cycle . Transcriptional shifts of these types are often associated with entry into altered states induced by a variety of environmental and temporal cues , including nutrient and membrane stress [36] . We hypothesized that the changes in Ids-excluded cells might also result in the secondary effect of increased antibiotic tolerance . We used susceptibility to antibiotics as a proxy for whether cells have entered an altered state . We conducted antibiotic tolerance assays on wildtype , the Δids strain , and a third strain containing an unmarked in-frame ( nonpolar ) deletion of the chromosome-encoded idsE within the wildtype background [21] . The Δids strain encodes a transgenic resistance gene to chloramphenicol; the deletion is stable in growth media without chloramphenicol . Neither the wildtype nor the ΔidsE strain encode transgenic genes for antibiotics resistance . All three strains are endogenously resistant to tetracycline . We predicted that swarms of the ΔidsE strain would display increased antibiotic tolerance compared to wildtype and the Δids strain . We first tested the beta-lactam antibiotic ampicillin to which the wild-type BB2000 is susceptible . We used independent , clonal swarms of either wildtype , the Δids strain , or the ΔidsE strain , each grown in the absence of antibiotics . We harvested cells after the entry to the third swarm ring . Cells were resuspended in LB media and immediately subjected to 100 μg ml-1 ampicillin exposure . Samples were extracted for viability assays on fresh media lacking antibiotics at regular intervals until eight hours and again at 16 hours . No clear difference was observed between wildtype and the Δids strain at any timepoint ( Fig 2A ) . Wildtype and the Δids strain exhibited a killing of approximately 105-fold after eight hours , while the ΔidsE strain experienced a killing of approximately 104-fold ( Fig 2A ) . Under these conditions , the ΔidsE strains had an approximately 50-fold increase in survival as compared to wildtype , even after sixteen hours incubation in ampicillin ( Fig 2A ) . We repeated this assay with three additional antibiotics to explore whether the ΔidsE cells exhibited a broad tolerance to antibiotics . We utilized the fluoroquinolone ciprofloxacin and the aminoglycosides streptomycin and kanamycin . No difference between wildtype , the Δids strain , and the ΔidsE strain was observed in viable cell counts following 50 μg ml-1 streptomycin or 1 μg ml-1 ciprofloxacin exposure after 12 hours ( Fig 2B ) . These antibiotics resulted in lower rates of cell killing as compared to ampicillin , likely reflecting a higher native resistance of P . mirabilis to these drugs [37 , 38] . However , the ΔidsE strain showed an increased number of viable cells as compared to wildtype or the Δids strain after 60 μg ml-1 kanamycin incubation for 12 hours ( Fig 2B ) . Therefore , the ΔidsE strain displays increased tolerance to both kanamycin and ampicillin under these conditions , indicating limited resistance to antibiotics . Since Ids functions through cell-cell contact-dependent secretion of the identity marker IdsD [15 , 17] , we next tested whether IdsD secretion was required for antibiotic tolerance to emerge . Cells secrete IdsD into growth media [15] . We performed this assay using MJT01 , which is an ΔidsE-derived strain containing a non-functional T6SS [17 , 39] and so does not secrete IdsD . The deletion of idsE and the disruption of the T6SS-encoding gene tssB/vipA are unmarked and nonpolar on the chromosome . We observed no clear difference between MJT01 and wildtype over three biological repeats ( S1 Fig ) . To examine whether social exchange was causative for antibiotics tolerance , we tested cells of the ΔidsE strain that were grown to stationary phase with shaking in liquid . IdsD transfer between cells is limited , if at all , during liquid growth . We observed no difference in antibiotics tolerance for the ΔidsE strains as compared to wildtype over three biological repeats ( S2 Fig ) . Therefore , antibiotics tolerance was induced by an Ids mismatch and caused by the transfer of IdsD between neighboring cells without a cognate IdsE present . We considered whether this Ids-mediated antibiotic tolerance might be due to entry into an irreversible state . To examine the dynamics of how cells exit from an Ids mismatch , we assayed co-swarms in which the GFP-producing Δids strain ( Δids-GFP ) was inoculated with an equal amount of a wild-type strain constitutively producing DsRed ( wildtype-DsRed ) . We let the swarm progress to the third swarm ring and then harvested the swarms . Cells of each strain were immediately sorted with FACS . Analysis of sorted cells showed that Δids-GFP formed ~10% of the sample from the third swarm ring . Equal numbers of particles of each strain were inoculated in LB media without antibiotics , and growth at 37°C was measured for 24 hours through optical density at 600 nm . No differences between co-swarmed wildtype-DsRed and co-swarmed Δids-GFP were observed at any time-point ( Fig 2C ) . We concluded that the effects induced by an Ids mismatch are transient outside of continual contact-mediated pressure . Consistent with this assertion , we found no differences between the growth of CCS02 ( the chimera Ids mismatch strain ) and strain CCS01 grown in liquid LB media at 37°C ( S3 Fig ) . IdsD transfer does not occur , or is limited , during liquid growth . Therefore , an individual cell within a swarm is shifted into a distinct transcriptional state when it has received non-self IdsD from the surrounding cells . We found that this cell state caused by an Ids mismatch is temporary and reversible . Entry into an antibiotic-tolerant state has been linked in other bacteria to the stringent response [36 , 40] , mediated by the alarmone messenger molecule ( p ) ppGpp . Although the stringent response has not been studied in P . mirabilis , the genome for wild-type BB2000 contains two canonical genes for production and degradation of ppGpp , relA and spoT [41] . We tested whether Ids mismatch was connected to the stringent response . Using quantitative high performance liquid chromatography ( HPLC ) [42] , we directly measured total ppGpp quantities in cells independently harvested from swarms of clonal wildtype , clonal Δids , or clonal ΔidsE . Nucleotide samples were purified , separated by HPLC , and quantified by measuring UV absorbance spectra using established methods [42] . We performed three biological repeats of ppGpp measurements and found that the samples from the ΔidsE strain contained nearly twice the ppGpp levels as wildtype and the Δids strain ( Fig 3A , S4 Fig ) . These results indicate that ppGpp levels and Ids mismatch are linked . We next examined whether ppGpp is necessary for Ids mismatch , specifically focusing on the ΔidsE swarm deficiency . Deletions of relA or spoT can prevent ppGpp accumulation and as such , prevent cells from activating the stringent response in several bacteria as discussed in [43] . We constructed three ΔidsE-derived strains , each with an independent , unmarked , nonpolar chromosomal deletion of relA , spoT , or both . We quantified ppGpp levels in each strain using HPLC as described above and found that ppGpp levels in each strain were minimal as compared to wildtype ( S5A Fig ) . We next assayed each strain for swarm colony expansion as compared to that of wildtype and the parent ΔidsE strain . We found that each newly constructed strain formed swarms of a diameter equivalent to wildtype and nearly twice that of the parent ΔidsE strain ( Fig 3C ) . We also found that in 1:1 co-swarms with the wild-type strain , the ppGpp-deficient strains were excluded from the swarm colony edges , similar to the ΔidsE strain ( S5B Fig ) . We interpret these results as interactions with the wild-type strain are not equivalent to interactions with clonal cells lacking ppGpp . Therefore , while ppGpp is required for Ids mismatch to have an effect in clonal swarms , it is likely not the only factor . Moreover , we posit that ppGpp might be a causative factor upstream of the observed transcriptional and physiological changes . Having observed that a response to Ids mismatch is only present under consistent pressure from neighboring non-self cells ( Fig 2 ) , we reasoned then that the effects of Ids mismatch control might be spatially and/or temporally attuned . The small molecule ppGpp might allow for such a rapid and transient response in cells . To interrogate this model , we took advantage of the genes newly identified as being induced in the presence of non-self ( Fig 1 ) to develop a fluorescent transcriptional reporter system . A gene encoding a variant of the fluorescent protein Venus [44] was engineered to be inserted immediately downstream of the gene BB2000_0531 , resulting in Venus production being controlled by the upstream promoter . The BB2000_0531 gene , encoding a putative sigma-54 modulation protein , was chosen as it displayed increased expression under different Ids mismatch conditions ( Fig 1B , S2 Table , S3 Table , S4 Table ) . The reporter construct was inserted unmarked and nonpolar in the chromosome of the Δids strain , resulting in strain MJT02; this strain had no apparent growth or swarm defects . We performed fluorescence microscopy time-course experiments on mixed swarms to measure transcriptional changes associated with BB2000_0531 over the course of a swarm-consolidation cycle . Two co-swarm conditions were used . In the first , a mixed culture of 50% MJT02 and 50% the Δids strain was used to inoculate swarm-permissive agar; in the second , a mixed culture of 50% MJT02 and 50% wildtype-DsRed was used . Swarms were grown at 37°C until the first swarm expansion was visible . Venus fluorescence intensity was measured at 30-minute intervals thereafter in swarm areas , and the mean fluorescence was calculated . The fluorescence intensity for both co-swarm conditions was graphed ( Fig 4A ) ; representative images are in Fig 4B . We observed a temporal spike in fluorescence associated with BB2000_0531 correlated with the consolidation cycle; this increase was only apparent when Δids-derived cells were intermingled with wild-type cells ( Fig 4 ) . Therefore , the gene expression response to Ids mismatch occurs during consolidation . In addition to causing territorial exclusion in mixed swarms , Ids mismatch determines boundary formation after collision between two clonal swarms [20] . Boundary formation following swarm contact is a complex process likely involving the contribution of several lethal and non-lethal systems [14 , 15 , 20 , 45] . To test whether equivalent transcriptional shifts were observed during the initial stages of boundary formation , when cells of each strain are first in contact with one another , we measured fluorescence intensity associated with BB2000_0531 in MJT02 swarms following collision with wildtype and Δids swarms . We observed a mean increase in fluorescence intensity over several hours after MJT02 encountered wildtype , but not in encounters with the Δids strain ( S6 Fig ) . The observed increase in fluorescence intensity occurred before a boundary was visually apparent . In fact , formation of a visible boundary between the two strains did not occur for a further 12–18 hours after the end of this experiment , which is consistent with previous observations [18] . Therefore , the Ids mismatch induces a response in the initial stages of self versus non-self recognition , after non-self cells have interacted . We reasoned that Ids mismatch control might be relevant for the formation and/or development of a swarm colony , which is consistent with our prior hypothesis that Ids impacts cooperative behavior [17] . Swarming is fundamentally a collective behavior . The spatial expansion of a wild-type swarm is connected to the oscillatory developmental cycle of outward migration and non-motile consolidation . To assess the hypothesis that the Ids system likely impacts local cell-cell interactions at the boundary and within an expanding swarm , we examined territorial exclusion in situ using epifluorescence microscopy and utilized co-swarms constructed with equal ratios of the Δids and wild-type strains . To allow visualisation of individual cells , 10% of Δids and wildtype cells in the starting mixture constitutively expressed GFP and DsRed fluorescent proteins , specifically strains Δids-GFP and wildtype-DsRed , respectively ( Fig 5 ) . The control experiment consisted of a co-swarm in which a starting inoculum consisted of wildtype doped with 10% wildtype constitutively producing GFPmut2 ( strain wildtype-GFP ) and 10% wildtype-DsRed ( Fig 5B ) . Once swarmer cells emerged from the inoculum , the proportion of cells expressing each fluorophore was measured at half-hour intervals ( Fig 5 ) . The developmental stages of active outward motility versus no outward motility ( i . e . , consolidation ) were noted by eye ( S7 Fig ) . We calculated the fluorophore ratios over time for both the Δids-GFP:wildtype-DsRed and wildtype-GFP:wildtype-DsRed co-swarms for three biological repeats . The GFP/DsRed ratios in the wildtype-GFP:wildtype-DsRed control experiment did not deviate over time , with approximately equal numbers of each strain observable in the swarm over the course of eight hours ( Fig 5B ) . The Δids-GFP:wildtype-DsRed co-swarm did not show measurable changes until well after swarm emergence ( Fig 5A ) , indicating that Δids-GFP cells arising from the inoculum were not excluded from swarm behavior . However , large decreases in the GFP/DsRed ratio were observed in consolidation periods between rounds of swarming , starting in the first consolidation phase ( Fig 5A ) . Later swarm rings often contained no observable Δids-GFP cells . To test whether the ratio changes observed over multiple swarm cycles were caused by lysis of the cells experiencing Ids mismatch , we monitored a single field of view of Δids-GFP:wildtype-DsRed or Δids-GFP:Δids co-swarms over one consolidation phase . Images were taken at 5-minute intervals from the start to end of the consolidation phase , with three biological repeats performed ( S8 Fig ) . Cell lysis rates were less than 0 . 1% for both conditions , with no difference in cell lysis rates apparent at any point . We did not observe an absolute decrease in Δids-GFP cell numbers ( S8A Fig ) . We also found that as the consolidation zone increased , the ratio of fluorescence associated with Δids-GFP decreased in co-swarms with wildtype and remained constant in co-swarms with Δids ( S8B Fig ) . Observation of individual cells in consolidation areas suggests that the Δids-GFP strain has apparent differences in cell division dynamics based on whether in a co-swarm with wildtype ( non-self ) or with the Δids strain ( no signal ) . Altogether , we found that Ids-mediated exclusion was increasingly effective over the course of the co-swarm with initial equal ratios of cells , eventually resulting in Δids-GFP cells being excluded from the leading edges of swarming colonies ( Fig 5 ) . Therefore , Ids mismatch results in cells unable to proceed through the swarm developmental cycle during the consolidation period . As Ids-mediated exclusion was correlated with consolidation in P . mirabilis swarms , we generated “hyperswarming” wild-type and Δids strains ( named “wildtype-HS” and “Δids-HS , ” respectively ) that continually swarm outwards without consolidation [28] , which leads to rapid surface coverage through swarm colony expansion . We performed co-swarms to test whether hyperswarming protected the Δids-derived strains from exclusion by wildtype . We observed that neither Δids-HS:wildtype-HS nor Δids-HS:wildtype co-swarms resulted in the exclusion of the hyperswarming Δids strain ( S9 Fig ) . However , Δids:wildtype-HS co-swarms , in which the Δids strain enters consolidation , did result in territorial exclusion of Δids-derived cells ( S9 Fig ) . We concluded that outside of the consolidation phase , the Δids-derived cells that received non-self signals were not effectively excluded , indicating that Ids mismatch does not affect swarm performance in hyperswarming cells . We propose that Ids mismatch induces the recipient cell to experience a growth arrest in the swarmer cell developmental cycle , which then prevents cells from re-entry into swarm-compatible states . Here we expand on models of kin discrimination [5 , 8] by showing that the Ids system encompasses a complex and subtle recognition that is attuned to the challenges of rapid migration as a collective along a hard surface . Ids-mediated recognition controls the spatial location of non-self cells over the lifetime of a swarm . It appears that for this robustly swarming bacterium , access to a social behavior is impeded via a non-lethal mechanism: the Ids self-recognition system selectively induces non-self cells into a growth-arrested lifestyle incompatible with cooperative swarming . Intriguingly , Ids-like proteins are encoded within the genomes of other members of the Morganellaceae family , suggesting that this mechanism might be more broadly found . Further , these data suggest a model for Ids territorial exclusion in mixed swarms ( Fig 6 ) . IdsD is likely primarily transferred during active swarming when the secretion machinery is produced and abundantly visible [21 , 25 , 39] . During consolidation phase , the presence of IdsD with an absence of a cognate IdsE ( resulting in unbound IdsD in recipient cells ) causes a shift into a distinctive transcriptional state ( Fig 1 ) that is partially due to activation of the stringent response via elevated ppGpp levels ( Fig 3 ) . This shifted state also causes a phenotype in affected cells that allows increased antibiotic tolerance ( Fig 2 ) . We propose that Ids mismatch functions by diverting cells , via growth arrest , from re-entry into P . mirabilis’ swarm-consolidation developmental cycle , which results in individual non-self cells being iteratively winnowed out of the migrating swarm front when initially present in equal ratios ( Fig 6 ) . Several potential models could explain exactly how unbound IdsD affects the recipient cell . One model is that the presence of unbound IdsD in a recipient cell interrupts a checkpoint in the differentiation from a swarmer to a consolidated cell or vice-versa . While the transcriptomic data provides a reasonable starting point , the list of differences for each Ids mismatch condition as compared to wildtype is quite large . These changes do not resemble those previously described during swarm-consolidation transitions [25] , suggesting that Ids mismatch induces entry into a novel expression state . It is also possible that IdsD might accumulate in the membrane over time , leading to a general stress response . However , several pieces of data contradict such a model . First , non-self cells are able to escape Ids-mediated territorial exclusion under laboratory conditions by overexpression of the master flagellar regulator flhDC , which abolishes consolidation to form hyperswarmer cells [31] . Hyperswarming Δids cells receiving a non-self signal are motile and able to swarm with wildtype ( S9 Fig ) . Excluded Δids cells are still able to grow and divide in situ [17] . Further , deletion of relA and/or spoT , which reduces ppGpp levels , allows for increased swarm colony expansion of the clonal Ids mismatch strain , ΔidsE , thereby bypassing Ids mismatch control ( Fig 3 and S5A Fig ) . The molecular mechanisms for how ppGpp levels might cause attenuated expansion remain to be uncovered . Although recognition signals need flagellar regulation and internal ppGpp levels to be effective , how these pathways intersect remains to be uncovered . Interpretation of our research is further complicated , because little is published about the stringent response and/or ppGpp activity in P . mirabilis . There are several transcriptional changes in the Ids mismatch-induced state that are not readily explained by the ppGpp response . We anticipate that ongoing studies into the emergence of bacterial dormancy and related phenotypes [46–49] in other species might help to untangle the order and hierarchy of the Ids-induced changes described here . Several candidate pathways for further analysis are apparent from the transcriptomics datasets . The role of the signaling molecule c-di-GMP in regulating motile/sessile lifestyle changes in many bacteria is well-studied and an attractive target for future work [50] . The SOS response , mediated by RecA , has been implicated in persister formation in E . coli K12 [51] and may also play a role here . Ids gene regulation in general has been linked with the MrpJ transcriptional network important for P . mirabilis virulence [27] . The observation of increased MR/P fimbrial expression ( Fig 1 ) suggests a potential link between Ids-induced changes , MrpJ , and changes in virulence . More generally , we have presented evidence of a peer pressure system for recognition that iteratively winnows non-self cells from participating in the collective behavior of swarming—a social activity between cells that is observed among many bacterial species . Ids-mediated macroscale territorial behavior emerges from the sum of cell-cell contacts within a swarm [15 , 17 , 20 , 52] . In this Ids model , cells do not receive any information about population composition and behavior other than that from their immediate neighbors , which is different when compared to other examples of bacterial collective behavior . For example , in bacterial quorum sensing , secretion of diffusible small molecules into the environment provides a global tracker accessible to all individuals in a group [53 , 54] . Therefore , each individual cell has potentially equal access to the external signal , because the signal molecule can freely diffuse between/among cells . Each P . mirabilis cell , however , has access only to the signal of a physically adjacent cell . As such , in Ids mismatch-mediated exclusion , any information about the swarming population as a whole is decentralized and distributed among every member of the swarm . Access to that information is restricted to clusters of adjacent , neighboring cells . In these respects , Ids-mediated control represents an orthogonal model for collective behavior in bacteria that provides new opportunities to explore cell-cell communication , especially as regards to spatial coordination . Any theoretical model of Ids-mediated behavior will likely need to differ from those describing quorum sensing of diffusible molecules . The Ids self-recognition system has distinct qualities from other contact-associated systems that have been described as bacterial communication . CDI systems [55] , for example , have been described as lethal [56] or inducing permanent persister-like states [57] . However , broad spatial analysis has yet to be more generally pursued . P . mirabilis swarms could provide an excellent framework for directly analyzing individual cells before , during , and after Ids communication and for examining the global spatial consequences to these local interactions . Tracking individual cell fates through swarming and consolidation will help in this regard . Moreover , the ability of the Ids system to temporally and spatially control non-self cells by altering cell state raises the question of which other mechanisms for contact-mediated signaling in bacteria enable sophisticated interactions between individuals . Finally , we find it unlikely that the Ids system is a specific adaptation to mitigate antibiotic pressure . This places it in contrast with Ghosh et al [57] who recently modelled CDI-mediated persistence in E . coli as a bet-hedging mechanism . We speculate instead that the antibiotic tolerance observed in this study represents an evolutionary "spandrel" of the type described by Gould and Lewontin [58 , 59] . Ids-induced antibiotic tolerance in this case would be a by-product of its primary adaptive feature , regulating clonal swarm composition . Under this view , the antibiotic tolerance data shown in this paper should be regarded as a method of measuring Ids-induced phenotypic shifts in the context of territorial exclusion .
All strains used in this study are described in Table 1 . P . mirabilis strains were maintained on LSW- agar [60] . CM55 blood agar base agar ( Oxoid , Basingstoke UK ) was used as a swarm-permissive agar . E . coli strains were maintained on Lennox lysogeny broth ( LB ) agar . All liquid cultures were grown in LB broth at 37°C with shaking . Swarm plates were grown either at room temperature or at 37°C . Antibiotics were added when appropriate at the following concentrations: kanamycin 35 μg ml-1 , chloramphenicol 50 μg ml-1 , carbenicillin 100 μg ml-1 , ampicillin 100 μg ml-1 , and tetracycline 15 μg ml-1 . All chromosomal mutations in BB2000 and the Δids strain were made as described in [17] with the following modifications for strains constructed de novo in this study: the suicide vector was pRE118 [61] and the conjugative E . coli strain was strain MFDpir [62] . The BB2000_0531 transcriptional reporter strain MJT02 includes a gene encoding the Venus fluorescent protein [44] immediately following the stop codon of gene BB2000_0531 . All chromosomal mutations were confirmed by PCR amplification followed by Sanger sequencing of the amplified product ( Genewiz , South Plainfield NJ ) or by whole genome sequencing as described in [17] . Strains were grown overnight at 37°C in LB broth with appropriate antibiotics . Overnight cultures were diluted in LB broth to an optical density at 600 nm ( OD600 ) of 1 . 0 , then mixed to the desired experimental ratio and inoculated with an inoculation needle onto a CM55 swarm agar plate . Plates were incubated at 37°C for 18 hours , ensuring that the swarm had covered most of the agar plate . After incubation , swarm composition was measured by using a 48-pin multi-blot replicator to sample the swarm and replica plate on non-swarming LSW- agar with relevant antibiotics as described in [15] . Strains were grown on swarm-permissive agar plates with appropriate antibiotics at 37°C . For consolidating cell samples of wildtype , swarm colonies were left to progress overnight and confirmed to be in consolidation phase by light microscopy . Cells from the swarm edge were then harvested by scraping with a plastic loop into 1 ml of RNA Protect solution ( Qiagen , Hilden , Germany ) . The ΔidsE and CCS02 samples were harvested after overnight incubation by scraping whole colonies into 1 ml RNA Protect solution . Total RNA was isolated using a RNeasy Mini kit ( Qiagen , Hilden , Germany ) according to the manufacturer’s instructions . RNA purity was measured using an Agilent 2200 Tapestation ( Agilent , Santa Clara , CA ) . To enrich mRNA , rRNA was digested using terminator 5’ phosphate dependent exonuclease ( Illumina , San Diego , CA ) according to the manufacturer’s instructions . Enriched RNA samples were purified by phenol-chloroform extraction [63] . The cDNA libraries were prepared from mRNA-enriched RNA samples using an NEBNext Ultra RNA library prep kit ( New England Biolabs , Ipswich , MA ) according to the manufacturer’s instructions . Libraries were sequenced on an Illumina HiSeq 2500 instrument with 250 bp single-end reads , and base-calling was done with Illumina CASAVA 1 . 8 in the Harvard University Bauer Core Facility . Sequences were matched to BB2000 reference genome PMID: 24009111 ( accession number CP004022 ) using TopHat2 using default arguments [64] . Differential expression data were generated using the Cufflinks RNA-Seq analysis suite [65] run on the Harvard Odyssey cluster . Specifically , the mRNA abundance data were generated using Cufflinks 2 . 1 . 1 with max-multiread-fraction 0 . 9 and -multi-read-correct . Samples were combined using cuffmerge with default arguments . Differential expression data were generated using Cuffdiff 2 . 1 . 1 with total-hits-norm . The data was analyzed using the CummeRbund package for R and Microsoft Excel . Gene functions were taken from the KEGG and COG databases [66 , 67] . The data shown in this paper represent the combined analysis of two independent biological and are available at NCBI GEO accession number GSE131647 . Samples of excluded Δids cells were obtained through fluorescent-activated cell sorting ( FACS ) . Fluorescent strains of wildtype-DsRed and Δids-GFP were grown in liquid at 37°C overnight and normalized to OD600 1 . 0 . Cultures were then mixed to the desired experimental ratio and spotted on swarm agar . After the emergence of the third swarm ring , swarm colonies were harvested into 1X phosphate-buffered saline ( PBS ) and sorted using a BD FACSAria cell sorter ( BD Biosciences , San Jose , CA ) into RNA-Protect solution . cDNA samples for RNA-Seq were prepared from sorted samples as described above . For experiments where strain ratio during swarming was examined , liquid cultures were grown overnight with shaking at 37°C . Cultures were normalized to OD600 1 . 0 , mixed to the desired experimental ratio , then used to inoculate swarm-permissive agar plates , which were incubated at room temperature overnight at room temperature to allow inoculum development . Plates were then imaged at 30-minute intervals , incubating at 37°C between measurements . For experiments with the BB2000_0531 transcriptional reporter , 1 μl of mixed , normalized overnight culture were used to inoculate a 1-mm swarm agar pad , which was incubated at 37°C for four hours prior to imaging . Images were taken in GFP ( 150 ms exposure ) , RFP ( 500 ms exposure ) , and phase contrast channels using a Leica DM5500B microscope ( Leica Microsystems , Buffalo Grove IL ) and CoolSnap HQ CCD camera ( Photometrics , Tucson AZ ) cooled to -20°C . MetaMorph version 7 . 8 . 0 . 0 ( Molecular Devices , Sunnyvale CA ) was used for image acquisition , and FIJI [68] was used for image analysis . Raw images are available on Open Science Framework . Strains were grown on swarm-permissive agar plates at 37°C until swarms reached the second round of consolidation ( approximately six hours ) . Swarm colonies were harvested into LB broth and diluted in LB broth to OD600 1 . 0 . Prior to antibiotic exposure , a sample was taken , serially diluted in LB broth , and plated on LSW- agar to count colony-forming units ( CFUs/ml ) in the sample . Antibiotics were added to the normalized culture at the following concentrations: ampicillin 100 μg ml-1 , kanamycin 60 μg ml-1 , streptomycin 50 μg ml-1 , and ciprofloxacin 1 μg ml-1 . Each mixture was incubated with shaking at 37°C . At the specified time-points , samples were taken , serially diluted in LB broth and plated on LSW- agar to measure CFUs/ml . LSW- agar plates were incubated for 16 hours or until visible colonies appeared . Colonies were counted using FIJI . Experiments were performed in triplicate . Co-swarms of Δids-GFP and wildtype-DsRed were inoculated onto CM55 plates and allowed to swarm at 37°C . Samples were harvested from swarm plates and sorted via FACS as described above , except cells were sorted into PBS solution . Immediately after sorting , portions of sorted cell suspension for each strain , containing equal numbers of sorted particles , were used as inoculum for overnight cultures grown at 37°C with shaking in a Tecan Infinite 200 Pro microplate reader ( Tecan , Männedorf , Switzerland ) . OD600 measurements were taken hourly . Experiments were performed in triplicate . For experiments with strains CCS01 and CCS02 , cell sorting was unnecessary . Instead , clonal swarms were directly harvested into PBS . The resulting suspension was diluted to OD600 0 . 1 and used as inoculum for cultures containing relevant antibiotics . A high-performance liquid chromatography ( HPLC ) -based method was used to quantify ppGpp levels , based on the work of Varik et al [42] . Swarm colonies of P . mirabilis were grown to the second swarm ring on CM55 agar . Samples for chromatography were obtained by harvesting cells in 1 ml 1M acetic acid and immediately flash-freezing in liquid nitrogen . Samples were thawed on ice for 1 hour 30 minutes with occasional vortexing , freeze-dried overnight and resuspended in 200 μl MQ-H2O , and then centrifuged at 4°C for 30 min to remove any insoluble fragments . Supernatants were run on a Spherisorb strong ion exchange chromatography column ( 80 Å , 4 . 6 by 150 mm , 5 μm , Waters , Milford MA ) . An isocratic program was used with flow rate 1 . 5 ml/min in running buffer consisting of 0 . 36 M ammonium dihydrogen phosphate , 2 . 5% acetonitrile ( v/v ) , pH 3 . 6 . Nucleotide concentrations were quantified by measuring UV absorbance at 252 nm , comparing peaks to those obtained from purified nucleotide and ppGpp samples ( Trilink Biotechnologies , San Diego , CA ) . | A resident of animal intestines , Proteus mirabilis is a major cause of catheter-associated urinary tract infections and can cause recurrent , persistent infections . Swarming , which is a collective behavior that promotes centimeter-scale population migration , is implicated in colonization of bladders and kidneys . A regulatory factor of swarming is kin recognition , which involves the transfer of a self-identity protein from one cell into a physically adjacent neighboring cell . However , how kin recognition regulates swarming was previously unclear . We have now shown a mechanism linking kin recognition , swarm migration , and antibiotics tolerance: cells induce a transient antibiotics-tolerant , persister-like state in adjacent non-identical cells which in turn prevents non-identical cells from continuing to participate in collective swarming . These affected non-identical cells continue to exhibit large-scale gene expression suggesting an active shift into a different expression state . These data provide two key insights for the field . First , kin recognition can be a regulatory mechanism that acts with spatial and temporal precision . Second , induction into an antibiotics-tolerant state , instead of occurring stochastically , can be physically and spatially regulated by neighboring cells . These insights highlight the importance of further developing four-dimensional ( time and X- , Y- , Z-axes ) model systems for interrogating cell-cell signaling and control in microbial populations . | [
"Abstract",
"Introduction",
"Results",
"Materials",
"and",
"methods"
] | [
"bacteriology",
"antimicrobials",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"pathogens",
"drugs",
"liquid",
"chromatography",
"microbiology",
"social",
"sciences",
"secretion",
"systems",
"animal",
"behavi... | 2019 | Peer pressure from a Proteus mirabilis self-recognition system controls participation in cooperative swarm motility |
Adaptation of molecular structure to the ligand chemistry and interaction with the cytoskeletal filament are key to understanding the mechanochemistry of molecular motors . Despite the striking structural similarity with kinesin-1 , which moves towards plus-end , Ncd motors exhibit minus-end directionality on microtubules ( MTs ) . Here , by employing a structure-based model of protein folding , we show that a simple repositioning of the neck-helix makes the dynamics of Ncd non-processive and minus-end directed as opposed to kinesin-1 . Our computational model shows that Ncd in solution can have both symmetric and asymmetric conformations with disparate ADP binding affinity , also revealing that there is a strong correlation between distortion of motor head and decrease in ADP binding affinity in the asymmetric state . The nucleotide ( NT ) free-ADP ( φ-ADP ) state bound to MTs favors the symmetric conformation whose coiled-coil stalk points to the plus-end . Upon ATP binding , an enhanced flexibility near the head-neck junction region , which we have identified as the important structural element for directional motility , leads to reorienting the coiled-coil stalk towards the minus-end by stabilizing the asymmetric conformation . The minus-end directionality of the Ncd motor is a remarkable example that demonstrates how motor proteins in the kinesin superfamily diversify their functions by simply rearranging the structural elements peripheral to the catalytic motor head domain .
Motor proteins in the kinesin superfamily play critical roles in a number of cellular processes such as vesicle and organelle transport , microtubule depolarization , chromosome and spindle organization during cell division [1] , [2] , [3] . They convert chemical energy associated with ATP hydrolysis into mechanical work to undertake their specific tasks in the cell . While sharing a very similar structure , especially in the catalytic motor head ( MH ) domain , proteins in the kinesin superfamily exhibit stark variations in their biological function [4] , [5] . Kinesin-1 moves from the minus-end of MTs to the plus-end [6] , [7]; kinesin-5 ( or Eg5 ) assembles antiparallel MTs to organize bipolar spindles [8] , [9]; kinesin-13 , responsible for the depolymerization of MTs , diffuses bidirectionally along the MT [10]; and kinesin-14 ( Ncd ) is minus-end-directed [11] . Among many outstanding questions related to motor dynamics [12] , it is of particular interest to ask how the directionality of the molecular movement is determined for a given structure of a motor . Among the motor proteins in the kinesin family , one of the best studied is the kinesin-1 that walks hand-over-hand towards the plus-end of MTs [13] , [14] , [15] . In kinesin-1 , NT-dependent affinity of the MH to MTs and order-disorder transition of the neck-linker , which connects the MH with C-terminal neck-helix , were suggested as the key structural elements responsible for the motor directionality and head-to-head coordination [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] . For a given binding interface between the MH and MTs , ATP-binding induced disorder-to-order transition of the neck-linker rectifies the diffusive dynamics of a tethered head towards the plus-end . In contrast to the kinesin-1 , Ncd motors retain its MH next to the N-terminal neck-helix . Cryo-electron microscope studies and a single molecule assay suggest that Ncd generates force from the plus to minus-end using only one head by keeping the remaining head away from the MTs during the entire motor cycle [25] , [26] , [27]; thus suggesting that the dynamics of Ncd is non-processive [11] , [26] . Although the reconstructed images of the kinesin-1 and Ncd on MTs provide partial glimpses of their functional mechanism [25] , [26] , [27] , a further study is needed to address the structural and dynamical origin of opposite directionalities for the two motors . Several findings from previous experiments provide valuable insights into Ncd dynamics . Stopped-flow experiment by Foster et al . suggested an apparent asymmetry between the two MHs in Ncd dimers [28] , showing biphasic ADP release kinetics with disparate time scales of 7 s−1 and 0 . 005 s−1 . An asymmetric Ncd conformer crystallized by mutating a residue ( N600K ) in the MT binding motif showed a structure , with a considerably low ADP affinity , distorted from the wild type [29] , suggesting that such an asymmetric conformation can be accessible for the wild type Ncd . As a plausible link between the biphasic ADP release kinetics and the presence of accessible asymmetric states , 13P-NMR spectroscopy showed that multiple Ncd dimer conformations , including asymmetric conformers , are accessible in solution [30] . Furthermore , according to mutation studies , the directionality of motor hinges on the structural elements outside of the MH domain [31]: a chimeric kinesin-1 with Ncd's neck is minus-end-directed , whereas an Ncd with kinesin-1's neck-linker exhibited the plus-end directionality although the motility was impaired in both cases 32 , 33 , which highlights the role of the neck domain for the motor directionality . Despite a number of studies on Ncd motors , still missing is a more integrated understanding that explicitly relates the scattered experimental findings mentioned above with the mechanochemical cycle of Ncd motors and their minus-end directionality in structural terms . A series of efforts have recently been made to decipher a number of crucially interesting issues involving molecular motors by performing large-scale molecular simulations [19] , [20] , [24] , [34] , [35] , [36] , [37] , [38] , [39] , [40] , [41] . Among them , our previous studies that employed the coarse-grained structure-based model of kinesin-1 , which retains its native-like configuration in the absence of external perturbation , could show that the conformational adaptation in response to the chemical and mechanical stresses renders the movement of kinesin-1 plus-end directed and highly processive [19] , [20] , [23] . Here , our model of the Ncd motor , adopting the computational strategy based on the theory of protein folding as the previous models [19] , [20] , [23] , [42] , [43] , clarifies how the repositioning of neck-helix to the N-terminal and the deletion of neck-linker from kinesin-1 drastically alter the intramolecular structural adaptation , giving rise to the minus-end directionality and associated mechanochemistry unique to the Ncd motors . Finally , our study show explicitly how the basic physical principle of protein folding [19] , [20] , [23] , [44] , [45] , [46] play roles for motor proteins in the kinesin superfamily to achieve a wealth of different behaviors needed for their biological functions .
Native contact maps and superimposed structures between the kinesin-1 ( Fig . 1a ) and symmetric conformer of Ncd ( Fig . 1b ) reveal notable similarity of the two molecules . The MH domains of two motors are very similar in both structure ( Fig . 1c ) and contact map ( Fig . 1d ) . The main difference between the Ncd and kinesin lies in the organization of structural motifs peripheral to the MH domain . While kinesin-1 has a C-terminal neck-helix followed by neck-linker and MH , Ncd has a N-terminal neck-helix that is linked to the MH via a short stretch of amino acids called neck-junction , which introduces slight variations in the contact maps . It is of particular note that the contacts between the neck-helix ( or coiled-coil stalk ) and the MH are unique to Ncd ( enclosed by the magenta box in Fig . 1d ) . Instead of these head-stalk contacts , kinesin-1 has multiple contacts between the MH and neck-linker ( enclosed by the green box in Fig . 1d ) . We show that this seemingly minor difference in contact map leads to stark differences in terms of motor functions between Ncd and kinesin-1 . In the asymmetric conformation of Ncd ( Fig . 2a ) , one of the two MHs has a structure rotated around the neck-junction and thus “distorted” from its symmetric counterpart that leads to a different contact map ( Fig . 2c & d ) . As we have discussed earlier , wild type Ncd can exists in either the symmetric or asymmetric form in solution . In our structure-based potential we combine the information of the native contacts from both symmetric and asymmetric states . As clearly depicted in the Ncd structure of Fig . 2b , it is impossible for the stalk ( residue indices around 50 ) to form contacts concurrently with MH in the symmetric state ( residue indices around 125 ) and in the asymmetric state ( residue indices 80–100 ) . Thus , the head-stalk contacts in the symmetric and asymmetric states are energetically frustrated and this leads to a competitive energy landscape ( see Methods for further details of how this is implemented in energy Hamiltonian ) [42] , [43] , [47] . Our simulation of the Ncd dimer in the absence of the MT reveals both symmetric and asymmetric conformations . Comparison of the contact maps explicitly demonstrates that the pattern of head-stalk contacts in the asymmetric state differs from that in the symmetric state ( Figs . 2c and 2d ) . Interestingly , while each MH can acquire distorted configuration with equal probability , we do not observe a conformation where both MHs are distorted . We found that only one of MHs is distorted towards its asymmetric conformation; the other MH still retains a conformation similar to the one in symmetric dimer . To quantify our findings from simulation , we follow the dynamics of the inter-residue distance ( E567 from each monomer ) in the dimer structure ( Fig . 3a ) . The distribution of inter-residue distance shows two peaks corresponding to the symmetric and asymmetric conformations with preference to the symmetric conformer . Since the probability for a monomer to sample one of the two asymmetric states is small ( prob . = exp ( −2 . 1 ) ∼0 . 12 from Fig . 3b ) , the odds to observe a conformation with both MHs being distorted are even smaller ( Fig . 3b ) . Two-dimensional free energy diagram ( Fig . 3b ) , plotted using the distances between K325 and E567 of the two monomers , shows that there are three free energy minima; one for the symmetric state , the others for the two asymmetric states ( asymmetric A and B in Fig . 3b ) . It is also evident from the free energy diagram that a configuration with both heads being distorted is inaccessible . In addition , principle component analysis [48] of our simulation trajectories ( see Methods ) reveals that the first two low frequency modes are associated with symmetry-breaking fluctuations that are responsible for bringing the symmetric Ncd structure towards the asymmetric state ( Fig . 3c ) ( also see the supporting movie S1 ) . To identify the correlation between the MH distortion and the NT binding affinity , we first monitor the change in the number of the head-stalk contacts and the RMS deviation of the NT binding pocket ( residues 432–451 , 535–555 , 576–590 ) ( Fig . 4a ) . Our simulation finds an interesting anti-correlation between these two quantities at the global level when the distortion dynamic between the symmetric and asymmetric conformers take place ( Fig . 4b ) . As the head-stalk contacts in the symmetric state decrease , the RMS deviation of the NT binding pocket region increases . This anti-correlation is observed in both heads , suggesting that there is an allosteric communication between the head-stalk contact region and the nativeness of the binding pocket . Furthermore , inspecting the details of NT binding pocket , i . e . ( residue 434–441 P-loop ( residue 434–441 ) , switch I ( residue 540–549 ) and switch II ( residue 580–585 ) [49] in both symmetric and asymmetric states , we find that there is a noticeable enhancement of fluctuation in switch I region whereas the change in the fluctuation dynamics of switch II and P-loop is relatively small ( Fig . 4c , 4d ) . This provides is a clear signature of the increased distortion ( especially for switch I ) of the ADP binding pocket when it acquires an asymmetric conformation . It may be argued that there is an allosteric communication between the head-stalk contact region and ADP binding pocket as symmetric-asymmetric transition involves changes in head-stalk contact pattern . Provided that the distortion in the NT-binding pocket leads to a weakening the ADP binding affinity [18] , [19] , our simulation results suggest a possible explanation for the origin of the biphasic ADP release kinetics observed in Ncd dimer in solution and also confirm the proposition by Foster et al [28] about an apparent asymmetry between the two MHs of the Ncd dimer . Based on our above findings , we also propose that the population shift ( or conformational selection ) mechanism is involved when Ncd binds to MTs . In the solution , Ncd can have a stable symmetric conformation with an asymmetric conformation . However , because the NT free-ADP ( φ-ADP ) state , attained due to fast ADP release from asymmetric conformation [26] , is a strong MT-binder , the equilibrium population associated with asymmetric ADP-ADP state in solution would be depleted as more Ncd motors bind to MTs . This MT-binding process produces a flux of population shift from the stable symmetric conformation to the less stable asymmetric conformation . Next , we perform the simulation of φ-ADP state of Ncd dimer with the NT free head being bound to the MT . Since no crystal structure is available in the protein data bank ( PDB ) for the MT bound Ncd , we use the binding interface ( contacts ) between MT and kinesin-1 by noting that MH of Ncd is homologous to that of kinesin-1 . This is also supported by docking studies that revealed similar orientation of the binding head for these two motors [25] . Thus , by incorporating the MH-MT contacts in our structure-based model we simulated the φ-ADP state on the MTs . Our simulation finds that the Ncd on the MT ( Fig . 5a ) prefers the symmetric conformation whose coiled-coil stalk points towards the plus-end , which is in agreement with the previous Cryo-EM measurement [27] . The distribution of an angle θ , defined using three residues ( LEU296 in the unbound monomer and D424 , E567 in the bound monomer; see Fig . 5 ) is bimodal with a dominant population near θ∼25° for symmetric and a small population near θ∼150° for asymmetric states , respectively . Of particular note is that the interaction of Ncd with MTs greatly reduces the population of asymmetric state in comparison to that of the solution state ( compare Figs . 3a and 5a ) . Subsequently , we perform the simulation of the MT bound Ncd dimer in the ATP-ADP state . Several mutation studies have suggested the importance of the neck-junction region for the directional motility of Ncd dimer [32] , [33] , [50] . Bioinformatics analysis based on COILS [50] , which assess the feasibility of the amino-acid sequence to form a coiled-coil structure , have also reported that the contacts near the neck-junction region are metastable [50] , [51] . Additionally , binding of ATP to the NT free MH may allosterically influence the residues in the junction region . Thus , to mimic the effect of ATP binding we increase structural flexibility by destabilizing the contacts in the junction region ( see Methods for details ) . Our simulation renders a striking inversion from a symmetric to an asymmetric biased landscape ( Fig . 5 ) . The distribution of the angle ( θ ) calculated above for the φ-ADP state ( Fig . 5b ) now shows a peak around 150° , suggesting that the Ncd dimer stays mostly in the asymmetric conformation when the neck-junction becomes flexible . This finding is in agreement with the previous cryo-EM study that the ATP-ADP state has a coiled-coil stalk pointing towards minus-end of MTs [27] . Thus , our simulation indicates a direct link between destabilization of head-stalk contacts and stabilization of the asymmetric Ncd on MTs . In fact , the conformational change of Ncd from symmetric to asymmetric state leads to the swinging of coiled-coil stalk from plus-end to minus-end , accounting for the structural origin of minus-end directionality of Ncd motors on MTs .
The results obtained from our simulation can be recapitulated in the mechanochemical cycle for Ncd motor in Fig . 6 . In solution , the Ncd dimer ( ADP-ADP state ) exists in either ( i ) symmetric or ( ii ) ( or ( ii ) ’ ) asymmetric conformations with more dominant population in the symmetric state . In the asymmetric conformation , distortion in one of the heads from its wild type structure is correlated with enhanced fluctuation near ADP binding pocket . Hence , the distorted head in an asymmetric conformation has a lower ADP affinity that facilitates the ADP release . Simulations from our Ncd model explicitly demonstrate that this symmetric-to-asymmetric conformational transition occurs occasionally and that there is a clear correlation between this transition and the increase of fluctuation at the binding pocket of the distorted head ( especially in switch I ) [49] . This transition explains how the bi-phasic ADP release kinetics occurs while Ncd is a homodimer . Following , the NT-free MH , which retains strong MT-binding affinity ( Kd∼nM ) , binds to the MTs and reaches ( iii ) a further stabilized symmetric conformation on the MT , whose coiled-coil stalk points towards the plus-end . ( iv ) ATP binding to the free head destabilizes the neck-head junction region of the MT-bound head , which overturns the ensemble of symmetric Ncd conformation in the φ-ADP or ADP-ADP state into the ensemble of asymmetric state that has a minus-end-directed coiled-coil stalk . Our studies have identified head-neck junction region as the critical structural element responsible for power stroke mechanism of Ncd , which in fact is amenable to further experimental studies . Furthermore , the step involving ( iii ) → ( iv ) corresponds to the power stroke , i . e . , the force production due to conformational changes . Our estimate of a lower bound of stall force from a probability P ( z ) , which provides a potential of mean force F ( z ) = −kBT log P ( z ) ( see the inset between step ( iii ) and ( iv ) in Fig . 6 ) , where z is the position of stalk tip projected along the MT axis is 0 . 42 pN . Compared to the stall force value of other molecular motors ( 2–3 pN for myosin V ( Biophys J . ( 2000 ) 79:1524–1529 , PNAS ( 2004 ) 101:5542–5546 ) and 5–6 pN for kinesin-1 [7] ) , this value is rather small; however , given that the Ncd is a non-processive motor that should cooperate with other motors to generate force , our estimate of the stall force for a single Ncd , which no experiment has measured to date , is not unreasonable . Finally , the ADP-ADP ( or ADP·Pi-ADP ) state of the Ncd with weak MT affinity ( Kd> mM ) [52] , produced as a result of ATP hydrolysis and subsequent release of inorganic phosphate from the bound head , leads to the dissociation of the Ncd motor from the MT ( ( iv ) → ( i ) ) . The minus-end-directed power stroke of Ncd motor is generated between the step ( iii ) and ( iv ) . The dissociation of Ncd motor after a single stroke suggests that the movement of Ncd on the MTs is non-processive . Thus the minus-end-directed cargo transport is realized by cooperation between multiple Ncd motors [26] . The results of the simulation can be understood as follows . In the case without MT , both the symmetric and asymmetric states were populated as we have constructed a dual-basin structure-based model with a bias towards the symmetric state . The symmetric state is more populated because the native bond , angle and dihedral parameters are derived from symmetric crystal structure ( see Methods ) . In the MT bound φ-ADP state , the presence of MT reduces the configuration space of the Ncd . This increases the stability of symmetric state whose coiled-coil stalk points towards the plus end of MT . Binding of ATP to the empty head , destabilizes few contacts in the head-stalk junction region which in turn destabilizes the symmetric basin . Therefore , it now preferentially populates the asymmetric basin whose coiled-coil stalk points towards the minus end of MT . Molecular motors in the kinesin superfamily share the common structural motifs , such as the MH , the neck-linker ( or the extended coiled-coil stalk ) , and the neck-helix that are used in their biological functions . The MH domain used for MT-binding is essentially identical for kinesin-1 , Ncd , Eg5 that use tetrameric MHs ( two pairs of dimer heads ) to bind the two anti-parallel MTs , and other kinesin-like proteins . Furthermore , the MT-binding affinity of MH in Ncd obeys a similar rule as in kinesin-1; the MH in ADP state has weak MT-affinity whereas the MH in φ or ATP state binds MT strongly [23] , [53] , [54] , [55] . Although there is a quantitative difference in the value of the binding affinity for different kinesin families due to the sequence variation , kinesin-1 and Ncd have the identical “footstep” on MTs and bind to MTs tightly during the processes of ATP binding and hydrolysis . Our study asserts that the determining factors of the directionality in kinesin-1 and Ncd are the structural motifs peripheral to the MH ( Fig . 1d ) , i . e . , neck-linker and neck-helix . This is evident from our Ncd model that is adapted from the previous model for kinesin-1 . The difference in directionality for the kinesin-1 and the Ncd is a remarkable example of how motor proteins in the kinesin superfamily diversify their functions by repositioning the structural motifs peripheral to the MH domain that are in charge of catalytic activity [31] . Our study provides unambiguous structural and dynamical understandings to the effect of such repositioning on the biological function .
Using the MT bound Ncd structures generated by aligning the crystal structures of Ncd on MTs , we simulated Ncd motor dimer with or without the MT track . Langevin dynamics simulations using the SB potential were performed where equation of motion was integrated using a Verlet algorithm . The energy function used for these simulations are given as follows: Here , total energy Hamiltonian is separated into intramolecular interaction for the Ncd and intermolecular interaction at the Ncd-MT interface . The subscripts N and MT denote the Ncd and MT , respectively and N-MT denotes the Ncd-MT interaction . We have fixed coordinates of MT in space throughout the simulation . The first and second terms associated with bond potential in define the backbone interactions . The bond distance ( ) between the neighboring residues i and i+1 are constrained harmonically with respect to its native bond distance ( ) with a spring constant ( ) of 20 kcal/ ( mol×Å2 ) . In the second term , the angle defined among residues i , i+1 and i+2 is constrained harmonically around its native state value ( ) with = 20 kcal/ ( mol×rad2 ) . The third term represents the dihedral angle potential with = 1 . 0 kcal/mol and = 0 . 5 kcal/mol that describes the rotation of the backbone involving successive residues from i to i+3 . The native values , , , and are taken from the symmetric conformer and the superscript ( S ) refer to the symmetric state . The 10–12 Lennard-Jones ( LJ ) potential is used in to describe the interactions that stabilize the non-bonded native contacts . A native contact is defined for a pair of residues ( i and j ) , if the distance between them is less than 8 Å in the native state and >3 . If i and j ( k for Ncd-MT interface ) residues are in contact in the native state , ( or ) = 1; otherwise ( or ) = 0 . It is of particular note that this non-bonded LJ potential retains three terms [42] , [43]: ( i ) the native bias mainly from symmetric state ( the first term ) with an equilibrium distance for i and j residue pairs , ( ii ) the perturbative bias from asymmetric state ( the second term ) with an equilibrium distance , and ( iii ) the repulsive potential for the residue pairs that belong to neither the symmetric nor the asymmetric state . The superscript of , i . e . , X = , , and , refers to the symmetric state , “purely” asymmetric state , and the union of symmetric and asymmetric state , respectively , so that = 1 when i and j residue pairs are in the state X , otherwise = 0 . Non-native pairs with ( or ) = 0 are under repulsive potential with a distance parameter σ = 4 Å . We assign εh = 1 . 8 kcal/mol for the upper part of intraneck , interneck helix ( residue <331 ) and Ncd-MT interactions to secure coiled-coil association and MT binding , respectively . For other residue-residue interactions , we set εh = εl = 1 . 0 kcal/mol regardless of sequence identity . The parameters determining the native topology , , and are determined from the symmetric and asymmetric crystal structure of Ncd . The intramolecular non-bonding energy term , , provides competition between the native contacts from symmetric and asymmetric states with more bias towards the symmetric conformer when NT-binding pocket is in φ or ADP state . This bias changes to the asymmetric state by removing a few native contacts from the neck-stalk junction ( see below ) . Initial structures in each case were relaxed under the SB Hamiltonian and subsequently Langevin dynamics simulations at low-friction limit were performed at T = 300 K to sample the equilibrium structural ensemble . The equation of motion for the Langevin dynamics used for integration iswhere ζ is the friction coefficient , is the conformational force . is the random force satisfying where integration time h is discretized . In this dynamics we chose ζ = 0 . 05 and h = 0 . 0025 with . Low friction was chosen for the purpose of effective conformational space sampling [58] . To simulate the Ncd dimer in solution we used the topological information from the crystal structures with PDB code of 1CZ7 ( symmetric state ) and 1N6M ( asymmetric state ) . All the native backbone and dihedral parameter was derived from the symmetric crystal structure . For each monomer nonbonded potential corresponding to the additional contacts for the asymmetric conformation are added to the potential . Here , we neglected the Ncd-MT interaction term from the Hamiltonian to simulate Ncd dimer in solution without MT . We superimposed the structure of each frame from the simulation onto the symmetric dimer to obtain the covariance matrix [48] . After diagonalization we examined the first few low frequency modes . The lowest frequency eigenvector was then represented residue-wise to illustrate the collective dynamics towards the asymmetric state . For the construction of the MT bound state of Ncd , we used three structures: single-headed kinesin ( KIF1A ) bound to the MT ( PDB ID 1IA0 ) ; Ncd dimer structure ( PDB ID 1CZ7 ) ; and three consecutive αβ tubulin dimers from 13-protofilament MT structure [57] . We overlapped chain A of 1CZ7 onto chain K of 1IA0 and αβ tubulin dimer of 1IA0 onto terminal αβ tubulin dimer of the tubular MT structure . This leads to the MT bound state of Ncd . Here , we assume that Ncd employs a similar binding interface on the MTs with kinesin-1 , which was shown in earlier studies [26] , [55] , [59] , [60] , [61] . Using the structure obtained from the construction of the MT bound state of Ncd , we extracted the Ncd-MT contacts at the interface and included in the total Hamiltonian to generate the conformations corresponding to the NT free-ADP state on MT . It was proposed earlier that the head-stalk junction of Ncd plays important role in determining the directional motility of motor [32] , [33] , [50] . While retaining the same Hamiltonian as the φ-ADP state on MT , we deleted a few native contacts in the junction region ( residues 341–345 of the coiled-coil stalk and residues 346–351 of the MH ) between the coiled-coil stalk and bound head of Ncd to mimic the binding of ATP . The junction region ( lower part of extended coiled-coil region ) of the Ncd motor is shown explicitly in the Fig . 1b . | Proteins belonging to the kinesin superfamily are responsible for vesicle or organelle transport , spindle morphogenesis , and chromosome sorting during cell division . Interestingly , while most proteins in kinesin superfamily that share the common catalytic motor head domain have plus-end directionality along microtubules , kinesin-14 ( Ncd ) exhibits minus-end directionality . Despite the several circumstantial evidences on the determining factors for the motor directionality in the last decade , a comprehensive understanding of the mechanism governing the Ncd minus-end directionality is still missing . Our studies provide a clear explanation for this minus-end directionality and the associated mechanochemical cycle . Here , we modeled an Ncd motor by employing structural details available in the literature to simulate its conformational dynamics . Simulations using our structure-based model of Ncd assert that the dynamics due to a simple rearrangement of structural elements , peripheral to the catalytic motor domain , is the key player in determining both the directionality and mechanochemistry unique to Ncd motors . Although Ncd has a different directionality , it uses a similar strategy to kinesin-1 of structural adaptation of the catalytic motor domain . Therefore using the same physical principle of protein folding and very similar structural elements , motors in the kinesin superfamily are able to achieve a variety of biological function . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"physics",
"theoretical",
"biology",
"biophysic",
"al",
"simulations",
"protein",
"folding",
"biophysics",
"theory",
"biology",
"computational",
"biology",
"biophysics",
"simulations",
"biophysics"
] | 2012 | The Origin of Minus-end Directionality and Mechanochemistry of Ncd Motors |
A reliable and effective human challenge model is needed to help down-select the most promising ETEC vaccines currently under development . Such a model would need to reliably induce diarrhea in a high proportion of volunteers using the lowest possible inoculum to maximize safety and sensitivity . Previously we validated a challenge model that utilized a dose of 2x107 CFU of ETEC strain H10407 ( LT+ , ST+ , CFA/I+ and O78+ ) to induce attack rates for moderate to severe diarrhea ( MSD ) of ~60–70% . Here we detail efforts to further refine the model in an attempt to determine if a lower challenge dose of H10407 can be used . Thirty subjects were randomized 1:1 to receive an oral administration of H10407 at doses of 106 or 105 CFU in bicarbonate buffer . After challenge , subjects were monitored for signs and symptoms of enteric illness and stool samples were collected to detect shedding of the challenge strain . Systemic and mucosal immune responses were measured using serum , antibody in lymphocyte supernatant and fecal samples . The attack rate was 13 . 3% ( 2/15 ) and 26 . 7% ( 4/15 ) for MSD in the 105 and 106 groups , respectively . Four MSD cases met criteria for early antibiotic treatment . All subjects but one shed the challenge strain in fecal samples . The frequency and magnitude of anti-LT toxin , CFA/I and LPS O78 immune responses were antigen , dose , severity of diarrhea and shedding levels dependent . Notably , although of lower magnitude , there were considerable immune responses in the subjects with no diarrhea . This may indicate that immune responses to asymptomatic infections of ETEC in children in the endemic countries may contribute to protection . Based on this and our prior studies , we conclude that a dose of 2x107 H10407 remains the lowest practical dose for use in future volunteer studies evaluating candidate vaccines and other preventive or therapeutic ETEC interventions . Trial registration: ClinicalTrials . gov NCT00844493 .
Enterotoxigenic Escherichia coli ( ETEC ) remain among the most common bacterial causes of diarrhea-associated morbidity and mortality [1–4] . In Africa and Southeast Asia , ETEC is also found to be an important cause of morbidity and mortality in age-groups older than 5 years of age [5] . Studies have shown that children infected with ETEC are at higher risk of becoming stunted [3 , 6 , 7] . As currently understood , the key virulence factors for ETEC are colonization factor antigens ( CFs ) and enterotoxins . CFs mediate bacterial attachment to host small intestinal epithelial cells and subsequent colonization , whereas enterotoxins including heat-labile ( LT ) and heat-stable ( ST ) toxins , occurring as either LT or ST alone or both LT/ST , disrupt fluid homeostasis in host epithelial cells , which leads to fluid and electrolyte hypersecretion and diarrhea [2 , 8] . ETEC vaccine development has been a long standing WHO priority [9] . However , there are no licensed vaccines for ETEC are currently available but several approaches to develop an effective vaccine are underway . A significant challenge to successful vaccine development is our poor understanding of the immune responses that correlate best with protection against ETEC illness . Although subclinical ETEC infections are very common in endemic countries , there is a gap of knowledge regarding the breadth and magnitude of immune responses to ETEC in asymptomatic cases which may prove helpful in guiding ETEC vaccine development strategies going forward . Historically , the ETEC human challenge model has been considered an important tool in facilitating decisions regarding the most promising ETEC vaccines . The human challenge models can provide a platform for the evaluation and screening of vaccine efficacy before more expensive and more long-term field trials are conducted in at-risk age-groups in low and middle income countries and in travelers . To maximize its utility in this regard , it is essential that the model be optimized to ensure its reliability as a decision making tool . An ideal challenge model would reproducibly induce diarrhea in a high proportion of volunteers ( >50% ) using the lowest possible inoculum to better maximize the sensitivity of the model , particularly in the evaluation of preventive or therapeutic interventions . If the attack rate is too low , a much larger sample size will be needed . However , an excessively high dose might achieve high attack rates but increase the risk that these high challenge doses could overwhelm any vaccine-induced protective immunity and lead to the down selection of vaccines that might have potential as public health tools [9 , 10 , 11] . Lower ETEC doses ( <108 ) yielded inconsistent attack rates [10 , 12 , 13]; however , lower challenge doses have been used effectively for other well-established enteric disease challenge models , like cholera , Shigella and Campylobacter [14 , 15 , 16] . In previous studies we validated a challenge model which utilizes an ETEC dose of 2x107 CFU of virulent ETEC strain H10407 with an overnight fast prior to challenge that resulted in MSD attack rates of 60–70% [8 , 17] . Our earlier studies also demonstrated that persons who had previously developed diarrhea following challenge were protected from illness when re-challenged with the homologous strain [8 , 17] . In this study , we attempted to further refine the challenge model by characterizing the clinical illness induced by lower ETEC challenge doses of 106 and 105 CFU . We also performed a comprehensive assessment of the serum and mucosal antibody responses induced by these lower challenge doses of ETEC as well as the shedding levels of the challenge organism in fecal samples from challenged subjects . In addition , to understand the kinetics and magnitudes of immune responses in the subjects with asymptomatic infection we compared antigen specific responses in subjects with MSD versus those who did not develop diarrheal illness post-challenge .
The protocol was conducted under BB-IND 12 , 243 at the Center for Immunization Research ( CIR ) , Johns Hopkins Bloomberg School of Public Health ( JHBSPH ) . Approval to conduct the study was provided by the Western Institutional Review Board ( Olympia , WA ) for JHBSPH and PATH and by the Institutional Biosafety Committee of the Johns Hopkins Institutions . The study was registered in clinicalTrials . gov under NCT00844493 , ( for detailed protocol https://clinicaltrials . gov/ct2/show/NCT00844493 ) . Healthy , 18- to 45-year-old , male or female subjects were recruited for the study . The study was explained to subjects in detail and signed , witnessed consent was obtained . The pre-challenge health status of subjects was assessed by written and oral medical history , physical examination , complete blood count ( CBC ) , urinalysis , urine toxicology , blood chemistries , and tests for liver and kidney function , HIV-1 , hepatitis B , and hepatitis C . Subjects were excluded if they had significant medical problems detected by history , physical examination , or screening laboratory tests; if an HIV-1 , hepatitis B , or hepatitis C test was positive; or if they had traveled to countries where ETEC or cholera infection is endemic within two years prior to receipt of investigational agent . The study was conducted between September 2011 and April 2012 . The subjects were admitted into the 30-bed CIR Isolation Unit . ETEC strain H10407 was used as the challenge strain , and all subjects received a challenge dose of either 105 or 106 CFU of this virulent organism . There were no placebo recipients in the study . A total of 30 subjects were challenged . The participants were divided randomly ( 1:1 ) into two groups of 15 subjects per group in a double-blinded manner ( S1 Text ) . Group A received a dose of 1x105 and group B received a dose of 1x106 CFU . Both the groups received bicarbonate buffer ( 2 gms ) with the dose . All subjects fasted overnight ( ~9 hours ) before dosing and for 90 minutes after dosing . The challenge strain H10407 is ETEC serotype O78:H11 , produces heat labile toxin ( LT ) and two forms of heat stable toxin ( STh and STp ) . The strain also produces colonization factor I ( CFA/I ) . It is sensitive to ampicillin , trimethoprim-sulfamethoxazole , and ciprofloxacin , which are used typically to treat ETEC H10407 infections . A current good manufacturing practices ( cGMP ) quality production cell bank ( PCB ) for ETEC H10407 was prepared by the Walter Reed Army Institute of Research ( WRAIR ) Pilot Bioproduction Facility ( Silver Spring , MD ) . The manufacturing information and production records for the PCB of these strains were provided to the FDA under BB-IND-7766 ( batch production record 285–000 , lot number . 0519 ) . The ETEC H10407 challenge strain was stored in 2-ml cryostorage tubes ( 1 ml per tube ) held at -80°C ±10°C in the bio-production facility at WRAIR . Cryovials containing organisms from this PCB were transferred on dry ice from WRAIR to the CIR Enterics Research Laboratory , JHBSPH , and stored at -80°C±10°C in a locked and temperature-monitored freezer . The challenge inocula were prepared as described previously [17] with modifications to achieve the expected doses . The inocula were prepared from fresh , plate-grown organisms , using a study-specific procedure . The number of CFU in the inocula was validated by titrating and plating on Luria agar plates ( Becton , Dickinson and Company , Sparks , MD ) before and after administration to volunteers . A sample of the final inoculum was also examined by Gram stain for purity and by agglutination in anti-H10407 antiserum before being administered to subjects . The bicarbonate buffer was prepared from USP-grade sodium bicarbonate ( Fisher Scientific , Fair Lawn , NJ ) by dissolving 13 . 35 g of sodium bicarbonate in 1 , 000 ml of sterile water for irrigation ( Hospira , Inc . , Lake Forest , IL ) . As with the earlier studies [17] , the bacterial challenge was administered in bicarbonate buffer after an overnight ( 9 hours ) fast . At approximately 0900 h on the day of challenge , subjects drank 120 ml of the sodium bicarbonate buffer to neutralize gastric acidity . Approximately 1 min later , subjects drank the ETEC H10407 inoculum dissolved in 30 ml of the same buffer solution . Medical interviews and physical examinations were performed daily and additional medical assessments and vital sign measurements were performed by the study team 3 times daily during the inpatient stay . Active solicitation regarding the following symptoms took place during the medical interview: fever , vomiting , nausea , abdominal pain , abdominal cramping , myalgia , malaise , bloating , flatulence , headache , lightheadedness , chills , constipation , and anorexia . Fever was defined as an oral temperature of 100 . 4°F or above . Severity of fever , vomiting , other symptoms and grades of stools were categorized as below [17] . Diarrhea was defined as 1 loose/liquid stool ( grade 3 ) of 300 g in any 24-hour period or 2 loose/liquid stools totaling 200 g during any 48-hour period within 120 h of challenge with ETEC H10407 . Diarrhea was classified as mild ( 1 to 3 diarrheal stools totaling 200 to 400 g/24 h ) , moderate ( 4 to 5 diarrheal stools or 401 to 800 g/24 h ) , or severe ( 6 or more diarrheal stools or 800 g/24 h ) . The subjects were treated with either oral rehydration solution ( Ceralyte; Cera Products , Inc . , Columbia , MD ) or intravenous ( i . v . ) fluid or administration of antibiotics as per protocol guidelines . All the subjects were treated with ciprofloxacin ( 500 mg twice daily ) approximately 120 h after challenge except those who meet the criteria for early antibiotic therapy and were treated earlier . To be eligible for discharge , subjects needed to have at least 2 negative stool cultures for ETEC H10407 . Subjects were seen as outpatients in the clinic 10 and 28 days after challenge and contacted by telephone about 3 months after challenge . Post challenge shedding levels of the challenge strain ETEC H10407 was detected in the fecal samples [17 , 18] . Qualitative and quantitative microbiology assessment were done as described [17] . For qualitative assessment , up to 5 colonies appearing to be E . coli on MacConkey agar plates were tested for agglutination with antiserum to H10407 . For semiquantitative microbiology , the fecal sample was serially diluted and spread onto MacConkey agar . After overnight incubation , the concentration of bacteria which appeared to be E . coli was calculated , and the proportion of these colonies ( of 5 colonies tested ) which agglutinated with anti-H10407 antisera was recorded . For analysis , subjects who did not shed H10407 , a value of 1 was used in place of 0 to calculate the geometric mean ( GM ) . If a subject had a positive qualitative sample , but negative quantitative sample , a value of 500 ( corresponding to half of the lowest detectable limit of the quantitative assay ) was used for the GM calculation . Blood and fecal specimens were collected from the challenged subjects to measure systemic and mucosal immune responses . Venous blood from subjects was collected in BD Vacutainer cell preparation tubes ( CPT ) with heparin ( Becton Dickinson , Franklin Lakes , NJ , USA ) and processed for the antibody in lymphocyte supernatant ( ALS ) assay as described before [8] . Briefly , peripheral blood mononuclear cells ( PBMCs ) were isolated and resuspended at 1x107 viable lymphocytes per ml . PBMCs were then incubated for 72 h with no antigenic stimulation . The supernatant fluid was cryopreserved and subsequently used in an enzyme linked immunosorbent assay ( ELISA ) to measure the concentration of antibody released by the PBMCs . The ALS samples were tested for antigen- specific IgA . An ALS response was defined as a ≥4-fold increase in antigen-specific IgA antibody titer over the baseline . Pre- and postchallenge venous blood samples were collected and processed for serum [8 , 17] . A serum response was defined as a ≥2 . 5-fold increase over the baseline . Antibodies from fecal samples were extracted as described before [8 , 19 , 20] . Thawed stool sample was mixed with protease inhibitor cocktail . Next , the mixture was centrifuged at 12 , 000 x g for 30 min . Aliquots of the supernatant were stored at -80°C until they were assayed by ELISA for total and specific IgA antibody contents . ELISAs with serum , ALS , or fecal samples were performed according to standard protocols [8 , 19] . In short , 96 well microtiter plates were coated with purified CFA/I or LPS diluted in PBS . A GM1-ELISA method was used for the determination of levels of LT-specific antibodies [8] . The plates were blocked and washed . Serum , ALS , and fecal samples were diluted 3-fold in the plates using 0 . 1% BSA-PBS-Tween as a diluent . Anti-human IgG or anti-human IgA conjugated with horseradish peroxidase ( KPL , Baltimore , MD ) followed by o-phenylenediamine dihydrochloride ( OPD ) ( Sigma , St . Louis , MO ) was added to each well . After 20 min , the plates were read at 490 nm in an automated ELISA reader . Levels of specific and total IgA for fecal samples were determined [8 , 19] . Antibody titers for fecal specimens were expressed as units per milligram of IgA , and these titers were calculated by using the specific titer ( in units per milliliter ) divided by the total IgA contents ( in micrograms per milliliter ) and then multiplying by 1 , 000 . The total IgA contents for given samples were determined by ELISA using a standard IgA preparation ( Sigma , St . Louis , MO ) with a known IgA concentration ( 1 mg/ml ) . Titers were calculated to interpolate the dilution of serum , which yielded an optical density ( OD ) above baseline of 0 . 2 for serum samples and of 0 . 4 for ALS and fecal samples . Prechallenge and postchallenge sera were tested simultaneously in the same plate . GM1 and LTB antigens were purchased from Sigma ( Sigma- Aldrich , St . Louis , MO ) , and CFA/I and LPS antigens were obtained from the laboratory of Ann Mari Svennerholm ( University of Gothenburg , Gothenburg , Sweden ) . Chi-square and t-tests were used to determine differences between groups as appropriate for categorical and continuous variables . Results of statistical analyses were considered significant only if p-values were less than 0 . 05 . We used GraphPad Prism ( GraphPad , CA ) software to analyze the results .
A total of 30 ETEC H10407-naïve subjects ( based on possible exposure history solicited at screening ) were enrolled in the study and randomly assigned to one of 2 dosing groups ( Fig 1 ) . Both groups had similar demographic characteristics ( Table 1 ) . Clinical outcomes and selected solicited adverse events for both groups are summarized in Table 2 and S2 Table respectively . The number of subjects with MSD was somewhat higher in group B , 4/15 ( 26 . 7% ) compared to 2/15 ( 13 . 3% ) in group A . Four of these subjects with MSD were given early antibiotics ( before day 5 after challenge ) . There was 1 subject with mild diarrhea in each group . The mean incubation period from challenge to initiation of diarrhea was similar in both the groups ( ~60hrs ) ( Table 2 ) and somewhat longer than reported previously for higher H10407 challenge doses ( ~48hrs ) ( 10 , 17 ) . The mean total diarrhea stool output and mean maximum diarrhea stool output in 24 hours were 2 . 5 fold and 1 . 9 fold higher in group B than in group A , respectively ( Table 2 ) . The mean duration of diarrhea was slightly higher in group B , 53 hours compared to 48 hours in group A . The mean maximum number of diarrhea stools in 24 hours was 10 in group B and 4 in group A ( Table 2 ) . Following the challenge , there were no serious unexpected adverse events ( S2 Table ) . The numbers of adverse events were slightly higher in group B than group A . Four of the 5 subjects with grade 2 or 3 vomiting were from group B . Six subjects from group B and two subjects from group A had moderate to severe nausea . One subject in each group had a grade 1 fever ( S2 Table ) . All but one of the 30 challenged subjects shed ETEC H10407 in their stool post challenge . The subject who did not shed was given the higher dose and this individual also did not have any gastrointestinal or systemic signs and symptoms associated with enteric illness . Five subjects did not start shedding the challenge organism until day 3 post challenge . Three of these subjects were in group B and two in group A . These 5 subjects also did not have any diarrhea throughout the study period . Overall the GM of H10407 ETEC shed on day 2 post challenge was the same ( 1 . 1x105 CFU ) in both the groups while the GM of the maximum shedding was ~2 fold higher in the group B ( 1 . 8 x 106 ) compared to group A ( 8 . 3 x 105 ) ( S3 Table ) To analyze shedding levels by clinical outcome post-challenge , subjects were divided into three groups: Group 1 included subjects with MSD ( n = 6 ) , Group 2 , subjects with mild diarrhea ( n = 2 ) and Group 3 , subjects with no diarrhea ( n = 22 ) ( S3 Table ) . The GM of H10407 ETEC shed on second day after challenge was 1 . 9x107 CFU in the subjects with MSD , 1 . 4x105 CFU in the subjects with mild diarrhea and 2 . 7x104 CFU in the subjects with no diarrhea ( p = 0 . 0360 , MSD vs mild + no diarrhea ) . The GM for maximum shedding was also significantly higher in the subjects with MSD ( p<0 . 0001 MSD vs mild + no diarrhea ) ( S3 Table ) . The systemic and mucosal antibody responses to the primary antigens ( CFA/I and LTB ) and “O” antigen of H10407 ( O78 ) were compared in the 15 subjects challenged with 105 dose ( group A ) and in the 15 subjects challenged with 106 dose ( group B ) . These comparisons were based on the frequency and magnitude of serum , ALS and fecal IgA antibody responses to the 3 antigens mentioned above . The highest response frequency for CFA/I IgA was found in ALS . The response frequencies in serum IgA and IgG were low and dose dependent , however , in ALS the frequency of responses to IgA were higher in group A compared to group B . ( Table 4 ) . For fecal extracts 4 ( 26 . 7% ) subjects in each group mounted anti-CFA/I IgA responses ( Table 4 ) . The magnitude of IgA and IgG responses in serum and IgA in ALS ( Fig 3 ) and fecal samples across both dosing groups were not significantly different . Interestingly , unlike other antigens , the serum GMT either plateaued ( IgA ) or continued to increase at day 84 ( IgG ) ( Fig 3A and 3B ) . The GMT of ALS IgA responses peaked at day 7 ( fold increase 2 . 9 , p = 0 . 0028 ) and remained high until day 9 , and dropped to baseline by day 28 in group A . However , the GMT also peaked at day 7 ( fold increase 2 . 5 ) in group B , but did not return to baseline levels until day 84 ( Fig 3C ) . The GMT for fecal extracts did not increase above baseline levels in either dosing groups ( S1A Fig ) . The frequency and magnitudes of anti-LTB antibody responses were lower than LPS but higher than CFA/I in both the groups . For serum , frequency of responses to anti-LTB specific IgA and IgG were similar in the two doses ( Table 5 ) . However , in ALS , similar to CFA/I , the frequency of responses to anti LTB IgA was higher in group A compared to group B . ( Table 5 ) . The GMT of the magnitude of responses to IgA also tended to be higher in group A compared to group B in both serum and ALS , but they were not significantly different ( Fig 4A , 4B and 4C ) . The serum IgA GMT of anti-LTB peaked at day 28 in group B ( 1 . 63 fold increase over baseline ) and at day 9 in group A ( 1 . 60 fold increase over baseline ) ( Fig 4A and 4B ) . The ALS anti-LTB response peaked at day 9 in group A ( 2 . 60 fold increase from baseline titer , p = 0 . 021 ) and returned to baseline by day 28 but again increased at day 84 ( 2 . 29 fold increase from day 28 p = 0 . 015 ) ( Fig 4C ) . Group B had similar kinetics with lower magnitude of response ( maximum fold increase 1 . 44 over baseline ) ( Fig 4C ) . The titers in the fecal samples were very low and were not different between the two groups ( S1B Fig ) . In general , in previous studies after challenge with ETEC , peak ALS IgA responses were most frequently observed at days 7 to 9 post-challenge [8 , 20 , 21 , 22] . In this study we looked at an earlier time point at day 4 for ALS samples to find any earlier responses . With the exception of very few subjects the earliest responses were on day 9 in these low dose groups ( Figs 2C , 3C and 4C ) . We compared immune responses in the subjects that had MSD , with subjects who had no diarrhea ( ND ) irrespective of the doses . There were 6 subjects with MSD , 22 with ND and 2 with mild diarrhea . Since there were only two subjects with mild diarrhea we have included those two subjects in the ND group for ease of analysis . All ( 100% ) of the MSD group responded to LPS IgA and IgG in serum . By contrast , only 10 ( 41 . 7% ) and 12 ( 50% ) subjects in the ND group mounted anti- IgA and IgG responses to LPS in serum respectively ( Table 6 ) . For ALS and fecal IgA responses , again more subjects responded in the MSD group than among the ND group ( Table 6 ) . The GMT of the anti-O78 serum IgA response in the MSD group was significantly higher at day 9 ( p = 0 . 044 ) , 28 ( p = 0 . 001 ) and 84 ( p = 0 . 0006 ) compared to the ND group ( Fig 5A ) . The GMT of IgG peaked on day 28 and increased 188 . 47 fold ( p = 0 . 002 ) in the subjects with MSD while increased only 6 . 76 fold ( p = 0 . 029 ) among the subjects with no or mild diarrhea ( Fig 5B ) . The GMT of the anti-O78 in ALS peaked at day 7 in both the groups however increased 3184 . 30 fold over baseline in the MSD group; while it increased only 42 . 14 fold over baseline in the ND group ( Fig 5C ) . In fecal extract the GMT of LPS IgA significantly increased to 10 . 20 fold over baseline at day 9 ( p = 0 . 0022 ) in the MSD group; while there was no increase in GMT titer in the ND group ( Fig 5D ) . The GMT of LPS IgA in fecal extracts was significantly higher in the MSD group compared to the ND group ( at day 9 , p = 0 . 0064 ) . Responses to CFA/I in serum were lower than responses to LPS . Serum , ALS and fecal responses to CFA/I were more common in the MSD group . ( Table 7 ) . The GMT of the serum IgA response in the MSD group increased 2 . 77 fold ( p<0 . 032 ) at day 28 from baseline and was significantly higher ( p<0 . 0004 ) than the GMT in the ND group ( Fig 6A ) . By contrast anti-CFA/I serum IgG GMTs were similar between the MSD and ND groups . Interestingly , the anti-CFA/I serum IgG response in subjects with MSD evolved more slowly over time , with the peak GMT being reached at day 84 post-challenge ( Fig 6B ) . For IgA ALS , the GMT in the MSD group increased 7 . 6 fold at day 7 ( p = 0 . 002 ) over baseline and , similar to LPS , stayed at the same level until day 9 , then decreased to baseline by day 28 . By contrast , for the ND group , the IgA ALS GMT only increased 2 . 1 fold ( p = 0 . 010 ) at day 7 compared to baseline ( Fig 6C ) . The GMT of anti-CFA/I IgA was also significantly higher ( at day 7 , p<0 . 0132 ) than that seen in the ND group ( Fig 6C ) . The GMT in fecal extracts never increased above the baseline in either group ( S2A Fig ) . The responses to LTB IgA and IgG in serum and ALS followed the same trend as LPS and CFA/I , the MSD group having higher responses than the ND group ( Table 8 ) . By contrast , the response rates to LTB IgA in fecal extracts were higher in the ND group compared to MSD group ( Table 8 ) . Serum IgA GMT of LTB in the MSD group peaked at day 28 with 2 . 89 fold increase from baseline; the magnitude was higher than seen in the ND group . The serum IgG GMT peaked at day 28 in the MSD group , increased 18 . 98 fold ( p = 0 . 0022 ) from baseline while the GMT only increased 2 . 39 fold in the ND group from baseline ( Fig 7A and 7B ) . The magnitude of ALS IgA responses to LTB was low and similar in both the groups ( Fig 7C ) . The GMT of IgA in ALS peaked at day 9 , and increased only 1 . 69 fold in the MSD group and 2 . 02 fold ( p = 0 . 033 ) in the ND group from baseline ( Fig 7C ) . In the ND group the GMT of IgA responses in ALS was bimodal and increased again at day 84 ( GMT of 2 . 0 fold increase from day 28 to 84 , p = 0 . 038 ) ( Fig 7C ) similar to the dose dependent response ( Fig 7C ) . The GMT in fecal samples never increased above the baseline in both the groups ( S2B Fig ) . We also analyzed the baseline GMTs ( day before challenge ) of all the antigens in serum , ALS and fecal extracts in the MSD and ND groups . The titers were similar in the two groups except the LTB IgG in serum was significantly higher in the ND group compared to the MSD group ( p = 0 . 0211 ) ( Table 9 ) . The CFA/I IgG titer in serum was higher in the ND group but was not significantly different from the MSD group .
In this study , using the same buffering and fasting procedures applied in our prior challenge model refinement efforts [17] we found that the diarrhea attack rate was lower when H10407 was given to subjects at doses of 105 ( 20% ) or 106 ( 33% ) than the ~70% attack rate for MSD we observed in previous studies when this strain was given at a dose of 107CFUs [17 , 20] . We have used an attack rate for MSD of 50–70% as a reasonable barometer in our efforts to better optimize the ETEC human challenge model for use in evaluating preventive and therapeutic interventions for ETEC . Based on our observation in this most recent model refinement study , we conclude that a dose of 107 H10407 remains the lowest practical dose for use in future volunteer studies evaluating the efficacy of candidate vaccines and other potential preventive tools targeting ETEC . Nevertheless , this study did demonstrate that lower H10407 doses ( 105 or 106 ) were capable of causing MSD in some individuals . Although the mean incubation periods were similar for both the doses , the stool output and number of subjects with MSD were higher in 106 dose . There were no unexpected serious adverse events in either group through the 180 day follow-up period . Solicited gastrointestinal and systemic signs and symptoms were only marginally higher in the 106 group overall . Five subjects had vomiting and two had a mild fever . In the study by Evans et al in 1978 , [23] subjects were challenged with H10407 , at a dose of 106 after fasting for 5 hours , and there was no discernible clinical diarrhea reported in these subjects following dosing . In the present study we found some severe cases at this dose as well as at the lower 105 dose suggesting that the overnight fast used in this study may have facilitated ETEC colonization and the subsequent development of illness . However , to determine the specific effect of the length of fasting period on clinical illness , a direct comparison of longer and shorter fasting periods is necessary . In our recent study , we have noted that an increased fasting time from 90 minutes before challenge to overnight can increase the diarrhea attack rate for another ETEC challenge strain , B7A ( unpublished ) . All the subjects except one shed the challenge strain even in this low dose challenge study . Although the overall shedding levels on day 2 were the same in both groups , the GM of the maximum shedding was higher in the subjects who received the higher dose . In this low dose study we found that 1 . 9x107 CFU/gm of stool on day 2 after challenge corresponds to the development of MSD . A similar correlation between a threshold level of shedding and a higher risk for MSD was shown by us using both quantitative stool culture and q-PCR in a previous study [18] . Similarly , in our previous ETEC vaccine immunization and challenge study using a higher challenge dose of 107 CFU of H10407 , we also found 1 . 5x107 CFU corresponds to MSD [20] . The association between level of colonization and diarrhea risk is also supported by several recent field studies in which quantitative TaqMan PCR has shown a similar threshold effect [4] . Not surprisingly , subjects challenged with H10407 seroconverted and had mucosal responses to known ETEC virulence antigens LPS , LTB and CFA/I . However , the frequency and magnitude of responses tended to be higher in subjects challenged with the 106 dose compared to the 105 dose suggesting a dose-dependent trend . The differences in response to these two doses varied depending on the antigen and sample types . Interestingly , responses to LTB were higher in the lower dose group compared to the higher dose group in all types of samples . In this low dose study the highest IgA and IgG responses to LPS in serum , ALS and fecal samples were generally at day 9 . For LTB and CFA/I IgA and IgG , the peak responses were either on day 9 or day 28 . Similar to our previous findings [8] with higher H10407 doses , the frequency and magnitude of responses to CFA/I were lower compared to LPS and LTB in all the samples . It is interesting to note that the responses to CFA/I IgG in serum peaked on day 7 and kept on increasing until day 84 . The magnitudes of the IgA responses to LTB and CFA/I in fecal samples were negligible in both dose groups . In general , immune responses induced by the lower doses used in this study were more modest than those seen earlier when a higher dose of H10407 ( ~107 CFU ) was used . A more detailed comparison of the humoral and cellular immune responses induced across the full spectrum of H10407 challenges doses studies by our investigative team in recent years ( 108 to 105 CFU of H10407 ) is currently underway . There is a growing concern of asymptomatic colonization of enteropathogens in the gut of the children in the endemic countries because of its impact on longitudinal public health issues , such as growth faltering , oral vaccine low efficacy and poor neurocognitive development through environmental enteropathy/environmental enteric dysfunction [24] . Asymptomatic ETEC infections are common in endemic countries and may impact the intestinal health as well as long-term developmental potential in the children in these areas [4 , 25] . This low dose CHIM model of ETEC may serve as a tool to decipher these critical interactions . Furthermore , the levels of immunity induced by these asymptomatic ETEC infections is not well studied and needs to be better understood since these may contribute to protection in the field and may serve to diminish vaccine take . In this study we compared antigen-specific ETEC immune response profiles in the subjects with clinical and subclinical infections after challenge with low doses of ETEC which may better mimic more natural levels of exposure in the ETEC endemic areas . Notably , although of lower magnitude , there were considerable number of responders and increased titers in the subjects with no diarrhea . In this study we found that the magnitude of serum and mucosal antibody responses after challenge were symptom dependent with subjects developing MSD generally showing stronger responses than those experiencing asymptomatic colonization irrespective of the ETEC challenge dose . Of note , the challenge bacteria were shed by all volunteers ( except one ) but the shedding level was related to the dose of the challenge and the severity of the symptoms . Thus , the magnitude of immune responses depended on the colonization levels , with the one exception being the responses to LTB where the responses were similar between those with MSD and those with ND . In natural infection it is often assumed that some degree of pre-existing immunity to a given pathogen may contribute to an individual’s susceptibility or resistance to infection and subsequent illness . In the present study we compared baseline serum and mucosal antigen-specific antibody levels to see if they are predictive of challenge outcome . Of the antigens evaluated , only LTB IgG levels in serum were significantly higher in subjects with no post-challenge diarrhea compared to subjects with MSD . A similar relationship has been previously noted by Porter and colleagues in their recent systematic review of ETEC CHIMs studies [10] . However , further investigation is required to establish if LTB IgG levels in serum can serve as an indication of immunity to ETEC . It is noteworthy that in our study , subjects with no history of exposure to ETEC and/or cholera ( >2 years ) given lower ETEC doses responded with very different clinical outcomes . Among those few subjects developing MSD , the incubation time was longer than has been generally seen at higher doses ( ~60 hrs versus 48 hrs ) [10 , 17] but the threshold level of shedding seen in these subjects was comparable to that previously associated with diarrheal illness in H10407 studies ( ≥107CFU per gram of stool ) [18 , 20]; while subjects with mild or no illness exhibited much lower levels of ETEC shedding ( approximately 104 to 105 CFU per gram of stool ) . The cellular and molecular basis for these differences deserve further investigation using new genomic , transcriptomic , proteomic , systems biology and microbiome tools . These studies are now underway [26 , 27] in the hope that they will lead to better defining host factors ( immunological or microbiological ) that may contribute to ETEC susceptibility and immunity . | Human challenge models can provide a platform for the initial evaluation of vaccine efficacy to down-select vaccine candidates before more expensive phase III trials are conducted . An ideal challenge model would need to reproducibly induce diarrhea in a high proportion of volunteers using the lowest possible inoculum to better maximize the sensitivity of the model . In this study , we attempted to refine the ETEC H10407 challenge model by characterizing the clinical illness induced by lower ETEC challenge doses . We also performed a comprehensive dose and clinical diarrhea severity dependent assessment of the immune responses . We found that some volunteers do become ill when challenged with these low doses , but the rates of moderate to severe diarrhea observed with these doses were too low to be useful for such volunteer studies . Therefore , we conclude a dose of 2x107 H10407 remains the lowest practical dose when evaluating candidate vaccines or other preventive or therapeutic ETEC interventions . Diarrheal illness , when it occurred , was associated with higher fecal shedding levels of H10407 . The magnitudes of immune responses were dependent on the dose , antigen and severity of diarrhea . This new knowledge may contribute to a better overall understanding of the immune responses to ETEC and its relation to diarrhea severity which will help facilitate and guide ETEC vaccine development strategies . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"immunology",
"vaccines",
"diarrhea",
"physiological",
"processes",
"preventive",
"medicine",
"signs",
"and",
"symptoms",
"gastr... | 2018 | Impact of lower challenge doses of enterotoxigenic Escherichia coli on clinical outcome, intestinal colonization and immune responses in adult volunteers |
Loss of retinoblastoma ( Rb ) tumor suppressor function is associated with human malignancies . Molecular and genetic mechanisms responsible for tumorigenic Rb downregulation are not fully defined . Through a forward genetic screen and positional cloning , we identified and characterized a zebrafish ubiquitin specific peptidase 39 ( usp39 ) mutation , the yeast and human homolog of which encodes a component of RNA splicing machinery . Zebrafish usp39 mutants exhibit microcephaly and adenohypophyseal cell lineage expansion without apparent changes in major hypothalamic hormonal and regulatory signals . Gene expression profiling of usp39 mutants revealed decreased rb1 and increased e2f4 , rbl2 ( p130 ) , and cdkn1a ( p21 ) expression . Rb1 mRNA overexpression , or antisense morpholino knockdown of e2f4 , partially reversed embryonic pituitary expansion in usp39 mutants . Analysis of pre-mRNA splicing status of critical cell cycle regulators showed misspliced Rb1 pre-mRNA resulting in a premature stop codon . These studies unravel a novel mechanism for rb1 regulation by a neuronal mRNA splicing factor , usp39 . Zebrafish usp39 regulates embryonic pituitary homeostasis by targeting rb1 and e2f4 expression , respectively , contributing to increased adenohypophyseal sensitivity to these altered cell cycle regulators . These results provide a mechanism for dysregulated rb1 and e2f4 pathways that may result in pituitary tumorigenesis .
The hypothalamic-pituitary axis regulates stress responses , growth , reproduction and energy homeostasis . Neuropeptides released from the hypothalamus via the hypophyseal portal plexus control synthesis and secretion of anterior pituitary hormones [1] . Different pituitary cell types secrete hormones that regulate post-natal growth ( growth hormone , GH ) , lactation ( prolactin , PRL ) , metabolism ( thyroid stimulating hormone , TSH ) , stress ( adrenocorticotrophic hormone , ACTH ) , pigmentation ( melanocyte-stimulating hormone , αMSH ) , sexual development and reproduction ( luteinizing hormone , LHβ , and follicle stimulating hormone , FSHβ ) [2] . Corticotropes and melanotropes produce proopiomelanocortin ( POMC ) , which is proteolytically cleaved to give rise to ACTH in corticotropes and αMSH in melanotropes . Central and peripheral signals including hypothalamic stimulatory hormones , growth factors and estrogen cause pituitary hyperplasia , genetic instability , subsequent monoclonal growth expansion and tumor formation [3] . Pituitary tumors are almost invariably benign , however if untreated , they are associated with increased morbidity and mortality due to tumor mass effect and/or hormonal disruptions leading to serious complications such as acromegaly and Cushing's disease [4] , [5] . How developmental or acquired signals elicit plastic change in pituitary cell growth resulting in hyperplasia or benign adenomas is not fully understood [6] . The pituitary gland is highly sensitive to cell cycle regulators including cyclins , cyclin dependent kinases ( CDKs ) , CDK inhibitors ( CKIs ) and retinoblastoma protein ( pRB ) , all of which are frequently dysregulated in pituitary tumors . pRB , a nuclear pocket protein , binds the E2F transcription factors and regulates the balance between cell quiescence and proliferation [7] . E2Fs control expression of genes crucial for cell cycle re-entry , DNA replication and mitosis . Dephosphorylated pRB binds to E2Fs and inhibits transcription of E2F target genes either by sequestration and inhibition of E2F cell cycle “activators” ( E2F1–E2F3 ) , or by formation of pocket protein complexes with “inhibitors” ( E2F4–E2F8 ) , which bind to E2F-responsive promoters and repress their transcription [7] . Accordingly , transcriptional repression of pRB activity prevents G1/S progression and promotes cell quiescence . In mice , Rb heterozygous mutations lead to early onset and increased incidence of endocrine neoplasma including pituitary , thyroid and adrenal tumors [8] , [9] . The 100% penetrance of pituitary tumors in Rb+/− mice [8] is partially reversed in Rb+/−; E2f4−/− double mutants , implicating the Rb/E2f4 pathway in pituitary tumorigenesis and also suggesting an E2F4 oncogenic activity [9] . E2F4 is also known as a key regulator associated with p130 in G0/G1 to promote quiescent G0 and terminal differentiation [10] , [11] . E2f4 null mice often die shortly after birth with defects of terminal differentiation resulting from an inability to establish cell cycle quiescence [12] . In response to cell cycle re-entry , E2F4 switches from p130 [10] , [13] to pRB [10] , [14] and p107 [10] , [14] , [15] , which inhibit E2F4 transactivation . Additionally , E2F4 overexpression has been shown to promotes cell proliferation and transformation [14] , [15] , which prevents growth arrest mediated by p130 [13] . Pituitary development and physiology are conserved in zebrafish [2] . Novel insights into developmental mechanisms have been obtained by in vivo analysis of transgenic zebrafish expressing GFP and RFP driven by regulatory elements of zebrafish pomc [16] and prl [17] , respectively . Through a forward genetic screen for novel zebrafish genes regulating adenohypophyseal pomc gene expression , we identified and characterized a mutant that harbors a nonsense mutation in usp39 , leading to expansion of all adenohypophyseal cell lineages . Usp39 encodes a conserved protein termed Sad1p in Saccharomyces cerevisiae and a 65 kDa ( 65K ) SR-related protein in humans [18] , [19] . Both yeast Sad1p and the 65K SR-related protein in humans are involved in assembly of the spliceosome , the RNA splicing machinery [18] , [19] . RNA splicing is crucial for eukaryotic gene expression and defective splicing can be detrimental since it leads to an altered genetic message [20] . The spliceosome consists of five small nuclear ribonucleoproteins ( snRNPs ) , U1 , U2 , U4 , U5 and U6 as well as a large number of non-snRNP proteins [20] . The yeast Sad1p is involved in splicing in vivo and in vitro and in the assembly of U4 snRNP to U6 snRNP [18] , while human 65K SR-related protein is essential for recruitment of the tri-snRNP to the pre-spliceosome and is known as a tri-snRNP-specific protein [19] , [21] . Additionally , Usp39 is also classified as a deubiquitinating enzyme but lacks protease activity due to absence of key active-site residues of cysteine and histidine [18] , [19] , [22] . In the present study , we aimed to define novel pathways regulating pituitary development through study of an usp39 mutation . Using microarray gene expression profiling followed by quantitative real time-polymerase chain reaction ( RT-PCR ) validation we observed a significant reduction of rb1 expression and increased e2f4 , rbl2 ( p130 ) and cdkn1a ( p21 ) expression in mutants . Zebrafish usp39 is predominantly expressed in the brain and represents a novel neuronal splicing factor . We show that zebrafish usp39 mutation leads to an rb1 splicing defect responsible for pituitary expansion . In addition , knockdown of e2f4 partially rescued pomc lineage expansion in usp39 mutants . Our finding that usp39 regulates expansion of all embryonic pituitary cell lineages through the rb1/e2f4 pathway may shed light on mechanisms underlying adult pituitary tumor formation .
To isolate genes required for adenohypophysis and hypothalamic development a standard forward genetics method was carried out using a three-generation ( F3 ) screen after mutagenesis with ENU , which mostly induces single nucleotide exchanges at random positions of the genome [23]–[25] . The genetic screen was performed using pomc expression as a specific marker . pomc is expressed in subepithelial pituitary cells , dorsal to the oral ectoderm roof and ventral to the ventral diencephalon . A subset of pomc-expressing cells is also located outside the adenohypophysis , corresponding to β-endorphin-synthesizing cells of the hypothalamic arcuate nucleus [2] . In zebrafish , spatial distribution of the six different hormone secreting pituitary cell types is subdivided into three regions along the antero-posterior adenohypophyseal axis of the rostral pars distalis , proximal pars distalis and pars intermedia [2] . pomc is expressed in corticotropes of the rostral pars distalis , in melanotropes of the pars intermedia and in the hypothalamus ( Figure 1A ) . We isolated an ENU-induced mutant , hp689 , which was characterized by reduced hypothalamic but increased pituitary pomc expression at 48-hours post fertilization ( hpf ) ( Figure 1B ) . In addition , hp689 mutants displayed microcephaly and smaller eyes starting at 33 hpf ( data not shown ) . Using segregation linkage analysis , the hp689 locus was mapped to zebrafish linkage group 5 with a critical interval of 0 . 03 centimorgan ( cM ) on marker ndrg3 ( Figure 1C , see Materials and Methods ) . This region contained 7 annotated genes and sequencing of usp39 from mutant embryos revealed a point mutation that converted a TAT codon into a TAA in exon 11 , resulting in a premature termination codon rather than a tyrosine amino acid ( Figure 1D ) . PROSITE database search of the Usp39 protein revealed two domains consisting of a zinc finger ( ZF_UBP ) and a Ubiquitin carboxyl-terminal hydrolases family 2 ( UCH_2_3 ) region [26] . As a result of the UCH_2_3 domain , Usp39 is classified as a deubiquitinating enzyme . However , it lacks protease activity due to the absence of key active site cysteine and histidine residues [18] , [19] , [22] . The single allele of hp689 carries a nonsense mutation within the UCH_2_3 region , resulting in a truncated Usp39 protein lacking amino acids after position 412 ( Figure 1D ) . To confirm that hp689 represents the usp39 mutation , a usp39 antisense morpholino ( MO ) oligonucleotide targeting the usp39 start codon and consequentially blocking translation was injected into wild-type ( wt ) embryos [27] . MO injected embryos showed increased pomc expression , similar to the hp689 phenotype ( Figure 1F compared to Figure 1B ) . Furthermore , injection of mRNA encoding wild-type usp39 rescued the mutant phenotype ( Table 1 ) , indicating that pituitary pomc upregulation in usp39 mutants results from the nonsense mutation in zebrafish usp39 . RNA whole-mount in situ hybridization was performed to determine the spatiotemporal expression pattern of usp39 during zebrafish development . Generally weak usp39 expression was detected in early cleavage embryos ( data not shown ) but tissue specific expression peaked by 36 hpf and decreased by 42 hpf ( Figure 2A–2C ) . Expression was detected predominantly within the brain , including the pituitary region and eyes . At 21 . 5 hpf , there was also expression in the intermediate cell mass , the site of embryonic zebrafish hematopoiesis ( Figure 2A , inset ) . However , mutants showed persistently lower usp39 expression that was completely lost by 42 hpf ( Figure 2D ) . In addition , loss of usp39 expression by 42 hpf corresponds to the time at which the usp39 mutant embryos fully develop a phenotype of microcephaly , smaller eyes and a pituitary abnormality , indicating the critical time point when usp39 is required for normal development . The zebrafish adenohypophysis consists of six different hormone-secreting cell types distributed along the anterior-posterior axis: lactotropes and corticotropes are located anteriorly in the rostral pars distalis , thyrotropes , gonadotropes and somatotropes are found medially in the proximal pars distalis whereas melanotropes are situated posteriorly in the pars intermedia ( Figure 3 ) . To determine if additional pituitary lineages are affected by the usp39 mutation , we performed double color RNA in situ hybridization analysis with combinatory pituitary lineage specific marker genes pomc , gh , prl , tsh , and with cga that encodes the glycoprotein α-subunit heterodimerizing with TSHβ , LHβ , or FSHβ subunit [2] , [16] . This analysis revealed expansion of all the analyzed cell lineages without apparent cell fate transformation in the usp39 mutant pituitary at 48 hpf ( Figure 3A–3P ) . Cell expansion was most marked in corticotropes and lactotropes , indicating that usp39 is important for regulating embryonic pituitary cell populations ( Figure 3B , 3D , 3F , 3H , 3J , and 3L ) . We examined expression of hypothalamic regulators to investigate whether pituitary lineage expansion in usp39 mutants is due to altered hypothalamic neuroendocrine input to the adenohypophysis . One of the primary hypothalamic inhibitory mechanisms controlling pituitary homeostasis is dopamine ( DA ) released from tuberoinfundibular neurons ( TIDA ) . Pituitary lactotrophs are almost exclusively regulated by tonic inhibition of dopamine , which inhibits lactotroph proliferation , PRL gene expression and secretion by activating D2 dopamine receptor subtype ( Drd2 ) [28] . We therefore processed 48 hpf whole-mount embryos for immunocytochemistry using an antibody against tyrosine-hydroxylase , the rate-limiting enzyme of dopamine synthesis in TIDA neurons , and detected no significant change of hypothalamic dopaminergic neurons in usp39 mutants compared with wt siblings ( Figure 4D and 4H ) . Corticotropin releasing hormone ( CRH ) as well as gonadotropin-releasing hormone ( GnRH ) stimulates cell growth , hormone synthesis and secretion of pituitary corticotropes and gonadotropes , respectively [29] , [30] . However , usp39 mutants exhibit no altered crh or gnrh expression ( Figure 4E–4G ) . Therefore , pituitary lineage expansion of usp39 mutants occurs independently of major hypothalamic neuroendocrine signals . We next studied expression of transcription factors important for adenohypophyseal development . Lim3/Lhx3 is one of the earliest pituitary specifying transcription factors and is required for progenitor proliferation and survival [2] . Pitx3 , a Pitx/Rieg homeodomain protein , defines the pituitary placode and is required for Lim3 expression [2] . Pit1 is a Pou domain homeoprotein and a lineage-determining factor for somatotropes , lactotropes and thyrotropes [2] . The Drosophila eye absent homolog , eya1 , is required for specification of gonadotropes , corticotropes , and melanotropes [2] . Zebrafish mutation of ascl1a , a homolog of the Drosophila MASH1 [31] , resulted in failed endocrine differentiation of all adenohypohyseal cell types [2] . Expression of these zebrafish pituitary regulators coincides within the pituitary placode of the anterior neural ridge ( ANR ) at 20-somite stage ( 18 hpf ) and persists in the adenohypophyseal anlage throughout 48 hpf ( for eya1 ) , or even later ( for pit1 , lim3 , pitx3 , and ascl1a ) . At 36 hpf , usp39 mutants exhibited no increased expression of lim3 , however pit1 , lim3 , pitx3 , eya1 and ascl1a showed a significant expression difference at a later state ( 48 hpf ) compared with wt ( Figure S1 , Figure 5 ) . These results suggest that the usp39 mutation did not affect initial embryonic pituitary progenitor specification but induced their expansion after 36 hpf . To distinguish whether the altered pituitary signals detected by whole-mount in situ hybridization is due to pituitary hyperplasia or higher expression of pituitary hormone levels we crossed usp39 +/− fish to POMC-GFP transgenic fish [16] . After identifying mutant usp39 in the POMC-GFP background , we sectioned usp39 and wt whole embryos and performed immunocytochemistry with anti-GFP followed by cell number quantification . Our results demonstrated that there was an increase in the number of POMC-GFP-positive cells in usp39 mutants compare to wt ( Figure S2A–S2F ) . In addition , counting the nuclei stained with DAPI indicated an increase in the total number of pituitary cells in the usp39 mutant compare to wt embryos ( Figure S2G ) . We carried out a BrdU incorporation study and demonstrated that the increase in pituitary cell number seen in usp39 mutants was due to an increase in proliferation ( Figure S2H and S2I ) . It is well established that cell cycle dysregulation is associated with pituitary pathology in animal models and human disease [32] . However , little is known about mechanisms underlying the sensitivity of differentiated pituitary cell lineages to cell cycle regulators . We therefore performed a microarray analysis and focused on cell cycle regulators , 11 of which were confirmed for altered expression in usp39 mutants by quantitative RT-PCR . In summary , expression levels of 3 genes were increased including e2f4 , rbl2 ( p130 ) and cdkn1a ( p21 ) and expression of the other 8 cell cycle genes including rb1 were downregulated ( Table 2; Figure S3 ) . We further examined e2f4 expression by RNA whole-mount in situ hybridization , which confirmed its upregulation in the adenohypophysis of usp39 mutants compared to wt embryos ( Figure 6A and 6B ) . To investigate whether e2f4 upregulation is responsible for adenohypophysis lineage expansion , we injected embryos with antisense MO oligonucleotide to knockdown e2f4 function . The overall usp39 mutant phenotype maintained after e2f4 MO injections , which resulted in partial reversal of pomc expansion in e2f4-MO-injected usp39 embryos at 48 hpf compared to control embryos ( Figure 6C–6E , mutant N = 20 , ∼60% showed rescue ) . The phenotypic rescue of usp39 embryos by e2f4-MO is pituitary specific since pomc hypothalamic expression was not altered . In addition , we analyzed prl expression and a partial rescue was also observed in e2f4-MO-injected usp39 embryos ( Figure S4 ) . These results indicate that loss of usp39 results in increased e2f4 expression , which at least partially contributes to the observed pomc lineage expansion in zebrafish adenohypophysis . Since usp39 is known to be an essential component of the RNA splicing machinery , the more than 70% decrease of rb1 expression in usp39 mutants may be attributed to defects in RNA splicing . We therefore examined rb1 splicing status by PCR amplification using primers corresponding to each end of 19 out of 27 exons of the rb1 gene . Primers designed for exon 3 and exon 4 resulted in a PCR product of 250 base pairs ( bp ) in wt and mutant embryos , representing a correctly spliced mRNA fragment . However , mutant embryos exhibited an additional larger PCR product of 343 bp ( Figure 6F ) . Further DNA sequence analysis revealed that the 343 bp PCR product derived from usp39 mutants contain the sequence of an unspliced intron between exon 3 and 4 . The splicing defect would lead to a premature stop codon in the intron between exon 3 and exon 4 of rb1 ( data not show ) , which would lead to nonsense-mediated mRNA decay . We then performed an rb1 mRNA overexpression experiment in usp39 mutants and observed partial rescue of the adenohypophysis phenotype ( Figure 6I , mutant N = 34 , ∼50% showed rescue ) , validating the importance of usp39-mediated rb1 mRNA splicing in controlling pituitary lineage expansion during development . In addition , we performed quantitative RT-PCR analysis on the rb1 mRNA-injected usp39 embryos and observed a 30% reduction of e2f4 expression compared to control uninjected usp39 mutants ( Figure S5A ) , indicating that e2f4 upregulation in usp39 mutant is secondary to Rb1 loss of function . Futhermore , this was confirmed by quantitative RT-PCR analysis on the e2f4 MO-injected usp39 embryos and observed that there was no change in rb1 expression compared to control uninjected usp39 mutants ( Figure S5B ) .
In this study , we identified and functionally characterized the zebrafish usp39 gene , important for human and yeast pre-mRNA splicing [18] , [19] . We demonstrated that loss of usp39 results in defects in rb1 mRNA splicing and downregulation of rb1 expression ( Figure 6F , Table 2; Figure S3 ) . Both loss of usp39 as well as rb1 downregulation in usp39 mutants may be explained by nonsense-mediated mRNA decay due to a premature termination codon . RB1 gene mutation leading to pre-mRNA splicing defects have been shown in human cancers [33] and our study suggests a novel mechanism resulting in rb1 splicing defects due to a usp39 mutation . Control of pituitary progenitor cell proliferation in concert with terminal differentiation during embryonic pituitary development is poorly understood . In mouse pituitary primordia , attenuated proliferation of cells destined to become hormone-expressing cell types occurs days before lineage-specific hormones start to express [34] . In contrast , zebrafish pituitary terminal differentiation is initiated while progenitor cells are still organized in a placodal fashion in the anterior neural ridge [2] , [16] . The usp39 mutants demonstrated no early difference of adenohypophyseal primordia compared with wt , until 48 hpf when terminally differentiated cells had already migrated to a mature pituitary destination . Pituitary lineage expansion became apparent at 48 hpf in usp39 mutants , as indicated by expression of pituitary transcription factors and lineage-specific hormone markers ( Figure 3 and Figure 5 ) . Similarly , it was found that inactivation of Rb in the small intestines of mice results in increased proliferation of differentiated cells in the villus but not in the stem cells located in the base of the crypts [35] . Therefore , our results suggest that loss of usp39 does not affect pituitary specification , initiation and early differentiation , but does induce lineage expansion at later development stages when the cells are terminally differentiated . Our results indicate that e2f4 overexpression has at least a partial but direct affect on adenohypophyseal cell lineages in usp39 mutants , as e2f4 antisense MO knockdown partially reverted the pituitary phenotype of usp39 mutants ( Figure 6 and Figure S4 ) . The usp39 mutants demonstrated persistently upregulated e2f4 expression , although molecular mechanisms leading to e2f4 overexpression remain to be determined . Overexpression of e2f4 may exert oncogenic activity promoting cell-cycle progression as previously indicated in pituitary , thyroid , lung neuroendocrine hyperplasia [36] , intestinal crypt cells , colorectal cancer cells [37] as well as in prostate cancer [38] . We demonstrated an increase of POMC-GFP-positive cells in the usp39 mutant embryos compare to wt ( Figure S2 ) . Consequently , e2f4 upregulation in usp39 mutants may contribute to increased proliferation of terminally differentiated pituitary cells leading to lineage expansion as seen in our BrdU studies ( Figure S2 ) . On the other hand , E2F4 is a key regulator associated with p130 to promote quiescent G0 and terminal differentiation [13] . The cyclin kinase inhibitor , p21 , inhibits decay of the E2F4-p130 complex , promotes senescence and restrains growth , contributing to the benign propensity of pituitary adenomas [39] , [40] . Our microarray and quantitative RT-PCR data showed increased cdkn1a ( p21 ) , e2f4 and rbl2 ( p130 ) , which may indicate an enhanced quiescent G0 phase inducing terminal differentiation and lineage expansion in usp39 mutants . Although the focus of this study was the role of usp39 in pituitary development , this gene is also expressed in neuronal tissues and when mutated , embryos show microcephaly and smaller eyes , therefore usp39 function may not be restricted to pituitary development . We propose that usp39 , through targeting a set of key regulatory genes by modulating RNA splicing , should have a broader role in regulating neuronal cell lineage development . Although how usp39 controls target mRNA splicing remains to be fully elucidated , the usp39 ortholog of the yeast protein Sad1p was found to have two roles: it is involved in the assembly of U4 snRNP to U6 snRNP and is also required for splicing [18] . Furthermore , previous reports have shown that a zebrafish RNA splicing factor , p110 , is required for U4 and U6 snRNPs recycling , and a mutation in p110 leads to thymic hypoplasia as well as eye and exocrine pancreas defects [41] . In addition , microarray analysis of p110 mutant shows a compensatory mechanism inducing increased expression of other splicing factors , which may reverse the recycling defects [41] . We observed a similar result in our microarray analysis with upregulation of other U4/U6 . U5 tri-snRNP proteins , which suggests a compensatory mechanism in usp39 mutants ( Table S2 and Figure 6F ) . The human tri-snRNP specific proteins include 65K , 110K and 27K are encoded by USP39 , SART1 and SNRNP27 , respectively and play a similar role in splicing [21] . Specifically , both sart1 and snrnp27 were found to be upregulated in our microarray analysis demonstrating a compensatory role due to the absence of usp39 ( Table S2 ) . Additionally , we discovered another neuronal gene , otx2 , which was also significantly downregulated due to a splicing defect ( Figure S6 ) . However , otx2 mRNA overexpression in usp39 mutant embryos did not rescue the pituitary phenotype ( data not shown ) , validating that the Rb1/E2F4 pathway is more specific to pituitary regulation . A systematic analysis of splicing variation of all mRNA transcripts affected by usp39 deficiency will uncover additional pathways that control neuronal and organ development by RNA splicing mechanisms . In summary , our findings indicate that usp39 plays an important role in pituitary development by regulating rb1 and e2f4 . Loss of usp39 leads to pituitary cell lineage expansion through rb1 downregulation due to a splicing defect . In addition , e2f4 overexpression contributes to increased pituitary cell mass , likely as a result of increased terminal differentiation or proliferation . Our studies reveal a novel role of usp39-mediated mRNA splicing of rb1 in pituitary cell growth control , which is critical for maintaining embryonic pituitary homeostasis .
Mutagenesis with ENU was performed as described [25] . Mutants including hp689 were identified from random sibling crossing from F2 families giving rise to 25% altered pomc expression at 48 hpf . Linkage analysis was established by mating hp689 heterozygote in an AB background to the WIK strain . Random sibling crossing identified F1 carriers , and mutants were identified phenotypically in F2 offspring at 48 hpf . We analyzed linkage between hp689 and simple sequence-length polymorphism markers [42] . Linkage analysis found the z34450 marker located 2 . 4 cM ( 4 recombinations in 168 meiosis ) , G40879 marker located 0 . 3 cM ( 1 recombination in 288 meioses ) , ephb4a located 2 . 2 cM ( 5 recombinations in 114 meioses ) and marker ndrg3 located 0 . 3cM ( 1 recombination in 310 meioses ) in linkage group 5 ( LG5 ) linked to the mutation . Total RNA derived from 48 hpf mutant and wild-type embryos was prepared by TRIZOL ( Invitrogen ) reagent extraction and used to generate cDNA by SuperscriptII reverse transcriptase ( Invitrogen ) with oligoDT primers ( Roche Applied Science ) . The region near marker ndrg3 contain 7 genes and sequencing the usp39 full-length cDNA with primers CGCGTTCACAGTGCGTTC and TTTCTCATTGTGTGTTTTACTCAGTC from mutant embryos revealed a point mutation that converted a TAT codon into a TAA in Exon 11 , resulting in a premature termination codon rather than a tyrosine residue . The usp39 full-length cDNA fragment was generated from wild-type embryos as described above and subcloned into pCRII-TOPO . Antisense MO'swere injected into embryos as described [27] . The sequence of usp39 MO is 5′-TTCACGCCTCTGATCATATTTTAAG-3′ and for e2f4 MO is 5′-ACTCTCCCATCGCTCCCAGGTCGTT-3′ ( Gene Tools , Inc ) . One- to two-cell stage embryo was injected at 3 . 9 ng for the usp39 MO and 1 . 4 ng of e2f4 MO . The usp39 overexpression construct was generated by subcloning full-length usp39 cDNA from vector pCRII-usp39 into the EcoRI site of vector pXT7 . The pXT7-usp39 vector was linearized with XbaI and mRNA was transcribed using T7 mMessage mMachine kit ( Ambion ) . The rb1 overexpression construct was generated by subcloning full-length rb1 cDNA ( Accession Number: BC125966 , Openbiosystems ) to vector pCS2+ in the StuI and XhoI sites . The rb1-pCS2+ vector was then linearized with XbaI and the mRNA transcribed using Sp6 mMessage mMachine kit ( Ambion ) . mRNA injections were performed at the one-cell stage at approximately 200 pg for usp39 and 267 pg for rb1 . Single and double whole-mount in situ hybridizations were performed as described [43] . usp39 antisense probe was synthesized from pCRII-usp39 with Sp6 RNA polymerase after linearization with NotI . The following riboprobes were generated from cDNAs as described: pomc [16] , gh , prl , tshβ , lim3 , pit1 and pitx3 [2] , eya1 [44] , crh [29] , gnrh2 and gnrh3 [30] , and ascl1a [31] . Full-length cDNA for cga ( Accession Number: BC116611 ) and e2f4 ( Accession Number: BC056832 ) were purchased from Openbiosystems . e2f4 was subcloned to pCRII-TOPO vector and linearized with SpeI whereas the cga-Express1 was linearized with EcoRI and riboprobes were synthesized with T7 RNA polymerase . Whole-mount antibody staining was performed using a rabbit anti-tyrosine hydroxylase ( TH ) primary antibody at 1∶200 dilution ( Chemicon ) and detected with an Alexa ( A594 ) -conjugated goat anti-rabbit secondary antibody at 1∶200 dilution ( Molecular Probes ) . Total RNA from 48 hpf usp39 mutants and wild-type embryos was prepared and microarray performed as described [45] . cDNA was generated as described above . RT-PCR was performed using the iCycler iQ Real-Time PCR Detection System ( BioRad ) and the iQ SYBR Green Supermix ( Biorad ) . Relative cDNA amounts were calculated using the iCycler program ( BioRad ) and gene expression levels measured by the 2−ΔΔCT method [46] , comparing usp39 mutant embryos to WT controls , with β-actin used as the reference gene . This procedure was repeated three times for each gene with three different experimental cDNA pools . At least three replicates were used for each cDNA pool . Gene expression was reported as relative expression change in usp39 mutants over WT embryos ± standard error ( for primer sequences see Table S1 ) . cDNA was generated as described above . We designed primers that covered the exon and intron region of Exon 3 to Exon 4 of the rb1 gene . The primers used were: CCGTATTCGAACAGACAGCA and GGTAGAGGGCCAAAGTCACA . After whole-mount in situ hybridization , embryos were washed in PBS , manually deyolked , and mounted on their lateral side in 4% low melting agarose ( Fisher Scientific ) in PBS . Thin 100 µm slices were cut using a vibratome ( Vibratome 1000 Plus ) and sections were stored in PBS until imaging . The in situ hybridization and the vibratome sections were imaged with an Axiocam digital camera ( Zeiss ) mounted on an Axioplan 2 compound microscope ( Zeiss ) . OpenLab 4 . 0 . 2 software ( Improvision ) was used to capture all images; Photoshop CS4 software ( Adobe Systems ) was used for further image processing . | Previous studies have shown that Rb+/− mice develop pituitary adenomas; however , RB1 mutations have not been found in human pituitary tumors . In the present study , we uncovered a novel genetic pathway that may lead to Rb downregulation through RNA splicing mediated by usp39 , a gene involved in assembly of the spliceosome . Our forward genetic study in zebrafish suggests that loss of usp39 results in aberrant rb1 mRNA splicing , which likely causes elevated expression of its target e2f4 , a key regulator known to have oncogenic activity when overexpressed . We established that e2f4 upregulation is a main factor responsible for the adenohypophyseal cell lineage hyperplasia observed in the zebrafish usp39 mutant . It should be of interest to investigate if mutations or downregulation of USP39 would contribute to pituitary tumorigenesis in humans . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology/organogenesis",
"diabetes",
"and",
"endocrinology/neuroendocrinology",
"and",
"pituitary",
"oncology",
"molecular",
"biology/rna",
"splicing"
] | 2011 | Zebrafish usp39 Mutation Leads to rb1 mRNA Splicing Defect and Pituitary Lineage Expansion |
Drosophila melanogaster mount an effective innate immune response against invading microorganisms , but can eventually succumb to persistent pathogenic infections . Understanding of this pathogenesis is limited , but it appears that host factors , induced by microbes , can have a direct cost to the host organism . Mutations in wntD cause susceptibility to Listeria monocytogenes infection , apparently through the derepression of Toll-Dorsal target genes , some of which are deleterious to survival . Here , we use gene expression profiling to identify genes that may mediate the observed susceptibility of wntD mutants to lethal infection . These genes include the TNF family member eiger and the novel immunity gene edin ( elevated during infection; synonym CG32185 ) , both of which are more strongly induced by infection of wntD mutants compared to controls . edin is also expressed more highly during infection of wild-type flies with wild-type Salmonella typhimurium than with a less pathogenic mutant strain , and its expression is regulated in part by the Imd pathway . Furthermore , overexpression of edin can induce age-dependent lethality , while loss of function in edin renders flies more susceptible to Listeria infection . These results are consistent with a model in which the regulation of host factors , including edin , must be tightly controlled to avoid the detrimental consequences of having too much or too little activity .
Drosophila has an effective innate immune system to combat infection . This response relies heavily on the Toll and Immune deficiency ( Imd ) pathways , both of which utilize NF-κB related transcription factors as central mediators of signaling: Dorsal and Dorsal-related immunity factor ( Dif ) in the case of Toll , and Relish ( Rel ) in the case of Imd ( reviewed in [1]–[3] ) . The Toll and Imd pathways have largely been characterized with respect to their role in the humoral immune response , a branch of immunity that is triggered through recognition of microbial molecular signatures by upstream components of both the Imd and Toll pathways and subsequent nuclear translocation and activation of the cognate NF-κB factor ( s ) . The activation of these transcription factors leads to transcription of hundreds of genes following infection [4]–[6] . The most studied are the antimicrobial peptide ( AMP ) genes , which are transcribed in the fat body , leading to secretion of these peptides into the circulating hemolymph ( reviewed in [7] ) . In addition to its role in AMP regulation , the Toll pathway is also known to participate in two other branches of immunity: the deposition of melanin and the cellular immune response [8]–[12] . The cellular response in particular has become of increasing interest , as studies of Drosophila immunity progress beyond the characterization of acute responses to non-pathogenic bacteria to those involving chronic infections that eventually kill the fly [13]–[16] . Many of these model infections proceed intracellularly within the phagocytic cells of the circulating hemolymph , thereby shielding the bacteria from the action of circulating AMPs . This provides a convenient model system for studying the molecular interactions between pathogens and their hosts , including the processes that eventually lead to the host's demise . One principle that has been understood in mammals for decades , and seems to also be true in Drosophila , is that an immune response can be both beneficial and detrimental to a host . Indeed , the same signals that are critical to containing a localized infection will kill the host if uncontrolled [17] . One such signal is Tumor Necrosis Factor ( TNF ) , which is both necessary to fight local infections of many organisms and sufficient to induce lethal septic shock if released systemically [18] , [19] . Homologous processes may also occur in Drosophila; loss of function mutations in the TNF family member eiger result in prolonged survival during infection with Salmonella typhimurium [14] , [20] . Thus Drosophila offers an appealing genetic system to uncover host genes that may have dual effects during the immune response , mediating deleterious consequences to both the pathogen and the host itself . Previously , we reported evidence that flies mutant for the Wnt family member wntD have a defective immune system and succumb prematurely to infection with the gram-positive , lethal bacteria Listeria monocytogenes [21] . Given that WntD acts as a feedback inhibitor of Toll-Dorsal signaling during embryonic development [21] , [22] , we presented a model in which wntD mutants exhibit a hyperactivated immune system , including the overexpression of specific Dorsal target genes that are deleterious to the flies' health . Here , we extend those observations by using Affymetrix oligonucleotide arrays to examine the whole genome transcriptional profiles of wntD mutants prior to and following infection with L . monocytogenes . We examine two groups of candidate mediators of the decreased survival of wntD mutants , and provide evidence that one of those genes , edin ( elevated during infection; synonym CG32185 ) , could be a novel effecter of pathogenesis .
In order to gain insight into the processes that are misregulated in wntD mutants and that may contribute to their susceptibility to L . monocytogenes infection , we collected RNA from wntD and control flies under two conditions: naïve and 24 hours following infection with L . monocytogenes . This time point was chosen because we had observed significant mortality of wntD mutants between 24 and 48 hours under these infection conditions , and hypothesized that misregulation of genes causally involved in this mortality would be seen most clearly at the beginning of this time window [21] . Previously , we showed that wntD mutants exhibit elevated expression of the AMP Diptericin prior to and following infection with the non-pathogenic bacterium Micrococcus luteus , while the AMP Drosomycin is expressed in wntD mutants at levels indistinguishable from wild type [21] . To test the idea that wntD mutants have a hyper activated basal immune system on a more global scale , we used our array data to look at the correlation between each gene's response to infection in wild type ( log2 ( infected controls/uninfected controls ) ) and its level of misregulation in wntD mutants prior to infection ( log2 ( uninfected wntD/uninfected controls ) ) . As shown in Figure 1 , the top thirteen genes most induced by infection all showed higher levels of expression in uninfected wntD mutants compared to uninfected controls . Of these thirteen genes , seven showed an average of greater than 2-fold difference between mutants and controls and had p-values less than 0 . 025 ( Figure 1 and Table 1 ) . This set of genes was comprised of the novel immunity gene edin , IM23 , AttD , AttB , AttA , DiptB , and Def , all of which are known to be induced by infection under various conditions [5] , [23] , [24] . It is worthwhile noting , however , that several known immune-regulated genes that were strongly induced by infection in our study showed no significant difference between wntD mutants and controls , including CG6639 , CecB , TotM and Dros ( Figure 1 and data not shown ) . Overall , the correlation coefficient ( r ) for these data sets was 0 . 14 , with a p-value<0 . 0001 . Calculating the coefficient of determination ( r2 ) suggests that approximately 2% of the variation within the data can be explained by the correlation between the two data sets . This corresponds to approximately 235 genes , a plausible number given previous studies have indicated that about 400 genes are significantly regulated by infection [4] . In a similar analysis looking at the misregulation of immune genes in wntD mutants following infection , no significant correlation was observed ( data not shown ) . As is evident from the cluster analysis presented below and the data in Table 1 , a subset of immune-induced genes were expressed more highly in wntD mutants following infection , but many of the most highly induced immunity genes were not significantly different between wntD mutants and controls , and some were expressed at lower levels in the mutants . This may have resulted from a lack of sensitivity from the array at these high levels of expression , saturation of the signaling processes leading to induction of expression , or dominant negative effects of activated Dorsal on the activity of other NF-κB proteins . To identify genes as candidate mediators of wntD mutants' infection sensitivity , cluster analysis was used [25] . Hierarchical clustering revealed several distinct groups of genes that showed correlation in their expression patterns across the four different conditions . However , two related clusters of genes were selected for further analysis based on the following rationale: the expression of genes actively contributing to pathogenesis will most likely be elevated following infection , and genes within this group that might be implicated in the more rapid lethality seen in wntD mutants would be expressed higher in these mutants . The average expression level under each condition for the two selected clusters ( Clusters A and B ) are shown in Figure 2 . The clusters differ in that Cluster A shows a greater overall change in response to infection than does Cluster B ( Figure 2 ) . Cluster A includes a number of known targets of infection , including several AMPs ( Table S1 ) . While it is certainly possible that several of these are contributing to pathogenesis in the fly , one uncharacterized gene in particular stood out based on its levels of expression . Confirmed by quantitative RT-PCR , edin shows strong induction by Listeria infection ( ∼45 fold ) , and dramatically higher levels of expression in infected wntD mutants versus infected controls ( ∼7 . 5 fold ) ( Figure 2B ) . Furthermore , only a 1 . 7 fold difference was seen between mutants and controls prior to infection , illustrating synergy between Listeria infection and the absence of wntD function on the regulation of edin . Cluster B is composed of genes that show less dramatic changes in response to infection , but are still elevated in wntD mutants versus controls ( Figure 2A , Table S2 ) . It seems likely that this set of genes would include those that are regulated by processes aside from those sensing acute infection ( Toll , Imd ) , and may include both mediators and markers of pathogenesis . Interestingly , this cluster includes the gene eiger , a TNF homolog known to mediate disease processes following Salmonella and Mycobacterium infections [14] , [20] . In this case , using quantitative RT-PCR , we see a statistically significant elevation of eiger expression only in infected wntD mutants ( Figure 2C ) . The edin gene is predicted to encode a secreted protein 115 amino acids in length ( http://flybase . bio . indiana . edu/ . bin/fbidq . htmlFBgn0052185 ) . The gene has homologs in other insects , but not in other Phyla . ( Figure 3 ) . Furthermore , no known conserved domains were identified in Edin or its putative ortholog in Drosophila pseudoobscura and secondary structure prediction failed to identify any similar proteins or motifs based folding patterns ( data not shown ) . To answer the question of whether edin misregulation in wntD mutants is specific to infection with Listeria , wntD and control flies were injected with the non-pathogenic gram-positive bacteria Micrococcus luteus . Analysis of Edin expression levels prior to and following infection were monitored using quantitative RT-PCR ( Figure 4A ) . The results are strikingly similar to those seen for Listeria infection; expression of edin is elevated 1 . 7-fold in wntD mutants compared to controls prior to infection , and 8-fold following infection . Again , a synergistic relationship is seen between infection and the presence of the wntD mutation . The smaller magnitude of edin induction seen in response to M . luteus compared to Listeria ( ∼10 fold versus ∼45 fold in wild-type flies ) may be explained by the shorter timecourse of infection ( 5 hours versus 24 hours ) , a smaller bacterial load at the time of assay , or intrinsic differences between the two species of bacteria . The strong regulation of edin in response to bacterial challenge raises the question of whether its transcription is regulated by the Toll and/or Imd pathways . To investigate this possibility , the induction of edin was monitored in genetic backgrounds each containing a loss of function mutation for a component in one of the pathways ( Figure 4B ) . Mutations in imd reduced the expression of edin following infection to approximately 25% of that seen in wild type . This indicates that the Imd pathway participates in edin regulation , but is not strictly required for its induction following infection . By contrast , loss of function mutations in the Toll ligand spatzle did not reduce the transcriptional induction of edin , and in fact resulted in higher than normal levels of expression . This has been seen for other genes ( such as diptericin ) that do not have a strong requirement for Toll signaling , and could be due to increased survival of the bacteria in these mutants ( data not shown; [4] ) . Levels of edin were slightly elevated ( 4-fold ) in naïve flies carrying a dominantly activated allele of Toll in the absence of infection ( Toll10b; Figure 4B ) . These data indicate that Toll signaling may be sufficient to induce low levels of edin expression , but is not required for its expression . In order to investigate whether Edin plays an essential role in disease progression , we knocked down its expression using two independently made UAS-driven RNA interference ( RNAi ) constructs . Edin expression was knocked down using the fat body driver Lsp2-Gal4 to ablate its activity in a major immune tissue . Edin knockdown flies displayed increased sensitivity to Listeria monocytogenes , with flies dying significantly faster than all controls ( p<0 . 001 ) ( Figure 5 ) . This demonstrates that edin is required for an effective host response against Listeria infection . Interestingly , bacterial loads in edin knockdown flies were not significantly different from controls ( data not shown ) . This places edin among several previously identified genes that affect a fly's endurance during Listeria infection rather than its ability to combat bacterial growth [26] . While the mechanism for this effect is unknown , we hypothesize that knockdown of edin expression alters the physiology of the fly in a way that makes it more susceptible to Listeria pathogenesis . Immunity can be a double-edged sword that has to be regulated precisely to help defend against infection while limiting damage to the body . Overexpression of genes misregulated during an immune response led us to edin and we found that it is required for fly survival during an L . monocytogenes infection . Next , we thought it was of great interest to determine whether Edin expression contributed to pathology . We first looked for more evidence that Edin was associated with pathology under different circumstances . We compared the expression of edin following infection of wild-type flies with wild-type Salmonella typhimurium or a SPI1 , SPI2 mutant strain of Salmonella that has decreased pathogenicity [14] . As shown in Figure 4C , edin was expressed at significantly higher levels during the course of a wild-type Salmonella infection compared to the less pathogenic strain at both time points tested . The more dramatic difference was seen later in infection , when edin transcript levels were over 5-fold higher in flies infected with wild-type Salmonella ( Figure 4C ) . These data add more correlative evidence that Edin is associated with pathogenesis . Do edin expression levels affect survival ? To answer this question , we first overexpressed edin using the UAS/Gal4 system . Two different insertions of the p-element carrying UAS-edin resulted in varied levels of expression , with one insert ( 19-3 ) causing expression levels ∼100 fold over wild type when combined with actin-Gal4 , and the other overexpressing edin over 500 fold ( Figure 6A ) . We observed that the higher level of expression resulted in significant levels of lethality prior to and following eclosion ( Figure 6B , C ) . Flies strongly overexpressing edin survived to adulthood at a frequency less than 50% of expected , compared to 111% for the lower expresser . The value greater than 100% can most likely be attributed to non-specific deleterious effects of carrying the CyO balancer . The average lifespan of those flies surviving to adulthood was also significantly reduced in the context of strong overexpression of edin ( Figure 6C ) . Given that wntD mutants infected with L . monocytogenes displayed similar levels of expression to the strong insertion of UAS-edin ( about 350 fold over uninfected wild-type flies; Figure 2B ) , it is possible that edin expression is contributing to the rapid mortality of these mutants . Taken together with the observation that edin loss of function mutants show increased sensitivity to L . monocytogenes , these data support a model in which edin expression must be tightly controlled during a host response to infection: moderate induction is essential to an effective response , but uncontrolled , high levels of expression become detrimental to the host animal .
We presented two experiments that compared the expression profiles of flies undergoing two different levels of pathogenesis: wntD versus control flies following L . monocytogenes infection , and wild-type S . typhimurium versus a SPI1 , SPI2 mutant strain . In both cases the gene edin was strongly elevated in the flies closer to death . In comparing wntD mutant versus control flies following Listeria infection , RNA samples were taken 1 day after infection , shortly before the mutants exhibit a sharp decrease is survival [21] . Expression of edin was about 8-fold higher in the wntD mutants . Similarly , at 7 days post Salmonella infection , flies infected with wild type have begun to die , while those infected with a SPI1 , SPI2 mutant strain will live for several more days despite carrying dramatically higher loads of bacteria [14] . In this case , we observed a 5-fold elevation in edin expression in the flies beginning to die . Thus , high edin expression is correlated with increased pathogenesis , although a causal relationship is not established by these data . Two results strongly suggest that edin induction is not downstream of pathogenesis . First , edin expression is elevated following infection with M . luteus , a non-pathogenic bacterium , and is more strongly induced in wntD mutants ( Figure 2A ) . These data demonstrate that pathogenesis is not required for edin expression . Second , the Imd pathway appears to play a significant role in regulating edin , and this pathway is acutely induced upon recognition of bacterial moieties and does not strictly depend on pathogenesis [29]–[31] . Could Edin play a causal role upstream of pathogenesis ? The induction of edin during M . luteus infection without any demonstrable pathogenesis suggests that the amount of Edin produced during this infection is not sufficient to elicit pathogenesis . However , these levels are approximately 5-fold lower than those seen for Listeria infection and persist for less than a day ( data not shown ) , in contrast to the chronic induction during infection with Listeria or Salmonella . Furthermore , the lethality induced by strong chronic overexpression of edin using the UAS/Gal4 system implies that this gene can induce processes detrimental to a fly's survival that could be affecting viability during persistent infections . Though Edin can be shown to cause pathology when overexpressed , it is difficult to produce clean evidence that this occurs during infection , because the overexpression of many genes can cause pathology; therefore it remains a suggestion . Edin shows several characteristics consistent with it being an AMP . First , it is strongly induced by infection; edin was the second most highly induced gene in wild-type flies following L . monocytogenes infection , and the most highly induced gene in wntD mutants . Second , edin is predicted to encode a short peptide and a processed form has been observed circulating in the hemolymph of infected flies [23] . However , edin also displays properties that would make it unique among AMPs , suggesting that it may be more broadly affecting physiology , perhaps in a cytokine-like role similar to that of eiger . For instance , the expression of this gene is required for normal survival following L . monocytogenes infection . While necessity for the signaling pathways controlling AMP expression is well documented , this is the first case of an individual putative AMP being necessary to fight infections {Ferrandon , 2007 #329} . This requirement during infection , combined with the toxicity observed upon overexpression suggests that Edin may be a powerful component of the immune response that must be tightly regulated to optimize survival . Further analysis of edin and other genes that are differentially regulated during pathogenesis could provide interesting clues into the complicated and evolving nature of the host-pathogen interaction .
The construction of wntD mutants was described previously [21] . Any reference to wntD mutant is the genotype yw; wntDKO1 . References to ‘wild type’ refer to yw; +/+; +/+ or w1118; +/+; +/+ if so noted . pP[UAS-edin] was constructed by amplifying the edin open reading frame using PCR , and cloning this fragment into the Xba-1 site of pPUAST [32] . UAS-RNAi ( edin ) 2 was created at the VDRC ( transformant 14289 ) . UAS-RNAi ( edin ) 1 was generated by PCR amplification of the complete cDNA with XbaI sites at both 5′ and 3′ ends . This fragment was subcloned into the pWIZ vector [33] in two sequential cloning steps on either side of a small intron in a 3′to 5′/5′to 3′ orientation . Expression of the double-stranded RNA is under the control of the UAS promoter and is transformed into a snapback hairpin upon splicing of the small intron . Flies carrying expression constructs were created using standard p-element transformation techniques . All injections were done using male flies aged one week post eclosion . A culture of Listeria monocytogenes was diluted to an OD ( 600 ) of 0 . 1 , and a 25 nL volume was injected abdominally using a pulled glass needle as previously described [15] . Groups of 20 flies of each genotype were injected in an alternating manner to control for variability over time . Flies were maintained on non-yeasted , standard dextrose medium at 25°C , 65% relative humidity , and survival was monitored daily . Micrococcus luteus and Salmonella typhimurium was injected as described for L . monocytogenes . For experiments on the regulation of edin , flies of different genetic backgrounds were injected with a mixture of M . luteus , L . monocytogenes , and E . coli , each at an OD ( 600 ) of 0 . 1 . Groups of 6 flies were collected , crushed in 150 µl of Trizol reagent , and RNA was extracted according to the manufacturer's recommendations . 1 µl RNA was used for subsequent reverse transcription using the ThermoScript RT-PCR system ( Gibco BRL ) , following the manufacturer's instructions and using a random hexamer as primer . Quantitative PCR was preformed in a LightCycler ( Roche ) , using the LightCycler FastStart DNA MasterPLUS SYBR green I kit ( Roche ) and following the manufacturer's recommendations . Primers used for PCR were as follows: Groups of 30 yw;wntDKO1 or yw flies ( some previously infected with Listeria monocytogenes as described above ) were collected in 1 . 5 mL microfuge tubes . Each experiment was done in triplicate , for 12 total samples . Conditions were: yw uninjected , yw;wntDKO1 uninjected , yw 24 hours post Listeria infection , yw;wntDKO1 24 hours post Listeria infection . Flies were crushed in 1 mL Trizol reagent , and RNA was isolated using the manufacturer's recommendations . 15 µg of each RNA sample was then used for cDNA synthesis , which was done using the one cycle cDNA synthesis ( Affymetrix ) and following the manufacturer's recommendations . cRNA was also synthesized using the manufacturer's protocol , and 20 ug was used for the subsequent fragmentation step . cRNA was hybridized to Affymetrix Drosophila Genome 2 . 0 arrays by the Stanford Protein and Nucleic Acid Biotechnology Facility ( http://pan . stanford . edu ) . Arrays were analyzed using the Affymetrix GCOS software to produce normalized values for each probe set on each array . Clustering was performed on a dataset in which genes were included only if they were marked as “present” by GCOS in all 3 samples of at least one condition . Clustering was done using Cluster 3 . 0 for Mac OS X ( http://bonsai . ims . u-tokyo . ac . jp/mdehoon/software/cluster/software . htm ) . Parameters used for clustering were: Data was log transformed and genes were centered . Data was filtered to include only genes where the difference between the highest and lowest values was greater than or equal to 1 ( representing a two-fold change or greater ) . Hierarchical clustering was performed using the centroid linkage algorithm . Clusters were viewed using Java Treeview software ( http://genetics . stanford . edu/alok/TreeView/ ) . Gene identities and annotations shown in Tables S1 and S2 were retrieved using the Netaffx analysis webpage ( http://www . affymetrix . com/analysis/index . affx ) . | Like any organism , fruit flies respond to invading microorganisms by mounting an immune defense . Many aspects of the immune defense in fruit flies are similar to the inflammatory response in mammals , including the harmful effects of a sustained response against persistent pathogenic infections . We found in the past that mutations in the gene wntD cause flies to succumb more easily to Listeria monocytogenes infections , apparently by losing an element of control over the inflammatory response . How does the wntD gene work ? In this paper , we have identified genes that may mediate the susceptibility of wntD mutants to lethal infection . These genes include eiger , a homolog of the mammalian TNF gene , and a previously uncharacterized gene called edin ( elevated during infection ) . Edin is expressed excessively in wntD mutant flies , and its expression also correlates with the level of pathogenesis induced by two different strains of Salmonella typhimurium . In its own right , overexpression of the edin gene can induce lethality , while losing edin function renders flies more susceptible to Listeria infection . Our results support a model in which the regulation of host factors , including edin , must be tightly controlled to avoid the detrimental consequences of having too much or too little activity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"immunology"
] | 2008 | Pathogenesis of Listeria-Infected Drosophila wntD Mutants Is Associated with Elevated Levels of the Novel Immunity Gene edin |
Dengue is a mosquito-borne viral disease caused by the four dengue viruses ( DENV-1 to 4 ) that can also be transmitted by blood transfusion and organ transplantation . The distribution of DENV in the components of blood from infected donors is poorly understood . We used an in-house TaqMan qRT-PCR assay to test residual samples of plasma , cellular components of whole blood ( CCWB ) , serum and clot specimens from the same collection from blood donors who were DENV-RNA-reactive in a parallel blood safety study . To assess whether DENV RNA detected by TaqMan was associated with infectious virus , DENV infectivity in available samples was determined by culture in mosquito cells . DENV RNA was detected by TaqMan in all tested blood components , albeit more consistently in the cellular components; 78 . 8% of CCWB , 73 . 3% of clots , 86 . 7% of sera and 41 . 8% of plasma samples . DENV-1 was detected in 48 plasma and 97 CCWB samples while DENV-4 was detected in 21 plasma and 31 CCWB samples . In mosquito cell cultures , 29/111 ( 26 . 1% ) plasma and 32/97 ( 32 . 7% ) CCWB samples were infectious . A subset of samples from 29 donors was separately analyzed to compare DENV viral loads in the available blood components . DENV viral loads did not differ significantly between components and ranged from 3–8 log10 PCR-detectable units/ml . DENV was present in all tested components from most donors , and viral RNA was not preferentially distributed in any of the tested components . Infectious DENV was also present in similar proportions in cultured plasma , clot and CCWB samples , indicating that these components may serve as a resource when sample sizes are limited . However , these results suggest that the sensitivity of the nucleic acid tests ( NAT ) for these viruses would not be improved by testing whole blood or components other than plasma .
Dengue is a febrile disease of global public health importance , affecting more than 100 tropical and subtropical countries and causing an estimated 390 million infections per year [1 , 2] . The disease is caused by any of four closely related dengue viruses ( DENV-1 to 4 ) from the genus Flavivirus , family Flaviviridae , and is transmitted by mosquitoes from the genus Aedes . Infection with any of the four DENV can be asymptomatic or cause an influenza-like illness ( dengue fever ) , that may progress to a potentially life-threatening condition ( severe dengue or dengue hemorrhagic fever; DHF ) , especially after secondary , heterotypic infections [3 , 4] . Although dengue primarily affects tropical and sub-tropical countries , the virus can be imported by infected travelers to non-endemic regions . Recent dengue epidemics have occurred in non-endemic areas of Europe and the United States of America ( U . S . ) , where the transmitting mosquitoes have become established , enabling local transmission cycles [5–7] . Dengue is endemic in the U . S . territory of Puerto Rico , where it has caused large epidemics in 2010 ( >26 , 700 suspected cases ) , 2012 ( >12 , 800 cases ) , 2013 ( >18 , 100 cases ) and 2014 ( >8 , 600 cases ) , for a total of more than 66 , 000 suspected cases reported during these last four epidemic years [5 , 8] . Transfusion-transmitted DENV ( TT-DENV ) has been reported in dengue endemic regions including Puerto Rico , where a transfusion of red blood cells caused DHF in a recipient [9–16] . DENV is also transmissible by solid organ transplant , thus posing a risk for recipients of these products [17–20] . At this time , there is no Food and Drug Administration ( FDA ) -approved nucleic acid test ( NAT ) for the screening of blood for DENV RNA . DENV , like many other arboviruses , can cause asymptomatic infections in up to 80% of infected individuals [21] . Prevalence studies conducted in endemic regions have found rates of asymptomatic infection among blood donors ranging from 0 . 07–0 . 45% [11 , 12 , 14 , 16 , 22 , 23] . In 2010 , Puerto Rico experienced a large epidemic with over 26 , 000 reported dengue cases [5 , 24] and blood donations were screened for DENV using an antigen-based immunoassay ( DENV NS1 Ag ) under an FDA-approved investigational new drug ( IND ) protocol [5 , 25] . Using this protocol , DENV was detected in blood units collected from asymptomatic blood donors from Puerto Rico and from Key West , Florida . Data acquired during that IND study indicated that the DENV NS1 Ag assay lacked proper sensitivity and specificity for blood donor screening [16] . Antibody tests for DENV are available and used for diagnostic purposes , but they are unsuitable for blood screening because the infectious viremic stage precedes seroconversion . Moreover , unlike chronic infections , identification of DENV-specific antibodies does not necessarily represent active viral infection . In addition , infection and antibody development can occur in the absence of clinically apparent infection . Therefore , NAT targeting DENV RNA has been considered the most appropriate approach for blood screening . Although there is no FDA-approved assay for the screening of blood for DENV , during the epidemic seasons of 2012–2013 NAT using transcription-mediated amplification ( TMA ) was used to test blood donation samples in Puerto Rico under an IND protocol [16 , 26] . In the absence of licensed DENV NAT for blood screening , developing measures to mitigate risk of TT-DENV and safeguard the blood supply is a challenge; even more so in endemic areas like Puerto Rico , and in areas with recurrent small outbreaks like Florida and Texas . The measures to be taken rely heavily on published data regarding the biology of infection , which are mostly obtained from clinical cases from epidemic areas . While the clinical course of dengue disease has been well-studied , many aspects of DENV infection in asymptomatic or pre-symptomatic infections are unknown or poorly understood . In particular , understanding the distribution of virions in the components of whole blood in asymptomatic or pre-symptomatic infections is critical to establishing and optimizing blood testing protocols , thus enabling the use of currently-discarded components to enhance sample availability . In this study , we used residual samples from specimens from blood donors confirmed positive for DENV RNA that were initially reactive by TMA performed as part of an IND study in Puerto Rico [16 , 26] , to determine the distribution of DENV in blood components , quantify viral loads during the asymptomatic period , and demonstrate infectivity of the various components in vitro . In addition , we investigated whether , as has been observed for WNV [27–29] , testing in blood components other than plasma or serum would result in increased assay sensitivity and enhancement of DENV detection .
The protocol for specimen collection and testing by investigational and research assays was approved by the American Red Cross ( ARC ) Institutional Review Board . The study protocol ( 13-001B ) was reviewed and considered exempt by the Research Including Human Subjects Committee at the U . S . FDA . This study was performed using available residual samples from blood donors who tested initially reactive for DENV RNA by an investigational NAT assay ( DENV TMA assay , Hologic , Inc . , formerly Gen-Probe , Inc . ) , in use under an FDA-approved IND to screen blood donations from Puerto Rico for DENV RNA as part of an independent blood safety study [16 , 26] . Date of specimen collection and demographical information were available for all blood donors . All subjects were asymptomatic at the time of donation . Paired EDTA whole blood samples and their corresponding plasma samples were received for a total of 165 donors . Additionally , 149 clot and 155 serum specimens from those donors were available for testing . Samples were collected between August 2012 and August 2013 , and the residual samples from tubes used for blood screening purposes by the ARC were unlinked and shipped either refrigerated ( for collection tubes containing EDTA whole blood or blood clots ) or frozen ( for tubes containing plasma or serum ) to the laboratory at the FDA for processing and testing . The cellular components of whole blood ( CCWB ) specimens consisted of all blood cells from the EDTA blood sample tube remaining after removal of the plasma for blood testing purposes including NAT . Plasma specimens were obtained from the co-component units from that same collection , which , after testing positive for DENV RNA , were discarded from the blood supply and labeled for research use only . Blood clots and serum specimens were available as residuals of samples tested for DENV antigen and serology as part of the IND protocol . Upon receipt , frozen specimens ( plasma and serum ) were thawed and divided into aliquots . One aliquot was used for cultivation , and the remaining aliquots were stored frozen at -80°C until processing . CCWB samples were prepared by centrifugation at 4°C , 3000 rpm , for 10 minutes , followed by removal of any remaining plasma , and divided into aliquots of 250 μL; one aliquot was used for cultivation and the remaining aliquots were mixed with 1 mL of Trizol and frozen at –80°C until RNA extraction . The blood clots were stored at 4°C until processing for cultivation and RNA extraction as described below . Negative control plasma and CCWB specimens were obtained from non-DENV-endemic areas of the contiguous U . S . and were processed and extracted as described above . For liquid specimens ( i . e . plasma , serum and cell culture supernatants ) , total RNA was extracted from aliquots of 140 μL from each sample using the QIAamp Viral Mini Kit ( Qiagen ) and eluted in either 70 or 140 μL of AVE buffer . Eluted RNA was kept at –80°C until testing . Total RNA extraction from CCWB and clot samples was performed using Trizol reagent ( Invitrogen ) as per the manufacturer’s instructions with some modifications . Briefly , the 250 μL CCWB aliquots previously mixed and stored frozen with Trizol were allowed to thaw at room temperature and the RNA extraction processed as per the manufacturer’s protocol , followed by RNA cleaning and concentration using the RNeasy Mini kit ( Qiagen ) , elution in 50 μL of RNase-free water and storage at –80°C until testing . Approximately 150 mg of each clot were placed into an RNAse-free microcentrifuge tube and 1 . 5 ml of Trizol were added and vortex-mixed . The mixture was transferred to a homogenization vial containing ceramic beads ( Precellys ) and was subjected to a homogenization cycle ( 2 cycles of 20 seconds at 5 , 000 rpm with a 5 second break between cycles ) in the Precellys 24 Dual homogenizer ( Precellys ) . The clot-Trizol homogenate was then transferred to an RNase-free microcentrifuge tube and the RNA extraction protocol with Trizol was continued as described above . RNA was eluted in 50 μL of RNase-free water and stored at –80°C until testing . RNA extracts were tested by an in-house dengue specific real-time polymerase-chain reaction ( TaqMan ) assay based on a modification of the protocol published by Johnson et al . [30] . Briefly , 10 μL of eluted RNA from each tested component were added to a reaction containing 12 . 5 pmol of primers and 6 . 25 pmol of probe for each DENV type in singleplex format in duplicate in a volume of 25 μL , using the 1-step RNA-to-Ct kit ( Applied Biosystems ) . Reactions were conducted in a 7300 Real-time PCR System ( Applied Biosystems ) . Standard curves for each DENV type , calibrated by endpoint dilution and covering a dynamic range of 7-log10 , were used and results were reported as PCR-detectable units per ml ( PDU/mL ) for RNA extracted from liquid or CCWB specimens or per mg ( PDU/mg ) for RNA extracted from clots [31 , 32] . Correction for extraction and elution volumes was done by using the formula: PCR detectable units per reaction x ( 1 ml [or 1 mg for clots]/sample extracted volume ) x ( elution volume/tested volume ) . At least two independent RNA extractions and TaqMan tests were conducted for each blood component evaluated . Samples with cycle threshold ( Ct ) <38 were considered positive [30] . Although samples with cycle threshold ( Ct ) ≥38 were not used for quantitative analysis , they were considered DENV RNA-positive if they tested positive in the co-component or were infectious in cell culture . The different blood components ( i . e . plasma , serum , CCWB and clot ) were used to infect monolayers of Aedes albopictus C6/36 cells in culture . The following volumes from each sample type were used for culture: 250 μL of undiluted cell suspension for CCWB; 500 μL of 1:20 diluted plasma or serum in MEM containing 2% FBS and 1% of fungizone and gentamycin ( MEM2% ) ; 150 mg of clots homogenized in 500 μL of MEM2% by extrusion using a syringe and needle . Dilution of liquid samples was performed to prevent non-specific cytopathic effect in the C6/36 monolayers . Samples for culture were added to T25 flasks containing ~85% confluent cell monolayers and incubated for one hour at 32°C and 5% CO2 with gentle rocking every 15 minutes . After incubation , 5 mL of MEM2% were added to each flask and the cell cultures were incubated at 32°C , 5% CO2 , and monitored daily . Cell cultures exposed to CCWB had their inocula removed and replaced with fresh MEM2% media 24 h post-infection and were maintained as indicated above . On day 7 post-infection ( d . p . i . ) , all cell cultures had their supernatants harvested and subjected to RNA extraction for testing by TaqMan as described above . An aliquot of supernatant was also used for titration of infectivity by focus-forming assay ( FFA ) using anti-DENV specific monoclonal antibody 4G2 obtained from hybridoma culture supernatants ( ATCC ) , essentially as described previously [33] . Infectivity titers were reported as focus-forming units/ml ( FFU/ml ) . RNA testing results were log-transformed to ensure normal distribution of the data and expressed as mean ± standard deviation . Differences between viral load or infectivity titer averages in the tested components from each DENV type were determined by t-test or by one-way ANOVA with Bonferroni’s post-test using GraphPad Prism version 5 . 04 ( GraphPad Software , San Diego , USA ) . Statistical significance was considered with p<0 . 05 .
Plasma samples from 165 Puerto Rican blood donors that tested initially reactive for DENV RNA by TMA assay as part of a parallel IND blood safety study were tested by an in-house TaqMan RT-PCR assay . A total of 69 ( 41 . 8% ) of these 165 samples were also DENV RNA-reactive on our assay . Of these 69 samples , 48 ( 69 . 6% ) tested reactive to DENV-1 and 21 ( 30 . 4% ) tested reactive for DENV-4 . Among the 165 CCWB tested , 130 ( 78 . 8% ) were DENV RNA-reactive on our assay; 98 ( 75 . 4% ) were reactive to DENV-1 , 31 ( 23 . 8% ) for DENV-4 , and one ( 0 . 8% ) was found to be reactive to both DENV-1 and DENV-4 , suggesting a co-infection ( Table 1 ) . All negative control plasma and CCWB specimens were non-reactive by TaqMan . Quantitative analysis was performed for all DENV RNA-positive samples by TaqMan with Ct≤38 [30] . The cutoff of Ct≤38 was chosen because reactivity outside of the linear range of the standard curve produces unreliable detection , inappropriate for quantification . Most plasma and CCWB samples had Ct≤38; 75% and 87% , respectively ( Table 1 ) . The average Ct value for DENV-1-positive plasma samples ( Ct≤38 ) was 31 . 3 ± 5 . 0 ( range 23 . 5–38 ) for an average titer of 4 . 778 ± 1 . 416 ( range 2 . 272–6 . 994 ) log10 PDU/mL , while the average Ct value for DENV-1-positive CCWB samples ( Ct≤38 ) was 32 ± 4 . 2 ( range 21 . 2–37 . 8 ) for an average titer of 3 . 824 ± 1 . 168 ( range 1 . 884–6 . 610 ) log10 PDU/mL . For DENV-4-positive plasma samples ( Ct≤38 ) the average Ct value was 30 . 4 ± 5 . 6 ( range 17 . 6–37 . 3 ) , with average titer of 5 . 163 ± 1 . 552 ( range 3 . 097–8 . 324 ) log10 PDU/mL , while the DENV-4-positive CCWB samples ( Ct≤38 ) had an average Ct value of 29 . 5 ± 4 . 6 ( range 19 . 1–37 . 3 ) for an average titer of 4 . 761 ± 1 . 356 ( range 2 . 716–7 . 691 ) log10 PDU/mL ( Fig 1 ) . Laboratory diagnosis of DENV is mostly performed using serum samples , and the blood clot from which the serum is recovered is usually discarded . When available , we have also tested sera and their respective blood clots for determination of DENV viral loads and infectivity , with the aim of defining the specimen that produces the highest rate of detection for DENV RNA as the most suitable testing sample for both diagnostic and blood screening settings . To assess DENV viral loads in clots , we standardized a process for extracting RNA from homogenized blood clots , and determined viral loads in clots from 90 samples that had tested positive for DENV RNA on plasma and/or CCWB by our TaqMan assay and had clots available for testing . Of these 90 clot homogenates , 66 ( 73 . 3% ) were TaqMan-reactive , of which 50/66 ( 75 . 8% ) were DENV-1 and 16 ( 24 . 2% ) were DENV-4 . Most tests ( 82% ) produced reactivity with Ct≤38 ( Table 1 ) . A total of 30 serum samples from the same donors who had the clots tested were available for testing . Of those , 26 ( 86 . 7% ) tested reactive to DENV , of which 16/26 ( 61 . 5% ) were reactive to DENV-1 , while 10/26 ( 38 . 5% ) were reactive to DENV-4 ( Table 1 ) . Overall , there was a correlation between the DENV viral loads found among the liquid components ( serum and plasma ) and among the cellular components ( clots and CCWB ) . However , DENV-1 viral loads were significantly higher in the liquid components ( plasma and serum ) than those found in cellular components , CCWB ( p≤0 . 001 ) and clots ( p≤0 . 05 ) ; while DENV-4 viral loads did not differ significantly between samples from different blood components tested ( Fig 1 ) . A side-by-side comparison of quantifiable DENV viral loads in all blood components from a subset of DENV-infected blood donors was performed to investigate whether , as reported for WNV , DENV would have a higher viral load associated with the cellular components of the blood . As stated above , quantitative analysis was only performed when all available specimens from each donor produced quantifiable TaqMan results with Ct values of 38 or lower ( Ct≤38 ) , which falls within the linear range of the standard curve . A subset of 29 donors fulfilled this criterion for plasma and CCWB , of whom 19 ( 65 . 5% ) had DENV-1 RNA and 10 ( 34 . 5% ) had DENV-4 RNA detected in their components . DENV viral loads did not differ significantly among the tested components ( ANOVA p>0 . 05 ) , ranging from 3 . 6–6 . 4 ( average ± SD: 5 . 0 ± 1 . 0 ) log10 PDU/mL in CCWB and 3 . 9–7 . 3 ( average ± SD: 5 . 5 ± 1 . 0 ) log10 PDU/ml in plasma . Similar results were obtained for samples from donors infected with DENV-4; 3 . 7–7 . 1 ( average ± SD: 5 . 4 ± 1 . 4 ) log10 PDU/mL in CCWB and 3 . 4–8 . 0 ( average ± SD: 5 . 7 ± 1 . 5 ) log10 PDU/mL in plasma and . Results from the DENV RNA quantitation in this group are shown in Fig 2 and S1 Fig and S1 Table . Of the 29 samples chosen for this analysis , 21 had serum and clot specimens available for testing . Serum viral loads did not differ significantly from plasma and CCWB viral loads , ranging from 3 . 6–6 . 8 ( average ± SD: 4 . 9 ± 0 . 9 ) log10 PDU/mL for DENV-1 and 3 . 6–7 . 4 ( average ± SD: 5 . 4 ± 1 . 1 ) log10 PDU/mL for DENV-4 . Notably , the clots tested for this analysis had been stored at 4°C since collection ( range 3 . 5–10 months of age ) . Surprisingly , all 21 tested clots contained detectable DENV RNA although at values slightly lower to those obtained from testing the other components . The viral loads detected from the clot homogenate extracts ranged from 3 . 3–6 . 0 ( average ± SD: 4 . 5 ± 0 . 9 ) log10 PDU/ml for DENV-1 and of 3 . 5–6 . 8 ( average ± SD: 4 . 8 ± 1 . 0 ) log10 PDU/ml for DENV-4 even after prolonged storage at 4°C , suggesting that only limited degradation of the viral RNA in the clot occurred during storage ( S1 Table , Figs 2 and 3A ) . The age of clot at the time of testing weakly correlated with the DENV viral loads detected , with fresher clots containing higher DENV viral loads . To confirm this apparent association , we tested an additional six fresher ( ≤2 . 5 months of storage at 4°C ) clot specimens that were available . We found that the DENV viral loads in these six clots were comparable and in some cases higher than those obtained from the plasma or CCWB co-component; of these , four were reactive for DENV-1 and two for DENV-4 . DENV-1 viral loads in these clots ranged from 4 . 3–5 . 7 ( average ± SD: 4 . 9 ± 0 . 5 ) log10 PDU/mL , and for DENV-4 from 4 . 8–6 . 1 ( average ± SD: 5 . 4 ± 0 . 7 ) log10 PDU/mL ( S2 Table , Figs 3B and S2 ) . The overall results indicate that a higher concentration of DENV RNA is present in fresher clots ( average storage age DENV-1: 1 . 32 months , DENV-4: 0 . 77 months ) in comparison to that observed in clots stored for longer periods of time ( average storage age DENV-1: 7 . 39 months , DENV-4: 5 . 82 months ) . This observation allows us to speculate that clots may harbor a high concentration of virions attached to or inside cells and although the viral load tends to drop over time , DENV RNA can still be detectable after 3 to 10 months of storage at 4°C ( S1 and S2 Tables ) . Plasma and CCWB samples were also cultured in the C6/36 mosquito cell line if sufficient volumes were available , in order to verify infectivity of DENV present in these components . Culture positivity was determined by TaqMan assay performed on cell culture supernatants . Of 111 plasma samples cultured , 29 supernatants ( 26 . 1% ) tested TaqMan-reactive , 19 for DENV-1 and 10 for DENV-4; while 32/97 ( 33 . 0% ) of CCWB culture supernatants tested TaqMan-reactive , 23 for DENV-1 and 9 for DENV-4 ( Fig 4A ) . Most specimens that were positive by tissue culture had also been reactive in the original tested component by TaqMan . For plasma specimens , 27/60 ( 45% ) component TaqMan-reactive specimens were positive by tissue culture ( 18/44 DENV-1 and 9/16 DENV-4 ) , and 2/51 ( 3 . 9% ) cultures for which the component specimens were TaqMan-negative grew in culture ( 1 DENV-1 and 1 DENV-4 ) . For CCWB specimens , 32/86 ( 37 . 6% ) component TaqMan-reactive specimens were positive by tissue culture ( 23/65 DENV-1 and 9/22 DENV-4 including one specimen that was dual-positive by TaqMan but negative by culture ) , and 0/11 ( 0% ) TaqMan-negative components produced infectious virus . In all cultures of TaqMan-reactive components , the DENV serotype , identified by TaqMan performed on the culture supernatant , matched that of the original component . Owing to the low rates of recovery for TaqMan-negative plasma and CCWB , only TaqMan-reactive serum and clot specimens were tested in tissue culture . Of 26 serum supernatants cultured , 16 ( 61 . 5% ) tested TaqMan-reactive , 11 DENV-1 and 5 DENV-4 ( Fig 4A ) . Clot homogenates that tested reactive to DENV RNA by TaqMan ( n = 66 ) were also cultured to determine whether infectious particles persisted after long-term storage at 4°C . Ten ( 15 . 2% ) clot homogenate culture supernatants tested TaqMan-reactive , 4 DENV-1 and 6 DENV-4 ( Fig 4A ) . A subset of TaqMan-reactive supernatants from plasma and CCWB co-cultures were tested by FFA to confirm that cultures positive for DENV RNA by TaqMan also produced infectious virions . A total of 22 supernatants from cells cultured with TaqMan-reactive plasma samples ( 18 DENV-1 and 4 DENV-4 ) and 28 supernatants from cells cultured with TaqMan-reactive CCWB samples ( 22 DENV-1 and 6 DENV-4 ) , yielded infectious virus as determined by the FFA assay ( Fig 4B and 4C ) . FFA performed on culture supernatants of CCWB specimens from DENV-4-positive blood donors produced significantly more infectious virus than cultures of plasma specimens ( mean 4 . 84 vs 3 . 51 , p = 0 . 007 ) , but there was no difference in the amount of infectious virus produced between CCWB and plasma from DENV-1-positive blood donors ( Fig 4B and 4C ) .
Dengue is the most important arboviral disease in the world and has become one of the most frequently reported travel-associated infections worldwide [2 , 4 , 34] . DENV can be transmitted by blood transfusion and solid organ transplants , and all blood products used in clinical practice , namely packed red blood cells , platelet concentrates and fresh frozen plasma , have been implicated in the transfusion-transmission of DENV [9 , 10 , 13–20] . Because approximately 80% of DENV infections can be asymptomatic [21] , DENV poses a threat for the blood supply , especially during outbreaks [11 , 12 , 14 , 16 , 22 , 23 , 35] . Overall , little is known about the biology of infection and pathogenesis of asymptomatic cases , especially regarding viral distribution in circulating blood , which is relevant for blood screening protocols . DENV has been endemic in Puerto Rico for many years , and during the most recent epidemic years ( 2010 , 2012 , 2013 and 2014 ) more than 60 , 000 cases of dengue have been reported [5 , 8 , 24] . DENV-1 and DENV-4 were the viruses predominantly detected in Puerto Rico during the epidemic years 2012–2013 [8] and from samples collected in this study ( Table 1 ) . Imported cases of dengue are reported every year in the U . S . , and autochthonous dengue transmission has been documented in the states of Florida , Texas , Hawaii and New York [5 , 36] . In the U . S . , DENV has been demonstrated to circulate among asymptomatic blood donors in endemic ( i . e . , Puerto Rico ) and non-endemic ( i . e . , Florida ) areas experiencing outbreaks [14 , 37] . Moreover , imported asymptomatic infections of travelers returning from endemic/outbreak areas may occur which are not identified and reported , and thus may serve as a source of localized outbreaks and pose an unknown level of risk to the blood supply . To better understand aspects of asymptomatic DENV infection which may affect the performance of blood screening assays , we analyzed the distribution of infectious DENV particles and viral RNA loads among different blood components ( i . e . , plasma , the cellular fraction of anticoagulated blood , serum and blood clots ) in infected blood donors from Puerto Rico . The distribution of both DENV RNA and infectious particles and their associated viral loads in blood may provide information to help increase the sensitivity of viral detection and assess relative transfusion-transmission risks for each component of blood . One of the main aims of this work was to determine the feasibility of use of samples other than plasma for the detection and quantification of DENV RNA . Probably due to easier collection logistics , serum has been the sample of choice for diagnosis of infection by DENV primarily in endemic areas where resources may be limited; however , plasma samples are also used although less frequently [4] . Conversely , FDA-approved blood donation screening assays for viruses such as West Nile Virus ( WNV ) , Human Immunodeficiency Virus , Hepatitis C Virus ( HCV ) and Hepatitis B Virus , and the investigational DENV NAT assay are performed using plasma as the testing specimen , because plasma is easy to handle and store and readily available for testing consistent with other infectious disease markers [16] . Additionally , hemoglobin interferes with most NAT assays . However , there has been no direct comparison of efficiency of detection of DENV for various sample types in diagnostic or blood screening settings . We compared the DENV RNA titers obtained from the plasma and serum with the cellular components of blood from all available DENV-TMA-reactive samples obtained from the ARC and found that at least for DENV-1 there is a significantly higher DENV viral load in the liquid components ( plasma and serum ) compared to those in the cellular components ( CCWB and clots ) ( Fig 1 ) . These differences between components observed for DENV-1 but not for DENV-4 , may correspond to differences in the performance of our TaqMan assay for each virus , to an intrinsic difference in the natural distribution of these two viruses in blood components , or to non-viral factors such as prior donor exposure to DENV or timing of donations . When a more detailed quantitative analysis of DENV RNA was performed in a subset of plasma and CCWB samples with Ct≤38 as the cutoff in our TaqMan assay , those differences were not detected and similar DENV viral loads were observed across all tested infected individuals ( Fig 2 ) . Our results suggest that DENV RNA may be equally distributed in all blood components , thus reinforcing the notion that blood products from viremic donors may be equally infectious for recipients , despite the presence of anti-DENV antibodies , as >90% of infected blood donors in Puerto Rico are antibody-positive [35] . The flaviviruses DENV and WNV have both been linked to cases of transmission by tissue and organ transplant [17–20 , 38 , 39] . We attempted to use blood clots ( acquired from tubes of blood collected without anticoagulants to obtain serum for screening purposes and other testing ) , as a source for the detection of DENV RNA and for its use for isolation of infectious viruses . DENV viral loads were in general lower in clots than in the other tested components , although not significantly ( Fig 2 ) . A surprising finding was the fact that we were able to detect DENV RNA and isolate viable viruses from clots that have been stored for several months at 4°C , thus suggesting a protective factor within the clot maintaining the integrity of virions and the retention of infectivity . This observation has prompted us to propose the use of fresh blood clots as a surrogate for tissues for the study of viral distribution in tissues and organs in research studies analyzing the infectivity of DENV and other viruses in tissues intended for transplantation . Although we tried to culture the CCWB and clots as soon as they were received , in some circumstances some samples were stored for longer periods at 4°C than others before culture , and therefore it is hard to make an accurate side-by-side comparison between the efficiency of these components for DENV isolation , with respect to the liquid co-components ( plasma and serum ) that were received and kept frozen until infection . Still , our observation of substantial culture positivity in the solid components along with the liquid components ( Fig 4 ) warrants further study since it may have important implications when analyzing the risk for the different blood components to efficiently transmit DENV , as well as for its detection during blood screening . It is also possible that a higher proportion of infectious cell culture supernatants would have been detected if subjected to additional passages in culture . This situation does not have the same repercussions for DENV viral load determination in CCWB , since for that purpose , aliquots of CCWB were frozen after addition of Trizol as soon as the samples were received . Overall , the rates of RNA-reactivity for plasma specimens from the TMA-reactive subjects in the study were noticeably lower than those observed for the other components tested . While the original TMA was performed on plasma obtained from the EDTA whole blood testing tube , insufficient plasma remained following blood screening for us to extract and test by TaqMan . Rather , we had to use fresh frozen plasma from the blood unit . While precautions were taken to avoid formation of cryoprecipitate during thawing , it is possible that the lower proportion of reactives in the plasmas is due to poor recovery of virus associated with cryoprecipitate in some specimens . However , viral load determination for samples with Ct≤38 appears to have been unaffected or minimally affected by cryoprecipitate , since viral loads in reactive plasma specimens were comparable or in some cases higher than in the co-components . Therefore , comparison between DENV viral loads in plasma , serum and CCWB can be done since each specimen had been treated similarly , although different RNA extraction protocols were used for the liquid and cellular components analyzed and cryoprecipitate may have affected the rate of reactivity in plasma . DENV viral load determinations in clots were performed after a variable time of storage ( 0 . 5–10 months ) at 4°C and thus were subject to greater degradation over time . This may account for the lower average titers observed in this component when compared to other specimens ( i . e . , plasma , serum and CCWB ) , since this trend was not observed in the fresher clot samples available for analysis ( S2 Fig ) . These findings are relevant for the blood transfusion setting , since packed RBCs for transfusion may be stored refrigerated at 4°C for up to 42 days after collection . Although the stability of virions in the clot may not reflect that in packed RBCs , these findings warrant further investigation of viral persistence and infectivity under long-term storage under proper conditions for transfusion . Our results on DENV distribution in blood samples differed from the results reported when analyzing the distribution of the related viruses WNV , a flavivirus related to DENV , and HCV , another member of the family Flaviviridae . Our laboratory has previously demonstrated that WNV associates with RBCs in blood donors and WNV RNA can be detected in the cellular component of blood samples in concentrations up to 1-log10 higher than that observed in the plasma co-component of the same sample [28] . Conversely , for HCV , significantly more viral RNA was detected in plasma than in RBC specimens from the same collection [40] . The results on the distribution of DENV-1 and DENV-4 in blood samples from this study revealed that there are no significant differences in the detection of DENV viral RNA in individuals with quantifiable amounts of viral RNA in all tested components ( S1 Table and Fig 2 ) . Although WNV association with human RBCs relative to plasma was significantly higher than the association of DENV or HCV with RBCs , it should be noted that in general , viral loads of WNV in human infections are much lower than those of DENV and HCV [41–44] . Taking into consideration the viral loads during the course of human infection for these viruses it is possible to speculate that the cellular compartment of the blood becomes saturated more quickly in human infection with DENV and HCV due to intense viral replication and release , resulting in higher viral loads in the plasma components relative to those seen in WNV infection , in which humans are an incidental host vs a required host for viral dissemination for DENV and HCV . Alternatively , ligands in the membrane of RBCs could have differing affinity to structural proteins of DENV ( prM/M and E ) than to structural proteins of WNV and HCV . Another explanation would be that HCV may also have higher affinity for proteins that are freely circulating in the plasma than either DENV or WNV , and binding competition would cause equilibrium in viral distribution . Multiple studies have demonstrated an association of DENV viremia levels with disease severity , with patients with severe dengue generally having higher viral loads , but these studies of patients have not included asymptomatic infected individuals [44–47] . The infecting DENV virus type may also influence viremia levels in some populations [46 , 48] . The samples used in this study represented only individual samples from a given infected individual , and as such , represent a snapshot of the infection and are not appropriate for determining peak viremia or viral persistence . However , the levels of DENV viremia observed in our specimens were within ranges previously reported for serum viremia in symptomatic individuals [44 , 46] . Overall , the finding of comparable titers between different components from the same blood sample from DENV-1 and DENV-4 infected blood donors suggests that sensitivity of NAT for these viruses would not be improved by testing whole blood or components other than plasma . The detection of infectious DENV virions in all tested blood components including clots is congruent with previous findings of TT-DENV in all transfused components [13–16] . Additionally , the detection of DENV RNA and even infectious virions in clots after extended refrigerated storage suggests that this material has potential for use in testing when resources and sample sizes are limited . | Dengue is a febrile disease caused by the four dengue viruses ( DENV-1 to 4 ) transmitted by mosquitoes from the genus Aedes that can also be transmitted by blood transfusion and organ transplantation . DENV is present in the blood of infected individuals without symptoms , meaning that infected donors may pose a risk to the safety of the donor blood supply . Current methods for detecting transfusion-transmitted viruses by nucleic acid testing use plasma as the testing specimen , and the number of tests that can be performed without reducing availability of blood for transfusion is limited . To determine whether blood components other than plasma could be suitable for testing , we quantified and compared the concentrations of DENV RNA in the residual components of blood collected from subjects previously identified as infected in a parallel blood safety study . Additionally , when available , samples were also evaluated for infectivity in tissue culture . The results showed that DENV RNA and infectious virions were detected comparably in all blood components , suggesting that using alternate specimens may improve sample availability but may not improve testing sensitivity . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"pathogens",
"immunology",
"rna",
"extraction",
"microbiology"... | 2016 | Distribution of Dengue Virus Types 1 and 4 in Blood Components from Infected Blood Donors from Puerto Rico |
Inputs to signaling pathways can have complex statistics that depend on the environment and on the behavioral response to previous stimuli . Such behavioral feedback is particularly important in navigation . Successful navigation relies on proper coupling between sensors , which gather information during motion , and actuators , which control behavior . Because reorientation conditions future inputs , behavioral feedback can place sensors and actuators in an operational regime different from the resting state . How then can organisms maintain proper information transfer through the pathway while navigating diverse environments ? In bacterial chemotaxis , robust performance is often attributed to the zero integral feedback control of the sensor , which guarantees that activity returns to resting state when the input remains constant . While this property provides sensitivity over a wide range of signal intensities , it remains unclear how other parameters such as adaptation rate and adapted activity affect chemotactic performance , especially when considering that the swimming behavior of the cell determines the input signal . We examine this issue using analytical models and simulations that incorporate recent experimental evidences about behavioral feedback and flagellar motor adaptation . By focusing on how sensory information carried by the response regulator is best utilized by the motor , we identify an operational regime that maximizes drift velocity along chemical concentration gradients for a wide range of environments and sensor adaptation rates . This optimal regime is outside the dynamic range of the motor response , but maximizes the contrast between run duration up and down gradients . In steep gradients , the feedback from chemotactic drift can push the system through a bifurcation . This creates a non-chemotactic state that traps cells unless the motor is allowed to adapt . Although motor adaptation helps , we find that as the strength of the feedback increases individual phenotypes cannot maintain the optimal operational regime in all environments , suggesting that diversity could be beneficial .
Escherichia coli cells navigate their environment by alternating straight runs with direction-changing tumbles to perform a random walk . During a run , the flagellar motors spin counterclockwise ( CCW ) and propel the cell at constant speed in one direction , which changes slowly due to rotational diffusion . Runs are terminated when one or more motors start rotating clockwise ( CW ) , which causes the cell to tumble [1]–[3] . Cells are able to bias their random walk toward favorable conditions using a two-component signal transduction pathway that detects changes in signal intensity during runs and modulates the probability to tumble accordingly , resulting in extended runs in the desired direction and net drift velocity in the direction of the gradient [4] . The sensory module of the chemotaxis pathway ( Figure 1A ) consists of large clusters of receptor proteins that bind signal molecules to modulate rapidly ( <0 . 1 s ) the activity of an associated histidine kinase , CheA [5]–[7] . The high gain of the receptor cluster is coupled to negative integral feedback control [8]–[10] , mediated by slow ( ∼1–30 seconds ) methylation and demethylation of the receptors by CheR and CheB , respectively [11]–[13] . This allows the receptors to adapt to a constant background signal while maintaining sensitivity over a wide range of concentrations [14] , [15] . For example , when cells are stimulated with a step of aspartate , the activity of the receptors returns nearly precisely to its pre-stimulus level after a transient response ( Figure 1B first line ) . While precise adaptation does not hold when receptors become saturated , adaptation with a precision above 80% has been measured for many relevant signals within the micromolar range [16] . Precise adaptation is an important feature of bacterial chemotaxis because it provides robustness by implementing a “maximin” strategy that guarantees at least minimum chemotactic performance in any environmental condition [17] . The activity of the sensory module is relayed through a diffusible response regulator CheY to the flagellar motors , which act as the actuator , ( Figure 1A ) . When phosphorylated by CheA , CheY-P binds to the motor subunit FliM and increases the probability of the motor to switch from CCW to CW [18] . Fast dephosphorylation of CheY-P by the phosphatase CheZ ensures rapid transfer of information from the sensor to the actuators . The CW bias of the flagellar motor , which defines the tumbling probability [2] , is a sensitive function of the CheY-P concentration ( Hill coefficient >10 , Figure 1A ) [19] , [20] . The capability of the system to maintain the CheY-P concentration within the tight dynamic range of the motor CW bias response function ( Figure 1A ) is often used to investigate robustness to fluctuations in protein concentrations and receptor activity [21] , [22] . An important underlying assumption is that performance is maximized when the motor converts small variations in CheY-P into large changes in CW bias . However , recent experiments and theory suggest that the coupling between sensor and actuator is more complex than previously thought . First , the flagellar motors partially adapt to persistent stimulus [23] , [24] . Second , the motor CW bias response to CheY-P is steeper than previously reported , further restricting the dynamic range of the motor response to CheY-P fluctuations [20] . Finally , in exponential ramps of chemoattractant , the CheY-P concentration reaches a dynamical equilibrium , Ym , hereafter called operational CheY-P , distinct from the adapted CheY-P concentration , Y0 , that the cell maintains in constant uniform environments [25] , [26] ( Figure 1B second line ) . For each of these three findings , the characterization of the internal dynamics of the signaling pathway was performed on immobilized cells using experimentally controlled input signals . However , for cells swimming freely in chemical gradients , the input signal dynamics are determined by the chemotactic response of the cell , creating a feedback of the behavior onto the input signal [27] ( Figure 1B third line ) . Because of this behavioral feedback , it remains unclear how the multiple time scales of the system , from signal detection to motor response , ultimately determine chemotactic drift performance . Here , we use analytical models and stochastic simulations of individual cells to examine the consequences of these new observations for our understanding of the bacterial chemotaxis strategy . Clonal populations of chemotactic E . coli grown in homogeneous conditions exhibit significant cell-to-cell phenotypic variability , with adaptation times ranging from 1 to 30 seconds [28]–[31] , and motor clockwise bias ranging from 0 . 1 to 0 . 4 [3] , [31] . Therefore , we consider how different combinations of adaptation times and motor clockwise biases , which define a cell's phenotype , affect individual cell chemotactic drift velocity in different environments . In a phenotypically diverse population , different phenotypes of that population may perform best in different environments . Focusing on how information transfer from sensor to actuator affects chemotactic performance , we analyze the dynamical relationship between the operational regime of CheY-P , Ym , and the drift velocity , VD , as a function of the phenotype and gradient steepness . We show that there is a unique operational regime of the sensor with respect to the motor that maximizes drift velocity in the direction of the gradient by maximizing the contrast between runs up and down the gradient , and not by maximizing the CW bias response . We characterize the performance trade-off faced by individual cells with different combinations of phenotypic parameters ( such as , adapted CheY-P concentration , Y0 , receptor adaptation time , τ , and cell resting tumble bias , TB0 ) .
Previous studies have examined how E . coli chemotactic drift velocity along a one dimensional gradient depends on the adaptation time [25] , [32] , the shape of the response function of the sensory module [17] , [33] , [34] and also behavioral feedback [27] . Instead , we first focus on how the coupling between the sensors and actuators by the response regulator CheY-P ( Figure 1A ) affects the chemotactic performance of individual cell phenotypes . What CheY-P concentration maximizes drift velocity of cells navigating exponential gradients of methyl-aspartate ? We examine this question using stochastic simulations of individual cells and an analytical model . Simulations were conducted using a standard model of the chemotaxis pathway in individual cells [2] , [15] as described in Methods . Receptor-kinase complex activity is modeled as an all-or-none response using quasi-equilibrium dynamics for fast ligand binding , chemoreceptor conformational changes , and phosphorylation cascade [15] . The slower ( de ) methylation kinetics follow simple negative integral feedback dynamics with adaptation rate τ−1 . The flagellar motor is modeled as an inhomogeneous Poisson process that switches cell behavior between runs and tumbles with rates defined as a function of CheY-P concentration that varies in time , Y ( t ) . The parameters of the motor model are calibrated to recent experimental measurements [19] , [20] , [23] , [24] . While motor adaptation [23] is not included at first , its effects are analyzed later in the paper . During runs , a cell swims with constant speed v = 20 µm−1 in a direction subjected to rotational diffusion ( rotation diffusion constant , Dr = 0 . 062 rad2 s−1 [1] ) . For simplicity the effects of multiple flagellar motors [2] , [35] or directional persistence [1] are not included but discussed in the Discussion section . Hence , in this model , motor clockwise bias ( CW ) and cell tumble bias ( TB ) are the same . We consider cells containing only Tar receptors and use methyl-aspartate as the ligand . Our results readily extend to more complex receptor cluster configurations . Three-dimensional trajectories of individual cells were simulated as described in [2] for various cell phenotypes , which are characterized by the receptor adaptation times τ and adapted CheY-P concentrations Y0 , in gradients of chemoattractant of different steepness , g . Following previous studies [25]–[27] , we used exponential gradients ( L ( x ) = L0egx ) so that cells experience an approximately constant “force” from the attractant field , as the chemotaxis system is a fold-change detector ( Eq . ( 2 ) below ) . This makes it possible to define a steady-state drift velocity , making the problem analytically tractable . The performance of each cell phenotype , which is defined by a unique adapted CheY-P concentration ( Y0 ) and receptor adaptation time ( τ ) , in each gradient steepness ( g ) is defined as the drift velocity VD ( Y0 , τ , g ) along the gradient direction calculated by averaging the velocity of 10 , 000 phenotypically identical cells over 4 minutes ( Methods ) . The first simulations were done in a relatively shallow gradient g−1 = 5 , 000 µm , with adaptation times of τ = 5 , 10 , and 30 s , and adapted CheY-P concentrations spanning the range Y0 = 1–4 µM . Plotting drift velocity as a function of the adapted CheY-P concentration reveals that maximal drift velocity is not achieved for CheY-P concentrations in the linear range of the CW bias response curve , where fluctuations in CheY-P result in large changes in clockwise bias ( Figure 2A ) . Instead , it occurs when adapted CheY-P is at the lower end of this curve ( around 2 . 4 µM in Figure 2A ) . To understand the underlying reasons of this result , we derived an analytical relationship between CheY-P concentration and drift velocity along a one-dimensional gradient . For simplicity we used a one-dimensional analytical representation of bacterial chemotaxis in two or three dimensions [27] , [33] , [34] , [36] , [37] . In this framework , cells either go up or down the gradient or tumble and the effect of rotational diffusion can be represented as a jump process between runs up and runs down the gradient with transition rate ( d-1 ) Dr , where d represents the number of spatial dimension [36] . At quasi-steady state ( for time scales longer than single run durations ) and with no directional persistence ( equal probability to run up or down the gradient ) , the drift velocity is proportional to the cell swimming speed v times the difference between the expected run durations up and down the gradient divided by the total time including the time spent tumbling [36]: ( 1 ) The only difference between d = 2 or d = 3 dimensions is a rescaling of the drift velocity and rotational diffusion ( factor d in the equations above; see Text S1 ) . The expected run duration up or down the gradient is controlled by the cellular concentration of CheY-P , Y . This quantity is in turn a function of the free energy difference between the inactive and active receptor complexes , F , such that F = ln ( α/Y-1 ) , where α is the gain of the phosphorylation cascade . The receptor activity follows simple spring-like dynamics around the adapted free energy difference F0 with adaptation time τ ( Methods , our Results still hold when considering asymmetric methylation/demethylation rates , see Text S1 and Figure S1 ) : ( 2 ) Here s = ±1 or 0 for cells running up , down the gradient , or tumbling , respectively . As the cell moves along a trajectory x ( t ) it encounters different concentrations of the ligand L ( x ( t ) ) . represents the magnitude of the change in free energy difference and depends on the local steepness of the gradient at the cell position . Here , v is the speed of the cell when it is swimming , N is the gain of the cooperative receptor system and Ki and Ka are the dissociation constants between ligand and receptors in the inactive and active conformation . In general , both s and f change as a function of time . For ligand concentrations Ki<<L<<Ka we have . If in addition , the gradient of ligand is exponential , L = L0egx , we see that f≈vNg becomes constant where g represents the inverse length scale of the gradient . Therefore , in an exponential gradient the free energy difference of the receptors , F , tends to increase at the constant positive rate +vNg when the cell swims up the gradient and to decrease at the constant negative rate −vNg when the cell swims down the gradient . This in turn causes the CheY-P concentration to decrease ( increase ) when cells swim up ( down ) the gradient . The exact CheY-P concentration trajectories can be calculated by integrating Eq . ( 2 ) as a function of time for different initial condition Fi ( Figure 2B ) , while a cell is swimming up ( s = 1 green curve ) or down ( s = −1 red curve ) a gradient . The expected durations of a run , , or a tumble , , are plotted as a function of CheY-P concentration in Figure 2B ( dashed lines; see definition in Methods ) . When a cell runs up or down the gradient , the rates of switching from one state to another change as a function of time and the direction of motion because they depend on the CheY-P trajectory . A run up or down the gradient can also be terminated by random reorientation from rotational diffusion with rate ( d-1 ) Dr [36] . Altogether , the rate at which a run is terminated by either rotational diffusion or a tumble is thus . In a shallow gradient , F deviates little from the adapted value F0 and the adapted value of τR provides a good approximation of the expected run duration along a direction ( black line in Figure 2B ) : , where . When swimming up or down the gradient , CheY-P fluctuates ( Figure 2B , green lines for up , red lines for down ) and the run lengths are modulated approximately following τR0 ( black line and red and green circles in Figure 2B ) . According to equation ( 1 ) , drift velocity is largest where the contrast between run duration up and down the gradient is the largest . Figure 2B reveals that this is the case where the slope of the expected run length as a function of CheY-P concentration is largest , which corresponds to the foot of the motor CW bias curve ( Figure 2A ) in agreement with the simulations . In contrast , for higher valued of CheY-P that are within the dynamic range of the CW bias response function , ( e . g . Y0 = 3 µM in Figure 2B ) run durations up and down the gradient have a smaller contrast and longer tumble duration ( dashed line Figure 2B ) , resulting in slower drift velocity . In the limit of shallow gradients , Equation [1] can be linearized around the adapted values F0 and Y0 to obtain the drift velocity ( Methods and Text S1 ) : ( 3 ) Here , represents the tumble bias of the cell as a function of the adapted CheY-P concentration Y0 and the subscript 0 indicates that the rates λR and λT and are all evaluated at the adapted state . The integral in Eq . ( 3 ) is the time-averaged input over the run durations , which in this approximation are exponentially distributed with characteristic time scale τR0 . For Ki<<L<<Ka and exponential gradients , the rate of change of the free energy difference is constant during a run . Equation ( 3 ) indicates that the drift velocity is proportional to the gradient steepness g and the gain of the receptor cluster N . From Equation ( 3 ) we also obtain the chemotaxis coefficient of an individual cell phenotype: . Plotting the drift velocity as a function of Y0 on top of the simulation results in Figure 2A shows that Equation ( 3 ) provides a good prediction of the drift velocity in shallow gradients and confirms that maximum velocity is reached for CheY-P values at the foot of the CW bias response curve . In this linear regime , the optimal CheY-P concentration is only weakly dependent on the cell adaptation time and does not depend on the gradient steepness . The factor in Equation ( 3 ) encapsulates the relationship between drift velocity and the CheY-P concentration ( or free energy difference F ) . For small adaptation times , τ<<τR0 , it increases linearly with adaptation time and is maximum where the slope is largest . For larger adaptation times , τ>>τR0 , this factor becomes . Because the response function of the motor ( defined in Methods ) is very steep ( dashed line in Figure 2B ) , the slope ( slope of black line in Figure 2B ) changes much faster as a function of Y0 than Y0 ( Y0/α-1 ) . Because rotational diffusion imposes an upper bound on the run length along a given direction [27] , [32] , it determines , along with the motor parameters , the optimal range for CheY-P fluctuations . As rotational diffusion becomes smaller , cells are able to maintain their original direction for a longer time . The upper bound on the run length therefore becomes longer ( dashed lines in Figure 2C ) and the optimal CheY-P concentration becomes smaller ( full lines in Figure 2C ) . Changes in the switching frequency of the flagellar motor , ω , ( see Methods ) also affects the optimal CheY-P concentration . This becomes apparent when considering that the rate of switching from run to tumbles scales linearly with the switching rate of the flagellar motor , ω ( see Methods ) . Therefore , the expected run duration of a cell scales like the inverse of ω . The result of this scaling is that for increasing values of ω the inflection point of the expected run length as a function of CheY-P shifts to lower values of CheY-P ( dashed lines in Figure 2D ) . Thus , increasing the switching rate of the flagellar motors tends to decrease the optimal CheY-P concentration ( full lines in Figure 2D ) . It also increases the maximum drift velocity that can be reached . The analytical model of drift velocity in Eq . ( 3 ) is different from previous results [27] in two ways . First , it takes into account both the adaptation time and the tumbling state of the cell . Taking the limit , in our model we recover the previous results . Second , in the previous study , the switching rate of the motor λR was a steep function of kinase activity centered at the adapted kinase activity level , or equivalently , at the adapted CheY-P concentration Y0 . This means that changing Y0 would also change the set point of the motor . However , the adapted CheY-P concentration and the set point of the motor are independent parameters . For this reason here we focused on the relationship between the set point of the motor and CheY-P activity . According to Eq . ( 3 ) the flagellar motor has its own sensitivity set point independent of the adapted CheY-P concentration of the sensory system Y0 ( see definition of λR in Methods; motor adaptation [23] , [38] is considered below ) . Experiments have shown that when immobilized cells are exposed to an exponential ramp of methyl-aspartate , CheY-P activity reaches a new steady-state , Ym , which is lower than its adapted activity , Y0 , because of the relatively slow adaptation rate of the system [9] , [26] ( Figure 1B second line ) . When cells are swimming in an exponential gradient ( Figure 1B third line ) , we expected a similar effect to take place because the average drift of an individual cell up the gradient will cause this cell to experience , on average , an exponential increase in ligand concentration as it makes its way up the gradient . While this effect should be minimal in a shallow gradient , it could become important in steep or rapidly changing gradients [27] , [39] , especially for cells with longer adaptation times . To investigate this issue we simulated cells swimming in a steeper exponential gradient ( g−1 = 1 , 000 µm ) . After less than one minute of simulation , cell populations ( 10 , 000 replicate trajectories for each phenotype ) reached a constant steady state drift velocity . We calculated the average ligand concentration that the cells encountered over time ( Figure 3A ) . This reveals that the swimming cells experience an average exponential increase in ligand concentration over time . This average input is similar to the signal experienced by immobilized cells exposed to temporal exponential ramps ( Figure 1B , second line ) [9] , [26] . However , for the swimming cells the ramp rate is dynamically determined by the average drift velocity in the direction of the gradient ( Figure 3A ) . Thus , for swimming cells the ramp rate depends on the feedback of the performance onto the input signal ( Figure 1B , third line ) . Consistent with experimental results obtained with immobilized cells exposed to exponential ramps [26] , the average CheY-P concentration in the swimming cells reaches a stable dynamical equilibrium , the operational value Ym , after an initial drop from the adapted CheY-P concentration ( Y0 ) ( Figure 3B ) . The fact that the operational CheY-P concentration is not the same as the adapted CheY-P concentration implies that an optimal choice of adapted CheY-P must take into account this behavioral feedback . While a phenotype may for example have an adapted CheY-P concentration equal to the optimal concentration ( ∼2 . 4 µM ) , during chemotaxis this level drops to an operational level lower than the optimum , hindering its performance ( Figure 3B , solid black line ) . This effect is intensified when the adaptation time of the receptor cluster increases ( Figure 3B , grey line ) . On the other hand , a phenotype with an adapted CheY-P concentration higher than the optimal concentration can approach the optimal operational CheY-P concentration as it reaches its steady-state drift velocity ( Figure 3B , dotted line ) . The difference between Y0 and Ym grows larger as drift velocity or the receptor adaptation time increase ( Figure 3C ) . To determine whether the adapted or the operational CheY-P concentration is the primary variable that controls the average drift velocity in exponential gradients , we simulated cell populations with different adapted CheY-P concentrations and calculated their respective operational CheY-P concentrations . In a steep gradient ( g−1 = 1000 µm ) , the optimal adapted CheY-P increased to ∼2 . 7 µM compared to ∼2 . 4 µM in a shallow gradient ( Figure 3D ) . However , the optimal operational CheY-P concentrations for steep and shallow gradients are identical ( Figure 3D ) . This suggests that a unique operational CheY-P concentration maximizes drift velocity in multiple gradients . The situation is different when the feedback is strong . In this case the signaling pathway fluctuates around the operational values Fm and Ym , rather than the adapted values F0 and Y0 . Therefore , we need to update the analytical model to describe the drift velocity , VD , as a function of Fm . If we linearize the drift velocity equation around Fm rather than F0 we obtain an equation identical to Eq . ( 3 ) but with the subscript 0 replaced by m and , , λRm , λTm , and TBm now functions of Fm . Knowing how VD depends on Fm is not enough to calculate the drift velocity . We also need an equation that describes how Fm depends on VD . To model the effect of a constant drift velocity along the chemical gradient on the activity of the receptor cluster , we can expand Eq . ( 2 ) around Fm and solve for quasi-steady state: ( 4 ) This expression quantifies the deviation between the operational free energy difference Fm and the adapted free energy difference F0 as a function of the drift velocity and is consistent with the results of our simulations ( Figure 3C ) and [27] . Equation ( 4 ) also makes clear that the strength of the feedback depends on adaptation time , the receptor cluster gain N , and the steepness of the gradient . Behavioral feedback strongly affects performance because it moves Ym away from the optimal operating point relative to the motor . This , in turn , affects the capability of the motors to best use the information carried by CheY-P . By explicitly taking into account the effect of the behavioral feedback onto the coupling between the operating regime of CheY-P and the motor , Eqs . ( 3 ) ( with 0→m ) and ( 4 ) extend previous studies [17] , [25] , [27] , [32]–[34] , [36] and reveal new possible dynamical regimes for the biased random strategy as shown below . For a given phenotype ( Y0 , τ ) and gradient length-scale , the steady state drift velocity is determined by the intersection of two curves ( Figure 4A ) . The first curve ( solid line in Figure 4A ) describes how the drift velocity depends on the operational CheY-P concentration , Ym . It is defined by Equation ( 3 ) ( with 0→m ) and its profile can be interpreted as follows . For very low values of CheY-P the cell never tumbles . Thus , the cell diffuses equally in all directions and the net drift along the gradient is zero . For high values of CheY-P , the cell tumbles all the time so drift is zero as well . In between these two extremes , drift velocity is maximized for a specific value of the operational CheY-P concentration . However , Ym is not an independent variable . As we showed above ( Eq . ( 4 ) ) , because of the feedback the behavior onto the input , the operational CheY-P concentration is itself a function of the drift velocity ( which can also be written as: ) . This equation defines the dashed line in Figure 4A , which intersects the horizontal axis at Y0 . Because each line in Figure 4A defines a relationship between VD and Ym , the intersection between the two lines fully determines the drift velocity and the operational CheY-P concentration ( black circle in Figure 4A ) for a given phenotype and gradient . When the feedback is weak ( τNg small , i . e . short adaptation time , small gain , or shallow gradient ) , the operational CheY-P concentration only exhibits a weak dependency on drift velocity and there is only one steady-state solution ( intersection ) . Therefore , an appropriate adapted CheY-P concentration could be selected to ensure that operational CheY-P concentration is approximately optimal at all times ( Figure 4A ) . When the feedback is stronger , drift velocity always acts as a negative feedback onto the operational CheY-P concentration . In contrast , the effect of the operational CheY-P concentration onto drift velocity depends on whether the operational CheY-P concentration is below or above the CheY-P concentration that maximizes drift velocity ( Figure 4A ) . Below this concentration , the system obeys negative feedback dynamics , whereas above it , the system obeys positive feedback dynamics . This positive feedback loops combined with the non-linear decrease of the drift velocity as a function of the operational CheY-P concentration , which arise from the extreme sensitivity of the flagellar motor , can lead to bistability [40] . Indeed , for a stronger feedback ( steeper gradient or longer adaptation time ) the slope of the feedback curve ( dashed line in Figure 4AB ) , which is proportional to 1/τNg , decreases . Thus , for phenotypes with high enough adapted CheY-P concentration ( Y0 is the intersection of the dashed line with the horizontal axis ) , the two curves can intersect more than once ( Figure 4B ) . In this case , a single phenotype can now experience three different chemotactic states . Two of these states , Ym1 and Ym2 ( filled circles in Figure 4B ) , are stable and are separated by one unstable state ( open circle in Figure 4B ) . For one of the stable solution , the drift velocity is nearly zero and Ym1 is high and very close to the adapted CheY-P concentration . For the other stable solution , the drift velocity is large and Ym2 is much smaller than the adapted CheY-P concentration . This analysis suggests that an individual phenotype might experience two different chemotactic states with dramatically different performance: a fast drifting state and a “trapped” state . To find evidence of these two behaviors , we simulated two cell phenotypes ( 10 , 000 replicates for each phenotype ) in a steep exponential gradient ( g−1 = 1 , 000 µm ) . One phenotype was predicted to operate closer to the bifurcation than the other ( red and blue dots in Figure 4C , respectively ) . Although both phenotypes reached the same average operational CheY-P ( Ym = 2 . 3 µM ) , cells with a phenotype closer to the predicted bifurcation point ( Y0 = 3 µM , τ = 30 s ) exhibited a distribution of behavior ( both drift velocity and diffusion ) significantly skewed toward the “trapped” state ( Figure 4C ) . Closer examination of the trajectories and CheY-P dynamics of individual cells in this simulation reveals that individual cells transition stochastically back and forth between the “trapped” and fast drifting state ( Figure S2 ) . For cases with higher feedback strength cells spend more and more time within the “trapped” state . When the feedback is strong and the system becomes multistable , the average includes cells in both the “trapped” and high drift states . This phenomenon explains the decreased average drift velocities observed when adaptation time is increased ( above 10 seconds ) in a relatively steep gradient ( g−1 = 1 , 000 µm ) . It also explains the resulting shift of the best operational CheY-P concentration to lower concentrations ( from ∼2 . 4 to 2 . 1 µM in Figure 4D ) , since for phenotypes with lower values of the adapted CheY-P only one stable state exists . Similar results are obtained when asymmetric methylation/demethylation rates of the receptors are taken into account ( Figure S3 ) . Recent experiments have shown that the number of FliM monomers in the C-ring of the flagellar motor slowly ( ∼minutes ) adapts as a function of the CW bias , affecting both the steepness and the half-maximum CheY-P concentration of the CW bias motor response curve [24] . To examine the effect of motor adaptation on the relationship between CheY-P concentrations and drift velocity , we added motor adaptation to our stochastic model of an individual chemotactic cell by taking into account recent experimental data [20] , [23] , [24] ( Methods ) . The resulting CW bias response curve of the adapted motor agrees well with both recent [20] and earlier [19] experimental measurements . In fact , it matches earlier experiments [19] better than a simple Hill function , suggesting that in these experiments the individual motors measured had adapted to the particular concentration of CheY-P expressed in the corresponding individual cells ( Figure 5A; Methods ) . Simulations of cells with motor adaptation in a shallow gradient ( g−1 = 5 , 000 µm ) show that motor adaptation changes the shape of the drift velocity curve as a function of operational CheY-P , especially at high CheY-P concentrations ( compare Figures 5B and 2A ) . These results are predicted by the analytical model ( Eq . ( 3 ) with 0→m ) once modified to include motor adaptation ( Methods; lines in Figure 5B ) . Setting the adapted activity of the motor for a given CheY-P concentration to lower or higher CW biases gives qualitatively equivalent results ( Figure S4 and S5 ) . How does the motor adaptation affect the bifurcation ? In a steep gradient the behavioral feedback ( Eq . ( 4 ) ) must be taken into account ( Figure 5C dashed line ) . Comparing Figure 5C and 4B we see that motor adaptation enable cells with high adapted CheY-P concentration to avoid the chemotactic trap improving performance ( see Figure S6 ) . This should provide a selective advantage because it helps buffer the functional consequences of inevitable cell-to-cell variability in the adapted CheY-P concentration , by increasing the range of CheY expression levels that allows effective chemotaxis . Motor adaptation also affects the optimal operational CheY-P concentration ( compare Figures 5B and 2A ) , shifting it to lower concentrations . When cells are drifting up a gradient , CheY-P drops to the operational CheY-P , causing the CW bias to drop . With motor adaptation , the lower operating CW bias causes the motor to shift its sensitivity to a lower CheY-P concentration . We see again that maximal performance is reached for Ym at the bottom of the CW bias response curve of the now adapted motor ( Figure 5A ) . However , the motor can only compensate partially for the shift in operational CheY-P concentration . As long as the system does not undergo bifurcation , maximum drift velocity is achieved by having a long adaptation time while maintaining the operational concentration of CheY-P in the optimal range . Therefore , the optimal adapted CheY-P concentration depends on the gradient length-scale and the adaptation time ( Figure 6 ) . In shallow gradients , the strength of the feedback is small , as is the difference between operational and adapted CheY-P . Thus , it is possible to select an adapted CheY-P concentration that will perform relatively well for multiple adaptation times ( Figure 6A blue line ) . In steeper gradients , the feedback is stronger ( Eq . ( 4 ) ) and the difference between Ym and Y0 grows larger with adaptation time . Maintaining the optimal operational CheY-P concentration requires a higher adapted CheY-P concentration ( Figure 5A green and red ) . The bifurcation of the system imposes an upper bound on the range of Y0 beyond which a portion of the cells spend a significant amount of time trapped into a non-optimal state even with motor adaptation ( Figure 6A dashed lines ) . Therefore , the optimal adapted CheY-P concentration is a function of both receptor adaptation time and gradient length-scale , making it difficult for a single phenotype to maximize drift velocity in multiple environments ( Figure 6A ) . To characterize the resulting performance trade-off and map it to phenotypic space , we calculated the contours of drift velocity relative to its maximum in each environment , as a function of adaptation time τ and the adapted cell tumble bias ( Figure 6B ) . In shallow gradients , cells benefit from a relatively long adaptation time and a low adapted CW bias . In steep gradients , cells benefit from a short adaptation time and a higher adapted CW bias . The best generalist phenotype can achieve at most 60% relative performance in all three gradients considered here . Motor adaptation , which was taken into account in generating Figure 6 , alleviates only partially the tradeoff faced by single cells .
The adaptive response and feedback control of the receptor cluster play a critical role in the robustness of the chemotaxis system [8] , [10] , [15] , [17] , [33] . However , chemotactic performance also relies on the optimal operation of the flagellar motors , which directly control cell behavior . By focusing on how the CheY-P concentration affects the coupling between sensors and actuators , we revealed the existence of an operational regime for CheY-P concentration , which is distinct from the adapted CheY-P concentration , that maximizes drift velocity in a wide range of gradient length-scales and receptor adaptation times . Fluctuations around the best operational CheY-P concentrations maximize the contrast between run duration up and down the gradients . This occurs outside the most sensitive region of the CW bias response curve of the motor . Thus , chemotactic performance relies on maintaining the operational CheY-P concentration within bounds [21] , [22] , [41] around this optimal value . The best operational CheY-P concentration is also determined by the cell rotational diffusion constant Dr , which imposes an upper bound on run durations in any particular direction ( Figure 2C ) [27] , [32] . In a more viscous environment or for longer cells , the lower rotational diffusion will result in a lower optimal operational CheY-P concentration . For an ellipsoid , rotational diffusion is inversely proportional the length of the major axis . Therefore , as cells grow the optimal range will shift to lower CheY-P concentrations . If the cell maintains a constant amount of CheY as it grew , the effective concentration of CheY-P would decrease , resulting in robust chemotactic performance during cell growth . The switching frequency of the flagellar motors also affects the best operational CheY-P concentration . Higher switching frequencies tend to increase drift velocity while shifting the maximum to smaller CheY-P concentrations ( Figure 2D ) . Therefore , the best operational CheY-P concentration is further away from the motor threshold . However , the range of CheY-P concentrations where the drift velocity is high becomes narrower ( because the expected run length becomes a steeper function of CheY-P ) . This tends to increase the performance trade-off between different gradient length-scales . Thus , while selecting a higher switching frequency for the flagellar motors may improve performance of some phenotypes it may be detrimental for the population overall . Another important consideration is that the switching frequency is bounded by the speed at which the motor and associated flagella can switch confirmation [2] , [42] . Directional persistence ( amount by which the swimming direction of a new run is biased towards the swimming direction of previous run ) has been shown to affect chemotactic performance in climbing shallow gradients of attractants [1] , [43]–[45] . However , previous modeling and simulations efforts have been done using cells with non-optimal CheY-P concentrations ( usually at 3 µM ) . In this regime , cells have a high tumbling rate , short run lengths , and low drift velocity . Directional persistence effectively reduces the reorientation rate of cells [45] , which is equivalent to reducing the tumbling rate slightly . Therefore , directional persistence will shift the optimal CheY-P concentration to higher concentrations and improve the drift velocity of frequently tumbling cells [44] . On the other hand , when cells operate at or close to the optimal CheY-P concentration , the tumbling rate is low . Therefore , the run length in a given direction is terminated by rotational diffusion and not by tumbles . For optimal phenotypes , the relative effect of directional persistence on chemotactic drift is thus less important . Previous studies have examined how the adaptation time affects chemotactic performance [12] , [25] , [27] , [32] , [46] . However , these studies only considered single values for the adapted CheY-P concentration ( typically set to a CW bias of 0 . 5 ) and concluded that adaptation time should decrease as gradients get steeper to keep the operational CheY-P concentration within the dynamic range of the motor CW bias response . We found that , as long as the cell can maintain the optimal operational CheY-P concentration , longer adaptation time is better because it enhances input signal over the course of a run . However , long adaptation reinforces the feedback from the cell drift velocity on the system and can lead to undesirable bistability . Therefore , the bifurcation boundary imposes an upper limit on adaptation time as a function of the gradient length-scale . Interestingly , the distribution of tumble bias typically observed during exponential growth in single E . coli cells ranges from 0 . 1 to about 0 . 4 and not many cells are found that have higher tumble bias [31] . Selection for cells with tumble bias below 0 . 4 is consistent with our finding that the performance of cells with higher tumble bias will suffer from the existence of the “trapped” chemotaxis state . Our results also provide a strong justification for the role of the recently-discovered flagellar motor adaptation . Indeed , we found that motor adaptation [23] , [38] plays a significant role in mitigating the behavioral feedback for cells with high tumble bias . When such feedback was included , cells with high tumble bias could escape the “trap” and gain access to a high drifting state in steep gradient . Our model also resolved an apparent contradiction between two sets of experimental measurements of the CW bias response of the flagellar motor as a function of CheY-P concentration . While one measurements reported a Hill coefficient of n = 10 [19] , newer experiments reported a Hill coefficient of n = 20 [20] . In this paper we used the new value n = 20 and showed that the previous measurements are fitted with the same parameter value if one makes the reasonable assumption that the motors had had time to adapt before each individual cell measurement ( Figure 5A ) . Because the difference between the operational and adapted CheY-P concentrations depends on the strength of the behavioral feedback , which itself is proportional to gradient steepness , different adapted CheY-P concentrations and adaptation times are required to perform optimally in different gradients . Thus , in conditions where drift velocity is important , cells are faced with a performance trade-off . Even though motor adaptation was included , the best compromising phenotype over the gradient steepness considered in this study achieved at most 60% of the theoretical maximal drift velocity in all gradients . The observed cell-to-cell phenotypic diversity in adaptation time and adapted tumble bias [29] , [31] in an isogenic population may resolve the performance trade-off faced by single cells to improve the chance of survival of a unique genotype in complex or varying environments . In addition , the negative correlation between tumble bias and adaptation time observed by Park et al . in an isogenic population of E . coli [31] , is consistent with our predictions about the most beneficial way to distribute phenotypes ( Figure 6B ) . At its core , the biased random walk relies on the dynamical control of the probability of reorientation . Overall , our analysis reveals limits to the use of negative integral feedback to control such strategy . Because the biased random walk strategy is used by many organisms , these results will inform our understanding of the constraints faced by other organisms as well .
We used a standard model of bacterial chemotaxis [15] as described in [2] . For a cell following the trajectory x ( t ) , the output of the sensory module , the CheY-P concentration , is where the free energy difference between inactive and active receptor complexes , , is a function of the methylation level , m ( t ) and ligand concentration L ( x ( t ) ) . With α = 6 µM , ε0 = 6 , ε1 = −1 , N = 6 , and Ki = 0 . 0182 mM , Ka = 3 mM for methyl-aspartate and Tar receptors in the inactive and active conformation . When the cell is adapted to its environment , . Adaptation mediated by methylation and demethylation of the receptors follows , where . The methylation level m is positive and bounded by the total number of methylation sites mmax = 48 available in a cooperative unit of receptors . The resulting adaptation dynamics fits recent experiments [26] . Cells switch between runs , R , and tumble , T , with rates . The motor is modeled as a bistable system with switching frequency ω = 1 . 3 s−1 ( unless otherwise stated ) and free energy difference where ε2 and ε3are non-dimensional constants that control the basal rate of switching of the motor when Y = 0 and the degree of cooperativity of the motor , respectively . K is the binding constant of CheY-P to FliM at the base of the motor . With ε2 = ε3 = 80 , and K = 3 . 06 µM , this coarse-grained motor model fits well recent experimental measurements of CW bias ( Hill coefficient 20 ) and switching frequency [20] , [23] , [24] . Motor adaptation is considered below . Eq . ( 2 ) follows by taking the time derivative of F and using the relations from the previous section . Integration of Eq . ( 2 ) gives: . The expected duration of a run along the direction s = ±1 is determined by the integral of the rate of terminating a run along the direction s by tumbling or because of rotational diffusion: ( 5 ) Because the average cell drift velocity in the direction of the gradient is determined by the contrast between expected run durations up and down the gradient ( Equation ( 1 ) ) , the quantity of interest to calculate from Equation ( 5 ) is . In a shallow gradient , the deviations from the adapted free energy difference F0 are small . Considering only first order deviations around F0 the change in free energy as a response to changes in ligand concentration is small and the inverse of the rate of run termination can be approximated by where the mean run duration along a direction and the gain are evaluated at F0 . Similar linear expansions are carried out for λR and λT . Linear expansion of the free energy difference in Eq . ( 5 ) and integration by part gives . For tumble , . Inserting in Eq . ( 1 ) and using the solution F ( t , s , Fi ) we obtain the drift velocity to first order in ΔF ( Eq . ( 3 ) ) . The number of FliM molecules in the motor , n , is modeled as a binding and unbinding process with CW bias dependent rates [38]: . The constants kon and koff define the rate of adaptation of the motor . n1 and n2 are the minimum and maximum FliM ring size that a motor can accommodate . Δn is an effective half max parameter that guarantees that the effective rates of unbinding and binding to the motor go to zero when n approaches n1 or n2 . When n changes it affects the steepness of the motor CW bias response , , which in our case is controlled by ε3 ( see above ) . We used a simple linear relationship where ε3 , 1 is the slope and n0 and ε3 , 0 are the pre-stimuli level of the number of FliM and motor steepness , respectively . kon = 0 . 025 s−1 , n1 = 34 , n2 = 44 from [24] . We choose ε3 , 0 = 80 , n0 = 36 to match the Hill coefficient of 20 measured for individual motor response curves [20] , and fit Δn = 4 . 16 , ε3 , 1 = 1 . 96 to reproduce [19] ( Figure 5B inset ) . koff = 0 . 0063 s−1 controls the CW bias that the motor adapts to ( 0 . 2 in this case , typical for wild type population of E . coli selected for swimming on agar plates [31] ) . At steady state , dn/dt = 0 defines CW ( n ( ε3 ) ) ( Eq . [S20] in Text S1 ) . On the other hand , assuming quasi-equilibrium between the motor and operational CheY-P concentration Ym , we have . Solving the two equations gives ε3 as a function of Ym from which we can calculate the drift velocity as Eq . ( 4 ) with motor adaptation ( Figure 5 ) . | The biased random walk is a fundamental strategy used by many organisms to navigate their environment . Drift along the desired direction is achieved by reducing the probability to reorient whenever conditions improve . In the chemotaxis system of Escherichia coli , this is accomplished with a sensory module that implements negative integral feedback control , the output of which is relayed to the flagellar motors ( the actuators ) by a response regulator to control the probability to change direction . The proper dynamical coupling between sensor and actuator is critical for the performance of the random walker . Here , we identify an optimal regime for this coupling that maximizes drift velocity in the direction of the gradient in multiple environments . Our analysis reveals that feedback of the behavior onto the system in steep gradients can constrain individual cell performance , by causing bi-stable behavior that can trap cells in non-chemotactic states . These limitations are inherent in the biased random walk strategy with integral feedback control , but can be alleviated if the output of the pathway adapts , as recently characterized for the flagellar motors in Escherichia coli . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"escherichia",
"coli",
"systems",
"biology",
"computer",
"and",
"information",
"sciences",
"prokaryotic",
"models",
"model",
"organisms",
"medical",
"microbiology",
"network",
"analysis",
"microbial",
"pathogens",
"biology",
"and",
"life",
"sciences",
"microbiology",
"c... | 2014 | Limits of Feedback Control in Bacterial Chemotaxis |
An organism's ability to thrive in changing environmental conditions requires the capacity for making flexible behavioral responses . Here we show that , in the nematode Caenorhabditis elegans , foraging responses to changes in food availability require nlp-12 , a homolog of the mammalian neuropeptide cholecystokinin ( CCK ) . nlp-12 expression is limited to a single interneuron ( DVA ) that is postsynaptic to dopaminergic neurons involved in food-sensing , and presynaptic to locomotory control neurons . NLP-12 release from DVA is regulated through the D1-like dopamine receptor DOP-1 , and both nlp-12 and dop-1 are required for normal local food searching responses . nlp-12/CCK overexpression recapitulates characteristics of local food searching , and DVA ablation or mutations disrupting muscle acetylcholine receptor function attenuate these effects . Conversely , nlp-12 deletion reverses behavioral and functional changes associated with genetically enhanced muscle acetylcholine receptor activity . Thus , our data suggest that dopamine-mediated sensory information about food availability shapes foraging in a context-dependent manner through peptide modulation of locomotory output .
Animals have a remarkable capacity for altering their behavior in response to changes in both their external environment and their internal physiological state . Such behavioral modulation is often achieved through the actions of neuropeptides . Neuropeptides often act by modulating the effects of fast synaptic signaling in order to alter neuronal excitability and neural circuit activity . Striking examples of this have come from pioneering studies of rhythmic motor activity underlying feeding in crustaceans where neuropeptides and other neuromodulators potently alter neural activity patterns [1] , [2] . The functional effects of specific neuromodulators can vary widely depending on levels of activity in target neurons as well as activity-dependent regulation of release , suggesting context-dependent modulation of circuit activity may enable flexible behavioral responses [3]–[5] . Local changes in food availability are among the most variable and significant environmental conditions that animals must cope with . Thus , mechanisms that regulate foraging behavior based on the availability of food are particularly important for survival . Recent studies using genetic approaches in worms , flies and mice have provided compelling evidence for state-dependent modulation of feeding and food-searching behavior [6]–[12] . In particular , these studies have elegantly demonstrated how neuromodulators signal information about internal factors , such as feeding state , and modulate sensory responsiveness to gustatory and olfactory stimuli . In contrast , we know relatively little about how sensory information signaling food availability is translated by the nervous system into alternative motor outputs such as those that underlie food searching . The nematode Caenorhabditis elegans provides an attractive system to address neuromodulatory mechanisms involved in generating context-dependent behaviors . C . elegans exhibit robust behavioral responses to changes in their environment and we have a growing understanding of the sensorimotor circuits involved . In particular , changes in food availability alter C . elegans movement via effects on both locomotion velocity and turning frequency . Most strikingly , removal of C . elegans from food initiates an alternative motor pattern in which animals restrict movement to their immediate environment . This behavior shares many features with a local foraging strategy , known as area-restricted search [13]–[16] that is observed across almost all animal species . Following removal from food C . elegans transiently increase their turning frequency and then , if unsuccessful in finding food , shift within minutes to longer runs of uninterrupted forward movement in order to disperse over larger areas . Thus , C . elegans movement is potently affected by external information about the availability of food in the local environment , and sensory information about food availability drives context-dependent behavioral transitions that are central to foraging . To elucidate mechanisms responsible for generating context-dependent modulation of behavior , we used a genetic strategy to identify candidate neuropeptides and assessed their role in modulation of C . elegans movement . We observed that dopaminergic regulation of neuropeptide signaling plays a central role in shaping context-dependent motor responses , such as those that underlie local food searching . These effects are mediated through the actions of the NLP-12 neuropeptide , a C . elegans homolog of mammalian cholecystokinin ( CCK ) . Under normal ( well-fed ) conditions , nlp-12 is not required for exploratory movement; however , a behavioral requirement for NLP-12 signaling is revealed by a genetic manipulation that produces increased synaptic activation of muscles . Release of NLP-12 , which is solely expressed in the interneuron DVA , is modulated by the D1-like dopamine receptor DOP-1 and both dop-1 and nlp-12 are required for normal foraging responses to food deprivation . While nlp-12 deletion impairs foraging through reductions in body bend depth and turning , nlp-12 overexpression is sufficient to stimulate deep body bends and enhance turning , and these effects require wild type acetylcholine receptor function in muscles . Thus , NLP-12/CCK shapes motor circuit responsiveness to sensory information through context-dependent modulation of locomotory output .
C . elegans moves sinusoidally by propagating waves of dorsoventral flexures along the length of its body . The pattern of C . elegans movement is profoundly affected by information from the environment . For example , mechanosensory stimuli such as those associated with mating and predation , or changes in food availability can each evoke alternative motor outputs . To precisely define how sensory stimuli conveying food availability affects sinusoidal movement , we monitored body bends immediately following removal from food using an automated tracking system . We found that removal of C . elegans from food produced a significant increase in body bend depth ( 40±5% increase , p<0 . 0001 ) ( Fig . 1A ) , suggesting that active modulation of body bends is an important component of C . elegans motor responses to changes in food availability . To further investigate how sensory perception of food may be translated into altered motor responses , we sought to develop a genetic strategy that would mimic aspects of the motor responses to food removal . We hypothesized that modulation of body bend amplitude may be achieved through increased synaptic activation of body wall musculature . To gain support for this idea , we used a genetic approach to enhance the activity of neuromuscular synapses . To increase neuromuscular synapse activity , we expressed an engineered form of a muscle ionotropic acetylcholine receptor ( iAChR ) with enhanced function . In C . elegans , two classes of iAChRs – levamisole sensitive ( L-AChR ) and nicotine sensitive ( N-AChR ) are expressed in muscle cells [17]–[19] . L-AChRs are heteropentameric complexes composed of the subunits UNC-38 , UNC-63 , LEV-1 , LEV-8 and UNC-29 [20] . Loss of L-AChR function leads to a characteristic reduction of cholinergic currents recorded from muscles and sluggish movement [18] , [21]–[23] . To determine if we could enhance synaptic activation of muscles by increasing L-AChR function , we introduced an amino acid change at a highly conserved position in the pore-lining M2 regions of cDNAs encoding the UNC-29 , UNC-38 and LEV-1 subunits ( Fig . S1A ) . Substitution of a polar amino acid ( e . g . serine ) for the native residue at this position ( typically leucine ) produces increased activation of mammalian iAChRs [24]–[27] . Specific expression of individual pore-modified L-AChR subunits in muscles led to phenotypic changes consistent with increased muscle excitability ( e . g . altered movement ) . To enhance this effect , we constructed a strain ( ufIs6 ) stably expressing the unc-29 ( L/S ) , unc-38 ( V/S ) and lev-1 ( L/S ) cDNAs under control of the muscle-specific myo-3 promoter , allowing for incorporation of 3 pore-modified subunits into each pentameric receptor complex . Hereafter , we refer to this transgenic strain as L-AChR ( gf ) . Visual analysis of muscle structure and neuromuscular synapses indicated L-AChR ( gf ) expression did not produce obvious muscle toxicity or disruption of neuromuscular synapse development ( Fig . S1B , C ) . Moreover , animals expressing L-AChR ( gf ) receptors were hypersensitive to the paralyzing effects of the L-AChR specific ligand levamisole , indicating this manipulation produced the predicted activating effects on the receptor ( Fig . S2A , B ) . To quantify behavioral changes produced by L-AChR ( gf ) expression , we monitored the exploratory movements of transgenic animals . L-AChR ( gf ) expression produced a significant increase in body bend amplitude compared to the wild type ( 48 . 3±5% increase , p<0 . 0001 ) , while movement velocity was decreased ( Fig . 1B–D and Fig . S2C ) . These results suggested L-AChR ( gf ) expression enhanced muscle contractions in response to synaptic ACh release from motor neurons , altering the sinusoidal motor pattern . Consistent with this , L-AChR ( gf ) animals were shorter in body length compared to the wild type , likely as a consequence of muscle hypercontraction ( Fig . S2D ) . In contrast , an integrated transgene ( ufIs47 ) encoding wild type copies of the same L-AChR subunits under control of a muscle-specific promoter did not produce strong effects on any of the measures we made , supporting the notion that the behavioral changes were caused by enhanced function of the engineered receptor . Finally , electrophysiological recordings from body wall muscles demonstrated that L-AChR ( gf ) expression prolonged the duration of current responses to levamisole , providing direct evidence that the engineered mutation altered muscle synapse activity by altering receptor functional properties ( Fig . 1E ) . Thus , genetically increasing synaptic excitation of muscles is sufficient to produce dramatic increases in body bend depth . The above results suggested that analysis of the alternative motor pattern produced by increased neuromuscular synapse activation might provide insights into similar motor patterns that occur during foraging behaviors . As neuropeptides and other neuromodulators often play instrumental roles in generating flexible behavioral responses [1] , [28] , we next investigated if loss-of-function mutations in the neuropeptide processing enzymes egl-3/PC2 proprotein convertase or egl-21/carboxypeptidase E [29]–[33] might modulate the behavioral effects of L-AChR ( gf ) expression . The extremely sluggish movement of either egl-3 or egl-21 mutants in the absence of L-AChR ( gf ) expression hindered our efforts to address this question on solid agar plates ( Fig . S3 ) . In contrast , we found that mutations in egl-3 and egl-21 caused more modest locomotory defects in a related assay that assesses movement in liquid ( swimming ) , enabling analysis of L-AChR ( gf ) effects . Transgenic L-AChR ( gf ) animals were significantly less active than either wild type animals or the neuropeptide-deficient mutants . Further , we observed that mutation of either egl-3 or egl-21 suppressed the effects of L-AChR ( gf ) expression , improving movement ( Fig . 2A ) . While neurotransmitters involved in fast synaptic transmission are packaged into small clear synaptic vesicles , neuropeptide signaling is mediated through release of neuropeptide-laden dense core vesicles ( DCV ) . To gain further support for neuropeptide involvement , we analyzed mutations in the pkc-1 gene . pkc-1 encodes a member of the novel Protein Kinase C family ( nPKC ) most closely related to vertebrate nPKCε/η [34] , and was previously implicated in the regulation of DCV release without obvious effects on fast synaptic transmission at the NMJ [35] . Mutation of pkc-1 caused only modest effects on movement in otherwise wild type animals; yet , it reversed L-AChR ( gf ) -mediated behavioral changes almost completely ( Figs . 2A and S3 ) . Thus , normal neuropeptide processing and secretion is required for L-AChR ( gf ) effects on movement . Together , our findings suggest that neuropeptide signaling modulates locomotory behaviors associated with high neuromuscular synapse activity . Loss of either EGL-3 or EGL-21 function impacts biosynthesis of a variety of peptides , raising the question of which peptides play primary roles in modulating neuromuscular signaling and movement . To address this question , we undertook a suppressor approach , using RNAi to knockdown expression of candidate neuropeptide precursors in animals expressing the L-AChR ( gf ) transgene . Sequence analysis of the C . elegans genome predicts more than 100 candidate neuropeptide precursors including insulin-like ( ins ) , FMRFamide-like ( flp ) and neuropeptide-like ( nlp ) protein family members [36] . In total , we targeted 66 of these precursors spanning each of the 3 gene families . Remarkably , downregulating the neuropeptide-like nlp-12 precursor , a C . elegans homolog of cholecystokinin produced dramatic effects . RNAi targeting nlp-12 increased the movement velocity of transgenic L-AChR ( gf ) animals by more than 4-fold ( L-AChR ( gf ) empty vector: 12 . 7±2 body bends/min , L-AChR ( gf ) nlp-12 RNAi: 49±3 body bends/min , p<0 . 0001 ) to a level that was indistinguishable from that of wild type animals ( WT: 41 . 5±2 body bends/min ) . In contrast , RNAi targeting nlp-12 in wild type animals had no obvious effects on movement [in the presence of bacterial food] ( not shown ) . To further support a role for nlp-12 in modulating locomotory behaviors , we analyzed the effects of L-AChR ( gf ) expression in a strain carrying a deletion mutation ( ok335 ) in the nlp-12 locus . nlp-12 gives rise to 2 distinct mature peptides that share similarity with mammalian cholecystokinin-8 , the predominant form of CCK in mammalian neurons [37] , [38] . The ok335 allele corresponds to a 1070 bp deletion that disrupts both of the mature NLP-12 peptides . We measured body bend amplitude , movement velocity and body length in nlp-12 mutants carrying the L-AChR ( gf ) transgene and found that each of these measures was brought to near wild type levels by deletion of nlp-12 ( Fig . 2B , C and Fig . S4A , B ) . Similarly , nlp-12 deletion reversed the effects of L-AChR ( gf ) expression in liquid movement assays ( Fig . S4C ) . Surprisingly , while loss of nlp-12 function suppressed the effects of L-AChR ( gf ) in locomotion , no locomotion defect was apparent in these animals in the absence of L-AChR ( gf ) ( under our normal growth conditions ) . This finding is consistent with the notion that NLP-12 signaling may be preferentially required during certain behavioral states , a hypothesis we tested below . To further define nlp-12 effects , we examined the expression of a reporter construct containing the native NLP-12 genomic sequence SL2 trans-spliced to the fluorescent reporter mCherry . Consistent with previous studies [37] , [39] , fluorescence was solely visible in DVA and was present from early embryonic stage ( not shown ) throughout adulthood ( Fig . 2D ) . DVA is an interneuron that relays mechanosensory information onto the motor circuit through synaptic outputs onto both premotor interneurons and motor neurons [40] , [41] . The DVA cell body is located in the dorsorectal ganglia of the tail and extends a single process along the length of the ventral nerve cord into the nerve ring . The restricted expression of nlp-12 to DVA suggested an essential role for this neuron in NLP-12 modulation of the motor pattern . We tested this idea in laser ablation experiments . DVA ablation in L-AChR ( gf ) animals normalized body bend amplitude ( Fig . 2E ) and movement velocity ( Fig . S4B ) . In contrast , ablation of DVA caused no obvious effects in either nlp-12;L-AChR ( gf ) or wild type animals under normal food conditions ( Fig . 2E and [42] , [43] ) . Blocking vesicular release from DVA by DVA-specific expression of Tetanus toxin ( Tetx ) produced similar effects ( Fig . S4D ) . Thus , nlp-12 deletion , cell-specific DVA ablation or block of vesicular release from DVA , the only neuron with detectable levels of nlp-12 expression , each similarly reversed the effects of enhanced muscle excitation . Prior work has suggested that one function for DVA and NLP-12 signaling may be to convey proprioceptive information that is signaled through stretching associated with muscle contraction and movement [39] , [40] . Our results support the model that NLP-12 effects are mediated via secretion from DVA , and demonstrate a behavioral requirement for NLP-12 and DVA with increased muscle synapse activity . Our experiments provided evidence for NLP-12 modulation of the motor pattern when muscle excitability was altered genetically; however , under normal growth conditions , DVA ablation or nlp-12 deletion had only minimal effects on exploratory movement . These observations suggest NLP-12 signaling may be preferentially active during particular behaviors . A major source of synaptic input to DVA comes from dopaminergic mechanosensory neurons ( PDE ) that contribute to food sensing [44] , [45] ( Fig . 3A ) . This pattern of connectivity suggests DVA may integrate proprioceptive information with sensory information from PDE . Our earlier behavioral analysis showed wild type animals deepen their body bends following removal from food ( Fig . 1A ) . To evaluate whether NLP-12 signaling may contribute to modulation of body bend amplitude during these responses , we investigated the effects of nlp-12 deletion in the context of food deprivation ( Fig . 3B ) . Interestingly , nlp-12 deletion significantly reduced body bend depth in food-deprived wild type animals ( 25±1% reduction , p<0 . 01 ) . Normal body bend modulation upon food deprivation was restored by expression of nlp-12 under the native promoter . These results indicate that NLP-12 modulation enables deepening of body bends in response to sensory information about food availability . During our analysis of wild type behavior , we noted that body bend depth was initially increased following removal from food but then normalized with increasing time off food . To quantify this effect , we compared the depth of body bends during an initial 5-minute period following removal from food to that measured 30 minutes later . We found that body bend depth was significantly reduced 30 minutes after removal from food ( 22 . 2±3% reduction , p<0 . 01 ) ( Fig . 3B ) . Removal of C . elegans from food induces an immediate switch to an alternative local searching motor pattern , often referred to as area-restricted search [13]–[16] . During this motor pattern , C . elegans increase the frequency of reorientations in their direction of movement ( high-angle turns ) , an effect thought to represent a strategy for localized searching . With increased time following removal from food , these reorientations decrease in frequency , allowing for longer runs of forward movement and dispersal of the animals . The transient increase in body bend depth we observed after removing wild type animals from food suggested that modulation of body bend depth might contribute to local food searching behavior . To investigate this possibility , we monitored movement during a 35 minute period immediately following removal from food and quantified reversals and turning behavior during the first ( 0–5 ) and last ( 30–35 ) five minutes ( Fig . 3C–E ) . Wild type animals displayed frequent , large reorientations in their movement trajectory during an initial five-minute time period immediately following removal from food ( 0–5 min: 16 . 1±1 . 1 reorientations ) , consistent with previous descriptions of local searching behavior [13] , [14] . By 30 minutes after removal from food , the movement pattern of wild type animals shifted to longer runs and the number of high angle turns ( 30–35 min: 6 . 5±0 . 6 reorientations ) decreased significantly from those measured during the initial five minute period of food deprivation ( 60±3% decrease , p<0 . 0001 ) ( Fig . 3D ) . Specific expression of TetX in DVA or nlp-12 deletion significantly reduced the frequency of high-angle reorientations during the initial five-minute period following removal from food ( 43±9% decrease and 42±6% decrease respectively , p<0 . 0001 ) ( Fig . 3C , D ) . These results suggest that the local search motor pattern is dependent on DVA and NLP-12 signaling . Neuropeptides generally act through G-protein coupled receptors ( GPCRs ) . NLP-12 directly binds the GPCR Cholecystokinin-like Receptor 2 ( CKR-2 ) in vitro and previously described roles for nlp-12 have indicated a requirement for ckr-2 [37] , [39] . Surprisingly however , we found that ckr-2 was not required for normal local food searching behavior ( Fig . 3D ) . This result suggests an interesting possibility that NLP-12 modulation of foraging behavior relies on additional GPCR signaling pathways which operate independently or in parallel with CKR-2 . There are several genes encoding GPCRs with significant homology to ckr-2 in the worm genome ( e . g . ckr-1 , npr-2 , npr-5 ) . While NPR-5 is activated by the neuropeptide FLP-18 when expressed in Xenopus oocytes [46] , CKR-1 and NPR-2 remain uncharacterized , so these genes are good candidates for additional components of the NLP-12 signaling pathway . Reorientations during local searching occur primarily through 2 mechanisms: reversals coupled to sharp , ventrally directed head-to-tail turns ( referred to as omega turns ) , or large head swings accompanied by deep body bends that establish a new heading [47] , [48] . For wild type animals , approximately 50% of the trajectory changes we observed were initiated by omega turns . Interestingly , the frequency of omega turns was not significantly decreased by nlp-12 deletion ( Fig . 3E ) . Instead , we found that deletion of nlp-12 caused a significant reduction in the frequency of trajectory changes that were not associated with omega turns ( 47±6% decrease , p<0 . 0001 ) . This reduction was rescued by nlp-12 expression under the native promoter . Previous work has demonstrated that olfactory information about food availability shapes local searching primarily by increasing the frequency of omega turning [13] . Our findings indicate that NLP-12 modulates local searching by affecting changes in trajectory that arise independently of omega turns . Thus , NLP-12-mediated locomotory changes during local searching are likely to be produced through a mechanism that is distinct from those elicited by changes in olfactory neuron activity . Our above behavioral experiments indicated that foraging responses to local changes in food availability were shaped by NLP-12 signaling . DVA receives strong synaptic projections from the dopaminergic PDE neurons and one proposed role for dopaminergic neurons is in food sensing [44] , [45] . Therefore , we next investigated the effects of dopamine on foraging behavior . Initially , we examined whether exogenous dopamine was sufficient to alter the motor pattern of wild type animals and , consistent with previous work [14] , found that a brief exposure to exogenous dopamine was sufficient to elicit increases in reorientations during movement ( 27±4% increase , p<0 . 01 ) . The effects of dopamine required nlp-12 , providing support for the notion that dopamine signaling may regulate NLP-12 release ( Fig . 4A ) . Prior studies in C . elegans have used fluorescently labeled neuropeptide precursors to monitor neuropeptide release [35] , [39] . Therefore to directly test the idea that dopamine signaling stimulates NLP-12 release , we expressed NLP-12-VenusYFP in DVA . In the absence of exogenous dopamine , we observed a punctate pattern of NLP-12-VenusYFP in the DVA process , consistent with prior work [39] . If dopamine signaling onto DVA leads to significant release of NLP-12 , we would expect acute dopamine treatment to reduce punctate NLP-12 fluorescence in DVA . Consistent with this , we found brief exposure to dopamine produced a significant decrease in NLP-12-VenusYFP fluorescence ( Fig . 4B , C ) To gain support for the idea that this was a direct effect of dopamine on DVA , we examined dopamine receptor expression and found that a transcriptional reporter for the dop-1 gene was strongly expressed in DVA ( Fig . 4D ) . dop-1 encodes a D1-like dopamine receptor previously implicated in modulation of sensory neuron responses to touch [49]–[51] . Acute exposure of dop-1 mutants to dopamine did not produce significant decreases in NLP-12-VenusYFP fluorescence , suggesting that dopamine stimulation of NLP-12 release required DOP-1 . In addition , we noted that mutation of dop-1 produced decreased levels of basal fluorescence , suggesting DOP-1 also regulates unstimulated NLP-12 levels . Specific rescue of dop-1 expression in DVA normalized basal NLP-12 levels and importantly , restored sensitivity to dopamine . Thus , our results support the notion that dopamine acts to stimulate NLP-12 release via direct activation of the D1-like receptor DOP-1 in DVA . To address whether DOP-1 stimulation of NLP-12 release played a central role in modulating foraging behavior , we examined the food-searching behavior of dop-1 mutants . Prior work has demonstrated that mutations in dop-1 alone do not impair exploratory movement in the presence of food [52] . We found that dop-1 mutants were defective in foraging responses to food deprivation , behaving similarly to nlp-12 mutants . Specifically , the frequency of high-angle reorientations following removal from food was significantly decreased in dop-1 mutants compared to the wild type ( 25±3% decrease , p<0 . 001 ) ( Fig . 4E ) . These effects were rescued by specific expression of wild type dop-1 in DVA and were not observed with mutation of the related dopamine receptor dop-3 . Conversely , behavioral changes elicited by increasing muscle synapse activation did not show a strong requirement for dopamine signaling ( e . g . L-AChR ( gf ) body bend depth: 72±4 µm , cat-2;L-AChR ( gf ) body bend depth: 73 . 4±2 µm ) , suggesting that proprioceptive signaling and dopamine mechanosensory signaling can each independently affect DVA activity and NLP-12 release . Taken together , our results support a model where dopamine signaling through DOP-1 regulates foraging by direct modulation of DVA activity and NLP-12 release . Our finding that NLP-12 signaling was required for animals to appropriately modulate body bends and turning during foraging suggested that altering endogenous NLP-12 levels would be sufficient to modify motor output . To test this idea , we expressed the nlp-12 genomic sequence and native promoter at high copy levels in wild type animals . We found that the locomotory pattern was dramatically altered in transgenic animals overexpressing a stably integrated nlp-12 transgene ( nlp-12 ( OE ) ; ufIs104 ) . To investigate effects of nlp-12 overexpression on foraging , we examined local search responses to food deprivation . nlp-12 overexpression increased high-angle turning in the initial five minutes following removing from food by more than 2-fold , and induced occasional bouts of coiling ( Fig . 5A , B ) . Notably , the effects of nlp-12 overexpression were not significantly attenuated by prolonged food deprivation and persisted even in the presence of food . Our overexpression analysis suggests that genetically elevating levels of NLP-12 signaling compromised modulatory control of behavioral responses to changes in food availability , producing a chronic local search-like behavioral state . To investigate the effects of nlp-12 overexpression in more detail , we also examined body bend amplitude and bending angles for both the wild type and nlp-12 ( OE ) strain ( Fig . 5C–E ) during exploratory movement in the presence of food . Body bend amplitude was increased by roughly 3-fold in nlp-12 ( OE ) animals ( Fig . 5C ) . We quantified bending angles as deviations from the midline using the midpoint of the body as reference . For wild type animals , body bend angles rarely exceeded 75° . In contrast , bending angles in the nlp-12 ( OE ) strain were more broadly distributed , often exceeding 150° ( Fig . 5D , E ) . These effects of nlp-12 overexpression were significantly attenuated after DVA ablation , yet were not reversed completely . Although we confirmed that the DVA cell body was eliminated in our ablations , we noted that residual mCherry signal associated with the DVA process often remained , suggesting continuing secretion of overexpressed NLP-12 from the process may account for the partial effects of DVA ablation . Alternatively , the partial effects of DVA ablation may indicate that the nlp-12 promoter drives very low levels of expression from other cells that were not detectable with our reporter . Taken together , our findings indicate that increasing NLP-12 levels is sufficient to alter locomotor activity by enhancing the depth of sinusoidal body bends and increasing bending angles . As our previous analysis indicated NLP-12 signaling was important for L-AChR ( gf ) effects on movement , we next asked whether normal L-AChR function was required for the effects of nlp-12 overexpression . UNC-29 is an essential L-AChR subunit and mutation of unc-29 leads to a complete absence of L-AChR mediated currents from body wall muscles . Loss of L-AChR function causes slowed movement , but does not disrupt the sinusoidal locomotory pattern [17] , [22] , [53] . The increases in body bend amplitude and bending angles that occurred with nlp-12 overexpression were significantly reduced by mutation of unc-29 ( Fig . 5C , E ) . These results suggest NLP-12 mediated effects on the motor pattern require intact L-AChR signaling at the NMJ . In contrast , deletion of acr-16 , an essential subunit of a second population of iAChRs that mediate much of the evoked current at the NMJ [17] , did not alter bending angles or body bend amplitude in nlp-12 ( OE ) animals . Thus , there is a specific requirement for L-AChR , but not N-AChR , function in NLP-12 enhancement of body bend amplitude and turning behavior . While the precise synaptic organization of L-AChRs and ACR-16 receptors at the neuromuscular junction has not been determined , our findings suggest the effects of NLP-12 are mediated through synapses where L-AChR signaling predominates . Neuromodulators can achieve their effects by regulating the release of classical neurotransmitters , altering post-synaptic receptor functional properties , or otherwise altering excitability . We sought to distinguish between these possible models for nlp-12 action . As our previous analyses showed the importance of L-AChR function in NLP-12 modulation , we first examined the effects of nlp-12 deletion on post-synaptic L-AChR functional properties . To facilitate our analysis , we returned to the L-AChR ( gf ) strain where a behavioral requirement for nlp-12 was most apparent . L-AChR ( gf ) ;nlp-12 animals paralyzed more rapidly to the specific agonist levamisole than the wild type , resembling transgenic animals that were wild type for nlp-12 ( Fig . 6A ) . These results suggested L-AChR function was not appreciably altered by mutation of nlp-12 . To examine this directly , we made whole-cell recordings of current responses to exogenous levamisole from body wall muscles . The time course of levamisole-evoked currents in L-AChR ( gf ) animals was prolonged dramatically compared to the wild type , while peak amplitude was unaffected ( Fig . 6B ) – a result consistent with the accelerated time course of levamisole-induced paralysis described above . However , mutation of nlp-12 did not produce obvious changes in the magnitude or duration of levamisole-evoked current responses ( Fig . 6B , C ) . These results demonstrate that the effects of NLP-12 are unlikely to be mediated through direct modulation of post-synaptic L-AChR function . A previous study found that treatment with the cholinesterase inhibitor aldicarb increased ACh release from motor neurons , and that these effects required nlp-12 [39] . To investigate whether the effects of NLP-12 we observed in our behavioral experiments may involve a similar mechanism , we recorded endogenous excitatory synaptic events as well as synaptic events evoked by photostimulation of motor neurons . We did not observe a significant change in the rate or amplitude of endogenous excitatory events with expression of the L-AChR ( gf ) transgene or with deletion of nlp-12 ( Fig . S5 ) . In contrast , L-AChR ( gf ) expression produced a dramatic change in evoked synaptic currents . These specific effects on evoked currents may suggest that the lifetime of ACh in the synaptic cleft is a primary factor in determining the duration of endogenous synaptic events , while the duration of evoked events are less strongly impacted by the action of cholinesterase . We measured evoked synaptic currents in response to photostimulation of motor neurons using a transgenic strain ( ufIs23 ) stably expressing Channelrhodopsin-2 ( ChR2 ) in cholinergic motor neurons . In wild type animals , excitatory evoked currents reflect synaptic activation of both L-AChRs and N-AChRs . The rapid peak phase of the synaptic current is primarily mediated by N-AChRs , and can be eliminated by deletion of the acr-16 gene or application of the N-AChR specific antagonist , dihydro-β-erythroidine ( dHβE ) [17]–[19] . Conversely , the prolonged component of the synaptic current is mediated through L-AChR activation . The peak amplitude of evoked synaptic currents to brief light exposure ( 10 ms ) was not altered appreciably by L-AChR ( gf ) expression; however , the duration of evoked current responses was increased significantly ( Fig . S6 ) . To investigate this effect in more detail , we isolated the L-AChR-mediated component of the evoked synaptic current using dHβE . The decay time of the L-AChR synaptic current was increased dramatically by L-AChR ( gf ) expression without a significant change in current amplitude ( Fig . 6D–F ) . The increased duration of evoked synaptic currents became even more apparent with prolonged ( 800 ms ) stimulation of motor neurons ( Fig . 6D , lower ) . Deletion of nlp-12 significantly reduced the duration of evoked responses in L-AChR ( gf ) animals ( 65±1% reduction , p<0 . 001 ) , such that the average decay time of evoked L-AChR currents in nlp-12 mutants expressing the L-AChR ( gf ) transgene was not statistically different from that of the wild type ( Fig . 6D , F ) . Thus , the duration of post-synaptic currents evoked by ACh release from motor neurons were normalized by deletion of nlp-12 , whereas current responses to drug application onto muscles were largely unaffected . These results support a model where NLP-12 increases the duration of synaptic currents by enhancing ACh release from cholinergic motor neurons .
Understanding how organisms appropriately adjust behavioral responses to their environment requires knowledge of how sensory information is transformed by neural circuits into motor outputs . Here , we provide a mechanism by which dopaminergic modulation of neuropeptide signaling alters neuromuscular excitability and shapes C . elegans foraging behavior . By elevating neuromuscular excitability , we revealed a state-dependent role for the C . elegans cholecystokinin ( CCK ) homolog NLP-12 in modulating motor behavior . Behavioral modulation by NLP-12 was particularly evident during local food searching , a motor program in which body bend amplitude and the frequency of high-angle turning is transiently increased in response to changes in food availability . NLP-12 signaling was regulated by the D1-like dopamine receptor DOP-1 , and DOP-1 signaling was also required for normal food searching behavior . Our results show that C . elegans dynamically modulate their motor pattern in response to changes in food availability , through dopaminergic regulation of NLP-12/CCK signaling . Remarkably , similar interactions between dopamine and cholecystokinin occur in the mammalian brain [54] , [55] , suggesting this may be a conserved mechanism for generating context-dependent behavioral responses . Behavioral deficits associated with loss of specific neuromodulators are often subtle and transient , owing to the fact that neuromodulatory effects are typically delimited to specific patterns of behavior . As a consequence of this , efforts to identify molecular components of specific neuromodulatory pathways using forward genetic approaches have met with only limited success . We developed a genetic strategy to overcome these hurdles . We engineered a mutation into the pore-forming domains of muscle L-AChR subunits to facilitate studies of neuromodulation in shaping C . elegans movement . Incorporation of mutant L-AChR subunits heightened sensitivity to pharmacological agents that increased post-synaptic iAChR activation , and prolonged the duration of evoked synaptic currents in electrophysiology studies . Using this approach to increase muscle synapse activity produced exaggerated sinusoidal movement , a behavioral phenotype distinct from previously reported mutations that enhanced muscle excitability ( e . g . by elimination of inhibitory signaling onto muscles [56] ) , suggesting that this manipulation would be informative for identifying novel modulators of excitatory transmission and movement . Consistent with this , a recent study employing a similar gain-of-function approach with the neuronal iAChR subunit ACR-2 identified roles for the flp-1 and flp-18 neuropeptides in regulating the activity of motor neurons [57] . The locomotory effects we observed with L-AChR ( gf ) expression were dependent upon neuropeptide signaling , as they were reversed by mutation of either pkc-1 nPKCε/η , egl-3 PC2 or egl-21 Carboxypeptidase E . Disruption of egl-3 leads to loss of the vast majority of mature nlp and flp neuropeptides in C . elegans , including NLP-12 , while mutation of either pkc-1 or egl-21 affects a smaller subset of neuropeptides [29] , [30] , [35] . The most potent neuropeptide modifier of L-AChR ( gf ) locomotory behavior in our assays was NLP-12 , and based on our analysis , we would predict that NLP-12 signaling is directly or indirectly dependent on normal PKC-1 and EGL-21 function . NLP-12 overexpression altered movement and produced increases in body bend amplitude that were qualitatively similar to those observed for L-AChR ( gf ) animals . Moreover , normal L-AChR function was required for the full effects of NLP-12 overexpression . These data suggested that one avenue for NLP-12 effects on the motor pattern was through direct modulation of neuromuscular transmission . In support of this , we found that L-AChR ( gf ) -mediated increases in the duration of synaptic currents evoked by photostimulation of motor neurons were reversed almost completely in nlp-12 mutants . Prior work has demonstrated that acute pharmacological treatment with the cholinesterase inhibitor aldicarb potentiates ACh release from motor neurons via a mechanism that requires NLP-12 [39] . As a primary effect of aldicarb treatment is muscle contraction , it was suggested that muscle contraction may trigger activation of the stretch–sensitive DVA neuron and subsequent release of NLP-12 , enhancing neuromuscular signaling . Thus , under certain conditions muscle contraction may modulate the motor pattern through a proprioceptive-like mechanism . Our genetic , electrophysiological and behavioral data support and extend this model . Expression of L-AChR ( gf ) causes chronic muscle contraction in a manner similar to aldicarb treatment , and both manipulations heighten activation of synaptic L-AChRs . We found that both NLP-12 and DVA are required for the behavioral effects of L-AChR ( gf ) expression , and nlp-12 deletion reduces the duration of L-AChR mediated synaptic currents . These results are consistent with the model that NLP-12 release from DVA acts in a feedforward pathway required for the behavioral phenotype of L-AChR ( gf ) animals . Our analysis also revealed novel roles for DVA and NLP-12 signaling in shaping alternative motor responses during local food searching . Neuromodulatory signaling allows anatomically fixed circuits to have flexibility in their outputs . Owing to detailed electron microscopy and laser ablation studies , we have long known the identity and connectivity patterns of the principal neurons involved in the control of C . elegans movement [13] , [15] , [45] , [58] , [59] . Nonetheless , defining neuromodulatory effects on motor circuit activity and behavior has remained a challenge . Our findings provide support for a model where sensory information about food availability is translated into context-dependent motor output by dopaminergic regulation of NLP-12 release from DVA and NLP-12 modulation of locomotory output . We propose that the DVA neuron integrates proprioceptive and dopaminergic mechanosensory information . In this model , NLP-12 release reflects contributions from both forms of sensory input . In support of this view , we show that dopamine elicits increases in reorientation in a dop-1 dependent manner without a requirement for increased muscle synapse activation . Conversely , mutation of cat-2/tyrosine hydroxylase does not alter behavioral changes elicited by increased muscle synapse activity . Thus , our findings suggest heightened signaling through either sensory modality ( proprioceptive or dopaminergic mechanosensory ) is sufficient to elicit altered NLP-12 release and behavioral changes . We envision that the precise nature of the behavioral changes reflect characteristics of the pattern of DVA activation as well as the activity of other neuronal pathways that contribute to each behavior . In principle , the context-dependent effects of NLP-12 could reflect altered sensitivity to the peptide during particular patterns of motor activity or altered release of the peptide in response to changing sensory information . Our analysis supports the latter possibility . Food availability is monitored through dopaminergic mechanosensory signaling ( ADE , PDE , and CEP neurons ) , as well as through activation of specific classes of olfactory and gustatory neurons ( AWA , AWC and ASE ) [13] , [14] , [44] . Neuropeptide signaling plays important roles in the regulation of olfactory neuron responses to food [46] , [60]; however , we have had only a limited understanding of how sensory representations of food availability are translated into altered patterns of behavior . A major source of synaptic input to DVA is through the dopaminergic neuron PDE . Prior work showed normal area-restricted search behavior requires dopaminergic signaling and implicated involvement of PDE [14] . Our results support the notion that an initial period of local searching immediately following removal from food is accomplished through DVA activation via dopaminergic inputs from PDE and context-dependent stimulation of NLP-12 release . This pathway may act in parallel with sensory information carried by olfactory neurons in order to reinforce the local searching motor pattern . Acute dopamine treatment rapidly decreased NLP-12 levels in DVA , and these effects were reversed with mutation of the D1-like dopamine receptor dop-1 . Indeed , acute exposure of dop-1 mutants to dopamine produced significant NLP-12 accumulation in DVA axons , suggesting additional DOP-1-independent effects of dopamine on DVA . DOP-1 was previously shown to promote ACh release from motor neurons by functioning antagonistically to the D2-like receptor DOP-3 and coupling to Gαq/EGL-30 and phospholipase C ( PLCβ/EGL-8 ) [52] , [61] . Our results suggest DOP-1 may similarly act in opposition to another D2-like dopamine receptor to regulate NLP-12 release from DVA during local searching . dop-3 expression has not been reported in DVA and we did not observe any effect of dop-3 mutations on local food searching behavior ( Figure 4E ) . A second D2-like dopamine receptor is encoded by the dop-2 gene , and dop-2 transcriptional reporters are expressed in a variety of head neurons , as well as some unidentified tail neurons [62] . Thus , dop-2 may be a good candidate as an additional participant in dopamine regulation of NLP-12 release from DVA . Alternatively , the NLP-12 accumulation in dop-1 mutants may reflect homeostatic changes due to loss of dop-1 . The olfactory neurons involved in food-sensing promote local searching behavior through synaptic contacts onto interneurons involved in the control of movement [13] , [46] , [60] . DVA has synaptic outputs onto interneurons that are postsynaptic partners of these neurons , in addition to motor neurons . This pattern of connectivity suggests that the modulatory effects of DVA and perhaps NLP-12 may extend beyond motor neurons , allowing for integrated regulation of neuronal activity at several layers of the circuit . The ligand and GPCR families underlying neuromodulatory signaling are highly conserved throughout the animal kingdom [42] , [63] , [64] . In particular , NLP-12 shares sequence similarity with mammalian gastrin and cholecystokinin peptides [64] . In mammals , the effects of CCK are mediated through two GPCRs , CCK1R and CCK2R [65] . CKR-2 is a C . elegans homologue of mammalian CCK/gastrin receptors and , to date , is the only GPCR that has been shown to bind NLP-12 [37] . We show that CKR-2 is not required for normal local food searching behavior . This result raises the possibility that , as is the case for mammals , NLP-12/CCK effects are not mediated through a single GPCR . There are several genes encoding GPCRs with significant homology to ckr-2 in the worm genome ( e . g . ckr-1 , npr-2 , npr-5 ) . Of these , ckr-1 and npr-2 remain uncharacterized and are good candidates as additional components in NLP-12 regulation of foraging . Mammalian CCK signaling regulates food intake by stimulating smooth muscle contractions , intestinal motility and secretion of digestive enzymes , as well as acting centrally to promote satiety [66] . Recent work in C . elegans has suggested NLP-12 plays similar roles in promoting digestive enzyme secretion and regulating fat storage [37] . Our work provides evidence that dopaminergic regulation of NLP-12 release and motor circuit activity is central to local food searching behavior . We suggest these dual roles for the conserved NLP-12/CCK signaling pathway in C . elegans enable a strategy for optimizing resource utilization through coordinated regulation of metabolic processes and food searching behavior .
All nematode strains were maintained on agar nematode growth media seeded with OP50 at room temperature ( 22–24°C ) as described previously [67] . The wild type reference animals for all cases are the N2 Bristol strain . The following strains were used or generated in this work: IZ818: ufIs47 [Pmyo-3:: unc-38 , Pmyo-3::unc-29 , Pmyo-3:: lev-1] , IZ236: ufIs6 [Pmyo-3:: unc-38 ( V/S ) , Pmyo-3::unc-29 ( L/S ) , Pmyo-3:: lev-1 ( L/S ) ] , VC731: unc-63 ( ok1075 ) I , IZ1260: unc-63 ( ok1075 ) ;ufIs6 , IZ801:ufIs23[Pacr-2::ChR2-GFP]; IZ806: ufIs23;ufIs6 , RB781: pkc-1 ( ok563 ) V , KP2342: pkc-1 ( nu488 ) V , IZ1001: pkc-1 ( ok563 ) ;ufIs6 , VC671: egl-3 ( ok979 ) V , IZ1006: egl-3 ( ok979 ) ;ufIs6 , KP2018: egl-21 ( n476 ) IV , IZ825: egl-21 ( n476 ) ;ufIs6 , IZ908: nlp-12 ( ok335 ) I , IZ681: nlp-12 ( ok335 ) ;ufIs6 , IZ1304: nlp-12 ( ok335 ) ;ufEx430[Pnlp-12::nlp-12::SL2::mCherry] , IZ1099: ufEx364[Pnlp-12::nlp-12::SL2::mCherry] , IZ1099: ufIs6; ufEx364[Pnlp-12::nlp-12::SL2::mCherry] , IZ1310: ufEx432[Pnlp-12::mCherry] , IZ1432: ufEx475[Pnlp-12::Tetanus] , IZ1480: ufIs6; ufEx475[Pnlp-12::Tetanus] , IZ1152: ufIs104[Pnlp-12::nlp-12 genomic locus] , IZ1313: ufIs104;ufEx433[Pnlp-12::mCherry] , LSC0032: ckr-2 ( tm3082 ) III , IZ74: unc-29 ( x29 ) I , IZ1729: unc-29 ( x29 ) ;ufIs104 , IZ57: acr-16 ( ok789 ) V , IZ1728: acr-16 ( ok789 ) ;ufIs104 , LX645: dop-1 ( vs100 ) X , IZ1512: dop-1 ( vs100 ) X;ufEx498 ( Pnlp-12::dop-1 ) , LX636: dop-1 ( vs101 ) X , LX703: dop-3 ( vs106 ) X , IZ1299: lin-15 ( n765ts ) X;ufEx428[Pnlp-12::nlp-12::Venus-YFP] , XP2: dop-1 ( vs100 ) lin-15 ( n765ts ) X , IZ1479: dop-1 ( vs100 ) lin-15 ( n765ts ) ;ufEx428[Pnlp-12::nlp-12::Venus-YFP] , IZ1513: dop-1 ( vs100 ) lin-15 ( n765ts ) ;ufEx428[Pnlp-12::nlp-12::Venus-YFP];ufEx499 [Pnlp-12::dop-1] , LX831: vsIs28[Pdop-1::GFP];vsIs33[Pdop-3::RFP] , IZ1477: vsIs28[Pdop-1::GFP]; ufEx440[Pnlp-12::nlp-12::SL2::mCherry] , IZ1605: cat-2 ( n4547 ) ufIs6 II . Transgenic strains were obtained by microinjection of one or more of the following plasmid DNAs ( 10–30 ng/ul ) into the germ line: pLJ2 [Pmyo3::unc-38] , pLJ3 [Pmyo3::unc-29] , pLJ6 [Pmyo3::lev-1] , pMF1[Pmyo3::unc-38 ( V/S ) ] , pLJ8[Pmyo3::unc-29 ( L/S ) ] , pLJ9 [Pmyo3::lev-1 ( L/S ) ] , pCL8 [Pnlp-12::Tetx] , pCL12 [Pnlp-12::NLP-12-VenusYFP] , pCL19 [Pnlp-12::NLP-12::SL2::mCherry] , pCL27 [Pnlp-12::mCherry] and pCL41 [Pnlp-12::Dop-1] . The nlp-12 ( OE ) strain ( ufIs104 ) was generated by integration of a 1 . 76 kb PCR product containing the nlp-12 promoter and genomic locus ( −354 bp to +1407 bp relative to the transcriptional start ) . Germline transformation was monitored either by coinjection with the lin-15 rescuing plasmid pL15EK , or by coinjection with fluorescent reporters expressed in the pharynx ( pHP6 [Plgc-11::GFP] ) or intestine ( Pelt-2::GFP ) . Multiple independent extrachromosomal lines were obtained for each transgenic strain and data presented are from a single representative transgenic line unless noted otherwise . Stably integrated lines were generated by X-ray integration and outcrossed at least four times to wild type . RNAi feeding experiments were essentially performed as described previously [68] . Briefly , bacteria containing RNAi clone targeting various proneuropeptide genes were cultured in LB media containing 50 ug/ml ampicillin . Bacteria were seeded on NGM agar plates containing 1 mM IPTG together with ampicillin and tetracycline and allowed to dry at room temperature . L4 stage wild type or L-AChR ( gf ) worms were placed on RNAi seeded plates and allowed to reproduce at 20°C . Staged young adult F2 worms were transferred to a fresh plate and visually assessed for altered movement compared to a control strain plated on bacteria expressing the empty vector . Scoring was performed blind and a positive hit was assigned to a plate if at least 25% of animals were scored as showing altered movement . Alterations in movement were scored by visually monitoring track amplitudes during spontaneous movement as well as motor responses to tail touch . Plates were scored in duplicate and movement velocity was quantified for the best candidates by counting body bend frequency in the presence of bacterial food . RNAi targeting the unc-22 gene was used as a positive control . From this analysis we identified five neuropeptide precursors ( nlp-2 , nlp-3 , nlp-7 , nlp-12 , nlp-15 ) that produced varying degrees of L-AChR ( gf ) suppression when downregulated by RNAi . We confirmed these effects in assays of the corresponding deletion strains in combination with L-AChR ( gf ) . Of these , nlp-12 deletion produced the strongest and most consistent effects . All behavioral analyses were performed using staged populations of young adult animals ( 24 h following L4 ) at room temperature ( 22–24°C ) . Strains were scored in parallel , with the researcher blinded to genotype . Movies and still images for behavioral analyses were obtained using an Olympus SZ61 upright microscope equipped with a FireWire camera ( Imaging Source ) and were acquired at a rate of 30 frames/s . For locomotion assays on agar , individual worms were transferred to fresh 5 cm NGM agar plates thinly seeded with bacteria and allowed to acclimate for 1–2 minutes prior to filming . In all cases 45 s digital video of individual worms were captured at 0 . 75× magnification and converted into AVI format prior to analysis using WormLab software ( MBF Bioscience ) . Each movie was analyzed using the backtracking module and mean body bend amplitude was calculated . All parameters were kept constant at manufacturer's recommended settings with the exception that thresholding was varied from movie to movie in order to maximize optimal recognition of worms . Track lengths and body lengths were calculated using ImageJ . For analysis of body lengths , the ventral midline was measured from synchronized young adult animals . Analyses of area-restricted food searching were performed following the protocol outlined in [14] with minor variations . Staged animals were transferred from food plates onto 11 cm NGM agar observation plates without food after removal of excess bacteria . Movement was recorded for a 5-minute period immediately following transfer to the observation plate and again after 30 min following transfer . Movement was analyzed using WormLab software to calculate the average body bend amplitudes and the paths of individual worms were reconstructed from the track baselines . To quantitate changes in direction during food search , reorientations were visually identified and counted manually from digital movies . Omega turns were defined as turns in which the head made contact with the tail , and typically followed long reversals of 3 or more head swings . Reorientations in the absence of omega turns were scored as directional changes with an angle change of >50° within a single body bend . In most cases these trajectory changes either followed a brief cessation of forward movement or a short reversal; however , they also occurred less frequently during forward runs . Dopamine effects on directional reorientation were performed using similar criteria , except that animals were monitored for 5 mins immediately after transfer to salt free plates , with or without 2 . 5 mM dopamine . Dopamine ( salt-free ) plates were prepared following the protocol outlined in [48] . Reorientation data for each worm was averaged over a 5 min observation period . Additional behavioral analyses of nlp-12 ( OE ) animals were conducted on bacteria seeded plates . Movies were analyzed with WormLab software using the Bending Angle – Single parameter to perform a frame by frame measurement of body curvature during movement . Bending angle is defined as the angle at the midpoint of the worm with respect to the head and tail centroids . For levamisole assays , staged populations of adult animals ( ≥10 ) were transferred to NGM plates containing levamisole ( Sigma ) at the concentration indicated and assayed for paralysis every 15 min for 2–3 hrs . Worms were scored as paralyzed when they failed to execute at least one body bend after prodding with a pick . In all cases , worms that were unresponsive to mechanical prodding were also paralyzed in the absence of mechanical stimulation . Movement assays in liquid were carried out essentially as described previously [69] . Briefly , staged single adult animal were transferred without food onto a PCR cap containing 20 µl of M9 buffer . After an equilibration period of 1 min , the number of body bends was counted manually over a 3 min period using a Leica MZ 6 series microscope at 1 . 5× magnification . For all pairwise comparisons , statistical significance was determined by two-tailed Student's t-test using GraphPad Prism software . For all multiple comparisons , statistical significance was determined by ANOVA followed by Sidak's test . Laser ablations were conducted using standard methods with a pulsed-UV laser ( Photonic Instruments , model: 337-USAS ) [70] . Briefly , animals were mounted on agar pads and anesthetized in 5 µl of 20 mM sodium azide . Both mock-ablated and laser-operated animals were processed in parallel . The DVA neuron was identified in transgenic L2/L3 larval animals expressing mCherry under the nlp-12 promoter . Ablation of DVA was confirmed by an absence of fluorescence in the cell body position in adult animals . Behavioral assays were conducted on young adult animals following a recovery period of roughly two days after ablation . Epifluorescent imaging was performed using a Zeiss Axioimager M1 microscope and Axiovision software ( Zeiss ) . Confocal microscopy was performed using a Zeiss Axioskop 2 microscope system and LSM Pascal 5 imaging software ( Zeiss ) . Worms were mounted on agarose pads and immobilized with sodium azide . All images were obtained from staged young adult animals ( 24 h after the L4 stage ) and processed using ImageJ software ( open source ) . For quantitative imaging of NLP-12-VenusYFP fluorescence in DVA axon , animals were orientated with their ventral side facing the objective and imaged immediately anterior to the vulva . Imaging was performed after transfer to salt-free plates containing either a thin lawn of bacteria or 10 mM dopamine for a period of ten minutes . Images were acquired as 0 . 5 µM confocal slices and maximum intensity projections were utilized for further fluorescence analysis . Raw YFP fluorescence values were normalized to the absolute mean fluorescence of ∼0 . 175 µm diameter green fluorescent microsphere beads ( Molecular Probes ) . Postsynaptic currents were recorded from body wall muscles using whole-cell patch clamp electrophysiology as previously described [71] , [72] . The extracellular solution typically consisted of 150 mM NaCl , 5 mM KCl , 15 mM HEPES , and 10 mM glucose ( pH 7 . 4 , osmolarity adjusted with 20 mM sucrose ) with 4 mM MgCl2 and 1 mM CaCl2 . Photostimulation experiments were performed using the same extracellular solution but containing 1 mM MgCl2 and 5 mM CaCl2 . The intracellular fluid ( ICF ) consisted of 115 mM K-gluconate , 25 mM KCl , 0 . 1 mM CaCl2 , 50 mM HEPES , 5 mM Mg-ATP , 0 . 5 mM Na-GTP , 0 . 5 mM cGMP , 0 . 5 mM cAMP , and 1 mM BAPTA ( pH 7 . 4 , osmolarity adjusted with 10 mM sucrose ) . At least 60–90 s of continuous data were used in the analysis of endogenous events . For light evoked recordings , strains expressing ChR2 were illuminated with blue light from Exfo X-Cite series 120 light source , filtered through a GFP excitation filter ( 450–490 nm ) . The duration of illumination was defined using a computer-controlled shutter ( SmartShutter , Sutter Instrument ) and HEKA Patchmaster software . For drug-evoked responses , drugs were pressure ejected using a Picospritzer II ( General Valve Corporation ) . For all recordings , series resistance was compensated 50% and only recordings in which the series resistance was stable throughout the course of the recording were included . Data analysis was performed using Igor Pro ( WaveMetrics , Inc . ) and Mini Analysis ( Synaptosoft , Inc . ) . Statistical comparisons were made by ANOVA with Sidak's post-hoc test or student's t-test using GraphPad Prism . | Animal behavior is profoundly affected by contextual information about the internal state of the organism as well as sensory information about the external environment . A class of signaling molecules known as neuropeptides have been implicated in driving transitions between behavioral states ( e . g . , from food seeking to satiety and back ) but we have only a limited understanding of how neuropeptide signaling modulates neural circuit activity and elicits context-dependent behaviors . Here we identify a novel mechanism by which C . elegans modulate their behavior in response to sensory information about food . We show that dopaminergic regulation of NLP-12 , a C . elegans homolog of the mammalian neuropeptide cholecystokinin ( CCK ) , shapes behavioral transitions that are central to food searching . Given the conserved nature of these signaling pathways , our work raises the interesting possibility that dopamine modulation of CCK signaling represents a general mechanism by which nervous systems shape context-dependent behavioral changes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"neuroscience",
"behavioral",
"neuroscience",
"anatomy",
"neurotransmission",
"nervous",
"system",
"biology",
"and",
"life",
"sciences",
"motor",
"system",
"neuroscience"
] | 2014 | A Conserved Dopamine-Cholecystokinin Signaling Pathway Shapes Context–Dependent Caenorhabditis elegans Behavior |
Catch-up growth after insults to growing organs is paramount to achieving robust body proportions . In fly larvae , injury to individual tissues is followed by local and systemic compensatory mechanisms that allow the damaged tissue to regain normal proportions with other tissues . In vertebrates , local catch-up growth has been described after transient reduction of bone growth , but the underlying cellular responses are controversial . We developed an approach to study catch-up growth in foetal mice in which mosaic expression of the cell cycle suppressor p21 is induced in the cartilage cells ( chondrocytes ) that drive long-bone elongation . By specifically targeting p21 expression to left hindlimb chondrocytes , the right limb serves as an internal control . Unexpectedly , left–right limb symmetry remained normal , revealing deployment of compensatory mechanisms . Above a certain threshold of insult , an orchestrated response was triggered involving local enhancement of bone growth and systemic growth reduction that ensured that body proportions were maintained . The local response entailed hyperproliferation of spared left limb chondrocytes that was associated with reduced chondrocyte density . The systemic effect involved impaired placental function and IGF signalling , revealing bone–placenta communication . Therefore , vertebrates , like invertebrates , can mount coordinated local and systemic responses to developmental insults that ensure that normal body proportions are maintained .
An important question in biology is how cells integrate intrinsic and extrinsic information such that their combined behaviours produce higher-order processes and structures , as seen during organogenesis and tissue repair . In Drosophila larvae , injured imaginal discs can undergo compensatory proliferation [1] as well as secrete an alarm signal that triggers both a systemic developmental delay and reduced growth of the spared imaginal discs [2–5] . Together , these processes allow the damaged tissue ( s ) to catch-up with other tissues , but the role of damaged versus undamaged cells remains controversial [6 , 7] . In vertebrates , systemic growth reduction after injury in a nonessential organ has not been reported . However , systemic catch-up growth has been described after transient impairment of whole-body growth [8–10] , and local growth compensation can occur after unilateral manipulation of long bones within the limbs [11] . Tight control of inter-limb and limb–body proportions are critical for efficient locomotion and interaction with the environment , and therefore long bones are an excellent model for studies of growth regulation . Growth of the long bones is driven by a process called endochondral ossification ( reviewed in [12 , 13] ) . Once mesenchymal cells condense into the template of the skeletal elements , they differentiate into collagen II–expressing chondrocytes that go through sequential differentiation steps from bone ends to centre . Round resting chondrocytes give rise to flat proliferative cells that form columns and , after a few rounds of duplication , cease proliferation and start to differentiate into hypertrophic chondrocytes . These cells increase their volume as they progress towards the centre of the shaft , lay down a collagen X–rich extracellular matrix , and secrete factors that recruit vasculature and bone precursors ( osteoblasts ) from the perichondrium , a fibrous layer that wraps the cartilage [14 , 15] . Some hypertrophic chondrocytes die by apoptosis , while others transdifferentiate into osteoblasts [16 , 17] . Osteoblasts form the primary ossification centre by replacing the original matrix with bone . This process is later repeated at the ends of the bone ( epiphyses ) , forming the secondary ossification centres . The so-called growth plate remains as a cartilage disc between the primary and secondary ossification centres and responds to both intrinsic and extrinsic factors that ultimately regulate bone length . For example , indian hedgehog ( IHH ) , secreted by pre-hypertrophic chondrocytes , and parathyroid hormone-like peptide , secreted by resting chondrocytes , form a negative feedback loop that couples chondrocyte proliferation and differentiation ( reviewed in [12 , 13] ) . This loop is the main conduit through which other local signals , such as fibroblast growth factors and bone morphogenetic proteins , exert their function , often impacting on the expression of key transcription factors . A number of systemic and local extrinsic signals ( growth hormone , insulin-like growth factors [IGFs] ) also play crucial roles in the modulation of chondrocyte activity and bone growth [18 , 19] . As per the regulation of growth after an insult , it has been proposed that bone catch-up growth is due to a cell-autonomous delay in the normal developmental decline of chondrocyte proliferation , such that when the insult is lifted , the formerly arrested chondrocytes retain a higher proliferative potential correlating with the stage at arrest [9 , 20] . It was suggested that a similar mechanism applies to other organs [21] . However , the possible contribution of unaffected cells has not been examined , which is important because a cell-autonomous mechanism does not account for cases in which catch-up growth is faster than expected for the observed maturation delay ( reviewed in [13 , 22] ) . Here , we developed new mouse models to transiently decrease long-bone growth in mice in order to determine the contributions of cell-autonomous and nonautonomous regulation during catch-up growth . Namely , we blocked proliferation in 50% of the cartilage chondrocytes that drive long-bone elongation , specifically in the left hindlimbs , such that the right limb remains as an internal control . Unexpectedly , left–right symmetry was maintained , revealing the deployment of compensatory mechanisms . Locally , we observed hyperproliferation of wild-type ( WT ) chondrocytes that mostly compensated for the lack of proliferation of their arrested neighbours . Systemically , a mild growth reduction affected the rest of the body , contributing to maintenance of limb–body proportions . In summary , our results reveal that long-bone catch-up growth shares some similarities with the response of imaginal discs to developmental insults in insects . We found that this response is mostly cell nonautonomous , representing a paradigm shift in the field that opens up new research avenues for basic and translational studies .
A major roadblock for studies of intra- and inter-organ growth regulation in mouse embryos has been a lack of models in which growth rate can be altered in a specific cell type within an organ , and ideally in only one of two paired organs , leaving the unmanipulated organ as an internal control . To address this deficiency , we devised new mouse models of inducible and transient growth inhibition in the left limb . We generated an Igs7TRE-LtSL-p21/+ allele , a variant of a double-conditional allele [23] , to achieve doxycycline ( Dox ) -tuneable misexpression of the cell cycle suppressor Cdkn1a ( p21 hereafter ) [24] in the cells in which activities of the bacteriophage recombinase Cre and of the ( reverse ) tetracycline transactivator ( [r]tTA ) intersect ( Fig 1A and 1B ) . Due to a floxed tdTomato-STOP sequence ( LtSL ) , expression of tdTomato ( tdT ) takes place in cells expressing ( r ) tTA but having no history of Cre activity , whereas p21 is expressed in the cell population with a history of Cre and current ( r ) tTA activity ( Fig 1A ) . We named the general type of allele Dox-controlled and Recombinase Activated Gene OverexpressioN ( DRAGON ) . By combining the DRAGON-p21 allele with an asymmetric-Pitx2-enhancer-Cre line expressing Cre in the precursors of the left limb mesenchyme ( S1A–S1F Fig ) [25] and a cartilage-specific Col2a1-rtTA line containing the reverse tetracycline transactivator under the control of a type II collagen promoter [26] ( Fig 1B ) , Dox-dependent ectopic p21 expression was achieved specifically in non-hypertrophic chondrocytes of the left limb cartilage elements ( Fig 1C and 1C’ ) . Consequently , any growth adjustment detected in the right limb of triple transgenic animals ( Pit-Col-p21 ) when compared to control littermates must be due to activation of a systemic effect or inter-organ communication . When Dox was administered from embryonic day ( E ) 12 . 5 until birth ( ePit-Col-p21 model ) , analysis at E14 . 5–E17 . 5 revealed the expected cartilage-exclusive expression of tdT , mainly in the right skeletal elements , and p21 expression preferentially in the left limb cartilage , albeit in a mosaic fashion . For example , 36%–67% versus 0 . 8%–23% of chondrocytes were found to be p21+ in left versus right proximal tibia ( S1G–S1K Fig , Fig 1C–1F; n = 3 E15 . 5; n = 5 E17 . 5 proximal tibias; n = 3 E17 . 5 proximal humerus ) . As we previously observed with the Cre transgene [19] , the activity of Cre and therefore p21 expression was more widespread in the left hindlimb than in the left forelimb ( S1I–S1K Fig; only 22%–38% of chondrocytes were p21+ in the left proximal humerus ) . Therefore , we focused our initial analysis on the hindlimb . As expected , proliferation was inhibited in p21+ proximal tibia chondrocytes at E15 . 5 and E17 . 5 ( Fig 1D and 1E and 1G; n = 3 and n = 5 , respectively ) . Although a potential consequence of p21 misexpression in proliferative zone ( PZ ) chondrocytes could have been their premature differentiation , we did not find precocious expression of chondrocyte maturation markers ( e . g . , Ihh , Col10a1 , Cdkn1c , S2A and S2B Fig ) . Moreover , p21 expression did not induce cell senescence ( monitored by expression of p19 and p16 ) by E17 . 5 ( S2D Fig ) , nor did it hamper chondrocyte survival at E15 . 5 or E17 . 5 ( S2E Fig ) . However , the normal expression domains of Ihh , Cola10a1 , and Cdkn1c in ( pre ) hypertrophic chondrocytes ( which do not express the transgene ) appeared slightly fainter in the distal femur and proximal tibia cartilage , and RNA sequencing ( RNA-seq ) analysis of the combined proliferative and pre-hypertrophic zones revealed that their normalized counts were diminished in the left versus right cartilage ( although only significantly for Cdkn1c . S2A and S2B Fig and S1 Data , S2 Data . See S6 Fig for description of the RNA-seq experiment ) . These results suggest that expression of p21 causes only a mild maturation impairment . We next examined whether mosaic chondrocyte arrest altered left bone growth . As a first test , we compared the lengths of several left forelimb and hindlimb bones of ePit-Col-p21 mice to Pitx2-Cre; Igs7TRE-LtSL-p21/+ control mice ( ePit-p21 hereafter ) . At E17 . 5 , the left bones were 0 . 2–0 . 3 mm ( approximately 10% ) shorter in animals misexpressing p21 ( Fig 2A , n = 7 for ePit-p21; n = 15 for ePit-Col-p21 femora and radii; n = 4 and 11 for humeri and tibiae ) , indicating that blocking chondrocyte proliferation resulted in decreased bone growth . However , the effect was milder than expected , given that between one-third and two-thirds of chondrocytes were being arrested . This result suggested that compensatory mechanisms that minimized the impact of the p21 insult had been activated in the left limbs . Indeed , at E15 . 5 or E17 . 5 , no major changes in the length of the proliferative or hypertrophic zones of the growth plate were found ( S2F and S2G Fig ) . We next took advantage of our unilateral approach that provides an experimental and a control limb within an animal and performed left–right intra-individual comparisons to determine the degree of asymmetry . Unexpectedly , most ePit-Col-p21 bones measured at E17 . 5 or birth ( P0 ) showed no obvious difference in their left/right length ratio compared to ePit-p21 control littermates . The one exception was a transient small reduction in the size of the left radius compared to the right ( Fig 2B and 2C , n ≥ 4 for ePit-p21; n ≥ 11 for ePit-Col-p21 at E17 . 5; n ≥ 5 and 6 at P0 ) . These results suggested that the right ( i . e . , control ) bones in ePit-Col-p21 mice responded to the impact of p21 expression in the left limbs via a systemic effect or inter-organ communication , such that their growth was reduced similarly to that of the left bones . The compensatory response observed in the limbs of ePit-Col-p21 mice could be due to both bone-intrinsic cellular mechanisms and the aforementioned apparent extrinsic regulation . We first tested for any organ-intrinsic responses and started by examining proliferation of the spared chondrocytes in the hindlimbs . Indeed , the left/right ratio of 5-ethynyl-2’-deoxyuridine ( EdU ) incorporation by p21− chondrocytes was higher in experimental animals as compared with controls at E17 . 5 and P0 but not E15 . 5 ( Fig 3A and S2H Fig ) , revealing cell-nonautonomous compensatory proliferation of p21− cells in the presence of p21+ neighbours . Because p21− cells did not differ in size from those of control mice ( S2I Fig ) , the hyperproliferation of these cells at E17 . 5 likely contributes to the lack of a left-specific growth reduction in ePit-Col-p21 embryos . In fact , overall EdU incorporation in left and right ePit-Col-p21 PZs ( without distinguishing between p21+ and p21− cells ) , while tending to be slightly reduced , was not significantly different , indicating that the compensatory proliferation phenomenon is quite effective ( Fig 3B ) . Moreover , the proliferative disadvantage of p21+ versus p21− chondrocytes in the left limb of ePit-Col-p21 mice resulted in dilution of p21+ chondrocytes from 45%–50% of PZ chondrocytes at E15 . 5 and E17 . 5 to approximately 20% at P0 ( Fig 3C and 3D; n = 3 , 5 , and 8 , respectively ) , and this depletion was not due to inactivation of rtTA activity ( Fig 3D; n = 3 at E17 . 5; n = 3 at P0 ) . As a means to examine whether compensatory proliferation was only dependent on local cell–cell interactions , we cultured left and right E15 . 5 ePit-Col-p21 tibiae ( together in the same well ) for 2 d with Dox , in the absence of surrounding mesenchyme ( Fig 4A ) . We found that the distal tibia showed chondrocyte proliferation throughout all section levels , with very few or no senescent cells ( S3A Fig , n = 3 ) , whereas EdU incorporation was not detected in the inner core of the proximal tibia ( bulkier than the distal one ) , probably because of insufficient nutrient diffusion . We therefore focused our analysis on the distal epiphysis . Similar to our findings in vivo , EdU incorporation in p21− chondrocytes was significantly higher in the left as compared to the right cultured cartilages ( Fig 4B and 4C ) , suggesting that compensatory proliferation is a bone-intrinsic phenomenon or at least does not require constant interaction with the surrounding mesenchymal tissues . As a control , we tested whether right cartilage proliferation was impaired due to the bone being cultured with the injured left tibia , since this would cause the impression of compensatory proliferation taking place in the left bone . We therefore cultured left and right tibiae from each embryo in different wells ( S3B Fig ) and found that EdU incorporation in distal right tibiae was not significantly different between bones cultured together ( n = 4 ) or separately ( n = 6 ) from the left bones . We next addressed whether the proportion of p21+ chondrocytes in the growing cartilage influences the extent of compensatory proliferation . Given that forelimb bones show a lower proportion of p21+ chondrocytes than hindlimb bones within each embryo analysed ( S1K Fig; 55 . 4% ± 11 . 2% in tibia versus 32 . 8% ± 7 . 1% in humerus; n = 3; p = 0 . 0267 for paired t test ) , we first tested whether compensatory proliferation was triggered in the proximal humerus . We found that although there was a trend towards increased proliferation in left p21− chondrocytes , the difference was not significant ( Fig 4D; n = 3 ) , suggesting that compensatory proliferation requires a minimum insult threshold to be triggered . Because intrinsic differences between forelimb and hindlimb bones might exist in regards to the compensatory proliferation response , we also tested the threshold hypothesis using only hindlimb bones . In order to induce p21 expression in fewer chondrocytes than in ePit-Col-p21 mice , we made use of a newly generated Col2a1-tTA line ( see Materials and methods ) in place of the Col2a1-rtTA transgene , such that p21 expression was achieved from approximately E12 . 5 onwards in the absence of Dox ( Pit-tTA-p21 model ) ( S4A Fig ) . tTA expression in this line is less extensive than rtTA in the Col2a1-rtTA line , as p21 misexpression was detected in fewer left tibial chondrocytes than in ePit-Col-p21 left tibia ( 30%–40% at E15 . 5 , 15%–35% at E17 . 5 , 10%–20% at P0; S4A Fig ) . Consistent with our prediction of a threshold being needed , compensatory proliferation was not detected in the Pit-tTA-p21 model ( S4B and S4C Fig ) . To further investigate whether a minimum insult threshold is required to trigger increased proliferation , we calculated the correlation coefficient between the percentage of p21+ chondrocytes and the extent of proliferation in the PZ from left and right ePit-Col-p21 ( in vivo and ex vivo ) and Pit-tTA-p21 tibiae , at E17 . 5 ( or E15 . 5 plus 2 d ex vivo ) . Segmental linear regression analysis revealed that the extent of EdU incorporation by p21− chondrocytes did not correlate with the proportion of p21+ neighbours when this proportion was below 35% , but beyond this threshold , there was a linear correlation between both parameters ( Fig 4E; n = 26 bones ) . These results suggest that compensatory proliferation is due to a signal produced in proportion to the number of arrested chondrocytes , that the signal needs to reach a certain threshold to be effective , and that it remains active until at least P0 despite the dilution of p21+ chondrocytes . We next asked whether additional cellular changes occur in the left limbs of ePit-Col-p21 mice that could contribute to the growth compensation and potentially correlate with the number of insulted chondrocytes . Because an alteration in cell density can influence organ size , we tested whether cell density was changed in the PZ of ePit-Col-p21 mice . That was indeed the case , and we found a temporal association between the occurrence of compensatory proliferation in the ePit-Col-p21 model ( i . e . , at E17 . 5 and P0 but not E15 . 5 ) and statistically significant reduction of cell density in the left PZ as compared to the right ( Fig 4F ) . Notably , left and right PZ cell densities were not significantly different at any stage in ePit-p21 mice ( Fig 4F , n = 12 ) . Moreover , in line with the threshold hypothesis , we found that , at E17 . 5 , there was a certain value of cell density below which EdU incorporation sharply increased in p21− chondrocytes ( Fig 4G , n = 20 bones ) . Given our initial finding that left and right limb bones exhibit reduced growth in ePit-Col-p21 embryos as compared to ePit-p21 littermates ( Fig 2 ) , we next investigated whether there was a systemic response to the p21 insult in the left limbs . Because the growth phenotype is quite mild ( an approximately 10% reduction ) , we first confirmed the finding by measuring micro-computerized tomography ( μCT ) -generated 3D reconstructions instead of flat micrographs ( S5A and S5B Fig; n = 7 for ePit-p21 and n = 13 for ePit-Col-p21 embryos ) . We found a very good correlation between both types of measurements ( S5C Fig , n = 80 bones ) and therefore used flat micrographs for all measurements in the study . We first tested whether the growth reduction affected the whole body . We found that , in addition to a decrease in right bone length , body weight of E17 . 5 and P0 ePit-Col-p21 mice—but not E15 . 5 or E16 . 5 embryos—was approximately 10% lower than in ePit-p21 littermates ( Fig 5A–5C , S5D Fig ) . Furthermore , the bone-length and weight effects required Dox treatment and therefore p21 expression ( Fig 5A–5C ) . As control experiments , we confirmed that there was no leakiness of the intersectional misexpression strategy ( S5E Fig ) that could account for the systemic growth reduction and that misexpression of tdT in all chondrocytes did not cause a systemic growth reduction ( Fig 5B and 5C ) . Our results thus revealed a whole-body response to an insult in a specific tissue in mice , similar to what has been described in Drosophila larvae [2–5] . In order to characterize the cartilage response , we performed an RNA-seq experiment to identify differentially expressed genes ( DEGs ) between left and right cartilage ( PZ plus pre-hypertrophic region of proximal and distal tibia and femur ) of single ePit-Col-p21 embryos at E17 . 5 ( S6A–S6E Fig , S1 Data and S2 Data ) . Indeed , overrepresentation analysis of the DEG ( adjusted p-value ≤ 0 . 05 ) showed enrichment of several pathways related to stress and immune responses in the left cartilage ( S6F Fig ) . In particular , we found several stress-related transcripts that shared a similar left–right pattern of expression within each embryo ( S6G Fig ) and verified their enrichment in the left cartilage by quantitative real-time polymerase chain reaction ( qRT-PCR ) ( Fig 5E ) or in situ hybridisation ( Fig 5F ) . Relaxin1 , the closest homologue to dilp8 , the recently identified [3 , 27] alarm gene in fly , was not expressed at significant levels in either limb ( S6E Fig ) , suggesting that the mechanism that links the local insult with a systemic response has diverged during evolution . With regards to the relationship between the extent of insult and the induction of a systemic response , Pit-tTA-p21 mice did not trigger a systemic growth defect at E17 . 5 or P0 ( S4D and S4E Fig , summary in Fig 6A ) , suggesting that the systemic growth reduction , like compensatory chondrocyte proliferation , is only triggered when a certain insult threshold is surpassed in the targeted cartilage . We reasoned that the most likely foetal organ to control systemic growth by responding to a circulating alarm signal is the placenta because in rodents it produces higher IGF levels than any other organ [28] and is considered the main organ controlling foetal growth [29] , whereas hepatic IGFs regulate systemic growth mainly after weaning [18] . Placental weight was not diminished in ePit-Col-p21 embryos ( n = 19 ) as compared to ePit-p21 controls ( n = 17 ) , such that the placenta/body weight ratio was increased ( Fig 6B ) . This result suggests that placental efficiency is reduced in response to the left-cartilage p21 insult . To determine the status of placental IGF signalling , we tested the expression of several pathway members by qRT-PCR and found that Igf2 levels were increased in the placentas of ePit-Col-p21 embryos as compared to ePit-p21 controls , whereas the levels of Igf1r did not vary significantly ( Fig 6C; n = 3 experimental and n = 4 control ) . Increased expression of IGF2 by the placenta has been seen as part of a placental stress response triggered by prenatal insults such as alcohol exposure , which is also associated with placental functional impairment , increased placental/body weight ratio , and foetal growth restriction [30 , 31] . Therefore , one possible interpretation of our results is that the left limb cartilage stress response is relayed to the placenta , which then indirectly impacts on foetal growth . Perhaps explaining why increased IGF2 expression does not correlate with enhanced embryo growth , we found an increase in the level of Igf2r ( Fig 6C; n = 6 experimental and n = 6 control ) , which encodes a decoy receptor that can decrease local IGF2 availability [32] . Furthermore , inhibition of IGF2R has been shown to boost placental efficiency [33] . As a means to test whether the systemic growth reduction in ePit-Col-p21 embryos was due to impaired IGF signalling within the placenta , we injected pregnant dams ( from E15 . 25 to E17 . 25 , 3 times per day , Fig 6D ) with an IGF2 analogue ( Leu27-IGF2 ) that does not cross the placental barrier , can only bind IGF2R ( and not IGF1R ) , and thus was shown to increase placental efficiency [33] . Body and placental weight and femur length of ePit-Col-p21 embryos were compared between litters that were either treated or not treated with Leu27-IGF2 to determine the degree to which body/organ size was rescued , and they were also compared with ePit-p21 embryos within treated litters to determine whether Leu27-IGF2 differentially affected experimental and control embryo growth . Boosting placental function led to the following results:
In summary , our results show that when the embryonic long bones experience mosaic inhibition of chondrocyte proliferation , an adaptive growth response can be triggered that involves cell-nonautonomous local mechanisms and systemic changes during the time frame of the insult , such that body proportions are maintained . We refer to this new type of catch-up growth that happens during an on-going insult as ‘adaptive growth’ ( S7 Fig ) . Our finding that a local compensatory response occurs during the insult and involves cell-nonautonomous mechanisms is distinct from previous models that proposed that compensation occurs after the insult is lifted and is cell-autonomous [9 , 11 , 20] . Therefore , our results introduce a new conceptual framework for interpreting studies of perturbed long-bone growth . Furthermore , the experimental approach we devised for the study of growth regulation in mice makes a strong case for using unilateral perturbation approaches when studying bilateral organs . Although a local response such as compensatory proliferation or reduced cell density could have been unveiled with a mosaic bilateral injury , a subtle body-weight effect would likely be ascribed to the reduced size of all limbs and not to inter-organ communication . Indeed , the hint that prompted us to explore inter-organ communication was the observed reduction in the unmanipulated limb between experimental and control mice . Below , we discuss the potential mechanisms and evolutionary conservation of local and systemic responses to developmental injury . We have shown that a few days into mosaic inhibition of proliferation affecting >35% of chondrocytes of the left limb , spared chondrocytes undergo increased proliferation , such that the overall proliferative rate in the left cartilage almost matches that of the right limb . We propose the following order of events , based on correlative data from our study: Our results reveal a mild but consistent systemic growth reduction ( approximately 10% ) in response to a local insult in the cartilage . We propose that the stress response generated in the left cartilage is somehow communicated to the placenta , which in turn systemically reduces growth ( S7 Fig ) . We suggest the following 2 mechanisms that could account for the injured cartilage–placenta communication: With regards to potential growth correction treatments , it would be important to determine whether all ePit-Col-p21 organs are equally reduced or whether the musculoskeletal system ( which is especially dependent on IGF signalling ) is primarily affected . Resolution of the latter question is currently difficult because the embryos are too small for individual organs to be weighed reliably . Volumetric analyses using mesoscopic techniques such as optical projection tomography [37] on embryos expressing fluorescent reporters in the tissue of interest will be necessary to achieve the necessary level of resolution . An unexpected result of our study is that when placental function is boosted in ePit-Col-p21 concepti through maternal Leu27-IGF2 treatment , long-bone growth is not enhanced to the same extent as body weight , resulting in a reduction in the ratio of bone length to body weight ( Fig 6 ) . Given that the right cartilage templates are not experiencing the same p21 insult as the left ones , the dampened response of the right skeletal elements to the systemic rescue suggests that the insult in the left cartilage influences growth of the right limb through some sort of left–right crosstalk . A similar crosstalk has been previously proposed in studies of amphibian limb regeneration , in which it was shown that amputation of the contralateral limb at the same rostrocaudal level as the originally amputated limb reduced the regenerative rate of the latter , whereas ipsilateral or contralateral amputation at a different rostrocaudal level did not [38] . Moreover , a study of tibial fracture repair in young rats showed that the healing environment of a fractured bone triggers the release of growth-promoting signals in the growth plate of the fractured bone and that the same signalling is induced in the contralateral growth plate [39] . As previously proposed [38] , the most obvious candidate system to mediate crosstalk between the left and right limbs is the nervous system . While the exact mechanism remains to be determined , a recent study showed that peripheral sympathetic nerves might inhibit bone growth in response to sustained social stress [40] . Regardless of the mechanism , these results suggest that the observed systemic growth reduction in ePit-Col-p21 embryos is a combination of 2 effects: reduced growth efficiency of the contralateral bones in response to the left-specific insult , and impairment of placental function that affects the rest of the body . Collectively , our results reveal that the processes leading to coordination of growth within and between organs to achieve normal proportions upon developmental insults are conserved across metazoans . However , the magnitude of the contributions of local and systemic mechanisms likely varies across phyla because the extent of the systemic growth reduction observed in mice seems to be less extreme than in Drosophila , with the caveat that different insults or tissues could elicit distinct responses . The exact underlying mechanisms also vary because we did not observe up-regulation of the dilp8 homologue Relaxin1 in the insulted cartilage . Different molecular mechanisms aside , the compensatory response in vertebrates shares some features with the response in insects , such as our finding that the injured tissue is able to catch up despite being exposed to an environment that stunts growth of the rest of the body . One explanation for this result is that local compensatory proliferation overrides a systemic effect . We further speculate that if the same ‘alarm’ signal were to trigger both the intrinsic and systemic mechanisms following injury , this would provide an evolutionarily advantageous strategy to achieve robust coordination of organ growth . While many unknowns remain in the field of organ growth and repair , further exploration of the mechanisms revealed by this study will open exciting new avenues for basic and translational research and lead to an understanding of human growth disorders .
All animal studies were performed under an approved Institutional Animal Care and Use Committee mouse protocol ( #07-01-001 ) according to MSKCC institutional guidelines . To correct for interlitter variability when studying the effect of p21 misexpression on systemic growth , we normalized each measurement from an experimental animal ( ePit-Col-p21 or Pit-tTA-p21 ) to the average measurement for its control littermates ( ePit-p21 or Pit-p21 ) . This is important because the absolute measurements vary significantly between litters , in part because they differ in exact developmental stage , number of embryos , and age of the mother [36] . For paired measurements , the use of left/right ratios allowed for intra-individual normalization . For each experiment , the minimum sample size was estimated using an online tool ( http://powerandsamplesize . com/Calculators ) , based on the average SD observed in pilot experiments , to achieve an effect size of 0 . 03 in the left/right bone length ratio , 0 . 5 in the left/right ratio of EdU incorporation , or 10% in normalized systemic measurements , with a power of 0 . 8 and a 95% CI . In Fig 5B and 5C , 2 embryos ( one from the ePit-Col-p21 and one from the eCol-tdT populations ) were abnormally small , possibly dead , and were excluded from the analysis . For comparison of qualitative expression , a minimum of 2 specimens per stage and 5 across several stages were used . The investigator measuring bone length was blinded to the treatment/genotype of the specimens . No blinding was done for other measurements . No randomization was used for animal processing . When data were available for control and experimental , a normalized measurement ( left/right ratio or percentage of average control mice ) was calculated for both . Between different time points , the normalized measurements were compared by multiple unpaired t test with Holm-Sidak correction for multiple comparisons . Within the same time point , comparisons were done by an unpaired Mann-Whitney test ( 1 variable and 2 conditions ) , by 1-way ANOVA ( 1 variable and ≥3 conditions ) , or by 2-way ANOVA ( 2 variables and 2 or more conditions ) following a matched ( paired ) design when possible ( indicated when not ) . When left and right measurements were compared within experimental animals only , paired 2-tailed t test was used . For all ANOVA , alpha = 0 . 05 . All relevant parameters for the statistical tests can be found in S1 Table . When parametric tests were used , data normality was confirmed by Shapiro-Wilk test and equality of variance by F-test . Prism7 software ( Graphpad ) was used for most analyses . Most graphs show individual values and mean ± SD , unless otherwise indicated . To generate the Igs7TRE-LtSL-p21 mouse line , the NruI-STOP-loxP-tdTomato-SnaBI fragment in the Ai62 ( TITL-tdT ) Flp-in replacement vector [23] was replaced by a custom NruI-tdTomato-STOP-loxP-MluI-HpaI-SnaBI cassette , to generate an empty DRAGON vector . A PCR-amplified Kozak-Cdkn1a cassette was subsequently cloned into the MluI and SnaBI sites to generate the DRAGON-p21 vector . This vector was then used for recombinase-mediated cassette exchange into Igs7-targeted G4 ES cells [23] . Two successfully targeted clones were injected into C2J blastocysts to generate chimeras , obtaining 27 chimeric males ( out of 30 born ) with 75% to 100% chimerism . Two males from each clone were crossed to BL6 albino mice ( Charles River , Wilmington , MA ) to assess germline transmission and to establish the new mouse lines . To generate the Col2a1-tTA line , a Kozak-tTA fragment was PCR-amplified from plasmid pEnt L1L3 tTA-3 ( Addgene plasmid #27105 , gift of Edward Hsiao ) and cloned into a vector containing the regulatory region of mouse Col2a1 obtained from plasmid p3000i3020Col2a1 [41] . Backbone-free vector DNA was injected into FVB zygotes to generate transgenic lines . Four out of 11 founders transmitted the Col2a1-tTA allele . The progeny of one of those ( founder number 92 ) expressed the tTA faithfully in the highest percentage of chondrocytes and was bred with Pitx2-Cre animals to generate breeders for the experiments . Col2a1-tTA mice were maintained on an outbred Swiss Webster background and genotyped using primers Col2a1-F ( CCAGGGTTTCCTTGATGATG ) and tTA-R ( GCTACTTGATGCTCCTGATCCTCC ) and a standard PCR program with 55°C annealing temperature . The Pitx2-Cre [25] ( kind gift of Dr . H . Hamada ) , Col2a1-rtTA [26] ( kind gift of Dr . K . Posey ) , Ai9 ( R26CAGGS-LSL-tdTomato ) [42] , and Ai62 ( Igs7TRE-LSL-tdTomato ) [23] mouse lines were maintained on an outbred Swiss Webster background and genotyped as previously described . Igs7TRE-LtSL-p21 animals were genotyped like Ai62 mice . Pitx2-Cre/Cre; Col2a1- ( r ) tTA/+ females and males homozygous for the conditional misexpression allele were crossed to generate experimental and control animals . Noon of the day of vaginal plug detection was considered E0 . 5 . The equivalent of E19 . 5 is referred to as P0 . Dox hyclate ( Sigma ) was added to the drinking water at a final concentration of 1 mg/ml , with 1% sucrose to increase palatability . A previously described protocol [43] was slightly adapted to culture embryonic long bones . Briefly , E15 . 5 tibiae were obtained from the embryos of Dox-treated pregnant females , dissected free of as much soft tissues as possible in ice-cold PBS , and then cultured ( at 37°C , 5% CO2 ) in 24-well plates with serum-free DMEM ( Gibco ) containing 0 . 2% Bovine Serum Albumin ( BSA ) , 0 . 5 mM L-glutamine , 40 U/ml penicillin/streptomycin ( Gibco ) , 0 . 05 mg/ml ascorbic acid ( Sigma ) , and 1 mM betaglycerophosphate ( Sigma ) . The medium additionally contained 1 ng/μl Dox to maintain transgene expression . After 2 d , the bones were incubated with 10 μM EdU for 90 min , then fixed in PFA and processed for histological analysis . Note that after 2 d , we consistently observed growth of 19% to 23% in control limbs as compared to the original length . This is less than in vivo ( approximately 87% growth between E15 . 5 and E17 . 5 ) , and the main difference seemed to be at the level of the proximal cartilage , which does not proliferate , likely due to insufficient diffusion of nutrients because it is larger than the distal cartilage . We therefore focused our analysis on the distal cartilage , which at these stages is expected to contribute one-third of total growth [44] , i . e . , approximately 29% , quite similar to the observed growth . Human Leu27-IGF2 ( GroPep , Australia ) was prepared at 500 ng/μl in sterile 0 . 01 N HCl solution and kept at 4°C in between injections . From E15 . 25 to E17 . 25 , the pregnant dam was subcutaneously injected every 8 h , for a total dose of 1 μg/g of body weight per day . Staining of cartilage and bone was performed as described [45] . Bone length was measured on digital micrographs using the Line tool in Adobe Photoshop . Unless otherwise indicated , only the mineralized region was measured . Whole femora and tibiae were scanned using a Scanco μCT 35 ( Scanco Medical , Brüttisellen , Switzerland ) system . Six-μm voxel size , 45 KVp , 0 . 36-degree rotation step ( 180-degree angular range ) , and a 400-ms exposure per view were used for the scans , which were performed in air . Scanco μCT software ( HP , DECwindows Motif 1 . 6 ) was used for 3D reconstruction and viewing of images . After 3D reconstruction , ‘Distance 3D’ tool was used for measuring the length of the ossified region . Three measurements were taken and the average derived for each bone . The observer was blinded to the genotype of the mouse . Mouse embryos were euthanized by hypothermia in cold PBS . Mouse pups were euthanized by decapitation after hypothermia-induced analgesia . Knees ( or isolated full tibiae and femora ) were dissected out , skinned , and fixed by immersion in 4% paraformaldehyde ( PFA; Electron Microscopy Sciences ) in PBS for 2 d at 4°C . After several washes with PBS , the tissue was then cryoprotected , first by brief incubation with a solution of 15% sucrose and then 30% sucrose in PBS for at least 4 h at 4°C , and then embedded in Cryomatrix ( Thermo ) using dry-ice-cold isopentane ( Sigma ) . The knees were oriented sagittally and facing each other , with the tibiae on the bottom of the block ( i . e . , closest to the blade when sectioning ) . Serial 7-micron sections were collected with a Leica Cryostat on Superfrost slides , allowed to dry for at least 30 min , and stored at −80°C until used . For all histological techniques , frozen slides were allowed to reach room temperature in a closed box , and Cryomatrix was washed away for 15 min with warm PBS ( 37°C ) . Sections were incubated in citrate buffer ( 10 mM citric acid , 0 . 05% Tween 20 [pH 6 . 0] ) for 15 min at 90°C , allowed to cool down , washed with PBSTx ( PBS containing 0 . 1% Triton X-100 ) , blocked with 5% BSA in PBSTx for 30 min at room temperature , and incubated with the primary antibody over night at 4°C ( see list of antibodies below ) . After PBSTx washes , incubation with Alexa647- and/or Alexa555-conjugated secondary antibodies ( Molecular Probes; 1/500 in PBSTx with DAPI ) was performed for 1 h at room temperature . After PBSTx washes , the slides were mounted with Fluoro-Gel ( Electron Microscopy Sciences ) . For TUNEL staining , endogenous biotin was blocked after antigen retrieval using the Avidin/Biotin blocking kit ( Vector #SP-2001 ) , and TdT enzyme and Biotin-16-dUTP ( Sigma #3333566001 and #11093070910 ) were subsequently used following manufacturer instructions . Biotin-tagged DNA nicks were revealed with Alexa488- or Alexa647-conjugated streptavidin ( Molecular Probes , 1/1000 ) during the incubation with the secondary antibody . Antibodies ( host species , vendor , catalogue number , dilution ) included tdT ( rabbit polyclonal , Rockland #600-401-379 , 1/500 ) , p21 ( rabbit polyclonal , Santa Cruz Biotechnology #sc-471 , 1/300 ) , p19Arf ( rat monoclonal , clone 12-A1-1 , Novus Biologicals #NB200-169 , 1/100 ) , and p16-INK4A ( rabbit polyclonal , Proteintech #10883-1-AP , 1/300 ) . The protocol described in [46] was followed . For embryos and newborns , samples were not decalcified . Except for Col2a1 , Col10a1 , and Ihh ( provided by Dr . Licia Selleri ) , the templates for most riboprobes were generated by PCR from embryonic cDNA , using primers containing the SP6 or T7 RNA polymerase promoters . Sequence of the primers is available upon request . After purification of the PCR product ( Qiagen PCR purification kit ) , DIG-labelled probes were transcribed following manufacturer instructions ( Roche ) , treated with DNAase for 30 min , and purified by LiCl-mediated precipitation in alcoholic solvent . Probes were kept at −80°C in 50% formamide ( Fluka ) . For immunohistochemistry after in situ hybridisation , sections were incubated in citrate buffer ( 10 mM citric acid , 0 . 05% Tween 20 [pH 6 . 0] ) for 15 min at 90°C , allowed to cool down , washed with PBSTx , and incubated with 1% H2O2 in PBSTx for 1 h to block endogenous peroxidases . After BSA blocking and primary antibody incubation , endogenous biotin was blocked using Avidin/Biotin Blocking kit ( Vector #SP-2001 ) , and then the slides were incubated with a biotinylated secondary antibody . A brown precipitate was obtained using a peroxidase-coupled streptavidin-biotin complex ( Vectastain Elite ABC Kit , Vector #PK-6100 ) and DAB substrate ( Vector #SK-4100 ) , following manufacturer instructions . For in vivo experiments , sagittal sections of the knees were imaged , focusing on the region between both menisci and analysing at least 2 , and typically 4 , sections per limb . For cultured distal tibiae , frontal sections were used because they allow for better identification of the different epiphyseal regions . The transition between round ( resting ) and flat ( columnar ) nuclei , roughly describing an arch between the upper point of the wedges formed by the groove of Ranvier , was chosen as the start of the PZ , while the transition towards bigger , more spaced nuclei ( pre-hypertrophic ) was chosen as the end of the PZ . The point where pericellular matrix is sharply reduced around enlarging chondrocytes was considered as the start of the HZ , while the end of the HZ was marked as the distal end of the last intact chondrocyte . Bright-field and fluorescence images were taken on a Zeiss inverted microscope ( Observer . Z1 ) using Axiovision software ( Zeiss ) . Mosaic pictures were automatically reconstructed from individual 10× ( bright-field ) or 20× ( fluorescence ) tiles . Five mg/ml EdU in PBS was injected ( 50μg/g body weight , s . c for pups , i . p . for adults and pregnant females ) 1 . 5 h before euthanizing the mice . EdU was detected using the Click-iT Alexa488 Imaging Kit ( Invitrogen , C-10337 ) once the immunohistochemistry and/or TUNEL staining were finished on the same slides . The fraction of nuclei that were positive for EdU , p21 , or tdT in the PZ of the cartilage was determined using ImageJ or CellProfiler , followed by manual curation . The PZ was cropped from 20×-imaged sections of left and right experimental and control proximal tibial cartilage , stained for DAPI , p21 , and EdU . The area of the region of interest ( PZ ) was measured in pixels using the Histogram tool in Adobe Photoshop and converted into mm2 using the resolution and scale information . The DAPI channel was segmented and quantified using Cell Profiler . Cell density was calculated as the number of chondrocytes per area unit . The PZ was cropped from 20×-imaged sections of left and right ePit-Col-p21 proximal tibial cartilage . tdT+ chondrocytes were segmented and counted , and their individual area was measured using Cell Profiler . The distal left or right femoral and proximal tibial cartilage from E17 . 5 Pit-Col-p21 embryos was dissected in cold PBS , the condyles and hypertrophic zones removed using a microknife , and the perichondrium removed by a combination of collagenase type II treatment ( Worthington , 2 mg/ml in DMEM , 2 min at room temperature ) and mechanical dissection . Left and right cartilage fragments from each embryo ( number 1 , 2 , and 3 ) were kept in separated tubes and flash-frozen in liquid nitrogen . RNA was extracted using Trizol ( Invitrogen ) and a mechanical tissue disruptor . High-quality RNA was deep sequenced ( ≥50 million paired-end reads ) by the New York Genome Center . Aligned reads were analysed using DESeq2 tool in R . A paired design was used , such that left and right comparison was performed for each specimen , which minimized the effect of sequencing batch and interspecimen variability . DEGs were obtained using a threshold of adjusted p-value ≤ 0 . 05 . NMF library tools were used to generate heatmaps . Enrichment analysis was performed using DAVID [47] and WebGestalt [48] . cDNA was synthesized from purified RNA using iScript reverse transcriptase ( RT ) as described by the manufacturer ( Bio-Rad ) . Each target was amplified in triplicate to obtain an average per sample , using SYBR Green ( Applied Biosystems ) on a StepOnePlus real-time PCR system ( Applied Biosystems ) . Primer sequences are shown in Table 1 . Negative controls ( no template ) and no-RT cDNA controls were included for each primer/sample combination . Relative expression on each sample was calculated by the 2−ΔCT method , with Gapdh ( for cartilage ) or Tbp ( for placenta ) as a reference . | The coordination of organ growth is necessary to attain correct individual organ sizes and body proportions . While extensive studies in insects have revealed that both intra-organ and inter-organ communication mechanisms are involved in regulating organ growth , vertebrate studies have lagged behind . Here , we developed a new mouse model to examine cellular mechanisms underlying growth regulation after a developmental insult . The cell cycle suppressor p21 was expressed in the cartilage that drives growth of the long bones , targeting the left limb exclusively and leaving the right limb as an internal control . By triggering the insult during the last gestational week , we found that left–right limb symmetry was maintained due to the following 2 compensatory mechanisms: ( 1 ) hyperproliferation of the spared cells within the targeted cartilage , which indicates that these cells respond to a signal coming from the arrested cells , and ( 2 ) a growth reduction in the rest of the body , an effect that correlates with changes in the levels of placental insulin-like growth factor ( IGF ) signalling and that can be rescued by boosting placental efficiency . These results reveal that the response to developmental insults is quite evolutionarily conserved across species as well as open new avenues of future research for the development of therapies to treat growth disorders . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"sequencing",
"techniques",
"tibia",
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"chondrocytes",
"developmental",
"biology",
"connective",
"tissue",
"cells",
"skeleton",
"molecular",
"biology",
"techniques",
"embryos",
"rna",
"sequencing",
"cartilag... | 2018 | Cell-nonautonomous local and systemic responses to cell arrest enable long-bone catch-up growth in developing mice |
An important issue associated with the control of visceral leishmaniasis is the need to identify and understand the relevance of asymptomatic infection caused by Leishmania infantum . The aim of this study was to follow the course of asymptomatic L . infantum infection in children in an area of Brazil where it is endemic . The children were assessed twice during a 12-month period . In this population study , 1875 children , ranging from 6 months to 7 years of age , were assessed . Blood samples were collected on filter papers via finger prick and tested by ELISA ( L . infantum soluble antigen and rk39 ) . Seropositives samples ( n = 317 ) and a number of seronegatives samples ( n = 242 ) were subjected to qPCR . After 12 months , blood samples were collected from a subgroup of 199 children and tested for Leishmania spp . to follow the course of infection . At baseline qPCR testing identified 82 positive samples . The prevalence rate , as estimated for 1875 children based on the qPCR results , was 13 . 9% . The qPCR testing of whole blood samples collected from a cohort of children after 12 months ( n = 199 ) yielded the following results: of the 44 ( 22 . 1% ) children with positive qPCR results at baseline , only 10 ( 5 . 0% ) remained positive , and 34 ( 17 . 1% ) became negative; and of the 155 ( 77 . 9% ) children with negative qPCR results , 131 ( 65 . 8% ) remained negative , and 24 ( 12 . 1% ) became positive at the follow-up measurement . The samples with positive findings at baseline ( n = 82 ) had a mean of 56 . 5 parasites/mL of blood; and at follow-up the mean positive result was 7 . 8 parasites/mL . The peripheral blood of asymptomatic children had a low and fluctuating quantity of Leishmania DNA and a significant decrease in parasitemia at 1-year follow-up . Quantitative PCR enables adequate monitoring of Leishmania infection .
Visceral leishmaniasis ( VL ) is a serious public health problem worldwide , and approximately 500 , 000 new cases are reported each year [1] . Brazil accounts for 90% of cases in the Americas , where the causative agent of this endemic disease is Leishmania infantum , an intracellular protozoan of the Leishmania donovani complex [2] . VL in Brazil has long been a typical rural zoonosis . However , since 1980 , this disease became a serious and emerging public health problem in the expanding urban centers of several Brazilian cities with different patterns of economic and social development [3] . In Brazil , 3526 new cases of visceral leishmaniasis were reported in 2010 . Of these cases , 13 . 3% were confirmed in the state of Minas Gerais [4] . In 2011 , in accordance with the Municipality Health Service of Belo Horizonte , state capital of Minas Gerais , the incidence rate of VL was 6/100 , 000 , and the fatality rate was 13 . 9% . Asymptomatic carriers are often reported in studies conducted in VL-endemic areas of Brazil , and they are usually more frequent than clinical cases [5]–[7] . Identifying and understanding the relevance of asymptomatic infection during the parasite transmission cycle is currently one of the key issues associated with the control of visceral leishmaniasis [2] . Difficulty in diagnosing asymptomatic VL because of low levels of antibodies and parasites has been reported [8] , [9] . Currently , molecular techniques such as polymerase chain reaction ( PCR ) are used to confirm the diagnosis of VL . These techniques have good sensitivity ( 75%–98% ) and excellent specificity ( 97%–100% ) . These methods are also more accurate than serological techniques in identifying infection in asymptomatic carriers [8]–[11] . Real-time quantitative PCR ( qPCR ) is a variation of conventional PCR and allows for not only the detection of parasite DNA , but also the accurate quantification of copies of the target DNA sequence . The aim of this study was to follow the course of asymptomatic L . infantum infection in children living in disease-endemic areas of the city of Belo Horizonte , Minas Gerais , Brazil , by using qPCR to detect parasite DNA as well as to monitor parasitemia over time .
This study was approved by the research ethics committees of the Universidade Federal de Minas Gerais ( No 253/09 ) , Belo Horizonte City Hall ( 080 . 2008 ) , and the Centro de Pesquisas René Rachou ( 01/2010 ) . Legal guardians of the children involved in this study were required to sign the Informed Consent Form at baseline and follow-up phases . Furthermore , the children were granted medical care and treatment when necessary . A population-based survey , followed by a cohort study , was carried out from 2009 to 2010 . The purpose of this study was to identify and monitor asymptomatic infection of L . infantum in children under the age of 7 years living in Belo Horizonte , the capital of Minas Gerais , using qPCR . Belo Horizonte is located 859 . 19 meters above sea level and between latitudes 44°03′47″and 43°51′27″ and longitudes 19°46′35″ and 20°03′34″ . It has a population of 2 , 375 , 244 inhabitants [12] . The regional climate is predominantly tropical , with rainy summers and dry winters . This study was conducted in three contiguous geographic subareas in the northwestern region of Belo Horizonte . These subareas were chosen because they have different histories of disease progression and control . The necessary sample size of approximately 1875 children was estimated according to the following parameters: prevalence of asymptomatic infection rate ranged from 2 . 4% to 5 . 6% , based on a study conducted in the metropolitan region of Belo Horizonte [13]; α = 0 . 05; 1−β = 0 . 80; and an estimated precision of 0 . 08 . Initially , the children living in the study area were randomly selected based on the municipal census , which is conducted through the Family Health Program , a strategy currently adopted by the Brazilian Unified Health System ( Sistema Único de Saúde - SUS] . These data are periodically updated by community health workers who conduct a census in the areas studied . Serological tests were performed to identify asymptomatic carriers , followed by qPCR for VL diagnosis and estimation of parasitemia in the sample group . After 12 months of follow-up , a subgroup of the survey sample was assessed to determine the course of infection . Blood samples were collected from these patients for further molecular , hematological , and biochemical tests . Further clinical evaluations were conducted by medical specialists at the health centers in the study regions to detect the presence or absence of signs and symptoms of VL . The collection of data and biological material at baseline ( survey ) was performed at the children's homes by trained staff . Finger-prick blood samples were collected on filter paper ( Whatman , number 4 ) . Data regarding the children were recorded on standardized forms . The samples were dried and stored at −20°C until testing . Twelve months after the initial survey , samples were collected at the health centers , the children were clinically examined , and the results were recorded on standardized forms . For this follow-up assessment , peripheral blood samples were collected in appropriate tubes , with ethylene diamine tetra-acetic acid ( EDTA ) . Enzyme-linked immunosorbent assays using L . infantum–soluble antigen ( ELISA-SLA ) and recombinant K39 antigen ( ELISA-rK39 ) were performed on blood samples collected on filter paper at baseline , according to the protocol proposed by Pedras et al . ( 2008 ) and Ho et al . ( 1983 ) [14] , [15] . At baseline , samples that had positive results for Leishmania on at least one serological test were also tested by real-time qPCR . In addition , a subgroup consisting of 15% of the serologically negative samples was also evaluated by molecular assay . At follow-up , all whole-blood samples collected with EDTA were tested by qPCR . The DNA from blood sample collected at both time points was extracted using the DNA extraction kit QIAamp DNA Mini ( QIAGEN GMbH; Hilden , Germany ) , according to the manufacturer's instructions . Three circular fragments , each approximately 5 mm in diameter , were punched from each filter paper blood sample with paper perforators . The protocol for disinfecting the paper perforators that was proposed by Bonne et al . [16] was followed to avoid contamination among the samples . The extraction kit was used at follow-up for 120 µL of whole blood , obviously without the step of filter paper elution . The volume of blood subjected to extraction ( 120 µL ) was equivalent to the volume of the eluate obtained from the filter paper used at baseline . The concentration and purity of the DNA extracted at both time points were determined by spectrophotometry . A260 and A280 were measured with the NanoDrop ND-1000 spectrophotometer ( Thermo Fisher Scientific; Wilmington , DE , USA ) . The ribosomal RNA small subunit gene ( SSU rRNA ) is conserved in all species of Leishmania and was chosen as the target of qPCR amplification [17] , [18] . The forward primer is LEIS . U1 ( 5′-AAGTGCTTTCCCATCGCAACT-3′ ) , and the reverse primer is LEIS . L1 ( 5′-GACGCACTAAACCCCTCCAA-3′ ) . These two primers amplify a 67–base pair ( bp ) fragment of SSU rRNA and were used together with a TaqMan probe , LEIS . P1 ( FAM 5′-CGGTTCGGTGTGTGGCGCC-3′ TAMRA ) [19] . The primers used in the qPCR protocol in this study are not specific to L . infantum and can amplify other organisms of the Leishmania genus . Standard curves were generated for each assay using known amounts of PCR 4 TOPO vector ( Invitrogen; São Paulo , Brazil ) containing the 67-bp fragment of L . infantum SSU rRNA . Serial dilutions ( ×10 ) of the recombinant plasmid were made . It is generally accepted that 160 copies of the SSU rRNA gene are present within a single cell of the parasite [17] . Each sample was tested in duplicate in a final volume of 20 µL , including 3 µL of extracted DNA . The final concentrations of the reagents used were as follows: TaqMan MasterMix ( Applied Biosystems; Foster , USA ) 1×; Probe LEIS . P1 , 0 . 25 pmol/µL; primer LEIS . L1 , 0 . 3 pmol/µL; and primer LEIS . U1 , 0 . 3 pmol/µL . The conditions used for amplification were as follows: 50°C for 2 min for activation of the UDG enzyme , and 95°C for 10 min followed by 40 cycles of 95°C for 15 s for denaturation and 60°C for 1 min for annealing . Values for the threshold of detection and the baseline were automatically determined using StepOne software , v2 . 1 ( Applied Biosystems ) . A blank consisting of the reaction mixture and water instead of DNA template was added in each qPCR run . Furthermore , as a control for DNA extraction , amplification , and quality , real-time qPCR assays of the human gene ACTB were performed . The primers Aco1 and Aco2 [20] , which generate 120-bp fragments , were used in this assay . The cycling parameters were universal , and the melting analysis was conducted based on the parameters of the StepOnePlus Thermal Cycler ( Applied Biosystems ) . SYBR Green ( Applied Biosystems ) was used for detection of the product . Samples from five children who received a diagnosis of the clinical form of VL before the treatment were used as positive controls for quantification . DNA was extracted from these samples with the same method used for extraction from whole-blood samples of study participants . The standard curve and the conditions of the qPCR assay were the same . The parasite load was expressed as the number of parasites present in 1 mL of blood , normalized per nanogram of extracted DNA . This approach neutralizes small variations in the quantification method . Calibrated pipettes were used to test the blood samples collected on filter paper , and the volume of blood in each paper circle was estimated . The estimated volume for the three fragments was approximately 100 µL of blood . For reproducibility analysis , 10% of the samples collected at both study time points were re-evaluated under the same conditions as those used in the first analysis . A PCR assay was performed to confirm infection by L . infantum . The chosen primers were sense RV1-CTTTTCTGGTCCCGCGGGTAGG and antisense RV2-CCACCTGGCCTATTTTACACCA , which amplify a fragment with 145 bp of kDNA from L . infantum . The protocol for the reaction was proposed by le Fichoux et al . ( 1999 ) [21] . Databases were generated using EpiData , version 3 . 2 ( EpiData Association; Odense , Denmark ) and data were analyzed with STATA 11 . 0 software ( Stata Corp . ; College Station , TX , USA ) . The qPCR-based prevalence of infection was calculated based on the following parameters: I ) the percentage of samples with positive qPCR results and negative serology results; II ) the expected number of positive qPCR results if all samples had been tested; and III ) the total and estimated numbers of positive qPCR results . Prevalence was assessed as the relationship between the estimated number of positive qPCR results and total number of samples ( n = 1875 ) . The median values of parasite load were calculated and compared by Kruskal-Wallis test at both time points . The agreement between qualitative tests was estimated by the kappa statistic .
In preliminary experiments , the performance of the Leishmania qPCR assay with the TaqMan detection system was assessed through serial dilutions of the linear plasmid DNA containing the 67-bp fragment , with concentrations ranging from 2 . 8×107 to 2 . 8×102 copies ( analogous to 1 . 76×105 to 1 . 7 Leishmania cells ) . The mean standard curve calculated from 17 independent experiments was linear over at least 6 log10 range of DNA concentrations , with a correlation coefficient of 0 . 989 . The PCR efficiency of amplification was 97 . 4% . The inter-assay coefficients of variation , calculated from triplicate of the 10-fold plasmid DNA dilutions , ranging from 1 . 7×105 to 1 . 7 parasites and performed on separated runs , were as follows: 2 . 35% , 3 . 05% , 3 . 64% , 3 . 27% , 3 . 35% , and 3 . 46% , respectively . The intra-assay coefficients of variation were also calculated from triplicate of the same six different concentrations and performed on the same plate . The results were as follows: 0 . 31% , 0 . 15% , 0 . 50% , 0 . 39% , 0 . 62% , and 0 . 98% , respectively . Leishmania qPCR demonstrated a good detection limit , since samples with 2 . 8×102 copies of linear plasmid DNA , equivalent to 1 . 7 Leishmania cells tested positive . The detection limit was also assessed through serial dilutions of L . infantum DNA in water to obtain the points of the curve spanning from 12 . 000 to 0 . 012 pg DNA/µL . To convert the amount of DNA detected into number of parasites , the genome size of a diploid L . infantum cell ( generally assumed to be 3 . 2×107 bp , [http://www . sanger . ac . uk] was converted into a mass equivalent , yielding a value of approximately 0 . 070 pg . As the qPCR assay uses 3 µL of DNA template , 0 . 036 pg of genomic DNA was detected , representing less than a single parasite cell . Results of melting curve analysis of the samples tested by qPCR of the ACTB gene were satisfactory ( mean melting temperature: 81 . 28°C ) . Therefore , the specific amplification of this gene was confirmed . The qPCR reproducibility assessment demonstrated regular concordance ( kappa = 0 . 59; 95% CI 0 . 36–0 . 80 ) after evaluation of 10% of the samples collected at both study time points . Parasite load was assessed at both time points . The samples with positive results at baseline ( n = 82 ) had a mean concentration of 56 . 5 ( ±33 . 4 ) parasites/mL of blood and a median of 60 . 9 ( range: 16 . 0–88 . 8 ) parasites/mL . At 12-month follow-up , the mean and median parasitemia of the children who remained positive were 7 . 8 ( ±7 . 0 ) and 5 . 8 ( range: 3 . 8–6 . 7 ) parasites/mL of blood , respectively ( Figure 1 ) . The mean and median values for the 24 children who became positive at follow-up were 7 . 9 ( ±6 . 5 ) and 5 . 0 ( range: 3 . 3–10 . 0 ) parasites/mL of blood , respectively . The Kruskal-Wallis test showed a statistically significant difference between baseline and follow-up values of parasitemia ( p = 0 . 0019 ) . Children with the clinically manifest form of VL , the positive control group , presented an estimated mean parasitemia of 2190 . 2 ( ±694 . 6 ) parasites/mL .
The results of this study showed a prevalence of Leishmania infantum circulating in 13 . 9% ( 95% IC 12 . 4–15 . 5 ) of asymptomatic children younger than 7 years living in an urban area that is endemic for VL . There was a significant decrease in parasitemia in these children after one year . A number of children had varying parasitemia values , but on average , their parasite levels were lower than those of the sick children in our control group . The implementation of effective measures to control VL is increasingly urgent because of the spread of this endemic disease in urban areas . Diagnostic tests in dogs and humans are essential for identifying the areas with the highest transmission rates . Although dogs are the main reservoir of L . infantum in urban areas , the prevalence of asymptomatic VL can serve as an indicator of the extent and maintenance of parasite transmission [5] , [8] , [22] , [23] . Therefore , in this study children were selected for the sample group to allow for the identification of relatively recent transmission events in the study area . The screening method chosen was serology , because it is more useful when testing many samples . However , several studies have already shown that ELISA is not an accurate technique for the diagnosis of asymptomatic infections; rather , it is better suited for identifying individuals who exhibit clinical signs and symptoms of VL . Therefore , when serology testing alone is used to identify asymptomatic infection , a large number of asymptomatic carriers are likely to be disregarded [24] , [25] . In light of this fact , molecular methods have been considered an important complement to serological detection of asymptomatic carriers . Thus , this study used a protocol for the detection of Leishmania DNA by qPCR in samples of peripheral blood collected on filter paper via finger prick or venipuncture . The baseline results showed a weak association between serology and qPCR findings . Of the samples with positive qPCR findings , 59 . 8% were also seropositive , and 40 . 2% were seronegative . Differences between these techniques have also been reported by other authors and may be due to serological testing in asymptomatic individuals for whom the tests are not accurate [23]–[26] . Furthermore , the techniques are based on different parameters for detecting infection: antibodies or genomic DNA . Variation in mode of transmission , such as simple contact/exposure versus genuine , established infections , may also be a factor . Moreover , results from the serum-reactive samples may include false-positives . Because this study aimed to identify areas with recent Leishmania transmission events , and given that qPCR identifies the presence of parasite DNA , this molecular method can be considered a technique that provides more robust results than other methods . A recent study demonstrated that Leishmania DNA rapidly degrades soon after death of the amastigote . This suggests that results obtained by qPCR are due to parasites that are alive and intact [27] . Standardization of blood extraction from the filter paper allowed us to use qPCR in an epidemiological survey involving a large number of participants . The methods used were adequate and maintained the quality of qPCR quantification . Blood collection on filter paper is an acceptable alternative to venipuncture , especially when the participants are children , because finger pricking is more practical and less invasive . Follow-up blood collection was carried out by venipuncture because of the smaller number of participants , who had previous positive results , and was performed at health centers . Because the children participating in the study live in VL endemic areas where transmission occurs continuously , their infection status may have changed during the follow-up period [28] , [29] . Recent infection may be responsible for the positive test results of children who had previously tested negative . In contrast , the 10 children who tested positive at baseline and follow-up may have maintained the original infection or been reinfected . None of the cohort subjects showed signs or clinical symptoms associated with VL throughout the duration of the study . The absence of clinical signs or symptoms of VL has already been observed in other studies of individuals with L . infantum infection , which was detected by molecular tests and evaluated during an extended follow-up period [5] , [11] , [24] . Quantitative PCR allowed for not only diagnosis , but also quantification of Leishmania DNA . Assessment of parasitemia at baseline and follow-up demonstrated a decrease in mean parasite load after 12 months ( from 56 . 5 to 7 . 8 parasites/mL ) . This finding suggests that , with regard to the samples that tested positive both times ( n = 10 ) , asymptomatic infection may be self-regulated , with the partial or total clearance of parasites , as demonstrated by the patients who tested positive at baseline and negative at follow-up ( n = 24 ) . Possible limitations of this study involve subjects lost to follow-up and the use of different procedures to collect biological material at baseline and follow-up . The main reasons for failure to present for follow-up examination were migration to other areas , denial , and lack of interest in continued participation . A comparison of demographic characteristics , such as age , sex , and duration of residence in the area , revealed no significant differences between those who were and were not included at follow-up . Therefore , selection bias may be ruled out when analyzing the results . It is important to highlight that despite the use of two different methodologies to collect biological material in both phases of the study , all procedures were performed in such a way as to minimize the possible variations intrinsic to the different techniques: the blood volume used for DNA extraction at follow-up was adjusted to be similar to the volume collected from the filter paper ( at baseline ) , which was approximately 120 µL . Furthermore , the parasite load values calculated at the end of experiments were standardized and normalized per nanogram of extracted DNA , which also minimized potential variations . The low number of circulating parasites in asymptomatic carriers may influence the reproducibility of the analytic technique ( kappa = 0 . 60 ) because of the decreased probability of finding parasites in peripheral blood samples . To investigate parasite load , we performed a similar assay , following the same criteria as those used in tests previously performed on samples from symptomatic children with confirmed clinical VL ( positive controls ) . Parasite load was significantly greater in the positive controls than in asymptomatic carriers , with thousands of parasites per milliliter of blood . This finding suggests that parasite load data should not be used alone , but rather in combination with monitoring of each patient or in a comparative manner between groups of individuals with clinical presentation of confirmed VL and asymptomatic individuals . Recent studies showed that the level of parasitemia estimated by qPCR is related to clinical patient profile [30] , [31] . Molecular techniques , especially qPCR , to estimate the prevalence and incidence of infection can be ideal for epidemiological studies because they can detect and quantify Leishmania DNA , which is a reliable marker of infection [32] . However , the greater complexity and cost involved in performing qPCR on a large number of samples , compared with the complexity and cost associated with immunological tests , must be taken into account . In conclusion , the findings of this study demonstrate the usefulness of qPCR for epidemiological studies monitoring parasitemia in asymptomatic individuals . Despite the cost , qPCR performance was satisfactory . It detected the clearance of parasite DNA , which may be relevant in cases of infection in which the tests are significant limitations on the interpretation of their results . Therefore , this method allows for better monitoring of Leishmania infection in areas where it is endemic . | In Brazil , visceral leishmaniasis ( VL ) is caused by the parasite protozoon Leishmania ( Leishmania ) infantum ( syn chagasi ) , which is transmitted to humans by sandflies that have widespread wild and domestic animal reservoirs . An important issue associated with the control of VL is the need to identify and understand the relevance of asymptomatic infection in the transmission cycle of the parasite . We evaluated the course of asymptomatic infection in children by using molecular methods to detect parasite DNA and to monitor parasitemia level over time . The prevalence of VL , as detected by means of a molecular method , for 1875 children was 13 . 9% . The data demonstrated a significant decrease in parasitemia after 1 year . Furthermore , the level of parasitemia was generally lower in asymptomatic children than in sick children . None of the asymptomatic children showed signs or clinical symptoms associated with VL throughout the duration of the study . This molecular method of measuring parasitemia is adequate for monitoring VL transmission in areas where it is endemic . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"molecular",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"leishmaniasis",
"neglected",
"tropical",
"diseases"
] | 2012 | Low Parasite Load Estimated by qPCR in a Cohort of Children Living in Urban Area Endemic for Visceral Leishmaniasis in Brazil |
Enterovirus 71 ( EV71 ) has caused great morbidity , mortality , and use of health service in children younger than five years in China . Vaccines against EV71 have been proved effective and safe by recent phase 3 trials and are now available in China . The purpose of this study was to evaluate the health impact and cost-effectiveness of a national EV71 vaccination program in China . Using Microsoft Excel , a decision model was built to calculate the net clinical and economic outcomes of EV71 vaccination compared with no EV71 vaccination in a birth cohort of 1 , 000 , 000 Chinese children followed for five years . Model parameters came from published epidemiology , clinical and cost data . In the base-case , vaccination would annually avert 37 , 872 cases of hand , foot and mouth disease ( HFMD ) , 2 , 629 herpangina cases , 72 , 900 outpatient visits , 6 , 363 admissions to hospital , 29 deaths , and 945 disability adjusted life years . The break-even price of the vaccine was $5 . 2/dose . When the price was less than $8 . 3 or $14 . 6/dose , the vaccination program would be highly cost-effective or cost-effective , respectively ( incremental cost-effectiveness ratio less than or between one to three times China GDP per capita , respectively ) . In one-way sensitivity analyses , the HFMD incidence was the only influential parameter at the price of $5/dose . Within the price range of current routine vaccines paid by the government , a national EV71 vaccination program would be cost-saving or highly cost-effective to prevent EV71 related morbidity , mortality , and use of health service among children younger than five years in China . Policy makers should consider including EV71 vaccination as part of China’s routine childhood immunization schedule .
Enterovirus 71 ( EV71 ) is one of the major agents that cause outbreaks of hand , foot , and mouth disease ( HFMD ) and herpangina worldwide[1] . Since the 1990s , the epidemic has mainly affected the Asia-Pacific region and EV71 has become a major public health issue across this region[2 , 3 , 4 , 5] . HFMD is characterized with fever and cutaneous lesions on hands , feet and buttocks , along with oral lesions . Although most cases are mild and self-limiting with an average duration of 7 days , approximately 1% can rapidly develop severe and even life-threatening complications such as encephalitis , aseptic meningitis , pulmonary oedema/hemorrhage and heart failure[1] . During the period from 2008 to 2012 , China reported more than 7 million children with HFMD , of which around 45% were associated with EV71[6] . During the period from May 2008 to December 2014 , China reported death of 2 , 225 children due to HFMD , with a case-fatality rate of 0 . 03% and 93% of them were associated with EV71 [6 , 7] . Current treatment is only to relieve symptoms . No specific drug to treat EV71 infection is available [1] . With limited impact of personal and environmental hygiene , vaccination is considered as the most effective and promising strategy to prevent HFMD and herpangina caused by EV71 [8] . Since 2013 , three phase 3 randomized clinical trials ( RCTs ) to evaluate efficacy of inactivated EV71 vaccines in infants and young children have been completed in China [9 , 10 , 11] . The vaccines showed high efficacy and satisfactory safety to provide protection against EV71-associated diseases and are now available in China . In 2010 , before the key clinical trials were initiated , an cost-effectiveness analysis estimated economic value of a future vaccine against EV71[12] . Here , to assist policy makers in evaluating the implication of a national EV71 vaccination program in China , we reassessed the public health impact and cost-effectiveness of EV71 vaccination , using new evidence on the vaccine safety and efficacy as well as updated clinical and economic data on EV71 associated infections .
Using Microsoft Excel , a decision tree model was built to calculate the net clinical and economic outcomes of EV71 vaccination compared with no EV71 vaccination ( Fig 1 ) . This model adopted Markov chain and hypothesized a 2012 birth cohort of 1 , 000 , 000 Chinese children . As most affected cases are younger than five years and the rates of severe illness and mortality decrease substantially in older children and adults[6] , the model’s time horizon was five years after birth . The time step was one year . If children experienced symptomatic infection of EV71 , they died or suffered from one of the following diseases: herpangina , mild HFMD , and severe HFMD[6] . Patients with HFMD were categorized as severe if they had any neurological complications ( encephalitis , aseptic meningitis , or flaccid paralysis ) , or cardiopulmonary complications ( pulmonary hemorrhage , pulmonary oedema , or myocarditis ) , or both; otherwise , they were classified as mild cases[6] . According to experience in China , almost all cases with HFMD make outpatient visits before deciding to receive home care or to be hospitalized for further treatment; a small number of cases with mild HFMD , almost all cases with severe HFMD and almost all death cases are hospitalized; herpangina alone is not an indication for hospitalization . Life years lost after the 5 years were taken into accounted . Accordingly , the model simulated events over a 5 year horizon but accounted for outcomes over the total lifetime of the simulated individuals . The primary result was presented as costs per disability adjusted life year ( DALY ) averted . The overall incidence of HFMD was 1 . 2 per 1 , 000 person-years from 2008 to 2012 , varied among provinces ranging from 0 . 2 in Tibet to 3 . 1 in Hainan according to the Chinese Center for Disease Control and Prevention ( China CDC ) [6 , 13] . EV71 accounted for 45% of mild , 80% of severe , and 93% of fatal cases and these proportions did not vary significantly with age among children aged 5 and under[6] . Thus , we calculated the annual incidences of mild , severe , and fatal EV71-associated HFMD by age from the corresponding overall incidences of HFMD by age and the proportions associated with EV71 . We used their average values of four years ( 2009–2012 ) in our base-case analysis ( Table 1 ) . Herpangina has not been included in surveillance system in mainland China . The reporting of cases of herpangina and EV71 are aggregated together in Taiwan; specific data on the epidemiology of herpangina are not available . Fortunately , studies supplied information to calculate the ratio of patients with EV71-associated herpangina to that of EV71-associated HFMD . The ratio ranged from 0 . 044 to 0 . 11 , with weighted mean of 0 . 069 , using study sample size as the weight[11 , 14 , 15] ( Table 1 ) . The incidence of EV71-associated herpangina was calculated using this ratio and the incidence of EV71-associated HFMD from the China CDC[6] . Several studies reported the spectrum of complications of severe EV71-associated HFMD; nevertheless , they were single-center in design , had small-sized sample or short duration of enrollment , or the cases were selected[14 , 15 , 16 , 17] . Chen et al summarized hospitalized cases of EV71-associated HFMD in Taiwan from 1998 to 2005[18] . These cases were reported to surveillance systems at the Taiwan CDC by 538 hospitals of various levels . Based on them , the proportion of each complication was calculated for this analysis ( Table 1 ) . The data from the largest pediatric infectious disease center in Shanghai between 2007 and 2010 showed a hospitalization rate of 14% for all 28 , 058 patients diagnosed as HFMD and 54% of the inpatients were positive for EV71[19] . The hospitalization rate of EV71-associated HFMD was calculated as: 0 . 14× ( 0 . 54/0 . 45 ) = 0 . 168 , in which 0 . 45 represented the proportion of EV71 in all HFMD cases according to the China CDC[6] ( Table 1 ) . Another survey from Guangdong reported a similar hospitalization rate [20] . The frequencies of outpatient visits for each symptomatic case were not available specifically for EV71-associated HFMD and herpangina . The frequency for overall HFMD patients was used in this analysis ( Table 1 ) . Recently , three multicenter , randomized , double-blind , placebo-controlled phase 3 trials evaluated the efficacy and safety of inactivated EV71 vaccines in healthy infants and young children in China[9 , 10 , 11] . The 1-year efficacies ranged from 90% to 97 . 4% against EV71-associated HFMD . We performed a meta-analysis using a random-effect model . The results showed an overall efficacy of 95% , with 95% confidence interval of 90%-98% . One of the trials reported efficacy against EV71-associated herpangina[11] . However , due to sparse events , no significant result was reached . In the absence of other data , this analysis assumed that the vaccine efficacy against EV71-associated herpangina was the same as that against HFMD . Extended follow-up of one trial showed that the antibody titers were maintained at a high level through two years post-vaccination [9 , 21] . There are no long-term results for the other two trials . However , one of them reported consistent titers from month 6 to month 12 post-vaccination [11] and the other one reported slightly waned titers at day 180 after vaccination [10] . Therefore , we assumed that the titers do not wane significantly in our model , of which the time horizon is just five years . The rate of adverse events within 28 days after vaccination was 56% on average ( range 47%-71% ) ( Table 1 ) [9 , 10 , 11] . Most of the adverse events of EV71 vaccines were mild . Serious adverse events , which were considered to be associated or most likely associated with vaccination , happened only in 0 . 04% of the participants ( range 0–0 . 1% ) ( Table 1 ) [9 , 10 , 11] . The schedule of vaccination against EV71 was two doses , 4 weeks apart[9 , 10 , 11] , given at 3 and 4 months of age [22] . As this schedule is the same as that for the first two doses of diphtheria , tetanus and pertussis ( DTP ) vaccine , DTP coverage was used to estimate EV71 vaccine coverage[22] . Due to the lack of data on the coverage of the second DTP dose , data on the third DTP dose was used . Data are limited to estimate the efficacy of a single dose of EV71 vaccine . It was assumed to be 50% in this analysis[22] . We estimated DALYs using 2010 life expectancy data of China[23] . DALYs are the sum of years of life lost ( YLLs ) and years of life lost due to disability ( YLDs ) [24] . The durations of herpangina and mild HFMD are both 7 days on average[25] . According to Xu et al , the mean duration of hospitalization of severe HFMD ( including critical cases ) was 16 days[26] . There is no data on the duration of disability after discharge from hospital . Therefore , the duration was underestimated . In the base-case analysis , we assumed no disability following discharge . In sensitivity analysis , we explored how its uncertainty influenced the cost-effectiveness results . Disability weights ( DW ) for each condition were taken from the World Health Organization’s estimates and a previous cost-effectiveness analysis ( Table 1 ) [12 , 27] . DWs for conditions with combined complications were not available . For simplicity , the highest DW was used if the patients suffered more than one complication . This cost-effectiveness analysis was conducted from a societal perspective . The costs for EV71-associated HFMD and herpangina incorporated direct medical costs and non-medical costs for physician visits , medications , lab tests , and transportation , and indirect costs for work loss ( S1 and S2 Tables ) . To today , seven studies have gathered these cost data from outpatient visits and hospitalizations in various regions of China[28 , 29 , 30 , 31 , 32 , 33 , 34] . The reported costs were weighted by the reported number of cases in each study to estimate average costs for each treatment setting ( outpatient or hospitalization ) ( Table 1 ) . In China , vaccines are either supplied by commercial market or Expanded Program on Immunization ( EPI ) . The latter is paid by the government . This analysis is to give an implication whether EV71 vaccines should be included in EPI in China . As the prices of vaccines in EPI are no more than $4 . 59 per dose , the analysis showed more concern for the case of $5 . 0 per dose ( close to $4 . 59 ) . As far as we know , recently EV71 vaccines have become commercially available in China and the price is around $30-$40 per dose , varied among regions . Right now the vaccines are paid by parents and the coverage is relatively low according to experiences from other commercial vaccines in China . The vaccine price may change in the future . Therefore , we performed the analysis at a range of prices for vaccines . Our analysis used the range from $2 . 5 to $40 ( $2 . 5 , $5 , $10 , $20 , $30 , $40 ) per dose because this range covers almost all prices of China made vaccines . The price of vaccine administration was estimated at 3 Chinese Yuan ( CNY ) per injection according to subsidy policies to health facilities for vaccine administration ( range 2–4 CNY , to cover the costs of nurse labor , syringe and transportation and storage of vaccine ) [35] . The costs of vaccine-associated adverse events were considered in this analysis and they were obtained from published studies ( Table 1 ) [12 , 36] . All costs were converted to 2012 US Dollars ( 1 US Dollar = 6 . 30 CNY ) using the medical care component of the Consumer Price Index[37] . Incremental cost-effectiveness ratio ( ICER ) was calculated using the following formula: ICER= ( Costno vaccination−Costvaccination ) / ( Effectno vaccination−Effectvaccination ) The numerator was the difference in total costs with or without vaccination . The denominator was DALYs that vaccination averted . There is no official guidance on discounting in China . All costs and DALYs were discounted to 2012 amounts at a rate of 3% annually ( range 0–10% ) according to Weinstein et al[38] . The cost-effectiveness thresholds were based on the WHO standard ( highly cost-effective , ICER < GDP per capita; cost-effective , GDP per capita < ICER < 3×GDP per capita; and not cost-effective , ICER > 3×GDP per capita ) [39] . The GDP per capita for China in 2012 was approximately $6 , 300[37] . To assess the robustness of the model and to identify influential model inputs for which additional data are warranted , one-way sensitivity analyses were performed at each level of vaccine price . There are substantial heterogeneities of disease incidence and costs among different regions in China . Thus , a two-way sensitivity analysis was performed to evaluate their influence on the base-case results for the case of $5 per dose . The ranges of model inputs for sensitivity analysis were all listed in Table 1 . All data used in this study are available through references .
Table 2 shows the clinical events in a birth cohort of 1 , 000 , 000 Chinese infants followed for five years with or without EV71 vaccination . EV71 vaccination would be expected to annually avert 37 , 872 cases of EV71-associated HFMD , 2 , 629 cases of EV71-associated herpangina , 72 , 900 outpatient visits , 6 , 363 admissions to hospital , 29 deaths , and 945 DALYs among children younger than five years . The economic burden of EV71-associated HFMD and herpangina incorporating direct and indirect costs is approximately 13 million dollars per 1 , 000 , 000 Chinese infants followed for five years . Table 3 shows the costs per DALY averted by the EV71 vaccination program at various prices per dose . According to WHO cost-effectiveness criteria , the vaccination program would be cost-saving at $2 . 5 and $5 . 0 per dose , cost-effective at $10 , and not cost-effective at $20 , $30 and $40 . The break-even price of the vaccine is $5 . 2 per dose . When the price is less than $8 . 3 or $14 . 6 per dose , the vaccination program would be highly cost-effective or cost-effective , respectively . Fig 2 shows the impact of HFMD incidence on the cost-effectiveness of EV71 vaccination at various prices per dose . As the incidence of HFMD falls below 0 . 3 , 0 . 5 , 0 . 9 , 1 . 6 , 2 . 4 , and 3 . 2 per 1 , 000 person-years , the vaccination program would be not cost-effective at the prices per dose of $2 . 5 , $5 , $10 , $20 , $30 , and $40 , respectively . Table 4 shows how other parameters influence the ICER comparing EV71 vaccination with no vaccination . At prices per dose less than $5 , EV71 vaccination is still cost-saving or highly cost-effective when the parameters are varied across their ranges . At $10 , the discount rate is the only influential parameter . When the cost and DALYs are both discounted at a rate more than 6% , EV71 vaccination would be no longer cost-effective . At $20 , EV71 vaccination would be cost-effective only when the cost and DALYs are both discounted at a rate less than 2% . At the prices more than $30 , EV71 vaccination would not be cost-effective when any parameter in Table 4 is varied across its range . A series of tornado diagrams show the rank of parameters’ influence on ICER at the prices per dose of $10 , $20 , $30 , and $40 , respectively ( S1 Fig ) . Fig 3 shows how HFMD incidence and disease costs influence the ICER when the vaccine price is $5 per dose . If the disease costs increase by 50% , the vaccination program would be not cost-effective in regions where the incidence of HFMD is below 0 . 4 per 1 , 000 person-years . If the disease costs decrease by 50% , the incidence making vaccination not cost-effective is below 0 . 6 per 1 , 000 person-years .
A national EV71 vaccination program would prevent a substantial portion of EV71 related morbidity , mortality , outpatient visits , and admissions to hospitals among children younger than five years in China . Within the price range of current routine vaccines paid by the government , the program is cost-saving or highly cost-effective . Policy makers should consider including EV71 vaccination as part of China’s routine childhood immunization schedule . | Enterovirus 71 ( EV71 ) has caused great morbidity , mortality , and use of health service in children younger than five years in China . Recently , effective and safe vaccines against EV71 have been approved . Whether EV71 vaccination should be included as part of China’s routine childhood immunization schedule is unknown . In this study , we built a decision model to evaluate the health impact and cost-effectiveness of a national EV71 vaccination program in China . We find that vaccination would annually avert 567 , 500 cases of hand , foot and mouth disease ( HFMD ) , 40 , 000 herpangina cases , 1 , 093 , 500 outpatient visits , 95 , 500 admissions to hospital , 435 deaths , and 14 , 000 disability adjusted life years based on the current Chinese birth cohort size . The break-even price of the vaccine was $5 . 2/dose . Within the price range of current routine vaccines paid by the government , a national EV71 vaccination program would be cost-saving or highly cost-effective . Policy makers should consider including EV71 vaccination as part of China’s routine childhood immunization schedule . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"children",
"medicine",
"and",
"health",
"sciences",
"cost-effectiveness",
"analysis",
"economic",
"analysis",
"china",
"immunology",
"geographical",
"locations",
"social",
"sciences",
"health",
"care",
"vaccines",
"preventive",
"medicine",
"age",
"groups",
"infectious",
... | 2017 | Cost-effectiveness of a national enterovirus 71 vaccination program in China |
The interferon-inducible transmembrane ( IFITM ) protein family represents a new class of cellular restriction factors that block early stages of viral replication; the underlying mechanism is currently not known . Here we provide evidence that IFITM proteins restrict membrane fusion induced by representatives of all three classes of viral membrane fusion proteins . IFITM1 profoundly suppressed syncytia formation and cell-cell fusion induced by almost all viral fusion proteins examined; IFITM2 and IFITM3 also strongly inhibited their fusion , with efficiency somewhat dependent on cell types . Furthermore , treatment of cells with IFN also markedly inhibited viral membrane fusion and entry . By using the Jaagsiekte sheep retrovirus envelope and influenza A virus hemagglutinin as models for study , we showed that IFITM-mediated restriction on membrane fusion is not at the steps of receptor- and/or low pH-mediated triggering; instead , the creation of hemifusion was essentially blocked by IFITMs . Chlorpromazine ( CPZ ) , a chemical known to promote the transition from hemifusion to full fusion , was unable to rescue the IFITM-mediated restriction on fusion . In contrast , oleic acid ( OA ) , a lipid analog that generates negative spontaneous curvature and thereby promotes hemifusion , virtually overcame the restriction . To explore the possible effect of IFITM proteins on membrane molecular order and fluidity , we performed fluorescence labeling with Laurdan , in conjunction with two-photon laser scanning and fluorescence-lifetime imaging microscopy ( FLIM ) . We observed that the generalized polarizations ( GPs ) and fluorescence lifetimes of cell membranes expressing IFITM proteins were greatly enhanced , indicating higher molecularly ordered and less fluidized membranes . Collectively , our data demonstrated that IFITM proteins suppress viral membrane fusion before the creation of hemifusion , and suggested that they may do so by reducing membrane fluidity and conferring a positive spontaneous curvature in the outer leaflets of cell membranes . Our study provides novel insight into the understanding of how IFITM protein family restricts viral membrane fusion and infection .
The interferon ( IFN ) system is the first line of host defenses against pathogen invasion , including viral infections . It protects by producing hundreds of IFN-stimulated genes ( ISGs ) that modulate diverse biological functions . A number of ISGs ( such as PKR , RNase L , ISG 15 , etc . ) have been characterized and shown to suppress viral replication , the mechanisms of which are still poorly defined ( reviewed in reference [1] ) . One exciting development in the last few years has been the discovery of some novel ISGs , also known as cellular restriction factors ( such as APOBEC3G , Trim5α and Tetherin , etc . ) , which intrinsically block different steps of retroviral replication [2] , [3] , [4] , [5] , [6] . It is notable that many viruses , including retroviruses , have evolved to acquire a variety of strategies that evade IFN-mediated restrictions [7] , [8] , [9] . This type of intrinsic immunity is believed to play crucial roles in virus-host co-evolution and viral pathogenesis [7] , [10] . The interferon-inducible transmembrane ( IFITM ) protein family belongs to a group of small ISGs ( ∼15 kD ) that has recently been shown to block early stages of viral replication [11] , [12] . Originally identified through RNAi genetic screening and shown to inhibit infections by influenza A virus ( IAV ) , West Nile virus and Dengue virus , the IFITM proteins are now known to potently restrict entry and infections by a number of highly pathogenic viruses , including HIV-1 , filovirus , and SARS coronavirus [12] , [13] , [14] , [15] , [16] , [17] , [18] . In humans , there are at least 4 functional members of IFITM proteins; IFITM1 , 2 and 3 are expressed in a variety of human tissues and cell lines , IFITM5 is limited to the bone and is involved in mineralization [11] . All of these human IFITM proteins have been shown to restrict viral entry and infection , with IFITM3 being generally thought to be the most potent [12] , [13] , [14] , [15] , [17] , [18] . A recent study demonstrated that the IFITM3 protein significantly restricts the morbidity and mortality associated with influenza , further underscoring the crucial role of IFITM3 in vivo [19] . Yount and colleagues recently showed that the mouse IFITM3 protein is not only palmitoylated but also ubiquitinated , and that these posttranslational modifications distinctly regulate the cellular localization of IFITM3 and its anti-influenza activities [20] , [21] . While it has been suggested that viral membrane fusion may be blocked by IFITMs [12] , [13] , [15] , direct evidence is still lacking and exactly how IFITM proteins restrict virus entry and infection is currently not known . Membrane fusion is an essential step for enveloped viruses to enter host cells and initiate infection , a process that is mediated by the viral fusion proteins present on the surface of virions [22] . To prevent premature activation , viral fusion proteins in the mature viral particles are normally metastable and exist at a high-energy state . Once triggered by specific cellular stimuli , such as receptor binding , a low pH , or both , they undergo a series of conformational changes , resulting in the insertion of the fusion peptide of the viral fusion protein into the target cell membrane , leading to hemifusion , pore formation , expansion , and ultimately , complete fusion [23] , [24] , [25] , [26] . While the general principle of viral membrane fusion has been extensively studied , the detailed molecular mechanisms governing this process are still poorly defined [22] . In particular , how viral membrane fusion is modulated by cellular factors other than the specific triggers ( such as receptor binding , low pH , cathepsin cleavage , etc . ) remains an emerging subject that needs to be explored . In this work , we sought to determine the mechanisms by which cellular IFITM proteins restrict viral membrane fusion and entry . We chose the Jaagsiekte sheep retrovirus ( JSRV ) envelope ( Env ) and IAV hemagglutinin ( HA ) proteins as the model system of study because of some of their advantages . JSRV is a simple retrovirus , with Env-mediated membrane fusion and entry requiring both receptor-binding and low pH; an initial receptor binding primes the subsequent low pH-dependent conformational changes required for full activation [27] , [28] , [29] , [30] . This unconventional two-step triggering mechanism was originally discovered in the avian sarcoma leukosis virus ( ASLV ) [31] , and has now been suggested to operate in other enveloped viruses , including HCV [32] . Compared to most pH-dependent viruses , JSRV has a relatively high pH threshold ( ∼pH 6 . 3 ) for fusion , the process of which likely occurs in a GPI-anchored-protein-enriched endosomal compartment ( GEEC ) or caveolae [27] , [33] , [34] . Thus , study of JSRV Env-mediated fusion should lead to new insights into the mechanism of action of the IFITM proteins . IAV is a prototype pH-dependent virus , the entry and infection of which has been shown to be significantly restricted by IFITM proteins , particularly IFITM3 , both in vitro and in vivo [12] , [13] , [19] . In addition to JSRV Env and IAV HA , which belong to class I fusion proteins , we also explored the inhibitory effects of IFITM proteins on membrane fusion induced by the Semliki Forest virus ( SFV ) E1/E2 and vesicular stomatitis virus ( VSV ) G proteins , which represent class II and III viral fusion proteins , respectively [22] . Hence , the mechanisms uncovered from this study are likely applicable to other viral fusion proteins , and collectively provide critical new insight into our understanding the mechanism by which IFITM proteins restrict viral membrane fusion and entry .
Prior studies focused on IFITM3 , and have suggested that it mainly acts in late endosomes or lysosomes to restrict viruses that fuse at lower pH ( ∼pH 5 . 5 ) than is present in early endosomes [12] , [13] , [14] , [18] . Here we examined if IFITM proteins also restrict entry of JSRV , whose Env-mediated membrane fusion readily occurs at pH 6 . 3 or even higher [27] , [33] . We did so by creating several stable lines expressing human IFITM1 , 2 or 3 and testing their effects on JSRV entry , along with that of several other viruses . We observed that all three IFITM proteins effectively inhibited the infections of MoMLV pseudovirions bearing either IAV HA/NA or VSV-G in HTX cells ( a subclone of the HT1080 cell line ) , with approximately equivalent efficiency ( Fig . 1A; p<0 . 01 ) . Interestingly , the JSRV pseudovirion infection in HTX cells was inhibited by IFITM1 ( p<0 . 01 ) to a much greater extent than by IFITM2 and 3 ( p<0 . 01 and p<0 . 05 , respectively ) ( Fig . 1A ) . IFITM1 also moderately , but consistently , inhibited the infection by amphotropic 10A1 MLV pseudovirions ( p<0 . 05 ) , yet IFITM2 and 3 did not inhibit and even somewhat enhanced entry ( Fig . 1A ) . Similar results were also obtained in 293 cells , where IFITM1 caused the greatest restriction of JSRV entry ( p<0 . 01 ) , although the overall restriction efficiency of these IFITM proteins on VSV and IAV entry was relatively low ( Fig . 1B ) , consistent with a previous report [13] . Immunoblotting revealed that all three IFITM proteins were expressed in both HTX and 293 cells ( Figs . 1C and D ) , with IFITM3 exhibiting a relatively low level of expression in HTX cells ( Fig . 1C ) , which might have contributed to its relatively low antiviral activities in this cell line . To ascertain that the observed phenotypes of IFITM proteins on viral entry was not due to the FLAG sequences attached to their N-termini , we created HTX cells stably expressing wildtype ( WT ) IFITM proteins . We observed similar patterns of restrictions by IFITM1 , 2 and 3 on all viral pseudotypes tested ( Fig . S1 ) . Altogether , these results demonstrate that these three human IFITM proteins effectively restrict IAV and VSV entry , with similar efficiency , while IFITM1 predominantly restricts JSRV entry as compared to that of IFITM2 and 3 . IFITM proteins are normally expressed in cells at a basal level , yet can be significantly induced by type I and type II IFN [35] . To examine if IFN blocks viral entry , we treated 293 or HTX cells with IFN-α2b , a subclass of type I IFN , and examined its effect on pseudoviral infections . We observed that IFN-α2b significantly inhibited infection of 293 cells by JSRV , VSV and IAV pseudovirions ( Fig . 1E; p<0 . 01; data not shown for HTX cells ) . Interestingly , entry of 10A1 MLV was also slightly but consistently blocked by the IFN-α2b treatment ( Fig . 1E; p<0 . 05 ) , similar to previous reports [13] , [36] . The greater inhibitory effect of IFN on IAV and VSV entry into 293 cells ( Fig . 1E ) relative to that of overexpressing individual IFITM proteins ( Fig . 1B ) was unexpected because IFITM proteins had higher levels when overexpressed than when induced by IFN ( Fig . S2 ) . Perhaps IFITMs induced by IFN synergistically cooperated to inhibit viral entry; additionally , ISGs other than IFITMs might have contributed to the observed effects of IFN treatment in 293 cells . No cytotoxicity was observed for the doses of IFN applied during the viral infection period . Given that HTX or 293 cells do not express a significant level of endogenous IFITMs , especially IFITM1 and 3 ( Fig . S2 ) , we next used a myelogenous leukemia line , K562 cells , to address if depletion of IFITM expression would enhance viral entry . We observed that , indeed , the entry of JSRV and IAV , and to a lesser extent that of VSV , was enhanced in K562 cells stably expressing shRNA against IFITM1 or 3 ( kind gifts of Michael Farzan and I-Chueh Huang , Harvard Medical School ) [13] as compared to the parental K562 cells ( Fig . 1F; p<0 . 01 or 0 . 05 ) . The entry of 10A1 MLV was also slightly enhanced by shRNA targeting IFITM1 , but the increase was not statistically significant ( Fig . 1F; p>0 . 05 ) . shRNA did not significantly reduce IFITM2 in K562 cells , and thus we could not assess the consequences of reducing of this protein ( data not shown ) . Overall , these results suggest that endogenous IFITM proteins intrinsically restrict JSRV , IAV and VSV entry . IFITM1 has been previously shown to be associated with caveolin-1 , a protein that is known to play an essential role in caveolin-mediated endocytosis [37] , [38] . Given that the JSRV receptor , hyaluronidase 2 ( Hyal2 ) , is a GPI-anchored protein that is localized in lipid rafts and that JSRV may use a GPI-anchored-protein-enriched endosomal compartment ( GEEC ) and/or a caveolar pathway for entry [33] , [34] , [39] , [40] , we considered the possibility that IFITM1 could preferentially interfere with the binding of JSRV Env to Hyal2 , thereby restricting viral entry . We took advantage of a soluble form of the JSRV SU-human IgG fusion protein ( JSU-hFc ) we previously created , and performed an in vitro binding assay on HTX cells expressing individual IFITM proteins and functional human Hyal2 [27] , [28] , [29] , [41] . Flow cytometry analysis revealed that the fluorescence shifts in HTX cells expressing IFITM proteins , including that of IFITM1 , were similar to those of parental cells ( Figs . 2A and B ) . This indicates that expression of IFITM proteins did not affect the binding of JSRV Env to HTX cells expressing the Hyal2 receptor . We also performed virus binding assays using Gag-YFP-expressing MoMLV ( kind gifts of Walter Mothes ) pseudoviral particles bearing JSRV Env [28] , [42]; again , similar fluorescence intensities were observed among cells expressing IFITM proteins and parental cells ( Figs . 2C and D ) . The expression of IFITM proteins on the surface of HTX cells was also examined by using an anti-FLAG antibody . IFITM1 had a relatively higher level of surface expression as compared to IFITM2 and 3 , but overall the fluorescence signals were low and their differences were not statistically different ( Figs . 2E and F; data not shown ) . We conclude that expression of IFITM proteins , including IFITM1 , does not affect the binding of JSRV Env to its Hyal2 receptor on the cell surface . JSRV Env uses a dual triggering mechanism in which receptor binding primes the Env to undergo low pH-dependent conformational changes that lead to fusion [28] , [29] . This unusual feature allowed us to examine if IFITM proteins may affect receptor-mediated priming for fusion . We performed metabolic labeling of 293T cells co-expressing IFITMs and JSRV Env , and determined shedding of JSRV SU in the presence or absence of a soluble form of Hyal2 ( sHyal2 ) . In our previous studies , we had established that shedding of JSRV SU into culture media is an important indicator of Hyal2 receptor-mediated triggering for the fusion activation of JSRV Env [28] , [29] , [30] . Here we observed that the levels of JSRV SU harvested from the culture media of 293T cells expressing IFITM1 , 2 or 3 were comparable to those of parental cells , and that they all increased with the presence of sHyal2 in a dose-dependent manner ( Fig . 2H ) . The total levels of JSRV Env expression in these radiolabeled cells were approximately equivalent , as evidenced by the intensities of Env precursors and processed TMs ( Fig . 2G ) . Collectively , we conclude that overexpression of IFITM proteins , including IFITM1 , does not affect the expression and trafficking of JSRV Env , nor does it impair receptor-mediated priming for fusion activation . Syncytia formation and cell-cell fusion assays have been instrumental in understanding membrane fusion , including viral membrane fusion [43] . We sought to obtain direct evidence that IFITM proteins may restrict viral membrane fusion mediated by JSRV Env and other viral fusion proteins . As JSRV Env requires Hyal2 overexpression for membrane fusion to be detected at low pH [27] , we generated stable HTX and 293 cell lines overexpressing Hyal2 and IFITM1 , 2 or 3 , which served as target cells for the syncytia formation and cell-cell fusion assays described below . For parental 293 cells overexpressing Hyal2 ( 293/LH2SN , mock ) , we observed almost complete syncytia formation ( ∼100% ) induced by JSRV Env and IAV HA within 5–10 min after a pH 5 . 0 pulse ( Fig . 3A ) . In sharp contrast , very little syncytia formation was detected in 293/LH2SN cells expressing IFITM1 , even after a 1 h recovery period ( Fig . 3A ) . Syncytia formation was also substantially reduced in 293/LH2SN cells expressing IFITM2 , but the reduction was much less in cells expressing IFITM3 , especially in the case of JSRV Env ( Fig . 3A ) . There is much less or no inhibitory effect of these IFITM proteins on entry of 10A1 MLV ( a virus that fuses at neutral pH , Fig . 2B ) [12] , [13] . We therefore measured syncytia formation induced by 10A1 MLV Env ( with its R peptide deleted ) at neutral pH and found , as predicted , that it was not significantly affected by IFITMs ( Fig . 3A ) . The fusion efficiency of 10A1 MLV Env in IFITM1 , 2 and 3-expressing cells , as quantified using fusion index ( 0 . 51±0 . 06 , 0 . 50±0 . 05 and 0 . 52±0 . 04 , respectively ) , was comparable to that in parental cells ( 0 . 52±0 . 06 ) . A similar order of syncytia inhibition by the IFITMs , i . e . , IFITM1>IFITM2>IFITM3 , on JSRV Env and IAV HA was also obtained in cells expressing WT IFITM proteins ( without the N-terminal FLAG tags ) ( Fig . S3 ) . The differential inhibitory effects of IFITMs on syncytia formation of JSRV Env and IAV HA in 293/LH2SN cells were unlikely due to their levels of IFITM expression , which were examined by Western blots ( Fig . 3B ) . Flow cytometry analysis of 293T cells co-expressing JSRV Env and WT IFITM proteins showed that the levels of JSRV Env on the surface of IFITM-expressing cells were comparable to that of the mock control ( Fig . 3C ) , indicating that the reduced syncytia formation was not due to a change in the Env surface expression . To evaluate if the differential effects of IFITM proteins on syncytia formation induced by JSRV Env were dependent over a limited pH range , we treated the JSRV Env-expressing cells with different pH values , i . e . , pH 6 . 2 , 5 . 7 and 5 . 0 , respectively . Under these pH conditions , IFITM1 consistently exhibited the strongest inhibition on syncytia formation induced by JSRV Env ( Fig . S4 ) . Further lowering the pH ( pH 4 . 0 ) or incubating 293 cells with an increased concentration of sHyal2 ( up to 30 µg/ml ) did not overcome the IFITM1-mediated restriction on fusion ( data not shown ) , suggesting that the block by IFITM1 does not occur at the triggering step . It has been previously established for influenza HA that progressively lowering pH causes the activation of more fusion proteins; rather than causing each individual protein to undergo increasingly extensive conformational changes [44] , [45] . Based on the results of JSRV Env described here , we conclude that the mechanism of IFITM inhibition is independent of fusion protein density . We also assessed if treatment of cells with IFN could suppress viral membrane fusion . We observed that , following a 24-h incubation of 293/LH2SN cells with IFN-α2b , syncytia formation induced by JSRV Env or IAV HA was greatly reduced in a dose-dependent manner ( Fig . 3D ) . In contrast , 10A1 MLV Env-mediated syncytia formation was not significantly affected by the IFN-α2b treatment ( Fig . 3D ) . No cytotoxicity was observed during the 24-h IFN treatment period; nor were there any changes in the expression of JSRV Env and Hyal2 , as examined by immunoblotting and flow cytometry ( data not shown ) . Since 293 cells do not express significant amounts of IFITMs , especially IFITM1 and 3 ( Fig . S2 ) , we have been unable to unambiguously determine if depletion of individual endogenous IFITMs by shRNA in 293 cells enhances syncytia formation . Nonetheless , these experiments clearly demonstrated that IFN can block viral membrane fusion . We applied a more quantitative cell-cell fusion assay to evaluate the effects of IFITM proteins on JSRV Env-mediated fusion as well as to understand the possible mechanisms for inhibition . We expressed JSRV Env in 293T effector cells stably expressing GFP , and labeled HTX/LH2SN target cells expressing IFITM proteins with a red-fluorescent dye , CMTMR; cell-cell fusion was measured by a fluorescence microscope and flow cytometry [27] . Similar to the syncytia formation results ( Fig . 3A ) , IFITM1 exhibited the strongest inhibition , reducing the cell-cell fusion efficiency of JSRV Env by ∼50% ( Figs . 4A , B and C; p<0 . 01 ) . IFITM2 and 3 also suppressed cell-cell fusion , but with much less efficiency ( Figs . 4A , B and C; p<0 . 01 ) . Given that the sizes of fused cells in parental HTX/LH2SN cells were much larger than those of fused cells in IFITM1-expressing cells ( Fig . 4A ) , and that the cell populations with larger size were excluded from flow cytometry analysis , it is likely that the percent of fusion reduction measured by flow cytometry ( Figs . 4B and C ) underestimates the inhibitory effect of IFITM proteins ( see another cell-cell fusion assay below ) . Nevertheless , these experiments provide clear indication that the content transfer resulting from cell-cell fusion by JSRV Env was inhibited by IFITM proteins . We explored whether IFITM proteins expressed in effector cells co-expressing the viral fusion proteins could also restrict cell-cell fusion . We expressed IFITM1 in 293T/GFP effector cells , in HTX/LH2SN target cells ( the stable cell lines described above ) , or in both , and compared their effects on JSRV Env-induced cell-cell fusion . Expressing IFITM1 in effector cells ( Fig . 4D , column 3; Fig . 4E ) was as effective in reducing cell-cell fusion by JSRV Env as expressing IFITM1 in target cells ( Fig . 4D , column 2; Fig . 4E ) , and co-expressing IFITM1 in both effector and target cells enhanced this inhibitory effect ( Fig . 4D , column 4; compare columns 2 and 3 with column 4; p<0 . 05 in both cases; Fig . 4E ) . Thus , the inhibition of cell-cell fusion by IFITM1 is not specifically related to its expression in effector or target cells . This result suggests that IFITM proteins are unlikely to suppress cell-cell fusion by directly acting on specific viral fusion proteins or their corresponding receptors , but rather through a common physical mechanism ( s ) ( see below ) . A critical question is whether or not the cell-cell fusion assay employed here is relevant to endosomal fusion; this is particularly important , given that IFITM proteins have been shown to predominantly restrict viruses that require low pH for membrane fusion and entry . To address this question , we took advantage of our previous finding that JSRV pseudovirions are virtually resistant to low pH inactivation and that an extracellular low pH pulse can overcome proton pump inhibitor–mediated block of JSRV entry [27] , [33] , [34] . We pretreated JSRV pseudovirion-bound HTX or HTX/IFITM1 cells with 20 nM bafilomycin A1 ( BafA1 ) , followed by a pH-5 . 0 pulse for 5–10 min; cells were then allowed for infection for 4 h in the presence of BafA1 . We observed that , while low pH substantially rescued the BafA1-mediated block in both the parental HTX and HTX/IFITM1 cells , as would be expected [33] , [34] , the low pH treatment did not increase the JSRV titer in HTX/IFITM1 cells to a level that was similar to that of parental HTX cells ( Fig . 4F ) . This result apparently differed from that of SARS coronavirus , whose fusion inhibition by IFITMs had been previously shown to be bypassed by trypsin [13] . Overall , our data suggest that IFITM1 expressed on the plasma membrane effectively blocks the forced entry of JSRV rendered by the low pH pulse , and this is consistent with the cell-cell fusion data . In order to assess possible broad effects of IFITMs on viral membrane fusion , we employed another cell-cell fusion assay for examination of IAV HA , SFV E1/E2 , and VSV-G , which represent class I , II and III fusion proteins , respectively [22] . In these experiments , effector cells were the NIH 3T3-derived HAB2 cell line stably expressing IAV HA ( kind gift of Judy White ) [46] or COS7 transiently transfected with plasmids encoding SFV E1/E2 or VSV-G; we labeled these cells with calcein-AM . We loaded the target 293/LH2SN cells expressing individual IFITM proteins ( the same cell lines as used for the syncytia formation assay shown in Fig . 3 ) with CMAC , allowing aqueous dye transfer during fusion to be monitored by a fluorescence microscope . We found that , somewhat surprisingly , both IFITM1 and 3 strongly inhibited membrane fusion induced by all three classes of viral fusion proteins , and their efficiencies were almost the same . IFITM2 also inhibited viral membrane fusion , but the efficiency was generally low ( Figs . 5A , B and C ) , consistent with the results of syncytia formation ( Fig . 3A ) and the GFP-transfer cell-cell fusion assay described above ( Fig . 4 ) . Even more surprisingly , we observed that when the JSRV Env protein was expressed in COS7 cells as effector , Env-mediated fusion was also markedly suppressed by IFITM3 ( Fig . 5D ) . This was in sharp contrast to the situation in which IFITM3 had essentially no effect on fusion when 293T cells were used as effector cells to express JSRV Env ( Fig . 5E ) . Perhaps COS7 cells express a specific cellular factor ( s ) that functionally promotes the inhibitory effect of human IFITM3 on viral membrane fusion . Consistent with this notion , COS7 cells expressing human IFITM3 ( also engineered to co-express human Hyal2 , because COS7 cells are not permissive to JSRV infection ) drastically suppressed JSRV entry , with efficiency almost equivalent to that of IFITM1 ( Figs . 5F and S5A ) . This observation was in sharp contrast to the situations in HTX and 293 cells , where IFITM3 had much less effect ( Figs . 1A and B ) . Similarly , the syncytia forming activity of JSRV Env , as well as that of IAV HA , was almost equally inhibited by human IFITM3 and IFITM1 in COS7 cells ( Fig . S5B ) as compared to that in 293/LH2SN cells ( Fig . 3A ) . This cell type-dependent effect on cell-cell fusion mirrored prior reports showing that the IFITM-mediated restriction on viral entry and infection is also cell type dependent [13] , [15] . Collectively , our data demonstrated that these three human IFITM proteins potently suppress cell-cell fusion induced by all three classes of viral fusion proteins . We next examined which steps of the viral membrane fusion process are inhibited by IFITM proteins . For this purpose , we used conditions that , in the absence of IFITM proteins , allow fusion to proceed up to and through the point of hemifusion while preventing the steps that lead to pore formation . We did so by creating an intermediate of fusion , referred to as a cold arrested state ( CAS ) . It has been previously shown that for viral fusion proteins that induce fusion at low pH , lowering pH at the low temperature of 4°C yields hemifusion; raising temperature at neutral pH then leads to pore formation and growth [47] , [48] , [49] . We created CAS for JSRV Env and IAV HA , and found , as expected , that membrane fusion did not occur ( Figs . 6A and B , middle panels ) . Adding CPZ at neutral pH and low temperature to the parental cells ( mock ) resulted in a significant aqueous dye transfer ( Figs . 6A , C , and D ) . Any further dye transfer upon subsequently raising of temperature to 37°C was not statistically significant ( Figs . 6A , C and D ) . For target cells expressing IFITM proteins , aqueous dye spread upon CPZ addition at CAS was much less than for the parental cells ( Figs . 6B , C and D ) . In contrast , raising the temperature led to a significant increase in dye spread; but the total dye spread was still less than for the parental cells ( Figs . 6B , C and D ) . The fact that the amount of dye spread induced by CPZ was relatively small showed that little hemifusion occurred for the IFITM-expressing cells . The greater amount of dye spread upon raising temperature indicated that the IFITM proteins in the target cell membrane block the creation of hemifusion . Temperature dependence of protein conformational changes is generally greater than for those of lipids that are not near a phase transition . We therefore assume that the presence of IFITM proteins blocked the conformations changes in JSRV Env and IAV HA needed for hemifusion . But it is possible that the decrease in membrane fluidity caused by IFITM proteins ( see below ) inhibits lipids from rearranging into a hemifusion configurations . Hemifusion is promoted by negative spontaneous curvature of monolayers that contact each other in binding [50] , [51] . Therefore , making the spontaneous curvature more negative should oppose the inhibitory actions of IFITMs and thereby promote hemifusion . We tested this by adding oleic acid ( OA ) to aqueous solutions for 15 min to allow them to incorporate into plasma membranes . We observed that creating CAS in the presence of OA led to more fusion both when adding CPZ ( Fig . 7A , compare second columns with first columns in each cell line ) and raising temperature ( Fig . 7B , compare second columns with first columns in each cell line ) . For the parental target cells , the addition of OA resulted in fusion between almost all cell pairs ( Figs . 7A and B ) . For target cells expressing IFITM proteins , OA-addition led to a great increase in the percentage of cells pairs that fused ( Figs . 7A and B ) . In sharp contrast , incorporating OA into membranes subsequent to creating CAS was much less effective in promoting fusion upon either the addition of CPZ or the raising of the temperature ( Figs . 7A and B , compare third columns with second columns in each cell line ) . The almost complete rescue of IFITM-mediated restriction on hemifusion by OA further supports the notion that the major block in fusion caused by ITIFM proteins in target membranes occurs upstream of hemifusion . The presence of IFITMs likely blocks hemifusion by making the spontaneous curvature of outer leaflets of plasma membranes more positive . In principle , IFITM proteins could alter spontaneous monolayer curvatures and thereby restrict the creation of hemifusion by influencing the membrane molecular order and membrane fluidity . To explore these possibilities , we labeled 293/LH2SN cells expressing IFITM1 , 2 or 3 or mock control with Laurdan , a hydrophobic fluorescent probe that is highly sensitive to lipid phases , and measured their generalized polarization ( GP ) values and florescence lifetimes using 2-photon laser scanning and fluorescence-lifetime imaging microscopy ( FLIM ) [52] , [53] . Because of its large excited state dipole moment to align surrounding water molecule in the energy dissipation process , Laurdan has been commonly used to report the extent of water penetration into the lipid bilayer , which correlates with lipid packing [52] , [54] . In general , higher GP values and longer lifetimes indicate that the membranes are more molecularly ordered , while lower GP values and shorter lifetimes mark membranes as less molecularly ordered [52] . In mock control cells , the GP distribution was characterized by two peaks , one with a lower GP , associated with intracellular membranes , and another with a higher GP , identified with plasma membranes; the less ordered populations were predominant in the mock controls ( Figs . 8A–D; first row ) . In cells treated with methyl-β-cyclodextrin ( MβCD ) , a cholesterol-depleting reagent known to make the membrane less ordered [55] , we observed dramatically reduced GP signals , resulting in an almost complete loss of the higher GP peak in the histogram ( Figs . 8A–D; second row ) . In sharp contrast , cells expressing IFITM1 , 2 or 3 all exhibited marked increases in the GP value , which was particularly evident in the higher GP peak ( Fig . 8A–D; third , fourth and fifth rows ) , suggesting that these cell membranes are more ordered than those of mock controls . Quantitative analysis showed that the averaged GPs values of IFITM1 and 3 in the plasma membranes were significantly different from that of mock controls ( p = 0 . 00672 and p = 0 . 00107 , respectively ) ( Fig . 8E ) . The averaged GP values for the intracellular membranes of IFITM1 and 3-epxressing cells were also increased , albeit not statistically significant from those of mock controls ( p = 0 . 08∼0 . 37 ) ( Fig . 8E ) . Unfortunately , we have been unable to distinguish the lipid order of endosomal membranes from that of total intracellular membranes in this analysis . Noticeably , IFITM2 exhibited modest increases in the GP value ( p = 0 . 1388 ) ( Fig . 8E ) , which correlated its less inhibitory effect on syncytia formation and cell-cell fusion ( Figs . 3 , 4 and 5 ) . The increased lipid order in cells expressing IFITM proteins was also evidenced by their longer FLIM lifetimes as compared to those of mock controls and MβCD-treated cells ( Fig . S6 ) . Collectively , these results showed that overexpression of IFITM proteins dramatically increases the lipid packing order of cell membranes and makes them less fluid and possibly less competent for membrane fusion ( see below the Discussion ) .
The IFITM protein family is the first and thus far only restriction factor known to block viral entry [12] , [56] . Previous studies have suggested that these proteins , particularly IFITM3 , predominantly restrict viruses that fuse in the late endosomal or lysosomal compartments at a lower pH [12] , [13] , [14] , [17] , [18] . Here we provide evidence that this family of proteins can also effectively restrict viruses that fuse with a higher pH threshold , such as JSRV ( ∼pH 6 . 3 ) which requires both receptor binding and low pH to co-trigger membrane fusion activation . This activation process likely occurs in the GPI-anchored-protein-enriched endosomal compartment or caveolin-associated compartments [27] , [29] , [33] . Interestingly , we found that , among the three human IFITM proteins examined , IFITM1 was more active than IFITM2 and 3 in restricting JSRV Entry ( Figs . 1A and B ) . This was not because of downregulation of JSRV Env or its Hyal2 receptor , nor due to a perturbation of receptor-mediated priming for fusion activation ( Fig . 2 ) . Instead , results of syncytia formation and cell-cell fusion experiments showed that JSRV Env-mediated fusion at low pH was profoundly inhibited by IFITM1 ( Figs . 3 , 4 and 5 ) . Given that an extracellular low pH pulse cannot bypass the IFITM1-mediated inhibition of endosomal entry of JSRV ( Fig . 4F ) , and that IFITM proteins do not inhibit 10A1 MLV Env-mediated fusion at neutral pH ( Fig . 3 ) , we conclude that the syncytia formation and cell-cell fusion assays employed in this study can reflect the situation of viral membrane fusion in endosomes , which is believed to be the predominant site of IFITM-mediated inhibition of viral entry . Somewhat surprisingly , we observed that IFITM1 was also generally more effective than IFITM2 and 3 in suppressing syncytia formation induced by IAV HA ( Fig . 3A ) and VSV-G ( not shown ) , even though these proteins restrict viral entry with almost comparable efficiency ( Figs . 1A and B ) . The exact mechanism underlying these observations is currently unknown , but could be related , in part , to the relatively higher levels of IFITM1 expression on the cell surface in 293 cells as compared to that of IFITM2 and 3 ( Fig . 1E and F ) . Quantitative cell-cell fusion analysis confirmed the syncytia formation data ( Fig . 5 ) , but also revealed that IFITM3 dramatically inhibited viral membrane fusion when COS7 , rather than 293T , was used as effector cells ( Figs . 4 , 5 and Fig . S5 ) . These results indicate that the effects of IFITM proteins on viral membrane fusion can be cell type dependent , which agrees with previous observations on viral entry [13] , [15] . Because all three human IFITM proteins tested exhibited potent restriction of viral membrane fusion induced by all three classes of viral fusion proteins that have different structures ( Fig . 5 ) , we suggest that a common physical mechanism , rather than specific interactions with viral fusion proteins , is responsible . Results from our series of experiments support this hypothesis . Our first line of evidence is the effect of CPZ on cell-cell fusion ( Fig . 6 ) . It has been previously established that CAS is a state of hemifusion , and that the addition of CPZ to cells at the CAS intermediate leads to full fusion [47] , [57] . It is also known that for cells brought to CAS by fusion proteins that are triggered by acidic conditions , raising the temperature at neutral pH leads to full fusion [47] , [48] , [58] . We observed that the extent of aqueous dye spread in target cells expressing IFITM proteins was much less upon CPZ addition or upon raising the temperature from CAS than was fusion induced by lowering pH at 37°C ( Fig . 6 ) . These findings provide strong evidence that IFITM proteins inhibit the creation of hemifusion . Because the effects caused by adding CPZ or raising the temperature were qualitatively similar for JSRV Env and IAV HA ( Figs . 6C and D ) , we conclude that the mechanism of inhibition by IFITM proteins is not dependent on the precise fusion protein . Although IFITM proteins may affect pore formation and/or expansion , their primary mechanism appears to be the prevention of hemifusion . The second line of evidence for the IFITM-mediated block on hemifusion came from the OA experiments ( Fig . 7 ) . It has been repeatedly shown that hemifusion is promoted by negative spontaneous curvature and is inhibited by positive spontaneous curvature [51] , [57] . Consequently , if IFITM proteins conferred positive spontaneous curvature to membranes that contain them , these proteins would naturally block hemifusion . As OA has a large negative spontaneous curvature [57] , we reasoned that it should overcome the inhibitory actions of IFITM proteins if the curvature was at the core of the action of IFITMs . Experimentally , we observed that the addition of OA virtually overcame all of the block of fusion by IFITM proteins ( Figs . 7A and B ) . That is , when hemifusion was induced by the addition of OA , pore formation readily resulted without any inhibition despite the expression of IFITM proteins in the target membrane . The fact that the addition of OA after establishing CAS had no apparent effect on IFITM-mediated inhibition further supports the conclusion that the block occurs at steps prior to the creation of hemifusion ( Figs . 7A and B ) . How can IFITM proteins block hemifusion ? Our Laurdan labeling experiments showed that IFITM-expressing cell membranes were more ordered than those of mock controls , as evidenced by their increased GP values and longer FLIM lifetimes ( Figs . 5 , 8 , and S6 ) . The increase in the lipid order of IFITM-expressing cells , particularly in their plasma membranes , correlates with the potency of IFITMs in suppressing viral membrane fusion ( Fig . 8 and Fig . S6 ) . Thus , IFITM proteins may block hemifusion by decreasing the fluidity of the membrane that contains them: a decreased fluidity would reduce the ability of lipids to undergo movements necessary for achieving hemifusion . It is also possible that the increased exclusion of water from the bilayer , as indicated by the higher GP values in the presence of the proteins ( Fig . 8 ) is due to an increased average area occupied by lipid headgroups relative to the area swept out by their acyl chains . This would be equivalent to a greater positive spontaneous curvature . It remains possible that expression of IFITM proteins alters the lipid composition of cell membranes , thereby influencing their fluidities and spontaneous curvatures . We emphasize that while changes in lipid order and membrane fluidity likely account for the general inhibitory effect of IFITMs on viral membrane fusion , they do not fully explain the virus-specific and somewhat cell type-dependent inhibition of IFITMs on viral entry as reported in this and previous studies ( Figs . 1 , 4 and 5; Fig . S5 ) [12] , [13] , [15] . Additional factors are likely to be involved , such as specific IFITM-binding partners and possibly viral elements that modulate IFITM-mediated inhibition of viral entry . In this respect , IFITM proteins may or may not influence cell-cell fusion mediated by developmental and cellular fusogens , depending on the specific cell systems that express IFITMs and cellular fusogens . The reason IFITM proteins promote greater lipid order remains unclear , but we offer a suggestion . IFITM proteins may directly change membrane curvature by adopting an unconventional membrane topology or topologies that function as a wedge to generate positive spontaneous curvature . There is an increased appreciation that lipid-binding proteins , along with lipids themselves , can influence membrane curvature , which has been shown to be crucial for vesicular trafficking and membrane fusion [59] , [60] . Results of continuum membrane mechanics show that the spontaneous curvature of the monolayer of the target membrane proximal to the membrane expressing the fusion proteins ( i . e . , outer leaflets ) affects hemifusion , but the spontaneous curvature of the distal monolayer ( i . e . , inner leaflets ) does not [61] . Given our experimental data showing that IFITM proteins block hemifusion ( Figs . 6 and 7 ) , we suggest that IFITM proteins affect the outer monolayer , probably by spanning part or all of the outer monolayer . This suggestion is in line with the predicted membrane topology of IFITM proteins [11] , in which N- and C-termini face the lumens of vesicles . It is also supported by prior and our current work showing that both the N and C-terminally tagged epitopes of IFITM3 can be , though not prominently , detected by flow cytometry and immunostaining without permeabilization ( Figs . 2E and F ) [12] , [14] . A recent study suggested that the mouse IFITM3 protein adopt an alternate topology [20] . In this report , the authors provided evidence that the originally predicted transmembrane domains of IFITM3 fold into a hairpin loop and span only the inner leaflets , resulting in an intramembrane topology with both N- and C-termini facing the cytosol [20] . While our data does not unambiguously demonstrate the existence of this alternate topology , it is possible that IFITM proteins adopts multiple and dynamic topologies . In fact , it has been shown that some transmembrane proteins , including those of viral glycoproteins , adopt dual or dynamic topologies because of “lipid flip-flop” and/or changes in the net charge of their cytosolic sequences [62] , [63] , [64] . Dynamic topologies could result from the cleavages of IFITM proteins that have been observed at both the N- and C-termini in mammalian cells [11] , [19] , [65] ( our unpublished data ) . It is therefore possible that IFITM proteins , including their orthologs in different species which differ significantly at the N- and C-termini [11] , adopt distinct topologies in mammalian cells; these sequence and topologic differences may account for , and contribute to , their somewhat distinct phenotypes in suppressing viral membrane fusion and entry into host cells .
293T , 293 , HTX ( a subclone of HT1080 ) , COS7 , 293T/GFP ( stably expressing GFP ) , HAB2 ( expressing IAV HA , kind gift of Judy White , University of Virginia , Charlottesville , VA ) , 293/LH2SN ( stably expressing Hyal2 ) , HTX/LH2SN ( stably expressing Hyal2 ) , and 293/GP-LAPSN ( expressing MLV Gag-Pol and alkaline phosphatase ( AP ) ) cells have all been described previously [27] , [40] , [46] , [66] . COS7 cells expressing human Hyal2 were generated by transduction with PT67/LH2SN retroviral vector encoding human Hyal2 [40] . 293 , HTX , 293/LH2SN , COS7/LH2SN , HTX/LH2SN cells stably expressing IFITM proteins were generated by transduction with pQCXIP ( Clontech , Mountain View , CA ) retroviral vectors encoding IFITM1 , 2 or 3 ( see below ) . K562 cells stably expressing control shRNA or shRNA targeting IFITM1 or 3 mRNA were kind gifts of Michael Farzan and I-Chueh Huang ( Harvard Medical School , Boston , MA ) . All mammalian cells used were grown in Dulbecco's modified Eagle's ( DMEM ) medium with 10% FBS ( Hyclone , Logan , UT ) . The human IFITM1 , 2 and 3 genes , with or without an N-terminal FLAG tag , were amplified by PCR from pRetro-Tet-IFITM constructs [15] . PCR products were digested and ligated into the EcoRI/BamHI restriction sites of pQCXIP vector , resulting in pQCXIP-IFITMs . Retrovirus packaging plasmid encoding the MoMLV Gag-Pol ( pCMV-gag-pol-MLV ) and transfer vector encoding the GFP ( pCMV-GFP-MLV ) were kind gifts of Francois-Loic Cosset ( INSERM U758-ENS , Lyon , France ) . Plasmids encoding JSRV Env with both N- and C-terminal FLAGs , the 10A1 amphotropic MLV Env , the vesicular stomatitis virus G protein ( VSV-G ) and SFV E1/E2 have been described previously [27] , [29] , [67] , [68] . The 10A1 MLV Env construct with the R peptide deleted was created by removing the last 16 amino acid of the R peptide using PCR . Plasmids encoding the IAV HA and NA ( Thailand KAN-1/2004 H5N1 strain ) were kind gifts of Gary Nabel ( NIH , Bethesda , MD ) . The MLV Gag-YFP construct was a kind gift of Walter Mothes ( Yale University , New Heaven , CT ) . The anti-FLAG monoclonal antibody beads ( EZview™-red ) , anti-FLAG antibody , anti-β-actin monoclonal antibody , anti-Tubulin , secondary anti-mouse immunoglobulin G conjugated to FITC , TRITC or HRP , chlorpromazine ( CPZ ) , bafilomycin A1 ( BafA1 ) , and oleic acid ( OA ) were all purchased from Sigma ( St . Louis , MO ) . Anti-IFITM1 , anti-IFITM2 and IFITM3 were purchased from Proteintech Group ( Chicago , IL ) . IFNα-2b , CMAC ( 7-Amino-4-Chloromethylcoumarin ) , calcein-AM , Methyl-β-cyclodextrin ( MβCD ) , CMTMR ( 5- ( and-6 ) - ( ( ( 4-Chloromethyl ) Benzoyl ) Amino ) Tetramethylrhodamine ) , and Lipofectamine 2000 were purchased from Invitrogen ( Carlsbad , CA ) . The Express 35S-Met/Cys protein labeling mix was purchased from Perkin Elmer ( Boston , MA ) . The JSRV SU-human IgG Fc fusion protein and sHyal2 were produced and purified as previously described [27] , [41] . The GFP- and AP-expressing MLV pseudovrions bearing JSRV Env , 10A1 MLV Env , IAV HA/NA and VSV-G were produced as previously described [33] . Target cells were infected with appropriate amounts of virus stock in the presence of 5 µg/ml Polybrene ( Sigma ) , and assessed for GFP expression by flow cytometry 48 h after infection or for AP activity by staining of cells 72 h after infection . To test the effect of interferon ( IFN ) on viral entry , 293 cells were treated with 200–1000 units of IFN-α2b or medium alone for 24 h before pseudovirus infection . Typically , an MOI of 0 . 05 to 0 . 2 was used for all infections . To create cell lines stably expressing IFITM1 , 2 or 3 , we produced retroviral pseudotypes by transfecting 293T cells with plasmids encoding IFITMs ( pQCXIP-IFITMs ) , MLV-Gag-Pol ( pCMV-gag-pol-MLV ) and VSV-G ( pMD . G ) using the calcium phosphate method . Supernatants were harvested 48–72 h post-transfection and centrifuged at 3 , 200 g to remove cell debris . Cells were infected with pseudovirions in the presence of 5 µg/ml Polybrene . Twenty-four hour after infection , cells were selected in growth medium containing 1 µg/ml puromycin ( Sigma ) . For production of MLV Gag-YFP pseudovirions bearing JSRV Env , 293/GP-LAPSN cells were co-transfected with plasmids encoding Gag-YFP and JSRV Env by the calcium phosphate method . MLV pseudovirions bearing JSRV Env and Gag-YFP were concentrated by centrifugation at 185 , 000 g on a 2 ml 20% sucrose cushion for 3 h , and were resuspended in phosphate-buffered saline ( PBS ) . Cells were detached with PBS plus 5 mM EDTA , and incubated with different amounts of purified pseudovirions on ice for 3 h . The cells were washed with PBS for 5 times and fixed with 3 . 7% paraformaldehyde before being analyzed by using flow cytometry . Cells were detached by PBS containing 5 mM EDTA and resuspended in PBS plus 2% FBS . To examine the binding of JSRV SU to cells expressing Hyal2 , 5×105 HTX cells were incubated with 10 µg purified JSRV SU-human IgG Fc proteins on ice for 3 h , washed 3 times , and incubated with FITC conjugated anti-human IgG Fc antibody for another 1 h . Cells were then washed , fixed and analyzed by flow cytometry . For detection of the expression of the FLAG-tagged JSRV Env and IFITMs , similar procedures were used except that cells were incubated with an anti-FLAG antibody on ice for 1 h , followed by incubation with FITC conjugated anti-mouse IgG antibody for 1 h before cells were analyzed by flow cytometry . Metabolic labeling was performed as previously described [27] . Briefly , 293T cells were transiently transfected with plasmids encoding JSRV Env and/or IFITMs by the calcium phosphate method . Twenty-four hours post-transfection , cells were starved in DMEM without cysteine and methinonine ( MP Biomedicals , Cost Mesa , CA ) for 30 min and pulse-labeled with a 62 . 5 µCi mixture of cysteine plus methionine ( Perkin Elmer , Waltham , WA ) for 1 h , followed by chase-labeling in complete growth medium . To examine shedding of JSRV SU , 3 h after the chase period , indicated amounts of sHyal2 were added and cells were incubated for another 3 h . Cell lysates and culture media containing the 35S-labeled JSRV Env were harvested and immunoprecipitated with anti-FLAG beads . Samples were then resolved by sodium dodecyl sulfate ( SDS ) -polyacrylamide gel electrophoresis ( PAGE ) and applied to autoradiography . Band intensities were quantified using Quantity One ( Bio-Rad , Hercules , CA ) . Syncytia induction assays were performed as described previously [27] . Briefly , 293/LH2SN or COS7/LH2SN cells , either parental ( mock ) or derivatives stably expressing IFITMs , were seeded in 6-well plates and cotransfected with 2 µg plasmids encoding JSRV Env or 10A1 MLV Env with the R peptide deleted ( 10A1 Env-R− ) , or 0 . 5 µg plasmids encoding IAV HA , plus 0 . 5 µg peGFP-N1 ( Clontech ) using the calcium phosphate method . Twenty-four hours post-transfection , cells were treated with pre-warmed buffer ( pH 6 . 2 , pH 5 . 5 , pH 5 . 0 etc . ) for 1 min and were examined for syncytium formation under a fluorescence microscope . If applicable , cells were incubated with indicated doses of IFN-α2b for 24 h before being treated with a low pH buffer to induce syncytia . For 10A1 Env-R− , IFN-α2b was added upon transfection and was maintained throughout the entire fusion assay . The fusion index was calculated by using f = [1- ( C/N ) ] , where C is the number of cells per phase field after fusion , and N is the total number of nuclei [69] . At least five phase-contrast microscopy fields were used for the analysis , with means and standard deviations calculated . Two cell-cell fusion assays were employed in this study . The first cell-cells fusion assay was used for JSRV Env as previously described [27] . In brief , 293T/GFP cells were transfected with 2 µg plasmids encoding JSRV Env alone or plus IFITM1 by Lipofectamine 2000 . Twenty-four hours after transfection , cells were detached with PBS plus 5 mM EDTA and co-cultured with CMTMR prelabeled effector HTX/LH2SN cell lines , either parental ( mock ) or derivatives expressing IFITM proteins . After co-culture for 1 h , cells were treated with a pH 5 . 0 buffer for 1 min , and recovered in complete growth medium for another 1 h . Cells were analyzed for fusion under a fluorescence microscope or by flow cytometry . For a second cell-cell fusion assay , the viral fusion proteins ( JSRV Env , VSV G , and SFV E1/E2 ) were separately expressed in COS7 cells to generate effector cells . In a few experiments , 293T cells were employed as effector cells to test whether the type of effector cell was of functional consequence . For IAV , the cell line HAB2 stably expressing influenza virus HA was used as the effector . 293/LH2SN cells stably expressing IFITM proteins , as described above , were the target cells . Effector cells were loaded with the fluorescent dye calcein-AM ( Invitrogen ) and targets were labeled by the dye CMAC ( Invitrogen ) . For fusion experiments , cells were allowed to bind for 30 min at room temperature and pH was then lowered , through exchange of aqueous solutions , for 10 min , to 5 . 0 for JSRV Env , 4 . 8 for IAV HA , 5 . 7 for VSV G , and 5 . 4 for SFV E1/E2 . The pH was then reneutralized to 7 . 2 , and 30 min later fusion between pairs of effector and target cells was scored by the transfer of both aqueous dyes , as observed by fluorescence microscopy . Experiments were performed as previously described [33] , [34] . Briefly , HTX or HTX/IFITM1 cells were pretreated with 20 nM BafA1 for 2 h and spininoculated with JSRV or IAV pseudovirions at 4°C for 1 h . Following three washes with cold PBS , the virion-cell complexes were either directly exposed to a pH 5 . 0 solution ( for IAV ) 5–10 min or were preincubated at 37°C for 1 h in the presence of 20 nM BafA1 ( for JSRV ) and then incubated with a pH 5 . 0 for 5–10 min . In both cases , the total period of infection in the presence of 20 nM BafA1 was 4 h . Noninternalized virus was inactivated using citrate buffer ( pH 3 . 0 ) after the infection period , and viral infectivity was determined by counting AP-positive foci 72 h after the initiation of infection . A pH 7 . 5-PBS buffer and 0 . 01% DMSO served as controls for the pH 5 . 0 buffer and 20 nm BafA1 , respectively . After binding labeled effector and target cells on cover slips within culture dishes at room temperature for 30 min , the dishes were placed on ice , bringing the solutions bathing the cells to 4°C . The pH was lowered ( ∼pH 5 . 0 ) for 10 min before reneutralizing at 4°C to pH 7 . 2 . At this point , the cells were in a “cold-arrested” stage ( CAS ) . After 3 min , one of two operations was performed . In the first , 0 . 5 mM CPZ was added ( through exchange of solutions at 4°C ) and 1 min later the CPZ was washed out with a solution containing delipidated-BSA . Each dish was kept on ice , and each cover slip was removed to monitor aqueous dye transfer by fluorescence microscopy . In the second manipulation , cells at CAS were placed in a 37°C incubator for 30 min and dye transfer was then monitored . In order to make spontaneous monolayer curvatures more negative , 285 µM oleic acid ( OA , Sigma ) was incorporated into cell membranes either prior to or subsequent to creating CAS . For prior incorporations , effector and target cells were incubated together at room temperature for 20 min and OA was then added; 15 min later , the solutions were lowered to 4°C and CAS was created . OA was removed at 4°C by washing the cells with a solution containing delipidated-BSA . CPZ was then added or temperature was raised to 37°C . To incorporate OA subsequent to CAS , OA was added to cells at CAS , and CPZ was added or the temperature was raised without removing OA . The membrane probe Laurdan ( 6-dodecanoyl-2-dimethylamino naphthalene , Invitrogen ) was dissolved in DMSO ( dimethylsulfoxide ) to make a stock concentration of 1 . 8 mM . 293/LH2SN cells expressing IFITM1 , 2 or 3 , or none ( Mock ) were incubated with 1 . 8 µM Laurdan for 40 min at 37°C . To deplete cholesterol , a 50 mM stock solution of methyl-β-cyclodextrin ( MβCD , Sigma-Aldrich ) was prepared by dissolving in nanopure water . Cells were incubated with 10 mM MβCD for 1 h at 37°C . All cells were rinsed with PBS once before being processed for imaging . In order to quantitatively assess the membrane order , a ratiometric method known as generalized polarization ( GP ) was developed [52] . The GP function or value characterizing the spectral properties of Laurdan is calculated through the following expression: ( 1 ) where Iblue and Igreen are the respective intensities conventionally centered at 440 nm ( the emission maximum for more ordered lipid bilayer ) and centered at 490 nm ( the emission maximum for less ordered lipids ) . GP and FLIM data were acquired with a Zeiss LSM710 META Laser scanning microscope , coupled to a 2-Photon Ti:Sapphire laser ( Spectra-Physics Mai Tai , Newport Beach , CA ) producing 80 fs pulses at a repetition of 80 MHz and a ISS A320 FastFLIMBox for the lifetime data . A 40× water immersion objective 1 . 2 N . A . ( Zeiss , Oberkochen , Germany ) was used for all experiments . The excitation wavelength was set at 780 nm . A SP 760 nm dichroic filter was used to separate the fluorescence signal from the laser light . For FLIM data , the fluorescence signal was directed through a 495 LP filter and the signal was split between two photo-multiplier detectors ( H7422P-40 , Hamamatsu , Japan ) , with the following bandwidth filters in front of each: blue channel 460/40 and green 540/25 , respectively . For image acquisition , the pixel frame size was set to 256×256 and the pixel dwell time was 25 . 61 µs/pixel . The average laser power at the sample was maintained at the mW level . For GP data the fluorescence signal was acquired from 416 nm to 474 nm for the blue channel and from 474 nm to 532 nm for the green channel , using the spectral detector of the LSM 710 by joining 6 channels of the detector each having a bandwidth of 9 . 7 nm . For image acquisition , the pixel frame size was set to 256×256 and the pixel dwell time was 177 . 32 µs/pixel . The average laser power at the sample was maintained at the mW level . SimFCS software developed at the Laboratory for Fluorescence Dynamics ( www . lfd . uci . edu ) was used to acquire FLIM data and to process FLIM and GP data . Calibration of the system and phasor plot for FLIM data was performed by measuring fluorescein ( pH 11 ) , which has a known single exponential lifetime of ∼4 . 04 ns . Calibration for GP data was performed with Laurdan in DMSO which has a reference GP value equal to zero . GP histograms of the images were calculated using SimFCS , and the histograms fitting was performed by using a MatLab routine . GP histograms were characterized by the presence of two Gaussian distributions . After highlighting the corresponding pixels in the images , it was possible to couple the first Gaussian distribution ( centered at lower GP values ) with the intracellular membranes of the cell , and the second distribution ( centered at higher GP values ) with the plasma membranes [70] . The two histogram distributions were used to calculate the averaged GP values of intracellular and plasma membranes . Paired student t test was used for statistical analysis unless otherwise noted . Typically data from three to eight independent experiments were used for the analysis . | Many pathogenic viruses contain an envelope that must fuse with the cell membrane in order to gain entry and initiate infection . This process is mediated by one or more glycoproteins present on the surface of the virions , known as viral fusion proteins . Recently , a family of interferon-inducible transmembrane ( IFITM ) protein has been shown to block viral infection , including those of highly pathogenic viruses . Here we provide evidence that these IFITM proteins potently suppress membrane fusion induced by representatives of all three classes of viral fusion proteins . Interestingly , we found that the block is not at the steps of receptor binding or low pH that triggers conformational changes of viral fusion proteins required for membrane fusion . Rather , we discovered that the creation of hemifusion , an intermediate in which the outer membranes of the two lipid bilayers have merged but the inner membranes still remain intact is blocked by IFITM proteins . We further demonstrated that overexpression of IFITM proteins rigidify the cell membrane , thereby reducing membrane fluidity and fusion potential . Our study provides novel insight into the understanding of how IFITM proteins restrict viral entry and infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"virology",
"biology",
"microbiology"
] | 2013 | IFITM Proteins Restrict Viral Membrane Hemifusion |
The interaction between signaling pathways is a central question in the study of organogenesis . Using the developing murine tongue as a model , we uncovered unknown relationships between Sonic hedgehog ( SHH ) and retinoic acid ( RA ) signaling . Genetic loss of SHH signaling leads to enhanced RA activity subsequent to loss of SHH-dependent expression of Cyp26a1 and Cyp26c1 . This causes a cell identity switch , prompting the epithelium of the tongue to form heterotopic minor salivary glands and to overproduce oversized taste buds . At developmental stages during which Wnt10b expression normally ceases and Shh becomes confined to taste bud cells , loss of SHH inputs causes the lingual epithelium to undergo an ectopic and anachronic expression of Shh and Wnt10b in the basal layer , specifying de novo taste placode induction . Surprisingly , in the absence of SHH signaling , lingual epithelial cells adopted a Merkel cell fate , but this was not caused by enhanced RA signaling . We show that RA promotes , whereas SHH , acting strictly within the lingual epithelium , inhibits taste placode and lingual gland formation by thwarting RA activity . These findings reveal key functions for SHH and RA in cell fate specification in the lingual epithelium and aid in deciphering the molecular mechanisms that assign cell identity .
The tongue is a muscular organ that plays critical roles in mastication , speech and taste , and the regulatory mechanisms that specify the diverse cell types and structures of the tongue are of great interest . The mature tongue ( Fig 1A and 1B ) is covered by a mucosa made of lingual epithelium ( LE ) and lingual mesenchyme ( LM ) . The dorsal LE of the anterior 2/3 of the tongue ( oral tongue ) is a stratified , squamous epithelium primarily comprised of mechanosensory filiform papillae . The gustatory units of the LE , the taste buds ( TBs ) , develop in three different types of papillae: fungiform papillae , foliate papillae , and circumvallate papillae . Fungiform papillae house a single TB and are distributed between filiform papillae over the dorsal surface of the oral tongue , whereas foliate papillae house several TBs and develop posteriorly at the lateral edges of the oral tongue . In rodents , a single circumvallate papilla harbouring numerous TBs forms at the junction between the oral tongue and the posterior 1/3 of the tongue , known as pharyngeal tongue [1] . The LE also produces minor salivary glands of a mixed sero-mucous type in the pharyngeal tongue , as well as a purely serous type , von Ebner’s glands , which arise from the epithelium of the circumvallate papilla [2 , 3] . The distribution of mature taste papillae depends on coordinated signaling events during early stages of tongue development [4] . From embryonic day ( E ) 10 . 5-E11 . 5 , Shh is expressed throughout the LE and signals both within the LE and to the LM [5–8] . At E12 . 5 , fungiform papilla development is heralded by fungiform placodes ( FPs ) , localized thickenings of the LE made of post-mitotic TB precursors expressing Shh , Wnt10b , and Bmps . At E12 . 5 to E14 , the interplacodal epithelium is devoid of Shh expression , which is restricted to taste placodes . Canonical signaling through these pathways is critical for regulating the size , number , and spacing of fungiform placodes [4] . As the tongue grows , new FPs develop until E14 . 5 , at which time fungiform papilla morphogenesis begins and differentiating TBs become morphologically and molecularly visible at the tip of fungiform papillae . After E14 . 5 and during early postnatal life , only developing TB cells within gustatory papillae produce SHH [4] . Several components of the Hedgehog signaling pathway play key roles to ensure properly calibrated spatio-temporal Hedgehog inputs [8–11] . Upon ligand binding to the Hedgehog receptor Patched ( PTCH1 ) , Smoothened ( SMO ) , an obligatory Hedgehog transducer , translocates to the primary cilium , unleashing a signaling cascade culminating in transcriptional regulation of Hedgehog target genes by GLI proteins [10] . Gli1 and Ptch1 are themselves direct targets of Hedgehog signaling , and hence their expression enables identification of cells responding to Hedgehog signals . Loss of SMO function abrogates all Hedgehog signaling even in the presence of copious amounts of ligands [9] . Studies of tongue organ cultures have shown that pharmacological inhibition or activation of SHH signaling causes increased or decreased size and number of FPs , respectively , indicating that SHH inhibits FP formation [12–15] . However , whether and how SHH performs such a function in vivo is unknown [6 , 16] . Moreover , it remains unclear which lingual tissue SHH acts upon , as in vitro manipulations may disrupt the LE and LM which are both SHH-responsive [5 , 6 , 8] . Furthermore , how different lingual structures such as TBs and glands are specified within a seemingly homogeneous epithelial sheet remains a critical open question . That SHH signaling is active in both the oral and pharyngeal tongue during the earliest steps of tongue formation [6 , 8] suggests this molecule plays a yet unknown role in cell fate specification within the LE . To address these issues , we utilized mouse genetics to modulate SHH signaling in the LE , LM , or both . We found that SHH plays a dual role during tongue development by acting in the LM to control growth and morphogenesis and in the LE to control patterning and cell fate determination . We revealed an unknown , highly sensitive balance between SHH and retinoic acid ( RA ) signaling that regulates the specification of taste placodes and glands . Upon loss of SHH signaling , the augmented RA inputs direct the LE to ( 1 ) form minor salivary glands in the oral tongue , ( 2 ) overproduce oversized FPs and TBs , and ( 3 ) sustainably generate taste placodes . We found that both FP specification and antero-posterior patterning of the LE are regulated by antagonizing activities of SHH and RA . We show that SHH antagonizes RA signaling by maintaining expression of Cyp26a1 and Cyp26c1 . Unexpectedly , we found that in the absence of SHH signaling , the LE aberrantly produced Merkel cells . Finally , we identified two novel TB markers , Homer1 and RALDH1 .
To decipher the function of SHH in patterning of and cell type specification in the LE , we first irreversibly ablated SHH signaling in epithelial cells that express Shh and their progeny by ShhGFPCRE-mediated removal of Smo in mutants carrying Smo floxed ( f ) alleles ( ShhGFPCRE/Smof/f ) . The ShhGFPCRE/Smof/f mutants die shortly after birth due to unknown causes . Littermates not carrying the ShhGFPCRE knockin and/or the Smo floxed alleles and heterozygotes , which are viable and phenotypically normal , were used as controls . Tongues from embryos carrying the R26R transgene ( ShhGFPCRE/R26R ) showed robust CRE activity in the LE ( S1A–S1D Fig ) . Probes targeting Ptch1 and Gli1 , which are readouts of Hedgehog activity [6 , 9 , 17–19] , confirmed the absence of Ptch1 and Gli1 transcripts in the ShhGFPCRE/Smof/f LE , indicating efficient Smo ablation in this tissue . By contrast , and as expected , the LM retained endogenous hedgehog signaling ( S1E–S1H’ Fig ) . Consistent with these findings , RT-qPCR revealed downregulation of Ptch1 in ShhGFPCRE/Smof/f tongues ( S1I Fig ) . To assess the impact of epithelial SMO ablation on tongue development , we compared control and ShhGFPCRE/Smof/f tongues at E17 . 5 , E18 . 5 , and shortly after birth ( postnatal day P0 ) . Alcian blue van Gieson staining and immunohistochemistry for Keratin 8 ( K8 ) revealed striking defects within the LE and its derivatives in the ShhGFPCRE/Smof/f cohort with 100% penetrance . Specifically , the oral tongue , a territory normally bereft of glands , exhibited heterotopic serous and sero-mucous glands ( Fig 1C–1F ) . Furthermore , the LE was hyperplastic , lacked filiform papillae ( Fig 1C’–1D’ ) , and displayed overgrown von Ebner’s glands transformed into a sero-mucous type ( Fig 1G and 1H; see also S6 Fig ) , as well as epithelial foci virtually devoid of K8 , Keratin 6 , and Keratin 15 immunostaining ( Fig 1I–1N ) . 7 . 6% of mutants also showed adhesions between the LE and epithelia of the floor of the oral cavity ( Fig 1F ) . Taken together , these findings indicate that an altered antero-posterior patterning and differentiation program in the SMO-deficient LE is translated into development of glands in the oral tongue , abnormal development of von Ebner’s glands , and formation of epithelial regions exhibiting a molecular signature different from that of the normal LE . Closer examination revealed abnormal formation of K8-positive ( + ) ( Fig 2A and 2B ) and SOX2+ ( Fig 2C and 2D ) solitary cells in the basal layer of the LE in all ShhGFPCRE/Smof/f mutants examined at E17 . 5 to P0 . We postulated that these cells could be stray TB cells or ectopic epithelial tactile cells called Merkel cells , as both cell types express high levels of Sox2 and Krt8 [20–22] . Merkel cells do not express Shh ( Fig 2P ) [21] , whereas all TB cells express SHH at least up to the perinatal period ( Fig 2E and 2F ) [4] . K8/SHH immunostaining revealed absence of SHH+ solitary cells in the basal layer of the ShhGFPCRE/Smof/f LE and showed that the K8+ solitary cells were SHH-negative ( – ) ( Fig 2F; S1 Table ) , suggesting that these were not TB cells . To further assess this notion , we immunostained sections for Homer1 , a scaffolding protein found in skeletal muscle and brain [23] which we serendipitously identified as a reliable TB marker in the course of this study; Homer1 expression in the LE of control and ShhGFPCRE/Smof/f tongues is not only restricted to the TBs , but also readily detectable in all TB cell types ( Fig 2G–2J; S1 Table ) , differing from other marker proteins that are only expressed in subsets of TB cells [1 , 4] . Neither orthotopic K8+ Merkel cells in whisker pads ( Fig 2Q ) [21] nor the K8+ solitary cells in the basal layer of the ShhGFPCRE/Smof/f LE ( Fig 2K; S1 Table ) were Homer1+ . Thus , two TB markers , SHH and Homer1 , excluded a TB cell type and supported a Merkel cell identity of the solitary cells in the mutant LE . To confirm that these cells were Merkel cells , we examined expression of Rab3c , a protein known to be enriched in Merkel cell cytoplasmic vesicles ( Fig 2R ) [21] . We found that while the basal layer of the LE in controls as well as K8+ TBs in controls and ShhGFPCRE/Smof/f mutants were Rab3c ( – ) , the ectopic K8+ solitary cells within the ShhGFPCRE/Smof/f LE were Rab3c+ ( Fig 2L and 2M; S1 Table ) . These data show that the K8+ , SOX2+ , and Rab3c+ cells were indeed ectopic Merkel cells , indicating that the LE , a tissue normally devoid of Merkel cells in humans and mice , underwent Merkel cell metaplasia upon loss of SMO . In mice , Merkel cells develop in whisker pads and touch domes of hairy and glabrous skin [24] and become innervated postnatally [24 , 25] . Remarkably , we found that the majority of the K8+ ectopic Merkel cells in the ShhGFPCRE/Smof/f LE were innervated by NF-200+ neurites ( Fig 2N; S1 Table ) , suggesting that they might be functional in this foreign environment . TBs in fungiform papillae are innervated by P2X2+ gustatory axons [26] , however , none of the K8+ basally located solitary cells in the ShhGFPCRE/Smof/f LE were innervated by P2X2+ axons ( Fig 2O; S1 Table ) , further supporting our conclusion that these cells are ectopic Merkel cells and not TB cells . Unlike in the glabrous skin where Merkel cells develop in the absence of SHH inputs , Merkel cell specification in the hairy skin is dependent on SHH signaling [27] . Here we showed that in the absence of SMO , a subset of cells in the LE are diverted to a Merkel cell fate . Taken together , these observations suggest a context-dependent requirement for SHH signaling for Merkel cell specification . Polycomb genes have been shown to repress Sox2 which encodes a crucial factor for Merkel cell specification in the epidermis [28] . However , the LE expresses Sox2 mRNA and protein from early stages of tongue development onwards [22 , 29] , and the ectopic Merkel cells in the ShhGFPCRE/Smof/f LE express SOX2 protein . These observations likely indicate that Polycomb genes are not effectors of the repression of Merkel cell formation in the LE . TBs and Merkel cells both degenerate after denervation [30 , 31] and share common markers such as Krt8 , K20 , Sox2 , Advillin , Satb2 , Cpe , Claudins 6/7 , Mash1 , Hes6 , and Snap25 [1 , 21 , 22 , 32–34] . This raises the possibility that a subset of TB progenitors converted into Merkel cells in the SMO-deficient LE . In vitro modulation of SHH signaling in the developing rodent tongue impinges upon FP patterning [12–15] , however , it is currently unclear whether this is due to altered SHH activities in the LE , LM , or both . Furthermore , whether SHH has a role in FP patterning in vivo is still unknown . To address these points , we analyzed the effects of loss of epithelial SMO on FP development by assessing the expression patterns of Shh mRNA and protein , an established taste placode marker . At E12 . 5 , the LE and LM showed robust and weak SHH immunostaining , respectively , in both control and ShhGFPCRE/Smof/f tongues . The staining of the LM and interplacodal epithelium was the result of protein diffusion from taste placodes which produce SHH ( Fig 3A and 3B ) . At E14 , instead of being restricted to FPs as in control tongues , SHH staining was expanded in the LE of the ShhGFPCRE/Smof/f tongues ( Fig 3C and 3D ) . At E14 . 5 FP induction normally ceases , and Shh transcripts become confined to developing TBs [4] . While at E14 . 5 and E15 . 5 the control tongues showed this expected Shh expression pattern , the ShhGFPCRE/Smof/f tongues displayed a remarkable expansion of Shh expression ( Fig 3E–3H ) and overproduced enlarged Shh-expressing spots ( FPs and/or TBs ) , some of which developed ectopically ( Fig 3G–3L ) . Quantification in E15 . 5 mutant and control tongues revealed increased numbers of Shh+ spots in the mutants ( 147 ± 8 in mutants vs 101 ± 7 in controls; mean ± SD; P = 0 . 000 ) . Thus , loss of SMO in the LE not only caused Merkel cell and glandular metaplasiae , but also engendered mispatterning of FPs as a result of their overproduction . Despite displaying FP patterning defects , the ShhGFPCRE/Smof/f tongues developed K8+ TBs that were innervated by P2X2 neurites ( Fig 3M and 3N; S1 Table ) . Having uncovered a stringent requirement for epithelial SHH signaling for patterning and differentiation of the LE , we sought to decipher the role of SHH inputs in the LE and LM by assessing the impact of loss of SHH production . We generated ShhCreERT2/Shhf mutant mice carrying both a Shh floxed allele and the tamoxifen ( TAM ) -inducible ShhCreERT2 knockin allele . Without exposure to TAM , the ShhCreERT2/Shhf mice are viable and phenotypically normal . Ablation of Shh in the mutants was induced at E10 . 5 . Immunostaining and in situ hybridization confirmed that TAM exposure led to a dramatic decrease of SHH protein in the LE and severe downregulation of Gli1 expression in the LE and LM ( Fig 4A–4D ) of the ShhCreERT2/Shhf tongues , indicating loss of SHH signaling . This was further confirmed by RT-qPCR for Ptch1 ( Fig 4O ) . Upon dissection , we found that relative to their control littermates , the E10 . 5 TAM-induced ShhCreERT2/Shhf mutants developed abnormally small tongues ( microglossia ) with a bifid tip as well as epithelial foci of squamous hyperplasia ( Fig 4A–4F ) . Of note , bifid tip of the tongue has been described in patients with Ellis-van Creveld syndrome , which is caused by mutations of the EVC and EVC2 genes that lead to decreased Hedgehog signaling [35–38] . In situ hybridization with a Shh riboprobe targeting both deleted ( exon2 ) and non-deleted ( exon1 ) Shh-coding sequences [39] revealed enlarged Shh+ spots , likely TBs or FPs , some of which were ectopically located along the midline of the E10 . 5 TAM-induced ShhCreERT2/Shhf tongues ( Fig 4E and 4F ) . Furthermore , these tongues formed K8+ TBs ( Fig 4H and 4L ) innervated by P2X2+ neurites ( Fig 4M and 4N; S2 Table ) . However , unlike the ShhGFPCRE/Smof/f tongues , the E10 . 5 TAM-induced ShhCreERT2/Shhf tongues were exempt from glandular metaplasia , i . e . no heterotopic glands developed in the oral tongue ( Fig 4L ) , and displayed only a few scattered ectopic Merkel cells ( Fig 4I and 4J ) . Thus , SHH is required for growth and morphogenesis of both the LE and LM , but it is dispensable for FP induction . Remarkably , in ShhCreERT2/Shhf mutants first exposed to TAM at E11 . 5 , the tongues grew normally ( Fig 5F–5I ) despite severely reduced SHH signaling as revealed by in situ hybridization with probes targeting Shh exon2 ( Fig 5A and 5B ) [40] and Gli1 ( Fig 5C and 5D ) as well as by Ptch1 RT-qPCR ( Fig 5E ) . In situ hybridization for Shh-exon1/exon2 ( Fig 5F and 5G ) and immunostaining for K8 and SOX2 ( Fig 5H–5N ) revealed that the E11 . 5 TAM-induced ShhCreERT2/Shhf tongues overproduced fungiform papillae harboring oversized TBs . Quantification in E15 . 5 control and ShhCreERT2/Shhf mutant tongues showed dramatically increased numbers of Shh+ spots ( FPs and/or TBs ) in the mutants ( 150 ± 6 in mutants vs 112 ± 4 in controls; mean ± SD; P = 0 . 000 ) . Furthermore , quantification of K8+ cell clusters ( TBs or parts of TBs ) within fungiform papillae at E18 . 5 revealed a higher number of K8+ cell clusters in ShhCreERT2/Shhf mutants when compared to controls ( 3 . 50 ± 0 . 50 K8+ clusters/section in mutants vs 1 . 43 ± 0 . 23 K8+ clusters/section in controls; mean ± SD; P<0 . 02 ) . However , the mutant tongues assessed by K8 immuostaining were free from glandular and Merkel cell metaplasiae ( Fig 5I and 5J; S2 Table ) . Similar to TBs in the E10 . 5-induced ShhCreERT2/Shhf tongues , TBs within fungiform papillae in the E11 . 5 TAM-induced ShhCreERT2/Shhf tongues were innervated by P2X2+ axons ( Fig 5M and 5N; S2 Table ) . Thus , unlike in the central nervous system where SHH can act as an axonal chemoattractant [41] , SHH is not required for local guidance of gustatory neurites towards fungiform papillae . To further dissect the role of SHH in lingual cell fate specification , we analysed Wnt1-CRE/Smof/f mutants lacking SHH signaling specifically in the LM . As shown previously [6] , the Wnt1-CRE/Smof/f tongues were abnormally small and cleft , and we found that they developed TBs and were free from glandular metaplasia ( S2A–S2D Fig ) . The overlapping tongue anomalies in the E10 . 5 TAM-induced ShhCreERT2/Shhf mutants and the Wnt1-CRE/Smof/f mutants demonstrate that SHH activity in the LM is necessary for tongue growth and morphogenesis but not for TB induction . The lack of glandular metaplasia in the oral tongue of the Wnt1-CRE/Smof/f mutants as well as in the E10 . 5 and E11 . 5 TAM-induced ShhCreERT2/Shhf mutants , as compared to the ShhGFPCRE/Smof/f mutants , suggests the requirement for SHH-dependent mesenchymal inputs for ectopic gland formation . This situation is akin to that of the skin where induction of glandular metaplasia was found to require SHH signaling in the mesenchyme [19] . We also found that proper tongue development necessitates SHH inputs during a restricted period since the ShhCreERT2/Shhf mutants and the ShhCreERT2/Smof/f mutants developed normal tongues following an initial TAM exposure at E12 . 5 ( S2E–S2H’ Fig ) . That loss of SMO in the interplacodal epithelium of K14-CRE/Smof/f mutant tongues caused no tongue defects ( S2I–S2J’ Fig ) further demonstrates a temporal requirement for SHH signaling in tongue development . K14-CRE activity in the LE is late onset ( ≈ E13 . 5 ) ( S2K–S2N Fig ) [42] , and taste placodes as well as TBs do not express K14 [29] . Together , these findings point to a restricted period before E12 . 5 during which loss of SHH signaling induces tongue defects and support previous findings in rat tongue organ cultures [15] . The active phase of FP patterning during which FPs are induced normally occurs between E12 . 5 and E14 , and the FPs subsequently form TBs . Strikingly , from E15 . 5 onwards , unlike in controls , the ShhGFPCRE/Smof/f tongues consistently displayed SHH/Shh-expressing placode-like structures underlain by a Ptch1+ mesenchyme , suggesting de novo formation of FPs well after their induction normally has ceased ( Fig 6A–6H ) . Basally located SHH+ cell clusters were also detected in the E11 . 5 TAM-induced ShhCreERT2/Shhf mutant tongues . They displayed weaker SHH immunoreactivity than TBs in controls , likely subsequent to Shh gene expression ablation shortly after they express Shh ( Fig 6I–6J’ ) . We next used a probe targeting Wnt10b which , like Shh , marks taste placodes [4] . In contrast to control tongues , which were devoid of Wnt10b transcripts after E14 . 5 , the ShhGFPCRE/Smof/f as well as the E10 . 5 and E11 . 5 TAM-induced ShhCreERT2/Shhf tongues abnormally exhibited Wnt10b expression separated by Wnt10b ( – ) regions within the basal layer of the LE ( Fig 6K–6P ) . That the basally located placodes expressing Shh mRNA and protein ( Shh/SHH ) and Wnt10b were overlain by Shh/SHH ( – ) and Wnt10b ( – ) suprabasal epithelial cells ( Fig 6A–6P ) supports de novo induction of FPs . These data show sustained formation of taste placodes upon loss of SHH signaling , at least up to birth , and are reminiscent of the finding that K14-CRE/Smof/f mutants undergo de novo hair follicle formation [19] . Since de novo taste placode induction occurred in the ShhGFPCRE/Smof/f and TAM-induced ShhCreERT2/Shhf mutant tongues , it could be concluded that loss of SHH signaling in the LE is the main cause of this phenomenon . Our work here utilized a genetic approach to reveal that upon loss of SHH signaling , the mutant tongues transcended the time when FP induction normally ceases ( E14 . 5 ) and formed de novo taste placodes . This phenomenon was not observed in previous studies using rodent tongue organ cultures , since in those studies the assessment of the impact of loss of SHH signaling on taste placode formation was limited to the active phase of FP patterning [12–15] . Furthermore , our findings that in vivo SHH inhibits FP induction during the active period of FP patterning not only concur with prior work in vitro [12–15] , but they also demonstrate that SHH signaling is required to prevent sustained taste placode induction and that SHH operates strictly within the LE to fulfill these functions , without inputs from the LM . Similarly , genetic studies have demonstrated that Wnt/β-catenin signaling within the LE is crucial for FP induction [1] . Loss of Wnt/β-catenin activity downregulates Shh in FPs and Ptch1 in FPs and their associated mesenchyme , whereas activation of Wnt/β-catenin signaling engenders expansion of Shh expression [14 , 43] . Moreover , findings from tongue organ cultures suggest that SHH inhibits Wnt/β-catenin activation , whereas Wnt/β-catenin signaling induces its own inhibitor , i . e . SHH [14] . However , how SHH impinges upon Wnt/β-catenin activity is unknown . Our findings of ectopic and anachronic induction of Wnt10b expression in mutants lacking SHH inputs suggest that SHH antagonizes Wnt/β-catenin signaling at least in part through inhibition of Wnt10b expression . Further supporting this notion is the finding that SHH inhibits Wnt10b expression in cultured teeth and whisker pads [44] . The ShhGFPCRE/Smof/f tongues not only underwent glandular metaplasia , but they also formed von Ebner’s glands displaying mucous metaplasia . We postulated that the dramatic tongue anomalies caused by disrupted SHH signaling were a result of enhanced retinoic acid ( RA ) signaling . Components of the RA pathway are expressed during tongue development [45–50] , and it is well-established that excess RA or vitamin A ( a RA precursor ) induces glandular/mucous metaplasia of various epithelia [51–56] . Strikingly , rodent embryos overexposed to RA/vitamin A exhibit tongue anomalies phenocopying those occurring in the Shhn/n [57] and the various mutants described in this work , including aglossia , microglossia , bifid tongue , and adhesion of the LE to epithelia of the oral mucosa [58–61] . To test our hypothesis , we first assessed the expression of RARb and RARg , well-established transcriptional targets of RA signaling that encode the nuclear retinoic acid receptors RARβ and RARγ , respectively [62] . We found that relative to controls , the LE of the ShhGFPCRE/Smof/f mutants ( Figs 7A–7D’ and S3A–S3L ) as well as that of the E10 . 5 ( Figs 7E–7F’ and S3M–S3P ) and E11 . 5 ( Fig 7G–7H’ ) TAM-induced ShhCreERT2/Shhf mutants displayed stronger RARb and/or RARg hybridization signals . Furthermore , RT-qPCR revealed significant upregulation of RARb and RARg expression in the mutants ( Fig 7I–7K ) . Consistent with these findings , anti-RARγ staining revealed that the ShhGFPCRE/Smof/f tongues exhibited stronger RARγ immunolabelling in the LE than that of controls from E11 . 5 to E14 . 5 ( S4A–S4G Fig ) except in epithelial foci , which could be developing squamous hyperplastic sites ( S4G Fig ) . Noticeably , RARγ was readily detectable in TBs which expressed RARg transcripts ( S4E and S4H–S4K Fig ) , and the lingual glands were also RARγ+ ( S4L–S4N Fig ) . These data show that the LE and its derivatives produce RARγ protein and that RA signaling is enhanced in the LE upon loss of SHH inputs . Transcripts encoding the RA synthesizing enzymes RALDH1-3 have been shown to be expressed in the LE at E14 . 5 and E16 . 5 [49] . To assess the distribution patterns of RALDH1-3 during tongue development and to determine whether they are altered in the ShhGFPCRE/Smof/f tongues , we immunostained sections from controls and mutants from E11 . 5 onwards . We found that RALDH1-3 were expressed in the LE from E11 . 5 onwards as well as in the orthotopic and heterotopic lingual glands ( S5 and S6 Figs ) . We also found that FPs were RALDH3+ ( S5Q and S5R Fig ) , whereas TBs were RALDH1+ ( S5S–S5T and S5Y–S5Z Fig ) . Thus , the LE and its derivatives in controls and mutants are exposed to RALDH-derived RA . However , besides abnormal enhancement of RALDH3 and RALDH2 staining in the LE of the mutants at E11 . 5 ( S5F Fig ) and E16 . 5 ( S5V Fig ) , respectively , there were no major alterations in RALDH1-3 distribution patterns in the mutants . RA signaling is tightly regulated by activities of RALDH1-3 and the RA catabolic enzymes belonging to the CYP26 family [63 , 64] . As the ShhGFPCRE/Smof/f mutant tongues had no major alterations in RALDH distribution , we reasoned that the increased RA signaling could be caused by decreased production of CYP26A1 and CYP26C1 which are expressed in the LE [65–67] . We found that Cyp26a1/c1 transcripts were dramatically reduced in developing tongues dissected from ShhGFPCRE/Smof/f mutants as well as from E10 . 5 and E11 . 5 TAM-induced ShhCreERT2/Shhf mutants compared to controls ( Fig 8 and Fig 9 ) . However , some Cyp26a1/c1-expressing epithelial foci persisted in the mutants ( Fig 8 and Fig 9 ) . These are likely developing focal squamous hyperplasiae as shown in tissue sections ( arrows in Fig 8 ) , which consistently develop in the ShhGFPCRE/Smof/f and TAM-induced ShhCreERT2/Shh f mutants . Strikingly , we found that during the active phase of FP patterning , Cyp26a1/c1 were excluded from FPs ( Fig 8 and Fig 9 ) . When FP induction is complete at E14 . 5 Cyp26a1/c1 levels started to diminish ( Fig 9I , 9K and 9M ) , and after E16 . 5 , Cyp26a1/c1 were no longer expressed in the LE [65 , 66] . Furthermore , Cyp26a1/c1 expression was excluded from the LE at the level of the developing circumvallate papilla and from the pharyngeal tongue ( Fig 8 and Fig 9 ) . Cyp26b1 , another CYP26 family member , is expressed exclusively in the LM [65] . We found that Cyp26b1 levels ( Fig 9U ) in the mutant tongues were either essentially unaltered ( ShhGFPCRE/Smof/f and E10 . 5 TAM-induced ShhCreERT2/Shhf ) or upregulated ( E11 . 5 TAM-induced ShhCreERT2/Shhf ) , confirming enhanced RARb and RARg levels in the LE of the mutants . Unlike in both the ShhGFPCRE/Smof/f and in the E10 . 5 and E11 . 5 TAM-induced ShhCreERT2/Shhf mutant tongues , Cyp26a1/c1 levels were unaltered in E12 . 5 TAM-induced ShhCreERT2/Shhf mutant tongues ( Fig 9O–9R ) . Thus , a critical temporal window of SHH activity is necessary for expression of Cyp26a1/c1 in the LE . These data , together with our findings indicating a temporal window during which loss of SHH signaling causes FP mispatterning , suggest that early loss ( before E12 . 5 ) of SHH signals is necessary for downregulation of Cyp26a1/c1 expression in the LE during FP patterning ( E12 . 5-E14 ) and for the entailing patterning defects . While our data indicate that Cyp26a1/c1 expression in the LE requires SHH inputs , it is unlikely that the transcripts are direct targets of SHH signaling . At E11 . 5 , despite the presence of significant SHH activity in the tongue primordium [6 , 8] , only weak Cyp26a1/c1 signals were detectable ( Fig 9A ) [67] . Furthermore , treatment of tongue/mandible explants with SAG , a SMO agonist , enhanced Cyp26a1 signals in the LE but failed to induce ectopic Cyp26a1 expression in epithelia of the pharyngeal tongue and mandible ( Fig 9S and 9T ) . These observations suggest that SHH signaling is required for maintenance and/or reinforcement of Cyp26a1/c1 expression in the LE but not for their initial induction . The virtual absence of Cyp26a1/c1 transcripts in taste placodes and TBs amid high expression levels in the interplacodal LE is a conundrum in view of the data indicating the requirement of SHH signaling for Cyp26a1/c1 expression . However , taste placodes and TBs , similar to other SHH-producing signaling centers including the notochord , neural floor plate and tooth enamel knots [18 , 68 , 69] , did not exhibit higher levels of Ptch1 and Gli1 than the nearby Shh-negative-SHH-responsive tissues . Furthermore , Gli1-lacZ mice revealed that Shh-expressing cells in taste placodes and papillae are less SHH-responsive than the neighbouring epithelial and mesenchymal cells [70] . Moreover , taste placodes , the notochord , and tooth enamel knots are induced after loss of Shh or Smo [our present data; 9 , 18 , 40 , 71] . In light of these observations , we propose that SHH signaling is attenuated in taste placodes and TBs and that this , combined with inputs from yet to be identified factor ( s ) , leads to absence of Cyp26a1/c1 in these structures . That RA signaling is activated upon loss of SHH signaling strongly suggests that the glandular metaplasia , i . e . cell fate switch prompting the LE in the ShhGFPCRE/Smof/f oral tongue to form heterotopic glands , is a direct consequence of hyperactivation of RA signaling . Indeed , it is well-established that excess RA and Vitamin A induces glandular/mucous metaplasia of various epithelia [51–56] . To functionally test this scenario , we cultured E11 . 5 tongues and the surrounding mandibular structures in the presence of retinoids and inhibitors of RA signaling . We found that treating explants from control embryos with retinoids , including all-trans RA , CD1530 ( RARγ selective agonist ) and CD2314 ( RARβ selective agonist ) induced formation of K8+ heterotopic glands in the oral tongue , mimicking the ShhGFPCRE/Smof/f in vivo phenotype ( S7A–S7E Fig; S3 Table ) . ShhGFPCRE/Smof/f tongue explants treated with vehicle ( DMSO ) recapitulated the in vivo defects , including glandular and Merkel cell metaplasiae as well as formation of K8 ( – ) epithelial foci ( S7F and S7I Fig; S3 Table ) . Importantly , exposure to BMS493 ( pan-RAR antagonist ) and DEAB ( RA synthesis inhibitor ) inhibited heterotopic gland formation in the ShhGFPCRE/Smof/f explants ( S7F–S7K Fig; S3 Table ) . Neither BMS493 nor DEAB inhibited lingual Merkel cell metaplasia in the mutant explants ( S7H and S7J Fig ) . Thus , lingual Merkel cell metaplasia is not caused by enhanced RA signaling . Furthermore , BMS493 and DEAB failed to inhibit development of the K8 ( – ) foci in the mutants ( S7H and S7J Fig ) , consistent with our findings suggesting that these are subjected to low RA signaling owing to persistent Cyp26a1/c1 expression . To determine whether RA signaling is required for development of von Ebner’s glands and posterior lingual glands , which normally develop in Cyp26a1/c1-free domains , we treated E12 . 5 control tongues/mandibles with DMSO or BMS493 ( S3 Table ) . While the DMSO-treated explants formed K8+ von Ebners’ glands and K8+ posterior lingual glands , the BMS493-treated explants failed to develop von Ebners’ glands and displayed developmentally arrested posterior lingual glands ( S7L–S7M’ Fig; S3 Table ) . These data demonstrate that abnormal RA activation in the ShhGFPCRE/Smof/f tongues causes glandular metaplasia and that development of von Ebners’ glands and posterior lingual glands requires RA inputs . They also show that the LE responds to RA . Despite this evidence and well-established expression of components of the RA pathway during tongue development , various transgenic mice reporting RA activity failed to show transgene activity in the LE and its derivatives [72–75] . Likewise , we were unable to detect RA activity in the LE except in subsets of desquamating cells of controls and ShhGFPCRE/Smof/f mutants carrying the RARE-hsp68-LacZ transgene , another reporter of RA activity [76] ( S8A–S8F Fig ) . Moreover , retinoid-treated tongue explants from RARE-hsp68-LacZ embryos failed to undergo epithelial RARE-hsp68-LacZ activation , whereas tails from these embryos displayed transgene activity ( S8G–S8I Fig ) . These observations and findings suggest that all the above transgenes are inactive in the LE and its derivatives . Accordingly , several studies pointed to lack of accuracy of such transgenic mice for visualization of RA activity in various tissues [77–79] . We found that loss of SHH signaling elicits enhanced development of FPs amidst increased RA activity . This strongly suggests that RA signaling promotes FP formation . To test for direct evidence of the phenomenon , we treated tongue explants from control embryos with retinoids and found that they induced development of oversized Shh+ placodes ( Fig 10C–10M; S4 Table ) . Furthermore , explants exposed to CD2314 and CD1530 exhibited ectopic Shh expression along the tongue midline and enhanced Shh signals , including in enlarged spots within two domains flanking the prospective intermolar eminence ( Fig 10J , 10L and 10M ) . By contrast , explants from controls ( Fig 10A and 10B; S4 Table ) and ShhGFPCRE/Smof/f mutants ( Fig 10N–10Q; S4 Table ) exposed to BMS493 or DEAB exhibited large areas with weak or no Shh signals . Taken together with our findings of occurrence of RA signaling in taste placodes ( expression of RALDHs amid absence of Cyp26a1/c1 ) , these data show that RA promotes taste placode formation . In this study , we adopted in vivo genetic approaches and experimental manipulation in tongue organ cultures in vitro which enabled us to reveal a temporal requirement for SHH signaling for proper tongue development . Before E12 . 5 , SHH signaling in the LE controls patterning and cell fate specification , whereas in the LM , SHH inputs are required to drive growth and morphogenesis ( S9 Fig ) . We uncovered unanticipated relationships between SHH and RA signaling during patterning of the LE ( S9 Fig ) . We discovered that genetic loss of SHH activity in the LE leads to enhanced RA signaling and that in the absence of SHH inputs , the LE overproduces oversized FPs and TBs and can even undergo de novo taste placode formation . We showed that Cyp26a1/c1 expression in the interplacodal epithelium requires SHH inputs and that retinoids or RA signaling inhibitors induce or inhibit FP formation , respectively . These findings provide compelling evidence that SHH inhibits FP induction by abating RA activity in the interplacodal epithelium . It is well-established that RA signaling plays a central role in controlling antero-posterior patterning of the limb , hindbrain , gut , and trunk [80–83] . Our genetic studies showed that loss of SHH signaling in the LE not only leads to formation of heterotopic glands in the oral tongue , but also causes overgrowth and mucous metaplasia of von Ebner’s glands . Furthermore , in tongue explants retinoids and RA inhibitors induced and inhibited lingual gland formation , respectively . These findings strongly suggest that antero-posterior patterning of the LE is controlled by antagonistic SHH and RA activities , where SHH inhibits and RA promotes lingual gland formation in the oral and pharyngeal tongues ( S9 Fig ) . To our knowledge , neither de novo taste placode induction nor cell fate switch causing glandular and Merkel cell metaplasiae in the oral tongue , as in the ShhGFPCRE/Smof/f mutants , have been reported in mouse models or in experimental studies in vitro , suggesting that SHH and RA pathways play key roles in cell fate specification within the LE ( S9 Fig ) . Our findings may provide insight into the etiopathogenesis of Merkel cell carcinoma of the tongue which remains an enigma [84] . We found that genetic loss of SHH signaling in the LE causes ectopic expression of Shh amidst abnormal activation of RA signaling and that retinoids induce expansion of Shh expression in tongue organ cultures . As normal levels of Shh expression in FPs require Wnt/β-catenin inputs [14 , 43] , these observations raise the question as to the relationships between RA and Wnt/β-catenin signaling during FP patterning . The Hedgehog and RA pathways are crucial for development and homeostasis of a vast array of tissues and organs , and deregulated Hedgehog signaling generates neoplasia [10 , 11 , 81–83] . Our findings thus constitute a basis for future research aimed at deciphering the interactions between these pathways during development and disease , a largely unexplored topic .
The experiments using mice were reviewed and approved by the Animal Research Ethics Committee in Göteborg , Sweden ( Dnr . 230–2010 , 174–2013 and 40–2016 ) . Mouse studies were also carried out under approved protocols in strict accordance with the policies and procedures established by the University of California , San Francisco ( UCSF ) Institutional Animal Care and Use Committees ( UCSF protocol AN084146 ) . Mice were euthanised by cervical dislocation . The different transgenic or knockin mice used in this work include the K14-CRE [40] , ShhGFPCRE and ShhCreERT2 [85] , Wnt1-CRE [6] , and R26R [86] mice . We crossed K14-CRE , ShhGFPCRE or Wnt1-CRE males with females carrying the Shh or Smo floxed alleles ( Shhf/fand Smof/f ) as described previously [17 , 18 , 40 , 87 , 88] . Thereafter K14-CRE/Smo+/f , ShhGFPCRE/Smo+/f , ShhCreERT2 , ShhCreERT2/Smo+/f , or Wnt1-CRE/Smo+/f males were crossed with Shhf/f or Smof/f females to generate mutants . K14-CRE , or ShhGFPCRE males were crossed with R26R females to generate reporter embryos carrying the CRE and R26R alleles ( K14-CRE/R26R and ShhGFPCRE/R26R ) . The RARE-hsp68-LacZ transgenic mice were as described [76] . We generated ShhGFPCRE/Smo+/f/RARE-hsp68-LacZ males which were thereafter crossed with Smof/f females to generate ShhGFPCRE/Smof/f/RARE-hsp68-LacZ mutants and controls ( without the CRE and/or the Smo floxed alleles ) . ShhGFPCRE/Smo+/f ( or without CRE ) without the RARE-hsp68-LacZ transgene were used as negative controls . For tamoxifen ( TAM ) induction of the ShhCreERT2/Shhf and ShhCreERT2/Smof/f mutants , pregnant females received intraperitoneal injections of TAM ( dissolved in corn oil ) every other day ( excluding the day of embryo harvest ) with the first injection consisting of 0 . 2 mg/g body weight ( bw ) TAM and the subsequent injections consisting of 0 . 1 mg/g bw . All mutants and their control littermates were identified by genotyping using DNA from the tip of the tail or a piece of embryonic tissue and/or , when possible , by their external phenotypes . The number of tongues from control and mutant embryos processed for in vivo studies is described in S1 Text . Heads or mandibles/tongues from the various mutants and their control littermates were prepared for sectioning or used as whole-mounts . For histological staining ( alcian blue van Gieson ) , immunohistochemistry , and immunofluorescence , specimens were fixed in ethanol-acetic acid as previously described [89 , 90] or in 4% paraformaldehyde in 1X PBS ( PFA/PBS ) . For in situ hybridization , specimens were fixed in 4% PFA/PBS at 4°C . The specimens were thereafter processed for whole-mount in situ hybridization or embedded in paraffin . Six-μm paraffin sections were used for immunostaining and in situ hybridization . For histology , immunostaining , and in situ hybridization the specimens were processed as previously described [18 , 19 , 89–91] . Sections were counterstained with Richardson’s azure II-methylene blue and methyl green after in situ hybridization and immunostaining , respectively . For green fluorescent protein imaging , tongues from ShhGFPCRE/+ ( controls ) and ShhGFPCRE/Smof/f mutants were fixed overnight in 1% PFA/PBS at 4°C , rinsed in PBS , and imaged as whole-mounts . For visualization of LacZ activity , specimens were fixed overnight in 2% PFA/PBS at 4°C . Whole-mounts and cryostat tissue sections were subsequently processed for β-galactosidase histochemistry as previously described [19 , 90] . The following antibodies were obtained and used for immunostaining: the Troma-1 ( anti-Keratin 8 ) rat monoclonal antibody ( 1:3000 ) developed by P . Brulet and R . Kemler from the DSHB developed under the auspices of the NICHD and maintained by the University of Iowa , Department of Biological Sciences , Iowa City , IA , USA; goat anti-ALDH1A2 ( RALDH2 , 1:400 ) and goat anti-SOX2 ( 1:3500 ) from Santa Cruz Biotechnology; rabbit anti-Keratin 6 ( 1:10 , 000 ) and chicken anti-Keratin 15 ( 1:10 , 000 ) from Covance; rabbit anti-Homer 1 ( 1:3000 ) and rabbit anti-Rab3c ( 1:1000 ) from Proteintech Group; rabbit Mab anti-ALDH1A1 ( RALDH1 , 1:400 ) and chicken anti-200 kD Neurofilament heavy ( NF200; 1:4000 ) from Abcam; rabbit anti-ALDH1A3 ( RALDH3 , 1:400 ) from Sigma Life Science; rabbit anti-RARγ ( 1:3000 ) from LifeSpan Biosciences; goat anti-SHH ( 1:200 ) from R&D Systems; rabbit anti-P2X2 ( 1:200 ) from US Biological; and rabbit anti-SHH ( Ab80; 1:800 ) [89 , 92] . Ab80 recognizes both SHH and Indian hedgehog ( IHH ) proteins [89 , 92] , hence staining of craniofacial cartilage which produces IHH was used as internal control in sections from mutants lacking Shh gene function . The following riboprobes were generated from linearized plasmids and used for in situ hybridization on tissue sections ( 35S-UTP-labelled riboprobes ) and/or whole-mount in situ hybridization ( Dig-labelled riboprobes ) : Shh targeting exon1 and exon2 [39; AP . McMahon , personal communication]; Ptch1 , Gli1 , and Shh/exon2 [40]; Cyp26a1 [93]; RARg and RARb [45]; and Cyp26c1 [66] . Tissue sections were processed for in situ hybridization with oligonucleotide probes targeting Mm-Gli1 ( NM_010296 . 2; target sequences: 25–1205 ) ; Mm-Cyp26a1 ( NM_007811 . 2; target sequences: 303–1556 ) , Mm-Ptch1 ( NM_008957 . 2; target sequences: 2260–3220 ) , Mm-RARg ( NM_007811 . 2; target sequences: 944–2623 ) and Mm-Wnt10b ( NM_011718 . 2; target: 989–2133 ) using the RNAscope technology ( Advanced Cell Diagnostics ) . Mandibular arches with the developing tongue were dissected from control embryos ( without the CRE and/or the Smo floxed alleles ) and ShhGFPCRE/Smof/f embryos . E11 . 5 and E12 . 5 explants were cultured for 6 and 9 days , respectively , in the presence of vehicle control ( DMSO ) , retinoids [all-trans retinoic acid ( RA ) and the retinoic acid receptor γ ( RARγ ) and RARβ selective agonists , CD1530 and CD2341 , respectively] , the pan-RAR antagonist BMS493 , or the retinoic acid synthesis inhibitor 4-diethylaminobenzaldehyde ( DEAB ) ( S3 Table ) . The specimens were thereafter processed for paraffin embedding and immunostaining for K8 as described above . All sections of explants from the ShhGFPCRE/Smof/f mutants treated with DEAB or BMS493 were processed for K8 immunostaining . The E11 . 5 specimens cultured for 6 days were exposed to DMSO , BMS493 ( 10 μM or 12 . 5 μM ) , DEAB ( 20 μM during the first 3 days of culture and 10 μM during the last 3 days of culture ) , CD1530 ( 1 μM ) , CD2314 ( 1 μM ) or RA ( 3 μM ) . The E12 . 5 explants cultured for 9 days were treated with DMSO or 10 μM BMS493 . Tongue and tail explants from E11 . 5 RARE-hsp68-LacZ embryos were cultured together in the same dishes for 6 or 2 days , respectively , in the presence of RA ( 3 μM; n = 2 ) , CD1530 ( 1 μM , n = 1 ) and CD2314 ( 1 μM , n = 1 ) . They were thereafter processed for LacZ staining following cryostat sectioning . E12 explants were cultured for 2 days in the presence of retinoids , BMS493 , or DEAB ( S4 Table ) , and processed for Shh in situ hybridization . These were treated with DMSO , DEAB ( 25 μM ) , RA ( 3 μM ) , BMS493 ( 12 . 5 μM ) , CD1530 ( 0 . 5 μM , 1 μM or 1 . 5 μM ) and CD2314 ( 0 . 5 μM , 1 μM or 1 . 5 μM ) . E12 explants from control embryos were cultured for 2 days in the presence of vehicle ( DMSO; n = 4 ) or 0 . 2 μM SAG ( n = 7 ) , a Smoothened agonist [94] , and processed for Shh in situ hybridization . The media containing freshly added compounds were changed daily . RA , SAG , BMS493 , and DEAB were from Sigma-Aldrich , St . Louis , MO , USA . CD1530 and CD2314 were from Tocris Bioscience , Abingdon , UK . All explants were cultured for the indicated periods at 37°C in an atmosphere of 5% CO2 in air . The culture medium consisted of DMEM/F-12 mix containing 2% fetal bovine serum , 2% vitamin A-free B27 supplement , 0 . 5% ITS , 0 . 5% glutamax , and penicillin/streptomycin ( 50 IU/ml and 50 μg/ml ) from Life Technologies . Tongues from ShhCreERT2/Shhf mutants ( first exposed to TAM at E10 . 5 or E11 . 5 ) and ShhGFPCRE/Smof/f mutants as well as their respective controls were dissected up to the level and including the circumvallate papilla . RNA was extracted using the RNAeasy Mini Kit ( Qiagen ) , and cDNA was synthesized by reverse transcriptase ( Qiagen ) . Quantitative PCR was performed using the iTAq Universal SYBR Green Supermix ( Biorad ) with the following primer sets: mRARg Forward 5'-GCAAGTACACCACGAACTCC , mRARg Reverse 3' AGGATGTCCAGACAAGCAGC; mRARb Forward 5'-GGAGAACTTGGGATCGGTGC mRARb Reverse 3'-TCTCGATGGCATTTTCCAGGC ) ; mCyp26b1 Forward 5’- CCGTGAGAAGCTGCAGTGTA , mCyp26b1 Reverse 3’- GGGTTCCATCCTTCAGCTCC [95]; mPtch1 Forward 5’- CTAGCAATAGGGACCGCTCA mPtch1 Reverse 3’- GTCTCAGGGTAGCTCTCATAGC [96] , and mActb Forward 5'-AGAGGGAAATCGTGCGTGAC , mActb Reverse 3'-CAATAGTGATGACCTGGCCGT ) . Student’s t-test was used for statistical analyses . | Knowledge of the biological mechanisms controlling cell fate specification is of paramount importance for cell-based therapies . Sonic hedgehog ( SHH ) and retinoic acid ( RA ) pathways play key roles in development and disease . The role of SHH during in vivo tongue development is a subject of great interest , and whether RA signaling has any function in the developing tongue is unknown . The tongue is covered by a mucosa made of lingual epithelium and lingual mesenchyme . Various structures , including mechanosensory filiform papillae , gustatory papillae harboring taste buds , and minor salivary glands , arise from the epithelium , but how these entities are specified remains unclear . Here we show that in the mesenchyme SHH signaling drives growth and morphogenesis , whereas in the epithelium , SHH controls patterning and cell fate specification . We demonstrate that SHH inhibits taste placode and lingual gland formation by antagonizing RA inputs . We also show that loss of SHH signaling elicits Merkel cell formation in the lingual epithelium , a tissue normally bereft of Merkel cells . This is at odds with the hairy epidermis where Merkel cell specification has been shown to be SHH-dependent . Our study establishes SHH and RA as key players in the control of cell identity within the lingual epithelium . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] | [
"merkel",
"cells",
"taste",
"buds",
"medicine",
"and",
"health",
"sciences",
"in",
"situ",
"hybridization",
"molecular",
"probe",
"techniques",
"retinoid",
"signaling",
"social",
"sciences",
"neuroscience",
"epithelial",
"cells",
"developmental",
"biology",
"tongue",
... | 2017 | Cell fate specification in the lingual epithelium is controlled by antagonistic activities of Sonic hedgehog and retinoic acid |
Prions are unconventional infectious agents thought to be primarily composed of PrPSc , a multimeric misfolded conformer of the ubiquitously expressed host-encoded prion protein ( PrPC ) . They cause fatal neurodegenerative diseases in both animals and humans . The disease phenotype is not uniform within species , and stable , self-propagating variations in PrPSc conformation could encode this ‘strain’ diversity . However , much remains to be learned about the physical relationship between the infectious agent and PrPSc aggregation state , and how this varies according to the strain . We applied a sedimentation velocity technique to a panel of natural , biologically cloned strains obtained by propagation of classical and atypical sheep scrapie and BSE infectious sources in transgenic mice expressing ovine PrP . Detergent-solubilized , infected brain homogenates were used as starting material . Solubilization conditions were optimized to separate PrPSc aggregates from PrPC . The distribution of PrPSc and infectivity in the gradient was determined by immunoblotting and mouse bioassay , respectively . As a general feature , a major proteinase K-resistant PrPSc peak was observed in the middle part of the gradient . This population approximately corresponds to multimers of 12–30 PrP molecules , if constituted of PrP only . For two strains , infectivity peaked in a markedly different region of the gradient . This most infectious component sedimented very slowly , suggesting small size oligomers and/or low density PrPSc aggregates . Extending this study to hamster prions passaged in hamster PrP transgenic mice revealed that the highly infectious , slowly sedimenting particles could be a feature of strains able to induce a rapidly lethal disease . Our findings suggest that prion infectious particles are subjected to marked strain-dependent variations , which in turn could influence the strain biological phenotype , in particular the replication dynamics .
Transmissible spongiform encephalopathies ( TSE ) , such as human Creutzfeldt-Jakob disease , sheep scrapie , bovine spongiform encephalopathy ( BSE ) and chronic wasting disease of cervidae , are infectious , fatal , neurodegenerative disorders caused by prions [1] . Prions are unconventional pathogens primarily composed of PrPSc , a rearranged conformer of the ubiquitously expressed prion protein ( PrPC ) , whose precise physiological function is largely unknown . Upon infection , PrPSc dictates the self-perpetuating conformational conversion of PrPC into nascent PrPSc . This conversion involves – without any apparent post-translational modification – the refolding of soluble , alpha-helix-rich PrPC molecules into beta-sheet enriched PrPSc polymers that form deposits in TSE-infected brains [2] , [3] and are assumed to be responsible for the observed neurodegenerative disorders [4] . The conversion reaction may proceed through a nucleated polymerization mechanism in which PrPSc multimers recruit PrPC molecules and trigger their conformational conversion into PrPSc ( for review [5] ) . The refolding/multimerisation process confers distinct physico-chemical properties to PrPSc , such as insolubility in non-denaturing detergents and partial resistance to proteolysis [6] . Distinct prion entities , referred to as strains , are known to self-propagate in the same host and exhibit distinguishable phenotypic traits that are heritable , such as incubation time , neuropathological and biochemical properties ( for reviews: [7] , [8] , [9] ) . Accumulating experimental evidence indicates that strain-specified properties are encoded within structural differences in the conformation of the PrPSc molecules , which are faithfully imparted to host PrPC during the conversion process [10] , [11] , [12] , [13] , [14] , [15] , [16] , [17] . However , the extent to which the aggregation state varies between different stains , and participates to strain-specific prion biology is unknown . The various fractionation methods and preparative procedures previously employed to estimate the size of the infectious particles [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] have led to a vast range of measured sizes , making it difficult to relate any variation to potential strain differences . Of note , almost all of these studies used substantially purified PrPSc as a starting material . In this study , we developed a specific protocol to fractionate PrP particles according to their sedimentation velocity properties in a viscous medium , characterized their relative levels of infectivity and looked for strain-specific variations . In contrast to previous reports , experiments were performed on crude brain homogenates , which a priori contain all TSE infectivity . We worked with a panel of strains that were biologically cloned on homogeneous genetic backgrounds , obtained after transmission of either classical and atypical ( Nor98 ) sheep scrapie and BSE , or hamster scrapie infectious sources in transgenic mice expressing ovine PrP ( VRQ allele; tg338 mice ) and hamster PrP ( tg7 line ) , respectively . We demonstrate that the sedimentation profile of the infectious component dramatically varies with the strain . We further show that the predominance of slowly sedimenting infectious particles that segregate from the bulk of proteinase K-resistant PrPSc particles may be a distinctive feature of strains able to induce a rapidly lethal disease .
PrPSc aggregates present in detergent-solubilised brain tissue homogenates were fractionated by sedimentation velocity centrifugation in an iodixanol gradient ( Optiprep ) . The experimental conditions were established with brain material from tg338 mice that were infected or not with LA21K fast strain ( referred to as LA21K ) , a prototypal , rapid strain that kills the mice within ∼2 months ( see Table 1 for information on the strains used in this study ) . As a first step , we tested a variety of detergents for solubilization , which showed variable efficacy in terms of partition of PrPC and PrPSc species . For example , the use of standard solubilization buffers containing Triton X-100 and sodium deoxycholate or sarkosyl led to sedimentation of both isoforms throughout the gradient ( Figure S1 ) , indicating an incomplete release of total PrP from cellular constituents . In contrast , the sequential use of dodecyl maltoside and sarkosyl resulted in more efficient separation of the two PrP isoforms . Thus , in the conditions eventually employed ( see Figure 1 for a summarizing flow diagram ) , the bulk of PrPC molecules remained in the upper fractions 1–4 ( Figure 2A and D , green line ) , while both PrPSc ( Figure 2B ) and proteinase K ( PK ) resistant PrPSc species ( Figure 2C–D , black line ) were mainly detected in fractions 6–20 of the gradient . Importantly , no pelleted PrP material was observed in the selected conditions . Increasing the ultracentrifugation time caused the majority of PrPSc to sediment toward the heaviest fractions of the gradient , indicating that this material had not reached its density equilibrium ( data not shown ) . Both dodecyl maltoside and sarkosyl are known to efficiently solubilize membrane structures , including rafts [26] , [27] , [28] , yet PrPSc could be attached to abnormal , prion-induced structures . To address this point , brain homogenates were solubilized using these detergents in more stringent conditions , i . e . at 37°C instead of 4°C [29] , however the sedimentation profile of PrPSc was affected only marginally ( Figure S2A ) . In order to assess the reproducibility of the partition and to enable quantitative analysis of the data , 7 independent fractionations were performed using different pooled or individual brains and the resulting data fitted ( Figure 3A , black line ) . This revealed that ∼80% of the PK-resistant PrPSc material sedimented as one major peak ( maximum in fractions 10–12 ) with a Gaussian-like distribution . Standard globular macromolecules and ovine recombinant PrP oligomers [30] loaded on gradients run in parallel enabled estimation of the approximate molecular mass of the PK-resistant PrPSc aggregates forming the peak in fraction 10–12: between 200 and 500 kDa ( by reference to the marker proteins , the sedimentation profile of which was affected only marginally in the presence of detergents ) , and ∼850 kDa based on the position of the 36-mer PrP oligomer ( Figure 3A ) . When solubilized brain material was PK-treated prior to ultracentrifugation , the PrPres sedimentation profile resembled that observed with intact brain material ( Figure S2B ) . However , when semi-purified PrPSc in the form of scrapie-associated fibrils [31] , [32] was resolubilized and centrifuged , a markedly different profile was obtained , with peaks in fractions 22 and 30 ( bottom fraction ) ( Figure S2C ) . Interestingly , fast sedimenting PrPSc material was also observed with Italian scrapie agent ( referred to as SSit ) , which in tg338 mice produces very long incubation times and abundant plaque-like PrPSc deposits in the brain [33] , in contrast to the LA21K agent . These plaques can be stained by thioflavin S ( Figure S3A–B ) , indicating the presence of amyloid fibrils . When SSit-infected brain material was fractionated , the majority of PrPSc multimers peaked in fractions 24 to 30 of the gradient ( Figure S3C ) . These results suggest that the experimental conditions employed preserve potential differences in the aggregation state of PrPSc thereby enabling the comparative analysis of sedimentation properties of “close to natural” PrPSc aggregates and of associated infectivity . The distribution of prion infectivity throughout the gradients was determined by an incubation time bioassay [34] . tg338 mice were inoculated intracerebrally with diluted aliquots from the different fractions . In terminally diseased mice , the PrPSc electrophoretic profile and regional distribution in the brain observed for representative fractions were both consistent and similar to that with the original brain material , indicating a conservation of the strain biological phenotype ( Figure S4A and data not shown ) . The mean survival time values resulting from the analysis of 2 independent gradients are shown in Figure 3A ( red line ) . Typically , the mice inoculated with the PK-resistant PrPSc-richest fractions ( 6–20 ) succumbed to disease in more than 80 days , whereas those inoculated with fractions 1–3 died in a markedly shorter time , ∼60–70 days . The correlation between the mean survival time values and infectivity was established by using a standard infectious dose/survival time curve previously established for this strain ( Figure S5 ) . This analysis indicated that fractions 1–2 were between 100- and 1000-fold more infectious than fractions 6–20 ( Figure 3A , blue scale ) . These upper fractions - within the sedimentation peak of aldolase ( 158 kDa ) and upstream of 12-mer PrP oligomer – totaled <10% of PK-resistant PrPSc molecules ( Figure 3A ) . There is substantial evidence to indicate that a fraction of PrPSc can exhibit low sedimenting properties and be PK-sensitive [28] , [35] , [36] . Recently , thermolysin has been used as a means to isolate PK-sensitive forms of PrPSc , while degrading PrPC [37] . When the upper fractions from LA21K gradients were thermolysin-digested , no enrichment in thermolysin-resistant species was observed by immunoblot as compared to unfractionated brain material ( Figure S6A–C ) . To further analyze the forms of PrPSc present in the upper fractions , aliquots were centrifuged at 100 000 g for 1 h to produce soluble ( supernatant ) and insoluble ( pellet ) fractions , before immunoblot analysis . The ratio of soluble and insoluble PrP species in LA21K versus uninfected fractions was determined based on signal intensities . As a result , the top two LA21K fractions were reproducibly shown to contain equivalent amounts of soluble material and about 2-fold more sedimentable material as compared to the corresponding uninfected fractions ( Figure S6D ) . Detergents and lipids have been proposed to increase the apparent infectious titer of PrPSc preparations non-specifically [38] , [39] and such compounds are relatively abundant in the upper fractions of the gradient . To test whether such an effect was responsible for the comparatively high infectivity levels of the top fractions , the PK-resistant PrPSc-enriched fractions 10 to 12 were mixed , incubated with either the top fractions 1 to 3 of a gradient made with uninfected tg338 brain or with dodecyl maltoside alone , and inoculated to mice . As a result , in either condition , the relative titer of these fractions was not significantly modified ( Figure 4 ) . To confirm that the differences in survival times observed between mice inoculated with the various fractions were correlated with differences in infectivity content , a mouse-free , cell bioassay was used . The distribution and level of LA21K infectivity in the gradient was measured using Rov cells [40] that were exposed in parallel to fraction aliquots and to serial tenfold dilutions of a LA21K brain homogenate prepared in the same conditions . Consistent with the bioassay data , the most infectious fractions were found at the top of the gradient and were ≥100-fold more infectious than the middle fractions ( Figure S7 ) . Overall these data indicate that the upper fractions were intrinsically highly infectious . The fact that the cumulated infectivity in the gradient fractions did not differ significantly from that present in the loaded material prior solubilization also supports the conclusion that the detergents used did not alter infectivity estimates . Brains of tg338 mice infected by another fast ovine strain named 127S ( Table 1 ) were also fractionated and analyzed for PrP and infectivity content . 127S PK-resistant PrPSc peaked in fractions 10–12 ( Figure 3B ) , as in the case of LA21K agent , despite some variation of the sedimentation profile in the bottom part of the gradient . Strikingly , the sedimentation profile of infectivity again largely segregated from that of PK-resistant PrPSc as assessed by mouse bioassay . The top two fractions were at least 50–100-fold more infectious than all the other fractions , including the major PK-resistant PrPSc peak ( Figure 3B ) . We next examined whether the decoupling of PK-resistant PrPSc and infectivity sedimentation profiles was a general feature of ovine strains . Three more strains were studied of which the incubation time in tg338 mice is at least twice that of LA21K and 127S: LA19K , Nor98 and sheep BSE ( see Table 1 ) . Four to five independent fractionations with different pooled or individual brains were performed for each strain . The combined curves resulting from the replicate analysis of PrP content indicated that a majority of PK-resistant PrPSc peaked in fractions 10–12 , similar to that seen with the two fast strains . However , faster sedimenting species were also observed , notably in fractions 16 , 20 for LA19K and fractions 22–24 for Nor98 ( Figure 3C–D ) . Remarkably , the infectivity sedimentation profile of these 3 strains , as established from bioassay of two independent gradients , tended to overlap PK-resistant PrPSc distribution , with a very small proportion of the total infectivity in the top fractions . LA19K most infectious fractions ranged from fractions 8 to 24 with a peak in fraction 20 ( range of mean survival time: 152 to 163 days ) , while the top and bottom fractions were ∼100-fold less infectious ( mean survival time ∼185 to 210 days; Figure 3C ) . Nor98 infectivity peaked in fraction 11 and to a lesser degree in fraction 17 and 22 ( mean survival time 222 , 240 and 245 days , respectively; Figure 3D ) . Fractions in the immediate vicinity of these peaks were among the most infectious , ( except fraction 13 ) . In contrast , the upper fractions were ∼100-fold less infectious ( survival time prolonged by >40 days ) . The most infectious sheep BSE fractions were found in fractions 6–12 , 16 and 20 ( mean survival times of 155–160 , 164 and 163 days ) while the top and bottom fractions were about 50-fold and 100-fold less infectious , respectively ( survival time of ∼175 days and >180 days; Figure 3E ) . To further explore the possibility that slow sedimenting infectivity could be a specific feature of fast prion strains , we applied the same sedimentation velocity protocol to three hamster strains passaged on tg7 transgenic mice expressing hamster PrP ( Table 1 ) . For fast strains 139H and Sc237 , the infectivity peaked in the top two fractions , which contained ∼10% of the total PK-resistant PrPSc material present in the gradient . The two 139H PK-resistant PrPSc peaks in fractions 11–12 and 16–18 and the Sc237 PK-resistant PrPSc peak in fraction 11–12 were ∼50-fold and <10-fold less infectious , respectively ( n = 2 independent experiments made with different individual brains; Figure 5A–B ) . ME7H strain , characterized by a longer incubation time , produced a different picture since the mice inoculated with the PK-resistant PrPSc peak in fraction 11 were the fastest to succumb to disease , i . e . ∼180 days , whereas those inoculated with the top 3 fractions had mean survival times significantly prolonged by 20 to 40 days ( Figure 5C ) . Therefore , much less infectivity was present in the upper region of the gradient than in the PK-resistant PrPSc containing fractions ( about 50–100-fold , based on the available results of the endpoint titration of ME7H , still ongoing ) . Collectively , the contrasted sedimentation properties of fast and slow hamster strains were reminiscent of the results obtained with the ovine strains .
Here we compared the sedimentation velocity properties of the infectivity and of abnormal PrP amongst several prion strains , using experimental conditions aimed at preserving as much as possible the “natural” multimerization state of the prion particles while minimizing artifacts due to improper membrane solubilization . To our knowledge this is the first study that allows a rigorous comparison of phenotypically distinct strains , cloned and propagated on the same genetic background . We found striking , strain-specific differences in the sedimentation profile of the infectious prion particles , which are not reflected in the sedimentation properties of the bulk of PrPSc . Fractionation of five tg338 mouse-passaged ovine prions revealed a major PK-resistant PrPSc population , which peaked at the same position of the gradient regardless of the strain . By comparison with standard molecular mass markers and recombinant ovine PrP oligomers [30] , we estimated that this population might correspond to approximately 12–30 PrP monomers averaging ∼30 kDa each , if constituted of PrP only , with the caveat that great caution must be exercised when attributing a size to a polymer by comparing its velocity to that of molecular mass markers . This result suggests that PrPSc is not a collection of multimers with a regular continuum of size . However , the overall sedimentation profiles were not uniform among the strains , indicating that the size distribution of PrPSc aggregates is strain-dependent . tg7 mouse-passaged hamster prions showed PrPSc sedimentation characteristics resembling that of ovine strains , with the same position of the major peak and limited variation . Greater differences in the size distribution of PrPSc aggregates may however exist depending on strain and/or PrP sequence , since one amyloid-forming ovine prion ( Italian scrapie; Figure S3 ) showed a clear shift of PrPSc toward heavier fractions of the gradient . In two studies , larger polymers were shown to be more PK-resistant than smaller ones [28] , [35] , indicating that the resistance to proteolysis of PrPSc largely depends on its quaternary structure . The PrPSc associated with Nor98 agent , a newly discovered strain responsible for an atypical form of field scrapie , is highly sensitive to PK digestion , when compared to the other ovine strains studied here [41] , [42] . Notwithstanding , Nor98 PrPSc exhibited not only the same predominant peak as the other strains , but also the highest proportion of faster sedimenting PrPSc species . This argues that the pronounced PK sensitivity of Nor98 PrPSc is not due to low size aggregates , but rather to its tertiary structure . Supporting this view , its C-terminal region is accessible to PK , in contrast to classical scrapie agents [43] , [44] . The infectivity sedimentation profiles unexpectedly contrast with that of the PrPSc aggregates . While infectivity and PK-resistant PrPSc roughly co-fractionated for LA19K , Nor98 and sheep BSE , their respective distribution was mostly decoupled for LA21K and 127S: with these two fast strains , infectivity essentially partitioned in the upper fractions , which were 2–3 logs more infectious than the PK-resistant PrPSc–richest fractions . LA21K , 127S and Nor98 are three strains exhibiting similarly high infectious titers in tg338 mice , making the observed differences particularly striking . Remarkably , a comparable situation was observed for hamster strains . Thus , ME7H infectivity and PK-resistant PrPSc sedimentation velocity profiles were broadly congruent , whereas for the fast strains 139H and Sc237 the infectivity peaked in the upper region of the gradient . Altogether , these findings lend support to the view that the predominance of slow sedimenting particles may be a common feature of prion strains with short incubation time . Such a decoupling between infectivity and PK-resistant PrPSc with respect to the size of the particles is to our knowledge unprecedented in the literature . Two earlier studies performed with fast hamster prions [18] , [23] also reported a relatively low PrPres content of the most infectious fractions . However , the level of infectivity in these fractions did not exceed that of the PrPres-richest fractions . The preparations used were fibrillar PrPres material under the form of SAF or Rods , disaggregated by sonication in the presence of anionic detergents [18] , [23] . Such a procedure is likely to destroy discrete subpopulations of infectious particles , which may explain the observed discrepancy . Which form ( s ) of PrPSc could support the high infectivity of the fast strains' slow sedimenting component ? One possibility is that it consists essentially of PK-resistant aggregates , with high specific infectivity [24] , [45] , [46] , [47] . If so , then ≥99% of LA21K or 127S infectivity would be supported by ≤10% of PK-resistant PrPSc molecules . Alternatively , infectivity could be mostly associated with a form of PrPSc with low resistance to PK . While a variable , strain-dependent proportion of abnormal PrP seems fairly PK-sensitive [28] , [35] , [36] , [48] , [49] , little is known about its specific infectivity , with two recent studies suggesting that it could be minimal [37] , [50] . However , a PK-sensitive and soluble form of PrPSc has been shown to support a substantial fraction of infectivity [51] and to have a good in vitro converting activity [35] . The abundance of both PrPC and other components in the upper fractions impeded further characterization of the most infectious PrPSc particles in the present study . Classical approaches based on PrPSc conformation-dependent assay [36] , [49] , [52] are unhelpful here; as already shown by others , low sedimenting , PrPC-rich fractions contain little or no conformation-dependent immunoreactive material [28] , [35] . Also , we failed to detect soluble or thermolysin-resistant PrPSc material that might be indicative of the presence of PK-sensitive molecules [48] , [51] in these fractions . Additional experiments including the titration of PK-treated and then fractionated infectious material are ongoing to further assess the protease-resistance of the slow sedimenting component . What physical properties could account for slow sedimentation of fast strains infectious component ? The detergents employed are known to produce a high degree of membrane solubilization [27] , [28] , [53] , [54] , [55] , [56] , including for GPI-anchored proteins of detergent-resistant microdomains [26] , and they led indeed to an efficient solubilization of PrPC ( Figure S6E ) . Dodecyl maltoside is also known to preserve activity of protein complexes in the detergent-solubilized state [27] . It is still possible that a tightly bound cellular component of low density , such as lipid molecule [45] , or low-density lipoprotein [57] , remains part of the prion particles in the most infectious fractions . This would imply that the tightness of such an interaction specifically differs between strains . Alternatively , the low sedimenting infectivity component could involve truly small size PrPSc particles . In this hypothesis , the data would indicate a size smaller than a PrP pentamer , which is compatible with that reported for PK-sensitive PrPSc aggregates [28] , [35] . As a means to distinguish between small size and lipid associated PrPSc aggregates , LA21K gradient centrifugation time was doubled . As a result , infectivity was found to peak in fraction 4 instead of fraction 2 ( data not shown ) . While arguing against lipid floatation of the most infectious component , this does not exclude the presence of tightly associated lipids . Additional experiments will be needed to address this issue . The neuropathology induced in mice by the fast and slow sedimenting particles did not differ for a given prion , therefore suggestive of structurally related multimers ( Figure S4 ) . The slow sedimenting infectious particles could reflect a stronger tendency of large PrPSc polymers to fragment . In the case of yeast prions [PSI+] , it has been proposed that the fittest strains are those whose large fibers break more easily into smaller oligomers that in turn act as new seeds for conversion [58] , a concept that was then extended to mammalian prions [59] . In this regard , our preliminary results indicate that LA21K and 127S PrPSc aggregates exhibit the lowest ‘stability’ among the ovine strains , as assayed by conformational stability assay . In addition to providing another measurable criterion of prion strain-related phenotypic variation , this study revealed the diversity of their infectious component . Further biochemical and biophysical investigations will be crucial for a mechanistic understanding of the replication dynamics of mammalian prions , in relation with the disease phenotype .
All the experiments involving animals were approved by the INRA Jouy-en-Josas ethics committee in accordance with the European Community Council Directive 86/609/EEC . The ovine prion strains used in this study have been obtained through serial transmission and subsequent biological cloning by limiting dilutions of classical and atypical field scrapie and experimental sheep BSE sources to tg338 transgenic mice expressing the VRQ allele of ovine PrP . The characterization of their phenotype in tg338 mice was performed as previously reported [41] , [60] , [61] . Pooled or individual tg338 mouse brain homogenates ( 20% wt/vol . in 5% glucose ) were used in centrifugation analyses . Three hamster strains , 139H , Sc237 and ME7H , were also studied . These strains ( kindly provided by R . Carp , Staten Island , NY , USA ) were serially passaged on tg7 transgenic mice expressing hamster PrP ( kindly provided by CSL-Behring ( Marburg ) ; [48] , [62] ) . Both 139H and Sc237 were subsequently cloned by limiting dilution on this genetic background . Individual tg7 infected brains ( 20% wt/vol . ) were used in centrifugation analyses . Non-infected brain tissue homogenates served as controls . The entire procedure was performed at 4°C . Mouse brain homogenates were solubilized by adding an equal volume of solubilization buffer ( 50 mM HEPES pH 7 . 4 , 300 mM NaCl , 10 mM EDTA , 2 mM DTT , 4% ( wt/vol . ) dodecyl-β-D-maltoside ( Sigma ) ) and incubated for 30 min on ice . Sarkosyl ( N-lauryl sarcosine; Fluka ) was added to a final concentration of 2% ( wt/vol . ) and the incubation continued for a further 30 min on ice . A volume of 150 µl was loaded on a 4 . 8 ml continuous 10–25% iodixanol gradient ( Optiprep , Axys-shield ) , unless specified otherwise , with a final concentration of 25 mM HEPES pH 7 . 4 , 150 mM NaCl , 2 mM EDTA , 1 mM DTT , 0 . 5% Sarkosyl . Gradient linearity was verified by refractometry . In standard experiments ( Figure 1 ) , the gradients were centrifuged at 285 000 g for 45 min in a swinging-bucket SW-55 rotor using an Optima LE-80K ultracentrifuge ( Beckman Coulter ) . Gradients were then manually segregated into 30 equal fractions of 170 µl from the bottom using a peristaltic pump . Fractions were aliquoted for immunoblot or bioassay analyses . The strains were fractionated in parallel to preserve as much as possible identical experimental conditions . To avoid any cross-contamination , each piece of the equipment was thoroughly decontaminated with 5 M NaOH followed by several rinses in deionised water after each gradient collection . Standard markers ( GE Healthcare , Little Chalfont , UK ) of aldolase ( 158 kDa ) , catalase ( 232 kDa ) , ferritin ( 440 kDa ) and thyroglobulin ( 669 kDa ) were run in parallel . The protocol used was as described above except that PK ( 100 µg/ml final concentration; Euromedex , Mundolsheim , France ) was added during the solubilization phase in sarkosyl ( 1 h at 37°C ) . Brain homogenates were treated with 20 µg/ml of PK for 1 h at 37°C . The digestion was stopped by the addition of 5 mM phenylmethylsulfonyl fluoride . The solution was added to 10% sarkosyl and 10 mM Tris-HCl pH 7 . 4 and then centrifuged at 175 000 g for 30 min at 20°C on a 10% ( wt/vol . ) sucrose cushion in a Beckmann TL100 ultracentrifuge . Pellets were resuspended in 2% ( wt/vol . ) dodecyl maltoside and the above solubilization/fractionation protocol was followed . Monomeric and oligomeric forms of purified ovine recombinant PrP [30] were resuspended at a final concentration of 7–10 µM in 20 mM citrate buffer . An aliquot ( 150 µl ) was loaded onto a 10–25% iodixanol gradient in citrate buffer and centrifuged at 285 000 g for 45 min in a SW-55 rotor . Gradients were segregated as described above . Fraction aliquots ( 20 µl ) were analyzed for PrP content by immunoblot ( see below ) . In these conditions , recombinant , monomeric PrP was found in the upper fractions 1–3 ( not shown ) . Aliquots of the collected fractions were treated or not with 50 µg/ml PK before methanol precipitation . The pellet was resuspended in Laemmli buffer and denatured at 100°C for 5 min . The samples ( 15 µl ) were run on 4–12% NuPAGE gels ( Invitrogen , Cergy Pontoise , France ) , electrotransferred onto nitrocellulose membranes , and probed with 0 . 1 µg/ml biotinylated anti-PrP monoclonal antibody Sha31 as previously described [60] . Immunoreactivity was visualized by chemiluminescence ( GE Healthcare ) . The amount of PrP present in each fraction was determined by the GeneTools software after acquisition of chemiluminescent signals with a GeneGnome digital imager ( Syngene , Frederick , Maryland , United States ) . All Bio-Rad TeSeE detection kit reagents were kindly provided by S . Simon ( CEA , France; [63] ) . Briefly , aliquots ( 75 µl ) of the collected fractions were digested with PK ( 50 µg/ml final concentration ) for 1 h at 37°C before B buffer precipitation and centrifugation at 28 000 g for 15 min . The pellet was resuspended in 25 µl of 5 M urea before denaturation at 100°C for 10 min . R6 buffer ( 200 µl ) was subsequently added to the samples and duplicates were analyzed in microtiter plates coated with anti-PrP antibody 11C6 . The plates were left at room temperature for 2 h . After 3 washes in R2 buffer , 100 µl/well of the enzyme conjugate ( Bar224 anti-PrP antibody ) was added for 2 h . The substrate ( 100 µl ) was added for 30 min and incubated in the dark . The absorbance was read at 450 nm . A dilution range of ovine , monomeric recombinant PrP was used for quantification of relative PK-resistant PrPSc levels . The PK-resistant PrPSc sedimentation profiles obtained by either immunoblot or ELISA were normalized to units and decomposed using multiple Gaussians fits procedures with a maximum entropy minimization approach . Fractions were methanol-precipitated . The pellet was resuspended in lysis buffer ( 2% sodium deoxycholate , 2% Triton X-100 , 200 mM Tris-HCl pH 7 . 4 ) and mixed with an equal volume of thermolysin diluted in lysis buffer to yield a final concentration of 125 µg/ml ( unless indicated otherwise ) for 1 h at 70°C . The samples were analyzed by electrophoresis ( 4–12% gels ) and immunoblotted as above . Blots were probed with either Sha31b or anti-octarepeat specific Pc248 anti-PrP antibody [64] at a final concentration of 0 . 1 µg/ml , before acquisition of chemiluminescent signals with a GeneGnome digital imager and analysis by the GeneTools software ( Syngene , Frederick , Maryland , United States ) . Aliquots ( 20 µl ) of the fractions were added to 80 µl of 5% glucose before centrifugation at 100 000 g for 1h at 4°C in a Beckmann TL100 ultracentrifuge to generate soluble ( supernatant ) and insoluble ( pellet ) fractions . Proteins in the supernatant were precipitated with 400 µl of cold methanol , centrifuged at 16 000 g for 30 min before denaturation in 100 µl of sample buffer . The insoluble pellet was resuspended in 20 µl of Laemmli buffer before denaturation . Samples ( 20 µl ) were analyzed by immunoblot as described above . Fractions 1 to 4 and then every other two fractions ( unless specified otherwise ) were diluted extemporarily in 5% glucose ( 1∶5 ) . This procedure was performed in a class II microbiological cabinet according to a strict protocol to avoid any cross-contamination . Individually identified 6- to 10-week old tg338 or tg7 recipient mice ( n = 6 mice per fraction ) were inoculated intracerebrally with 20 µl of the solution . Recipient mice inoculated with fractionated uninfected mouse brain were euthanized while still healthy at >400 days post-infection . Their brain was negative for PrPres content . Mice showing TSE neurological signs were monitored daily and euthanized in extremis . Brains were removed and analyzed for PrPres content by either immunoblot or histoblot ( see below ) as a confirmatory test . The survival time was defined as the number of days from inoculation to euthanasia . The survival times of tg338 or tg7 reporter mice was measured for each tenfold dilution tested during endpoint titration experiments performed with all but ME7H strains . Animals inoculated with 2 mg of infectious brain tissue were assigned a relative infectious dose of 0 . From these data , curves representing the relative infectious dose to survival time were established ( [41] and Figure S5 ) . The different patterns in survival time distribution among the gradients can thus be looked at as a function of relative infectious dose so as to estimate what difference in survival times between inoculated fractions means in terms of infectivity . The scrapie cell assay technique will be fully described elsewhere . Briefly , LA21K gradient fractions aliquots ( typically 20–30 µL ) were methanol precipitated before resuspension in culture medium ( alpha minimal essential medium supplemented with 10% fetal bovine serum , 100 U/ml penicillin and 10 µg/ml streptomycin ) . We verified that methanol precipitation did not affect the overall level of infectivity . Rov cell [40] monolayers established in a 96 well plate were exposed to the fractions for one week . After several washes , the cells were further cultivated for two weeks before fixation and PrPSc detection by immunofluorescence as previously described [65] . Immunofluorescence signals were acquired with an inverted fluorescence microscope ( Zeiss Axiovert ) . A program in NIH Image J software was designed to quantify the levels of PrPSc signal per cell in each well . Serial tenfold dilutions of LA21K infected brain homogenates were prepared in the same conditions and run in parallel experiments to establish a tissue culture infectious doses curve that directly relates to the percentage of PrPSc content . Brains were rapidly removed from euthanized mice and frozen on dry ice . Cryosections were cut at 8–10 µm , transferred onto Superfrost slides and kept at −20°C until use . Histoblot analyses were performed on 3 brains per infection , using the 12F10 anti-PrP antibody as described [60] . For thioflavin-S binding , formalin- or methanol-fixed sections were incubated with 0 . 01% thioflavin-S for 1 hour as previously described [66] . Sections were then incubated with nuclear marker 4′ , 6-diamidino-2-phenylindole ( Sigma ) , mounted in fluoromount-G ( Interchim ) before acquisition with an inverted fluorescence microscope ( Zeiss Axiovert ) and analysis with the Metamorph software . The Swiss-Prot accession numbers for the proteins mentioned in the text are sheep ( P23907 ) and hamster PrP ( P04273 ) . | Prions are unconventional transmissible agents causing fatal neurodegenerative diseases in human and animals . They are thought to be formed from polymers of abnormal conformations of the host-encoded prion protein ( PrP ) , but little is known about the physical organization of the infectious particles and any relationship between packing order and infectivity . As an additional layer of complexity , different PrP conformational variants associated with distinct biological phenotypes , or ‘strains’ , can propagate in the same host . We subjected PrP polymers from eight different ovine and hamster prion strains to sedimentation velocity centrifugation , which allows separation of macromolecular complexes according to their size , density or shape . We showed that , whereas the PrP sedimentation profiles share common features , the infectivity profiles exhibit striking differences amongst the strains . For four of them , the infectious component was predominantly associated with slowly sedimenting particles , suggestive of small size oligomers and/or low density PrP aggregates . Such particles appeared to be a feature of strains able to induce a rapidly lethal disease in the recipient host . Our findings suggest that prion infectious particles are subjected to marked strain-dependent variations , which in turn could influence the strain biological phenotype , in particular the replication dynamics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/prion",
"diseases",
"biochemistry/protein",
"folding",
"neurological",
"disorders/prion",
"diseases"
] | 2010 | The Physical Relationship between Infectivity and Prion Protein Aggregates Is Strain-Dependent |
Assessing the impact of the social environment on health and disease is challenging . As social effects are in part determined by the genetic makeup of social partners , they can be studied from associations between genotypes of one individual and phenotype of another ( social genetic effects , SGE , also called indirect genetic effects ) . For the first time we quantified the contribution of SGE to more than 100 organismal phenotypes and genome-wide gene expression measured in laboratory mice . We find that genetic variation in cage mates ( i . e . SGE ) contributes to variation in organismal and molecular measures related to anxiety , wound healing , immune function , and body weight . Social genetic effects explained up to 29% of phenotypic variance , and for several traits their contribution exceeded that of direct genetic effects ( effects of an individual’s genotypes on its own phenotype ) . Importantly , we show that ignoring SGE can severely bias estimates of direct genetic effects ( heritability ) . Thus SGE may be an important source of “missing heritability” in studies of complex traits in human populations . In summary , our study uncovers an important contribution of the social environment to phenotypic variation , sets the basis for using SGE to dissect social effects , and identifies an opportunity to improve studies of direct genetic effects .
Social interactions contribute to health and disease ( e . g . peer smoking increases one’s risk of taking up smoking ) . So far , quantifying social effects has required a clear hypothesis about the mechanisms mediating the influence of the social environment ( in the example above peer smoking is the trait that mediates the social influence ) . For many phenotypes however , such hypotheses do not exist . Therefore we propose an alternative strategy to study social effects: we investigate effects on an individual's phenotype that arise from genotypes of social partners ( social genetic effects , SGE , also called indirect genetic effects[1 , 2] ) . SGE constitute the genetic basis of social effects and can be detected without prior knowledge of the phenotypes through which the social influence is exerted . SGE have been reported for interactions between mothers and offspring ( maternal genotypes indirectly affect offspring phenotypes ) [3–8] and more recently for interactions between adult individuals , in livestock and wild animals[2 , 9–17] For example , growth rate in farm pigs has been found to be in part determined by the genetic makeup of the other pigs in the pen[2] . However , the extent to which SGE explain variation in biomedical traits is largely unknown . If SGE do contribute to such traits , they are a promising approach to quantify effects of the social environment . Additionally , they provide an anchor to investigate causal paths and dissect the mechanisms underlying social effects . Finally , in studies of direct genetic effects carried out by the broad community , SGE may be used to account for social environmental effects . Our study aimed at quantifying the contribution of SGE to multiple biomedical traits . We uncover unexpectedly large social genetic effects on multiple organismal and molecular phenotypes .
We first carried out an experiment with two inbred strains . We chose C57BL/6J ( B6 ) and DBA/2J ( D2 ) , the progenitor strains of the largest mouse reference population , the BxD recombinant inbred panel[18] . We co-housed 86 mice at weaning as B6/B6 , B6/D2 or D2/D2 pairs . After a period of six weeks during which the mice interacted undisturbed in their home cages , we collected 50 organismal phenotypes relevant to unconditioned anxiety , helplessness ( a measure of depressed mood ) , general locomotor activity , stress , social dominance , wound healing and body weight ( Fig 1A , S1 Table ) . We also profiled genome-wide gene expression in the prefrontal cortex ( PFC ) . The PFC was selected because it is involved in coordinating behavioural responses based on sensorimotor information , motivation and affect , all of which may be affected by the social environment; as a result individual differences in PFC expression levels may reflect behavioural responses to the social environment [19–21] . In this design with two inbred strains there are three potential genetic sources of phenotypic variation: differences in the strain of the focal ( i . e . phenotyped ) mice ( DGE ) , differences in the strain of their cage mates ( SGE ) and an interaction between the two , whereby the effect of the strain of the cage mate depends on the strain of the focal mouse ( Fig 1B , S1 Fig and Methods ) . Variance partitioning and model selection ( see Methods ) provided evidence that interactions between DGE and SGE are common across traits ( S2 Fig ) , with SGE typically affecting a specific trait in one strain ( B6 or D2 ) but not the other ( e . g . Fig 1B , S3 Fig ) . Thus , for each trait , we modeled the measurements collected in B6 mice and D2 focal mice separately ( S1 Fig ) . For eleven out of 50 organismal phenotypes we found significant SGE in either B6 or D2 mice ( P < 0 . 05 , FDR < 43%—see Methods; Table 1 , S2 and S3 Tables ) . Strain-specific differences existed as different measures were affected by SGE in B6 and D2 mice . Of these eleven phenotypes , two were measures of stress and six were measures of anxiety , providing evidence that variation in the genetic makeup of cage mates causes variation in stress-related phenotypes . The direction of effect was consistent across the six measures of anxiety ( S3 Fig ) . We also detected strong SGE on the rate of wound healing ( measured from an ear punch , P = 6 . 6 10−3 , Q = 0 . 36 ) , showing that SGE are not limited to behaviors . ( Table 1 ) . Importantly , SGE explained a considerable proportion of phenotypic variance ( up to 18% ) , showing that social effects of genetic origin are important contributors to phenotypic variation in these two strains . We also found evidence that SGE affect gene expression in the PFC . A gene set enrichment analysis based on the contribution of SGE to gene expression levels ( see Methods ) revealed “integrin-mediated signaling pathway” ( P = 2 . 5 10−6 , Q = 0 . 024 ) and “regulation of dopamine metabolic process” ( P = 7 . 0 10−4 , Q = 1 ) as the most significant Gene Ontology ( GO ) terms in D2 and B6 mice respectively ( Table 2 ) . The latter enrichment , although statistically less robust , is strikingly consistent with our finding that SGE affect dopamine levels measured by HPLC in B6 mice ( Table 1 ) . Our results therefore converge to show that variation in the genetic makeup of cage mates causes variation in behavioral , biochemical , and gene expression traits relevant to stress . Although the large number of tests limits the statistical power to detect SGE on individual genes ( 12 , 898 genes tested ) , we identified three genes with significant effects ( Srsf2 , Phlpp2 , and Ppid; FDR < 33%; S4 Table ) . We next investigated SGE in a dataset from outbred ( Heterogeneous Stock ) mice[22–24] . This design better represents traditional mouse housing conditions ( groups of four , five and six mice mostly rather than pairs; S4 Fig ) and genetic variation in natural populations ( high genetic diversity and genetically unique individuals rather than two inbred strains ) . The dataset comprises more than 100 organismal phenotypes measured in 2 , 448 mice ( S5 Table ) , and gene expression in hippocampus for a subset of 457 mice . To accommodate the genetic design of this study , we fitted random effects models with variance components for DGE , SGE and their covariance . Our models are inspired from models published in the literature[2 , 25] yet differ in some important ways , which we explain and justify in S1 Note . Because genome-wide genotypes were not available for a subset of mice ( 526 out of 2 , 448 ) , we used pedigree information to estimate pairwise genetic covariance for the mice with no genotype information ( see Methods ) . We found that the results obtained using both pedigree and genotype data for estimation of genetic similarity were in agreement with those obtained using genotype data only from the subset of mice that were in cages where all mice had been genotyped ( S5 Fig ) . All models were fitted using LIMIX[26 , 27] . Simulations showed that DGE and SGE were unbiased ( S6 Fig and S7 Fig ) . Of the 117 organismal phenotypes available in this dataset , 43 were significantly affected by SGE ( P < 0 . 05 , FDR < 5 . 7%; Table 3 , S5 Table ) . SGE explained up to 29% of phenotypic variation and an average of 8 . 9% across the 43 significantly affected traits . Importantly , the estimated contribution of SGE was greater than that of DGE for 8 of the 43 traits . Among the organismal phenotypes most significantly and strongly affected by SGE in this dataset were measures of lymphocyte activation ( in particular size and number of CD4+ T cells and B cells , collected by fluorescence-activated cell sorting and full blood count , Table 3 ) . In contrast , measures related to other leucocytes ( neutrophils , basophils and monocytes ) and natural killer T cells . Altogether , these results indicate that genotypes of cage mates influence humoral immunity , although this is unlikely that this represents spread of a disease as the mice were kept in a clean mouse facility . In addition , rate of wound healing , the measure most significantly and strongly associated with SGE in the experiment with two inbred strains , was also significantly affected by SGE in the outbred dataset ( P = 1 . 6 10−3 , Q = 3 . 7 10−3 ) . Three measures of body weight ( collected on weeks 6 and 7 ) also figured among the traits significantly affected by SGE . Other measures of body weight , collected at weeks 9 and 10 , showed no SGE ( S5 Table ) . The two sets of measures likely reflect a different phenotype as the normal physiology of the mice was disrupted between weeks 7 and 9 by aggressive phenotyping ( tests of conditioned anxiety involving foot shocks , airway sensitization by allergen , intraperitoneal injection of glucose ) . Another potential but unlikely explanation for the higher contribution of SGE on earlier body weight measures is a sharp decrease in social effects between weeks 7 and 9 . Finally , a subset of measures of anxiety , blood biochemistry , and lung function were also affected by the genotypes of cage mates . In contrast , there was no statistical evidence for SGE on gene expression levels in the hippocampus ( smallest nominal P value 4 . 1 10−5 corresponding to a Q value of 64% , S6 Table ) . This is most likely the result of a much smaller sample size ( 457 or 5% of mice ) for expression traits , and is consistent with published power analyses[28] . We next explored whether studies focused on DGE can safely ignore SGE , focusing on the estimation of the collective effect of additive DGE on phenotypic variation ( narrow-sense heritability ) . In the outbred dataset , cage mates are genetically more similar to each other than average ( S8 Fig ) . As a result , DGE and SGE are correlated ( an example of gene-environment correlation ) . Thus , we hypothesized that failing to account for SGE would bias estimates of DGE . We also hypothesized that fitting cage effects might be sufficient to eliminate this bias . To investigate both hypotheses , we compared DGE estimates obtained using a linear mixed model for DGE that does not account for cage effects nor SGE , a model that accounts for DGE and cage effects , and one that accounts for DGE , SGE , corresponding social environmental effects , and cage effects ( “full model” ) . Models that did not account for SGE led to substantially larger DGE estimates , and estimates from the model with DGE and cage effects were intermediate between those from the model with DGE only and those from the full model . In simulated traits based on the real genotypes and generated from DGE , SGE and cage effects ( see Methods ) , we found that models that did not account for SGE yielded inflated DGE estimates , whereas joint modeling of DGE , SGE and cage effects resulted in unbiased estimates ( Fig 2C and S6 Fig ) . Importantly , fitting cage effects but no SGE did not eliminate the bias . The simulation results strongly suggest that , in the real data , the estimates obtained from the full model are most accurate , and that models that ignore SGE overestimate heritability . This problem is particularly acute when direct and social random genetic effects are positively correlated ( i . e . σADS>0 , see Methods; Fig 2A and 2B ) . The problem we highlight here is general and likely to affect other studies in which social partners are related , including twin and family studies used to estimate heritability in humans ( see Discussion ) .
Using two complementary genetic designs–one using two mouse inbred strains and one using outbred mice—we estimated the contribution of social genetic effects to a variety of organismal phenotypes and gene expression traits . The experiment with two inbred strains was designed to investigate SGE and focused on behaviours ( anxiety and helplessness ) , as there is strong evidence that behaviours are socially affected[29–33] . To test whether SGE can be detected in outbred populations and survey a broader range of phenotypes , we re-analysed a large dataset from outbred mice and quantified the contribution of SGE to more than 100 phenotypes . The design of our study raises important questions: are positive results ( i . e . evidence of SGE ) in the experiment with two inbred strains expected to replicate in the outbred dataset ? Are some phenotypes expected to be affected by SGE and some not ? We now discuss these points . Some phenotypes were measured in both experiments with similar protocols . Wound area , the measure of wound healing , was collected in both experiments using the exact same protocol . It is significantly affected by SGE in both experiments . The protocols for measuring body and adrenal gland weight are fairly simple thus reducing technical variation between experiments , and body weight was measured at about the same time point ( around 50 days of age ) in both experiments . There was strong evidence for SGE on body weight in the outbred dataset but no evidence in the experiment with two inbred strains . No SGE on adrenal gland weight were detected in either dataset . Finally , a partially overlapping set of measures of unconditioned anxiety was significantly affected by SGE in both experiments . While reviewing results from the two experiments in parallel is informative , positive results in one experiment are not strictly expected to replicate in the other . Indeed , although the variants that give rise to SGE in the experiment with two inbred strains also segregate in the outbred population ( the two strains used in our experiment were among the eight founders of the outbred population ) , they have recombined with many additional variants from the six other founders . Moreover , the housing conditions were very different in the experiment with inbred strains and outbred experiment ( group size of 2 vs . 2 to 7 respectively , and unfamiliar mice vs . familiar mice housed together ) . Therefore , one should not expect the overall contribution of SGE be the same in the two experiments . Rather , combining the two experiments provides a first hint at the generalizability of our results . Published studies of SGE provide additional information on this matter , and suggest that SGE may contribute to variation in body weight across species[2] . That social genetic effects contribute to variation in anxiety probably does not come as a surprise but their contribution to wound healing maybe more so . This result is however supported by significant p-values in both experiments of our study and large effect sizes ( 18 and 6% ) . When interpreting this result , it is important to bear in mind that social effects on wound healing can ( and will , necessarily ) be mediated by traits of cage mates that are different from wound healing . For example , social effects on wound healing could be mediated by social grooming , which could either mechanically disrupt the healing process or chemically enhance it[34] . Any traits of cage mates that may induce a systemic stress response in the focal animal could also mediate social effects on wound healing[35 , 36] . Thus , social effects on wound healing are not unlikely , and , similarly , social effects may affect any phenotype ( e . g . by through the induction of a systemic stress response , ) . Because any phenotype may a priori be affected by social effects and the mechanisms at play are rarely known , SGE offer an attractive alternative to investigate social effects . First , as we have shown , they can be used to quantify social effects , effectively providing a lower bound estimate of social effects ( as only the genetic component is captured ) . Second , SGE can be used to test whether a particular trait of social partners has an effect on a phenotype of interest . Establishing a causal relationship between two phenotypes is always difficult because of the risk of reverse causation and independent action of hidden confounders on both traits; SGE provide an anchor to test causality . Independent of their relevance for studying social effects , we show that ignoring SGE can lead to biased estimates of heritability ( i . e . the collective effect of DGE ) . In our study ( outbred dataset ) , DGE and SGE are correlated by design ( mice that share a cage are more genetically similar than average ) , and we show that this correlation leads to biased estimates of heritability if unaccounted for . Fitting cage effects , which has the primary goal of accounting for environmental effects shared by cage mates ( e . g . noise levels ) , does not eliminate the bias . Our results are of interest to the broad genetics community as DGE and SGE are correlated in most if not all experimental designs traditionally used to estimate heritability in humans and model organisms , and SGE may thus have caused widespread bias . For example , in twin designs , MZ twins not only share 100% of their genotypes but they also share 100% of the genotypes of their sibling; DZ twins in comparison share both 50% of their genotypes and 50% of the genotypes of their sibling . Thus , SGE can contribute to increased concordance between MZ twins compared to DZ twins . If SGE are not modelled , heritability may be overestimated and appear “missing” when compared to genome-wide association results obtained from unrelated individuals[37] . Note that when the covariance σADS between direct and social random genetic effects is negative ( competition effects ) , ignoring SGE may lead to underestimating heritability . SGE in humans were considered once before ( “sibling effects” [38] ) but were never , to the best of our knowledge , modelled in heritability studies . Because we found that fitting cage effects was not sufficient to eliminate the bias due to SGE , we suspect that accounting for a “common environment” shared by family members , as is commonly done in human studies[39–41] , will not eliminate SGE-induced bias . It is not the first time that unaccounted for gene-environment correlations are put forward as potential causes of bias ( e . g . Conley et al . investigated the correlation between genetics and urban setting [42] ) . However , the impact of the correlation between DGE and SGE is likely to be particularly severe as we have shown that SGE affect a wide range of phenotypes and DGE and SGE are correlated in most experimental designs used to estimate heritability . Our study sheds light on an important component of the genetic architecture of complex traits , one that lies outside the individual , in social partners . Social genetic effects have already been shown to play an important role in artificial selection of livestock[43] and have important evolutionary consequences[44 , 45] . Our results provide evidence that SGE are also an important component of health and disease .
Gene expression data from the experiment with inbred strains are available from ArrayExpress E-MTAB-5276 . Phenotype data for the same experiment are provided as S7 Table . | Daily interactions between individuals can influence their health both in positive and negative ways . Often the mechanisms mediating social effects are unknown , so current approaches to study social effects are limited to a few phenotypes for which the mediating mechanisms are known a priori or suspected . Here we propose to leverage the fact that most traits are genetically controlled to investigate the influence of the social environment . To do so , we study associations between genotypes of one individual and phenotype of another individual ( social genetic effects , SGE , also called indirect genetic effects ) . Importantly , SGE can be studied even when the traits that mediate the influence of the social environment are not known . For the first time we quantified the contribution of SGE to more than 100 organismal phenotypes and genome-wide gene expression measured in laboratory mice . We find that genetic variation in cage mates ( i . e . SGE ) explains up to 29% of the variation in anxiety , wound healing , immune function , and body weight . Hence our study uncovers an unexpectedly large influence of the social environment . Additionally , we show that ignoring SGE can severely bias estimates of direct genetic effects ( effects of an individual’s genotypes on its own phenotype ) , which has important implications for the study of the genetic basis of complex traits . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Accession",
"numbers"
] | [
"body",
"weight",
"medicine",
"and",
"health",
"sciences",
"population",
"genetics",
"random",
"variables",
"covariance",
"physiological",
"processes",
"mathematics",
"physiological",
"parameters",
"experimental",
"organism",
"systems",
"mammalian",
"genomics",
"population"... | 2017 | Genetic Variation in the Social Environment Contributes to Health and Disease |
Kaposi's sarcoma associated herpesvirus ( KSHV ) is etiologically associated with endothelial Kaposi's sarcoma ( KS ) and B-cell proliferative primary effusion lymphoma ( PEL ) , common malignancies seen in immunocompromised HIV-1 infected patients . The progression of these cancers occurs by the proliferation of cells latently infected with KSHV , which is highly dependent on autocrine and paracrine factors secreted from the infected cells . Glutamate and glutamate receptors have emerged as key regulators of intracellular signaling pathways and cell proliferation . However , whether they play any role in the pathological changes associated with virus induced oncogenesis is not known . Here , we report the first systematic study of the role of glutamate and its metabotropic glutamate receptor 1 ( mGluR1 ) in KSHV infected cell proliferation . Our studies show increased glutamate secretion and glutaminase expression during de novo KSHV infection of endothelial cells as well as in KSHV latently infected endothelial and B-cells . Increased mGluR1 expression was detected in KSHV infected KS and PEL tissue sections . Increased c-Myc and glutaminase expression in the infected cells was mediated by KSHV latency associated nuclear antigen 1 ( LANA-1 ) . In addition , mGluR1 expression regulating host RE-1 silencing transcription factor/neuron restrictive silencer factor ( REST/NRSF ) was retained in the cytoplasm of infected cells . KSHV latent protein Kaposin A was also involved in the over expression of mGluR1 by interacting with REST in the cytoplasm of infected cells and by regulating the phosphorylation of REST and interaction with β-TRCP for ubiquitination . Colocalization of Kaposin A with REST was also observed in KS and PEL tissue samples . KSHV infected cell proliferation was significantly inhibited by glutamate release inhibitor and mGluR1 antagonists . These studies demonstrated that elevated glutamate secretion and mGluR1 expression play a role in KSHV induced cell proliferation and suggest that targeting glutamate and mGluR1 is an attractive therapeutic strategy to effectively control the KSHV associated malignancies .
Kaposi's sarcoma-associated herpesvirus or human herpesvirus-8 ( KSHV/HHV-8 ) infection is etiologically associated with Kaposi's sarcoma ( KS ) , a vascular endothelial tumor , and two B-cell lymphoproliferative diseases , primary effusion lymphoma ( PEL ) or body-cavity based lymphoma ( BCBL ) and multicentric Castleman's disease [1] , [2] , [3] . These cancers occur more frequently in the setting of immunosuppression , including HIV-1 infected patients , and develop from cells latently infected with KSHV . In vivo KSHV has a broad tropism and viral genome and transcripts are detected in a variety of cells such as B cells , endothelial cells , monocytes , keratinocytes , and epithelial cells [4] , [5] . Latent KSHV DNA is present in vascular endothelial and spindle cells of KS lesions , associated with expression of latency associated ORF73 ( LANA-1 ) , ORF72 ( v-cyclin D ) , K13 ( v-FLIP ) , and K12 ( Kaposin ) genes and microRNAs [5] . Cell lines with B cell characteristics , such as BC-1 , BC-3 , BCBL-1 , HBL-6 and JSC have been established from PEL tumors [4] , [5] . In PEL cells , in addition to the above set of latent genes , the K10 . 5 ( LANA-2 ) gene is also expressed [4] , [5] . About 1–3% of PEL cells spontaneously enter the lytic cycle and virus induced from these cells by chemicals serve as the source of virus . Multiple genome copies of both KSHV and EBV exist in latent form in BC-1 , HBL-6 and JSC cells while BCBL-1 and BC-3 cells carry only the KSHV genome [4] , [5] . KSHV infects a wide variety of human cell types in vitro , including fibroblasts , keratinocytes , B cells , endothelial , and epithelial cells [4] , [5] , [6] . Following infection , KSHV establishes latency within the target cells and the expression of the viral latent ORF71 , ORF72 , ORF73 , and K13 genes continues to maintain latency [7] . In addition , host genes required for regulating apoptosis , signal induction , cell cycle regulation , inflammatory response , and angiogenesis are also highly upregulated in the latently infected cells [8] . Studies have linked the expression of KSHV latency genes ORF71 ( v-FLIP ) , -72 ( v-Cyclin ) , -73 ( LANA-1 ) and K12 ( Kaposin A ) to the oncogenic activity of latently infected cells [9] , [10] , [11] . These genes induce the oncogenic potential of KSHV by increasing proliferative potential , growth and chromosome instability as well as by preventing apoptosis of the infected cells [9] , [10] , [11] . A hallmark of KSHV associated cancers is the excessive secretion of cytokines and growth factors [12] , [13] , [14] . Modulation by viral proteins and virally induced cellular proteins promote the secretion of autocrine and paracrine cytokines and growth factors leading into the proliferation , survival , and growth of the latently infected cells [12] , [13] , [15] , [16] , [17] , [18] . However , the mechanism behind KSHV induced cancer progression is not completely understood . Glutamate is a major excitatory neurotransmitter in the mammalian brain . It also plays a central role in several cellular functions , including cell survival and death by its interaction with receptors [19] . Glutamate released into the extracellular space binds and activates two classes of cell surface receptors , ionotropic ( iGluRs ) and G protein-coupled metabotropic glutamate receptors ( mGluRs ) . There are three groups of mGluRs , and group I mGluRs have been extensively studied in relation to cell survival and death . Group I consists of mGluR1 and mGluR5 subtypes ( mGluR1/5 ) which are coupled to Gαq/11 proteins . Agonist stimulation of group I mGluRs activate PLC , which results in the activation of PKC , PKC dependent pathways , and ERK1/2 [20] , [21] , [22] . The group I mGluRs are distributed in a variety of non-neuronal cells including human B and microvascular dermal endothelial cells , the natural target cells of KSHV [23] , [24] , [25] . However , the involvement of excess glutamate secretion and glutamate receptor expression in cell proliferation is an unexplored area of research in the KSHV oncogenesis field . This current study was undertaken with a rationale that identifying and defining the role of glutamate in KSHV biology will lead into targeted specific treatments for KSHV-associated malignancies . Our studies demonstrate that KSHV infection induces the secretion of glutamate and expression of mGluR1 receptor , and increased mGluR1 expression was detected in KS and PEL tissue sections . Most notably , glutamate secretion and mGluR1 activation in KSHV latently infected cells occurred through two independent pathways regulated by two individual viral latent proteins , LANA-1 and Kaposin A . Our data highlight how KSHV LANA-1 and Kaposin A proteins contribute to the generation of glutamate , activation of mGluR1 , and strongly suggest the possibility of exploiting the glutamatergic system for the therapeutic intervention of KSHV dependent cancers .
To determine the role of glutamate in KSHV infection , we first evaluated the secretion of glutamate during de novo KSHV infection of primary human microvascular dermal endothelial cells ( HMVEC-d ) . Kinetics of glutamate secretion showed that KSHV infection robustly increased glutamate release as early as 8 h post-infection ( p . i . ) which continued to increase throughout the 5 d p . i . observation period ( Figure 1A ) . In contrast , when the cells were infected with replication defective UV treated KSHV for 5d , there was no significant difference in glutamate secretion between uninfected and UV-KSHV infected cells , ( Fig . 1A ) suggesting that viral gene expression is required for the increased secretion of glutamate . To determine whether the secretion of glutamate is specifically induced by KSHV , cells were infected with KSHV pre-incubated with heparin ( Hep-KSHV ) , which is known to block the binding and entry of KSHV to the target cells [26] . In contrast to the untreated virus , heparin treated virus ( Hep-KSHV ) considerably reduced the secretion of glutamate ( Figure 1A ) . This suggested that KSHV entry and infection is required for the increased secretion of glutamate . KS is an endothelial tumor , whereas PEL is of B-cell origin [2] , [27] , [28] . The telomerase immortalized endothelial cell line ( TIVE ) latently infected with KSHV ( TIVE-LTC ) , and the PEL derived B-cell line BCBL-1 are well-established in vitro models to study KS and PEL , respectively [2] , [29] . In addition , BJAB-KSHV , a Burkitt's lymphoma B-cell line carrying latent KSHV DNA , has also been used as an additional model for studying KSHV pathogenesis [30] . To test whether the process of glutamate generation is relevant to KS , we measured the secretion of glutamate in KSHV TIVE-LTC cells as well as in uninfected control TIVE cells . Similar to de novo KSHV infection , higher levels of glutamate release were observed in KSHV ( + ) TIVE-LTC cells than in KSHV ( − ) TIVE cells ( Figure 1B ) . When the association of glutamate to PEL was assessed , high levels of glutamate release were observed in KSHV ( + ) BCBL-1 and BJAB-KSHV cells compared to the KSHV ( − ) B-cell line BJAB ( Figure 1C ) . To elucidate the mechanisms of glutamate generation in the infected cells , we next determined the expression of glutaminase , the major enzyme responsible for glutamate production [31] . Compared to the uninfected cells , a time dependent increase in glutaminase expression was observed during 8 h , 24 h , 48 h and 5 d of de novo infection of primary HMVEC-d cells by KSHV ( Figure 1D , lanes 1–6 ) . In contrast , at 5 d p . i . with UV-KSHV , no significant difference in glutaminase expression from uninfected cells was observed ( Figure 1D , lanes 1 and 7 ) . These results demonstrated that the increased glutamate secretion is linked with increased glutaminase expression in the infected cells . This link was further confirmed by the detection of a higher level of glutaminase expression in the latently infected TIVE-LTC ( 2 . 8 fold ) , BJAB-KSHV ( 2 . 5 fold ) , and BCBL-1 cells ( 3 . 4 fold ) than in their respective uninfected control TIVE and BJAB cells ( Figure 1E , lanes 1–5 ) . To further investigate the role of glutaminase in glutamate secretion , we used a glutaminase specific inhibitor , L-DON ( 6-diazo-5-oxo-norleucine ) [32] . Cells were treated with L-DON at a concentration of 500 µM and 1 mM and the supernatants were analyzed for glutamate release . We found that 500 µM of L-DON inhibited glutamate secretion by >50% and 1 mM of L-DON by >65% . Dose dependent inhibition of glutamate secretion in L-DON treated cells strongly suggested that glutaminase is the major enzyme that contributes to the generation of excess glutamate in KSHV infected cells ( Figure S1A ) . L-DON had no significant cytotoxicity on BJAB cells at 500 µM and 1 mM concentrations ( data not shown ) . KSHV latency-associated ORF73 gene product LANA-1 has been shown to induce c-Myc expression [33] . Since c-Myc has also been shown to activate the expression of glutaminase [34] , we hypothesized that the increased glutamate secretion observed in KSHV infected cells could be mediated by LANA-1 through its c-Myc activation , which in turn stimulates the expression of glutaminase . To test this hypothesis , when BJAB cells were transduced with lentivirus constructs of LANA-1 , we observed increased secretion of glutamate in LANA-1 transduced cells compared to vector alone ( Figure 2A ) . We also observed ∼2-fold increase in c-Myc and glutaminase protein expression in LANA-1 transduced cells ( Figure 2B ) . To support our finding that LANA-1 mediated c-Myc activation is directly involved in glutaminase expression and glutamate secretion , we used lentiviruses encoding shRNAs to knock down c-Myc in BJAB cells over expressing LANA-1 . As shown in figure S1B , LANA-1 over expression induced the secretion of glutamate , and this induction was abolished by the knockdown of c-Myc ( Figure S1B ) . Since no considerable increase in glutamate release was observed in the absence of c-Myc in LANA-1 expressing cells ( Figure S1B ) , these results suggested that LANA-1 mediated c-Myc activation is required for glutamate release . To confirm the functional relationship of c-Myc expression with glutaminase expression in KSHV infected cells , we transduced TIVE-LTC cells and BCBL-1 cells with c-Myc and control shRNA lentiviral vectors . A significant reduction in glutaminase expression was observed in c-Myc knockdown TIVE-LTC cells ( 61% ) and BCBL-1 cells ( 67% ) compared to control shRNA transduced cells ( Figures 2C and D ) . These results suggested that LANA-1 mediated c-Myc activation plays a crucial role in the expression of glutaminase and glutamate secretion in cells latently infected with KSHV . Among the several types of glutamate receptors , mGluR1 is considered an oncogenic protein due to its ability to regulate the functions related to cancer cell proliferation [35] , [36] . Hence , we theorized that the biological effect of glutamate in latent KSHV induced oncogenesis may be mediated through the expression of mGluR1 receptors . To test this , we first determined mGluR1 expression by RT-PCR in primary endothelial cells infected for 5 d with KSHV and UV-KSHV . Compared to uninfected cells , KSHV infection increased the expression of mGluR1 ( Figure 3A ) . In contrast , UV treated virus had no significant effect on mGluR1 expression ( Figure 3A ) and suggested that sustained mGluR1 receptor expression probably depended upon KSHV gene expression . When the relative expression levels for the mGluR1 receptor in KSHV latent TIVE-LTC , BJAB-KSHV and BCBL-1 cells as well as control BJAB and TIVE cells were determined by RT-PCR , upregulation of mGluR1 in both KSHV ( + ) TIVE-LTC cells and BCBL-1 cells was observed compared to uninfected TIVE and BJAB cells ( Figure 3B ) . Western blot ( Figure 3C ) and immunoprecipitation analysis ( Figure S2A ) confirmed the higher levels of mGluR1 protein in de novo KSHV infected primary cells compared to the uninfected and UV-KSHV infected cells . Similarly , high levels of mGluR1 expression were also observed in BJAB-KSHV , BCBL-1 and TIVE-LTC cell lines by Western blots ( Figure 3D ) and by immunoprecipitation analysis ( Figure S2B ) . The expression of mGluR1 in KSHV infected primary cells and latent cells was also examined by immunofluorescence assay ( IFA ) . Increased mGluR1 staining was detected in LANA-1 expressing spindle shaped HMVEC-d cells infected with KSHV ( Figure 3E ) , as well as in TIVE-LTC , BJAB-KSHV and BCBL-1 cells compared to their respective uninfected controls ( Figures 3F and G ) . These results clearly demonstrated that KSHV infection results in the increased mGluR1 expression in latently infected cells . To verify the pathological association of mGluR1 in KSHV associated cancers , we immunostained normal as well as KSHV infected KS and PEL tissues by dual labeled IFA for mGluR1 and KSHV LANA-1 as a marker for infection . Strong positive immunostaining for both mGluR1 and LANA-1 were detected in the spindle shaped endothelial cells of KS tissue ( Figure 4A ) , and in the stomach PEL ( Figure 4B ) samples . In contrast , only a basal level of mGluR1 was detected in control normal skin and stomach samples ( Figures 4A and B ) . These results clearly demonstrated the in vivo association of increased mGluR1 expression with KSHV infection . The expression of mGluR1 in non-neuronal cells is regulated by RE-1 silencing transcription factor/neuron restrictive silencer factor ( REST/NRSF ) [37] . Binding of REST to a DNA recognition sequence called the neuron restrictive silencer elements ( NRSE or RE-1 ) repress the expression of neuronal genes such as mGluR1 in non-neuronal cells [37] , [38] , [39] . To analyze whether REST expression plays any role in mGluR1 expression in KSHV infected cells , we determined the expression of REST mRNA and protein . real-time RT-PCR analysis of REST revealed similar levels of REST expression in both uninfected and KSHV infected latent cells ( Figures S3A and B ) . However , REST protein expression determination by Western blots showed 55% , 72% and 42% reduction in BJAB-KSHV , BCBL-1 and TIVE-LTC cells , respectively , compared to the respective controls ( Figures S3C and D ) . This suggested that REST expression in KSHV infected cells is probably modulated at the post-transcriptional level . To decipher the mechanism regulating REST expression at the post-translational level , we first determined the subcellular localization of REST in TIVE and TIVE-LTC cells by IFA . In the uninfected TIVE cells , REST was highly expressed and was predominantly localized in the nucleus ( Figure 5A ) . In contrast , REST distribution was markedly decreased in the nucleus of TIVE-LTC cells and was predominantly localized to the cytoplasm ( Figure 5A ) . A similar cytoplasmic relocalization of REST was observed in almost all KSHV infected BJAB-KSHV and BCBL-1 cells compared to the uninfected BJAB cells where it was exclusively localized in the nucleus ( Figure 5B ) . Western blot analysis of cytoplasmic and nuclear fractions of the KSHV positive cell lines confirmed that REST localization is significantly decreased in the nucleus ( Figures 5C and D ) with a concomitant increase in the cytoplasm of infected cells , whereas it was undetectable in the cytoplasm of uninfected TIVE and BJAB cells ( Figures 5C and D , lane 1 ) . Interestingly , analyses of REST in the cytoplasmic fractions from the infected cells showed a small shift in molecular weight in both TIVE-LTC ( Figure 5C , lane 2 ) and BCBL-1 cells ( Figure 5D , lane 3 ) . We reasoned that this small shift in band size could be due to phosphorylation of REST , which is known to result in migration differences on SDS-PAGE . To determine whether the shifted band detected in the cytoplasm is indeed the phosphorylated form of REST , we first treated the cytoplasmic extracts from TIVE-LTC cells with lambda phosphatase or with lambda phosphatase and phosphatase inhibitor , and then the extracts were Western blotted . Treatment with lambda phosphatase resulted in the disappearance of the modified band , suggesting that the shift in band size was due to phosphorylation ( Figure S4A ) . It has been reported that serine phosphorylation of REST in the conserved phosphodegron motif promotes recognition by the E3 ubiquitin ligase β-TRCP and ubiquitination [40] , [41] . As our data suggested an unexpected decrease of REST in the infected cells , we next asked whether the phosphorylation of REST in the cytoplasm was followed by its phosphorylation-dependent ubiquitination . To examine this regulatory role , we first verified the serine phosphorylation of REST in the cytoplasm of infected cells by immunoprecipitating with phosphoserine antibody and Western blotting with REST antibody . Consistent with the Western blot results ( Figures 5C and D ) , a significant level of serine phosphorylation of REST was detected in KSHV-infected TIVE-LTC , BJAB-KSHV and BCBL-1 cells compared to a very low level of phosphorylation in KSHV-negative TIVE and BJAB cells ( Figure 6A ) . We next determined whether β-TRCP could be associated with phosphorylated REST in the cytoplasm of infected cells . Immunoprecipitation of cytoplasmic extracts of BJAB , BJAB-KSHV and BCBL-1 cells with REST and Western blots with anti-β-TRCP antibodies showed increased interaction of REST with β-TRCP in the infected cells , whereas it was barely detectable in uninfected cells ( Figure 6B ) . We next determined whether REST degradation occurs in the cytoplasm of KSHV-infected cells . Analysis of cytosolic fractions from TIVE-LTC , BJAB-KSHV and BCBL-1 cells by immunoprecipitation with REST and Western blots for polyubiquitin revealed a higher level of ubiquitination in TIVE-LTC , BJAB-KSHV and BCBL-1 cells compared with TIVE and BJAB cells displaying lower levels of ubiquitination ( Figure 6C ) . Thus , the ubiquitination levels of REST correlated with REST phosphorylation and the association of REST with β-TRCP in the cytoplasm . In order to confirm that the ubiquitin proteasome system is involved in the degradation of REST in KSHV infected cells , BCBL-1 and TIVE-LTC cells were treated with the proteasome inhibitor MG132 , and the cell lysates were Western blotted for REST . As shown in figure 6D , compared to the untreated cells ( lanes 3 , 5 , and 7 ) , MG132 treatment increased the protein level of REST in the infected BJAB-KSHV , BCBL-1 , and TIVE-LTC cells ( lanes 4 , 6 , and 8 ) . However , MG132 treatment had no significant effect on the REST protein level in uninfected BJAB cells ( lanes 1 and 2 ) . This result further supported our finding that the degradation of REST observed in the infected cells was probably mediated by the ubiquitin proteasome pathway . Since REST was more localized in the cytoplasm of latently infected cells , we hypothesized that latent KSHV protein ( s ) in the infected cells binds and sequesters REST in the cytoplasm , which in turn leads to overexpression of the mGluR1 gene . To determine the identity of the KSHV latent protein responsible for this , BJAB cells were transduced with the lentiviral constructs of KSHV latent ORF71 , -72 , -73 , and Kaposin A genes , expression levels assessed by real-time PCR ( Figure S4B ) , and mGluR1 level analyzed by Western blot . ORF K12 or Kaposin A transduction led to a robust increase in mGluR1 expression in BJAB cells , indicating the involvement of Kaposin A in the regulation of mGluR1 expression , whereas the other latent genes did not significantly induce the expression of mGluR1 ( Figure 7A ) . mGluR1 expression in Kaposin A transduced BJAB cells was further confirmed by immunoprecipitation experiments ( Figure S4C ) . Transduction efficiencies were determined by control lentiviral GFP expression ( Figure S4D ) . We also observed higher levels of mGluR1 protein expression in primary HMVEC-d cells transduced with Kaposin A which further demonstrated the Kaposin A dependency of mGluR1 expression ( Figure 7B ) . To determine whether Kaposin A is responsible for the observed cytoplasmic relocalization of REST in the infected cells , we transduced HMVEC-d cells with a lentiviral Kaposin A construct ( ORF K12 ) and localization was determined by IFA using anti-Kaposin A antibodies . This analysis revealed that a major portion of endogenous REST was translocated into the cytoplasm and colocalized with Kaposin A in the transduced cells ( Figure 7C ) . To verify that REST binds to Kaposin A in the cytoplasm of KSHV infected cells , we immunoprecipitated REST from cytoplasmic fractions of both uninfected BJAB and KSHV-infected BCBL-1 cells and then Western blotted with anti-Kaposin A antibodies , which detected specific bands of Kaposin A at approximately 16–18 kDa ( Figure 8A ) . BCBL-1 cell lysates used as positive control also identified 16–18-kDa immunoreactive bands of Kaposin A in the infected cells ( Figure 8A ) . The predicted molecular weight of Kaposin A is 6-kDa; however , WB analyses often detect specific bands of about 16–18-kDa and above [42] , [43] , [44] . Similar immunoprecipitation analysis using TIVE and TIVE-LTC cells also revealed that REST interacts with Kaposin A in the infected cell cytoplasm ( Figure 8B ) . We also observed the colocalization of REST and Kaposin A in the cytoplasm of TIVE-LTC ( Figure 8C ) and BCBL-1 cells ( Figure 8D ) . To further verify the physical interaction of REST with Kaposin A , we co-transduced 293T cells with Kaposin A and retroviral FLAG tagged REST and the cytoplasmic and nuclear lysates were immunoprecipitated with Kaposin A and Western blotted with anti-FLAG antibodies . This co-immunoprecipitation experiment demonstrated the ability of Kaposin A to interact with REST in the cytoplasm , but not in the nucleus ( Figure 8E ) . We next examined the staining pattern and colocalization of REST and Kaposin A in KS and PEL patient samples and in normal tissues by immunofluorescence analysis . As shown in Figures 9A and B , strong nuclear staining of REST was observed in normal skin tissues as well as in normal stomach tissues . In contrast , cytoplasmic localization of REST and notable colocalization with Kaposin A were observed in the endothelial cells of KS as well as in the cells of PEL tissues , presumably the B cells . Together , these results suggested that Kaposin A expression regulates mGluR1 expression through interaction with REST in the cytoplasm of KSHV infected cells . As shown in figure 6B and C , phosphorylated REST interacts with β-TRCP and promotes the ubiquitination and degradation of REST in the cytoplasm of infected cells . It has previously been reported that REST has a degron motif and the phosphorylation of REST at serine 1024 , 1027 , and 1030 of the degron motif is required for the interaction of REST with β-TRCP during oncogenic transformation [41] . Because Kaposin A is a protein involved in transformation of infected cells [45] , [46] , we postulated that Kaposin A binding with REST phosphorylates REST at the 1024 , 1027 , and 1030 residues , leading to the interaction with β-TRCP and ubiquitination of REST . To investigate this , we co-transduced 293T cells with Kaposin A and FLAG REST-WT or FLAG-REST triple mutant ( where all three phosphodegron residues are mutated-FLAG-REST-S1024/1027/1030A ) , cytoplasmic fractions immunoprecipitated with anti-phosphoserine antibodies and Western blotted with anti-FLAG antibodies . A significant level of serine phosphorylated REST was detected in Kaposin A and FLAG REST-WT transduced cells ( Figure 10A , upper panel ) . In contrast , the serine phosphorylation of REST was severely impaired in Kaposin A and FLAG-REST triple mutant transduced cells suggesting that Kaposin A mediates REST phosphorylation in its phosphodegron sites . As Kaposin A is involved in REST phosphorylation in the conserved degron sites , we next determined whether phosphorylated REST binds to endogenous β-TRCP . As shown in Figure 10A , middle panel , immunoprecipitation with FLAG and Western blot with β-TRCP showed a markedly increased interaction of REST with endogenous β-TRCP in Kaposin A and FLAG REST-WT transduced cells , whereas no interaction was observed in Kaposin A and FLAG-REST triple mutant transduced cells . These data demonstrated that blocking Kaposin A mediated phosphorylation of REST weakens its association with endogenous β-TRCP . To further investigate which specific degron site is phosphorylated by Kaposin A , we transiently transduced 293T cells with lentiviral Kaposin A and 48 h after transduction , the cells were transfected with pCMV-FLAG-REST WT plasmid or pCMV-FLAG-REST individually mutated at serine 1024 , ( pCMV-FLAG-REST-S1024A ) , 1027 ( pCMV-FLAG-REST-S1027A ) , or 1030 ( pCMV-FLAG-REST-S1030A ) . Cell lysates were immunoprecipitated using anti-phosphoserine antibodies followed by Western blotting with anti-FLAG antibodies or vice versa . As shown in Figure 10B , the serine 1027 mutant completely abolished the capacity for phosphorylation ( Figure 10B , lane 5 , first and second panel ) . However , the serine 1024 and 1030 mutants had no effect on phosphorylation compared to wild type REST ( Figure 10B , lane 4 and 6 , first and second panel ) , indicating that the phosphorylation of these two sites are not directly mediated by Kaposin A . The defect in REST phosphorylation in mutant serine 1027 suggested that Kaposin A initially phosphorylates REST on serine 1027 . Phosphorylation on 1027 may provide the signal to phosphorylate the other degron residues . In order to determine whether the serine 1027 induced phosphorylation is responsible for REST ubiquitination , the cell lysates immunoprecipitated with anti-FLAG antibody were analyzed by Western blotting with an anti-polyubiquitin antibody . Consistent with the increased phosphorylation , the ubiquitination was markedly increased in wild type REST , as well as in serine 1024 and 1030 mutants transfected cells ( Figure 10B , lanes 3 , 4 , and 6 , third panel ) . In contrast , the phosphorylation defective mutant 1027 failed to induce ubiquitination ( Figure 10B , lane 5 , third panel ) . These results suggest that serine 1027 mediated phosphorylation is required for the ubiquitination of REST . We also observed that the phosphorylation defective mutant FLAG-REST-S1027A stabilized REST ( Figure 10B , lane 5 , fourth panel ) . The observed reduction of REST in FLAG-REST WT and FLAG-REST-S1024A and -S1030A ( Figure 10B , lanes 3 , 4 and 6 , fourth panel ) , after Kaposin A stimulation may be due to the degradation of phosphorylated REST at the 1027 residue . Taken together , our studies demonstrated that Kaposin A regulates REST phosphorylation in the conserved phosphodegron motif which enhances the ubiquitination of REST and thus reduces the level of REST . To further verify the role of Kaposin A in REST phosphorylation , 293T cells transduced with vector alone or Kaposin A were transfected with FLAG-REST S1027 or FLAG-REST WT first and then transfected with control or Kaposin A specific siRNA . After 48 h post transfection , levels of REST phosphorylation were assessed by immunoprecipitating with anti-FLAG antibody followed by Western blotting with anti-phosphoserine antibody . We observed that compared to control siRNA transfected cells , Kaposin A siRNA transfected cells abolished the phosphorylation and degradation of REST in REST WT transfected cells ( Figure 10C , lane 4 and 5 , first and second panel ) . Kaposin A specific siRNA efficiently knocked down the expression of Kaposin A in the transduced cells ( Figure 10C , lane 5 , third panel ) . As expected , cells transfected with phosphorylation defective mutant REST-S1027 had no effect on phosphorylation ( Figure 10C , lane 3 , first panel ) . These data confirm that Kaposin A is essential for the phosphorylation of REST . We next focused on the biological response of glutamate release and binding to its receptors . We postulated that the glutamate released by infected cells binds to mGluR1 permitting cellular signaling and the proliferation of glutamate secreting infected cells . To determine the effects of glutamate and mGluR1 on cell proliferation , primary HMVEC-d cells infected with KSHV for 3 d were cultured for 2 d in the presence or absence of glutamate release inhibitor riluzole , and mGluR1 antagonists A841720 and Bay 36-7620 , pulsed with BrdU for 2 h and BrdU incorporation determined by IFA . As shown in Figure 11A , HMVEC-d cells infected with KSHV for 5 d showed a much higher rate of proliferation than the uninfected cells . This increased proliferation of HMVEC-d cells was significantly reduced by exposure to riluzole , A841720 and Bay 36-7620 ( Figure 11A ) . These results were also confirmed by BrdU cell proliferation ELISA ( Figures S5A and B ) . We further tested the involvement of riluzole , A841720 , and Bay 36-7620 in TIVE and TIVE-LTC , BJAB and BCBL-1 cell proliferation by BrdU cell proliferation ELISA . No treatment and vehicle treatment were used as controls . As shown in Figure 11B , treatment with riluzole , A841720 and Bay 36-7620 showed a concentration dependent decrease in the proliferation of both TIVE-LTC and BCBL-1 cells ( Figures 11B and C ) . Due to the absence or low level of expression of mGluR1 receptors , only a minimal effect was observed in the proliferation of uninfected TIVE and BJAB cells ( Figures S5C and D ) . To further confirm the effect of inhibitors on cell proliferation , cells treated with riluzole , A841720 or Bay 36-7620 were monitored using a vibrant MTT cell proliferation assay kit . Riluzole , A841720 , and Bay 36-7620 caused a 60–70% decrease in cell growth compared to the untreated control ( Fig . 11 D and E ) . Next , we confirmed the role of mGluR1 on the proliferation of infected cells by using mGluR1 shRNA . TIVE- LTC cells transduced with mGluR1 shRNA or control shRNA were assayed for BrdU incorporation . Compared to control shRNA cells , mGluR1-shRNA significantly reduced the proliferation of TIVE-LTC cells ( Fig . 11G ) , indicating that mGluR1 plays a key role in the proliferation of KSHV infected cells . Collectively , these results suggested that riluzole and mGluR1 antagonists suppressed the binding of glutamate to the receptors of infected cells and thereby arresting the activation of receptors by glutamate leading into the proliferation of KSHV infected cells .
Glutamate release along with autocrine and paracrine glutamate receptor signaling has been demonstrated to accelerate cell proliferation and tumor progression [47] , [48] . During the latent phase of KSHV infection , the cytokines and growth factors released into the extracellular milieu play significant roles in the long term proliferation , survival , and maintenance of the infected cells which probably results in KSHV associated malignancies [8] , [12] , [13] , [15] , [16] , [17] , [18] , [49] . Our comprehensive studies demonstrating the increased secretion of glutamate into the cytokine milieu in response to KSHV infection suggest that glutamate could be acting as an autocrine and paracrine growth factor during KSHV induced oncogenesis . Secretion of glutamate occurs in uninfected and infected cells , with comparatively low levels in uninfected cells . We have demonstrated that KSHV infection and appropriate viral gene expression are critical for the generation and release of glutamate in the infected cells . As the viral genome persists in a latent state in the infected cells , the expression level of the latent genes may affect glutamate secretion . Our current study clearly suggested a mechanism whereby the latent ORF73 gene expression affect the stability of c-Myc activation and the depletion of which resulted in reduced glutaminase expression and glutamate secretion . This implies that the level of infection and consistent expression of viral genes are required for the continued secretion of glutamate . Our studies also show that KSHV infected cells induced the highest levels of glutaminase expression and caused a moderate increase in glutamate release . This difference could be attributable to glutamate transporters and the uptake of glutamate into cells . The glutamate taken up by the cells is converted into glutamine via the glutamine synthetase pathway [50] . Since there are several evidences to indicate that glutamate uptake and its enzymatic conversion are significant steps to maintain extracellular glutamate concentration [50] , [51] , [52] , it is possible that expression or functional impairment of glutamate transporters may also be involved in the maintenance of extracellular glutamate levels in the infected cells . c-Myc has numerous significant effects on cancer cell metabolism by modifying expression of proteins involved in metabolic pathways [53] . It is known to stimulate increased expression of its target proteins and glutaminase expression by transcriptional repression of mir23a/b in cancer cells [34] . The increased c-Myc activity may also significantly alter the metabolism of glutamine in the infected cells . These changes in glutamine metabolism may profoundly influence the synthesis of molecules involved in growth and survival of infected cells . Although increased glutaminolysis is a supplementary source of energy and may provide significant benefits in terms of the survival of the infected cells , they require additional factors for the induction of cell proliferation or transformation . Thus , while the role of increased metabolism and the components involved in metabolism remains to be determined , it is clear from our study that the secreted glutamate is being used to activate mGluR1 which contribute to the proliferation of infected cells . Interestingly , we report that KSHV infected cells also upregulate the expression of glutamate receptor mGluR1 , which in turn results in increased proliferation as a result of glutamate binding to mGluR1 in the infected cells . Enhanced expression of mGluR1 , and the intracellular signaling pathway activated by mGluR1 , has the ability to induce cell proliferation and oncogenic transformation [35] , [36] . Our data provide evidence that mGluR1 is upregulated in in vitro latently infected cells and in vivo patient samples . Mechanistically , mGluR1 overexpression involves relocalization of REST from the nucleus to the cytoplasm and loss of REST expression in the infected cells . Decreased REST expression , relocalization of REST , and degradation of REST are possible adaptations to antagonize REST-mediated effects to accomplish the overexpression of mGluR1 [39] , [54] , [55] . A remarkable difference in the pattern of REST localization observed in the infected cells indicates that mGluR1 expression may be regulated via the relocalization of REST . Translocation of REST to the cytoplasm relieves the NRSE or RE1 mediated transcriptional repression in the promoter regions of mGluR1 and upregulates its transcription ( Figure 12 ) . Another one of our major findings is that the KSHV latent protein Kaposin A is responsible for cytoplasmic relocalization of REST and mGluR1 activation ( Figure 12 ) . Kaposin A mediated oncogenesis has been demonstrated in vitro in Rat3 fibroblasts and in nude mice [45] , [46] . Previous studies have suggested that Kaposin A regulates oncogenesis by influencing the phosphorylation of signaling molecules involved in cellular processes , such as cell proliferation and gene transcription [42] , [46] . Our findings suggest that sequestration of REST in the cytoplasm by Kaposin A modulates phosphorylation-dependent ubiquitination of REST by altering the phosphorylation status of REST ( Figure 12 ) . Kaposin A regulates REST phosphorylation at the specific degron sites which are essential for binding to β-TRCP and degradation of REST during oncogenic transformation . Thus , the downregulation of REST , which is seen in actively proliferating cancer cells [56] , [57] , might be involved in the regulation of mGluR1 and in cellular transformation during KSHV induced cancer development . Since Kaposin A does not have a known protein kinase domain , how Kaposin A binding to REST induces phosphorylation of REST needs to be elucidated . Several mechanisms are possible to account for the phosphorylation of REST by Kaposin A . Kaposin A has been reported to phosphorylate a number of kinases involved in cell proliferation [46] . Therefore , it is possible that Kaposin A may couple through one of these kinases for the activation of REST and recruitment of β-TRCP . It is also possible that the interaction of Kaposin A with REST may induce the phosphorylation of REST by allowing a conformational change . These modifications would create a favorable molecular environment for the cross talk between REST and β-TRCP . In addition to Kaposin A mediated mGluR1 expression , we observed that over expression of LANA-1 also lead to a slight increase in mGluR1 expression . This increase in mGluR1 expression may be a result of an alteration in the N-terminal repressor domain of REST on the mGluR1 promoter . The N-terminal repression domain of REST represses target gene expression by recruiting the transcriptional corepressor mSin3 and then forming a complex with histone deacetylase ( HDAC ) [58] , [59] . Since LANA-1 has already been shown to be associated with mSin3 co-repressor as well as with HDAC [60] , it is expected that LANA-1 may be able to bypass REST mediated repression by sequestration of the mSin3/HDAC complex which results in the expression of mGluR1 genes . It is also known that mSin3/HDAC regulated repression is not sufficient for complete transcriptional repression of REST target genes [58] , [59] , [61] . Therefore , the slight induction of mGluR1 expression in LANA-1 expressing cells could be due to the partial derepression of REST target genes by LANA-1 . Glutamate receptor antagonists and glutamate release inhibitors were shown to be effective in suppressing the proliferation of non-neuronal cancer cells [62] , [63] . Identification of the activity of glutamate and mGluR1 in glioma and melanoma development has been the rational approach for testing glutamate release inhibitors talampanel and riluzole in clinical trials for the treatment of glioma and melanoma , respectively [64] , [65] . Our functional data shows that the increased production of glutamate and expression of mGluR1 in response to KSHV infection promotes the proliferation of infected cells . Several studies have demonstrated that the signaling pathways activated by mGluR1 contribute to the proliferation and survival of cancer cells [22] , [66] . Further studies are essential to determine the role of glutamate and mGluR1 activity in signal induction , viral gene expression , and viral genome maintenance in cells latently infected with KSHV . The blocking effect of riluzole , and the mGluR1 antagonists on proliferation of KSHV infected cells suggests that these molecules could potentially be used for the treatment of KSHV associated malignancies by directly targeting the glutamatergic system in the infected cells .
Primary human dermal microvascular endothelial cells ( HMVEC-d cells CC-2543 ) were purchased from Clonetics , Walkersville , MD . KSHV negative B-lymphoma cell line BJAB , and the KSHV latently infected B-cell line BCBL-1 , were obtained from ATCC . BJAB-KSHV ( KSHV–GFP recombinant virus in BJAB ) was a gift from Dr . Blossom Damania ( University of North Carolina , Chapel Hill ) . TIVE ( telomerase-immortalized vein endothelial cell line ) and TIVE LTC cells ( TIVE cells carrying KSHV in a latent state ) were a gift from Dr . Rolf Renne ( University of Florida ) . These cell lines were maintained as described previously [67] . Induction of the KSHV lytic cycle with TPA in BCBL-1 cells , and KSHV purification procedures have been previously described [68] . UV-treated replication-defective KSHV was prepared by exposing the purified virus stock to UV light ( 365 nm ) for 20 min at a 10-cm distance . KSHV DNA was extracted from live KSHV and UV-treated KSHV , and the copies were quantitated by real-time DNA PCR using primers amplifying the KSHV ORF73 gene as described previously [7] . Unless stated otherwise , primary cells were infected with KSHV at 50 MOI ( multiplicity of infection ) per cell at 37°C . Rabbit anti-mGluR1 and β-TRCP antibodies as well as mouse anti-BrdU antibodies were from Cell Signaling , Beverly , MA . Mouse anti-glutaminase and rabbit anti-mGluR1 and -TATA binding protein ( TBP ) antibodies were from Abcam , Cambridge , MA . Mouse anti-tubulin and β-actin antibodies were from Sigma , St . Louis , MA . Mouse anti-c-Myc ( 9E10 ) and REST antibodies were from Santa Cruz , Santa Cruz , CA . Rat anti-Kaposin A/C and mouse anti-polyubiquitin antibodies were from Millipore , Temecula , CA . Mouse anti-ORF73 antibodies were generated in Dr . Chandran's laboratory . Anti-rabbit and anti-mouse antibodies linked to horseradish peroxidase were from KPL Inc . , Gaithersburg , Md . Alexa 488 and 594 conjugated secondary antibodies were from Invitrogen . Protein A and G–Sepharose CL-4B beads were from Amersham Pharmacia Biotech , Piscataway , NJ . Lambda phosphatase ( λPPase ) , and L-DON ( 6-diazo-5-oxo-norleucine ) were from Santa Cruz . Riluzole , A841720 and Bay 36-7620 were from Tocris Bioscience , Minneapolis , MN . Plasmids encoding FLAG-tagged human REST wild-type and site-specific REST mutant plasmids ( pCMV-FLAG-REST-S1024A , pCMV-FLAG-REST-S1027A , pCMV-FLAG-REST-S1030A ) , wild type FLAG-REST and triple mutant FLAG-REST-S1024/1027/1030A cloned into retroviral vector pQCXIN were provided by Dr . Stephen Elledge [41] ( Harvard Medical School ) . Lentiviral constructs of KSHV ORF71 ( vFLIP ) , ORF72 ( vCyclinD ) , ORF73 ( LANA-1 ) and ORFK12 ( Kaposin A ) were obtained from Dr . Chris Boshoff at the UCL Cancer Institute [69] . A plasmid encoding c-Myc shRNA sequence ( plasmid #29435 ) was from Addgene . Transfection was performed using 5 µg of plasmid DNA and lipofectamine 2000 ( Invitrogen ) as per the manufacturer's instructions . Lentivirus was produced by transfection with a four-plasmid system , as previously described [70] . Briefly , 293T cells were transiently transfected with lentiviral constructs and the plasmid packaging system ( Gag-Pol , Rev and VSV-G ) , the supernatants were collected , and filtered . Infections were carried out by incubating the virus preparation with cells in the presence of polybrene . The infection efficiency was estimated by analyzing GFP-expressing lentiviral vectors as positive controls . The expression levels of transduced viral genes were assessed by real-time PCR . For mGluR1 knockdown , lentiviruses encoding mGluR1 shRNA or control shRNA were purchased from Santa Cruz Biotechnology . TIVE-LTC cells were transduced with control lentivirus shRNA and mGluR1 lentivirus shRNA according to the manufacturer's instructions and selected by puromycin hydrochloride . An equal number of uninfected and infected cells were used for the experiments . Supernatants harvested at different times were centrifuged and glutamate levels were determined in 96-well plates by using a glutamate assay kit as per the manufacturer ( Biovision , Mountain View , CA ) . The concentration of glutamate was determined by measuring the absorbance at 450 nm with a microplate reader . Total RNA was isolated with TRIzol Reagent ( Invitrogen ) and treated with DNase I ( Ambion ) at 37°C for 30 min . Reverse transcription was performed using a High-Capacity cDNA reverse transcription kit ( Applied Biosystems ) . Regular PCR for mGluR1 was performed using 5 µl of the synthesized cDNA using appropriate forward and reverse primers as described by Choi et al [71] . PCR primers were as follows: mGluR1 5′-GTGGTTTGATGAGAAAGGAG-3′ ( forward ) and 5′-GTTGCTCCACTCAAGATAGC-3 ( reverse ) . β-actin 5′-GCTCACCATGGATGATGATATCGCC-3′ ( forward ) and 5′GGATGCCTCTCTTGCTCTGGGCCTC-3′ ( reverse ) . Quantitative real time-PCR was performed with SYBR Green and an ABI prism 7000 sequence detection system ( Applied Biosystems , Foster City , CA ) . The comparative Ct method was used to quantitate gene expression relative to the uninfected control . The following primer set was used: REST ( forward 5′-GAGGAGGAGGGCTGTTTACC-3′; reverse 5′-TCACAGCAGCTGCCATTTAC-3′ ) . Primers used for qRT-PCR of viral genes: ORF71 ( forward 5′-AGGTTAACGTTTCCCCTGTTAGC-3′; reverse , 5′-AGCAGGTCGCGCAAGAG-3′ ) , ORF72 ( forward-5′-AGCTGCGCCACGAAGCAGTCA-3′; reverse , 5′-CAGGTTCTCCCATCGACGA-3′ ) , ORF73 ( forward 5′-CGCGAATACCGCTATGTACTCA-3′; reverse 5′-GGAACGCGCCTCATACGA-3′ ) , Kaposin A ( forward 5′ GGATAGAGGCTTAACGGTGTTT-3′; reverse 5′-CAGACAAACGAGTGGTGGTATC-3′ ) . A pool of two siRNAs synthesized by Integrated DNA technologies ( IDT ) were used to knockdown Kaposin A . siRNA sequences were as follows: siRNA1- 5′-r ( UUGCAACUCGUGUCCUGAAUGCUACGG ) -3′ , siRNA2-5′- r ( CCACAAACACCGUUAAGCCUCUAUCCA ) -3′ . Cells were transfected with siRNA at 100 pmol ( 50 pmol each ) using siLentFect ( Biorad ) according to the manufacturer's instructions . Cell lysates were collected at 48 h post-siRNA transfection for immunoprecipitation and Western blot analysis . Formalin-fixed , paraffin-embedded tissue samples from healthy subjects and patients with KS and primary effusion lymphoma were obtained from the ACSR ( AIDS and Cancer Specimen Resource ) . Sections were deparaffinized with HistoChoice clearing reagent and rehydrated through ethanol to water . For antigen retrieval , the sections were microwaved in 1 mmol/l EDTA ( pH 8 . 0 ) for 15 min , permeabilized with 0 . 5% Triton X-100 for 5 min , and then blocked with blocking solution ( Image-iT FX signal enhancer-Invitrogen ) for 20′ at RT . Immunostaining was performed using anti-mGluR1 and anti-mouse LANA-1 antibodies , followed by Alexa-488 and Alexa-594 conjugated secondary antibodies . Nuclei were stained with 4′ , 6-diamidino-2-phenylindole ( DAPI ) ( Molecular Probes , Invitrogen ) , and stained cells were viewed under a fluorescence microscope with a 20× objective and the Nikon Metamorph digital imaging system . Cells grown on 8 well chamber slides ( Nalge Nunc International ) were fixed with 4% paraformaldehyde for 15 min , permeabilized with 0 . 2% Triton X-100 , and blocked with Image-iT FX signal enhancer ( Invitrogen ) for 20 min . Cells were then incubated with primary antibodies against the specific proteins and subsequently stained with Alexa 488 or 594 conjugated secondary antibodies . Cells were mounted in mounting medium containing DAPI . Images were acquired using a Nikon 80i fluorescent microscope equipped with a Metamorph digital imaging system . Cells were lysed in RIPA buffer containing 15 mM NaCl , 1 mM MgCl2 , 1 mM MnCl2 , 2 mM CaCl2 , 2 mM phenylmethylsulfonyl fluoride , and protease inhibitor mixture ( Sigma ) . The cell lysates were centrifuged at 13 , 000× g for 20 min at 4°C . Samples mixed with sample buffer containing β-mercaptoethanol , heated at 95°C for 5 min , and separated by SDS PAGE . The protein samples were then Western blotted with the indicated primary antibodies followed by incubation with species-specific HRP-conjugated secondary antibodies . Immunoreactive bands were visualized by enhanced chemiluminescence ( Pierce , Rockford , IL ) according to the manufacturer's instructions . To determine the fold change , blots were scanned , and quantified by densitometric analysis ( Alpha Innotech Corporation , San Leonardo , CA ) and normalized with respect to the amount of β-actin . For immunoprecipitations , 300–500 µg of cell lysates prepared in RIPA buffer or in NP-40 buffer were incubated with the appropriate primary antibody for 4–8 h with end-over-end rotation at 4°C , and the precipitated proteins captured by Protein A or G-Sepharose . The samples were Western blotted with specific primary and secondary antibodies . Statistical significance was calculated using a two tailed Student's t-test . P<0 . 05 was considered significant . | Kaposi's sarcoma associated herpesvirus ( KSHV ) , prevalent in immunosuppressed HIV infected individuals and transplant recipients , is etiologically associated with cancers such as endothelial Kaposi's sarcoma ( KS ) and B-cell primary effusion lymphoma ( PEL ) . Both KS and PEL develop from the unlimited proliferation of KSHV infected cells . Increased secretion of various host cytokines and growth factors , and the activation of their corresponding receptors , are shown to be contributing to the proliferation of KSHV latently infected cells . Glutamate , a neurotransmitter , is also involved in several cellular events including cell proliferation . In the present study , we report that KSHV-infected latent cells induce the secretion of glutamate and activation of metabotropic glutamate receptor 1 ( mGluR1 ) , and KSHV latency associated LANA-1 and Kaposin A proteins are involved in glutaminase and mGluR1 expression . Our functional analysis showed that elevated secretion of glutamate and mGluR1 activation is linked to increased proliferation of KSHV infected cells and glutamate release inhibitor and glutamate receptor antagonists blocked the proliferation of KSHV infected cells . These studies show that proliferation of cancer cells latently infected with KSHV in part depends upon glutamate and glutamate receptor and therefore could potentially be used as therapeutic targets for the control and elimination of KSHV associated cancers . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"oncology",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"cell",
"biology",
"antiangiogenesis",
"therapy",
"cell",
"growth",
"oncology",
"agents",
"cancer",
"treatment",
"biology",
"and",
"life",
"sciences",
"cell",
"processes",
"viral",
"disease... | 2014 | Glutamate Secretion and Metabotropic Glutamate Receptor 1 Expression during Kaposi's Sarcoma-Associated Herpesvirus Infection Promotes Cell Proliferation |
Human cytomegalovirus ( HCMV ) is a widely distributed herpesvirus that causes significant morbidity in immunocompromised hosts . Inhibitors of viral DNA replication are available , but adverse effects limit their use . Alternative antiviral strategies may include inhibition of entry . We show that soluble derivatives of the platelet-derived growth factor receptor alpha ( PDGFR-alpha ) , a putative receptor of HCMV , can inhibit HCMV infection of various cell types . A PDGFR-alpha-Fc fusion protein binds to and neutralizes cell-free virus particles at an EC50 of 10–30 ng/ml . Treatment of particles reduced both attachment to and fusion with cells . In line with the latter , PDGFR-alpha-Fc was also effective when applied postattachment . A peptide scan of the extracellular domain of PDGFR-alpha identified a 40mer peptide that inhibits infection at an EC50 of 1–2 nmol/ml . Both , peptide and fusion protein , were effective against various HCMV strains and are hence promising candidates for the development of novel anti-HCMV therapies .
Human cytomegalovirus ( HCMV ) is a pathogenic human beta-herpesvirus that shares the property of other beta-herpesviruses to replicate only in its specific host . Primary infection is followed by lifelong latent persistence with occasional reactivation of the virus , which usually goes unnoticed by the infected individual . However , under conditions of insufficient immune responses , HCMV can cause severe or even life-threatening disease , e . g . in AIDS patients , transplant recipients , and fetuses infected in utero . Although antiviral drugs are available , their use is limited due to associated adverse effects and the development of resistance [1 , 2] . Therefore , alternative treatment options are desired . One powerful antiviral strategy is the inhibition of entry into the host cell , as exemplified by the effective neutralizing activity of anti-HCMV antibodies [3–10] . While the therapeutic use of antibodies may be limited as they are difficult to engineer , other entry inhibitors are also conceivable for HCMV . Small molecules and peptides have already been approved for antiretroviral therapy [11] , and a peptide-based entry inhibitor against Hepatitis B virus is in clinical trial [12] . In the case of picornaviruses , an Fc-CAR fusion protein inhibits viral entry and is effective in animal models , but has not yet been developed for clinical use [13–15] . HCMV is an enveloped virus and thus requires membrane fusion with the host cell to deliver its nucleocapsid into the cytoplasm . Several glycoprotein complexes in the envelope of HCMV particles have been described to contribute to the entry of HCMV into its target cells and are therefore potential targets for entry inhibitors [16–21] . By analogy with other herpesviruses , homotrimers of glycoprotein B ( gB ) are assumed to exert the fusion between viral envelope and cellular membrane , while heterotrimers of gH , gL and pUL74 ( gO ) are necessary to promote this fusion process [22–26] . On certain cell types like endothelial and epithelial cells , an additional pentameric complex is required for effective entry , which consists of gH , gL , and three accessory proteins from the viral UL128 gene locus [26–30] . On the cellular side , numerous proteins were proposed as entry receptors of HCMV , but have been controversially discussed: e . g . the epithelial growth factor receptor ( EGFR ) and the platelet-derived growth factor receptor alpha ( PDGFR-alpha ) [31–36] ( reviewed in [28] ) . Here , we show that the extracellular part of PDGFR-alpha is a highly potent entry inhibitor of HCMV in fibroblasts and endothelial cells , which represent the pentamer-independent and the pentamer-dependent entry pathway , respectively . PDGFR-alpha-derived small peptides are also effective , thus providing a rationale for the development of PDGFR-alpha based anti-HCMV therapeutics .
Two cellular growth factor receptors , PDGFR-alpha and EGFR have been reported to promote HCMV infection in fibroblasts [31 , 34] . However , their relevance for HCMV infection was questioned in subsequent studies [32 , 36] . With regard to PDGFR-alpha , a very recent report confirmed its significance for HCMV infection [37] . As we aimed at exploring the potential of these molecules to serve as a basis for the development of HCMV entry inhibitors , the first step was to confirm their contribution to HCMV infection . To address the diverse entry pathways of HCMV , we applied a virus strain expressing both gH/gL complexes to infect two model cell types representing the restricted tropism ( fibroblasts ) and the extended tropism ( endothelial cells ) . Using an siRNA approach , the respective growth factor receptors were targeted in human foreskin fibroblasts ( HFFs ) and endothelial hybrid cells ( EA . hy926 ) two days before infection with HCMV strain TB40/E at a multiplicity of infection ( MOI ) of 1 infectious unit per cell , corresponding to about 60% infection . Cells treated with non-targeting siRNAs served as negative controls while an siRNA targeting viral immediate-early ( IE ) transcripts was included as a positive control . One day after infection , cell cultures were fixed and viral IE antigens were immunostained to determine the fraction of infected cells . In each of three experiments , the degree of infection obtained with the various treatments was normalized to the non-targeting control . As expected , the IE siRNA partially reduced the infection efficiency ( by 60% in HFFs and 80% in endothelial cells , p-values > 0 . 01 ) . Treatment with PDGFR-alpha siRNA almost completely prevented HCMV infection of fibroblasts ( 95% reduction; highly significant with a p-value < 0 . 001 ) whereas it had no inhibitory effect in endothelial cells ( Fig 1 ) . In line with these results , PDGFR-alpha was only found on the surface of HFFs but not of endothelial cells in FACS analyses , and surface expression in HFFs was substantially suppressed upon transfection with PDGFR-alpha siRNA ( Fig 1D; for immunofluorescence stainings see S1 Fig , panel A ) . In contrast , no EGFR-specific signal was detected on fibroblasts , and EA . hy926 cells had only a very weak signal . Treatment with EGFR-specific siRNA did neither reduce the amount of EGFR detectable on the cell surface ( S1 Fig , panel B ) nor reduce infection efficiencies in these cell cultures ( S1 Fig , panel C ) . Hence , no conclusion on the role of EGFR for HCMV-infection was possible in our cell culture systems . In summary , of the two growth factor receptor molecules that had previously been reported to promote HCMV entry , only the contribution of PDGFR-alpha was confirmed in our experimental setting . The strong dependence of HCMV infection on expression of PDGFR-alpha suggested an interaction between this cellular growth factor receptor and HCMV particles during the entry process in HFFs . Therefore , we hypothesized that pretreatment of viral particles with soluble forms of this cellular molecule might block the putative interaction sites on the surface of HCMV virions and thereby inhibit infection . To test this hypothesis , we preincubated cell-free preparations of HCMV strain TB40/E with variable concentrations of a soluble PDGFR-alpha-Fc chimera for two hours prior to infection of HFFs and a human endothelial cell line ( HEC-LTTs; denoted as HECs ) . After another two hours , the virus was removed and replaced with the appropriate cell culture medium for an overnight incubation . Cultures were then fixed , and the fraction of infected cells was determined by indirect immunofluorescence staining of viral IE antigens . In fact , the PDGFR-alpha-Fc chimera inhibited infection of HFFs in a dose dependent manner with a half maximal effective concentration ( EC50 ) of about 10–20 ng/ml and a complete abrogation of infection at 200 ng/ml ( Fig 2 , left panel ) . Unexpectedly , infection of HECs was also reduced , albeit at slightly higher concentrations ( EC50 = 20–30 ng/ml ) and inhibition was incomplete ( Fig 2 , right panel ) . To address the possibility that the effect is non-specifically mediated by the Fc part of the chimeric molecule , we compared PDGFR-alpha-Fc with PDGFR-beta-Fc and EGFR-Fc chimeras regarding their inhibitory potential on HCMV infection . Again , cell-free preparations of TB40/E were preincubated with increasing concentrations of the various Fc chimeras for two hours . HFFs and HECs were then incubated with the mixtures for 2 h followed by a medium exchange and an overnight incubation . Evaluation of the degree of infection by immunofluorescence staining of viral IE antigens showed that only PDGFR-alpha-Fc blocked infection in a dose dependent fashion , whereas neither PDGFR-beta-Fc nor EGFR-Fc had an effect ( Fig 3A ) . As the Fc-part is identical with all three molecules , the growth factor receptor part of the PDGFR-alpha-Fc chimera is obviously required for the inhibitory effect . Next , we wondered whether soluble PDGFR-alpha-Fc would inhibit not only strain TB40/E but also other HCMV strains . We therefore prepared cell-free stocks of TB40/E , its fibroblast-restricted variant TB40/F and five other HCMV strains ( AD169 , Towne , Merlin , VR1814 , VHL/E ) that represent different phylogenetic glycoprotein variants described for HCMV [38] . Virus preparations were pretreated for two hours with PDGFR-alpha-Fc at a concentration that was sufficient for complete inhibition of strain TB40/E in HFFs in the previous dose response experiments ( 250 ng/ml ) . The mixtures were added to HFFs and incubated overnight . The fraction of infected cells was determined by immunofluorescence staining of viral IE antigens . All strains were strongly inhibited by pretreatment with the soluble receptor ( Fig 3B ) , and the reduction of infection was almost complete . Strains VR1814 and TB40/E showed a residual infection in 2 . 3% and 1 . 2% of cells , which only reached significance for VR1814 when compared to the other strains ( p-values > 0 . 05 ) . In conclusion , all tested strains were susceptible to inhibition by PDGFR-alpha-Fc irrespective of whether they contain the pentameric glycoprotein complex ( VR1814 , VHL/E , TB40/E ) or not ( AD169 , Towne , Merlin , TB40/F ) . Finally , to test whether this inhibitory effect was specific for HCMV we repeated the experiment and included another species of Herpesviridae , HSV-1 strain F . While the inhibitory effect on HCMV could be again reproduced , HSV infection was not affected by PDGFR-alpha-Fc ( S2 Fig ) , indicating that the effect is specific for HCMV . The finding that pretreatment of virus with soluble PDGFR-alpha abrogated infection supported the idea of a direct physical interaction between PDGFR-alpha and virions . To test this , we aimed to detect bound PDGFR-alpha-Fc molecules on HCMV particles that have been pretreated with this molecule , using a fluorescence-labelled antibody against human IgG . Cell-free preparations of HCMV strain TB40/E were preincubated with PDGFR-alpha-Fc , PDGFR-beta-Fc or EGFR-Fc for two hours at 37°C . Virus particles were then allowed to attach to HFFs for 90 min on ice followed by fixation with acetone in order to immobilize them and to facilitate fluorescent staining . To visualize all virus particles , the abundant viral capsid-associated protein pp150 ( pUL32 ) was stained in red via indirect immunofluorescence . Fc-fusion proteins were stained in green by direct immunofluorescence detecting the Fc part of the PDGFR-alpha-Fc fusion protein . Only after pretreatment with PDGFR-alpha-Fc , green fluorescence signals co-localized with attached virus particles whereas PDGFR-beta-Fc and EGFR-Fc did not yield signals ( Fig 4 ) . This is in accordance with the functional data regarding inhibition of infection and indicates that only PDGFR-alpha-Fc but not the other receptor-chimeras bind to HCMV particles . Mock-infected cells did not show any staining , proving that PDGFR-alpha-Fc did not bind to the cell surface in the absence of virus . We then sought to further analyze the mode of action by comparing the dose-dependent binding of PDGFR-alpha-Fc with the dose-response curve for inhibition of infection . If all available binding sites contributed to infection ( by attaching the virions to PDGFR-alpha on the cell surface ) and inhibition was simply due to blocking of these binding sites , we expected both dose-response curves to be in the same concentration range . If binding of PDGFR-alpha-Fc not only blocks interaction sites but also induces alterations of virions , e . g . premature activation of fusion proteins , we expected inhibition already at concentrations far below saturation of binding sites . Similarly , the presence of different binding sites ( e . g . active or inactive conformation of glycoproteins ) might result in a difference between dose response curves for inhibition and binding . Cell-free virus preparations were pretreated with a dilution series of PDGFR-alpha-Fc for two hours and virus particles were subsequently attached to HFFs on ice as described before . Again , attached particles were visualized by red staining of pUL32 and green staining of the Fc fusion part . Pictures were taken of both channels and a total of 100 viral particles per dilution was randomly selected according to the red pUL32 channel . For each particle , the maximum grey value of the Fc-staining was measured and median values representing the extent of binding at the respective concentration of PDGFR-alpha-Fc were calculated ( Fig 5 ) . In parallel , the same virus/PDGFR-alpha-Fc-mixtures were used to generate a dose-response curve of inhibition of infection as described above . Both curves converged towards a maximum ( Fig 5 , linear graph ) but differed greatly regarding the half maximal concentration ( Fig 5 , logarithmic graph ) , with an EC50 for binding to HCMV particles of approximately 100 ng/ml while the EC50 for inhibition of HCMV was more than tenfold lower ( approximately 10 ng/ml ) . This showed that PDGFR-alpha-Fc inhibits HCMV infection efficiently already at concentrations where only a minority of the total interaction sites are bound . To further investigate which of the initial steps of infection are blocked , we performed a series of experiments in HFFs and HECs to dissect adsorption and penetration . We used the dual fluorescent virus TB40-BACKL7-UL32EGFP-UL100mCherry as it allows to discriminate between enveloped ( both EGFP- and mCherry-positive ) and non-enveloped ( only EGFP-positive ) particles [39] . We compared adsorption and penetration of untreated particles to particles preincubated with 100 ng/ml PDGFR-alpha-Fc or PDGFR-beta-Fc by counting the number of enveloped ( = adsorbed , but not penetrated ) versus non-enveloped ( = penetrated ) particles . On both cell types , adsorption of PDGFR-alpha-Fc-treated particles was reduced ( by 50% on HFFs and 75% on HECs; Fig 6 ) , indicating that soluble PDGFR-alpha-Fc generally hinders HCMV attachment . The reduction of binding was significant in both cell types ( p-values < 0 . 05 ) . In contrast , inhibition of penetration was cell type-specific , with PDGFR-alpha-Fc-treated particles penetrating HFFs 75% less efficiently than untreated controls ( p-value < 0 . 05 ) whereas penetration was not specifically inhibited in HECs . As these results indicated that pretreatment with PDGFR-alpha-Fc inhibits fusion of the viral envelope with cellular membranes in HFFs , we tested whether this inhibitor could also block entry of virus particles that have already attached to HFFs . HCMV virus particles were adsorbed to HFFs for 1 h on ice . The virus-containing medium was then exchanged by medium containing PDGFR-alpha-Fc at a concentration of 200 ng/ml . Inhibition of preadsorbed virus was performed for additional 2 h on ice , before the cells were shifted to 37°C to allow entry . After 2 h of incubation at 37°C , the cells were supplied with fresh medium without inhibitor and further incubated overnight . After 24 h , the cells were fixed and stained for the viral IE antigens . PDGFR-alpha-Fc reduced infectivity of already attached viruses to 60% and this reduction was significant ( p-value < 0 . 05 ) ( Fig 7 ) . To corroborate the idea that this postattachment-effect was due to inhibition of fusion , we tested whether the reduced fusion capability can be overcome by addition of the chemical fusogen polyethylene glycol ( PEG ) . Therefore , cells with preadsorbed viruses were identically treated with PDGFR-alpha-Fc , but then PEG was added for 30 s before it was replaced with PDGFR-alpha-Fc containing medium for the 2 h incubation at 37°C . Postattachment inhibition was completely reversed by addition of PEG ( p-value < 0 . 01 ) , whereas PEG did not increase the infection efficiency of untreated control virus , suggesting that PDGFR-alpha-Fc actually inhibits the fusion step of HCMV entry under these experimental conditions . Alternatively , considering a recent report on endosomal uptake into fibroblasts [40] , PDGFR-alpha-Fc might interfere with an endocytic process and thus inhibit the downstream fusion step . Regarding possible viral interaction partners of PDGFR-alpha , it is currently unclear whether it binds to gB ( pUL55 ) [31] or gO ( pUL74 ) [37] or both . We compared wild type HCMV particles with pUL74-deficient viral particles regarding their ability to bind PDGFR-alpha-Fc chimera ( Fig 8 ) , thereby enabling a discrimination between the two proposed interaction partners . If gB was the binding partner , PDGFR-alpha-Fc should bind to both wild type and mutant . If gO ( pUL74 ) was the binding partner , it should bind only to wild type particles . One experimental challenge was posed by the fact that deletion of pUL74 strongly reduces infectivity , which is partially explained by a reduction of the ability of mutant virions to bind to cells ( S3 Fig ) . Therefore , TB40-BAC4-UL74stop virus preparations had to be concentrated by ultracentrifugation in order to achieve sufficient particle numbers on cells for a comparison with wild type virus . Following this adjustment procedure , wild type or UL74stop particles were incubated with 500 ng/ml ( i . e . a concentration that would yield saturated binding , see Fig 5 ) of PDGFR-alpha-Fc for two hours prior to attachment to the cells on ice . The particles on the cells were again visualized with an antibody recognizing the structural protein pUL32 and bound Fc-chimeras were stained with an Alexa488-conjugated anti-human-IgG antibody ( Fig 8A ) . Only virus particles containing the glycoprotein pUL74 were stained with the fluorescent antibody against the Fc-part of the soluble receptor molecule , indicating that the trimeric gH/gL/pUL74 complex but not gB is involved in binding of PDGFR-alpha-Fc to virions . To investigate whether lack of the trimeric complex not only abrogates binding but also renders virus insensitive to the inhibitory effect of PDGFR-alpha-Fc , UL74stop virus and wild type virus were compared regarding the PDGFR-alpha-Fc-mediated inhibition of infection in HFFs and HECs ( Fig 8B; see also S4 Fig ) . As deletion of pUL74 from the virus greatly reduced infectivity of viral progeny , mutant virions had to be concentrated 50-fold by ultracentrifugation to achieve infection efficiencies up to 40% in HECs and 5% in HFFs . Wild type virus was used at dilutions that yielded a comparable level of infection . As expected , infection with wild type virus was significantly reduced by PDGFR-alpha-Fc in a dose-dependent fashion ( p-values < 0 . 0001 ) . In contrast , the infectivity of the UL74stop virus did not significantly change with increasing doses of PDGFR-alpha-Fc , indicating that the inhibitory effect of PDGFR-alpha-Fc is mediated via gH/gL/pUL74 . The finding that only PDGFR-alpha-Fc but not PDGFR-beta-Fc or EGFR-Fc inhibits HCMV infection indicated that the inhibitory effect is specific for the PDGFR-alpha part of the chimeric molecule . PDGFR-alpha-Fc contains only the extracellular domain of the native PDGFR-alpha transmembrane molecule . We hypothesized that peptides derived from this part of the protein could also inhibit infection . Therefore , we tested a set of overlapping 40mer peptides , which cover the whole sequence of the extracellular PDGFR-alpha domain , regarding their inhibitory potential . Cell-free preparations of the Gaussia luciferase-expressing HCMV strain TB40-BAC4-IE-GLuc were preincubated with the individual peptides for two hours at concentrations ranging from 0 . 05–50 nmol/ml . The mixtures were then used to infect HFFs and HECs for two hours at 37°C . Virus-Peptide mixtures were replaced with the appropriate growth medium and cells were further incubated overnight . Luciferase containing supernatants were harvested and luminescence indicating the extent of infection was counted in a plate reader using coelenterazine as a substrate . The degree of neutralization was calculated as 1- ( luminescence / maximal luminescence ) . The various peptides differed greatly regarding their inhibitory potential: the C-terminal region from peptide no 11 to peptide no 17 ( corresponding to aa300-aa505 ) was ineffective ( Fig 9 ) . The peptide GT-40 ( #4 in Fig 9 ) , ranging from aa91 to aa130 ( GRHIYIYVPDPDVAFVPLGMTDYLVIVEDDDSAIIPCRTT ) , was particularly effective with an EC50 of 2 nmol/ml and almost complete inhibition at 10 nmol/ml , irrespective of the cell type . Like the full-length protein , this peptide was also effective against other strains of HCMV ( S6 Fig ) . At higher concentrations , some of the other peptides in the N-terminal part of the extracellular domain of PDGFR-alpha also inhibited infection in both cell types to some extent .
The finding that soluble platelet-derived growth factor receptor alpha binds to HCMV virions only if they contain the gH/gL/gO trimer , thereby inhibiting entry into fibroblasts and endothelial cells , has implications for our basic understanding of HCMV entry and may provide the starting point for a novel antiviral strategy . PDGFR-alpha has repeatedly been reported as a cellular factor that can mediate HCMV infection [31 , 32 , 37] . However , for the first time we directly visualize the binding of PDGFR-alpha to virions of HCMV . The fact that binding is abrogated in virions of pUL74 deletion mutants , which lack the viral complex gH/gL/gO in their envelope , suggests that the trimer is the only viral interaction partner of the cell surface protein PDGFR-alpha . This finding is in line with a recent report showing that soluble gO can bind to this growth factor receptor [37] . As gO deletion mutants have been shown to contain normal amounts of gB in their envelope [41] our data directly argues against a binding to the envelope complex formed by gB , which has initially been suggested [31] . This fits with a previous report questioning a receptor-binding activity of gB [42] . Concerning EGFR , the fact that soluble EGFR-Fc did not inhibit infection in either cell type argues against a role for EGFR as a binding receptor on the cell surface , although it does not preclude a more indirect contribution of this molecule during HCMV entry , e . g . via signaling [43 , 44] . While inhibition of fibroblast infection by pretreatment of virions with soluble PDGFR-alpha-Fc can be explained by interference with the binding of gH/gL/gO on virions to PDGFR-alpha on the cell surface , it appears surprising at a first glance that infection of endothelial cells was also inhibited by almost 90% . This seems to conflict with the fact that treatment of endothelial cells with PDGFR-alpha-specific siRNA had no effect on infection of endothelial cells , similar to previous findings in epithelial cells [32] . In support of the siRNA data , the lack of detectable PDGFR-alpha staining in our FACS analyses indicates that this receptor is not available for the promotion of HCMV entry into endothelial cells , which is concordant with available mRNA expression data from primary cell cultures ( GEO data set accession #GDS1402 ) . Hence , an interference of soluble PDGFR-alpha with infection of endothelial cells was not expected . On the other hand , Zhou et al . recently showed that the gH/gL/gO trimer is also necessary for efficient entry into endo- and epithelial cells [26] . Hence , it is conceivable that soluble PDGFR-alpha bound to the trimer might impede its penetration-promoting function . However , our mode-of-action analysis , in which only the binding step but not the subsequent penetration step was significantly reduced in endothelial cells , indicates that PDGFR-alpha-Fc does not interfere with this contribution of gH/gL/gO . Alternatively , PDGFR-alpha-Fc molecules that are bound to trimers in the virion envelope might sterically interfere with other interactions between HCMV and endothelial cells , e . g . binding of the pentamer to a yet undefined receptor . Regardless of the exact mode of action , the fact that PDGFR-alpha-Fc can block infection in both cell types obviously increases the potential to exploit this inhibitory molecule for antiviral treatment strategies . One problem of current HCMV treatment options is that all approved drugs ( ganciclovir , valganciclovir , foscarnet and cidofovir ) target the same step of the viral life cycle , which is replication of the viral genome [1 , 45] . They all frequently cause severe adverse effects including myelosuppression and nephrotoxicity , and this is particularly disadvantageous in transplant recipients with preexisting dysfunction of the kidney and/or the bone marrow . Furthermore , mutations have been reported that confer resistance to all of these drugs . Hence , alternative approaches are needed to provide treatment options in cases where the currently available drugs should be avoided . An inhibitor of the viral terminase that contributes to genome packaging is currently in a phase III trial [45] . This inhibitor causes little adverse effects but also induces resistance mutations [46 , 47] . Another possible target for antiviral strategies is the beginning of the replicative cycle when HCMV must attach to target cells and fuse its envelope with a cellular membrane for entry into the cytoplasm . One well established way to interfere with these initial steps of viral infection is blocking of viral envelope proteins by neutralizing antibodies . In case of HCMV , the envelope glycoproteins gB , gH , and the accessory proteins of the pentameric gH/gL complex are major targets of neutralizing antibodies [6 , 48 , 49] . Remarkably , such neutralizing antibodies are only moderately effective against infection of fibroblasts as compared to their effect on epi- and endothelial cells [6 , 8 , 50 , 51] . Entry into fibroblasts might be blocked specifically by antibodies against gO [37 , 52] , which are usually not found in HCMV-seropositive individuals , or by soluble PDGFR-alpha-derivatives . Based on our data with the UL74-deletion mutant , PDGFR-alpha-Fc can be assumed to specifically neutralize gO and thus act as a mimic of anti-gO antibodies . Noteworthy , the interaction of HCMV virions with cell surface receptors can be targeted in the extracellular space , thus greatly reducing the risk of interference with intracellular pathways and opening possibilities for molecules that cannot easily penetrate cells , like proteins and peptides . Importantly , PDGFR-alpha-Fc inhibited all tested HCMV strains almost completely , including three strains that express both the trimer and the pentamer ( VHL/E , TB40/E and VR1814 ) . In a recent report , VR1814 was about 10-fold less sensitive to inhibition by soluble PDGFR-alpha than AD169 [37] , which might suggest that this difference is due to expression of the pentamer . Our data argue against the assumption that the pentamer would render HCMV resistant against inhibition by PDGFR-alpha-derivatives , as VHL/E was completely inhibited and the residual infectivity with TB40/E was around 1% at an inhibitor concentration of 10 x EC50 . Only VR1814 differed significantly from the other strains by a slightly higher residual infectivity , suggesting that strain-specific differences might exist independent from pentamer expression . As all three proteins of the trimer show polymorphisms , minor interstrain differences in the binding to PDGFR-alpha would not be too surprising . It is also noteworthy that the soluble PDGFR-alpha used by Kabanova et al . was more than a log step less efficient that the PDGFR-alpha-Fc molecule that we applied , which may indicate a contribution of the Fc part . In line with this interpretation , an anti-gO antibody was effective against VR1814 in both fibroblasts and epithelial cells [37] , which fits perfectly with our finding that PDGFR-alpha-Fc is also effective in endothelial cells . Regarding the distinct steps during viral entry that are inhibited , PDGFR-alpha-Fc did not reduce binding during incubation on ice ( Fig 4 and S5 Fig ) suggesting that the initial low affinity binding to heparan sulfate proteoglycans ( HSPGs ) is unaffected . In contrast , PDGFR-alpha-Fc reduced binding to both cell types during incubation at 37°C ( Fig 6A ) . One possible explanation is that binding of PDGFR-alpha-Fc to the trimer cannot impede the initial attachment to HSPGs but interferes with a secondary high affinity binding to other cell surface molecules , e . g . receptors of the gH/gL-complexes . Unexpectedly , adsorption was also affected in endothelial cells although they lack PDGFR-alpha . One explanation is that trimer-bound PDGFR-alpha-Fc sterically hinders interaction of the pentamer with its ( unknown ) cellular receptor simply by increasing the distance between virion and cell surface . Another possibility is that the trimer itself is involved in binding of virions to endothelial cells via an unidentified receptor , which is supported by the unexpected finding that adsorption to endothelial cells is reduced with UL74stop virions ( S3 Fig ) . At present , too little is known about the exact role of the trimer for entry into endothelial cells to favor one of these options . Finally , the penetration efficiency of adsorbed particles was only affected in fibroblasts but not in endothelial cells ( Fig 6B ) . Taken together , it appears that binding of soluble PDGFR-alpha to virus particles inhibits the high affinity attachment to both cell types and in addition blocks fusion of the viral envelope with cellular membranes specifically in fibroblasts . Further support for an effect on fusion is provided by the finding that PDGFR-alpha-Fc blocks infection of HFFs even when applied after initial attachment and that this block can be overcome by the chemical fusogen PEG ( Fig 7 ) . The fact that two different steps of the entry process are inhibited in HFFs but only one step is affected in HECs may also explain the difference regarding the maximal effect that can be achieved with soluble PDGFR-alpha in the respective cell type . While inhibition in HFFs is complete , a small residual PDGFR-alpha-Fc-resistant infectivity remains in HECs ( Figs 2 , 3 and 8 ) . It is tempting to speculate that this residual infection occurs via a pathway that is independent of the trimer and only relies on the pentamer . It has recently been reported that the trimer is necessary for efficient fusion in both cell types [26] . Our data support this conclusion but also suggest that there is a minor trimer-independent entry pathway in HECs whereas the trimer is almost indispensable in HFFs . This pathway is probably irrelevant during natural infection by wild type virus but becomes apparent when the trimer-mediated pathways are not available , as in case of the UL74stop mutant . Somewhat unexpectedly , inhibition of infection is already complete at concentrations that saturate only a low proportion of possible binding sites ( Fig 5 ) . This may be explained by the presence of at least two different binding sites , with the higher affinity binding site contributing more to virus entry than the lower affinity binding site . Alternative explanations are that binding to only few of the trimeric complexes can trigger a premature inactivation of the fusion protein gB or sterically hinder interaction with the cell by preventing the binding of other active trimeric complexes to PDGFR-alpha on the cell surface . Concerning the therapeutic potential , this may be relevant because compounds that rely on complete saturation usually do not work for treatment of patients whereas compounds that are fully active at lower concentrations are more promising regarding the ratio of desired and adverse effects [53 , 54] . In case of PDGFR-alpha-Fc , inhibition of HCMV already occurs at concentrations that are 10-100fold below those reported for binding to the natural ligands in the information of the reagent provider ( R&D systems ) . If this applies also in vivo , an antiviral effect can be expected at doses that would not significantly bind and sequester the natural ligand , thus limiting unwanted effects . Due to its mode of action as an inhibitor of viral adsorption and penetration , which is analogous to neutralizing antibodies , PDGFR-alpha-Fc can only be expected to block cell-free virus transmission but will probably be ineffective regarding cell-to-cell spread . A PDGFR-alpha-derived entry inhibitor would fall into the class of biologicals , which are in general characterized by their low toxicity [54] . This could be particularly advantageous , when considering treatment of pregnant women with primary infection , treatment of fetuses infected in utero and long-term treatment of congenitally infected newborns [2] . Remarkably , a similar approach has been successful in an animal model of coxsackie virus , where Fc-CAR was effective in preventing viral dissemination and disease [13 , 14 , 55] . Clinical experience with surface receptor-Fc chimeras is already available in rheumatic disease , where a tumor necrosis factor receptor-Fc fusion protein is well established in treatment regimens [56] . Beside the whole extracellular domain of PDGFR-alpha , small peptides derived from its sequence are an alternative therapeutic option . There is an increasing amount of evidence proving the excellent tolerability of host-derived peptides [57] and peptides have been shown to be effective particularly against viral entry [12 , 58] . Hence , our finding that a PDGFR-alpha-derived 40mer can also efficiently reduce infection in both cell types at a concentration of 3–30 nmol/ml provides a promising starting point for further optimization [54] . The fact that inhibition achieved with this peptide ( 95% in HFFs ) was lower than with PDGFR-alpha-Fc ( 99% in HFFs ) could for example be due to a lower affinity of the peptide as compared to the complete extracellular domain of the receptor , and this indicates that further improvements might be possible . Considering their therapeutic application , both PDGFR-alpha-Fc and PDGFR-alpha-derived peptides may offer a number of advantages: ( i ) they are completely host-derived and therefore assumed to be non-immunogenic , ( ii ) an additive effect with the established anti-HCMV drugs can be expected due to the different modes of action; ( iii ) in contrast to most antibodies [5 , 8] they are almost equally effective against infection via the pentamer-dependent and the pentamer-independent entry pathway and ( iv ) resistance-conferring mutations would most likely affect the entry potential of the virus and hence reduce viral fitness . In conclusion , soluble derivatives of PDGFR-alpha can effectively inhibit entry of HCMV into various cell types , which might pave the way for the development of a therapeutic entry inhibitor .
Primary human foreskin fibroblast ( HFFs ) were isolated from tissue samples that were residuals from routine procedures . Samples were obtained anonymized after written informed consent of the parents in agreement with articles 21 and 23 of the recommendations of the council of Europe ( 2006 ) . HFFs were propagated in MEM supplemented with GlutaMAX ( Life Technologies ) , 5% fetal calf serum ( FCS; PAN Biotech ) , 100 μg/ml gentamicin and 0 . 5 ng/ml basic fibroblast growth factor ( bFGF; Life technologies ) . Experiments were carried out in HFF-medium without bFGF ( denoted as MEM5 ) . Conditionally immortalized human endothelial cells ( HEC-LTT , denoted as HECs ) , were kindly provided by D . Wirth [59 , 60] . HEC-LTTs are human umbilical vein endothelial cells ( HUVECs ) that contain doxycycline-controlled expression cassettes for the human telomerase catalytic subunit ( hTERT ) and the simian virus 40 large T-antigen ( SV40-TAg ) [59] . In the presence of the doxycycline , hTERT and SV40-TAg expression are activated , resulting in high cell proliferation and unlimited expansion . HECs were cultured in gelatin-coated vessels using endothelial cell growth medium ( EGM bullet kit; Lonza ) supplemented with 2 μg/ml doxycycline . For infection experiments , HECs were seeded the day before in the absence of doxycycline , resulting in a growth-arrested state resembling primary HUVECs . The efficiently transfectable and HCMV-susceptible hybrid endothelial cell line EA . hy926 ( ATCC CRL-2922 ) [60–62] was expanded in DMEM ( life technologies ) supplemented with 10% FCS . The HCMV strain TB40 has been previously isolated in our lab from a bone marrow transplant patient [63] . The highly endotheliotropic version TB40/E was the result of long-term propagation on endothelial cells whereas TB40/F lost the pentameric complex during cultivation on fibroblasts and is therefore non-endotheliotropic . AD169 [64] , Towne [65] and Merlin [66 , 67] are widely used HCMV strains that also lack the pentameric complex . VR1814 [68] , VHL/E [69] represent endotheliotropic HCMV strains . TB40-BACKL7-UL32EGFP-UL100mCherry is a BAC-cloned derivative of TB40/E that was fluorescently labelled to allow differentiation between enveloped and non-enveloped virus capsids [39] . TB40-BAC4 is a highly endotheliotropic BAC-clone based on TB40/E [70] and served as the basis for mutant BAC4-UL74stop which lacks the expression of pUL74 ( gO ) [71] . The reporter virus TB40-BAC4-IE-GLuc contains a Gaussia luciferase expression cassette under control of the major immediate-early enhancer/promoter at an ectopic position [72] . Infectious supernatants of TB40 variants , AD169 , Towne , VHL/E and VR1814 were harvested from infected HFFs at five to seven days postinfection . Supernatants were cleared from cells and cellular debris by centrifugation at 2 , 700 x g for 10 min before storage at -80°C . Cleared UL74stop supernatants were 50-fold concentrated by ultracentrifugation at 70 , 000 x g for 70 min . Virus stocks of TB40-BAC4-IE-GLuc were first cleared from cellular debris as described above and virions were washed twice by ultracentrifugation at 70 , 000 x g for 70 min to minimize background levels of secreted luciferase . All recombinant Fc-fusion proteins used in this study were obtained from R&D: PDGFR-alpha-Fc , PDGFR-beta-Fc and EGFR-Fc . Proteins were dissolved in phosphate buffered saline ( PBS ) at a concentration of 500μg/ml . The 40mer peptides based on the extracellular domain of human PDGFR-alpha isotype 1 were obtained from Phtdpeptides ( Shanghai , China ) at a purity of > 95% . Depending on their physiochemical properties , peptides were dissolved in either water , 0 . 1 M ammonium carbonate , 10% acetic acid or dimethyl sulfoxide to a concentration of 1 mmol/l . Infection efficiencies were determined by indirect immunofluorescence staining of viral IE proteins pUL122/123 . Cells were fixed with 80% acetone for 5 min and incubated sequentially with primary mouse antibody E13 ( Argene ) and secondary antibody Cy3-goat anti-mouse IgG F ( ab′ ) 2 ( Jackson ImmunoResearch ) . To locate nuclei , cells were counterstained with 4′ , 6-diamidino-2-phenylindole ( DAPI ) . The percentage of infection was determined by counting the number of IE-positive cells , as well as the total number of nuclei per image . For each condition , three images were evaluated . To determine infection efficiencies with the reporter virus TB40-BAC4-IE-GLuc , a 20 μl aliquot of each cell culture supernatant was transferred to a 96-well luminescence reader plate . 50 μl of the substrate coelenterazine ( 0 . 2 μg/ml; PjK ) were injected using a microplate reader with a built-in injection system ( Chameleon , Hidex ) and luminescence was measured . For all obtained values , background luminescence was subtracted and neutralization efficiencies were calculated . For reverse transfection of siRNAs , pools of four different siRNAs ( siGENOME Dharmacon ) per target were complexed with Lipofectamine RNAiMAX ( Life Technologies ) in 96-well plates . HFFs and EA . hy926 cells were added at a density of 10 , 000 cells per well . A highly efficient HCMV-IE siRNA [62] ( Sigma-Aldrich ) served as a positive control and the siGENOME non-targeting pool #2 ( Dharmacon ) was used as a negative control . Duplicate wells of all conditions were prepared . Two days posttransfection , cells were infected with HCMV-TB40/E at a multiplicity of infection ( MOI ) 1 . The next day , cells were fixed with 80% acetone and stained for viral IE antigens . To analyze the effect of siRNA-mediated knockdown on cell surface expression of PDGFR-alpha and EGFR , HFFs and EA . hy926 cells were transfected with the respective siRNA or non-targeting siRNA , harvested 2 d posttransfection , and incubated on ice with PE-conjugated anti-PDGFR-alpha-antibody ( FAB1264P , R&D ) , Fluorescein-conjugated anti-EGFR-antibody ( FAB10951F , R&D ) or the respective isotype controls . Cells were then washed twice in PBS with 2% FCS and analyzed in a FACS Calibur ( BD Biosciences ) . The obtained data sets were processed with Flowing Software version 2 . 5 . 1 ( by Perttu Terho ) . For testing the inhibitory effect of Fc-fusion proteins or peptides on HCMV , agents were diluted in MEM5 and mixed with virus preparations at a concentration resulting in a final MOI of 1 . The mixtures were incubated for 2 h at 37°C before addition to the cells . HFFs and HECs seeded on gelatin-coated 96-well plates at a density of 1 . 5 x 104 cells per well were preincubated with MEM5 for 30 min , medium was exchanged against virus mixtures and incubated for 2 h . Virus mixtures were removed and replaced by the respective growth medium . Cells were incubated overnight and infection efficiencies were measured as described above . EC50 values were determined by non-linear regression applying sigmoidal dose response curve fitting ( Sigma Plot ) . To assess the binding of Fc-Proteins to virus particles , HFFs were seeded on gelatin-coated 8-well μ-slides ( Ibidi ) at a density of 4 x 104 cells per well one day prior to infection . Virus preparations were preincubated with Fc-fusion proteins at indicated concentrations for 2 h at 37°C and subsequently precooled on ice for 15 min . Medium of precooled cells was exchanged against virus mixtures and incubated on ice for 90 min . The cells were washed once with MEM5 prior to fixation with 80% acetone . For staining of viral particles , cells were incubated with a mouse monoclonal antibody recognizing the abundant viral protein pUL32 ( generously provided by W . Britt ) [73] . As a secondary antibody , Cy3-conjugated goat anti-mouse IgG F ( ab’ ) 2 ( Jackson ImmunoResearch ) was used . Visualization of bound Fc-proteins was achieved by applying Alexa488-conjugated goat-anti-human IgG ( H+L; Invitrogen ) . For better orientation , cell nuclei were stained with DAPI . For quantification of PDGFR-alpha-Fc binding to HCMV particles , the grey values of 100 particles per condition were analyzed using AxioVision Software ( Zeiss ) . HFFs and HECs were seeded on gelatin-coated 8-well μ-slide ( Ibidi ) at a density of 4 x 104 cells per well . Freshly produced cell-free infectious supernatant of TB40-BACKL7-UL32EGFP-UL100mCherry was mixed with MEM5 ( untreated control ) or MEM5 with 100 ng/ml soluble Fc-chimeras and incubated for 2 h at 37°C . Following a preincubation of the cells with MEM5 for 30 min , medium was exchanged against virus mixtures and incubated for 20 min at 37°C . Virus mixtures were exchanged against MEM5 ( untreated control ) or MEM5 with 100 ng/ml of the respective Fc-chimera and incubated for another 100 min at 37°C . Cells were fixed with 80% acetone for 5 min and sequentially reacted with anti-GFP mouse IgG2a ( clone 3E6 , Invitrogen ) , Alexa Fluor488-conjugated goat anti-mouse IgG F ( ab′ ) 2 fragments ( Invitrogen ) and DAPI . Pictures of each individual channel ( native red fluorescent signals; stained green pUL32-EGFP signals and blue nuclei ) were taken and particles of a total of 15 cells per condition were counted . HFFs were seeded on two gelatin-coated 8-well μ-slides ( Ibidi ) at a density of 4 x 104 cells per well . Cells were precooled on ice for 15 min and medium was subsequently exchanged against ice-cold , freshly produced cell-free infectious supernatant of TB40/E . Attachment of viral particles was allowed on ice for 1 h . Supernatant was exchanged against precooled MEM5 with or without 200 ng/ml PDGFR-alpha-Fc and incubated on ice for 2 h . One of the plates was shifted to 37°C for 2 h , whereas cells of the replica-plate were treated with prewarmed 50% polyethylene glycol 1500 PEG; Roche ) for 30 sec . The PEG was immediately removed by five-times washing with prewarmed PBS and cells were incubated in prewarmed MEM5 with or without 200 ng/ml PDGFR-alpha-Fc for 2 h at 37°C . Medium was exchanged to MEM5 on both replica plates and infection was allowed to proceed for 24 h before infection efficiencies were assessed by staining of viral IE antigens . Datasets were analyzed by one-way-ANOVA using the build-in data analyses function of Excel to test whether there are significant differences between the various conditions . If ANOVA indicated significant differences between groups within the data set , post-hoc analyses were performed to identify which of the groups differ from others . Depending on the structure of the data sets , post-hoc analyses were done by t-test ( Figs 1 and 3 ) , Dunnett's test ( Fig 6 ) or Tukey test ( Fig 7 ) , using the build-in data analyses functions of Excel ( t-test ) or Sigmaplot ( Dunnett’s , Tukey ) . Only the p-values of post-hoc tests are mentioned in the results section . The error of the median ( Fig 5 ) was determined by 1000fold resampling ( bootstrapping ) from our data sets of 100 cells per condition . | Human cytomegalovirus ( HCMV ) depends on expression of platelet-derived growth factor receptor alpha ( PDGFR-alpha ) for infection of fibroblasts whereas this cell surface protein is not required for infection of endothelial cells . Surprisingly , pretreatment of HCMV with a soluble derivative of PDGFR-alpha prevents infection of both cell types , most probably via specific binding to the trimeric gH/gL/pUL74 complex . While adsorption is inhibited in both cell types , an additional penetration block occurs only in fibroblasts . The finding that an essential molecular interaction of HCMV with fibroblasts can be subverted for inhibition of the virus provides an antiviral strategy that may be hard to circumvent by the virus . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"cell",
"binding",
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"endothelial",
"cells",
"pathogens",
"immunology",
"microbiology",
"viral",
"structure",
"fibrob... | 2017 | A derivative of platelet-derived growth factor receptor alpha binds to the trimer of human cytomegalovirus and inhibits entry into fibroblasts and endothelial cells |
Aspergillus species are a major worldwide cause of corneal ulcers , resulting in visual impairment and blindness in immunocompetent individuals . To enhance our understanding of the pathogenesis of Aspergillus keratitis , we developed a murine model in which red fluorescent protein ( RFP ) -expressing A . fumigatus ( Af293 . 1RFP ) conidia are injected into the corneal stroma , and disease progression and fungal survival are tracked over time . Using Mafia mice in which c-fms expressing macrophages and dendritic cells can be induced to undergo apoptosis , we demonstrated that the presence of resident corneal macrophages is essential for production of IL-1β and CXCL1/KC , and for recruitment of neutrophils and mononuclear cells into the corneal stroma . We found that β-glucan was highly expressed on germinating conidia and hyphae in the cornea stroma , and that both Dectin-1 and phospho-Syk were up-regulated in infected corneas . Additionally , we show that infected Dectin-1−/− corneas have impaired IL-1β and CXCL1/KC production , resulting in diminished cellular infiltration and fungal clearance compared with control mice , especially during infection with clinical isolates expressing high β-glucan . In contrast to Dectin 1−/− mice , cellular infiltration into infected TLR2−/− , TLR4−/− , and MD-2−/− mice corneas was unimpaired , indicating no role for these receptors in cell recruitment; however , fungal killing was significantly reduced in TLR4−/− mice , but not TLR2−/− or MD-2−/− mice . We also found that TRIF−/− and TIRAP−/− mice exhibited no fungal-killing defects , but that MyD88−/− and IL-1R1−/− mice were unable to regulate fungal growth . In conclusion , these data are consistent with a model in which β-glucan on A . fumigatus germinating conidia activates Dectin-1 on corneal macrophages to produce IL-1β , and CXCL1 , which together with IL-1R1/MyD88-dependent activation , results in recruitment of neutrophils to the corneal stroma and TLR4-dependent fungal killing .
Fungal infections of the cornea ( i . e . fungal keratitis ) account for approximately 1% of corneal ulcers in temperate regions of industrialized nations [1] . However , in tropical regions of developed countries , such as in the southeastern United States , fungal infections of the cornea account for up to 35% of all corneal ulcers resulting in severe visual impairment and blindness [1]–[3] . Globally , the impact of fungal keratitis on visual health is even greater , with reports of up to 60% of corneal ulcers attributable to fungal infection in developing nations including China , Nepal , India , Bangladesh , Ghana and Mexico [4]–[11] . The etiological agents of these corneal infections are most commonly filamentous Fusarium ( F . solani , F . oxysporum ) [1] , [12] , and Aspergillus species ( A . flavus , A . fumigatus , nomius [13] , A . tamarii [14] , A . terreus , and A . tubingensis [15] , [16] ) , which are prevalent in hot , humid climates , and where the predominant risk factor is traumatic injury associated with agricultural work . As Aspergillus is ubiquitous in the environment , the population at large is constantly exposed to this opportunistic pathogen [17]–[19] , and conidia are inhaled at an estimated 200 conidia per day [19]–[21] . In areas of high endemicity , A . fumigatus can also be isolated from the conjunctival sac of healthy individuals [22] . Thus , in the setting of a disrupted corneal epithelium , there is broad opportunity for inoculation of A . fumigatus conidia ( 2–3 µm diameter ) or conidiophores ( >100 individual conidia per hyphal stalk ) from airborne vegetative matter or the conjunctiva into the corneal stroma in association with traumatic injury [8] , [23] . Subsequently , fungal virulence factors that facilitate hyphal invasion of the tissue ( e . g . toxin and protease secretion ) [24]–[29] , together with low efficacy of anti-mycotic therapy [30]–[32] , and the resulting inflammatory response all converge to induce destruction of corneal tissue . Treatment failure also occurs in up to 60% of patients , who may require at least one and sometimes repeated corneal transplantation , and in severe cases results in enucleation of the infected eye [23] , [31] . In contrast to pulmonary aspergillosis , which is associated with immune suppression , fungal keratitis occurs in otherwise healthy , immunocompetent individuals . Despite the impact of filamentous fungal infections on global blindness , and a recent outbreak of contact lens-associated Fusarium keratitis in the USA [12] , the pathogenesis of this disease is not well understood . We therefore developed a murine model of corneal infection using Aspergillus that constitutively express dsRed fluorescent protein at each stage of development , and are therefore readily detected in the transparent cornea . Our findings demonstrate that Dectin-1 mediates the initial wave of pro-inflammatory and chemotactic cytokine production and cellular infiltration into the cornea . We also show that the IL-1R1/MyD88 pathway further propagates cellular recruitment into the cornea during Aspergillus infection , and that in contrast to Dectin 1 , TLR4 does not regulate cellular infiltration , but is essential for anti-fungal activity .
To visualize A . fumigatus during infection of the transparent mammalian cornea , we generated a monomeric dsRED RFP-expressing A . fumigatus strain . Figure S1 illustrates the structure of the plasmid , and shows expression of RFP at each lifecycle stage of Aspergillus . To determine the role of the host response to these organisms in the cornea , C57BL/6 mice were either treated systemically with cyclophosphamide ( cyc ) , or were left untreated prior to injecting 1×105 Af293 . 1RFP conidia into the corneal stroma ( Preliminary studies revealed that 1×105 conidia was the smallest inoculum in which organisms could be recovered from the cornea after 24h ) . Corneal opacification , fungal growth , survival , and cellular infiltration were assessed at each time point . Figure 1A shows that at 24h post-infection , immunocompetent C57BL/6 mice developed significant corneal opacity , which peaked at 48h , and persisted up to 72h post-infection . However , cyclophosphamide-treated mice had significantly lower corneal opacification scores than immunocompetent mice at 48 and 72h post-infection . To characterize the host response in Aspergillus keratitis , 5µm corneal sections were stained with PASH and examined by brightfield microscopy . As shown in Figure 1B , there was a pronounced cellular infiltration in the corneas of immunocompetent mice between 24 and 72h , with neutrophils also present in the anterior chamber ( for comparison , a PASH stained section of a normal mouse cornea is shown in Supplementary Figure S2A ) . In contrast , there was less cellular infiltration in corneas of cyclophosphamide-treated mice at each time point . There was also enhanced fungal penetration though Descemet's membrane and into the anterior chamber at 48 and 72h post-infection , and increased epithelial debridement in cyclophosphamide treated mice compared with immunocompetent mice . Figure 1C shows that cells in the C57BL/6 cornea are NIMP-R14+ neutrophils , and that the numbers increased over time . Neutrophils were not detected in corneas of cyclophosphamide-treated mice at any time post-infection , and in contrast to immunocompetent mice , longer-term infection of treated mice results in corneal perforation ( data not shown ) . To quantify corneal opacification and hypertrophic and filamentous fungal growth during corneal infection , we used image analysis software that provides numerical output for corneal opacification and for fungal dsRed expression for all animals in the experiment rather than representative mice . Figure 1D shows a representative image analysis derivation , which shows color ( pixel ) intensity . Using this method , we found that opacification was significantly lower in cyclophosphamide treated animals ( Figure 1E ) . Conversely , fungal dsRed expression was significantly elevated in corneas of cyclophosphamide-treated mice compared with immunocompetent mice at each time point ( Figure 1F ) . There was no detectable red fluorescence in uninfected mice ( Figure S2 ) . To examine fungal viability , eyes were homogenized , and the number of colony forming units ( CFU ) was determined by standard methods . Figure 1G depicts a 1 log-fold reduction in fungal CFU in immunocompetent mice between 48h and 72h post-infection , whereas CFU were not reduced in cyclophosphamide-treated C57BL/6 corneas over this time period . Together , these findings identify a resistant phenotype in immunocompetent animals , where there is a pronounced neutrophil infiltration to the corneal stroma , resulting in controlled fungal growth . These findings also show a susceptible phenotype illustrated by cyclophosphamide treated mice , where neutrophil recruitment to the corneal stroma is impaired , fungal growth continues unabated , and the cornea eventually perforates . The predominant cells in the corneal stroma are keratocytes , which produce collagen and proteoglycans that comprise the extracellular matrix; however there is also a population of resident macrophages and dendritic cells at this site [33] , [34] . To determine if resident macrophages and dendritic cells mediate the initial recognition of A . fumigatus in the cornea , we utilized Macrophage Fas Induced Apoptosis ( Mafia ) mice , which express eGFP and a membrane bound suicide protein under control of the myeloid-lineage specific c-fms promoter [35] , [36] . In these mice , all macrophages and dendritic cells express eGFP constitutively , and undergo cell-lineage specific apoptosis after cross-linking the FK506 binding domain of the membrane-bound suicide protein using the FK506 dimerizer AP20187 [36] . We showed previously that eGFP+ macrophages/dendritic cells are readily identified in the corneas of naïve untreated Mafia mice , and that eGFP expressing cells are depleted in AP20187-treated Mafia mice [35] . To ascertain the role of macrophages and dendritic cells in Aspergillus keratitis , Mafia mice were treated with AP20187 , and infected with 1×105 Af293 . 1RFP conidia . Corneas of untreated Mafia mice developed opacification at 24 and 48h after infection , consistent with increased eGFP+ cellular infiltration to the corneal stroma and dsRed expressing Aspergillus hyphae ( Figure 2A ) . In contrast , corneas of AP20187-treated Mafia mice had decreased opacification , and significantly increased dsRed expressing Aspergillus hyphae ( Figures 2A ) . Cellular infiltration to the corneal stroma was absent in infected AP20187-treated Mafia mice as shown by the absence of eGFP+ cells ( Figure 2A ) and in histological sections ( Figure 2B ) . To determine if the impaired cellular infiltration is related to pro-inflammatory and chemotactic cytokine production , corneas were dissected from untreated and AP20187-treated Mafia mice at 10h after intrastromal injection of Af293 . 1RFP conidia or PBS ( trauma control ) , and prior to detectable cellular infiltration . Corneas were homogenized , and CXCL1/KC and IL-1β production were examined by ELISA . Figure 2C shows that both CXCL1/KC and IL-1β were elevated in untreated , but not AP20187-treated corneas from Mafia mice . Figures 2D–F show image analysis based quantification of corneal opacification , eGFP+ cell infiltration , and RFP expression for individual corneas in the experiment and reveal significantly lower corneal opacification ( Figure 2D ) and eGFP+ cell infiltration ( Figure 2E ) in AP20187-treated Mafia mice . Conversely , dsRed hyphae ( Figure 2E ) and CFU ( Figure 2F ) were significantly higher in corneas of AP20187-treated Mafia mice compared with untreated Mafia mice , indicating that cfms+ cells regulate fungal growth and survival ( Figure 2G ) . Taken together , these data show that resident c-fms+ macrophages and dendritic cells in the naïve cornea produce IL-1β and CXCL1/KC , which likely mediate neutrophil recruitment into the cornea , and subsequently limit Aspergillus growth and survival . As resident cfms+ macrophages and dendritic cells are essential for early cytokine production in the cornea after Aspergillus infection , they are likely the first cells to respond to germinating conidia . We therefore examined the pathogen recognition molecules initiating this response , including Dectin-1 , TLR2 and TLR4 . Dectin-1 is a C-type lectin expressed by myeloid-derived cells that recognizes β-glucan when it is exposed on the cell wall , and β-glucan expression in Aspergillus occurs in germinating , but not dormant conidia , and in hyphal stages [37]–[39] . We examined β-glucan and Dectin-1 expression and activation in the cornea after Aspergillus infection . As shown in Figure 3A , PASH stained corneas revealed swollen conidia ( 6h ) and hyphae ( 24h ) ; further , β-glucan expression was apparent in both forms , especially at 24h when hyphal forms predominated , indicating that Aspergillus expresses the ligand for Dectin-1 during corneal infection . To determine the effect of Aspergillus on Dectin-1 protein levels in the cornea , we dissected corneas 10h after infection or after injection with PBS ( trauma controls ) , and processed the corneas for western blot analysis . Dectin-1 was not detected in naïve and PBS-injected corneas; however , as early as 10h after Aspergillus infection , Dectin-1 was clearly expressed in the cornea ( Figure 3B ) Further , phosphorylated Spleen Tyrosine Kinase ( Syk ) was elevated in infected corneas , compared with naïve and trauma controls ( Figure 3C ) . Given that Syk phosphorylation is indicative of Dectin-1 activation [40] , these findings indicate that not only is Dectin-1 expressed in the cornea , but that it is also activated during Aspergillus infection . As we found β-glucan expression in vivo , and increased Dectin-1/pSyk activation in infected corneas , we next examined the role of Dectin-1 during Aspergillus keratitis . Dectin-1−/− and control 129SvEv mice were injected intrastromally with Af293 . 1RFP conidia , and corneal opacification , cellular infiltration , and fungal survival were measured as before . As shown in Figure 4A , corneal opacification was evident in 129SvEv corneas by 24h post-infection , increased at 48h and decreased at 72h , whereas corneal opacification was significantly lower in Dectin-1−/− mice at 24h and 48h . However , there were no significant differences between Dectin-1−/− and control , 129SvEv corneas in fungal dsRed expression ( Figure 4A ) . To determine the role of Dectin-1 in cytokine production and cellular infiltration to the cornea , Dectin-1−/− and 129SvEv mice were infected as before , eyes were processed for histology , and 5µm sections were stained with PASH . As shown in Figure 4B , cellular infiltration in 129SvEv mice was similar to infected C57BL/6 mice , with pronounced cellular infiltration at each time point . In contrast , Dectin-1−/− corneas had impaired cellular infiltration at each time point . To examine if the decreased cellular infiltration in Dectin-1−/− mice was associated with pro-inflammatory and chemotactic cytokine production , corneas were dissected and homogenized at 10h post-infection ( prior to detectable cellular infiltration ) , and CXCL1/KC and IL-1β were measured by ELISA . Consistent with impaired cellular infiltration , infected Dectin-1−/− corneas had significantly less CXCL1/KC and IL-1β compared with 129SvEv mice ( Figure 4C ) . Quantification by image analysis shows elevated corneal opacification in 129SvEv compared with Dectin-1−/− mice 24h and 48h after infection ( Figures 4D ) , but no difference in fungal RFP expression ( Figures 4E ) . Consistent with the latter observation , there were no differences in fungal CFUs between Dectin-1−/− and 129SvEv mice at any time point examined ( Figure 4F ) . These findings demonstrate that Dectin-1 regulates cytokine production in the cornea , cellular infiltration to the corneal stroma and development of corneal opacification; however , Dectin-1 expression was not required to regulate growth and survival of strain Af293 . 1RFP . Failure to detect increased survival of Af293 . 1RFP in Dectin-1−/− mice , despite decreased cellular infiltration at all stages post-infection , led us to hypothesize that fungal survival in Dectin-1−/− mice is dependent on the virulence of the infecting A . fumigatus strain . We therefore infected Dectin-1−/− and 129SvEv mice with a clinical isolate from a patient with fungal keratitis . Injection of Strain Af-BP into the 129SvEv cornea induced corneal opacification at 24h post-infection , which increased after 48 and 72h ( Figure 5A ) . In contrast to strain Af293 . 1RFP , the Af-BP isolate also caused engorgement of limbal blood vessels after 48h , and hemorrhage after 72h post-infection ( Figure 5A ) , and may be reflective of the characteristic angio-invasiveness of clinical A . fumigatus isolates [41] . Dectin-1−/− mice exhibited impaired cellular infiltration ( Figure 5B ) , and had significantly lower corneal opacification at 24 and 48h post-infection with Strain Af-BP ( Figure 5C ) , which was similar to infection with strain Af293 . 1RFP . However , in contrast to infection with Af293 . 1RFP , fungal CFUs recovered from infected corneas were significantly elevated in Dectin-1−/− compared with 129SvEv mice at 48 and 72h post-infection ( Figure 5D ) . There were no apparent structural differences between naïve Dectin-1−/− and wild type corneas ( data not shown ) . To determine if Dectin-1 regulation of Af-BP but not Af293 . 1RFP survival is related to β-glucan surface expression , we cultured strains Af293 . 1RFP , Af-BP , and the lung isolate B-5233 in SDA media for 6h or 10h , and examined β-glucan expression as described above . β-glucan was not detected on resting conidia of Af293 . 1RFP or the clinical isolates ( data not shown ) . However , at 6 h ( swollen and germinating conidia ) , there was significantly higher surface β-glucan expression in the clinical isolates Af-BP and B-5233 compared with laboratory strain Af293 . 1RFP ( Figure 5E , F ) . β-glucan surface expression on hyphae after 10h growth was also quantified; as shown in Figure 5G , H , there were no significant differences among these strains at this time point . These findings indicate that strains with high surface β-glucan expression on germinating conidia are more likely to be detected by Dectin-1 and to induce a more pronounced cellular infiltrate , resulting in increased fungal killing . Since macrophage depleted Mafia mice and Dectin-1−/− mice exhibited impaired cellular recruitment to the cornea during Aspergillus infection , we next examined the role of Dectin-1 in activation of bone marrow derived macrophages ( BMMs ) by A . fumigatus . Af-BP swollen conidia were isolated after 6h incubation with Sabouraud dextrose media , fixed , and incubated with 129SvEv or Dectin-1−/− BMMs at an MOI = 100 . After 15 , 30 , and 60 min , cells were lysed , and Syk and Iκb phosphorylation were detected by western blot analysis , and NFκB nuclear translocation was examined after incubation with anti-p65 antibody . Figure 6A shows Syk phosphorylation in 129SvEv BMMs after 15 min incubation , which was sustained for 30 and 60 min . In contrast , p-Syk was not detected in Dectin-1−/− BMMs until 60 min post-exposure . P-Syk was not detected in naïve BMM or after LPS stimulation . Similarly , IκB phosphorylation was elevated in 129SvEv BMMs after 15 min incubation with swollen conidia compared with naïve BMM; however , elevated P-Syk was not detected in Dectin-1−/− BMMs until 60 min incubation ( Figure 6A ) . Additionally , A . fumigatus-induced CXCL1/KC production by BMMs from 129SvEv mice was significantly higher than BMMs from Dectin-1−/− mice ( Figure 6B ) . As shown in Figure 6C , D , NF-κB p65 nuclear translocation was apparent in 129SvEv BMM after 30 min and 60 min , but not in Dectin-1−/− BMMs within this time period . As cell-associated conidia were detected in the p65 translocation experiments , and Dectin-1 – mediated phagocytosis has been shown to inhibit cell signaling [42] , we examined the effect of Dectin-1 on cell associated conidia . BMM were incubated with swollen conidia at 100 MOI as described above , then incubated with Calcofluor white to identify conidia [43] . Cell association was examined by DIC and fluorescence microscopy , and quantified by direct counting . As shown in Figures 6E-G , Dectin-1−/− BMM had significantly less conidia associated with each macrophage than 129SvEv BMMs ( Figure 6F ) . Similarly , the percent of total Dectin-1−/− BMM macrophages with associated conidia was also significantly lower than 129SvEv BMMs ( Figure 6G ) . Together , these data indicate that A . fumigatus conidia bind Dectin-1 on resident corneal macrophages , which then stimulates Syk and NFκB–dependent production of CXCL1/KC that is important for neutrophil recruitment to the cornea . To determine the role of TLR2 and TLR4 in Aspergillus keratitis , C57BL/6 , TLR2−/− and TLR4−/− mice were injected intrastromally with Af293 . 1RFP conidia , and corneal opacification , cellular infiltration , fungal growth and fungal survival were measured as described above . As depicted in Figure 7A–C , there were no significant differences between TLR2−/− and C57BL/6 mice in any of these parameters , indicating that TLR2 has no role in corneal infection with these organisms . Similar results were obtained with clinical isolate Af-BP ( data not shown ) . Figure 7D , E illustrate that as with TLR2−/− mice , there were no significant differences in corneal opacification , cellular infiltration or fungal RFP expression between TLR4−/− and C57BL/6 corneas; however , in marked contrast to TLR2−/− mice , significantly more CFU were recovered from TLR4−/− mice after 48h compared with C57BL/6 mice ( Figure 7F ) , indicating an impaired ability of TLR4−/− mice to clear the infection , and suggesting a role for TLR4 in fungal killing . Increased CFU was also detected in TLR4−/− corneas infected with the Af-BP clinical isolate at 48h and 72h after infection ( Table 1 ) , indicating that the role for TLR4 is consistent among Af strains . Despite the difference in CFU , there was no difference in cellular infiltration between TLR4−/− and C57BL/6 and TLR2−/− corneas ( Figure 7G , H ) , nor any differences in fungal RFP expression ( Figure 7I , J ) . Given the role for TLR4 in fungal viability , we also examined the role of TLR4 co-receptor MD-2 , which binds lipid A of Gram negative bacteria . Corneas of C57BL/6 , TLR4−/− , and MD-2−/−corneas were infected with Af293 . 1RFP , and CFU were quantified after 24 , 48 , and 72h . As before , CFU recovered from corneas of TLR4−/− mice were significantly higher than C57BL/6 at 72h post-infection Figure 7K . However , despite the documented role for MD-2 in responding to LPS , there was no difference in fungal survival in MD-2−/− compared with C57BL/6J mice , indicating that the role of TLR4 fungal killing is MD-2 independent . Given that fungal survival is dependent on TLR4 , and TLR4 signaling involves the adaptor molecules MyD88 , MAL/TIRAP and TRIF ( MyD88: myeloid differentiation primary-response gene 88; TIRAP: toll-interleukin 1 receptor domain containing adaptor protein , which is also called MAL:MyD88-adaptor-like protein; TRIF: TIR-domain-containing adaptor protein inducing IFNβ [44] ) , we next examined the role of these adaptor molecules in cellular infiltration and fungal killing . C57BL/6 , MyD88−/− , TIRAP−/− and TRIF−/− mice were injected intrastromally with Af293 . 1RFP conidia , and markers of infection were examined as before . Figure 8A shows that MyD88−/− mice had less corneal opacification scores at 24h compared with C57BL/6 mice , but not 48h after infection; conversely , MyD88−/− corneas also had increased fungal RFP ( Figure 8A ) . We found impaired cellular infiltration in the corneal stroma of MyD88−/− mice compared with C57BL/6 mice at 24h ( Figure 8B ) ; however , at 48h , there was an intense cellular infiltration into the MyD88−/− corneas . Image analysis shows significantly lower corneal opacification ( Figure 8C ) , but higher fungal RFP expression ( Figure 8D ) in MyD88−/− mice versus C57BL/6 mice . Consistent with the latter observation , Figure 8F shows significantly increased fungal CFU at 48h post-infection in MyD88−/−mice . These data demonstrate that MyD88 regulates early cellular infiltration and fungal survival in Aspergillus keratitis . Further , even though cellular infiltration is detected in MyD88−/−mice after 48h , there was no difference in CFU , indicating a role for MyD88 on the ability of neutrophils and infiltrating macrophages to kill Aspergillus . To determine if MyD88 is due to TLR4 signaling , we infected mice deficient in TIRAP , which is essential for MyD88 signaling by TLR4 . We also infected mice deficient in TRIF , which mediates the TLR4 ( and TLR3 ) , MyD88-independent signaling pathway [45] . Figure 9A shows that TRIF−/− and TIRAP−/− mice develop corneal opacification that is not different from C57BL/6 mice . Similarly , the presence of dsRed Aspergillus in the corneas of TRIF−/− and TIRAP−/− mice was similar to C57BL/6 mice . Image analysis also showed no difference in opacity or dsRed expression , respectively , between C57BL/6 , TRIF−/− , and TIRAP−/− mice ( Figure 9B , C ) . In addition , there was no difference in fungal CFU among the three strains at either 24h or 48h ( Figure 9D ) or in cellular infiltration among these strains ( Figure 9E ) . These findings indicate that although MyD88 has a critical role in Aspergillus keratitis , there is no apparent involvement of TIRAP or TRIF . The role of MyD88 , but not TIRAP in Aspergillus keratitis led us to conjecture that the IL-1 receptor ( IL-1R1 ) , which signals through MyD88 independently of TIRAP , could mediate cellular recruitment during Aspergillus keratitis . To test this hypothesis , we injected 1×105 A . fumigatus conidia into the corneas of C57BL/6 and IL-1R1−/− mice . Figure 10A shows that IL-1R1−/− mice had significantly lower corneal opacification and increased fungal RFP expression and survival at 24 and 48h post-infection . Conversely , there was significantly less cellular infiltration into the corneal stroma of IL-1R1−/−mice compared with C57BL/6 mice ( Figure 10B ) . Image analysis revealed decreased corneal opacity at 24h ( Figure 10C ) and increased fungal dsRed expression at 48h post-infection in IL-1R1−/− mice compared to C57BL/6 mice ( Figure 10D ) . Correspondingly , Figure 10E shows increased fungal CFU 48h post-infection in IL-1R1−/− mice compared to C57BL/6 mice . Taken together , these observations indicate that IL-1R1 mediates early cellular infiltration and fungal survival in Aspergillus keratitis , which is a similar phenotype to MyD88−/− mice .
Aspergillus is a major cause of visual impairment and blindness worldwide; however , the nature of the host response to these organisms in the cornea is not well understood . Our findings show that critical components of the innate immune response in the cornea include c-fms+ macrophages and dendritic cells in addition to Dectin-1 , TLR4 , MyD88 , and IL-1R1 . We also showed that there was no role for MD-2 , TLR2 , TIRAP or TRIF . Taken together , these observations are consistent with a sequence of events that is initiated by expression of β-glucan on germinating conidia in the corneal stroma , and recognition by Dectin-1 expressed on resident corneal macrophages and dendritic cells . Dectin-1 mediated activation of p-Syk , p-IκB , and translocation of NFκB to the nucleus of these cells results in production of CXCL1/KC and IL-1β within 10h of infection , and recruitment of neutrophils to the corneal stroma . Neutrophils kill Aspergillus by attaching to hyphae and releasing cytotoxic proteases , antimicrobial peptides , and reactive oxygen species [46] . Although this inflammatory response causes corneal opacification , the corneas eventually heal , leaving scarified tissue . However , blockade of this response at any of these stages allows the organisms to grow unimpaired , resulting in corneal perforation . Germinating Aspergillus conidia encounter a dense , highly organized matrix in the corneal stroma , with anti-parallel layers of collagen separated by keratan sulfate proteoglycans that are essential for corneal transparency; however , the growing hyphae migrate through the stromal matrix and penetrate the basement ( Descemet's ) membrane of the corneal endothelium , which forms the barrier to the anterior chamber . Unless killed by neutrophils entering the anterior chamber from iris vessels ( seen as a hypopyon in infected individuals ) , hyphae can penetrate the posterior eye and cause endophthalmitis , at which point enucleation of the infected eye is often indicated . Although the cornea was long considered to be an immune privileged tissue , it has been well established that macrophages and dendritic cells are resident in the corneal stroma and epithelium [47]–[52] . Recently , a role for macrophages in mediating Fusarium and Candida keratitis was identified [53] . Similarly , in the present study we identified c-fms+ resident macrophages and dendritic cells as essential mediators of cellular recruitment and fungal survival into the cornea during A . fumigatus infection , implicating bone marrow derived cells rather than corneal epithelial cells or fibroblasts as the cellular mediators of innate immune recognition [53] . In the current study , we found that β-glucan is not only expressed on Aspergillus germinating conidia and hyphae in vitro as shown previously [37]–[39] , but is also expressed during corneal infection . Furthermore , Syk phosphorylation was detected in infected corneas , and Syk and IκB were phosphorylated in bone marrow macrophages in a Dectin-1 dependent manner . Correspondingly , both NFκB translocation to the nucleus , and CXCL1/KC and IL-1β production were significantly lower in Dectin-1 deficient compared with control macrophages . Similarly , Dectin-1−/− corneas had impaired pro-inflammatory cytokine production and cellular infiltration compared with control corneas . These findings indicate that expression of Dectin-1 on resident corneal macrophages is essential for the initial innate immune recognition of A . fumigatus , for the subsequent pro-inflammatory cytokine response , and ultimately for cellular recruitment into the cornea . We also found that the physical association of swollen conidia with bone marrow macrophages , which includes attachment and phagocytosis , is dependent on Dectin-1 expression . This observation is consistent with reports showing Dectin-1 dependent phagocytosis of β-glucan coated particles [42] , [54] , and that phagocytosis inhibits Dectin-1 signaling [42] , indicating that macrophage interactions with hyphae that are too big to be ingested may have enhanced signaling compared with conidial forms . Our findings also indicate a possible role for Dectin 1 in early killing of swollen conidia and germ tubes rather than hyphae , where we found no differences in β-glucan staining among the A . fumigatus strains . Taken together , it is likely that during corneal infection , macrophages and neutrophils utilize Dectin-1 to bind conidia and germ tube developmental stages of A . fumigatus , which can be phagocytosed and presumably killed . Results from the current study are consistent with reports on pulmonary aspergillosis in which germinating Aspergillus conidia in the lungs express β-glucan , and Dectin-1 mediates cellular infiltration and fungal killing [38] , [39] . Our data are also in agreement with a recent study showing that Dectin-1−/− mice have lower cytokine and chemokine production in the lungs after intratracheal infection , resulting in impaired neutrophil recruitment and increased susceptibility to Aspergillus [56] . The role of Dectin-1 in responding to Aspergillus therefore appears to be conserved in the cornea and lungs . Although studies with fungal cell wall components indicate that Dectin-1 and TLR2 collaborate in recognizing β-glucan [57]–[59] , we found no role for TLR2 in Aspergillus keratitis , indicating that Dectin-1 mediates cellular recruitment independently of TLR2 . Additionally , TLR2-independent receptor collaboration occurs in a sequential manner during non-opsonic phagocytosis of C . albicans by macrophages , where Dectin-1 , CR3 , and mannose receptors collaborate in recognition and phagocytosis [60] . In the current study , phagocytosis of swollen A . fumigatus conidia was impaired in Dectin-1−/− bone marrow macrophages , raising the possibility that a similar collaboration occurs during macrophage phagocytosis of this pathogen . In support of this notion , the mannose receptor was one of the few genes found to be upregulated in a microarray analysis of Aspergillus –infected corneas [61] . Future studies will determine if a similar mechanism occurs for macrophage phagocytosis of Aspergillus conidia . In pulmonary aspergillosis , TLR2 and TLR4 have no role in otherwise immunocompetent animals , although Aspergillus CFU were elevated in vinblastin- or cyclophosphamide-treated TLR2−/−and TLR4−/− mice [62] , [63] . In contrast , we showed increased Aspergillus CFU in TLR4−/− , but not TLR2−/− mice . Although this is consistent with our earlier observation that TLR4−/− mice have impaired clearance of Fusarium in the cornea [64] , the mechanism of TLR4 regulation of Aspergillus survival has yet to be determined . Given that TLR4−/− mice exhibit defects in fungal killing but not cellular recruitment , we predict that the primary role for TLR4 is on infiltrating neutrophils and macrophages that mediate fungal killing rather than on resident macrophages , which regulate cell recruitment to the cornea . Further , as fungal clearance is unimpaired in MD-2−/− corneas , and MD-2 is the binding site for the lipid A moeity of LPS [65] , [66] , TLR4 likely recognizes Aspergillus hyphae at an MD-2 independent site of the receptor . TLR4 recognizes C . albicans o-linked mannosyl residues [59] and Cryptococcus neoformans glucuronoxylomannan [67] , although the ligand on Aspergillus hyphae has yet to be identified . In addition to being the receptor for LPS , MD-2 mediates TLR4 dimerization and cell signaling [44] , [68]; therefore , it is possible that TLR4 recognition of the fungal cell wall does not induce signaling . In support of this notion , we found that the absence of TIRAP or TRIF did not affect cellular infiltration or fungal survival . We also noted that although TLR4−/− mice have elevated Aspergillus CFU , there was no difference in RFP expression between TLR4−/− and C57BL/6 mice . RFP quantification measures total fungal load , but not viability , and CFU and RFP measurements correlated well in most studies; however , in TLR4−/− mice , more RFP+ hyphae appear to be viable , possibly due to production of fungistatic rather than fungicidal mediators in the tissue , including lactoferrin and lipocalin [69] . In the current study , MyD88−/− mice had delayed cellular infiltration and unimpaired fungal growth , which is similar to the role of MyD88 in pulmonary aspergillosis [63] , [70] , and indicates an essential role for this adaptor molecule in Aspergillus keratitis . However , although TLR4 signaling through MyD88 requires the accessory adaptor molecule TIRAP [44] , we found no detectable effect on the progression of Aspergillus keratitis in the absence of TIRAP . We therefore examined the role of IL-1R1 , which signals through MyD88 in the absence of TIRAP [44] , and showed that IL-1R1−/− mice have a similar phenotype as MyD88−/− mice . As IL-1β is produced in the cornea early after infection , it seems reasonable to assume that the MyD88 dependence is due to IL-1R1 signaling . In the corneal stroma , IL-1R1 is expressed not only by macrophages and dendritic cells , but also by resident keratocytes in the corneal stroma , which can differentiate into fibroblasts and produce pro-inflammatory and chemotactic cytokines , including CXCL1/KC [71]–[73] . IL-1R1 also mediates the host response in pulmonary candidiasis [63] , and we identified a similar role for IL-1R1 and MyD88 in trauma-induced and biofilm-associated Fusarium keratitis [64] , [74] , indicating that the cornea employs similar responses to regulate infection by filamentous fungi . In conclusion , results of the current studies using a murine model of fungal disease demonstrate essential , though distinct roles for Dectin-1 and TLR4 . There also appears to be a role for these receptors in human fungal disease as specific polymorphisms in Dectin-1 and TLR4 genes are associated with susceptibility . A Dectin-1 polymorphism resulting in an early stop codon was associated with reduced β-glucan binding and increased susceptibility to Candida albicans infections [75] , [76] , whereas TLR4 polymorphisms correlate with susceptibility to aspergillosis in recipients of stem cell transplants [77] . Taken together , results from human genetics studies and animal model studies combine to demonstrate that these receptors have an essential role in fungal infection , and are therefore potential targets for immunotherapy that are common to multiple fungal pathogens . Future studies will examine the potential of targeting receptors in prevention and treatment of fungal keratitis .
All animals were treated in accordance with the guidelines provided in the Association for Research in Vision and Ophthalmology ARVO statement for the Use of Animals in Ophthalmic and Vision Research , and were approved by Case Western Reserve University IACUC . C57BL/6 mice ( 6–12 wk old ) , and IL-1R1−/− mice on a C57BL/6 background were purchased from The Jackson Laboratory ( Bar Harbor , ME ) . 129SvEv mice were purchased from Taconic Farms ( Hudson , NY ) . Dectin 1−/− mice on a 129SvEv background were provided by Dr . Gordon Brown , University of Aberdeen , UK , MD-2−/− mice were provided by Dr K . Miyake , University of Tokyo , and TLR2−/− , TLR4−/− , TRIF−/− , and MyD88−/− mice on a C57BL/6 background were provided by Dr . Shizuo Akira , Osaka University , Osaka , Japan , TIRAP−/− mice were provided by Dr Ruslan Medzhitov , Howard Hughes Medical Institute , Yale University , NewHaven , CT . MAcrophage Fas-Induced Apoptosis ( Mafia ) mice were obtained from Sandra Burnett , Dept . of Microbiology and Molecular Biology , Brigham Young University , Provo , UT [36] . Aspergillus fumigatus strains used in this study were cultured on Vogel's Minimal Media ( VMM ) w/wo 4% agar +/− supplementation with 10 mM uracil and 5 mM uridine at 37°C/5% CO2 unless otherwise stated . A . fumigatus Strain Af-BPis a fungal keratitis clinical isolate from a patient treated at Bascom Palmer Eye Institute ( Miami , FL ) provided by Dr . Darlene Miller . A . fumigatus Strain B-5233 is a clinical strain isolated from the lungs of a patient with severe neutropenia provided by Dr . Kwon-Chung ( NIAID ) [78] . The uracil auxotroph strain Af293 . 1 ( ΔpyrG1 ) [79] was a gift from Gregory May ( M . D . Anderson Health Science Center , Texas ) . Complementation of pyrG1 and constitutive RFP expression by strain Af293 . 1 was achieved via transformation with the plasmid pRG3AMA1-RFP forming strain Af293 . 1RFP ( Figure S1 ) . All primers used in this study are listed in Supplementary Figure S1A . The first step in the construction of Af293 . 1RFP , entailed PCR amplification of the glyceraldehyde 3 phosphate dehydrogenase promotor ( gpdA ) on the plasmid pDV2 [80] using a forward primer bearing a 5′ Kpn1 overhang ( Kgpd-F ) and a reverse primer bearing a 5′ complementary DNA sequence to the improved monomeric dsRed fluorescent protein- encoding gene ( rfp ) , expressed on the plasmid pMT-RFP ( Rgpd-R ) [81] , [82] . Next , the rfp gene on pMT-RFP was PCR amplified using a forward primer bearing a 5′ complementary sequence to the 3′ end of the gpdA promotor on pDV2 ( Grfp-R ) and a reverse primer harboring a 5′ Kpn1 cut site overhang ( Krfp-R ) . Both PCR amplicons ( gpdA promotor = 1084 Af-BP; rfp = 769 Af-BP ) underwent fusion PCR , using primers Kgpd-F and Krfp-R to yield a single fusion PCR amplicon gpdA:rfp ( Af-BP ) . Lastly , Kpn1-digested gpdA:rfp was ligated to Kpn1-digested pRGAMA-1 yielding an 11 . 8 kb plasmid ( Figure S1B ) . The resulting plasmid , pRG3AMA1-RFP , was transformed into the uracil auxotroph strain , Af293 . 1ΔpyrG according to the method described by May [83] . Figure S1C shows that the resulting strain , Af293 . 1RFP , constitutively expressed dsRED fluorescence ( 580 nm ) at each morphological stage of the organism . Additionally , there was no difference in the growth kinetics of this strain with the parental Af293 . 1 strain ( data not shown ) . A . fumigatus strains were cultured for 2–3 days on VMM without uracil or uridine in 25 cm2 tissue culture flasks . Fresh conidia were disrupted with a bacterial L-loop and harvested in 5 ml PBS . Pure conidial suspensions were obtained by passing the culture suspension through sterile PBS-soaked cotton gauze positioned at the tip of a 10 ml syringe . Conidial suspensions were quantified using a haemacytometer , and adjusted to a final concentration of 5×104 conidia/µl in PBS . Mice were anaesthetized by intraperitoneal ( IP ) injection of 2 . 25 mg ketamine and 0 . 45 mg xylazine , and the corneal epithelium was abraded using a 30-gauge needle . Through the abrasion was inserted a 33-gauge Hamilton syringe from which a 2 µl injection containing 1×105 conidia ( Optimal inoculum size for induction of keratitis based on preliminary studies; data not shown ) was released into the corneal stroma . Mice were examined daily under a stereomicroscope for corneal opacification , ulceration , and perforation . At set time points , animals were euthanized by CO2 asphyxiation , and eyes were either placed in 10% formalin and embedded in paraffin and sectioned at 5 µm intervals , or excised and placed in 1 ml of sterile saline and homogenized for quantitative culture . For cyclophosphamide immunosuppression , mice were given 180 mg/kg cyclophosphamide ( Sigma-Aldrich ) via I . P . injection at days 3 and 1 prior to infection [84] . All animals were bred under specific pathogen-free conditions and maintained according to institutional guidelines . MAcrophage Fas-Induced Apoptosis ( Mafia ) mice are C57BL/6J mice harboring an eGFP and suicide-protein expressing transgene downstream of the macrophage and dendritic cell lineage-specific c-fms promoter[36] , [85] . As such all MACs/DCs in these mice , constitutively express eGFP and a membrane-bound suicide protein . To deplete c-fms expressing cells , Mafia mice received five daily ip injections ( 10 mg/kg ) of the covalently-linked dimer ( AP20187; Ariad Pharmaceuticals ) followed by two days rest before the experiment . AP20187 has affinity for the FK506 binding protein region of the suicide protein; and homodimerization of the suicide proteins activates the intramolecular cytoplasmic Fas domains , with the subsequent induction of caspase-8 mediated apoptosis in all MACs/DCs . Our recent study showed that this protocol results in loss of viable eGFP+ cells in the cornea [36] , [85] . Mice were sacrificed by CO2 asphyxiation and positioned in a thee-point stereotactic mouse restrainer . Corneal opacity ( Brightfield ) , fungal proliferation ( RFP; 580nm ) and cellular infiltration ( eGFP; 488nm ) , were visualized in the intact cornea using a high-resolution stereo fluorescence MZFLIII microscope ( Leica Microsystems ) and Spot RT Slider KE camera ( Diagnostics Instruments ) . In some experiments , corneas were dissected and examined using an inverted Leica DMI 6000B microscope . All images were captured using SpotCam software ( RT Slider KE; Diagnostics Instruments ) . To quantify corneal opacity objectively , brightfield images of mouse corneas were analyzed using Metamorph software ( Molecular Devices ) . Briefly , a constant circular region encompassing the cornea was defined , and the pixel intensity within this region summed to yield a numerical value , called the pixel intensity corresponding to the total amount of light reflected from the cornea ( i . e . opacity ) . Similarly , fungal dsRed RFP and c-fms+ cell eGFP expression were quantified via Metamorph image analysis . All images were obtained with the same Spot RT Slider KE camera using the same Spot Advanced Software under the same magnification , exposure ( BF = 0 . 4s; RFP = 10s; eGFP = 2s ) , gain ( BF = 1; RFP/eGFP = 16 ) , and gamma ( BF/RFP/eGFP = 1 . 85 ) parameters . For assessment of fungal viability , whole eyes were homogenized under sterile conditions in 1 ml PBS , using the Mixer Mill MM300 ( Retsch , Qiagen , Valencia , CA ) at 33 Hz for 4 min . Subsequently , serial log dilutions were performed and plated onto bacteriologic-grade Sabouraud dextrose agar plates ( Becton Dickenson ) . Following incubation for 24h at 37°C , the number of CFUs was determined by direct counting . Eyes were enucleated and fixed in 10% formalin in PBS ( Fisher ) for 24h . Tissues were then dehydrated in graded ethanol concentrations at room temperature ( 65% 1× , 80% 2× , 95% 1× , 100% 3×; 1 h for each change of solution ) , followed by three 1h changes of xylene , and 4 changes of paraffin at 60°C under 15mm Hg vacuum to remove air bubbles . Five µm sections from the center of the cornea ( as determined by noncontiguous iris morphology ) were cut and stained with Periodic-Acid Schiff ( PASH ) for identification of fungi and inflammatory cell recruitment . To detect β-glucan expression during live corneal infection , 5-µm paraffin sections were deparaffinized . Slides were blocked with 1 . 5% normal rabbit serum in PBS for 1h , then incubated with primary mouse anti-fungal β-glucan IgM ( BF-Div; Biothera/Brown Univ . ) diluted to 24 µg/ml with 1% BSA ( Fisher ) for 1h at 37°C . The slides were then washed 3× in PBS plus 0 . 05% Tween 20 ( PBS-T; Sigma ) and incubated with alexafluor-488 rabbit-anti-mouse IgM ( Invitrogen ) diluted to 1µg/ml in PBS for 1h at 37°C . The slides were washed 3× in PBS-T and imaged by fluorescence microscopy ( magnification , 400× ) . A similar process was performed to stain 5µm corneal sections for neutrophils using monoclonal rat-anti-mouse neutrophil IgG ( NIMP-R14 , AbCam , Cambridge , MA ) , and alexafluor-488 tagged rabbit-anti-rat IgG ( Invitrogen ) . A . fumigatus strains were cultured for 3 days in VMM+4% agar . Pure conidial suspensions were prepared from the 3-day culture , and 5 million conidia were added to 50 ml SDA broth in vented-cap 250 ml tissue culture flasks . The conidia were grown at 37°C/5% CO2 for 0 , 6 , and 10h . At indicated timepoints , A . fumigatus was fixed with 4% paraformaldehyde for 30 min , washed 3× with PBS and spun onto Superfrost microscope slides ( Fisher ) using a Cytospin centrifuge . Subsequently , slides were blocked with 1 . 5% normal rabbit serum in PBS for 1h , then incubated with primary mouse anti-fungal β-glucan IgM ( BF-Div; Biothera/Brown Univ . ) , diluted to 24 µg/ml with 1% BSA ( Fisher ) for 1 h at 37°C . The slides were then washed 3× in PBS-T and incubated with alexafluor-488 rabbit-anti-mouse IgM ( Invitrogen ) diluted to 1µg/ml in PBS for 1h at 37°C . The slides were washed again 3× in PBS-T and imaged by fluorescence microscopy ( magnification , 400× ) . Subsequently , images were analyzed using Metamorph software and the numerical output of average 488nm pixel intensity/area of fungal cells was used to quantify fungal β-glucan surface expression . To minimize the confounding variable of enhanced fluorescence due to cell clumping , only 488nm fluorescence emanating from isolated fungal cells was used for quantitation . Cornea protein extracts were prepared by homogenization in cell lysis buffer ( Cell Signaling ) plus 1mM phenylmethylsulphonyl-fluoride ( PMSF ) using a Tissue-Lyser ( described above ) . Similarly , cell culture protein extracts were prepared via lysis in cell-lysis buffer+PMSF . Total protein was quantified via Bicinchoninic Acid Assay ( Pierce ) , denatured with 2× Laemmli buffer ( Sigma ) and heated to 95°C for 5 min . 20 µg total protein was loaded onto each well of a 10% polyacrylamide gel ( BioRad ) , separated via electrophoresis , and transferred to nitrocellulose . Blots were stained accordingly with anti-Dectin-1 ( RandD; MAB1756 ) , anti-phosphoSyk ( Cell Signaling; 2710 ) , anti-pIκB ( Cell Signaling; 2859 ) , and anti-β-actin ( Cell Signaling; 4967 ) . HP-tagged secondary antibodies were purchased from Santa Cruz Biotechnologies . All blots were developed with GE Healthcare ECL Western Blotting Detection Reagent ( Amersham ) or Supersignal West Femto Maximum Sensitivity Substrate ( Pierce ) . Mice underwent euthanasia by CO2 asphyxiation , and femurs and tibias from 129SvEv and Dectin-1−/− mice were removed , cleaned , and centrifuged at 5000×g for 45 s at 4°C . Any contaminating red blood cells were lysed in 5 ml RBC Lysis Buffer ( eBioscience ) , and remaining bone marrow cells were cultured in bacteriologic grade petri dishes in 6 ml Macrophage Growth Medium ( MGM:DMEM w/L- Glutamine , Na-Pyruvate , HEPES , 10% FBS , P/S , 30% L929 cell conditioned medium ) . On day 5 , and every 2 days thereafter , the cell supernatant was aspirated , and fresh MGM media was added . Adherent cells were harvested between 7–14 days of culture , and counted . 2×104 cells were cultured onto sterile 18mm2 coverslips ( Corning ) in a 6 well-plate , and treated with LPS ( 100 ng/ml; positive control ) , or 6h A . fumigatus strain Af-BP swollen conidia ( MOI = 100 ) for 15 , 30 and 60 min . Following activation , BMM were fixed with 4% paraformaldehyde for 15 min at room temperature , permeabilized using 0 . 1% Triton X-100 in PBS for 1 min at RT , and incubated with rabbit anti-mouse p65 ( 1∶100; eBioscience Ltd ) in PBS containing 10% goat serum for 1h at RT . Coverslips were washed 2× with PBS and cells were incubated with Alexa Fluor 488-labeled goat anti-rabbit IgG antibody ( Molecular Probes Inc . ) in PBS at RT for 1h and washed 2× with PBS . The cells were mounted on glass slides using Vectashield mounting medium with DAPI ( Vector Laboratories , UK ) , and examined by fluorescence microscopy , and average pixel intensity ( green fluorescence ) per nucleus ( DAPI ) was calculated using Metamorph . For cell association studies , bone marrow macrophages incubated with A . fumigatus strain Af-BP swollen conidia were washed and fixed in 4% paraformaldehyde as described above , then incubated 5 min with Calcufluor white ( Sigma ) at a 1∶1 ratio with 10% KOH . After washing 2× with PBS , cells were examined by DIC and fluorescence microscopy , and associated conidia per 100 cells , and the percent cells with associated conidia was determined after direct examination of at least 300 cells per coverslip . Two coverslips were examined , and the mean and SD were calculated . Corneas were homogenized in 150 µl Reagent Diluent ( R & D Systems , Minneapolis , MN ) using the Retsch MM 300 ball miller at 33 Hz for 4 min ( Qiagen ) . For analysis of bone marrow macrophage cytokine production , cell culture supernatants were obtained and assayed directly . Half-well cytokine assays were performed using Duoset ELISA kits ( R & D Systems ) according to the manufacturer's directions . Statistical analysis was performed for each experiment using an unpaired t test ( Prism , GraphPad Software ) . A p value<0 . 05 was considered significant . | Corneal infection with filamentous fungi , including Aspergillus species , is a common cause of visual impairment and blindness in the southern USA and worldwide . The incidence in India and China greatly increases during harvest season when infection occurs after traumatic injury with fungal spores ( conidia ) . In contrast to pulmonary aspergillosis , keratitis occurs in immunocompetent individuals . To characterize the host response , we injected Aspergillus fumigatus conidia expressing red fluorescence into the transparent mouse cornea , and showed that cytokine production , neutrophil and monocyte recruitment to the corneal stroma and fungal killing is dependent on the presence of macrophages and dendritic cells , and on expression of the β-glucan receptor Dectin-1 . We also found that fungal killing , but not cellular infiltration , is dependent on expression of the LPS receptor TLR4 . In addition , we demonstrate that IL-1R1 and MyD88 regulate neutrophil recruitment and fungal killing in Aspergillus keratitis , whereas TLR4 associated adaptor molecules TRIF and TIRAP have no role . Together , these findings identify specific mediators of the innate immune response to these organisms that regulate disease severity and survival of Aspergillus that may have potential application as targets for therapeutic intervention . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/fungal",
"infections",
"immunology/leukocyte",
"activation"
] | 2010 | Distinct Roles for Dectin-1 and TLR4 in the Pathogenesis of Aspergillus fumigatus Keratitis |
The diverse , specialized genes present in today’s lifeforms evolved from a common core of ancient , elementary genes . However , these genes did not evolve individually: gene expression is controlled by a complex network of interactions , and alterations in one gene may drive reciprocal changes in its proteins’ binding partners . Like many complex networks , these gene regulatory networks ( GRNs ) are composed of communities , or clusters of genes with relatively high connectivity . A deep understanding of the relationship between the evolutionary history of single genes and the topological properties of the underlying GRN is integral to evolutionary genetics . Here , we show that the topological properties of an acute myeloid leukemia GRN and a general human GRN are strongly coupled with its genes’ evolutionary properties . Slowly evolving ( “cold” ) , old genes tend to interact with each other , as do rapidly evolving ( “hot” ) , young genes . This naturally causes genes to segregate into community structures with relatively homogeneous evolutionary histories . We argue that gene duplication placed old , cold genes and communities at the center of the networks , and young , hot genes and communities at the periphery . We demonstrate this with single-node centrality measures and two new measures of efficiency , the set efficiency and the interset efficiency . We conclude that these methods for studying the relationships between a GRN’s community structures and its genes’ evolutionary properties provide new perspectives for understanding evolutionary genetics .
The evolutionary history of a gene can be mapped in various ways . The absolute evolutionary rate , for example , can be computed from observed differences in orthologs across species in the context of their phylogenetic relationships [1] , whereas the age of a gene can be measured by tracing when the gene first appeared in the organism’s phylogenetic tree [2] . Quantities such as these allow researchers to chronicle the journey of individual genes across evolutionary history . But genes do not exist , and therefore do not evolve , in isolation . Mutations in a transcription factor may affect the expression of the genes it regulates , since changes in a protein’s amino acid sequence can cause it to lose compatibility with former binding partners , and gain compatibility with new partners . Accumulation of these alterations can lead to changes in fitness and , eventually , speciation . The evolution of individual genes is thus coupled with the evolution of the structure of the organism’s gene regulatory network ( GRN ) , and network properties should be related to the evolutionary properties of its constituent nodes and edges . It has been proposed that GRNs grow and evolve incrementally via gene duplication followed by mutation and functional divergence [3–7] , although changes may have occasionally arrived in bursts , as in whole-genome duplication [8] . This time-dependent network formation suggests that GRNs are composed of a core of ancient , conserved genes with fundamental functions , and younger , peripheral genes with species- or cell type-specific function , which mutate frequently until the functions of the newly created pathways are optimized . These mutations can alter GRNs by creating , removing , reassigning , or changing other properties of nodes and edges . Fraser et al . demonstrated that interacting pairs of proteins have similar evolutionary rates [9] . This constraint is likely driven by the necessity of coevolution , since a change in one protein’s sequence may require a corresponding change in its partner’s sequence in order for the pair to remain compatible . Daub et al . showed that genes which are part of many biological pathways have lower evolutionary rates than genes which belong to few or no known pathways , further supporting the idea that related genes share similar evolutionary properties [10] . It has also been shown that evolutionary rates are weakly , but significantly , negatively correlated with degree , closeness centrality , and betweenness centrality ( network measures which quantify the location of individual nodes in different ways ) [11 , 12] , and that essential genes have high centrality and low evolutionary rates [13] . Here , our goal is to establish quantitative relationships between the evolutionary history of genes and their topological properties in an acute myeloid leukemia GRN , AML 2 . 3 [14] , as well as for a general human GRN , HumanNet [15] . In contrast to the earlier studies above , we go beyond an analysis based on single-node centrality and pairwise measures by studying the connection between topology and evolution from the point of view of network community structures . We demonstrate that the evolutionary rates and ages of genes are not randomly distributed across the networks , but are naturally organized in communities with well-defined evolutionary characteristics: old genes cluster with old genes , and young cluster with young . Likewise , “cold genes” ( genes with low evolutionary rates ) cluster with cold genes , and “hot genes” ( genes with high evolutionary rates ) cluster with hot genes . This segregation also exists for groups of enriched genes identified by DAVID [16] within the communities . In terms of network topology , we show that genes and DAVID groups which are old and cold tend to be central , and those which are young and hot tend to be peripheral . We demonstrate this with traditional single-node centrality measures as well as two new network measures , the set efficiency [14] and the interset efficiency , which quantify the mean distance between all nodes within a single set and between two sets , respectively ( see Methods ) . We find that PageRank [17] , a finite-range centrality measure , shows stronger biological significance than degree ( a local measure ) and betweenness centrality ( a global measure ) , and that the set efficiency and interset efficiency correlate strongly with the evolutionary histories of individual genes and DAVID groups .
We have shown that slowly evolving , old genes tend to interact with each other , and frequently evolving , young genes tend to interact with each other , whereas edges between those groups are less common . This naturally creates communities of genes with relatively homogeneous evolutionary attributes . Analyzing the networks in terms of communities and DAVID groups rather than single genes provided a new perspective which allowed us to establish clear relationships between network topology and evolution . The abundance of connections between old DAVID groups and the relative scarcity between old-and-young and young-and-young DAVID groups suggests that during the course of human evolution , the primitive gene regulatory network began as a core of fundamental genes and pathways . As genes duplicated and mutated , novel functions arose and eventually , through selective duplications , deletions , mutations , and rewirings , novel regulatory pathways emerged , growing outward from these ancient genes . This would place the oldest genes near the middle of the network and the youngest genes toward the periphery . These findings were mainly derived from an AML network and a general human network , and they were broadly confirmed in a normal hematopoietic network and are consistent with previous reports [13] . No gene is an island . A real understanding of the evolution of a genome only comes from studying its constituent genes in the context of the underlying complex network of interactions rather than as independent units . As network reconstruction methods continue to improve and more high quality networks become available , we expect to find more evidence of how evolution shapes the topology of gene regulatory networks .
To compute the evolutionary rate ( ER ) of a gene , we first calculated the absolute ER for each amino acid position of the protein it encodes using the method from Kumar et al . [1] . Given the multiple alignment at an amino acid position in 46 species [30] , its ER equals the number of different residues divided by the total evolutionary time span , based on a known phylogenetic tree [1] . The ER of a gene is the average of ERs over all amino acid positions , in units of the number of substitutions per amino acid site per billion years . The ER value ranges from ~0 . 011 ( most conserved ) for LSM2 to ~6 . 928 ( least conserved ) for CDRT15 . Ages , taken from Chen et al . [2] , were estimated from comparing the human genome to the genomes of 13 major clades with origins at different points along the human clade , indexed 0 ( oldest ) through 12 ( youngest ) . A gene’s age was determined by searching for the earliest time at which an orthologous gene appears in an organism which branched from the human clade . One gene regulatory network used in this analysis , “AML 2 . 3” , is a partially directed , weighted acute myeloid leukemia ( AML ) GRN [14] . This network was chosen primarily for its quality . It was constructed from more than 1 , 800 patients across 12 studies from both microarray and RNA-seq gene expression measurements in AML cells . Edges were inferred via gene expression correlation within each study , and each edge was assigned a weight based on the number of times it was detected across all studies . Edge directionality was taken from the TRANSFAC [31] and HIPPIE [32] databases . A second network was built using five studies for healthy hematopoietic stem cells ( HSCs ) . The limited amount of data means that the HSC network is a lower confidence network than AML 2 . 3 . Finally , the network “HumanNet” [15] was built from 21 different methods using diverse data types , including microarray co-expression , databases and mass spectrometry proteomics . A weighted , directed , modularity-based community-finding algorithm was used to divide the genes into communities of various sizes [33] . A spy plot of the adjacency matrix after community sorting is shown in S5 Fig . The ten largest communities , indexed 0 through 9 , were selected for further analysis ( see Tables 1 and 2 ) . The individual communities were then provided to the DAVID functional annotation tool to identify enriched DAVID groups in the communities [16] . The top three distinct enriched DAVID groups with Benjamini values less than 10−4 in each community are also included in Tables 1 and 2 . Communities and DAVID groups in Tables 1 and 2 labeled “cold” and “hot” have significantly lower and higher evolutionary rates ( ERs ) than the network average , respectively . Likewise , groups of genes labeled “old” and “young” are significantly older and younger than the network’s average age , respectively . A one-tailed significance level of p < 10−3 in the difference from the mean was chosen for both ER and age . The Kolmogorov-Smirnov ( KS ) statistic and p-value were also computed for each community and DAVID group to quantify the difference between the distribution of all genes and the distribution of each set of genes . KS statistics are reported in S1A Table . Some example DAVID group distributions are shown in Fig 1A and 1B , and all distributions are shown in S7 and S8 Figs for ERs and ages , respectively . A summary of the ERs and ages of the enriched DAVID groups in Table 1 is shown in Fig 1C . The same analysis was conducted for normal hematopoietic stem cell network built from five studies ( GSE48846 , 2666 , 33223 , 24759 , and 30376 ) using the same method as for AML 2 . 3 , with the data reported in S1A Table . To determine the significance of the correlation between ERs and gene-gene interactions , the difference in evolutionary rates between all gene pairs connected by an edge was computed for AML 2 . 3 as well as for degree-preserving randomizations of AML 2 . 3 . S1 Fig shows the distribution of ( ERj − ERi ) for all gene pairs ( i , j ) which are connected by an edge j→i in AML 2 . 3 ( green distribution ) , as well as for all pairs of genes in one degree-preserving randomization of the same network ( purple distribution ) . Note that the distributions are asymmetric because AML 2 . 3 is a directed network . The real distribution of ER differences has a smaller standard deviation than for the randomized network , meaning that difference in evolutionary rates between interacting genes is small on average , in agreement with Fraser et al . [9] . To quantify the significance of this difference , AML 2 . 3 was randomized 20 , 000 times and the standard deviation of each set of ER differences was recorded , as shown in S2 Fig . This gave a z-score of –96 . 8 for the ER differences in the real network . Since none of the sampled randomized networks had an ER difference width less than that of the real network , an upper limit of 5 . 0 × 10−5 was placed on the p-value . The same procedure was used to find the significance in the age difference between connected genes , which resulted in a z-score of –72 . 0 and an upper limit of 5 . 0 × 10−5 for the p-value ( see S3 and S4 Figs ) . The global efficiency [34] of a network is defined as Eglobal=1n ( n−1 ) ∑i≠j1dij where n is the number of nodes in the network , dij is the distance from node j to node i , and 0 ≤ Eglobal ≤ 1 for unweighted networks . We define the set efficiency ( SE ) of a set of nodes M as EM=1|M| ( |M|−1 ) ∑i , j∈M , i≠j1dij where |M| is the number of nodes in M , and 0 ≤ EM ≤ 1 for unweighted networks . EM > Eglobal implies that nodes in M are closer to each other than average in the network , and EM < Eglobal implies that the nodes are more dispersed than average . Note that dij is calculated using the full network , so shortest paths from j to i may pass through nodes which are not in M . The SE was used to examine the topological distribution of ERs and ages in AML 2 . 3 . The ERs were sorted from coldest to hottest , and the SE of the first 500 genes was computed , increasing the window size in steps of 10 genes from the beginning to the end of the ER list ( i . e . the 500 coldest , 510 coldest , etc . ) . The resulting curve is shown in blue in Fig 3 . As a control , the order of the genes was randomized ( but the underlying network , AML 2 . 3 , remained unchanged ) and the SE was computed for the first 500 genes in the randomized list , then the first 510 genes , etc . in steps of 10 . 100 of these curves were generated . Fig 3 shows the mean of these 100 controls ( solid green line ) plus/minus one standard deviation ( dashed green lines ) . See S6 Fig for the same plot using age rather than ER . We define the interset efficiency ( IE ) from node set J to node set I as EIJ=1|I||J|−|I∩J|∑i∈I , j∈J , i≠j1dij where |I ∩ J| is the number of nodes shared by sets I and J , and 0 ≤ EIJ ≤ 1 for unweighted networks . As with the set efficiency , shortest paths may pass through nodes which are neither in I nor J . Note that this formulation is defined when sets I and J have a non-empty intersection , and that the diagonal terms of the interset efficiency reduce to the set efficiency , i . e . EII = EI . A large EIJ implies that the average distance from nodes in J to nodes in I is small , and a small EIJ implies large distances . EIJ is asymmetric for directed networks . This measure was used in Figs 4 and 5 to quantify the proximity of the DAVID groups from Tables 1 and 2 , respectively . See S4 File for a more detailed explanation of the interset efficiency . | We found strong relationships between the community structures and evolutionary properties of an acute myeloid leukemia gene regulatory network ( GRN ) and a general human GRN . Interacting genes tend to have similar evolutionary ages and rates , causing the GRNs to segregate into slowly-evolving ( “cold” ) , old gene communities and rapidly-evolving ( “hot” ) , young gene communities . The coldest , oldest communities are centrally located and are highly enriched for gene groups related to fundamental cellular functions , whereas the hottest , youngest communities are peripheral and enriched for gene groups related to higher order functions . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Methods"
] | [
"acute",
"myeloid",
"leukemia",
"leukemias",
"medicine",
"and",
"health",
"sciences",
"genetic",
"networks",
"myeloid",
"leukemia",
"gene",
"regulation",
"computational",
"biology",
"cancers",
"and",
"neoplasms",
"oncology",
"hematologic",
"cancers",
"and",
"related",
... | 2016 | Evolutionary and Topological Properties of Genes and Community Structures in Human Gene Regulatory Networks |
We report here a study of regeneration in Drosophila larval wing imaginal discs after damage by ionizing radiation . We detected faithful regeneration that restored a wing disc and abnormal regeneration that produced an extra wing disc . We describe a sequence of changes in cell number , location and fate that occur to produce an ectopic disc . We identified a group of cells that not only participate in ectopic disc formation but also recruit others to do so . STAT92E ( Drosophila STAT3/5 ) and Nurf-38 , which encodes a member of the Nucleosome Remodeling Factor complex , oppose each other in these cells to modulate the frequency of ectopic disc growth . The picture that emerges is one in which activities like STAT increase after radiation damage and fulfill essential roles in rebuilding the tissue . But such activities must be kept in check so that one and only one wing disc is regenerated .
Ionizing radiation ( IR ) is one of three main modalities in the treatment of cancer , the others being surgery and chemotherapy . Therapeutic effect of IR relies on its ability to kill cells . But what remains could regenerate a tumor , leading to treatment failure . In parallel , incorrect healing of normal tissues after collateral damage by radiation therapy contributes to side effects such as fibrosis and ulcers [for example , [1 , 2]] . Understanding how tumors and normal tissues regenerate after damage by IR could help us make radiation therapy of tumors more effective while reducing unwanted side effects . This study aims to use a powerful genetic model , Drosophila , to identify and characterize genes needed for faithful regeneration after radiation damage . Imaginal discs of Drosophila melanogaster larvae are precursors of adult organs . Because of their high regenerative ability , imaginal discs have been used in studies of tissue regeneration . In classical and recent studies , imaginal discs were surgically fragmented and cultured in adult or larval hosts ( e . g . [3]; reviewed in [4 , 5] ) . Faithful regeneration restored the original structure but abnormal regeneration that produced duplications , transdeterminations and multiplications were also recorded . In leg discs , for example , medial ( anterior ) fragments typically regenerated whole leg discs whereas lateral ( posterior ) fragments typically duplicated the surviving structure [3] . Transdetermination in which the regenerated parts assume the identity of another disc was also frequent in the classical studies of surgically ablated leg discs ( taking on wing fate ) . Multiplication produced too many copies of a normal part in the regenerated structure , for example sensilla trichodea ( sense organs ) in the leg [3] . Another model of regeneration uses genetic means to kill cells in a specific compartment , for example by temporally regulated expression in the wing pouch of pro-apoptotic genes such as eiger , hid or rpr [6 , 7] . The discs with ablated parts reside in situ and regeneration is monitored after shutting off the death-inducing gene . Genetic and molecular analysis identified Wg ( Drosophila Wnt-1 ) and STAT92E as key players in regeneration in both surgically and genetically ablated discs ( e . g . [5 , 7–11] ) . STAT92E ( to be called STAT here ) is the sole Drosophila STAT and is an orthologue of mammalian STAT3/5 . Wg and STAT , for example , were found to be activated by JNK activity at the wound site and promoted cell proliferation and blastema formation that underlies regeneration after surgical ablation [8] . We have been studying regeneration of imaginal discs after damage by ionizing radiation ( IR ) . Unlike surgical or genetic ablation , IR induces apoptosis that is scattered throughout the disc . Larvae in which about half of the cells in each imaginal discs have been killed by IR ( typically 4000R of X-rays ) regenerate to produce a viable fertile adult fly [12 , 13] . While IR-induced apoptosis is scattered , the distribution is not random . We found , for example , that in wing imaginal discs , cells of the future hinge , particularly those in the dorsal half , are more resistant to IR-induced apoptosis than cells of the wing pouch [14] . The hinge cells then participate in the regeneration of the wing pouch that suffers more IR-induced apoptosis than the hinge . The hinge also displays other characteristics such as different apical-basal organization of cells and the presence of ‘tumor hot-spots’ that are particularly prone to neoplastic transformation [15] . We identified two functions of Wg and STAT in the hinge . First , Wg and STAT protect hinge cells from IR-induced apoptosis . Second , Wg and STAT are needed to promote the translocation of these cells towards the pouch , thereby facilitating regeneration of the wing disc after radiation damage . Here , we report that regeneration after IR damage can occur faithfully to restore the wing disc or produce abnormal structures , just like abnormal regeneration after surgical ablation produces duplications , transdeterminations and multiplications . We studied regeneration of irradiated larval wing discs in a time course using a lineage tracing system , and identified a sequence of cellular events that produced a second , ectopic wing disc . We used tissue-specific expression of RNAi to identify cell-autonomous roles for epigenetic regulators , STAT , and Wg in this process . We found that faithful regeneration ( to restore the primary wing disc ) and abnormal regeneration ( to produce an ectopic wing disc ) have common as well as distinct genetic requirements . All genetic components we identified are conserved in mammals , making it possible to test their roles in normal and abnormal regeneration after radiation damage in those systems in future studies .
In our studies of wing disc regeneration after radiation damage , we used the 30A-GAL4 driver , in conjunction with GAL80ts , to direct gene expression to the hinge , which is the site of IR-resistant cells that participate in regeneration [14] . 4000R of X-rays were used in all experiments in this work . We expressed a published G-trace system in which UAS-RFP serves as a real-time marker ( Fig 1B and 1H ) , while clonally stable GFP expression is induced by UAS-FRT mediated recombination [16] . When used in conjunction with a tissue/cell type-specific GAL4 driver such as 30A-GAL4 , cells in which GAL4 is active express RFP . 30A-GAL4 is active in parts of the hinge ( indicated with white lines in Fig 1B , see also Fig 1H ) and a few cells of the notum ( arrow in Fig 1B , see also Fig 1 H ) . These cells also undergo FRT-mediated recombination to stably express GFP from a Ubiquitin promoter . If these cells or their descendants change fate and lose GAL4 activity , RFP would cease to express but GFP would still be expressed , allowing us to follow cells through fate changes . GAL4 activity is further controlled temporally by repressing it with a temperature sensitive GAL80 . Only upon a shift to 29°C to inactivate GAL80ts ( Fig 1I ) , would GAL4 become active to lineage mark cells of interest . The larvae were then irradiated and the fate of labelled cells followed in a time course . In these studies , regeneration restores normal appearance by DNA stain ( Fig 1 , compare 1A to 1C ) , but regenerated wing pouches were composed in part of cells derived from the hinge but have lost their hinge fate ( RFP-GFP+; arrow in Fig 1D ) [14] . Without irradiation , cell fates are relatively more stable ( Fig 1B ) . In this experimental system , we noted that some of the wing discs in irradiated samples had ectopic growths ( Fig 1E , arrowhead , quantified in G and classified in H ) . Wing discs from similarly treated wild type w1118 larvae also produced ectopic growths ( Fig 1F , arrowhead ) . IR-induced ectopic growths were dependent on the temperature shift protocol ( Fig 1I ) . This protocol allowed larval growth and disc development before shifting to 29°C to inactivate GAL80ts and activate GAL4 . Ectopic growths did not form if the larvae were kept at 25°C throughout the experiment; 45/45 discs were class 0/I in two separate experiments . While we do not know why the production of ectopic discs depended on a temperature shift , the ability to produce them provides us with an opportunity to study this mode of regeneration . It should be possible to test the role of temperature-induced molecular changes such as the heat shock response in the future . We used 30A-GAL4>RFP expression pattern to classify discs with ectopic growths . RFP was predominantly in the hinge but was also in a few scattered cells in the notum ( Fig 1H ) . All classes of discs showed similar RFP signal in the hinge but differed in the number and organization of RFP+ cells in the notum . Class 0 discs showed <10 RFP+ cells ( arrow ) , almost all in the anterior ( A ) half of the notum . Class I discs showed more RFP+ cells that have spread into the posterior ( P ) half of the notum ( arrows ) . Class II discs showed a cluster of RFP+ cells near the dorsal-posterior edge of the notum ( arrowhead ) in addition to scattered RFP+ cells in the notum ( arrow ) . Class III discs showed RFP+ cells in a ring that is often incomplete ( arrowhead ) , also near the dorsal posterior edge of the notum , that still shares a continuous boundary with the rest of the disc ( the disc boundary traced with a yellow line appears continuous ) . Class IV discs showed similar rings of RFP+ cells ( arrowhead ) that have grown into a structure no longer confined within the notum ( the disc boundary traced with a yellow line is dis-continuous ) . Quantitative analysis showed that all un-irradiated discs were class 0 or I; classes II-IV are strictly IR-dependent ( Fig 1G , compare 72 h -/+IR , the last two bars ) . To understand how different classes of ectopic growth relate to each other , we followed 30A-GAL4>G-trace discs in a time course after irradiation . Immediately before irradiation ( 0 h ) , the discs were predominantly class 0 ( Fig 1G , IR+ 0h ) . All discs were class I at 24 h after IR . Classes II and III appeared at 48 h after IR , followed by classes IV at 72 h after IR . All irradiated larvae entered the pupal stage subsequently . The complete absence of class 0 discs at 24 h after irradiation is striking because by 72 h after irradiation , all five classes were represented ( Fig 1G ) . At the intermediate time point of 48 h , we saw classes II and III but not yet IV . We interpret these data to mean that between 24 and 72 h after irradiation , some class I discs regressed to class 0 while others progressed to classes II and III and , subsequently , class IV . The frequency of class III and IV discs at 72 h after irradiation in 30A-GAL4>G-trace discs was 20 . 9% ( Fig 1G ) . w1118 discs lacked RFP and could not be classified as in Fig 1H , but the frequency of discs with obvious overgrowths as in Fig 1E and 1F was similar; 38/183 or 20 . 8% in four independent experiments using the protocol in Fig 1I showed obvious ectopic growths , while none did in un-irradiated samples ( 0/153 ) . Similar frequency of ectopic discs in 30A-GAL4>G-trace and w1118 discs means that ectopic growths are not due to 30A-GAL4>G-trace transgenes . IR-induced ectopic growths lacked Ubx and were therefore not haltere discs ( Fig 2A–2C , arrowhead ) [17] . The expression of Wg protein in a theta ( θ ) pattern in the ectopic growths ( Fig 2D–2F , arrowhead ) matched that of wing and haltere discs but not leg discs where it is a wedge . Wg pattern and the absence of Ubx indicated that ectopic growths were ectopic wing discs . All ectopic wing discs were found at the dorsal-posterior edge of the notum ( n>300 discs examined ) , at the same location as RFP+ cell clusters in class II discs ( Fig 1H , arrowhead in ‘class II’ ) . We hypothesized that such cell clusters matured into ectopic discs to produce class III and IV discs . To test the idea , we followed the fate of RFP+ cells using the GFP lineage tracer [16] ( Fig 2H and 2I ) . We found that ectopic pouches included descendants of cells that used to express RFP ( RFP-GFP+ , arrowhead in Fig 2H ) . This is similar to what we saw in the regenerated wing pouch after irradiation , that it was composed of GFP+RFP- former hinge cells that lost their hinge fate [14] ( arrow in Fig 1D ) . The key difference was that fate change and pouch-building occurred days after irradiation and in an ectopic wing disc that did not suffer IR-induced damage . Furthermore , ectopic wing discs also included RFP-GFP- cells ( * in Fig 2I ) . These never expressed RFP ( because they lacked the GFP lineage marker ) and therefore must have originated either from RFP- cells of the primary notum or the primary pouch . We had used the G-trace system previously to test if pouch cells relocated after irradiation; they did not [14] . Therefore , the notum was the likely origin of RFP-GFP- cells in the ectopic disc . These findings are consistent with the idea that ectopic wing discs originated from a combination of RFP+ cell cluster near the dorsal posterior margin of the notum in class II discs and un-related ( RFP-GFP- ) notum cells . The abundance of class 0 discs at the time of irradiation and the complete incidence of class I discs at 24 h after irradiation ( Fig 1G ) reflects the fact that RFP+ cells in the notum increased in number during this period . Quantification of RFP+ cell number confirmed this ( Fig 3A , compare -/+IR 24 h in GAL4 only samples , p = 1 . 05E-05 in 2-tailed t-test ) . Sample images used in the quantification are shown in S1 Fig . To ask if this increase was due to mitosis , we co-expressed Rux in RFP+ cells . Rux is an inhibitor of cyclin/cdk complexes , with particular activity towards cyclin A/cdk1 , which plays a mitotic role in Drosophila [18 , 19] . Rux is expected to block mitosis without affecting S phase . We confirmed this by staining for phospho-S10-Histone H3 , a mitotic marker; we saw reduced mitoses in the 30A domain ( Fig 3B and 3C , brackets ) and cells with large nuclei that resulted from repeated S phases without mitosis ( Fig 3 , arrows compare red nuclei in D and E ) . Co-expression of Rux abolished the IR-induced increase in RFP+ cells in the notum ( Fig 3A , compare UAS-Rux -IR 24h to +IR 24 h , p = 0 . 16 by 2-tailed t-test ) , suggesting that cell number increase within 24 h after irradiation was due to increased mitoses . Increased proliferation after IR-induced cell death would be similar to the phenomenon of ‘accelerated proliferation’ in which surviving cancer cells proliferate faster than before irradiation ( pg . 384 in [20] ) . Mitotic index also increased in larval imaginal discs recovering from irradiation , with the increase occurring throughout the wing disc [12 , 13] . Cell death by other means can also result in increased proliferation in Drosophila imaginal discs , as in the phenomenon of compensatory and apoptosis-induced proliferation ( e . g . [5 , 21–23] for review ) . Even with Rux-induced inhibition of mitosis , classes II-IV discs were produced , although with a delay and with reduced frequency ( Fig 4A–4C , compare 4D to Fig 1G ) . Lineage tracing experiments showed that ectopic discs formed under Rux expression were also composed of RFP+GFP+ , RFP-GFP+ and RFP-GFP- cells , just like the ectopic discs of GAL4 only control larvae ( Fig 4E–4H ) . Furthermore , ectopic disc formation accompanied a large increase in RFP+ cell number in the notum of 30A>Rux discs between 24 and 72 h after irradiation; we counted 281±61 nuclear RFP+ cells in the ectopic growths of class II-IV 30A>Rux discs at IR+72 h ( n = 6 in three biological replicates ) . In contrast , RFP+ cells in the notum in 30A>Rux discs numbered only 57±16 at IR+24 h ( Fig 3A ) . This increase occurred without apparent changes in the rest of the disc . For example , Fig 4A–4C were threshold-adjusted and shown in Fig 4I–4K . Where did these extra ~200 RFP+ cells come from in the span of 48 h while mitosis was prevented by Rux ? We do not favor the possibility that they are migrants from the primary hinge , for the following reasons . At 48–72 h after irradiation , un-irradiated hinges showed 443±111 cells ( n = 10 in two different biological replicate experiments; Fig 4M shows an example of an optical slice used for nuclear counting ) . To produce ~200 cells in the notum , about half must have translocated . But in all our time courses ( Fig 1G ) , we saw no evidence of hinge cells translocating into the notum/ectopic pouch even though the translocation in the other direction was readily detected ( Fig 1D , arrow ) . We also attempted to count RFP+ cells in the primary hinge , to detect any depletion . But unlike in the ectopic hinge where RFP+ nuclei were discrete ( e . g . Fig 4O ) , RFP+ nuclei in the primary hinge of irradiated discs were often fragmented , perhaps due to IR-induced cell death ( arrows in Fig 4N ) . This prevented us from counting nuclei in the primary hinge with confidence . Instead , we quantified the RFP+ area in the hinge from threshold-adjusted images such as those in Fig 4I–4K . While a significant increase in the notum was detected ( Fig 4L ) , there was no evidence of depletion of RFP+ cells in the hinge of class III/IV discs ( compare–IR to +IR class III/IV ) . We also saw no signs of hypertrophy in irradiated hinges of class III/IV discs ( compare nuclei size in Fig 4M and 4N ) , ruling out the possibility that while half the hinge cells translocated to the notum , the remainder underwent extra S phases to fill in the gap . Taken together , these results do not favor the possibility that RFP+ cells from the primary hinge translocated into the notum to produce ectopic discs . This leaves de novo induction of the RFP+ fate in notum cells as a likely possibility , as discussed in DISCUSSION . Our previous work found that Wg and STAT protected the hinge from IR-induced apoptosis and promoted the translocation of hinge cells into the pouch [14] . Therefore , we investigated whether Wg and STAT also has a role in the generation of ectopic wing discs after irradiation . Depletion of STAT with RNAi in the 30A domain reduced the frequency of classes II-IV ( Fig 5A , compare the first two bars ) . We conclude that STAT activity is required in the RFP+ cells to produce ectopic wings . We also inhibited Wg by expressing the antagonist Axin using the protocol in Fig 1I . We found , however , that the resulting discs were malformed even without irradiation . We reasoned that Axin , being an inhibitor , would not require the lag time RNAi needs to take effect in other experiments . As such , Wg , we believe , was being inhibited too early in larval development . In contrast , STAT RNAi produced wing discs that were indistinguishable from GAL4 only controls in un-irradiated samples ( S2 Fig ) . Therefore , we reared larvae for an additional 24 h before shifting to 29° in UAS-Axin experiments . This protocol produced discs that were indistinguishable from GAL4 only controls or STAT RNAi discs ( S2 Fig ) . Under these conditions , UAS-Axin reduced the frequency of ectopic discs compared to similarly treated GAL4 only controls ( Fig 5B ) . The generation of ectopic wing discs thus described involves changes in cell fate , gene expression status ( RFP+ to RFP- ) or both . Consistent , we found that depletion of epigenetic regulators of transcription altered the frequency of ectopic wings ( Fig 5A ) . We depleted Enhancer of Polycomb , E ( Pc ) ; eggless ( egg ) and Set2 , encoding histone methyl transferases; or Nurf-38 , encoding a member of the Nucleosome Remodeling Factor ( NURF ) complex . These regulators were chosen because prior studies showed them to have a role in the apoptosis of larval salivary gland [24] , and we wanted to know if these also affect IR-induced apoptosis in the wing discs . Depletion of each using published RNAi constructs produced inconclusive effects on apoptosis . Instead we saw a significant increase in the frequency of ectopic discs at 72 h after irradiation upon depletion of egg , Set2 or Nurf-38 , but not E ( Pc ) , compared to GAL4 only controls ( Fig 5A ) . No ectopic discs were produced without IR ( Fig 5C ) . Of these , Nurf-38 had the strongest effect and was confirmed using a second RNAi line ( ‘Nurf#2’ ) . Therefore , subsequent analyses focused on Nurf-38 . Two genes that showed the strongest effect on the frequency of ectopic growths are STAT and Nurf-38 . Further , they act in an opposing manner; depletion of STAT inhibited ectopic growths while depletion of Nurf-38 promoted them . To ask if these are functionally related , we depleted both simultaneously . Reduction of STAT using heterozygotes for a STAT severe loss-of-function allele [25 , 26] rescued the incidence of ectopic growth caused by 30A>Nurf-38 RNAi ( Fig 5A , the second to last bar ) . This interaction was confirmed with simultaneous RNAi for STAT and Nurf-38 ( Fig 5A , the last bar ) . These results are consistent with STAT functioning downstream or parallel to Nurf-38 to produce ectopic growths , with each providing activities that oppose the other . Our published studies found that STAT was needed for hinge cells to translocate to the pouch during normal regeneration [14] . This is seen by quantifying the extent of translocation of RFP-GFP+ hinge cells into the pouch as in Fig 1D ( quantification method described in S3 Fig ) . Under conditions where depletion of Nurf-38 led to increased ectopic discs , there was no significant effect on the movement of hinge cells into the pouch ( Fig 5E , compare GAL4 only to Nurf ) , suggesting that depletion of Nurf-38 was unable to promote regeneration beyond what was normally seen . Thus , normal and abnormal regeneration show different dependence on Nurf-38 . We next asked if Nurf-38 depletion could promote regeneration under conditions where regeneration was defective . To this end , we analyzed discs in which Nurf-38 and STAT were depleted simultaneously in the 30A domain . Such discs show reduced ectopic growth suggesting that STAT was successfully depleted ( Fig 5A , the last bar ) . In the same discs , translocation of hinge cells into the pouch was significantly greater than in discs with STAT RNAi alone ( Fig 5E , compare STAT to Nurf STAT ) . We conclude that depletion of Nurf-38 could indeed rescue regeneration defects that result from depletion of STAT . Genetic interaction data described above suggest that STAT functions parallel or downstream of Nurf-38 in ectopic disc formation . To distinguish between these possibilities , we assayed the effect of Nurf-38 depletion on STAT . STAT activity could be monitored using a GFP transcriptional reporter with 10X STAT binding sites from a known STAT target gene , Socs36E ( ‘STAT-GFP’ , [27] ) . This reporter is highly expressed in the hinge as reported previously ( Fig 5D , GAL4 only -IR ) . Irradiation increased STAT-GFP reporter expression throughout the disc , but more obviously in the notum ( brackets ) and the pouch ( arrows ) that normally display close to background levels ( Fig 5D , quantified in F; see S3 Fig for quantification method ) . In discs expressing 30A-GAL4>Nurf-38 RNAi , STAT-GFP also increased after irradiation and this change was similar to what was seen in GAL4 only controls ( Fig 5D , quantified in F ) . We conclude that while depletion of Nurf-38 could modulate a STAT-dependent process in the growth of ectopic discs , this effect is not through changes in STAT activity as detected by the Socs36E-GFP reporter . In other words , these data favor the possibility that STAT and Nurf-38 function in parallel in ectopic disc formation . To ask whether 30A-GAL4 expression in the notum and later in the ectopic discs indicates hinge fate , we stained for the hinge marker Zfh2 . Zfh2 is a transcription factor and a downstream effector of STAT during wing development [28] and during regeneration of genetically ablated wing pouch [9] . Zfh2 expression is confined to the hinge in the 3rd instar wing disc where gray Zfh2 signal encompasses and extends beyond the red 30A>RFP domain ( compare Fig 6E and 6M ) . Without irradiation , we saw little or no signal in the notum despite the presence of RFP+ cells in the region ( Fig 6 , -IR panels ) . In irradiated discs , Zfh2 was detectable in the notum . Specifically , RFP+ cells in the posterior half of the notum in class I discs show faint Zfh2 signal , which also encompassed the RFP signal and extended to neighboring cells ( arrowhead in Fig 6B and corresponding panels below ) . In class II/III discs where the posterior RFP+ cell cluster was clearly visible , Zfh2 also encompassed and extended beyond the RFP signal ( arrowhead in Fig 6C and corresponding panels below ) . The RFP+ ectopic hinge in class IV discs likewise showed Zfh2 staining ( arrowhead in Fig 6D and corresponding panels below ) . Based on the presence of Zfh2 , we conclude that 30A>RFP-expressing cells in the ectopic discs did indeed take on the hinge fate . These data lead us to propose a model for ectopic disc growth ( Fig 6Q ) . In this model , irradiation increases RFP+ cells in the notum through increased mitosis such that all discs are class I by 24 h after irradiation ( Fig 3A ) . Class I discs could remain as class I , revert to class 0 , or form a posterior cell cluster to progress to class II . Cells of the cluster then produce an ectopic hinge/pouch to progress to classes III/IV . We propose that the formation of the cluster and subsequent maturation of it into an ectopic disc requires STAT; depletion of STAT not only reduced the fraction of class II-IV discs compared to GAL4 only controls but also abolished classes III and IV . Therefore , STAT likely functions in all step starting with the formation of class II . Inhibition of Wg also reduced the overall fraction of Class II-IV ( like STAT RNAi ) , but still was compatible with the formation of classes III/IV ( unlike STAT RNAi ) . This suggests that Wg activity is needed in the 30A domain for the formation of the cluster , but less so for subsequent steps . We cannot rule out , however , that Wg activity is needed in cells outside of the 30A domain for the cluster of RFP+ cells to mature into an ectopic wing disc . We further propose that Set2 , Nurf-38 and Egg oppose the formation of class II discs as well as subsequent steps; thus , depletion of Set2 , Nurf-38 or Egg increased classes II-IV .
Regeneration inevitably requires cell fate changes . During blastema formation in vertebrate limb development , for example , cells of the stump de-differentiate , migrate to the wound site , proliferate , and re-differentiate to new fates [29] . Cell fate changes would require destabilizing the gene expression landscape of differentiated cells and , later , re-establishing and stabilizing a new gene expression landscape . In different experimental systems including Drosophila and vertebrates , post-translational modifications in Histones correlate with cell fate changes during tissue regeneration ( e . g . [30] ) . Where studied , there are also causal relationships , for example the role of Polycomb/Trithorax group of genes in transdetermination and regeneration ( e . g . [31–36] ) . In this context , our results contribute to our understanding of regeneration after damage by ionizing radiation as follows . We report a remarkable feat of cellular gymnastics that occur during a 48 h period , from 24 to 72 h after irradiation , to produce an ectopic wing disc . This can happen without the full benefit of cell multiplication ( in Rux-expressing discs ) . While we cannot rule out translocation of hinge cells into the notum in the formation of ectopic discs , our findings do not favor this possibility as described above ( Fig 4 ) . Instead , a possible mechanism is the de novo induction of the hinge marker 30A-GAL4 in the notum cells to produce the posterior RFP+ cell clump . This , we believe is the critical step that is governed by STAT , Wg , Set2 , Egg and Nurf-38 ( model in Fig 6Q ) . Once formed , this cluster of RFP+ cells can lose their hinge fate ( become RFP-GFP+ ) and recruit un-related cells ( RFP-GFP- ) to form an ectopic wing disc ( Fig 2I ) . Moreover , the genetic requirements we identified are cell-autonomous; we expressed dsRNA and Axin using the 30A-GAL4 driver that also drove RFP expression . Thus , while STAT activity increases throughout the notum after irradiation ( Fig 5D ) , cell autonomous depletion within RFP+ cells was sufficient to interfere with duplication , thereby identifying a cell population critical for the formation of ectopic wing discs . Are ectopic discs duplications or transdeterminations ? Because we see two wing discs instead of one , we may classify ectopic discs as duplications . On the other hand , if duplication means copying of existing structures , ectopic discs are not exactly duplications; ectopic pouch , for example , is not derived from the primary pouch . Likewise , notum to hinge/pouch fate change we see may be classified as transdetermination . On the other hand , if transdetermination is defined as when one disc takes on the fate of another ( e . g . leg->wing ) , ectopic discs that remain as wing disc do not fit this definition . Therefore , we will refer to what we see simply as ‘ectopic discs’ , to avoid confusion . The frequency of ectopic wing discs after irradiation ( ~20% in both GAL4 only and w1118 ) seems high but is in fact lower than the frequency of abnormal regenerations recorded in classical studies of surgically ablated discs . Schubiger saw , for example , that among lateral ( posterior ) leg fragments cultured in the adult abdomen , 33/41 ( 80% ) regenerated discs with duplications [3] . In the same study , among medial ( anterior ) leg fragments cultured in the adult abdomen , 21/42 ( 50% ) regenerated discs with transdetermination . Classical regeneration studies using X-rays to cause damage also describe disc duplications but these observations were limited to irradiation of ‘early imaginal discs consisting of only about 20 cells’ [13] . We describe a step-wise sequence of events that can lead to mistakes during regeneration , i . e . ectopic discs . A key step in this sequence , which we propose is targeted by epigenetic regulators , STAT and Wg , is the formation of the posterior cell cluster . Wg , STAT and their homologs have conserved well-studied roles during normal development as well as in tissue regeneration . Wg and STAT are also implicated in pattern duplication and transdetermination of larval imaginal discs . In fact , ubiquitous expression of Wg in second and third instar larva is sufficient to induce transdetermination in imaginal discs [37 , 38] . Molecular analysis of regeneration after surgical ablation of leg discs also identified prominent roles for both STAT and Wg [8] . Here , JNK activation at the wound site activates JAK/STAT and Wg , both of which then contribute to extra proliferation to form a blastema . We also identified STAT and Wg is essential for the hinge cells to participate in regeneration of the pouch after IR damage [14] . The current study further solidified the role of STAT and Wg in regeneration , by identifying their contribution to abnormal regenerations induced by IR . All ectopic wing discs we saw grew out of the dorsal posterior notum . Their location and appearance resemble ‘supernumerary wings discs’ that result from ectopic expression of Wg [39 , 40] or STAT [41] , or mutations in Drosophila EGFR ( DER ) [42] . In the first case , expression of UAS-Wg , with Dpp-GAL-4 driver or in Ubx-driven clones , produced a wing disc that grows out of the notum . In the second study , expression of Dpp-GAL4>UAS-Upd , a ligand for JAK/STAT , or en-GAL4>UAS-hop ( Dm JAK ) , also led to the growth of ectopic wings out of the notum . The key difference was that in the published studies re-programming of cells in the notum to wing pouch was limited to a very early stage in wing development . For example , Wg overexpression at 36±6 h after egg laying induced ectopic wing discs but could not at 48±6 h . The authors concluded that Wg has an early role in specifying wing pouch at the expense of the notum but that at later times , Wg has another role , in D/V patterning . Likewise , induction of ectopic wings was achieved by expression of Upd in second instar larvae . Likewise , ectopic wing discs result from the loss of DER only in young larvae ( up to 120 h after egg laying at 17°C or 72 h after egg laying under ‘normal culture conditions’ ) [42] . In our studies , larvae at the time of irradiation were 72–96 h old . Taken together , these results suggest that in irradiated discs , Wg and STAT revert to their potential seen earlier in development . NURF is known to oppose JAK/STAT signaling in the innate immune response in Drosophila , specifically in the larval hemocytes and the fat body [43] . In those tissues , NURF is recruited to a subset of STAT target genes by physical association with Zinc-finger protein Ken , and provides a repressive function . JAK/STAT signaling is activated in response to infection in order to mount an effective immune response . NURF is proposed to temper JAK/STAT such that an immune response occurs only when needed and is shut off when no longer necessary . This , we propose , parallels the activation of STAT after irradiation that serves an essential regenerative function but must be tempered in order to prevent excessive regeneration , i . e . the production of ectopic structures . Nurf-38 , we found , also opposes STAT in this context , in genetics and RNAi analyses . Depletion of Nurf-38 , however , did not affect bulk STAT activity as seen with the STAT-GFP reporter derived from one STAT target gene . It remains possible that Nurf-38 instead affects STAT activity in the context of another STAT target gene ( s ) , which is important for regeneration . While our data implicate epigenetic regulators in abnormal regeneration , we do not yet know whether epigenetic regulation is at play . Identification of STAT and Nurf-38 targets and analysis of their epigenetic status would be required to address this possibility . We note that we do not believe regeneration in the wing disc occurs by polyploidization . Rather , polyploidization in our experiments was simply an outcome of using Rux to block mitosis . We were surprised by the result that mitosis was dispensable for the formation of an ectopic wing disc . But there is precedent for regeneration with polyploid cells [44] . For example , wound closure in Drosophila adult abdomen occurs partly through polyploidization and cell fusion [45] . Ectopic discs in UAS-Rux experiments in our experiments served a useful purpose in that these provide strong evidence that ectopic pouches were made of both cells from the 30A-GAL4 domain ( large , RFP+ ) and recruits from outside ( small , GFP-RFP- , arrow in Fig 4G ) . In this regard , the RFP+ posterior cluster of cells may resemble the Spemann-Mangold Organizer , identified by early Experimental Embryologists as a region of a frog embryo that , when surgically implanted at an ectopic location , could induce a second body axis . The resulting supernumerary organs included not only the cellular descendants of the Organizer but also host cells near the implant site that assumed new fates and were organized into new structures . Loss or gain-of-function studies that test the effect of losing the RFP+ cell cluster or its appearance at ectopic locations would be needed to further address this parallel .
These stocks are described in Flybase: w1118 , 30A- GAL4 ( on Ch II , Bloomington stock# or BL37534 ) , Ptub-GAL80ts on Ch III , 10XSTAT-GFP ( on Ch II , [27] ) , UAS-Axin-GFP ( on Ch III , BL7225 ) , UAS-STAT RNAi ( on X , BL26899 ) , Set2 RNAi ( BL55221 ) , egg RNAi ( BL31352 ) , Nurf-38 RNAi ( #1 = BL35444; #2 = BL31341 ) , E ( Pc ) RNAi ( BL28686 ) ; UAS-Rux ( BL9166 ) ; STAT92E06346 [25 , 26] . All Nurf-38 RNAi experiments used BL35444 with the exception of Fig 5A ‘Nurf-38 #2’ . The stock used for lineage tracing is also described in Flybase; w*; P{UAS-RedStinger}4 , P{UAS-FLP . D}JD1 , P{Ubi-p63E ( FRT . STOP ) Stinger}9F6 /CyO ( BL28280 ) . Genotypes for BL stocks are in S2 Table . 30A-GAL4>UAS-RFP , G-trace/CyO-GFP; GAL80ts/GAL80ts virgin females were crossed to w1118 males ( GAL4 only controls ) or UAS-dsRNA males . STAT RNAi virgin females were crossed to 30A-GAL4>UAS-RFP , G-trace/CyO-GFP; GAL80ts/GAL80ts males . Progeny bearing G-trace ( RFP+GFP+ larvae ) were sorted for use . Larvae were raised on Nutri-Fly Bloomington Formula food ( Genesee Scientific ) at 25°C unless otherwise noted . The cultures were monitored daily for signs of crowding , typically seen as ‘dimples’ in the food surface as larvae try to increase the surface area for access to air . Cultures were split at the first sign of crowding . Larvae in food were irradiated in a Faxitron Cabinet X-ray System Model RX-650 ( Lincolnshire , IL ) at 115 kv and 5 . 33 rad/sec . Antibodies to cleaved Phospho-S10-Histone H3 ( 1:1000 , rabbit monoclonal , Upstate Biotech ) , Wingless ( 1:100 , mouse monoclonal , Drosophila Hybridoma Bank Cat#4D4 ) , Ubx ( 1:750 , Developmental Hybridoma Bank Cat#UbxFB3 . 38-c ) , and Zfh2 ( 1:400 , rat polyclonal , [46] ) . Secondary antibodies ( Jackson ) were used at 1:100 ( rabbit and mouse ) or 1:200 ( rat ) . For antibody staining , wing discs were dissected in PBS , fixed in 4% para-formaldehyde in PBS for 30 min , and washed three times in PBS , permeabilized in PBTx ( 0 . 5% Triton X-100 ) for 10 min and rinsed in PBTx ( 0 . 1% Triton X-100 ) . The discs were blocked in 5% Normal Goal Serum in PBTx ( 0 . 1% Triton X-100 ) for at least 30 min and incubated overnight at 4°C in primary antibody in block . The discs were rinsed thrice in PBTx ( 0 . 1% Triton X-100 ) and incubated in secondary antibody in block for 2 h at room temperature . Stained discs were washed in PBT . The discs were counter-stained with 10 μg/ml Hoechst33258 in PBT or PBTx ( 0 . 1%TritonX-100 ) for 2 min , washed 3 times , and mounted on glass slides in Fluoromount G ( SouthernBiotech ) . Discs that were imaged without antibody staining ( in quantification of disc classes ) were fixed in 10% paraformaldehyde in PBS for 10 min , washed 1xPBS for 10 min and washed in PBTx for 5 min . The discs were stained with Hoechst 33258 and mounted as described above . With the exceptions noted below , the discs were imaged on a Perkin Elmers spinning disc confocal attached to a Nikon inverted microscope , using a SDC Andor iXon Ultra ( DU-897 ) EM CCD camera . The NIS- Elements acquisition software’s large image stitching tool was used for the image capture . 20–21 z-sections 1 um apart were collected per disc and collapsed using ‘maximum projection’ in Image J . The exceptions are Fig 3 , Fig 4E–4H and 4M–4O that show a single z-section each; Fig 1A , 1B , 1E , 1F and 1H and Fig 2D–2F which were acquired on a Leica DMR compound microscope using a Q-Imaging R6 CCD camera and Ocular software . The distribution of disc classes was analyzed using the Chi-square test , using the numbers for GAL4 only control +IR to generate the expected ratios of classes . Cell number in the notum , RFP+ area , and STAT-GFP were analyzed using a 2-tailed t-test . For sample size justifications , we used a simplified resource equation from [47]; E = Total number of animals − Total number of groups , where E value of 10–20 is considered adequate . When we compare two groups ( -/+IR or GAL4 vs RNAi , for example ) , 6 per group or E = 11 would be adequate . All samples subjected to statistical analysis meet or exceed this criterion . | Accuracy in regeneration ensures that the original structures are restored , no more and no less . Prior studies in the wing primordia of Drosophila melanogaster larvae that have been damaged by high energy radiation show that regeneration occurs to restore the original structure . We report here that , in the same experimental system , abnormal regeneration can also occur to produce extra wing structures . We describe a series of cell rearrangements and fate changes that underlie abnormal regeneration , and identify genes responsible for these events . Modulation of such genes have the potential to mitigate abnormal regeneration that occurs after radiation damage to produce such side effects as ulcers and fibrosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"invertebrates",
"cell",
"death",
"rna",
"interference",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"animals",
"animal",
"models",
"mitosis",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"organism",
"development",... | 2017 | STAT, Wingless, and Nurf-38 determine the accuracy of regeneration after radiation damage in Drosophila |
In Bangladesh , increases in cholera epidemics are being documented with a greater incidence and severity . The aim of this prospective study was to identify the prevalence and importance of V . cholerae O1 and enterotoxigenic Escherichia coli ( ETEC ) as causal agents of severe diarrhea in a high diarrhea prone urban area in Dhaka city . Systematic surveillance was carried out on all diarrheal patients admitted from Mirpur between March 2008 to February 2010 at the ICDDR , B hospital . Stool or rectal swabs were collected from every third diarrheal patient for microbiological evaluation . Of diarrheal patients attending the hospital from Mirpur , 41% suffered from severe dehydration with 39% requiring intravenous rehydration therapy . More diarrheal patients were above five years of age ( 64% ) than those below five years of age ( 36% ) . About 60% of the patients above five years of age had severe dehydration compared with only 9% of patients under five years of age . The most prevalent pathogen isolated was Vibrio cholerae O1 ( 23% ) followed by ETEC ( 11% ) . About 8% of cholera infection was seen in infants with the youngest children being one month of age while in the case of ETEC the rate was 11% . Of the isolated ETEC strains , the enterotoxin type were almost equally distributed; ST accounted for 31% of strains; LT/ST for 38% and LT for 31% . V . cholerae O1 is the major bacterial pathogen and a cause of severe cholera disease in 23% of patients from Mirpur . This represents a socioeconomic group that best reflects the major areas of high cholera burden in the country . Vaccines that can target such high risk groups in the country and the region will hopefully be able to reduce the disease morbidity and the transmission of pathogens that impact the life and health of people .
Epidemics of acute watery diarrhea are on the increase in many countries around the world with major outbreaks being seen in Asia and Africa [1] in both rural and urban areas . In Bangladesh , such increases in diarrheal epidemics have been documented [2] , [3] , [4] , [5] with a greater incidence in the capital city , Dhaka . Not only has the number of diarrheal patients seeking care increased but also the severity of disease [2] , [4] , [5] . The epidemics in Bangladesh have occurred during floods , cyclones and other natural disasters [6] , as well as in the biannual seasonal periods [7] . Vibrio cholerae is the most frequently isolated bacterial pathogen from patients presenting with diarrhea to hospitals [2] , [3] , [5] . Comparing diarrheal epidemics in Dhaka from 2007 , 2004 , and 1998 , more severe dehydration due to cholera was seen in 2007 ( 35% ) compared with 2003 ( 25% ) and 1998 ( 22% ) [2] . During both the 1998 and 2004 flood associated epidemics; there was an approximate doubling in the proportion of patients with V . cholerae infection compared with the seasonally matched control period ( 1998; 42% from 20%; 2004 23% from 11% ) . A similar trend was observed in epidemics in rural Bangladesh , with over 70% of patients presenting with severe diarrhea in 2006 , compared to 30–40% in the late 1990s [4] . The aim of this prospective study was to identify the prevalence and importance of V . cholerae O1 and Escherichia coli ( ETEC ) as causal agents of severe diarrhea in Dhaka . The objective was to focus on acute watery diarrhea which required hospitalization to obtain information that could serve as baseline data for carrying out vaccine related as well as water and sanitation based preventive measures in the near future . Our earlier retrospective data from the ICDDR , B have showed that of patients from about 13 areas of Dhaka city that seek help at the diarrheal hospital , the highest hospitalization is from the Mirpur area of Dhaka city [8] . We have in the present study carried out a systematic surveillance of patients attending the ICDDR , B diarrheal hospital in Dhaka city from Mirpur . This area is a densely populated area of a mixed income neighborhood with a population of about 2 . 5 million , representing other populated areas in the region . To better understand the etiology of diarrheal disease we carried out a demographic , clinical and microbiological outcome analyses . We placed emphasis on determining the prevalence of V . cholerae O1 induced acute watery diarrhea over the 24 month study period between 2008 to 2010 . The prevalence of another major cause of watery diarrhea , ETEC which is a common pathogen in these setting was also studied [7] , [9] .
This study was carried out in patients attending the diarrheal hospital at the ICDDR , B , Dhaka , Bangladesh . Previous data analysis of hospitalized patients due to severe diarrhea showed the highest percentage coming from the Mirpur area of the city [8] . Systematic surveillance was carried out on all diarrheal patients admitted from Mirpur between March 2008 to February 2010 . The metropolitan area of Dhaka has a total area of 153 . 84 Sq . Km , is divided into 10 zones . [10] . The Mirpur area is in the north-west of the city consists of two zones that are subdivided into 16 metropolitan wards . The hospital surveillance activities were approved by the Research review committee and ethical review committee of ICDDR , B . Informed oral consent was obtained due to most of the participants were illiterate . According to the ICDDR , B hospital surveillance system , we only require verbal consent from patients undergoing routine investigation for collecting only stool specimens . Consent was documented in the surveillance questionnaire . Demographic , socioeconomic , and clinical data for each patient was captured in the ICDDR , B database and clinical examination and history forms were completed by trained hospital physicians . All patients were assessed for clinical conditions , including degree of dehydration according to WHO guidelines [11] with treatment being given according to the ICDDR , B protocol [12] . The clinical criteria for admission were moderate to severe diarrhea requiring hospitalization . The WHO categorization for exclusive breast feeding practice in infants 6 months and younger was also applied [13] , [14] . Stool or rectal swab specimens were collected from every third diarrheal patient coming from Mirpur who were admitted to the ICDDR , B hospital and were evaluated for V . cholerae O1 and O139 and also tested for ETEC [9] , [15] . In addition specimens were tested for Salmonella spp . , and Shigella spp . using standard microbiological techniques [16] Other diarrheagenic E . coli were not analyzed in the study . For isolation of V . cholerae , specimens were cultured on taurocholate-tellurite-gelatin agar plates . Specific monoclonal antibodies were used to detect V . cholerae O1 , Ogawa and Inaba serotypes , as well as O1 or O139 serogroups [17] , [18] . For microbiological evaluation of V . cholerae , specimens were also enriched in alkaline peptone water for 4 hours and then cultured [3] . The variant phenotype of cholera toxin expressed by V . cholerae O1 was detected by PCR [19] . For this purpose , a random collection of every 100th strain of V . cholerae O1 in the collection ( n = 57 ) were typed as the El Tor or the altered phenotype [20] . For ETEC detection , E . coli was cultured overnight on MacConkey agar plates and six freshly lactose-fermenting colonies were isolated and tested for the presence of heat labile toxin ( LT ) and heat stable toxin ( ST ) [9] , [18] . Detection of LT was carried out using a ganglioside GM1 ELISA test [21] and ST was detected by an inhibition ELISA [22] , [23] . Colonies that tested positive for either toxin were plated onto colonization factor antigen ( CFA ) agar with bile salts to identify the CFs using a dot blot immunoassay technique with specific monoclonal antibodies [9] , [24] . Information concerning the prevalence of rotavirus was obtained for the study period from the available ICDDR , B database but was not prospectively determined for patients coming from the Mirpur area . Nutritional status of the children was expressed in terms of Standard Deviation Score ( SD ) of an anthropometric index such as weight for age , Height for age or weight for height . The anthropometric measurement for children's WAZ ( “Weight-for-age z-score” ) , HAZ ( “height-for-age z-score” ) or WHZ ( “Weight-for-Height z-score” . ) was calculated in relation to the new World Health Organization growth standards [48] . We considered <−2 Z score as underweight , stunted and wasted respectively for the children . Statistical analyses were performed using Statistical Package for Social Sciences ( SPSS , Chicago , IL ) version 12 . 0 . Associations were carried out by calculating the odds ratio ( OR ) with 95% confidence intervals ( CI ) using EpiInfo 3 . 3 . 2 and χ2 ( chi-square ) tests . Areas maps were prepared using Adobe Photoshop 7 . 0 .
Of all patients admitted to the ICDDR , B hospital during the study period , 31 , 588 ( 12% ) came from the Mirpur area . Diarrheal patients from all 16 wards of the area came to the ICDDR , B hospital for treatment . We however , identified 6/16 metropolitan wards within Mirpur which had a high rates of acute diarrhea based on the patient hospitalization information while the rest had moderate to lower rates . The median age was 18 years ( 1 mo-96 yr ) with 54% male and 46% female patients ( Table 1 ) . In terms of the Mirpur patients , 89% lived in low income housing , and only 8% lived in independent homes or in high income residential areas . About 20% of adult males , either the patients or the parent's of admitted children , had received formal schooling , while for females this percentage was lower , falling , between 11–12% . Over 98% of patients used tap water and around 1% used tube-well water for bathing or drinking purposes . About 33% of patients did not use treated drinking water , and 65% reported that they boiled their water . The majority of the study population ( 92% ) used shared sanitary latrines in the community . Of the diarrheal children aged up to 6 months of age , only 14% were exclusively breastfed . About 77% of the study population attending the hospital from Mirpur had a low monthly income ( ≤10 , 000 taka ∼US$ 150 ) . Of the diarrheal patients attending the hospital from Mirpur , 41% suffered from severe dehydration and 39% were given intravenous rehydration therapy ( Table 1 ) . About 4% of the patients had fever ( >37 . 7°C ) ; 58% had abdominal pain; and 79% suffered from vomiting . More than half of the patients ( 52% ) took antibiotics or had oral rehydration ( 92% ) before they were hospitalized . There were more diarrheal patients above 5 years of age ( 64%; median: 27 yr ) than those below five years of age ( 36%; median: 1 yr ) who were admitted to the hospital from Mirpur . About 60% of the patients above 5 years of age had severe dehydration compared with only 9% of patients under 5 years of age . Among the children under 5 years of age , 40% ( N = = 1717 ) were moderate to severely under weight , 52% were stunted ( N = 2225 ) and 22% ( N = 902 ) were wasted . One of the definitions of HAZ is “height-for-age z-score” , WAZ is “Weight-for-age z-score” and WHZ is “Weight-for-Height z-score” . Among cholera and non cholera pediatric patients WAZ<−2 Z score 43% vs . 39% ( p = 0 . 03 ) , HAZ<−2 Z score 53% vs . 50% ( p = ns ) , WHZ<−2 Z score 27% vs . 19% ( p = 0 . 001 ) . The most prevalent pathogen isolated from diarrheal patients coming from Mirpur during the study period was Vibrio cholerae O1 ( 23%; n = 2647 ) . This was followed by ETEC ( 11%; n = 1248 ) . In addition , 230 patients ( 2% ) were positive for both V . cholerae O1 and ETEC . Cholera rates showed two peaks , one between April and May and the second between August and September with lower rates in the winter months beginning in November . In terms of ETEC rates , the disease increased from April and continued until September with lower levels in the winter months ( Figure 1 ) . Less than 1% of the specimens were positive for Salmonella spp and Shigella spp . [20] . The Mirpur area comprises zones 7 and 8 of the city which include 16 metropolitan wards . Diarrheal hospitalization rates were highest from wards 2 , 4 , 5 , 14 , 16 and 41 , moderate ( >2 to 4/1000 ) from wards 6 , 7 , 8 , 10 , 11 , 12 and 13 , and lower ( less than 2/1000 ) for the rest of wards ( 1 , 3 , 9 and 15 ) from where patients attended the hospital ( Figure 2 ) . ETEC diarrheal hospitalization rates were lower and ranged from 2 . 3–3 . 5/1000 population with wards 5 , 11 , 14 , 16 and 41 showing higher rates . In terms of infection by both V . cholerae and ETEC , patients came from wards 5 , 14 , 16 and 41 , although rates of cholera were higher in wards 2 and 4 . About 23% ( n = 2647 , median: 20 yr ) of patients had culture-confirmed cholera . Severe dehydration was seen in 70% of cholera patients . About 77% were over five years of age ( median: 23 yr ) . Among cholera patients , 80% of patients from Mirpur above 5 years of age suffered from severe dehydration , but children less than 5 years of age ( median: 2 yr ) this percentage was lower at 35% . About 8% of cholera infection was seen in infants ( median: 3 mo ) with the youngest children being only 1 month of age ( n = 186 ) . Of the V . cholerae O1 strains isolated from Mirpur patients , 82% were Ogawa and 18% were Inaba serotypes . Screening of a random number of V . cholerae O1 strains ( around every 100th strain isolated during the period; 42 strains in total ) , showed these were all variant strains that produced the classical phenotype V . cholerae O1 toxin . In terms of overall ETEC rates among Mirpur patients , 11% had ETEC diarrhea ( ( N = 1248 , median: 18 yr ) with 38% suffering from severe dehydration . Of these , 59% were more than 5 years of age ( median: 28 yr ) while 41% were under 5 years of age ( median: 1 yr ) . Of the under 5 year old in age , 10% suffered from severe dehydration where as 57% adults with ETEC diarrhea were severly dehydrated . Among children less than 5 years of age ( median: 1 yr ) , 12% had ETEC diarrhea . The cholera hospitalization rate for children below 1 year of age was 8% while in the case of ETEC , it was 11% ( Figure 3 ) . All severely dehydrated patients received azithromycin in the hospital with an average duration of stay of 9 hours . We analyzed data to determine if there were any differences between disease presentation of patients admitted with cholera or ETEC diarrhea . There were more cholera patients than ETEC diarrheal patients admitted to the hospital ( P<0 . 001 ) . The highest numbers of cholera and ETEC diarrheal patients were above 18 years of age . The second highest age group of cholera patients was in the 1–17 year range , whereas ETEC diarrhea patients fell into the 1–5 year range ( Table 2 ) . Fever was more common in ETEC than V . cholerae O1 infection ( 4% versus 1%; P = 0 . 001 ) . Cholera patients had significantly higher rates of watery stools than ETEC diarrheal patients ( 99% vs . 97% , respectively; P = 0 . 001 ) but no difference in abdominal pain was seen between the patient groups . Cholera patients showed higher rates of profuse vomiting at more than ten times prior to hospital admission when compared with ETEC patients translating into a higher risk for vomiting of 20% versus 11% , respectively ( P<0 . 001 ) . Severe dehydration was also more common in cholera than ETEC diarrheal patients ( 70% vs . 38% , P<0 . 001 ) as were the rates for intravenous fluid requirement ( 65% vs . 37% , P<0 . 001 ) . Of the isolated ETEC strains , the enterotoxin type were almost equally distributed; ST accounted for 31% of strains; LT/ST for 38% and LT for 31% . The isolation was carried out more or less equally throughout the study period and no seasonality in ETEC enterotoxin type was detected . In terms of colonization factors ( CFs ) , 47% were positive among the 13 CFs that were tested for in this study . The major CFs detected were CS5+CS6 ( 15% ) , CS7 ( 13% ) , CS14 ( 13% ) , CS17 ( 13% ) , CS6 ( 13% ) , and CFA/I ( 9% ) . The prevalence of CF positive ETECs based on the toxin phenotypes was also determined . When analyzing CFs by toxin type , 35% of LT-producing , 24% of ST-producing and 41% of LT/ST producing ETEC strains were positive for the CFs . The CS7 and CS17 types were expressed mostly by LT expressing ETEC and only about 1% and 3% were LT/ST-ETEC respectively ( Table 3 ) .
Although mortality rates due to diarrhea have decreased globally over the years , diarrheal diseases still range among the highest causes of child and adult morbidity in developing countries in Asia and Africa [25] , [26] . The aim of this study was to carry out surveillance for severe diarrhea especially cholera in Dhaka city , which is facing problems of rapid urbanization and the associated lack of public health intervention , sanitation and safe water availability to meet the needs of the growing population [27] . As a result , rates of infectious diseases including cholera have increased tremendously [28] , [29] . The equal numbers of male and female patients that sought care from Mirpur were predominantly from lower socioeconomic groups , which reflect the demographic picture that has been seen for diarrheal patients being treated at the ICDDR , B hospital over recent years [30] . The patients who came from the area lacked appropriate water and sanitation facilities in their homes . Almost all patients obtained tap water from the government source . However tap water was collected by household members and then stored in homes for use for drinking , bathing and other purposes [7] . Although 60% reported of having used boiled drinking water , all other usage was from untreated tap water suggesting high risk from such sources . In an earlier study in Mirpur , only 21% of people were found to treat their drinking water and most people also used sanitary latrines [7] . During the study period , more adults than children from Mirpur area attended the hospital , although data from the ICCDR , B 2% surveillance system has shown that 60% of total patients treated are children [31] . The discrepancy could be due to the fact that there are numerous health clinics in and around the Mirpur area , which may cater to the treatment of children when compared with treatment facilities that are available for adults [32] , [33] . In terms of causal bacterial pathogens for watery diarrhea in Mirpur patients , V . cholerae O1 was the most prevalent 70% in adults with 72% of patients suffering from severe dehydration . However , about 34% of children less than five years diagnosed with cholera suffered from severe diarrhea and required IV fluid requirement; when considering children below the age of 2 years , 12% showed symptoms of severe dehydration . Earlier epidemiological studies in cholera patients have indicated that the disease is more prevalent in older children and adults , which is supported by these study data . However , findings from the Mirpur patients reveal that children are also susceptible to the disease and suffer from severe dehydration [34] , [35] , [36] . In a 2008 study in Bangladesh , 86% of children 12 years and younger in age that had cholera suffered from severe dehydration [37] while another study has reported cholera in neonates [34] . These reports and the present analyses show the changing trend in cholera epidemiology , which mirrors data published recently in other studies in India and Africa [35] , [36] , [38] . The biannual seasonality observed for cholera in Mirpur is similar to that seen in most other areas in Bangladesh [3] , [5] , [30] . Such a typical seasonality suggests that the rising temperature from spring onwards , as well as other environmental factors such as lowering levels of surface water and the increased spread of the pathogen in the community by the fecal oral route may be the major causes of the biannual epidemics in the urban area of the city . Therefore , high prevalence of cholera is likely due to the increased bacterial load in the water , as well as the high transmission rates among the population . In a previous study involving cholera patients and their household contacts , the rates of cholera infection were high [39] , [40] with two contacts in a household being infected from a hospitalized index case on average [40] . Since the Mirpur area lacks access to pond and river water , the contribution of these systems to increased cholera outbreaks in Mirpur as has been postulated earlier for rural areas in Bangladesh , cannot be extrapolated to the urban area [41] . A reason for the high rates of cholera in children in Mirpur , despite local clinics offering health care to children in the area , may be due to the immunocompromised status of the children . A relatively high proportion of children studied were stunted ( HAZ-<2Z score; 48% ) or moderately ( 54% ) to severely malnourished ( 44% ) . Earlier studies carried out in Mirpur showed that by the age of two years 38% of children were stunted while 58% were underweight [7] . An important limiting micronutrient in children in developing countries is zinc [42] . It is known that over 50% of children are zinc deficient in Bangladesh [43] , as well as in the Mirpur area [44] , in addition to varying rates of other malnutrition indices and micronutrients [45] . During the surveillance study , identification of etiological agents other than V . cholerae O1 was carried out with emphasis on ETEC , the other major bacterial cause of acute diarrhea . Overall , among the two most common bacterial etiologies , V . cholerae O1 infections were the most prevalent and was about the twice the rate seen for ETEC diarrhea in the Mirpur area . The overall trend at the ICDDR , B hospital data was also comparable with that of Mirpur ( V . cholerae O1: 23%; ETEC: 11% ) . The rates of ETEC diarrhea seems to have undergone a decrease compared to what has been described earlier although it is still the second most common cause of bacterial diarrhea [2] , [5] , [30] , [46] . ETEC diarrhea was seen in around equal proportions of adults and children . Earlier analyses have shown ETEC to be more prevalent in children than adults [9] , [46] . The three enterotoxin types of ETEC were isolated in similar proportions in strains obtained from the study patients . Data from the ICDDR , B hospital over the last decade has shown that the prevalence of the different ETEC toxin phenotypes have undergone a change . In studies carried out between 1996–1998 , the ST toxin type was more prevalent . In studies carried out more recently in 2007 , the LT phenotype predominated [2] . In the present analyses , all the three phenotypes were evenly distributed among the patients . The phenotypic changes in ETEC have been monitored using the similar methods and techniques over the years and therefore not due to changes in assay procedures . Rotavirus was not prospectively tested in specimens obtained from Mirpur but a 2% systematic surveillance data from the ICDDR , B showed that it is also very high among children from the Mirpur area ( 42% ) [47] . Studies on ETEC have shown that the ST , ST/LT phenotype of ETEC are predominantly CF positive while the LT only strains are not [7] , [22] . In contrast , the current data shows that the CF expressing strains belonged to the LT phenotype alone in relatively high proportion of ETEC , an observation that supports results from our study in 2007 [2] but different from our analyses earlier in 2000 where LT-ETEC were generally more negative for CFs [9] . In terms of the CFs detected on strains , the major types were CS7 and CS17 , as well as CS5+CS6 and CS6-ETECs . The CS1 , CS2 , CS3 phenotypes were low in prevalence . Therefore , it appears that ETEC diarrheal agents are undergoing changes in both toxin and CF profiles and ETEC vaccine development should take into consideration the changing phenotypes that are being seen . The prospective analyses of hospitalized patients from Mirpur in urban Dhaka , shows that V . cholerae O1 is the major bacterial pathogen and a cause of severe cholera disease . Strategies for the implementation and use of available cholera vaccines are needed in the area to decrease the annual burden of disease . Mirpur represents a socioeconomic group that best reflects the major areas of high cholera burden in the country . Vaccines that can target such high risk groups in the country and in other similar regions will hopefully be able to reduce the disease morbidity and the transmission of pathogens that impact on the life and health of people . | Bangladesh is a country where acute dehydrating diarrhea or cholera is common and is seen at least two times every year and additionally in natural disasters . In addition cholera cases have increased in the country , especially in urban settings such as in the capital city , Dhaka , where the number of hospitalized patients with more severe disease has tremendously increased . In the present observation , we have concentrated on determining the occurrence of diarrhoea caused by the two most common bacterial agents V . cholerae O1 and enterotoxigenic Escherichia coli ( ETEC ) in a densely populated , disease prone area Mirpur in Dhaka for two years from March 2008 to February 2010 . Stool or rectal specimens from diarrheal patients coming to the ICDDR , B hospital from Mirpur were tested for the two bacterial pathogens . We found that V . cholerae O1 was the major bacterial pathogen and a cause of severe cholera disease in 23% of patients ( 2 , 647 of a total of 11 , 395 patients ) from Mirpur . We surmise that cholera vaccines , as well as other public health tools that can target such high risk groups in the country , will be able to reduce the disease morbidity and the transmission of pathogens to improve the quality of life in urban settings . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/bacterial",
"infections",
"microbiology/immunity",
"to",
"infections"
] | 2011 | Impact of Rapid Urbanization on the Rates of Infection by Vibrio cholerae O1 and Enterotoxigenic Escherichia coli in Dhaka, Bangladesh |
Since 2004 , treatment of Mycobacterium ulcerans disease , or Buruli ulcer , has shifted from surgery to daily treatment with streptomycin ( STR ) + rifampin ( RIF ) for 8 weeks . For shortening treatment duration , we tested the potential of daily rifapentine ( RPT ) , a long-acting rifamycin derivative , as a substitute for RIF . BALB/c mice were infected with M . ulcerans in the right hind footpad and treated either daily ( 7/7 ) with STR+RIF or five days/week ( 5/7 ) with STR+RIF or STR+RPT for 4 weeks , beginning 28 days after infection when CFU counts were 4 . 88±0 . 51 . The relative efficacy of the drug treatments was compared by footpad CFU counts during treatment and median time to footpad swelling after treatment cessation as measure of sterilizing activity . All drug treatments were bactericidal . After 1 week of treatment , the decline in CFU counts was significantly greater in treated mice but not different between the three treated groups . After 2 weeks of treatment , the decline in CFU was greater in mice treated with STR+RPT 5/7 than in mice treated with STR+RIF 7/7 and STR+RIF 5/7 . After 3 and 4 weeks of treatment , CFU counts were nil in mice treated with STR+RPT and reduced by more than 3 and 4 logs in mice treated with STR+RIF 5/7 and STR+RIF 7/7 , respectively . In sharp contrast to the bactericidal activity , the sterilizing activity was not different between all drug regimens although it was in proportion to the treatment duration . The better bactericidal activity of daily STR+RIF and especially of STR+RPT did not translate into better prevention of relapse , possibly because relapse-freecure after treatment of Buruli ulcer is more related to the reversal of mycolactone-induced local immunodeficiency by drug treatment rather than to the bactericidal potency of drugs .
Buruli ulcer ( BU ) is a devastating skin infection caused by Mycobacterium ulcerans [1] . The standard antibiotic treatment combines rifampin ( RIF ) and streptomycin ( STR ) given for 8 weeks [2] . Since foci of the disease are primarily found in tropical areas of rural Africa and the majority of patients are children , a treatment with shorter duration would be more acceptable to the patients and facilitate the delivery of drugs in settings with limited public health infrastructure [3] , [4] . In addition , shortening treatment duration will reduce the risks of oto- and nephro-toxicity and opportunities for bloodborne pathogen transmission associated with daily injection with streptomycin [5] . Replacing RIF with the long-lived rifamycin derivative rifapentine ( RPT ) in the standard first-line regimen for tuberculosis results in more rapid culture conversion ( a measure of bactericidal activity ) and a shorter treatment duration needed to prevent relapse ( a measure of sterilizing activity ) . Consequently , the RPT-containing regimen allowed in mice a reduction of treatment duration from 6 months to 3 months [6] , [7] . In the murine model of M . ulcerans disease , daily RPT at the lower dose of 5 mg/kg was shown to be as active as or even more active than daily RIF at 10 mg/kg [8] . In addition , RPT alone at 10 mg/kg was more active than RIF and STR given alone and almost as active as the STR+RIF combination [9] . Thus we hypothesized that the substitution of RPT for RIF would increase rifamycin exposure and antimicrobial activity , and consequently reduce the current duration of treatment for BU . To test this hypothesis the reduction of CFU counts during treatment and reactivation of the disease after cessation of treatment with STR+RPT given 5 days/week ( 5/7 ) were compared to those after the standard STR+RIF regimen in mice , given 5/7 . Since in humans the standard treatment [10] is given 7 days/week ( 7/7 ) and STR+RIF regimen given daily 7/7 in mice was shown [11] more active than STR+RIF given 5/7 , the 7/7 STR+RIF regimen was also included in the present experiment . Drugs were given for 4 weeks and doses were equivalent ( similar AUC ) to the human doses [12] .
STR and RIF were purchased from Sigma ( St . Louis , MO ) and RPT was a gift from Sanofi-Aventis pharmaceuticals ( Paris , France ) . Stock solutions of RIF and RPT were made in sterile 0 . 05% agarose solution and STR was prepared in sterile normal saline . All stock solutions were prepared weekly and stored at 4°C . RIF and RPT were administered , at 10 mg/kg body weight , orally using an esophageal cannula ( gavage ) , and STR was administered by subcutaneous injections at 150 mg/kg body weight . Mice were infected with M . ulcerans 1615 ( Mu 1615 ) ( ATCC 35840 ) , obtained from Dr . Pamela Small , University of Tennessee , that was originally isolated in Malaysia from a patient and is a part of the Trudeau collection [13] , [14] The Mu 1615 strain was grown on Middlebrook 7H11 + OADC ( Becton-Dickinson , Sparks , MD ) for 8–12 weeks . A uniform homogenous suspension was prepared from the colonies by suspending them in sterile phosphate buffered saline ( PBS ) with Tween 80 ( 0 . 05% ) and vortexing them with sterile glass beads . The suspension was allowed to stand for 15–20 minutes till the large clumps settled , the supernatant was adjusted to an optical density at 600 nm ( O . D . 600 ) of 1 , serially tenfold diluted from 0 to 10−6 in sterile PBS , and 0 . 5 ml of each dilution was inoculated on 7H11 agar + OADC plates containing twofold concentrations of RIF or RPT ranging from 0 . 007 to 2 µg/ml along with drug free controls . Plates were incubated at 32°C for 12 weeks . An aliquot of a twice-mouse-passaged Mu 1615 strain stored at −80°C was thawed and inoculated in mouse footpads . Once the footpads were swollen to a lesion index of 2–3 , defined as inflammatory footpad/hind foot swelling [15] , mice were sacrificed and footpad tissue was harvested , minced and suspended in sterile PBS . The solution was vortexed briefly , allowed to stand for 30 minutes , and the supernatant was used for footpad infection . Prior to infection , the inoculum was checked qualitatively for acid-fast bacilli , serially diluted , and plated for CFU counts on selective 7H11 plates [9] . The kinetic method developed by CC Shepard for assessing the activity of anti-leprosy drugs was used to assess drug activity [9] , [16] , [17] . In brief , 355 female BALB/c mice aged 4 to-6 weeks ( Charles River , Wilmington , MA ) were infected in the right hind footpad with 0 . 03 ml of the Mu 1615 suspension . After infection , mice were randomized to an untreated negative control group ( n = 55 ) and two positive control groups treated with ( i ) STR+RIF given 5 days a week or 5/7 ( n = 100 ) or ( ii ) STR+RIF given 7 days a week or 7/7 ( n = 100 ) . Finally , to assess the benefit of substituting RPT for RIF , the test group was the STR+RPT combination ( n = 100 ) , given five days a week or 5/7 . The study was conducted adhering to the Johns Hopkins University guidelines for animal husbandry and was approved by the Johns Hopkins Animal Care and Use Committee , protocol MO08M240 . The Johns Hopkins program is in compliance with the Animal Welfare Act regulations and Public Health Service ( PHS ) Policy and also maintains accreditation of its program by the private Association for the Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) International Five mice from the untreated group were sacrificed the day after infection ( D1 ) and 10 mice were sacrificed 26 days later ( D26 ) at treatment initiation to establish baseline CFU counts in the footpads . The bactericidal activity of each drug regimen was monitored by performing weekly CFU counts from the foot pad of 5 mice per drug regimen and 5 untreated mice . In mice treated 5/7 , the last gavage was on Friday morning and mice were sacrificed on the following Monday afternoon . In mice treated 7/7 , the last gavage was on Sunday morning and mice were sacrificed the next day , on Monday afternoon . However to rule out any carryover by rifapentine ( RPT ) , mice treated with streptomycin-rifapentine ( STR+RPT 5/7 ) group were held for a wash-out period of one week after stopping treatment before being sacrificed . The sterilizing potential of each drug regimen was monitored by determining the time to footpad swelling in negative control mice and in 20 mice treated for 1 , 2 , 3 , and 4 weeks from each of the three treatment groups , according to the protocol given in Table 1 . Footpad swelling was considered to have occurred when the footpad appeared inflamed to the naked eye , i . e . , reached a lesion index of 2 [15] . Additionally to confirm whether the relapse was really due to multiplication of M . ulcerans , at least one mouse from each group which had appreciable swelling was cultured for CFU count . For quantitative footpad CFU counts , mice were sacrificed by cervical dislocation and footpads were harvested after thorough disinfection with soap and sterile PBS followed by 70% alcohol swabs . Footpad tissue was then homogenized by fine mincing and suspended in 2 ml sterile PBS . Appropriate dilutions of 0 . 5 ml were plated on selective 7H11 plates in duplicate and incubated at 32°C for 12 weeks before CFU were enumerated . Survival analysis , with footpad swelling as the measurement , was performed using the Kaplan-Meier method [18] . The log rank test was used to determine the level of statistical significance when comparing survival curves of the different treatment groups with the control group . p values were two-tailed , and a value of p<0 . 05 was considered statistically significant . CFU counts were log-transformed before analysis . Culture-negative footpads were assigned a log value of 0 . Group means for experimental treatment groups were compared with that of the standard treatment control by one-way analysis of variance with Dunnett's post-test . Paired t-tests were also used to compare groups of equal size . All analyses were performed with GraphPad Prism version 4 . 01 ( GraphPad , San Diego , CA ) .
For M . ulcerans strain 1615 , the MIC of RIF was 0 . 25 µg/ml and the MIC of RPT was 0 . 125 µg/ml , one dilution lower than that of RIF . The CFU counts by D-27 , the day after infection , and by D0 , the day on treatment initiation , and after 1 , 2 , 3 , and 4 weeks of treatment are given in table 2 and figure 1 . In the negative control untreated mice , CFU counts were 4 . 88±0 . 51 , one log10 higher on treatment initiation ( D0 ) than on the day after infection ( D-27 ) , and further increased to 5 . 72±0 . 68 , by another one log10 during the following week , indicating that the organisms were in a replicating state on treatment initiation . Thereafter they plateaued at the 5 . 5 log10 level while the foot pad swelling that began at week 5 after infection , was present in almost all mice by week 6 after infection and progressively worsened up to week 8 . The foot pad swelling occurred very late in few untreated mice , possibly because the foot pad inoculation did result in the implantation of a limited number of M . ulcerans in these mice . At week 1 of treatment , the log10 CFU counts , were significantly ( p<0 . 0001 ) reduced , by at least one log10 , in all three treated groups of mice by comparison with baseline counts at treatment initiation . However they were similar in mice treated with STR+RIF 5/7 or 7/7 , respectively while they were slightly lower ( p>0 . 05 ) by about 0 . 5 log10 , in mice treated with STR+RPT 5/7 . At weeks 2 , 3 , and 4 , the CFU counts continued to decrease at a similar rhythm in mice treated with STR+RIF either 5/7 or 7/7 , with a trend in favor of 7/7 over 5/7 , though the difference was not statistically significant ( p>0 . 05 ) . At the end of 4 weeks of treatment , 2 out of 5 mice treated with STR+RIF 7/7 were culture-negative , while all mice in the group treated 5/7 were culture-positive . In mice treated with STR+RPT , the regression in CFU counts was logarithmic , and resulted in complete culture conversion to negative of the foot pads by week 3 . The bactericidal activity of STR+RPT was thus spectacular and much more pronounced than that of STR+RIF even given 7/7 ( p<0 . 01 ) . The time to foot pad swelling after cessation of treatment with each of the different regimens given for 1 , 2 , 3 , and 4 weeks is depicted in figure 2 . This figure illustrates several important and rather unexpected findings . First , more than 50% ( median ) of untreated control mice exhibited foot pad swelling by week 6 after infection , providing the reference to which median time to swelling in treated mice were compared . Second , in all treated mice whatever the drug regimen and the duration of treatment , the median time to foot pad swelling was considerably delayed , by at least 12 weeks , compared to that in the untreated controls This finding reflects the potent bactericidal activity and/or the post-antibiotic effect of the tested drug regimens , even when they were administered for one week only . Third , as shown in figure 2 , the longer the duration of treatment , the longer the time to swelling and the lower the proportion of mice that developed swelling during the 40 weeks of the experiment . The median times to swelling in mice treated for one and two weeks were 18 weeks and 20–22 weeks , respectively , indicating that a 2-week course of treatment was insufficient to prevent reactivation of the disease after treatment cessation . However , only 30 to 40% of mice treated for 3 weeks exhibited foot pad swelling at week 40 , indicating that a treatment of 3 weeks duration was enough to cure more than half of the infected mice . Only a few mice treated for 4 weeks whatever the drug regimens exhibited foot pad swelling between weeks 25 to 30 , suggesting that a treatment of 4 weeks duration is able to cure close to 100% of mice . Finally , and most surprisingly , none of the drug regimens tested over a period of 1 to 4 weeks was significantly better than the others in preventing foot pad swelling , e . g . reactivation of the disease . In other words , in the current experimental conditions , the RPT-containing regimen did not result in a higher rate of relapse-free cure than the RIF-containing regimens despite its much better bactericidal activity . It should also be noticed that the relapse-free cure rate was not better in mice receiving the STR+RIF regimen given 7/7 than in mice receiving the same regimen given 5/7 .
The primary objective of the current study was to determine whether the increased rifamycin exposure associated with RPT administration would substantially increase the bactericidal activity and decrease the rate of reactivation after treatment cessation which is considered to be a measure of sterilizing activity , of the RIF-based regimen for the treatment of M . ulcerans infection in mice . As expected the STR+RPT combination resulted in better bactericidal activity than the STR+RIF combination . This was demonstrated by the weekly CFU counts and the culture conversion rate to negative by week 3 in all mice treated with the RPT-containing regimen whereas all mice treated with RIF-containing regimens were still culture-positive . Such a finding is in agreement with previous results [9] . However , contrary to our expectations [7] , the more rapid culture conversion rate did not translate into a better cure rate in the RPT treated group as evidenced by the lack of significant differences with the RIF-treated groups in the time to foot pad swelling i . e . , reactivation after treatment cessation . There was also no difference in time-to-swelling between mice treated with STR+RIF given either five or seven days a week . Such unexpected findings raise numerous issues related to the experimental model and the cure potency of the drug regimens used . One possible explanation is technical and related to the one week delay after stopping treatment before cultures from the foot pad of the RPT-treated mice were performed . This procedure aimed to eliminate the risk of RPT carryover due to its long half-life of 15 hr . However , because of RPT's long half-life , M . ulcerans likely continued to be exposed to RPT during the days following treatment cessation . The better bactericidal activity of RPT after one , two and three weeks of treatment may thus result , at least in part , from continuation of RPT exposure after treatment cessation . If mice receiving RPT were effectively treated longer than mice receiving RIF , then that could explain why the bactericidal activity of the RPT-containing regimen was found better than that of RIF-containing regimens . But that cannot explain why the RPT-containing regimen was not better than the RIF-containing regimens in preventing relapse , except if there was no or limited correlation between bactericidal activity and sterilizing activity [19] against M . ulcerans . It has been suggested [20] , [21] that the high protein binding of RPT , 97% versus 85% for RIF [7] , may be partially responsible for its suboptimal activity . It is even possible that such a significant protein binding compensates for the slightly lower MIC of RPT ( 0 . 125 µg/ml ) for M . ulcerans than that of RIF ( 0 . 25 µg/ml ) , especially if rifamycin activity is concentration-dependent [22] . Finally , it also is possible that the potential better sterilizing activity of rifapentine over rifampin which is well established in the murine model of tuberculosis [7] does not matter in the treatment of Buruli ulcer . For the cure of Buruli ulcer , the important thing may not be increasing the antimicrobial sterilizing potency but stopping the production of mycolactone to reverse as soon as possible the local mycolactone-induced immunodeficiency . In that role , rifampin and rifapentine , may have equivalent activity . Blocking the production of mycolactone for a sufficient length of time may , in fact , be more important than the speed of killing the microbe . Eventually , the cure after treatment likely results from the decline in mycolactone content and the development of a specific immune response , both phenomena being closely interlinked . One may even go further and consider that the concept of sterilizing activity which is crucial in the treatment of tuberculosis [23] and leprosy [24] is irrelevant in the treatment of Buruli ulcer , a toxin-derived disease like diphtheria and tetanus . All of these considerations give support to the WHO sponsored trial on the comparative curing effect on Buruli ulcer of the bactericidal STR + RIF combination and the less bactericidal RIF + clarithromycin combination [25] Another point of concern is that the STR+RIF combination had only slightly more antimicrobial activity when it was administered on a daily basis ( 7 days a week ) than when it was administered 5 days a week . This is in contradiction with the findings by Ji et al [11] that the relapse rate after 7 days a week treatment with the STR+RIF combination was significantly lower than after 5 days a week treatment with the STR+RIF combination . The likely explanation of such a difference between the Ji et al . results and ours may be found in the different microbial burdens on treatment initiation . In the Ji et al . experiment , the treatment of M . ulcerans infection was initiated when the infected mouse footpad was clearly swollen and the log10 CFU count has reached a peak at 6 . 09±0 . 19 per foot pad whereas in our experiment treatment was initiated when the foot pad was just beginning to swell and the log10 CFU count was only 4 . 88±0 . 51 per foot pad . The higher CFU count at treatment initiation might have biased the comparison in favor of a 7/7 regimen . Similar observations have been made in the murine model of tuberculosis in which the differences in the initial bacterial burden also influence the relapse rate after treatment cessation [26] . The differences between the Ji et al . data and our data also raise the issue of the more reliable experimental model to assess comparatively the bactericidal potential of drug regimens . Possibly the high microbial burden model with plateauing CFU and the low microbial burden model with CFU in the logarithmic period of growth provide insights that complement one another . As we do not know which model is the most predictive of relapse-free cure in humans , , better information would likely be gained by the simultaneous use of the two models . It is perhaps useful to remind ourselves of the limitations inherent to any experimental model in quoting Georges Canetti: “One should never forget the limitations of experimental studies in animals . They provide useful hypotheses and certain facts not observable in man , but in no case can they replace observations in man for the ultimate understanding of the disease ( and its treatment ) in human beings . ” [27] . | Until 2004 , the treatment of Buruli ulcer ( BU ) was surgical excision followed by skin grafting . Now an 8-week daily regimen of streptomycin and rifampin ( STR+RIF ) is recommended by the World Health Organization , supplemented when necessary by surgical intervention . However , such an antimicrobial treatment is still a heavy burden in settings with limited public health infrastructure . Thus it would be beneficial to shorten the duration of treatment without reducing antibacterial activity . Among the more active drugs having the potential to reduce the duration of treatment is rifapentine ( RPT ) , a rifamycin derivative with a much longer half-life than RIF , which permitted in mice the reduction of treatment duration for tuberculosis from 6 months to 3 months when it is substituted for RIF tuberculosis . We therefore compared in the mouse daily treatment of BU with STR+RIF and with STR+RPT . As expected , treatment with RPT was much more bactericidal than treatment with RIF , but surprisingly the relapse rate and the time to relapse after stopping treatment were not different between treatment groups . Such findings raise numerous issues on the mechanisms involved in the cure of Buruli ulcer and on the impact of bactericidal activities of rifamycin derivatives in this disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"bacterial",
"diseases",
"infectious",
"diseases",
"buruli",
"ulcer",
"neglected",
"tropical",
"diseases"
] | 2013 | Bactericidal Activity Does Not Predict Sterilizing Activity: The Case of Rifapentine in the Murine Model of Mycobacterium ulcerans Disease |
From the microscopic to the macroscopic level , biological life exhibits directed migration in response to environmental conditions . Chemotaxis enables microbes to sense and move towards nutrient-rich regions or to avoid toxic ones . Socio-economic factors drive human populations from rural to urban areas . The effect of collective movement is especially significant when triggered in response to the generation of public goods . Microbial communities can , for instance , alter their environment through the secretion of extracellular substances . Some substances provide antibiotic-resistance , others provide access to nutrients or promote motility . However , in all cases the maintenance of public goods requires costly cooperation and is consequently susceptible to exploitation . The threat of exploitation becomes even more acute with motile individuals because defectors can avoid the consequences of their cheating . Here , we propose a model to investigate the effects of targeted migration and analyze the interplay between social conflicts and migration in ecological public goods . In particular , individuals can locate attractive regions by moving towards higher cooperator densities or avoid unattractive regions by moving away from defectors . Both migration patterns not only shape an individual’s immediate environment but also affects the entire population . For example , defectors hunting cooperators have a homogenizing effect on population densities . This limits the production of the public good and hence inhibits the growth of the population . In contrast , aggregating cooperators promote the spontaneous formation of patterns through heterogeneous density distributions . The positive feedback between cooperator aggregation and public goods production , however , poses analytical and numerical challenges due to its tendency to develop discontinuous distributions . Thus , different modes of directed migration bear the potential to enhance or inhibit the emergence of complex and sometimes dynamic spatial arrangements . Interestingly , whenever patterns emerge , cooperation is promoted , on average , population densities rise , and the risk of extinction is reduced .
In traditional public goods games N individuals gather in a group and each individual chooses to cooperate and invest into a common pool at a cost c , or to defect and shirk the investment . The common pool is multiplied by a factor r > 1 and equally divided among all individuals within the group . This results in a payoff of PD = nC ⋅ rc/N for defectors and PC = PD − ( 1 − r/N ) c for cooperators when facing nC cooperators among the N − 1 other group members . For r < N every participant is tempted to withhold their investment because PC < PD and hence to forego the benefits of the public good to the detriment of all . Had everyone cooperated the payoff would be ( r − 1 ) c > 0 . This conflict of interest between the group and the individual constitutes the social dilemma at the heart of public goods interactions [16–18] . Only for r > N these interests align . In well-mixed populations with normalized densities ( frequencies ) of cooperators , u , and defectors , v = 1 − u , both strategies encounter , on average , nC = u ( N − 1 ) cooperators among their interaction partners in randomly assembled interaction groups . In evolutionary game theory payoffs represent fitness and hence determine the rate of change of the population composition as captured by the replicator dynamics ∂tu = u ( 1 − u ) ( PC − PD ) , where ∂t denotes the time derivative [19] . Consequently , cooperation dwindles , ∂tu < 0 , whenever cooperators obtain lower payoffs in the public goods game than defectors , i . e . if PC − PD = −c ( 1 − r/N ) < 0 . Indeed , for r < N defection dominates and ∂tu < 0 always holds . Conversely , for r > N cooperation spreads in the population , ∂tu > 0 . The replicator equation does not take ecological quantities such as variable population densities into account . The interplay between ecological and evolutionary dynamics is captured by ∂ t u= u [ w ( b + f C ) − d ] ( 1a ) ∂ t v= v [ w ( b + f D ) − d ] ︸ecologicaldynamics , ( 1b ) where w = 1 − u − v reflects reproductive opportunities that diminish for increasing population densities [11] . Selection is based on ecological public goods interactions affecting the birth rates of cooperators and defectors through their average payoffs fC and fD , respectively , while the death rate is constant , d ( see S1 Appendix for further details on dynamics ) . Even in the absence of spatial dimensions , rich dynamics are observed especially under environmental stress where the death rate exceeds the baseline birth rate , d > b [10] . We focus on this case because the survival of the population hinges on the availability of the public good . Varying population densities introduce an interior equilibrium Q , which undergoes a Hopf-bifurcation [20] when increasing the rate of return of the public good , r , giving rise to stable and unstable limit cycles [11] . For r < rHopf , Q is unstable , and well-mixed populations are doomed . However , for r > rHopf , the equilibrium Q allows cooperators and defectors to coexist but the basin of attraction of the equilibrium Q remains limited . More specifically , if cooperators are too rare for the public goods production to offset the death rate d , or if defectors abound , the population may still go extinct . Our baseline scenario can be interpreted as a microbial population in a biocide where the multiplication factor r reflects the effectivity of the public good to oppose the detrimental effect of the toxic environment . The following extensions explore the effects of different modes of motility on the public goods production and survival of the population .
The spatial dynamics of ecological public good interactions with undirected ( diffusive ) migration can be formulated as a selection-diffusion process ∂ t u= u [ w ( b + f C ) − d ]+ D C ∇ 2 u ( 2a ) ∂ t v= v [ w ( b + f D ) − d ] ︸ecologicaldynamics+ D D ∇ 2 v ︸ undirectedmigration ( 2b ) where the diffusion constants DC and DD reflect the migration of cooperators and defectors , respectively [12] . In spatial settings , undirected migration promotes coexistence of cooperators and defectors even if the coexistence equilibrium Q is unstable . In particular , if defectors run faster than cooperators , DD > DC , the dynamics exhibit Turing instabilities , where cooperators serve as activators and defectors as inhibitors . This enables the population to survive through spontaneous pattern formation . Close to the Hopf-bifurcation with r < rHopf ( Q unstable ) , the influence of temporal oscillations gives rise to chaotic spatio-temporal dynamics [21] , which facilitate population survival even for equal diffusion rates , DD = DC . Conversely , for slower diffusion of defectors , DD < DC , activation through cooperation is no longer tenable , spatial effects disappear and either results in homogeneous coexistence for r > rHopf ( Q stable ) or extinction for r < rHopf . This implies that attempts of cooperators at outrunning defectors are futile . Here , we extend the spatial ecological public goods model , Eq ( 2 ) , by introducing directed migration , which enables cooperators and defectors to bias their movement towards more attractive regions . ∂ t u= u [ w ( b + f C ) − d ] + D C ∇ 2 u − A C ∇ · ( u w ∇ u ) + R C ∇ · ( u w ∇ v ) ( 3a ) ∂ t v= v [ w ( b + f D ) − d ]︸ecologicaldynamics+ D D ∇ 2 v ︸ undirectedmigration − A D ∇ · ( v w ∇ u ) ︸ attraction towardcooperation + R D ∇ · ( v w ∇ v ) ︸ repulsion fromdefectors , ( 3b ) with AC , AD , RC , RD ≥ 0 and where terms of the form −K∇⋅ ( ϕw∇ψ ) reflect that individuals of type ϕ are attracted to the gradient of type ψ proportional to reproductive opportunities , w , at a non-negative rate K ( for a detailed microscopic derivation see S2 Appendix ) . The density of cooperators directly translates into the rate of production of public goods and hence serves as a proxy for its availability and the quality of the environment . For example , the term −AD∇ ⋅ ( vw∇u ) reflects hunting defectors in search of public goods that are attracted to higher densities of cooperators at a rate AD . Note that the negative sign indicates movement toward higher densities . Similarly , −AC∇ ⋅ ( uw∇u ) represents aggregating cooperators that are attracted to their kind . In contrast , the density of defectors serves as a proxy to avoid exploitation and reduce competition . Instead of directly sensing the density of defectors their excretions from public goods consumption could reflect the poor quality of the environment . Thus , RC∇⋅ ( uw∇v ) reflects fleeing cooperators that avoid higher densities of defectors at a rate RC , whereas RD∇ ⋅ ( vw∇v ) refers to spreading defectors that steer clear of their kind . Rich dynamics unfold under directed migration as showcased in Fig 1 , ranging from the spontaneous formation of quasi-stable or stable patterns ( Fig 1A ) and ever-changing , chaotic spatio-temporal dynamics ( Fig 1B ) to cooperators aggregating under a self-reinforcing migration response ( positive feedback , Fig 1C ) .
Public goods production and population survival are both crucially linked to the spontaneous emergence of spatial patterns . Here we derive necessary conditions for the onset of pattern formation . An analytical understanding of the pattern formation process is obtained by considering homogeneous population densities reflecting the coexistence equilibrium Q of the well-mixed population dynamics , Eq ( 1 ) . A small perturbation vector ϵ ( exp ( ik ( x + y ) ) , exp ( il ( x + y ) ) ) of mode k ( for cooperators ) and l ( for defectors ) may get exponentially amplified by the dynamics and give rise to the emergence of heterogeneous density distributions , also called Turing patterns [13] . The temporal mode is reflected by k = l = 0 whereas k , l > 0 refers to spatial modes . Because defector survival depends on the public goods production by cooperators , the stationary spatial density patterns of the two types are highly correlated resulting in similar periodicity of the two types ( see Fig 2 ) . For this reason , we restrict the perturbation analysis to k = l ( for more information , see S3 Appendix ) . The linearized dynamics in the vicinity of Q = ( ueq , veq , weq ) is given by ( ∂ t u ^ ∂ t v ^ ) = ( J I + k 2 J S ) · ( u ^ v ^ ) , ( 4 ) where u ^ = u − u eq , v ^ = v − v eq and JI , JS represent the Jacobians from public goods interactions and spatial migration ( directed and undirected ) , respectively: J I =[ a CC a CD a DC a DD ] =[ − u eq d / w eq + u eq w eq ∂ u f C − u eq d / w eq + u eq w eq ∂ v f C − v eq d / w eq + v eq w eq ∂ u f D − v eq d / w eq + v eq w eq ∂ v f D ] , ( 5a ) J S =[ s CC s CD s DC s DD ] = 2 ·[ A C u eq w eq − D C − R C u eq w eq A D v eq w eq − R D v eq w eq − D D ] . ( 5b ) The largest eigenvalue ( real part ) of JI + k2JS is a function of k and is called the dispersion relation λ ( k ) ( see S3 Appendix ) . Modes with λ ( k ) > 0 are unstable indicating the potential for spatial patterns to emerge . If the spatial Jacobian JS admits eigenvalues with positive real parts then λ ( k ) > 0 holds for all sufficiently large k , i . e . all spatial perturbations of sufficiently high frequency are unstable . This is not meaningful in natural systems . In order to ensure λ ( k ) < 0 for k → ∞ , the condition det ( JS ) > 0 or , more precisely , DC > ACueqweq is required , which means that at the homogeneous equilibrium , the effects of diffusion ( undirected migration ) need to outweigh effects of cooperator aggregation . In spite of the above restrictions , intermediate modes can nevertheless become unstable through a combination of selection and migration . A necessary condition for pattern formation is − a DD ︸ > 0 ( A C u eq w eq − D C ) − a DC ︸ > 0R C u eq w eq ︸ activation throughcooperator migration > − a CD ︸ > 0A D v eq w eq + a CC ︸ > 0 ( − R D v eq w eq − D D ) ︸ inhibition throughdefector migration , ( 6 ) see S3 Appendix for details . Consequently , faster aggregating cooperators or spreading defectors ( increased AC , RD ) or slower undirected cooperator migration ( decreased DC ) promote the development of heterogeneous patterns by actively or passively promoting cooperator aggregation . In contrast , faster directed migration of defectors towards cooperation or of cooperators away from defection ( increased AD , RC ) as well as slower undirected defector migration ( decreased DD ) all increase exploitation of the public good , reduce the efficacy of cooperator aggregation and hence suppress pattern formation . Fig 3 depicts the competing effects of activating and inhibiting forms of directed migration . The most unstable mode determines the characteristic length scale of the emerging spatial patterns ( see S3 Appendix ) . Hunting defectors , AD , and fleeing cooperators , RC , exhibit similar homogenizing effects . In contrast , significant qualitative differences arise in the promotion of pattern formation due to aggregating cooperators , AC , and spreading defectors , RD , in the sense that aggregating cooperators have the potential to fully suppress defection . In this case neither increases in AD nor in RC are capable of counteracting the positive feedback induced by AC to recover coexistence . The positive feedback between cooperator densities and cooperator migration is intrinsic to aggregating cooperators . Regions of higher cooperator densities attract cooperators , which further increases their densities and , more importantly , the gradient along the periphery . As a result the region exerts an even stronger attraction . In fact , for sufficiently large AC the attraction prevents cooperators from exploring the available space thus the population remains localized . The potential of aggregating cooperators to give rise to discontinuous distributions through positive feedback lies at the core of the necessary condition that diffusion must outweigh aggregation at the homogeneous equilibrium: DC > ACueqweq . However , this condition is conservative and no longer sufficient for heterogeneous distributions . The gradient in cooperator densities , ∇u , promotes further aggregation of cooperators despite being limited by reproductive opportunities , w , see Eq ( 3 ) . As a consequence the gradient further increases and discontinuities in the density distribution can develop . Interestingly , this effect turns out to be strong enough to completely suppress defection ( c . f . Fig 3 ) . However , increasing gradients require finer discretization to numerically integrate the selection-migration dynamics , Eq ( 3 ) , and hence the emerging distributions depend on the discretization . The manifestation of discontinuous distributions becomes increasingly likely for larger values of AC ( see S3 Fig ) but can also be triggered by pattern formation or large gradients in initial distributions . Increased and differing diffusion rates , DD > DC , or spreading defectors , RD , lower the threshold for pattern formation and are additionally required to reliably observe smooth patterns through aggregating cooperators , AC . Directed migration not only promotes or inhibits spontaneous pattern formation but also affects the frequency of cooperators as well as the population density and hence the survival of the population . Whenever patterns develop , cooperation is promoted and population densities consequently rise , see Fig 4 . This effect extends to multiplication factors below rHopf , which cannot sustain unstructured populations . In this domain , undirected migration is capable of inducing spatial heterogeneity for DD > DC and thereby promotes the survival of the population . Aggregating cooperators , AC , and spreading defectors , RD , further promote pattern formation and population survival even under conditions such as DC ≥ DD , where undirected migration alone is unable to support the population , c . f . Eq ( 6 ) and Fig 2A and 2D . Moreover , the positive feedback of aggregating cooperators coupled with limited reproductive opportunities , offers the intriguing prospect that defectors can potentially get crowded out and driven to extinction , see Figs 2A and 4A .
Cooperation is doomed in traditional public goods games—as paraphrased by Hardin’s “Tragedy of the Commons” [9] . In contrast , in ecological public goods interactions cooperators and defectors can coexist through the feedback between variable population densities and interaction group sizes [10] even if the reproductive performance and survival of the population hinges on the availability of public goods . The introduction of continuous spatial dimensions allows population densities and social composition to vary across the domain . For undirected migration the ratio between the rates of cooperator and defector migration is crucial [12 , 22] . For DC > DD , the population either settles in a homogeneous state with densities corresponding to the well-mixed coexistence equilibrium Q , if it is stable ( r > rHopf ) , or goes extinct if Q is unstable ( r < rHopf ) . In either case , attempts of cooperators at avoiding exploitation by outrunning defectors are futile . For DC = DD , chaotic spatio-temporal dynamics enable the population to survive for r slightly below rHopf [21] . However , for DC < DD cooperation is promoted through Turing instabilities that result in quasi-stable or dynamical patterns that increase cooperator frequency , overall population density , and enhance population survival . Through undirected migration heterogenous density distributions of cooperators and defectors spontaneously arise [12 , 22] . Consequently the abundance of public goods varies across space and renders some regions more attractive than others . This provides a natural incentive for directed migration—either to avoid poor areas or to seek out better ones . For example , chemotactic bacteria aggregate in patches in response to excreted attractants [23] or try to escape oxidative stress [2] . Instead of explicitly modelling the concentration of public goods and its waste products to assess the quality of the environment , we use the density of cooperators and defectors as proxies for the availability of public goods and the degree of exploitation , respectively . Thus , seeking cooperation and avoiding defection allows individuals to increase their access to public goods and improve their reproductive potential . Interestingly , even though the two migration patterns appear very similar , they may actually trigger movements in opposite directions . For example , cooperators that seek their kind aggregate at the centre of cooperative areas . The resulting increase in public goods also benefits defectors in that same location and may attract more defectors , which increases competition and exploitation . In contrast , cooperators that avoid defectors tend to migrate towards the periphery of cooperative areas and thereby effectively counteract cooperator aggregation . Analogous arguments apply to defectors seeking cooperators or avoiding their kind , respectively , but with opposite effects . More specifically , spreading defectors indirectly support cooperator aggregation by creating ( temporary ) refuges with low exploitation . As a consequence , directed migration can both enhance as well as inhibit pattern formation . In particular , aggregating cooperators or spreading defectors both promote or even trigger pattern formation . This extends into unfavourable parameter regions that otherwise result in extinction , such as when cooperators outpace defector , DC ≥ DD , or small multiplication factors r , see Fig 2A and 2D . Conversely , fleeing cooperators and hunting defectors suppress pattern formation by levelling out population densities . For r < rHopf this increases the risk of extinction . The complementary effects of the different modes of directed migration are captured in Eq ( 6 ) , which provides an analytical threshold for the onset of pattern formation . Regardless of whether spatial heterogeneities arise through directed or undirected migration , they invariably increase both the average frequency of cooperators as well as the average density of populations as compared to unstructured populations , see Fig 4 , and thus improves the odds of population survival . Traditionally the effects of spatial structure in evolutionary games have been investigated based on lattices or more general network structures [24 , 25] . Such discrete spatial arrangements are capable of supporting cooperation because they enable cooperators to form clusters and thereby reduce exploitation by defectors . This effect is further enhanced by success-driven migration [26] . In contrast , in our setup space is continuous and the state of the population represented by density distributions of cooperators and defectors . This difference is crucial and renders cooperation even more challenging because , in general , the density of defectors may become arbitrarily small but never and nowhere zero within finite times . As a consequence cooperators are always subject to exploitation by defectors . In fact , in the absence of ecological dynamics and constant population densities cooperators invariably disappear ( if r < N ) or , conversely , take over ( if r > N ) , just as in well-mixed populations . Notably , the only exception to this rule refers to the aggregation of cooperators because this mode of directed migration can trigger a positive feedback between densities and migration of cooperators: increases in density due to cooperator aggregation result in steeper gradients , which heightens the attraction and hence further increases their density . As a result , cooperator densities may not only remain localized but also eliminate defectors . However , this positive feedback results in analytical and numerical challenges because it gives rise to discontinuous distributions . Yet , those challenges are common among models incorporating aggregation . For example , the Keller-Segel chemotaxis model predicts infinite population densities after finite times [27–30] . Even though a mathematical artifact , those singularities are associated with the inherent feedback between chemotaxis and the secretion of chemical attractants [15] . Nevertheless , the most striking feature of directed migration is the potential of aggregating cooperators to crowd out and eliminate defectors altogether , see Figs 2A , 3A , 3C and 4A . Unfortunately this intriguing phenomenon is linked to the positive feedback that gives rise to discontinuous distributions and hence eludes further analysis based on the present framework , Eq ( 3 ) . Somewhat surprisingly , the reduced aggregation rate due to a lack of reproductive opportunities , represented by the term w∇u , turns out to be insufficient to maintain smooth numerical solutions . Motility plays an important role in biofilms . Microbes excrete extracellular substances to generate and maintain this protective film . Free-riders benefit from the protection without contributing , which gives rise to the public goods dilemma . Experiments indicate that not only the inherent social conflict plays a vital role in the effective secretion of biofilms [31] but also the microbes motility [32–34] . Biofilms are no longer generated when deactivating the movement apparatus through deliberate mutations [35] . When specifically targeting the ability to chemotact , biofilm production significantly varies across microbes and experimental setups [32] . This sensitivity of public goods production in response to different types of migration is reflected in our model . The complex interplay between ecological public goods and motility shapes population densities and distributions as well as their social composition . Not only does migration affect the production of the public good but some public goods also alter the motility of their producers . For example biofilms increase viscosity and reduce the motility of microbes or even segregate populations [36] . Conversely , Paenibacillus collectively lubricate hard surfaces to enhance population expansion [37] . Either scenario creates intriguing opportunities for novel feedback mechanisms where migration not only shapes the production and availability of the public good but where the public good represents the very infrastructure needed to stay put or migrate more efficiently . | The production and maintenance of shared environmental resources such as access to nutrients in microbial communities or potable water in human societies require the cooperation of groups of individuals . However , cooperation is costly and prone to exploitation . If too many individuals follow selfish interests and spoil their environment , the group and possibly the entire population suffers . Nevertheless , many forms of biological life—from humans to microbes—migrate in response to resource availability . Here , we analyze the interplay of the social conflict in public goods production and targeted migration . In particular , we find that aggregation of cooperators can enhance or trigger the spontaneous formation of heterogeneous spatial distributions , which promote cooperation and result in higher population densities . Conversely , attempts to avoid defectors increases the risk of extinction because it tends to homogenize population distributions and lower population densities . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"biofilms",
"cell",
"motility",
"ecology",
"and",
"environmental",
"sciences",
"population",
"dynamics",
"applied",
"mathematics",
"microbiology",
"social",
"sciences",
"extinction",
"risk",
"developmental",
"biology",
"mathematics",
"animal",
"behavior",
"public",
"goods... | 2019 | Directed migration shapes cooperation in spatial ecological public goods games |
Bacterial capsules are common targets for antibody-mediated immunity . The capsule of Bacillus anthracis is unusual among capsules because it is composed of a polymer of poly-γ-d-glutamic acid ( γdPGA ) . We previously generated murine IgG3 monoclonal antibodies ( mAbs ) to γdPGA that were protective in a murine model of pulmonary anthrax . IgG3 antibodies are characteristic of the murine response to polysaccharide antigens . The goal of the present study was to produce subclass switch variants of the γdPGA mAbs ( IgG3→IgG1→IgG2b→IgG2a ) and assess the contribution of subclass to antibody affinity and protection . Subclass switch antibodies had identical variable regions but differed in their heavy chains . The results showed that a switch from the protective IgG3 to IgG1 , IgG2b or IgG2a was accompanied by i ) a loss of protective activity ii ) a change in mAb binding to the capsular matrix , and iii ) a loss of affinity . These results identify a role for the heavy chain constant region in mAb binding . Hybrid mAbs were constructed in which the CH1 , CH2 or CH3 heavy chain constant domains from a non-protective , low binding IgG2b mAb were swapped into the protective IgG3 mAb . The IgG3 mAb that contained the CH1 domain from IgG2b showed no loss of affinity or protection . In contrast , swapping the CH2 or CH3 domains from IgG2b into IgG3 produced a reduction in affinity and a loss of protection . These studies identify a role for the constant region of IgG heavy chains in affinity and protection against an encapsulated bacterial pathogen .
Bacillus anthracis , the causative agent of anthrax , is a gram-positive , spore-forming bacterium [1] . Due to the ease of infection and high lethality , B . anthracis has been listed by the Centers for Disease Control and Prevention as one of the Category A agents of bioterrorism . Virulent strains of B . anthracis carry two large plasmids , pXO1 and pXO2 , that encode genes needed for toxin production and capsule formation , respectively [2] , [3] . Anthrax toxins are composed of protective antigen ( PA ) combined with lethal factor ( LF ) or edema factor ( EF ) to form active toxins [4] , [5] . The polypeptide capsule is composed of poly-γ-d-glutamic acid ( γdPGA ) [6] and is both poorly immunogenic and antiphagocytic [7]–[9] . The current vaccine , anthrax vaccine adsorbed ( AVA , Biothrax ) , is composed of aluminum hydroxide-adsorbed , formalin-treated , cell-free filtrate of a non-capsulated B . anthracis strain [10] . The active component of AVA is believed to be PA . However , there has been recent interest in targeting γdPGA as an addition to a vaccine [11]–[16] . Antibodies to γdPGA are opsonic [11] , [12] , [15] , [17] . γdPGA-based immunity is attractive because such immunity would interdict the infection when the bacterial load is low and would prevent infection from reaching the stage where large amounts of toxin are formed . As a consequence , γdPGA-specific antibodies could exhibit potent synergy with toxin-targeted immunity [11] . We recently reported production of monoclonal antibodies ( mAbs ) specific for the γdPGA capsule [13] , [14] . Passive immunization with murine mAbs of the IgG3 subclass was protective in a murine model of pulmonary anthrax [13] , [14] . This was the first report of protection mediated by B . anthracis capsular antibodies . The ability of capsular antibodies to protect is related to the level of antibody generated in response to immunization as well as qualitative factors such as affinity and effector functions . Antibody affinity for antigen is largely determined by the variable regions of the amino-terminal domains of the heavy and light chains , whereas the effector functions and the IgG subclass are determined by the constant region of the heavy chain . In the mouse , the IgG subclasses are IgG1 , IgG2a , IgG2b and IgG3 . The murine constant region of the IgG heavy chain consists of three domains , CH1 , CH2 , and CH3 . The region between the CH1 and CH2 domains is called the hinge region and permits flexibility in the chain . A successful active or passive immunization strategy that targets γdPGA will require an understanding of the roles of IgG subclass in protection . The overall aim of this study was to assess the contribution of the constant region of the murine IgG heavy chain to protection in a murine model of inhalational anthrax . The results showed that murine IgG3 is highly protective , but the IgG1 , IgG2a and IgG2b subclasses are poorly or non-protective and have a markedly reduced affinity compared to IgG3 antibodies . Hybrids of the protective IgG3 antibody having CH1 , CH2 or CH3 domains of the non-protective IgG2b antibody were constructed in an effort to better understand the contribution of heavy chain domains to protection and antibody affinity . The results showed that the CH2 domain and to a lesser extent the CH3 domain were major determinants of both antibody affinity and protection .
The IgG3 mAbs F24F2 and F26G3 were produced from mice immunized with γdPGA in combination with agonist mAbs to CD40 [13] . Full subclass-switch families ( IgG3→IgG1→IgG2b→IgG2a ) from the parental F24F2 and F26G3 IgG3 mAbs were generated using sequential sib selection [18] . Once all eight mAb-secreting hybridoma cell lines were obtained , total mRNA was isolated and used for cDNA synthesis and PCR amplification . Sequencing of PCR products verified that the heavy- and light-chain variable regions were identical in all four subclasses ( data not shown ) , and the heavy chain sequences corresponded to those reported for murine IgG3 , IgG1 , IgG2b , and IgG2a [19] . An initial experiment determined protection afforded by the subclass switch variants in a murine model of pulmonary anthrax . Mice were treated with various doses of each mAb and challenged 18 h later with a lethal dose of B . anthracis ( Ames ) spores . The results from evaluation of subclass switch families of both mAbs F24F2 and F26G3 showed that only the IgG3 mAbs were protective; similar results were found with subclass switch families of both parental cell lines ( Fig . 1A ) . Treatment with the IgG3 mAbs at doses between 125–2000 µg/mouse increased the overall percent survival of mice in a dose-dependent fashion . Treatment with IgG1 , IgG2b , or IgG2a mAbs did not significantly increase the overall percent survival at any treatment dose ( P>0 . 05 ) . The capsular quellung reaction is a means to evaluate interactions of antibodies with an intact capsule [20] . In previous studies of the opportunistic yeast Cryptococcus neoformans , we found that protective antibodies cross-link the capsule edge to produce an annular rim that is visible by differential interference contrast ( DIC ) microscopy [21] . Non-protective antibodies bind throughout the capsular matrix and form a puffy-type reaction that does not cross-link the edge . In this experiment , killed B . anthracis Ames bacilli were incubated with the subclass switch families of mAbs F24F2 and F26G3 , and capsule reactions were analysed by DIC microscopy . The results ( Fig . 1B ) showed that the IgG3 mAbs produced capsule reactions with annular rim-type patterns at the outer edge of the capsule that are typical of protective mAbs in the C . neoformans system ( Fig . 1B , red arrow ) . In addition , the IgG3 mAbs produced the previously described dual-capsule reaction [14] where there was a distinct reaction at a layer that was well beneath the capsule surface in the region of the cell wall ( Fig . 1B , blue arrow ) . In contrast , switching from the IgG3 to IgG1 , IgG2b , or IgG2a subclasses was accompanied by a switch to a puffy type of capsular reaction that was characteristic of non-protective mAbs in the C . neoformans system . Neither the outer rim pattern nor the pronounced inner layer of reactivity near the cell wall was observed with the switch variants . Similar results were found with subclass switch families derived from both parental mAbs . One explanation for the failure of the IgG1 , IgG2b and IgG2a variants to both protect and produce the annular rim seen via DIC microscopy is the possibility that subclass switching was accompanied by a reduction in affinity such that the mAb could not cross-link the capsular matrix . As a consequence , the functional affinity of each member of the two subclass switch families was determined by surface plasmon resonance analysis . Functional affinity is defined as the overall strength of binding between a multivalent antigen and a multivalent antibody [22]–[24] . Equilibrium binding was assessed using sensor chips coated with a 25-mer of γdPGA . The results are shown in Fig . 2A as a Scatchard plot in which RUeq/concentration is plotted against RUeq . Fig . 2B provides a summary of the Scatchard analysis where the results are shown as KD . Results in Figs . 2A and 2B show that switching from IgG3 to IgG1 , IgG2b or IgG2a was accompanied by a drop in functional affinity . An alternative means to estimate relative functional affinity is the ELISA in which microtiter plate wells are coated with an excess of native γdPGA and antibody binding is determined via standard ELISA techniques . The results ( Fig . 2C ) showed that the amount of mAb required to produce a standard signal ( OD450 = 2 . 0 ) with the IgG3 subclass was at least 15-fold less than was required for the subclass switch variants . These results confirm that switching from IgG3 to other subclasses produces a dramatic drop in functional affinity . Analysis of mAb binding via surface plasmon resonance used chips that were coated with γdPGA oligomers of 25 residues . Experiments done under these conditions will allow for bivalent binding and therefore only measure functional affinity [25] . As a consequence , a second experiment was done in which fluid phase binding was assessed using ( γ-d-Glu ) 5 . A 5-mer peptide interacts with γdPGA mAbs via a monovalent binding process and measures intrinsic affinity [14] , [23]–[25] . Intrinsic affinity is the interaction between one antibody paratope ( binding site ) and one epitope ( antigenic determinant ) [24] . Intrinsic affinity was measured using the change in intrinsic fluorescence ( fluorescence perturbation , ΔF ) of the mAb when incubated with increasing amounts of synthetic ( γ-d-Glu ) 5 . The results ( Fig . 3 ) showed that the loss of functional affinity in switching from IgG3 to other subclasses shown in Fig . 2 is reflected in a loss of intrinsic affinity ( 18–120 fold loss relative to the parent IgG3 ) . Indeed , the loss in intrinsic affinity with subclass switching was several-fold greater than the loss in functional affinity . Results in Figs . 1 , 2 and 3 indicated that some portion of the murine IgG heavy chain constant region influences both affinity for γdPGA and the ability of an antibody to protect . The constant region of the murine IgG heavy chain consists of the hinge region and three constant region domains: CH1 , CH2 and CH3 . A series of experiments was done in an effort to identify the portion of the heavy chain that contributes to the affinity and protective activity of the IgG3 γdPGA mAbs . To this end , hybrid antibodies were constructed that contained the framework of the IgG3 heavy chain substituted with the CH1 , CH2 and CH3 domains of the low-affinity , non-protective F26G3 IgG2b subclass switch variant . A schematic that illustrates the composition and nomenclature of the various hybrid antibodies is shown in Fig . 4 . The protective and binding properties of the hybrid mAbs were then evaluated in the same manner as described for the subclass switch variants . Passive protection was assessed using a dose of 800 µg/mouse . This dose was chosen because it fell slightly below the maximal level of protection produced by the 1000 and 2000 µg doses of the F26G3 IgG3 and would be sensitive to a loss of protection produced by swapping of domains from the non-protective IgG2b . The results ( Fig . 5 ) showed a high level of protection following treatment with the parent IgG3 or an IgG3 that contained the CH1 domain of the IgG2b . There was a significant ( P = 0 . 02 ) loss of protection if the IgG3 contained the CH2 domain of the IgG2b . Capsular quellung reactions were determined for the parent F26G3 IgG3 or hybrid IgG3 mAbs that contained each of the heavy region constant domains from the IgG2b . The results ( Fig . 5 ) showed rim reactions or binding patterns that resemble a bead of pearls for the IgG3 and the IgG3 hybrid mAb containing CH1 from IgG2b ( IgG3-CH1 . 2b ) . Binding to an inner layer of the capsule near the cell wall was also evident with the IgG3 hybrid mAb containing the CH1 domain from IgG2b . In contrast , hybrid IgG3 substituted with either the CH2 ( IgG3-CH2 . 2b ) or CH3 ( IgG3-CH3 . 2b ) domains of the IgG2b produced a puffy reaction that was identical to capsule reactions produced by the non-protective IgG1 , IgG2b and IgG2a subclass switch variants . Binding activities of the parent F26G3 IgG3 and the hybrid mAbs engineered with the CH1 , CH2 or CH3 domains from IgG2b were assessed via surface plasmon resonance . The Scatchard plot is shown in Fig . 6A . There was a clear hierarchy of dissociation constants ( KD ) for the antibodies with IgG3∼IgG3-CH1 . 2b<IgG3-CH3 . 2b<IgG3-CH2 . 2b<IgG2b ( Fig . 6B ) . Finally , binding behavior of the parental and heavy-chain hybrid antibodies was assessed by ELISA . Microtiter plates were coated with γdPGA and incubated with serial dilutions of each mAb . The data are reported as the concentration of each mAb that produced an OD450 = 2 . 0 . The results directly tracked results from the Scatchard analysis with the amounts of mAb produced to reach an endpoint: IgG3 = IgG3-CH1 . 2b<<IgG3-CH3 . 2b<IgG3-CH2 . 2b<IgG2b ( Fig . 6C ) . The differences in charges between the various CH domains might contribute to the binding activity; polyglutamic acids have considerable negative charges ( −1 per residue ) . As a consequence , we calculated the electrostatic potentials of CH1 , hinge , CH2 , and CH3 of each murine IgG subclass as well as the net charge of the full molecules . The results of the calculations are reported in Table 1 . The KD values of the F24F2 and F26G3 subclass switch variant mAbs determined by fluorescence perturbation assays ( Fig . 3 ) were converted to the free energy of dissociation ( ΔG0 ) , plotted against the respective charge of CH1 , CH2 , CH3 , the hinge region and the full antibody molecule of each subclass each mAb , and a correlation was determined ( Fig . 7 ) . There was no significant correlation between the net charge of the full antibody molecule and binding energy ( r = 0 . 53; P = 0 . 18 ) . Further analysis also showed no correlation between charge and binding energy for the CH1 ( r = 0 . 23; P = 0 . 59 ) and the hinge regions ( r = 0 . 64; P = 0 . 085 ) . The two domains that comprise the Fc portion of the antibodies had higher but opposing correlation coefficients: CH2 ( r = 0 . 87; P = 0 . 005 ) and CH3 ( r = −0 . 85; P = 0 . 008 ) .
Many bacteria are readily killed by phagocytic cells . As a consequence , pathogens may produce extracellular capsules that inhibit phagocytosis and killing . Mammalian hosts in turn produce antibodies that bind to capsules to facilitate phagocytosis or complement-mediated killing of extracellular bacteria . As a consequence , targeting capsules through active or passive immunization is a key strategy for control of many bacterial diseases . Immunity that targets bacterial capsules is , therefore , critically dependent on the interaction of antibodies with capsular antigens [26] . The immune responses to different capsular antigens share many features ( reviewed in [26] ) . Most capsules are polysaccharide in composition and have identical repeating units composed of one to six monosaccharides . Most capsular polysaccharides lack T cell-dependent immunogenicity , do not induce a booster antibody response , and fail to produce protective serum antibody in infants and young children . Finally , the antibody response to immunization with capsular polysaccharides is highly restricted; mice produce primarily IgG3 [27] , [28] and humans produce primarily IgG2 ( reviewed in [29] ) . This restriction has led to speculation as to the advantage to the host of murine IgG3 or human IgG2 over other IgG subclasses in interaction with capsular polysaccharides [22] , [30] . The B . anthracis capsule is composed of a polypeptide ( γdPGA ) rather than a polysaccharide . However , γdPGA has many structural and immunological properties in common with capsular polysaccharides described above , including being a T-independent antigen [8] , [11] . Given the similarities between γdPGA and capsular polysaccharides , a major goal of this study was to evaluate the nature of the interaction between γdPGA and its cognate antibodies and to place these results in the larger context of antibody binding to capsular polysaccharides . Results from the present study identify a key role for antibody constant regions in i ) intrinsic affinity , ii ) functional affinity , iii ) patterns of binding to the capsule and iv ) protection in a murine model of pulmonary anthrax . Construction of hybrid antibodies in which heavy-chain domains of the protective IgG3 mAb were replaced with their respective domains from a low-affinity and non-protective IgG2b mAb showed that the CH1 domain did not contribute to either protection or affinity . The primary contribution of heavy chain domains to protection and affinity was found to be in the CH2 domain with a lesser contribution by the CH3 domain . The classical explanation for binding between antigen and antibody is that of complementarity between the antibody binding site ( paratope ) and a monovalent ligand ( epitope ) . CH domains contribute to functional affinity due to the polyvalent nature of intact immunoglobulins . However , this classical view has been challenged in two series of studies of antibody binding to polysaccharide antigens and a recent report of antibody binding to HIV-1 [31] . The binding of mAbs to γdPGA adds to this body of evidence for a role of CH domains in antibody binding , but the mechanism appears to be fundamentally different from previous reports . In one example , Casadevall and colleagues examined a subclass switch family of mAbs against the glucuronoxylomannan ( GXM ) capsule of C . neoformans . GXM IgG2b , IgG2a , and IgG1 mAbs were protective in a murine model of cryptococcosis , whereas IgG3 mAbs provided no protection [32] , [33] . Evaluation of thermodynamic and kinetic parameters of mAb binding found the non-protective IgG3 mAb to have the lowest binding affinity within the set [34] . The binding properties of the parent antibodies were largely reflected in the binding of Fab fragments , suggesting that structural differences in the CH1 domain are responsible for the differences in affinity constants . Our studies of γdPGA mAbs differ from those of GXM mAbs in two fundamental respects . First , γdPGA mAbs of the IgG3 subclass showed the greatest affinity and highest level of protection; IgG3 mAbs binding to GXM showed the lowest affinity and the lowest level of protection . Second , switching of domains from the low affinity , non-protective IgG2b variant of γdPGA mAb F26G3 into the IgG3 framework localized both the protective effect and high affinity binding to the CH2–CH3 domains rather than the key role played by CH1 in binding of GXM mAbs . In the second example , seminal studies by Greenspan and colleagues found that IgG3 mAbs reactive with the N-acetyl-glucosamine ( GlcNAc ) residues of streptococcal group A polysaccharide had a higher functional affinity than variable region-identical subclass switch IgG1 and IgG2b mAbs [35]–[37] . The high functional affinity of IgG3 antibodies was due to Fc-dependent cooperative interactions . Enhanced functional affinity required an intact Fc region , identifying CH2–CH3 as the region of the IgG3 heavy chain required for cooperative interaction . Binding of mAbs to γdPGA was similar to binding of GlcNAc mAbs in two key respects . First , IgG3 mAbs showed the highest functional affinity relative to other subclasses . Second , the enhanced functional affinity of both the γdPGA and GlcNAc IgG3 antibodies was due to the CH2–CH3 domains . Indeed , the available evidence ( Fig . 6 ) suggests that the CH2 domain plays a larger role in affinity than CH3 . However , there is a key fundamental difference between the binding activities of γdPGA and GlcNAc mAbs that suggests different mechanisms . In the case of the GlcNAc mAbs , the binding advantage of the IgG3 subclass was attributed to enhanced functional affinity that results from an increased valence due to Fc cooperativity . In contrast , the binding advantage of the intact IgG3 γdPGA mAb ( functional affinity determined via SPR or ELISA , Fig . 2 ) was directly reflected in the higher intrinsic affinity ( fluorescence perturbation with a 5-mer peptide ) of the mAb relative to other subclasses ( Fig . 3 ) . Indeed , the differences in intrinsic affinity of mAbs of different subclasses were considerably greater than the respective differences in functional affinity . These results indicate that cooperative binding alone cannot explain the contribution of heavy chain domains to affinity of γdPGA mAbs . A third potential contribution by CH domains to antibody binding was suggested by Morelock et al . [38] who found that differences in the functional affinity of a family of murine/human chimeric mAbs ( huIgG1>huIgG4>huIgG2 ) reflected the flexibility of the hinge region ( IgG1>IgG4>IgG2; [39] ) . In contrast , our studies of variation in the functional affinity of γdPGA mAbs ( IgG3>IgG1>IgG2b∼IgG2a ) did not reflect the hierarchy in flexibility of the murine IgG ( IgG2b>IgG2a>IgG3>IgG1; [39] ) . Our studies support previous arguments that the heavy chain constant regions of IgG3 contribute to binding to antigen [35] , [40]–[42] . The evolutionary or adaptive value for development of enhanced binding by IgG3 is not known . As noted by Greenspan and Cooper [42] , the IgG3 C region gene is the most highly conserved of the Ig heavy chain C region genes among inbred mice [43] , [44] . Involvement of CH domains in binding ( positive or negative effects ) may be a general characteristic of antibodies to antigens with repeating subunits such as those in GXM , GlcNAc or γdPGA . Notably , GXM and γdPGA are both T-independent antigens . Production and evaluation of CH domain hybrids found the greatest contribution of the IgG3 heavy chain to antibody binding to be in CH2 in two distinct assay systems – surface plasmon resonance and ELISA ( Fig . 6 ) . The available data do not provide insights into a mechanism by which CH2 can contribute to affinity . One possibility is glycosylation . CH2 is heavily glycosylated; however , our preliminary studies have found that deglycosylation of IgG3 has no effect on functional affinity ( unpublished results ) . A second possible explanation for the contribution of CH domains to antibody binding is that of charge . Charge could contribute to binding by direct electrostatic or coulombic attraction between the antibody and the negatively charged γdPGA . However , the measured binding energy of each antibody varied little with the net charge of the whole antibody ( Fig . 7 ) . Further dissection of the electrostatic charges that are present on individual antibody domains showed no correlation between electrostatic charge of the CH1 and hinge . We found a significant correlation between electrostatic charge and the respective binding energies when the CH2 and CH3 were analysed; however , the correlation is opposite for the two domains . In unpublished studies , we have found that neither mAb F24F2 or F26G3 showed any appreciable binding to poly-l-glutamic acid . Because poly-l-glutamic acid carries the same charge as γdPGA , it is unlikely that simple coulombic interactions are major contributors to the differences in binding energy seen among the four subclasses of murine IgG . However , our results do not exclude the possibility that charge in CH2 influences conformation or permissivity in the antibody paratope . An alternative mechanism by which the C domain could influence antibody binding is by imposing structural constraints on the V region that alter the chemical and/or electronic environment within the antibody paratope or the ability to undergo a conformational change on epitope binding [45] , [46] . For example , Janda et al . recently reported that the murine IgG constant regions influenced the energy landscape of the variable region [46] . Our data do not provide information on possible effects of CH on the antibody paratope , but the results map the effects of CH on binding to the CH2/CH3 regions . Considerable effort has gone into mutagenesis of antibody variable regions to produce mAbs with enhanced affinity . Enhancement of affinity may improve the therapeutic efficacy of many mAbs or enhance the sensitivity of diagnostic tests in which antibodies are the test reagent . Identification of a region in CH2 that contributes to affinity may allow for affinity-enhancement of γdPGA mAbs of the IgG1 , IgG2b or IgG2a subclasses if the active regions of the IgG3 were swapped into mAbs of other subclasses . More interesting is the possibility that swapping the affinity-enhancing region of IgG3 CH2 into other mAbs with specificity for polysaccharides or proteins could produce a significant affinity enhancement . If this were the case , a single antibody engineering solution might improve antibody performance without the need for affinity-enhancement of each mAb variable region . Finally , limitations to the experimental design should be noted . The experimental approach we used was that of loss of function , which showed a striking congruence between loss of protection and loss of binding activity as shown by i ) interactions with the capsular matrix and ii ) binding activity via SPR and ELISA . Future studies will examine gain of function and produce hybrid mAbs in which smaller segments of CH2 from IgG2b are exchanged into IgG3 ( and vice versa ) to further refine the structural requirements for affinity enhancement by the IgG3 heavy chain . We also recognize that the protective activity of the IgG3 mAbs likely includes a mosaic of factors of which affinity is one critical element . The biological activities of IgG antibodies are heavily influenced by the distinct biological activities of each subclass and the engagement of different Fc receptors by antibodies of different subclasses . However , the biological activities of IgG3 do not distinguish themselves in a manner that would suggest greater protective activity relative to the other subclasses . For example , mouse IgG3 has a segmental flexibility that is slightly greater than IgG1 but less than IgG2a or IgG2b [39] . Like IgG2a and IgG2b , mouse IgG3 is a potent activator of the classical complement system [39] , [47] . Mouse IgG3 binds to mouse FcγRI with low affinity but shows limited or no binding to mouse FcγRIIB , FcγRIII or FcγRIV [48]–[50] . In summary , our studies identify important parameters for antibody-mediated protection that targets the B . anthracis capsule . The CH2 domain of murine IgG3 is critical for protection , intrinsic affinity and functional affinity . The involvement of the CH2 domain in affinity was found to be a novel mechanism by which the IgG heavy chain contributes to antibody affinity and protection . Finally , these results support arguments that the influence of antibody heavy chain regions on antibody binding should be an important consideration in the development of genetically engineered antibodies for therapeutic use [45] , [51] . Optimization of mAbs as therapeutic agents will require a better understanding of the interactions between antibody constant and variable regions and the effects of such interactions on antigen binding .
This study was carried out in accordance with recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . These specific protocols were approved by the University of New Mexico Institutional Animal Care and Use Committee ( Animal Welfare Assurance Number AS3350-01 ) . Full subclass-switch families ( IgG3→IgG1→IgG2b→IgG2a ) were generated from parental F24F2 and F26G3 anti-γdPGA IgG3 mAbs by the sequential sublining method of Spira et al . [18] . Both mAbs are VH IMGT subgroup IGVH10S2*02 and VL subgroup IGKV1-135*01; however , the mAbs differ in the JH and JL regions [14] . Accession numbers for the V sequences are available online at the GenBank database ( accession numbers EF030730 and EF030736 for VH and VL of F24F2 , respectively , and EF030731 and EF030737 for VH and VL of F26G3 ) . Switching from one subclass to another follows the germline order of heavy chain exons ( γ3 , γ1 , γ2b , γ2a ) . The procedure was done in a sequential manner to obtain hybridomas that secrete mAbs of the IgG1 , IgG2b and IgG2a subclasses . Once hybridoma cell lines were established , total mRNA was isolated from cells using a Straight A's mRNA Isolation System ( Novagen Inc . , Madison , WI ) and cDNA synthesis using a First Strand cDNA Synthesis Kit ( Novagen Inc . ) with poly-T primers . cDNA was PCR amplified using 5′ end primers specific for the heavy and light chain variable region of F24F2 and F26G3 . PCR products were sequenced by Nevada Genomics Center to verify identical variable region sequence and correct subclass sequence . CH domain hybrid mAbs were generated with an overlap PCR technique using F26G3 IgG3 and IgG2b template DNA . PCR primers were designed with overlapping sequences and are listed in Table S1 . Platinum PCR SuperMix High Fidelity ( Invitrogen , Carlsbad , CA ) was used for all PCR reactions . Briefly , for generation of F26G3 IgG3-CH1 . 2b mAb the following PCR reactions were performed , i ) PCR1 forward and reverse primers were combined with F26G3 IgG3 template , ii ) PCR2 primers were combined with F26G3 IgG2b template , and iii ) the resulting PCR1 and PCR2 fragments were purified , combined and amplified with PCR3 primers to generate a full length IgG3-CH1 . 2b PCR fragment . This fragment was TA cloned to pGEM-T easy vector and transformed to NEB 5-alpha competent E . coli cells ( New England BioLabs Inc . , Ipswich , MA ) . Positive clones were sequenced at the Nevada Genomics Center . The F26G3 IgG3-CH1 . 2b sequence was cleaved from the pGEM vector with XbaI and AgeI restriction enzymes ( New England BioLabs Inc . , Ipswich , MA ) and cloned to pcDNA3 . 3-TOPO expression vector ( Invitrogen/Life Technologies , Grand Island , NY ) that was digested with the same enzymes . After transformation to E . coli cells , positive clones were purified and sequenced . The additional heavy chain hybrid mAbs ( IgG3-CH2 . 2b and IgG3-CH3 . 2b ) were cloned in a similar fashion . F26G3 light chain was PCR amplified with primers ( Table S1 ) and cloned to pOptiVEC-TOPO vector ( Invitrogen ) according to product manual . Each heavy chain hybrid plasmid was separately transfected with the F26G3 light chain pOptiVEC-TOPO plasmid into a DG44 cell line , and the hybrid mAbs were produced according to the OptiCHO Antibody Express Kit product manual ( Invitrogen ) . All hybrid CH mAbs were isolated from growth medium by affinity chromatography on protein A ( GE Healthcare , Piscataway , NJ ) . γdPGA for immunochemical assays was isolated as described [14] from culture filtrates of B . licheniformis strain 9945 that was grown on Medium E that contained 2 mM MnCl2•4 H2O to stimulate maximal production of PGA in the D isoform [52] . An acid hydrolysate of the purified γdPGA exhibited a specific optical rotation ( −25 . 2° ) indicating that ∼84% of the glutamic acid was the D isomer . The amino acid sequences of the CH1 , hinge region , CH2 , and CH3 of each murine IgG subclass were obtained from the ImMunoGeneTics online database ( www . IMGT . org ) . These sequences were used to calculate the net charge of each murine IgG subclass as a whole , and the net charge of each antibody domain/hinge region separately . The net charges were calculated assuming an environment of pH 7 . 4 . Net charge of each immunoglobulin domain was determined by use of PepCalc ( Innovagen AB ) . The binding affinities ( KD ) of each murine IgG subclass of mAbs F24F2 and F26G3 were used to calculate the thermodynamic free energy of dissociation ( ΔG0 ) using the equation: ΔG0 = −RTlnKD , where R is the gas constant ( 1 . 986 cal K−1 mol−1 ) , T is the temperature ( K ) , and KD is the dissociation constant ( M ) of each antibody as determined by fluorescence perturbation ( from Fig . 3 ) . Graphs plotted net charge and the charge of each domain against ΔG0 for each antibody . Correlation was evaluated using the Pearson product-moment correlation coefficient ( r ) . BALB/c mice were treated intraperitoneally with mAbs diluted with DPBS or with vehicle ( DPBS ) alone as a control . The mice were infected 18 h after mAb treatment by intratracheal challenge with 104 ( 10 LD50 ) B . anthracis Ames strain spores in 50 µl . The actual number of spores deposited in the lung was determined for each experiment by sacrificing three mice following infection , homogenizing the lungs in 1 ml of DPBS , culturing serial dilutions on sheep blood agar plates , and averaging the number of colony forming units . Survival and clinical signs were monitored daily for 14 days post-infection . Percent survival was compared to the vehicle-treated control by the Fisher exact test ( SigmaStat 3 . 5 , Systat Software , Inc . ) . | The ability of an antibody to recognize and bind to its target is classically viewed as a function of the variable region of the molecule; this region distinguishes an antibody with one specificity from an antibody with a different specificity . We examined binding of antibodies to an outer coat of the biothreat Bacillus anthracis that is essential for bacterial virulence . We identified regions of the antibody constant region which contribute to antibody binding and the ability of the antibody to protect the host . These constant regions are distinct from the variable regions that directly mediate antibody binding . The results of the study have implications for i ) understanding how antibodies function in protection against anthrax and possibly other diseases , ii ) understanding how the host responds to a key bacterial virulence factor , iii ) selection of antibodies that might be used to treat anthrax , and iv ) design of vaccines to protect against anthrax . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"adaptive",
"immunity",
"immunity",
"gram",
"positive",
"immunity",
"to",
"infections",
"immunology",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"microbial",
"pathogens",
"bacterial",
"pathogens",
"immunotherapy",
"immunoglobulins"
] | 2013 | IgG Subclass and Heavy Chain Domains Contribute to Binding and Protection by mAbs to the Poly γ-D-glutamic Acid Capsular Antigen of Bacillus anthracis |
Dengue is an emerging infectious disease that has become the most important arboviral infection worldwide . There are four serotypes of dengue virus , DENV-1 , DENV-2 , DENV-3 , and DENV-4 , each capable of causing the full spectrum of disease . rDEN1Δ30 is a live attenuated investigational vaccine for the prevention of DENV-1 illness and is also a component of an investigational tetravalent DENV vaccine currently in Phase I evaluation . A single subcutaneous dose of rDEN1Δ30 was previously shown to be safe and immunogenic in healthy adults . In the current randomized placebo-controlled trial , 60 healthy flavivirus-naive adults were randomized to receive 2 doses of rDEN1Δ30 ( N = 50 ) or placebo ( N = 10 ) , either on study days 0 and 120 ( cohort 1 ) or 0 and 180 ( cohort 2 ) . We sought to evaluate the safety and immunogenicity of this candidate vaccine in 50 additional vaccinees and to test whether the humoral immune response could be boosted by a second dose administered 4 or 6 months after the first dose . The first dose of vaccine was well tolerated , infected 47/50 vaccinees and induced seroconversion in 46/50 vaccinees . Irrespective of dosing interval , the second dose of vaccine was also well tolerated but did not induce any detectable viremia or ≥4-fold rise in serum neutralizing antibody titer . Only five subjects had an anamnestic antibody response detectable by ELISA following a second dose of vaccine , demonstrating that the vaccine induced sterilizing humoral immunity in most vaccinees for at least six months following primary vaccination . The promising safety and immunogenicity profile of this vaccine confirms its suitability for inclusion in a tetravalent dengue vaccine .
Dengue fever has emerged as the world's most important mosquito-borne viral disease . Four antigenically distinct serotypes of dengue virus ( DENV-1 , DENV-2 , DENV-3 , and DENV-4 ) are transmitted by Aedes aegypti and Ae . albopictus mosquitoes , and the geographical spread of both mosquito vectors and the four viruses has led to an increased number of countries experiencing epidemic dengue fever [1] . In dengue endemic countries , hyperendemicity in many urban centers as well as focal outbreaks in rural areas are a major public health concern [2] , [3] . Up to 3 billion people are at risk of infection in tropical and sub-tropical countries , and an estimated 50 – 100 million people develop dengue illness annually [1] , [4] , [5] . Although most dengue-infected individuals are treated on an outpatient basis , hundreds of thousands of hospitalizations and approximately 20 , 000 deaths per year are attributable to dengue . In many countries , children bear much of this disease burden [6] . The spectrum of DENV disease ranges from subclinical disease or undifferentiated febrile illness to classic dengue fever ( DF ) and to life-threatening dengue hemorrhagic fever/dengue shock syndrome ( DHF/DSS ) . All four DENV serotypes are capable of causing the full spectrum of disease , although differences in virulence might exist [4] , [7] , [8] . Long-term homotypic immunity is induced by a single infection with DENV [9]; however , heterotypic protection is less durable [10] , and pre-existing immunity to one DENV serotype has been identified as a risk factor for more severe disease upon secondary , heterotypic infection [8] , [11] , [12] , [13] . Therefore , a DENV vaccine needs to induce long-lived protective immunity against all four DENV serotypes . In order to do so , more than one dose of a live attenuated tetravalent vaccine may be needed [14] , [15] , [16] , [17] , [18] , [19] , [20] . For example , two doses of the Mahidol/US Army/Pasteur PDK passaged tetravalent vaccine , given 180 days apart , were needed to achieve greater than 75% seroconversion against 3 or more serotypes [15] . Similarly , three dose of ChimeriVax are needed to induce a trivalent or tetravalent response in >80% of flavivirus-naive children and adults [14] . In preparation for tetravalent DENV vaccine studies , others have conducted two dose studies to evaluate the safety , infectivity and immunogenicity of a second dose of monovalent DENV vaccines [17] . Sun et al . reported 50% plaque reduction neutralization test ( PRNT50 ) seroconversion rates between 46% and 100% following the first dose of their live attenuated monovalent DENV vaccines and limited seroconversion following a second dose given one or three months later [17] . The Laboratory of Infectious Diseases ( LID ) at the NIH has developed monovalent vaccines against all four DENV serotypes using a variety of strategies [21] , [22] , [23] , [24] , [25] , [26] . The NIH monovalent DENV-1-4 vaccines were found to be safe , highly attenuated , and immunogenic in healthy adult volunteers when given as a single dose of 103 plaque-forming units ( PFU ) [22] , [23] , [24] . Dengue-like illness was not observed in the more than 500 vaccinees evaluated thus far , and vaccine-related serious adverse events ( SAEs ) did not occur . The most common reactogenicity events were an asymptomatic faint maculopapular rash over the trunk and proximal extremities and a transient neutropenia that lasted 2–3 days without associated clinical signs . Specifically , the rDEN1Δ30 vaccine , when evaluated as a single dose of 103 PFU administered subcutaneously , caused low-level viremia ( 10 PFU/mL ) in 9/20 vaccinees , starting around day 10 and lasting for approximately three days [22] . None of the volunteers developed a dengue-like illness , nor did any report vaccine-related reactogenicity that interfered with daily activities . Compared to placebo recipients , a significant increase in the occurrence of any solicited clinical sign or laboratory finding other than asymptomatic rash and neutropenia , was not observed in vaccinees . The vaccine induced a four-fold or greater rise in PRNT60 antibody titers in 95% of the vaccinees [22] . Here we present the results of a two-dose study of the monovalent rDEN1Δ30 vaccine in healthy adult volunteers . This study was designed to evaluate the safety and immunogenicity of a new lot of rDEN1Δ30 in 50 vaccinees in preparation for tetravalent vaccine trials and to determine a suitable interval for administration of a second dose of vaccine . The 4-month dosing interval was chosen because evaluation of tetravalent dengue vaccine formulations containing rDEN1Δ30 demonstrated that neutralizing antibody responses in rhesus macaques could be boosted with a second dose of vaccine given at 4 months ( but not at 1 month ) post dose one [10] , [27] . The 6-month interval was chosen because clinical trials with ChimeriVax indicated that boosting at an interval longer than 4 months was preferable [14] . Two cohorts of 30 subjects were recruited to receive two doses of the rDEN1Δ30 vaccine or placebo with the second dose given either on Day 120 or on Day 180 after dose 1 . While the first dose of DEN1Δ30 was safe , infectious , and immunogenic in 92% of the vaccinees , the second dose of the vaccine , given at either interval , did not cause detectable viremia and did not boost neutralizing antibody titers , indicating that a single dose of monovalent DEN1Δ30 vaccine induced sterilizing humoral immunity against a second dose of this vaccine for at least 6 months .
The study was performed under a NIAID-held investigational new drug application ( BB-IND #11677 ) reviewed by the US Food and Drug Administration . The clinical protocol , consent form , and Investigators' Brochure were developed by Center for Immunization Research and National Institute of Allergy and Infectious Diseases ( NIAID ) investigators and were reviewed and approved by the NIAID Regulatory Compliance and Human Subjects Protection Branch ( RCHSPB ) , the NIAID Data Safety Monitoring Board ( DSMB ) , the Western Institutional Review Board ( WIRB ) , and the Johns Hopkins University Institutional Biosafety Committee . Written informed consent was obtained from each volunteer in accordance with the Code of Federal Regulations ( 21 CFR 50 ) and International Conference on Harmonisation guidelines for Good Clinical Practice ( ICH E6 ) . The DSMB of the NIAID Division of Clinical Research reviewed all safety data at 6-month intervals . This single-institution parallel group Phase 1 trial was conducted as a randomized double-blind placebo-controlled study at the Center for Immunization Research ( CIR ) at the Johns Hopkins Bloomberg School of Public Health ( JHSPH ) . Study subjects were enrolled between May 2007 and July 2008 under study protocol CIR-229 , registered at ClinicalTrials . gov as Study NCT00473135 . The study was designed to evaluate the safety of two doses of the rDEN1Δ30 vaccine , given at day 0 and again at day 120 or day 180 , and to evaluate the virologic and serologic response to the vaccine after dose 1 and dose 2 . Viremia was characterized by mean peak titer , day of onset , and duration . The serologic response was characterized by PRNT60 at study day 28 and 42 following both first and second vaccination . Two cohorts were evaluated; cohort 1 subjects received vaccine or placebo at study day 0 and again at study day 120 , subjects in cohort 2 received vaccine or placebo at study day 0 and again at study day 180 . Within each cohort , subjects were randomized such that 25 would receive vaccine and 5 would receive placebo . A sample size of 25 vaccinees and 5 placebo recipients in each cohort was chosen based on our previous Phase I trial of this candidate DENV-1 vaccine [22] and to expand the safety evaluation of this vaccine . Subjects were randomized using a random number generator , by a member of the CIR who was not involved in the clinical or laboratory evaluation of the subjects . Subjects who had been enrolled could be replaced until study day 42 following first vaccination if they did not complete study day 28 or study day 42 , or if their specimens were not usable in the immunological assessment at study day 42 due to protocol-defined reasons . If a volunteer was replaced , the safety data and viremia data for that subject , up to the point of replacement , were included in the analysis . Subjects retained the same treatment assignment as the subject they replaced and the treatment assignment remained blinded . Healthy adult male and non-pregnant female volunteers were recruited from the Baltimore , Maryland and Washington DC metropolitan areas . Written informed consent was obtained from each volunteer in accordance with the Code of Federal Regulations ( 21 CFR 50 ) and International Conference on Harmonisation guidelines for Good Clinical Practice ( ICH E6 ) . Healthy volunteers 18 to 50 years of age were enrolled if they met the following eligibility criteria: normal findings during physical examination; negative for antibodies to all DENV serotypes , yellow fever virus , West Nile virus , St . Louis encephalitis virus; negative for hepatitis B and C; negative for HIV; normal values for complete blood count ( CBC ) with differential , aspartate aminotransferase ( AST ) , alanine aminotransferase ( ALT ) , creatinine , coagulation studies , and urinalysis . Additional safety-related exclusion criteria were also applied . Female volunteers were required to have a negative result on a urine pregnancy test at screening and on each vaccination day and were required to use a reliable method of contraception . On study day 0 , volunteers reported to the CIR outpatient clinic and were randomly assigned to receive either vaccine or placebo ( vaccine diluent ) given as a 0 . 5 mL subcutaneous injection . The appearance of the vaccine and placebo was identical . All study staff involved in the clinical and laboratory assessment of the subjects remained blinded to the treatment assignment until the study was unblinded at study day 42 following the second vaccination . Subjects were monitored for immediate adverse reactions for at least 30 minutes after vaccination . Each subject was given a digital thermometer and a diary card to record his/her oral temperature three times per day for 16 days . Clinical assessments were performed every other day through study day 16 and again on study days 21 , 28 and 42 post-vaccination . During clinical visits , a medical provider performed a focused physical examination to evaluate for local reactogenicity ( pain , erythema and swelling at the injection site ) and systemic reactogenicity ( lymphadenopathy , photophobia , rash , and hepatomegaly ) and questioned subjects specifically about dengue-related symptoms ( fever , headache , retro-orbital pain , nausea , photophobia , fatigue , myalgia , arthralgia , and rash ) . These signs and symptoms as well as the specific clinical laboratory studies described below were classified as solicited reactogenicity . In addition , subjects were questioned at each visit about other intercurrent illnesses . For each dose , blood was drawn at each assessment for detection of viremia through study day 16 post-vaccination and for antibody assay on study days 0 , 28 and 42 post-vaccination . Prior to vaccination , baseline CBC with differential , ALT and/or AST and coagulation studies were obtained . After each vaccination , a CBC with differential was obtained every other day through study day 16 . Coagulation studies were done every fourth day through study day 16 post-vaccination and again at day 28 post-vaccination . Serum for ALT testing was collected every other day from study day 4 through day 16 and again on study day 21 post-vaccination . Subjects returned to the clinic at study day 90 ( cohort 1 ) or study day 150 ( cohort 2 ) for repeat HIV , hepatitis B and hepatitis C testing and to assess for continued eligibility for second vaccination . Exclusion criteria for second vaccination included positive HIV-1 serology , active hepatitis B or hepatitis C infection , pregnancy , receipt of a live vaccine within 4 weeks prior to second vaccination or a killed vaccine within 2 weeks prior to second vaccination , use of immunosuppressive doses of corticosteroids ( excluding topical or nasal ) or immunosuppressive drugs within 30 days prior to second vaccination , receipt of an investigational agent in the 30 days prior to second vaccination , or history of severe allergic reaction or anaphylaxis associated with first vaccination . Clinical and laboratory assessments following the second vaccination were identical to the first vaccination throughout the 42-day follow-up period . All adverse events were graded for intensity and relationship to vaccine . Fever was determined by oral temperature recorded on two consecutive measurements 1 hour or more apart and was defined as grade 1 ( 38 . 0°C–38 . 6°C ) , grade 2 ( 38 . 7°C–39 . 1°C ) , or grade 3 ( >39 . 1°C ) . Adverse events were graded as mild ( no effect on daily activity ) , moderate ( required an intervention or interfered with daily activity ) , or severe ( prevented daily activity ) . Abnormal hematology , coagulation and serum chemistry findings were also graded as mild , moderate , or severe , using standardized toxicity tables . A dengue-like syndrome was defined as infection associated with fever and ≥ 2 of the following symptoms: Grade 2 headache lasting 12 hours , Grade 2 photophobia lasting ≥12 hours , Grade 2 generalized myalgia lasting ≥12 hours . Grade 2 retro-orbital pain lasting ≥12 hours , or sustained or intermittent epistaxis lasting ≥24 hours . Dengue infection was defined as recovery of vaccine virus from the blood and/or seroconversion to DENV-1 as measured by 60% plaque-reduction neutralization titer ( PRNT60 ) . All adverse events and abnormal clinical findings were collected for the duration of the study and were followed to resolution . Serious adverse events were defined in accordance with 21 CFR312 , 32 . rDEN1Δ30 is a recombinant , live attenuated virus derived from the DENV-1 Western Pacific ( WP ) wild-type virus by deletion of 30 nucleotides in the 3′ UTR [28] . The vaccine virus in the current clinical lot of this investigational vaccine ( Lot DEN1#104A , Charles River Laboratories , Malvern PA ) is identical at the amino acid level to that evaluated in an earlier phase I study ( Lot DEN1#1 ) [22] . A cDNA clone was generated to match the sequence of the previously manufactured rDEN1Δ30 ( LotDEN1#1 ) and vaccine virus was derived and manufactured in qualified Vero cells . The genome sequence of the vaccine virus in Lot DEN1#104A differed from that of Lot DEN1#1 at seven nucleotide positions , however all substitutions were translationally silent . Vaccine virus was stored at -80±15°C until use . Just prior to vaccination , vaccine virus was removed from the freezer and diluted to a concentration of 3 . 3 log10 PFU/mL with vaccine diluent ( 1X Leibovitz L-15 medium lacking phenol red ) . L-15 was also used as the placebo . The 1X Leibovitz L-15 medium was prepared from a qualified lot of 2X Leibovitz L-15 ( Lonza , Walkersville , MD ) mixed 1∶1 with sterile water for injection . Diluted vaccine was administered as a 0 . 5 mL subcutaneous injection within four hours of removal from the freezer . The virus titer of the diluted and undiluted vaccine was determined to confirm the potency of the vaccine . Virus titers were determined by plaque assay after inoculation of undiluted or serial 10-fold dilutions of serum onto Vero cell monolayer cultures , as described previously [21] . The lower limit of virus detection in this assay is 0 . 5 PFU/mL . Antibody responses to DENV-1 virus were determined by PRNT60 as described previously [21] , using DENV-1 ( WP ) as the target virus in the assay . For comparison of the PRNT60 values induced by the different vaccine lots , the day 42 samples collected during the previous clinical trial of rDEN1Δ30 were assayed in parallel with those collected during this trial . Seroconversion was defined as a ≥4-fold rise in serum neutralizing antibody to wild-type DENV-1 by study day 42 following each vaccination . Following first vaccination , this corresponded to a PRNT60 of ≥1∶20 as all subjects were flavivirus naïve ( PRNT60<1∶5 ) . Following second vaccination , the titer obtained on day 42 post-second vaccination ( day 162 or day 222 ) was compared with the titer from Study Day 120 or Study Day 180 . An IgG ELISA against whole virus was used to detect the presence of non-neutralizing antibody against DENV-1 ( WP ) virus . The present study is largely descriptive . Comparisons of the incidence of solicited adverse events between groups , neutrophil count parameters ( baseline , absolute and relative decline , day of nadir ) between groups , and comparisons of the ages of vaccinees and placebo recipients were performed using a 2-tailed Fisher Exact Test . Comparisons of mean peak titer , onset of viremia , duration of viremia , were performed using Tukey-Kramer HSD test . Mean values ± standard error ( SE ) are indicated . Statistical analysis was performed using JMP software ( version 5 . 0 . 1 . 2; SAS Institute ) .
A total of 145 subjects were screened and 62 subjects were enrolled in the trial ( Figure 1 ) . All subjects who received vaccine ( or placebo ) were included in the safety assessment even if they did not complete day 42 and were replaced . Two subjects were replaced; one had received vaccine and the other had received placebo . Fifty-one subjects received the first dose of vaccine ( 26 in cohort 1 and 25 in cohort 2 ) and 11 subjects received the first dose of placebo ( 6 in cohort 1 and 5 in cohort 2 ) . Forty-six subjects ( 23 in each cohort ) received a second vaccination and 10 subjects received placebo ( 5 in each cohort ) at second vaccination; all were included in the safety assessment . One vaccinee in cohort 1 withdrew prior to study day 222 and was not included in the immunological assessment ( Figure 1 ) . A statistically significant difference in age was neither observed between vaccinees ( 32±1 . 4 years ) and placebo recipients ( 36±2 . 5 years ) , nor between the two cohorts ( data not shown ) . Twenty-six subjects ( 42% ) were female and 36 subjects ( 58% ) were male . This difference was not significant . Thirty-one of 51 vaccinees ( 61% ) self-reported as Black , 17 ( 33% ) as White , and the remaining 6% were comprised of multi-racial , Pacific Islander , or unknown racial identity . In comparison , 9/11 placebo recipients ( 82% ) self-reported as Black and the remaining 18% self-reported as White . The type and frequency of adverse events reported by subjects who received this new lot of rDEN1Δ30 ( Lot DEN1#104A ) were compared with those reported by subjects who received the Lot DEN1#1 to establish bioequivalence and to expand our safety database . Two SAEs occurred following dose 1 in a single vaccine-recipient . This subject fractured his ankle and , while hospitalized for surgical repair of the ankle fracture ( SAE #1 ) , had surgical repair of a pre-existing spinal stenosis ( SAE #2 ) . Both SAEs were judged to be unrelated to vaccination . The vaccine was well tolerated by all vaccinated subjects . None of the subjects developed a dengue-like illness . One vaccinee developed mild injection site erythema starting 4 days post-vaccination and lasting for 4 days . None of the other subjects developed injection site erythema , induration , or tenderness during the intensive follow-up period ( 16 days post-vaccination ) . The frequency and severity of solicited reactogenicity events following vaccination with rDEN1Δ30 , Lot DEN1#104A , was comparable to that reported following vaccination with the previous lot , DEN1#1[22] , and selected reactogenicity events are shown in Table 1 . Overall , 42/51 subjects ( 82% ) who received rDEN1Δ30 , Lot DEN1#104A developed at least one solicited clinical or laboratory adverse event , compared with 5/11 placebo recipients ( 45% ) , p = 0 . 02 . The most commonly observed adverse events were rash , headache , and neutropenia . Since the frequencies of adverse events were similar for cohort 1 and cohort 2 ( shown separately in Table 1 ) , the pooled data is discussed here . The only solicited adverse event that showed a trend towards more common occurrence in vaccine recipients was rash ( 27% of vaccinees , 0% placebo recipients , p = 0 . 06 ) . A mild asymptomatic maculopapular rash , similar in character to what we have described previously , was detected in 14 vaccinees ( Table 1 ) [22] , [23] , [24] . The mean onset of rash was study day 12 . 8±0 . 4 and the mean duration was 11±2 . 2 days . Because the rash was not noticed by the majority of the volunteers , the reported duration of rash may be longer than its actual duration because subjects were not seen between study days 21 and 28 and the rash was not determined to be resolved until the subject was evaluated by a clinician at study day 28 . There were 39 episodes of headache reported by 21 vaccinees throughout the 42-day post-vaccination period . The mean duration of headache was 2 . 2±0 . 6 days with a mean day of onset of 10 . 1±1 . 2 . Sixty-nine percent of headache episodes were of mild severity and 31% were of moderate severity . There were four episodes of headache reported by 2 placebo recipients; this difference was not statistically significant when compared with vaccinees . Seven vaccinees ( 14% ) reported mild ( 6 subjects ) or moderate ( 1 subject ) retro-orbital pain following dose 1 , six of them between day 13 and 16 . None of the subjects developed fever during the 16-day follow-up . Four vaccinees reported myalgia that was determined to be possibly , probably , or definitely related to vaccine . Three vaccinees ( 6% ) reported a mild myalgia lasting 3–5 days; one vaccinee reported moderate myalgia lasting 2 days . One placebo recipient ( 9% ) reported myalgia lasting 3 days . Twenty-three of 51 vaccinees ( 45% ) in the current study ( Lot DEN1#104A ) developed transient neutropenia , similar to the frequency observed in our previous study ( Table 1 ) . Neutropenia in vaccinees was graded as mild ( range 1 , 000–1 , 500/mm3 ) in 15 subjects , moderate ( range 750–999/mm3 ) in 3 subjects , and severe ( range 500–749/mm3 ) in 5 subjects . The onset of neutropenia ranged from Study Day 4 to Study Day 16 , [mean day of onset = 11 . 8±0 . 7 ( SE ) ] and resolved in all subjects . The mean duration of neutropenia in vaccinees was 2 . 4 days ±0 . 3 days . Four of 5 subjects with severe neutropenia had an ANC<750/mm3 on a single day and one subject had an ANC<750/mm3 on days 12 and 14 . Two of 11 placebo recipients ( 18% ) developed neutropenia , both episodes were mild . In the current study , vaccinees who became neutropenic had a statistically significant lower mean baseline ANC ( 2 , 663/mm3 ) than vaccinees who did not become neutropenic ( 3 , 855/mm3 ) , α = 0 . 01 ( Table 2 ) . Individuals who had a baseline ANC of ≤3 , 000/mm3 were more likely to become neutropenic than individuals with baseline ANCs > 3 , 000/mm3 ( p<0 . 0001 ) . Sixteen of the 23 vaccinees ( 70% ) who became neutropenic had a baseline ANC of ≤3 , 000/mm3 compared with 4/28 ( 14% ) vaccinees who did not become neutropenic . The mean absolute decline in ANC was not significantly different between those vaccinees who became neutropenic and those who remained non-neutropenic . However , the mean decline in ANC among all vaccinees was significantly greater than in placebo recipients ( Table 2 ) . One vaccinee developed a moderate elevation in serum ALT activity that was possibly related to vaccine . The ALT level peaked on day 15 at 123 IU/L and was not associated with nausea , vomiting , abdominal pain , or hepatomegaly . Although at screening this subject denied taking any prescription drugs , she later indicated that she had been taking anti-depressive medication for several weeks , which may have contributed to the elevation in serum ALT level . Following Dose 2 , none of the subjects developed dengue-like illness , fever , retro-orbital pain , or rash ( Table 1 ) . A statistically significant difference in the number of volunteers reporting headache between vaccinees and placebo recipients was not observed . The incidence of neutropenia following second vaccination was similar in vaccine and placebo recipients; four vaccine recipients ( 9% ) developed mild neutropenia and one placebo recipient ( 10% ) developed moderate neutropenia . Two vaccinees developed a mild transient elevation in serum ALT level ( peak of 82 IU/L on day 4 and peak of 83 IU/L on day 7 post dose 2 , respectively ) . The elevated ALT level in both subjects resolved within 4 days . Neither of these subjects met the definition of infection following second dose nor did they have any increase in ELISA antibody titer following the second dose . Thirty-four of 51 vaccinees ( 67% ) who received rDEN1Δ30 Lot DEN1#104A had detectable viremia following the first vaccination ( Table 3 ) . The percent of volunteers with viremia , the mean peak virus titer , and the mean number of viremic days following dose 1 in the current study were comparable to that observed in the previous study rDEN1Δ30 Lot DEN1#1 [22] . Following dose 1 , 94% vaccinees were infected with rDEN1Δ30 Lot DEN1#104A ( 88% in cohort 1 and 100% in cohort 2 ) and 95% with rDEN1Δ30 LotDEN1#1 . Amongst the 48 subjects infected with rDEN1Δ30 Lot DEN1#104A , one was viremic on day 12 but did not meet criteria for seroconversion , resulting in an overall seroconversion rate of 92% to wild-type DENV-1 virus by study day 42 ( 84% in cohort 1 and 100% in cohort 2 , Table 4 ) . With regard to immunogenicity , PRNT60 titers against wild-type DENV-1 induced by the two lots of vaccine were comparable ( less than a 2-fold difference ) ( Table 4 ) . Twenty-three vaccinees per cohort received a second dose of vaccine ( Figure 1 ) . Following a second dose of vaccine at either interval , viremia was not detected in any subject ( Table 1 ) , and none of the subjects had a 4-fold or greater rise in serum neutralizing antibody titer against DENV-1 , indicating that none of the subjects met the protocol definition of infection following a second dose of rDEN1 ? 30 ( Table 5 ) . Geometric mean PRNT60 titers against wild-type DENV-1 following the second dose of vaccine were similar in cohorts 1 and 2 ( Table 5 ) . Three of the four subjects who did not seroconvert after the first dose of vaccine received a second dose of vaccine four months later; none of them developed a ≥4-fold rise in serum neutralizing antibody titer or in ELISA antibody titer . Of the subjects who seroconverted to DENV-1 after the first vaccination , one who received a second dose of vaccine four months after the first dose developed a ≥4-fold rise in ELISA antibody titer following the second dose , as did four subjects who received a second dose six months after the first dose ( Table 5 ) . These findings suggest that a single dose of rDEN1Δ30 was highly infectious and was able to provide sterilizing humoral immunity to infection with the vaccine virus for at least 6 months in 83% of vaccinees .
The development of a safe and effective tetravalent dengue vaccine has been a goal for decades , yet a licensed vaccine is still not available . At a minimum , such a vaccine must have an acceptable safety profile and must induce long-lived protective immunity against all four serotypes of wild type dengue virus . Ideally this would be accomplished with a single dose of vaccine . However , tetravalent vaccines currently in clinical development require two or three doses [14] , [18] . A number of dengue viruses have been attenuated either by serial passage in tissue culture or by the introduction of attenuating mutations and/or chimerization of dengue viruses using recombinant DNA technology [26] , [29] , [30] , [31] , [32] . In addition , a tetravalent vaccine based on chimerization of DENV-1-4 with the yellow fever vaccine virus is currently in Phase II/III clinical development [14] , [20] . Other investigational vaccines have been abandoned because they were either under or over-attenuated [15] , [16] , [17] , [33] , [34] . In addition , there are concerns that viral interference may affect the ability of a live attenuated tetravalent dengue vaccine to induce a balanced immune response to all four serotypes . For these reasons , we have attempted to carefully examine the safety profile and humoral immune response to each monovalent DENV vaccine candidate virus prior to formulating a tetravalent vaccine . Clinical examination and laboratory studies performed every other day for the first 16 days post vaccination have helped to fully characterize the reactogenicity and kinetics of replication of these viruses when given as monovalent vaccines . The ability of live attenuated dengue vaccines to induce a satisfactory antibody response without clinically significant reactogenicity has been a hurdle to DENV vaccine development . There are few published studies of monovalent DENV-1 vaccines to which we can compare the safety and immunogenicity of rDEN1Δ30 in healthy adult subjects . Three such studies describe the DENV-1 vaccine candidates 16007 , 45AZ5 PDK20 , and 45AZ5 PDK27 [17] , [35] , [36] . These candidates were attenuated by serial passage in tissue culture , and the parent virus of 45AZ5 PDK20 and 45AZ5 PDK27 underwent chemical mutagenesis prior to passage in tissue culture . Five subjects received DENV-1 16007 as a monovalent vaccine and 12 subjects received 45AZ5 PDK20 . Vaccine candidate 16007 was well tolerated: none of the subjects developed an oral temperature >38°C , one subject developed a rash and 2 subjects developed elevated liver function tests [35] . However , only 60% of vaccinees seroconverted to the vaccine as defined as a PRNT50 ≥1∶10 . In contrast , 100% of vaccinees seroconverted to DEN1 after receipt of the monovalent DENV-1 vaccine 45AZ5 PDK20; however , 5/12 subjects developed a temperature >38°C and 3/12 developed dengue-like illness [17] . Additionally , 20% to 35% of subjects who received a tetravalent formulation containing 45AZ5 PDK20 developed a temperature of >38°C [18] . The further passaged DENV-1 candidate vaccine 45AZ5 PDK27 was evaluated as a monovalent DENV-1 vaccine in 10 subjects and was found to be more attenuated than 45AZ5 PDK20 [36] . While the reactogenicity profile of 45AZ5 PDK27 was acceptable , only 40% of vaccinees seroconverted to DENV-1 following a single dose of 45AZ5 PDK27 [36] . Although 45AZ5 PDK27 was less immunogenic than 45AZ5 PDK20 , it has replaced 45AZ5 PDK27 in the tetravalent formulations currently under evaluation , illustrating the trade-off that is sometimes necessary to balance reactogenicity and immunogenicity . Thus , achieving a satisfactory balance between attenuation and immunogenicity has been difficult to attain by passage and/or mutagenesis of DENV-1 . In the rDEN1Δ30 vaccine virus , the attenuating 30 nucleotide deletion is located in the 3′ UTR but the virus is otherwise wild-type . This might have contributed to its high level of infectivity and immunogenicity . The promising safety profile of the rDEN1Δ30 vaccine described in a previous single-dose study was affirmed in the current two-dose study [22] . None of the subjects developed fever , a dengue-like illness or any study-related SAE . All of the clinical solicited adverse events were mild or moderate in severity and were transient in nature . As we have reported for our previous dengue vaccine studies , neutropenia and rash were the most commonly reported adverse events in subjects who received rDEN1Δ30 . As was seen in the previous study of this vaccine candidate , the onset of neutropenia and rash generally followed the onset of viremia . The definition of neutropenia used in our studies is more inclusive than that utilized in many other dengue vaccine studies . We defined neutropenia as an ANC of ≤1500/mm3 and obtained blood counts every other day for the first 16 days after vaccination; other studies of leading dengue vaccine candidates have defined neutropenia as an ANC<1 , 000/mm3 with less frequent monitoring , e . g . only on day 15 , or on days 4 , 8 , 12 , and 16 post vaccination [14] , [17] , [18] , [20] . Had we used an ANC<1 , 000/mm3 to define neutropenia our studies , only 2/71 ( ANC measured on day 16 only ) or 5/71 ( ANC measured on days 4 , 8 , 12 , and 16 only ) vaccine recipients would have been described as neutropenic , a rate of neutropenia comparable to that reported in the studies indicated above . In addition , more than 60% of our study subjects were of African descent , a population known to maintain a lower mean baseline ANC [37] , [38] , [39] . In comparison , subjects of African descent made up only 5–33% of volunteers in the studies referenced above . Most importantly , the very short duration of neutropenia observed in recipients of rDEN1Δ30 was not associated with fever or other clinical signs or symptoms suggestive of neutropenia-related illness . In contrast to neutropenia associated with neoplasia or therapeutic interventions , we assume that neutrophil function during DENV infection or following vaccination with a live-attenuated DENV vaccine is not compromised and that the short duration of neutropenia does not compromise protective immune functions . Our monovalent DENV vaccines have been evaluated in approximately 500 subjects thus far and fever of unknown origin or any sign or symptom suggestive of opportunistic infections has not been observed . To our knowledge , neutropenia due to natural dengue illness has not been associated with secondary bacteremia . A review of cases of concomitant bacteremia in patients diagnosed with DHF found that leukopenia was not associated with an increased incidence of bacteremia compared with control patients [40] . The investigational vaccine rDEN1Δ30 appears to have a well-balanced safety and immunogenicity profile . In addition to the acceptable reactogenicity of rDEN1Δ30 described above , the vaccine was able to induce an overall seroconversion rate of 93% ( 65/70 subjects ) when administered as a single subcutaneous dose to flavivirus-naïve healthy adults . Importantly , the immunogenicity end-point used in both this and our previously reported study of rDEN1Δ30 was a 4-fold rise in serum neutralizing antibody ( i . e . , a PRNT60 titer of ≥1∶20 defined seroconversion ) and is more stringent compared with a PRNT50 titer of ≥1∶10 used by others in the studies described above . In addition to the high seroconversion rate that was achieved following a single dose of vaccine , rDEN1Δ30 induced sterilizing humoral immunity to infection with a second dose of vaccine administered 4 or 6 months later in the vast majority of vaccinees , a finding that could be indicative of long-term homotypic protection as has been described following natural DENV infection [10] , [41] . In the present study , the four subjects who did not develop detectable neutralizing antibody after the first dose of rDEN1Δ30 could not be infected by a second dose of vaccine given 4 months later . Following dose 1 , one of the 4 subjects had laboratory findings consistent with infection by the vaccine virus , and another subject had findings that may be suggestive of infection by the vaccine virus . The former had detectable viremia on study day 12 and a mild neutropenia on study day 16 , and the latter subject was mildly neutropenic on study day 4 following first vaccination . The other two subjects did not have clinical or laboratory findings suggestive of infection . Of these four subjects , three received a second dose of vaccine but did not develop a detectable neutralizing antibody response following dose 2 . It is unclear whether these individuals were vaccine non-responders , completely resistant to infection , or whether their innate immune system was able to abort infection prior to the induction of an antibody response . Loss of vaccine potency was excluded as a cause for the absence of infection following dose 2 . This study has important implications for the development of a live attenuated tetravalent dengue vaccine . First , the expanded safety evaluation of the monovalent rDEN1Δ30 vaccine confirmed it to be well tolerated without vaccine-related fever or dengue-like illness . Second , the vaccine was able to induce seroconversion to DENV-1 in 93% of recipients when given as a single subcutaneous dose of 103 PFU . The 103 PFU dose required for the induction of seroconversion to DENV-1 and the ability to grow this virus to high titer ( >107 PFU/mL ) , make this a very economical vaccine component to produce . Third , the vaccine virus behaved immunologically much like wild-type virus in that sterilizing homotypic immunity was induced after a single dose . It is unknown whether this will make boosting of the humoral response to DENV-1 more difficult should multiple doses of the tetravalent vaccine be required . It is also not known whether the observed protection against a second dose of vaccine virus will translate into protection against infection/disease following exposure to wild type DENV-1 virus after vaccination . Future studies with tetravalent vaccine will need to evaluate whether boosting is necessary and will need to be designed to determine a suitable interval for boosting . Lastly , because the replication kinetics of this virus have been very well characterized as a monovalent vaccine , any effect of viral interference on the replication kinetics of the vaccine when administered as part of a tetravalent formulation should be discernable . In summary , rDEN1Δ30 is an excellent candidate for inclusion in live tetravalent dengue vaccine formulations , and clinical evaluation of tetravalent dengue vaccines with this rDEN1Δ30 as DENV-1 component have been initiated . Our study has at least two potential limitations regarding the infectivity , safety , and immunogenicity of rDEN1Δ30 when administered as a component of a tetravalent formulation . First , sterilizing humoral immunity to a second dose of monovalent rDEN1Δ30 vaccine at 6 months might not translate into sterilizing humoral immunity six months after vaccination with a tetravalent vaccine . It is not known whether or not the high seroconversion rates achieved following a single dose of the rDEN1Δ30 vaccine , when given as a monovalent virus , will be achievable when the candidate vaccine is included in a tetravalent formulation . Secondly , the results of our studies in flavivirus-naive healthy adults do not necessarily mirror how this investigational vaccine will behave in flavivirus-naive and partially immune children , one of the target groups for vaccination in several hyperendemic areas . Tetravalent vaccine studies will need to carefully proceed from adults into children and from flavivirus-naive into partially immune individuals to evaluate the safety and immunogenicity of this vaccine in these vulnerable populations . | Globally , dengue fever has become the most common clinically significant mosquito-transmitted viral illness . Dengue viruses exist as four serotypes , and increasingly several serotypes co-circulate in the same region . Infection with one serotype increases the risk of severe illness following infection with a second serotype . Therefore , any dengue virus vaccine needs to protect against all four serotypes . We and others are working to develop a live-attenuated tetravalent dengue vaccine that contains four monovalent vaccine viruses . Since two or more doses of such a vaccine are thought to be necessary for induction of long-lasting protective immunity , a feasible dose interval needs to be determined . Here , boosting with a second dose of a monovalent dengue type 1 ( DENV-1 ) vaccine at four months or six months was compared in flavivirus-naïve healthy adult subjects with regard to safety , infectivity , and immunogenicity . We found that both doses of the vaccine were safe and well tolerated . While the first dose infected 92% of recipients , the second dose was neither infectious nor immunogenic , irrespective of the dose interval . These findings indicate that in most subjects , a single dose of this monovalent vaccine confers sterilizing humoral immunity against a second dose for at least six months . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"public",
"health",
"and",
"epidemiology",
"clinical",
"research",
"design",
"immunology",
"dengue",
"phase",
"i",
"global",
"health",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"immunizations",
"infectious",
"diseases",
"dengue... | 2011 | A Single Dose of the DENV-1 Candidate Vaccine rDEN1Δ30 Is Strongly Immunogenic and Induces Resistance to a Second Dose in a Randomized Trial |
Tubulointerstitial kidney disease is an important cause of progressive renal failure whose aetiology is incompletely understood . We analysed a large pedigree with maternally inherited tubulointerstitial kidney disease and identified a homoplasmic substitution in the control region of the mitochondrial genome ( m . 547A>T ) . While mutations in mtDNA coding sequence are a well recognised cause of disease affecting multiple organs , mutations in the control region have never been shown to cause disease . Strikingly , our patients did not have classical features of mitochondrial disease . Patient fibroblasts showed reduced levels of mitochondrial tRNAPhe , tRNALeu1 and reduced mitochondrial protein translation and respiration . Mitochondrial transfer demonstrated mitochondrial transmission of the defect and in vitro assays showed reduced activity of the heavy strand promoter . We also identified further kindreds with the same phenotype carrying a homoplasmic mutation in mitochondrial tRNAPhe ( m . 616T>C ) . Thus mutations in mitochondrial DNA can cause maternally inherited renal disease , likely mediated through reduced function of mitochondrial tRNAPhe .
End stage renal disease ( ESRD ) is common , affecting over 660 , 000 individuals in the United States in 2013 and is increasing in incidence[1] . Renal replacement therapy is life-saving for these patients , but is associated with increased morbidity and mortality , and high cost . In a proportion of cases ( currently less than 1 in 10 ) a genetic cause is identified on the basis of a clear family history and/or specific clinical features . Genetic factors are also presumed to make a major contribution to the development of ESRD in patients without a clear-cut family history of renal disease . Chronic tubulointerstitial nephritis is a significant , but poorly understood , cause of progressive renal disease characterised by bland urinary sediment and histological evidence of interstitial fibrosis and tubular atrophy . To date , mutations in UMOD , HNF1B , REN and MUC1 have been shown to cause autosomal dominant tubulointerstitial kidney disease ( ADTKD ) [2] . While these account for a substantial proportion of inherited tubulointerstitial kidney disease , further mechanisms remain to be identified . Here we show that tubulointerstitial renal disease in a large family without other clinical features is caused by a substitution in mitochondrial DNA . We found a single base substitution in the promoter sequence directing heavy-strand transcription . Homoplasmic variants in the mtDNA control region are not traditionally considered to be pathogenic[3] , but we show that this substitution alters transcription , and reduced levels of mitochondrial tRNAPhe . This was accompanied by a defect in mitochondrial protein synthesis and a biochemical defect affecting multiple components of the respiratory chain . Examination of additional families with unexplained tubulointerstitial kidney disease led to the identification of two pedigrees with a substitution that directly alters the sequence and concentration of the mitochondrial tRNAPhe . , suggesting that defects in mitochondrial transcription or translation can cause tubulointerstitial kidney disease .
We investigated the inheritance of renal disease in a large multiply affected white English family ( Fig 1 ) . Affected individuals developed progressive renal impairment from early adulthood , with increased urinary N-acetyl-beta-D-glucosaminidase ( NAG ) excretion ( NAG to creatinine ratio 73 . 9–183 . 8 μmol/h/g , normal < 28 ) , without amino aciduria or renal Fanconi syndrome . Renal biopsy tissue was available from four individuals , showing changes consistent with tubulointerstitial kidney disease . Three individuals had received renal transplants with no evidence of recurrent renal disease up to 26 years following transplantation . Across the pedigree , penetrance of significant renal disease by the fourth decade of life was 100% , although the age of onset and speed of progression varied between individuals . Physical and intellectual development has been normal . Neurological examination in a national referral centre for mitochondrial disease was normal , and imaging of the CNS did not reveal any distinctive features . Multipoint linkage analysis failed to identify a shared chromosomal haplotype , using both autosomal dominant and X-linked models of inheritance ( S1 Fig ) . This was confirmed by conflicting homozygosity analysis of SNP data , which demonstrated no regions >1 . 0 cM that were identical by descent in affected individuals , even allowing for a 20% phenocopy rate[4] . Whole exome sequencing of two affected individuals failed to demonstrate any shared substitutions or variants that were predicted to be deleterious in genes associated with ADTKD . No cytosine insertions were identified in the MUC1 tandem repeat sequences by probe-extension assay . Given the absence of paternal transmission in the pedigree , we sequenced the mitochondrial genome in urinary epithelial cells from affected patients . Mitochondria contain a 16 . 5 kb circular genome , encoding two mitochondria-specific ribosomal RNAs , 22 tRNAs and 13 transmembrane proteins which form part of the core of the multi-protein oxidative phosphorylation ( OxPhos ) complexes I , III , IV and V [5] . Two promoters in the control region drive transcription of the polycistronic light and heavy strand transcripts[6] . We detected a total of 41 benign polymorphisms , one novel insertion located in a small hypervariable region downstream of the light strand promoter ( m . 16038+1 insC ) and a previously unreported substitution ( m . 547A>T ) located within the heavy strand promoter ( HSP ) of the mtDNA control region ( S1 Table ) [7] . We focussed our attention on the m . 547A>T substitution , which was present at homoplasmic levels in blood , saliva , urinary epithelial cells , and fibroblast samples in all affected individuals [8] , in 14 individual skeletal muscle fibres from one patient , and in blood samples taken over 15 years apart . No mtDNA deletions or duplications were detected on long-range PCR of DNA from blood , fibroblast or skeletal muscle . A diagnostic skeletal muscle biopsy[9] indicated a subclinical reduction in the enzymatic activity of complexes I and IV relative to citrate synthase ( S3 Table ) . Cultured dermal fibroblasts derived from four different patients showed a significantly reduced proliferation rate when galactose was substituted for glucose in the growth medium , suggesting a defect in the mitochondrial electron transfer chain ( S2 Fig ) . Mitochondrial DNA copy number was increased by ~50% ( S3 Fig ) , but citrate synthase activity was not significantly changed relative to control cells ( S4 Table ) . To further characterise the effects of the m . 547A>T substitution , we directly measured the oxygen consumption of cultured patient-derived dermal fibroblasts and of trans-mitochondrial cybrids made by fusion of patient platelets ( free of nuclear DNA but containing mitochondria ) and mitochondria-depleted 143B ρ0 cells[10] . Basal and maximal respiration rates were significantly reduced in fibroblasts and cybrids , confirming a mitochondrially inherited respiratory defect ( Fig 2 ) . To investigate the mechanism of impaired mitochondrial respiration we examined the expression of mitochondrial transcripts in patient-derived and control cells ( Fig 3 ) . We observed a significant reduction of heavy strand encoded RNR1 and MT-CO1 expression compared to light strand ND6 , consistent with reduced function of the HSP . We also performed quantitative Northern blot analysis to assess the expression of mitochondrial tRNA species in fibroblasts and cybrids ( Fig 3 ) . This demonstrated a highly reproducible reduction in the expression of heavy strand tRNAPhe and tRNALeu relative to light strand encoded tRNAGln . tRNAVal was also reduced but this was less marked . To gain insight into effects on mitochondrial protein homeostasis , we developed a quantitative proteomic approach to analyse mitochondrially-enriched fractions of SILAC-labelled fibroblasts , which we term MitoSILAC . This showed a reduction of mtDNA-encoded components in complex I , III , IV , and V in patients ( Fig 4 , annotated dots ) , but not in the exclusively nuclear-encoded complex II components . In addition , the majority of nuclear-encoded components of the respiratory chain complexes I , III and IV were also reduced , implying complex instability secondary to a deficiency of mtDNA-encoded subunits . By contrast , nuclear-encoded subunits of complex V ( ATP synthase ) were unchanged or up-regulated . The reduction of mitochondrially-encoded proteins was also pronounced in patient-derived cybrid lines as assessed by metabolic labelling ( S3 Fig ) . To investigate if m . 547A>T directly influences HSP activity , we employed a reconstituted in vitro transcription system , containing purified mitochondrial transcription factors ( Fig 5 ) [11] . We used a shortened linear mtDNA template that contained both the LSP and HSP . The relative expression of the two different length run-off transcripts allowed direct assessment of HSP transcription relative to the LSP signal ( Fig 5A ) . The m . 547A>T substitution decreased HSP transcription activity by 27 ± 7 . 2% compared to control , confirming a defect in HSP function ( Fig 5B and 5C ) . Analysis of mitochondrial DNA of 62 other pedigrees with previously unexplained , but potentially inherited , renal disease led to the identification of a different homoplasmic substitution , in the anticodon stem of mt tRNA Phe ( m . 616T>C ) in three families ( S2 Table ) . Two of these families ( Pedigrees II and III , Fig 6A ) originated from a study of 10 families with tubulointerstitial kidney disease , which only identified a genetic cause in 7 of them[12] . We also found the m . 616T>C substitution in a family with tubulointerstitial renal disease which was originally reported in 1982[13] . Further genealogic enquiries demonstrated that this family is actually connected to Family 8 in Ref[12] , and the combined kindred is shown in Pedigree II ( Fig 6A ) . Currently we have no genealogic evidence that pedigrees II and III are also connected . However , both are from Australia , and their mtDNA sequence is identical , implying a recent common maternal ancestor . Similar to the family with the m . 547A>T HSP substitution , none of the affected individuals in these kindreds had clear evidence of extra-renal disease . To explore the functional consequences of the tRNAPhe substitution we studied oxygen consumption of patient-derived cybrids which showed reduced basal and maximal respiration , indicating that the patient-derived mitochondrial DNA was responsible for impaired respiration ( Fig 6B ) . Analysis of patient fibroblasts revealed a reduction of the level of tRNAPhe to an extent similar to that which we observed in patients with the HSP substitution ( Fig 6C ) .
We consider that the m . 547A>T substitution in the HSP of mtDNA is a pathogenic substitution for the following reasons . First , it segregated with renal disease in 18 members of a maternal lineage and was not seen in 29 , 867 population controls[7] . Second , it was associated with decreased mitochondrial respiration in skin fibroblasts from multiple affected individuals and in cybrid cell lines . Third , it was associated with deficiency of multiple respiratory chain complexes in a skeletal muscle biopsy from an affected individual and four different fibroblast cell lines , supporting an abnormality of intra-mitochondrial protein synthesis . Fourth , it was associated with reduced MT-CO1 and mitochondrial tRNA transcripts , consistent with reduced function of the mtDNA HSP . Fifth , it resulted in decreased HSP transcription in vitro . Finally , a related mtDNA substitution was identified in other pedigrees with a similar clinical and biochemical phenotype . To the best of our knowledge m . 547A>T is the first substitution in the mitochondrial transcriptional control region that has been shown to cause disease . Interestingly , a m . 547A>G substitution has been recorded previously without any reported disease association . It is therefore likely that this alternative substitution does not have a functional effect , consistent with the structural impact of such a purine to purine transition being less severe than a purine to pyrimidine transversion [14] . The m . 616T>C substitution alters an evolutionarily conserved base , the only pseudouridine in mitochondrial tRNAPhe , and will disrupt the last base pairing before the anticodon loop . Reinforcing the pathogenic nature of this substitution , Zsurka et al . described an individual with m . 616T>C at near-homoplasmic levels who had intractable epilepsy and developmental delay[15] . Interestingly , she had chronic renal insufficiency when she died at the age of 17 , and a maternal relative had died from kidney failure and had epilepsy starting in childhood . Two of the m . 616T>C individuals in Pedigree II who are now deceased had epileptic seizures which are described in Burke et al . [13] . Although we have not detected neurological abnormalities in living individuals from this family or in Pedigree III , taken together with the report of Zsurka et al , this suggests that m . 616T>C may be associated with a reduced threshold for epileptic seizures . Previous reports of substitutions in mitochondrial tRNAPhe provide some further support for the notion that renal function is sensitive to perturbing this mitochondrial tRNA . Tzen et al . described two siblings with m . 608A>G substitutions ( the partner nucleotide of 616T in the anticodon stem ) , neurological features and a urinary concentrating defect[16] . D’Aco et al . reported a child with a heteroplasmic m . 586G>A substitution , persistent lactic acidosis , failure to thrive and progressive renal failure[17] . Mitochondrial tRNAPhe is one of two tRNAs that can form a structural component of the mitoribosome[18] which could exacerbate the translational defect . Although minor abnormalities in proximal tubular function are frequent in patients with other mitochondrial mutations , clinically overt renal disease is not usually observed[19] . Taken together with our study , these observations suggest that renal function is particularly susceptible to specific alterations in tRNAPhe . In all the individuals described in this report it is striking that despite the mitochondrial defect being homoplasmic and present in all tissues clinical disease was restricted to the kidney , although we acknowledge that other organs may be affected at a subclinical level . This situation is similar to the organ specific effect of mutations in mt tRNAIle , which primarily cause hypertrophic cardiomyopathy [20][21] and of mutations in NADH dehydrogenase causing Leber’s hereditary optic neuropathy [22] . Taken together , the contrasting effects of these mutations should offer a powerful route to understanding how mitochondrial defects cause disease in different tissues . From a clinical perspective , we propose the term Mitochondrially Inherited Tubulointerstitial Kidney Disease ( MITKD ) to complement the recently defined entity , Autosomal Dominant Tubulointerstitial Kidney Disease ( ADTKD ) [2] . MITKD may account for a significant proportion of unexplained familial tubulointerstitial kidney disease and we recommend that kindreds with compatible patterns of transmission should be offered analysis of mitochondrial DNA as a route to establishing a diagnosis and appropriate genetic counselling .
DNA was extracted from peripheral-blood leukocytes , urinary epithelial cells , and saliva . 1 μg of genomic DNA at 75 ng/μl was used for SNP genotyping . Individuals were genotyped using 300 , 000 single nucleotide polymorphism ( SNP ) markers with an average separation of 6 . 1 kb on the Illumina CytoSNP 12 chip ( Illumina , Cambridge , UK ) . Genotyping was performed at the automated , high-throughput system available at UCL Genomics . Linkage analysis was performed on pedigrees using the easyLINKAGE Plus v5 . 05 graphical user interface for two-/multi-point linkage analysis [23] , with analysis by GeneHunter[24] and SimWalk [25] . Pedigrees were confirmed using HaploPainter V . 1 . 043[26] and PedCheck[27] . For calculation of parametric LOD scores a ‘rare dominant’ model was utilised , with disease allele frequency set to 10–4 and disease penetrance vector set to 1 . 0 . Conflicting homozygosity was assessed in this genotype data , using a program written by Dr Daniel Gale and Dr Adam Levine at UCL . mtDNA was analysed by Sanger sequencing . Long-range PCR was performed using published primers[28] . Homoplasmy analysis was performed using labelled amplicons from the D4 primer pair[28] and tetra-primer amplification refractory mutation system PCR designed to detect the m . 547A>T substitution . A quadriceps muscle biopsy was taken from one affected individual , snap frozen , and analysed using enzyme histochemistry and respiratory chain complex analysis as described[29] . Individual skeletal muscle fibres were isolated by laser capture for quantitative pyrosequencing to determine whether there was any intracellular heteroplasmy[30] . Fibroblasts from four different patients and four healthy volunteers were obtained by skin biopsy and cultured in Dulbecco’s Modified Eagle’s medium , supplemented with pyruvate and uridine . Cybrids were created by fusion of 143B ρ0 cells and enucleated patient fibroblasts or platelets as described[10] . Cell lines were confirmed to contain the promoter variant by sequencing . All lines tested negative for mycoplasma . All donors consented to research use of their skin-derived primary cells . Volunteers were enrolled at UCL London , study reference 06/Q0406/151 . Oxygen consumption rate was determined with a Seahorse XF24 Analyser ( Seahorse Bioscience , MA ) and normalised for cell number . Respiratory profiles were generated by serial treatment with optimised concentrations of oligomycin ( 0 . 5–5 μg/ml ) , p-[trifluoromethoxy]-phenyl-hydrazone ( FCCP ) ( 1–7 . 5 μg/ml ) , and antimycin[31] ( 2–4 μg/ml ) , with fibroblasts receiving the higher and cybrids the lower concentration . Cell number normalisation was done by sulforhodamine B staining of TCA fixed cells in the seahorse plate as described[32] . Quantitative reverse-transcriptase-polymerase chain reaction ( qRT-PCR ) TaqMan assays ( Applied Biosystems , CA ) were used to measure the expression of mitochondrially encoded messenger RNA . The following primers were used: RNA was prepared in Trizol ( Invitrogen ) according to the manufacturers recommendations , resuspended in formamide and 1 μg separated on a 15% polyacrylamide/TBE gel . RNA was then transferred by wet blotting onto Immobilon NY+ nitrocellulose membranes ( Millipore , Watford , UK ) and crosslinked in a Stratalinker at 120 mJoules . The crosslinked membranes were transferred to hybridisation tubes and pre-hybridised for an hour at 42°C in hybridisation solution ( 50% formamide , 10% dextran sulfate , 1% SDS , 5x Denhardt's solution in 5x SSPE buffer ) . Random-labelled probes were generated by incubating 100 ng PCR amplified tRNA and 5S rRNA templates with 1 . 85 MBq α-32P dCTP ( Perkin Elmer , Seer Green , UK ) , random decanucleotides and Klenow polymerase exo- fragment ( Thermo Scientific DecaLabel DNA Labelling Kit , Waltham , MA ) . The following primers were used to generate the PCR products: Probes were purified on NICK columns ( GE healthcare , Sale , UK ) and a specific activity of 2x10^6 cpm was added to the membranes in hybridisation solution and incubated over night at 42° C . Up to 3 probes were used together ( two tRNAs and 5S rRNA ) . The membranes were rinsed three times with 2x SSPE at room temperature and once with 2X SSPE/2%SDS for 15 minutes at 65° . Bound probe was quantified by phosphorimaging . The transcription reactions were performed as described previously ( Posse et al . 2014 ) with 100 mM NaCl , 400 nM TFAM , 20 nM POLRMT , 60 nM TFB2M and 4 nM template . The templates used were obtained by restriction cleavage ( MscI and HindIII ) of pUC18 plasmids containing a mitochondrial DNA insert ( human mtDNA 1–741 , 547A or 547A>T ) generating run-off products of 90 and 190 nucleotides from the light and heavy strand promoters . The α-32P-UTP labelled transcription products were separated on a 4% Urea PAGE and exposed on a phosphorimager plate . HSP transcript levels were quantified and normalized to LSP transcription from the same reaction . Patient derived fibroblasts were incubated with normal ( light ) amino acids while control fibroblasts of the same low passage number were incubated with 13C6 Arginine and 13C8 Lysine ( heavy ) for at least 5 passages . Pairs of patient and control fibroblasts were mixed at 1x10^7 cells each and lysed by dounce homogenisation . Mitochondria were enriched as previously described by differential centrifugation followed by further enrichment on a Percoll gradient[33] . 50 μg of enriched mitochondria were resolved approximately 6 cm into a pre-cast 4–12% Bis-Tris polyacrylamide gel ( Novex , Thermo Fisher Scientific , East Grinstead , UK ) . The lane was excised and cut in 8 approximately equal chunks and the proteins reduced , alkylated and digested in-gel . The resulting tryptic peptides were analysed by LC-MSMS using a Q Exactive coupled to an RSLCnano3000 ( Thermo Fisher Scientific ) . Raw files were processed using MaxQuant 1 . 5 . 2 . 8 using Andromeda to search a human Uniprot database ( downloaded 03/12/14 ) . Acetyl ( protein N-terminus ) , oxidation ( M ) and deamidation ( N/Q ) were set as variable modifications and carbamidomethyl ( C ) as a fixed modification . SILAC data was loaded in R to process it with the Microarray-oriented limma package to call for differential expression[34] , relying on the original normalisation processes produced by MaxQuant as reported previously[35 , 36] . The proteomics data was deposited to the ProteomeXchange Consortium via the PRIDE[37] partner repository with the dataset identifier PXD004809 . A linear model was fitted with limma for the log-2 normalised expression ratios; moderated t-statistics , moderated F-statistics , log-odds and associated p-values for differential expression were produced with the same package . P-values were corrected for multiple hypothesis testing . Proteins were matched to the different mitochondrial respiratory complexes based on the gene family annotation available at HGNC[38] and plotted using ggplot2 [39] . Low variance of the data between individual donors is indicated by the low incidence of standard deviation greater than 20% as shown in S3 Fig . Quantitation of mitochondrial protein translation was done as described by Chomyn[40] . Cells were grown to 80% confluence in 6 well dishes , starved for 20 min in medium without methionine and cysteine , cytosolic translation was blocked with emetine ( 100 ug/ml ) for 30 min before adding 6 . 8 MBq radiolabelled 35S methionine/cysteine ( EasyTag Express35S , Perkin Elmer , Seer Green , UK ) for 30 min . Washed cells were harvested and separated on TGX Stain-Free precast gels ( Bio-Rad , Marnes La Coquette , France ) , equal loading was confirmed by UV imaging ( ChemiDoc , Bio-Rad ) prior drying and exposure to phosphorimager screens Unless otherwise stated , Student’s t-test was used to assign significance , assuming normal distribution of variance ( two-tailed , non-paired ) . Mean and standard deviation are indicated where appropriate . The traces of the whole mitochondrial sequencing for the m . 547A>T variant can be found in the NCBI trace archive , accession number TI 2344039844–2344039995 . The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE[37] partner repository with the dataset identifier PXD004809 . | Mitochondria provide the cell’s energy through respiration , using glucose and oxygen to produce ATP . Critical components for this process are encoded by maternally inherited mitochondrial DNA . Mutations in mitochondrial DNA usually affect organs that use energy intensively , notably the brain , eyes and muscles . Here we have discovered a variant in the promoter region of mitochondrial DNA in a large family with kidney disease and normal function in other organs . We show that patient-derived mitochondria do not generate energy normally and have reduced promoter activity , leading to loss of mitochondrial gene expression . We also identify similar mutations that might provide a mechanism for other cases of unexplained inherited kidney disease: two families with reduced cellular respiration carry mutations affecting mitochondrial phenylalanine tRNA , normally required for mitochondrial protein translation . Together , this study provides the first example of a disease-causing mutation in the mitochondrial promoter , and also establishes that the kidney is particularly sensitive to different mutations in mitochondrial DNA . As these can arise from a mutation in the promoter region of mitochondrial DNA or in the tRNA itself , and both reduce the amount of mitochondrial phenylalanine tRNA , this may provide a potential unifying mechanism for mitochondrially inherited kidney disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"transfer",
"rna",
"medicine",
"and",
"health",
"sciences",
"cultured",
"fibroblasts",
"biopsy",
"mitochondrial",
"dna",
"biological",
"cultures",
"fibroblasts",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"connective",
"tissue",
"cells",
"forms",
"of",
"d... | 2017 | Mutations in mitochondrial DNA causing tubulointerstitial kidney disease |
The decision to replicate its DNA is of crucial importance for every cell and , in many organisms , is decisive for the progression through the entire cell cycle . A comparison of animals versus yeast has shown that , although most of the involved cell-cycle regulators are divergent in both clades , they fulfill a similar role and the overall network topology of G1/S regulation is highly conserved . Using germline development as a model system , we identified a regulatory cascade controlling entry into S phase in the flowering plant Arabidopsis thaliana , which , as a member of the Plantae supergroup , is phylogenetically only distantly related to Opisthokonts such as yeast and animals . This module comprises the Arabidopsis homologs of the animal transcription factor E2F , the plant homolog of the animal transcriptional repressor Retinoblastoma ( Rb ) -related 1 ( RBR1 ) , the plant-specific F-box protein F-BOX-LIKE 17 ( FBL17 ) , the plant specific cyclin-dependent kinase ( CDK ) inhibitors KRPs , as well as CDKA;1 , the plant homolog of the yeast and animal Cdc2+/Cdk1 kinases . Our data show that the principle of a double negative wiring of Rb proteins is highly conserved , likely representing a universal mechanism in eukaryotic cell-cycle control . However , this negative feedback of Rb proteins is differently implemented in plants as it is brought about through a quadruple negative regulation centered around the F-box protein FBL17 that mediates the degradation of CDK inhibitors but is itself directly repressed by Rb . Biomathematical simulations and subsequent experimental confirmation of computational predictions revealed that this regulatory circuit can give rise to hysteresis highlighting the here identified dosage sensitivity of CDK inhibitors in this network .
Understanding the mechanisms of plant growth and differentiation is an important task , given the global biomass of land plants with approximately 600 billion tons of carbon [1] . Although cell proliferation is a main determinant of growth , relatively little is known about cell-cycle regulation in plants in comparison to yeast or metazoans . The typical eukaryotic cell cycle , as found also in plants , is divided into four phases: the S ( synthesis ) phase in which the nuclear DNA is replicated; the M ( mitosis ) phase in which sister chromatids are separated and distributed to the newly forming daughter cells; and two gap ( G1 and G2 ) phases that separate the M and S phases . The control of the G1-to-S-phase transition is a key step in cell-cycle regulation because cells typically become committed to divide after they have replicated their DNA [2]–[4] . In all eukaryotes , S-phase entry is tightly regulated by various mechanisms , incorporating intrinsic information , such as nutrient status and developmental program , with extrinsic , environmental conditions , such as temperature . Intrinsic and extrinsic cues are integrated through a complex control of the central driving force of cell-cycle progression , i . e . the cyclin-dependent kinases ( CDKs ) . Only when the sum of the different input systems is positive , CDKs become activated and entry into the next cell-cycle phase will be promoted once a certain threshold of activity is reached [5] . The four major input systems that regulate CDK activity are binding of positive cofactors ( i . e . cyclins ) and negative regulators ( i . e . CDK inhibitors ) , and positive and negative phosphorylation events ( i . e . at threonine and/or tyrosine residues of the T- and the P-loop , respectively ) . All four modules are themselves under elaborate control , for instance through regulation of the protein stability of CDK inhibitors [6] . Work in yeast and metazoans has shown that these regulatory modules are typically wired so that the activated CDK complex promotes its activators , while inhibiting its counter players . This circuitry leads only to two stable steady states , inactive or active , that generate a biological switch . Through the feedback wiring , the system becomes buffered against small changes in regulator concentrations , namely the activator concentration must be higher to switch from G1 phase to S phase than to remain in the S phase . This property of the feedback wiring , called hysteresis , greatly reinforces the switch-like behavior and is important in many biological processes; in the case of cell-cycle regulation , it is thought to be critical to promote the unidirectional progression of the cell cycle . In general , cell-cycle regulation appears to be conserved among eukaryotes . Analysis of the Arabidopsis thaliana genome has revealed that the majority of the core cell-cycle regulators known from yeast and/or metazoans is also present in plants , such as homologs of the transcriptional regulator E2F and its counter player Retinoblastoma ( Rb ) [7] . In particular , the general theme of the CDK-cyclin-regulated cell-cycle progression seems to be conserved . Functional analyses have shown that CDKA;1 , the only homolog of Cdk1/Cdc2+/CDC28p , is required throughout the Arabidopsis life cycle 8–12 . Nevertheless , the mammalian and plant cell-cycle control differs pronouncedly . For instance , the CDKA;1 regulation by phosphorylation and dephosphorylation through WEE1 and CDC25 , respectively , is not used in the cell-cycle control of Arabidopsis [11] , [13] , [14] . In addition , cell-cycle regulators are present in plants that are unknown in animal or yeast model systems or are only very distantly related to their metazoan or microbial counterparts . The plant CDK inhibitors , represented by two plant-specific groups , the INHIBITOR/INTERACTOR OF CDK or KIP-RELATED PROTEIN ( ICK/KRP ) and the SIAMESE RELATED are both only very little similar to the animal CDK inhibitor p27Kip1 [15]–[18] . A paradigm for the importance of precise cell-cycle control is the generation of gametes in flowering plants during the plant-specific gametophytic life phase that starts after meiosis with the formation of four monocellular haploid spores [19] . The male gametophyte has been analyzed in detail because it is more easily accessible than the female part , especially the cell proliferation and differentiation of the microspore into a mature pollen grain . The microspore undergoes strictly two rounds of mitotic cycles [20] . A first division of the microspore ( pollen mitosis I [PMI] ) that is asymmetric results in a smaller generative cell that is engulfed by a larger vegetative cell . The vegetative cell will exit the cell cycle and the plant retinoblastoma homolog RETINOBLASTOMA RELATED 1 ( RBR1 ) is required to terminate this lineage , because pollen with multiple vegetative cells are formed in rbr1 mutants [21] , [22] . The generative cell represents the beginning of the short male germline and will undergo one final division ( PMII ) , leading to two sperm cells , whereas RBR1 seemingly restricts its cell proliferation and/or the sperm cells , as supernumerary sperm cells can be found in rbr1 pollen [21] , [22] . The CDKA;1 activity is of key importance for PMII . The cdka;1 mutant pollen develops a vegetative cell similar to the wild type , but only one generative/sperm-like cell , see below [9] , [10] . A similar phenotype was observed also in mutants of the F-BOX-LIKE 17 ( FBL17 ) gene [23] , [24] . FBL17 was found to act as an adaptor protein in an SKP-CULLIN-F-BOX ( SCF ) complex and to mediate the degradation of KRP6 and KRP7 [24] . Consistently , mutants of KRP6 could partially restore the second mitotic division in fbl17 mutants [23] . KRP6 seems to be regulated during plant reproduction by at least one other mechanism involving the RING E3 ligases , RING-H2 group F 1a ( RHF1a ) and RHF2a , that also target KRP6 for degradation [25] . In rhf1a rhf2a double mutants , embryo sac development was arrested at early stages and , likewise , pollen development was defective at PMI and PMII . These gametophytic cell-cycle defects could be phenocopied by KRP6 overexpression [25] , consistent with previous experiments in which pollen mitosis in Brassica napus ( rapeseed ) was blocked by ectopic expression of ICK1/KRP1 from Arabidopsis [26] . However , a detailed molecular genetic framework of cell-cycle control is still missing in flowering plants . Especially , cell-cycle control during female gametophyte development is far from being understood . Here , we identified a regulatory cascade that functioned during all divisions in female and male gametophyte development . Subsequently , the biomathematical modeling of this network revealed that this circuitry can generate hysteresis . In this network , the CDK inhibitors are of central importance . We postulate that this cascade builds a general G1/S-phase module that probably operates in all cells of Arabidopsis and other plant species as well .
As CDKA;1 is the only functional Arabidopsis homolog of the yeast Cdc2+/CDC28p protein , this kinase might plausibly be involved in cell-cycle control in every cell . This assumption was supported by hypomorphic cdka;1 mutants , in which many , if not all , cells were affected , such as mitotically dividing cells in the epidermis of leaf primordia , as well as endoreplicating leaf hairs ( trichomes ) [8] , [11] , [27] . However , heterozygous null mutants were arrested or even delayed only in the second mitotic division during male gametogenesis , although the CDKA;1 promoter was active throughout the male gametophyte development [9] , [10] , [28] ( Figure 1A , 1B , 1D ) . Moreover , female gametogenesis that comprises three divisions was not affected at all in the heterozygous mutant . To determine the localization pattern of CDKA;1 , we analyzed the accumulation of a CDKA;1-YFP fusion protein during the development of the female and male germlines ( Figure 2A–2AVI ) . Previously , the production of the CDKA;1-YFP fusion protein from the endogenous CDKA;1 promoter had been found to completely rescue the cdka;1−/− mutants [29] . On the male side , CDKA;1-YFP occurred in the microspore mother cell ( Figure 2A ) and , subsequently , in the nucleus of the single-celled microspore ( Figure 2AI , 2AII ) . After PMI , CDKA;1-YFP was present in both the vegetative and generative cells ( Figure 2AIII , 2AIV ) . Besides in the nucleus of the vegetative cells , CDKA;1-YFP accumulated also in the two sperm cells after PMII ( Figure 2AV ) . At anthesis , CDKA;1-YFP could no longer be detected in the vegetative nucleus and the two sperm cells became exclusively marked by the fusion protein , consistent with a terminal state in the vegetative cell and the observation that the two sperm cells are still in S phase in mature pollen grains ( Figure 2AVI ) [30] . Analysis of the female gametophyte development revealed an overall similar accumulation pattern of CDKA;1-YFP . From meiosis onward , CDKA;1-YFP was found in all nuclei at all developmental stages of the developing embryo sac ( Figure 2D–2DVI ) . At maturity , CDKA;1-YFP activity withdrew from the accessory cells and was present only in the gametic cells , i . e . the egg cell and the central cell ( Figure 2DVI ) . As this expression pattern is consistent with a function of CDKA;1 throughout the female and male gametogenesis , we analyzed plants that were homozygous for the cdka;1 mutation and carried a single allele of the CDKA;1-YFP rescue construct . This situation mimicked heterozygous mutants in which approximately half of the pollen was arrested at PMII [29] . CDKA;1-YFP could be detected in all meiocytes and in all single-celled microspores ( Figure 2B–2BII ) . However , shortly before PMII , the YFP signal diminished in approximately half of the pollen and was not , or only hardly , visible at anthesis in almost 50% of the pollen ( Figure 2BIII–2BVI ) . Taken together , these observations are consistent with a carry over of CDKA;1 mRNA and/or protein from maternal , i . e . premeiotic and meiotic stages , and a subsequent reduction of CDKA;1 levels during male gametophyte development falling below the S-phase threshold concentration around the second mitotic division in cdka;1 mutant pollen ( Figure 2C ) . In contrast , almost all embryo sacs were CDKA;1-YFP positive ( data not shown ) , indicating a higher level of maternal CDKA;1 mRNA and/or protein inheritance at least partially accounting for the absence of a mutant phenotype during female gametogenesis . To unravel the function of CDKA;1 in early female and male gametogenesis , we assessed the possibility to additionally deplete CDKA;1 function in a heterozygous cdka;1 mutant background . Recently , the cdka;1 mutant pollen that is bicellular at anthesis , has been shown to still undergo a second division [28] . However , whereas the egg cell could still fuse with one of the generated sperm cells , karyogamy of the second sperm with the central cell failed for yet unknown reasons . Thus , the transmission rate of the mutant cdka;1 allele can be severely distorted and is not necessarily a good measure of the primary division activity and we therefore focused in the following analyses only on the pollen phenotypes . First , we generated artificial micro RNAs ( amiRNA ) against CDKA;1 ( amiCDKA;1 ) and expressed these amiCDKA;1 constructs in a heterozygous cdka;1+/− mutant background under the native CDKA;1 promoter . Indeed , the cdka;1 mutant phenotype was enhanced in these plants with 14% more bicellular pollen ( 57% ) than in cdka;1+/− heterozygous plants ( 43% ) and concomitantly , CDKA;1 protein levels were reduced in these plants ( Figure 1B , 1D; Figure S1; Table 1 ) . The observed phenotypic enhancement was consistent with inheritance of the CDKA;1 mRNA/protein , but the effect was small . When the amiCDKA;1 constructs were expressed in a wild-type background , CDKA;1 protein levels could be reduced to approximately the level of the heterozygous plants ( Figure S1 ) but only 5% of the pollen showed a cdka;1 mutant phenotype ( Table 1 ) . Next , we generated plants that produced a CDKA;1 mutant version in which Asp146 was replaced by Asn ( CDKA;1D146N ) driven by the CDKA;1 promoter . In mammalian and yeast kinases , homologous substitutions are known to abolish ATP access to the catalytic cleft , while cyclins and substrates are still bound , thus functioning as dominant-negative proteins [31] . Like in mammals and yeast , previous studies in plants have shown that this substitution has no kinase activity [32] . Surprisingly , the expression of CDKA;1D146N in heterozygous cdka;1+/− mutants partially rescued the pollen phenotype , namely between 85% and 93% of the pollen were tricellular and only 7% to 15% bicellular versus the typical 57% tricellular/43% bicellular pollen in heterozygous cdka;1+/− plants ( Figure 1B , 1D; Table 1 ) . Similarly , a StrepIII-CDKA;1D146N version also partially rescued the cdka;1 pollen phenotype ( data not shown ) . To test whether this effect was limited to the CDKA;1D146N construct , we generated two additional transgenic plants , another dominant-negative allele CDKA;1K33R fused to a StrepIII-tag and one CDKA;1PSTAIRE-dead version in which the archetypically conserved PSTAIRE domain in the C-helix of the N-terminal cyclin-binding domain ( residues Glu42–Glu57 ) and neighboring residues ( from Gly43–Lys56 ) had been replaced by 14 alanines ( designated PSTAIRE-dead ) , resulting in a protein that could presumably not bind to cyclins any longer . Similarly to CDKA;1D146N , CDKA;1K33R could partially rescue the pollen development of cdka;1+/− heterozygous mutants ( Table 1 ) . In contrast , cdka;1+/− mutants expressing the CDKA;1PSTAIRE-dead construct showed the typical cdka;1 pollen arrest at anthesis ( Table 1 ) . Thus , the PSTAIRE domain and , probably , the binding to the cyclin partner are important for the observed effect of the dominant-negative protein version . We conclude that , in contrast to the expected titration of cyclins or the blocking of the phosphorylation of substrates required for cell-cycle progression , the expression of the dominant-negative CDKA;1 versions from the CDKA;1 promoter interfered with negative factors of the cell cycle . Important negative regulators of plant cell-cycle progression are CDK inhibitors of the KRP class [15] , [16] . Moreover , KRP6 and KRP7 had been found previously to be expressed during male gametogenesis and overaccumulation of these inhibitors to be associated with cell-cycle arrest during pollen development [23]–[25] . Indeed , CDKA;1D146N could bind not only to cyclins but also to KRPs in bimolecular fluorescence complementation ( BiFC ) assays , whereas CDKA;1PSTAIRE-dead bound only to a cyclin-dependent kinase subunit ( CKS ) ( Figure 3A , 3B , 3C ) . The results suggested that the stoichiometric ratio between CDK inhibitors and CDKA;1 is crucial for the formation of two sperm cells . Therefore , we isolated null mutants in KRP3 and KRP4 ( Figure S2A , S2B ) and combined them as well as the previously described mutants for KRP1 , KRP2 , KRP5 , KRP6 and KRP7 [23] , [33] , [34] with cdka;1+/− mutants . The krp6−/− mutants and , to a lesser extent krp1−/− , could rescue the cdka;1 mutant pollen phenotype ( Table 1 ) . These results raised the question how cell-cycle progression is driven in the absence of CDKA;1 . Therefore , homozygous cdka;1−/− mutants complemented with only a hemizygous CDKA;1-YFP rescue construct were reassessed ( see above ) . We found that mutant pollen had occasionally only one single sperm-cell-like cell that still contained residual levels of the CDKA;1-YFP fusion protein ( Figure 2BVI ) . Thus , the observed rescue of cdka;1 by the krp1 and krp6 mutants is , at least partially , due to a liberation of remaining CDKA;1 levels in mutant pollen , additionally implying a previously not recognized role for KRP1 during gametogenesis . As FBL17 had been found to control KRP6 and KRP7 levels during male gametogenesis [23] , [24] , we asked whether KRP1 , as well as the other KRPs , might be also controlled by FBL17 . Therefore we combined mutants in all 7 KRPs with fbl17+/− mutant plants . In addition to the previously reported partial rescue of fbl17+/− by krp6+/− mutants [23] , we found that mutants in krp1 , krp3 , krp4 and krp7 could partially rescue fbl17+/− , as seen by an increased proportion of tricellular pollen at anthesis in comparison to heterozygous fbl17+/− plants ( Table 1 ) . To further test a possible role of FBL17 in degrading other KRPs than KRP6 and KRP7 , we co-expressed FBL17 with YFP fusion proteins of all seven KRPs in tobacco leaf cells and monitored the fluorescence intensity in comparison with the co-expression of KRP-YFP fusion proteins and CKS1 ( Figure 4A , 4B , 4C ) . In this assay , co-production of FBL17 reduced the fluorescence intensity especially of KRP3 , KRP4 , KRP5 and KRP7 , and to a lesser extent of KRP2 and KRP6 . This finding is consistent with the mediation of the proteasomal degradation of all or almost all KRPs through FBL17 . In further support of this conclusion , we found that , after CDKA;1D146N had been combined with fbl17+/− mutants , the proportion of wild-type-like tricellular pollen increased presumably due to a titration of overaccumulating levels of KRPs and , thus , at least in part , to a liberation of CDKA;1 ( Table 1 ) . Given its central importance , we next asked how FBL17 is regulated . The promoter of FBL17 contains a putative binding site for the transcription factor E2F [35] , and FBL17 transcripts strongly accumulate in plants that co-overexpress the transcription factor E2FA ( also designated E2F3 ) along with its dimerization partner DPa [23] . In support of a direct regulation by E2F , we could amplify the promoter fragment that contains the predicted E2F binding site after chromatin immunoprecipiation ( ChIP ) with an antibody against E2FA ( Figure 5A , 5B ) . Fragments further upstream and downstream of the putative E2F binding site could not be amplified , while a strong enrichment for a promoter fragment of the known E2F target PROLIFERATING CELL NUCLEAR ANTIGEN 1 ( PCNA1 ) as a positive control was obtained ( Figure 5B ) . The action of E2F is counterbalanced by Rb in mammalian cells [36] , [37] , and to test whether FBL17 is a direct target of the Rb homolog RBR1 in plants , we performed ChIP experiments with RBR1 using plants that express a functional RBR1-mRFP fusion protein [38] . As a positive control , RBR1 binding to the known RBR1 target PCNA1 was observed , while RBR1 ChIP did not precipitate heterochromatic loci used as negative controls ( Figure 5A , 5C; data not shown ) . As a proof of direct binding of RBR1 to the FBL17 promoter , we amplified one fragment of the FBL17 promoter from the precipitated DNA of the RBR1-mRFP-producing plant ( Figure 5A , 5C ) . Notably , this fragment contained the predicted E2F binding site ( Figure 5A ) and was the same fragment that could be precipitated in the E2F ChIP experiments ( Figure 5B ) . In contrast , fragments further upstream and downstream of the putative E2F site could not be amplified after ChIP suggesting that the interplay between E2F and RBR1 regulates FBL17 expression . Furthermore , the FBL17 transcript level was significantly upregulated in rbr1–2−/− mutants ( Figure 5D ) . To check then the biological importance of this wiring , we generated the fbl17+/− rbr1+/− double mutant . The mutant pollen phenotype of fbl17+/− plants was partially rescued , i . e . 43% bicellular pollen in fbl17+/− versus 30% in fbl17+/− rbr1+/− ( Table 1 ) . This is consistent with the higher level of FBL17 expression observed in rbr1−/− mutants ( Figure 5D ) and , although the fbl17 allele used here is a transcriptional null allele [23] , a modulation of FBL17 activity appears to be possible , assuming that similar to CDKA;1 , FBL17 mRNA and/or protein is carried over from meiotic stages . Conversely , we hypothesized that mutants in e2fa should enhance the fbl17 mutant phenotype . Analysis of embryo sac development in e2fa−/− fbl17+/− mutants did not reveal any deviation from the wild type , but a new class of mutant pollen was found that consisted of only one single cell versus the tricellular composition of pollen at anthesis in the wild type ( Figure 1E; Table 1 ) . Thus , the direct control of E2FA and RBR1 is decisive for FBL17 action at least during male gametophyte development . Moreover , the new phenotype of the e2fa−/− fbl17+/− mutants demonstrates that FBL17 already functions in the first division cycle of pollen development and likely , similar to the situation in cdka;1+/− mutants , is masked by a maternal carry over of FBL17 transcript and/or protein . The early function of FBL17 reinforced the idea that the interplay between CDKA;1 and FBL17 also controls PMI . Therefore , we combined the cdka;1 and fbl17 mutants that are linked by 3 cM on chromosome 3 . As no double heterozygous mutants could be recovered in the progeny of a backcross of cdka;1+/− fbl17+/− with the wild type , we used the above-described CDKA;1-YFP rescue line in cdka;1+/− for combinations with fbl17+/− . We could stably generate plants in which mutations in CDKA;1 and FBL17 were in cis located in the presence of a heterozygous CDKA;1-YFP rescue construct in trans . In these plants , approximately 25% of the gametophytes were expected to be devoid of both FBL17 and CDKA;1 ( either the endogenous transcripts or the CDKA;1-YFP rescue construct ) . However , due to the identified inheritance of CDKA;1 and probably of FBL17 as well , a mutant phenotype might be underrepresented . Analysis of pollen of these double mutants revealed that in nearly 10% the pollen contained only one single cell ( Figure 1C , 1E , Table 1 ) . Quantification of the fluorescence intensity of 4′ , 6-diamidino-2-phenylindole ( DAPI ) -stained mutant versus wild-type pollen showed that the single-celled pollen had a DNA content of 1C , i . e . arrested in G1 ( Figure 1F ) . Thus , CDKA;1 together with FBL17 control S-phase entry during both the first and second mitotic division during male gametogenesis . Next , we analyzed the female gametophyte in cdka;1+/− fbl17+/− ProCDKA;1:CDKA;1-YFP+/− plants . Approximately 25% of embryo sacs in these plants had a different developmental pattern than in the wild type . At maturity , embryo sacs contained only one or two nuclei of similar size and presented no sign of cellularization ( Figure 6A , 6D , 6G ) . The mutant embryo sacs remained unfertilized after pollination with wild-type pollen ( Figure 6B , 6E , 6H ) , indicating that , as expected , they were not functional . Consistently , we found aborting ovules in differentiating siliques ( Figure 6C , 6F ) . Notably , in the absence of the CDKA;1-YFP rescue construct , cdka;1 together with the fbl17 mutant allele was never transmitted through either the female or the male parent ( data not shown ) , implying that the combination of both genes is essential for the development of the two gametophytes . Taken together , CDKA;1 and FBL17 also control the first and second mitotic division cycle during embryo sac formation . The data presented above show that the CDKA;1-KRPs-FBL17-RBR-E2F pathway builds a general G1/S module that controls entry into S phase in Arabidopsis . This module involves four steps of negative regulation , i . e . RBR1 repressing FBL17 ( this study ) , FBL17 mediating the degradation of KRPs ( this study and [23] , [24] ) , KRPs inhibiting CDKA;1 [15] , [16] , [39] , and CDKA;1 phosphorylating RBR1 and inhibiting it [12] ( Figure 7A ) . Biomathematical simulations of this G1/S module revealed that this wiring gives rise to stable and self-maintaining steady states with a pronounced hysteresis ( Figure 7B , Dataset S1 ) . Moreover , the decision whether to move from G1 into S phase strongly depended on the concentrations of KRPs , consistent with the experiment with dominant-negative CDKA;1 variants . This model also predicted that the expression of FBL17 should intensely rely on CDKA;1-cyclin activity levels . In a cdka;1 mutant background , the system not only runs out of kinase to promote S-phase entry , but the remaining kinase activity might also be inhibited , because less FBL17 might be expressed due to the reduced repression of RBR1 , and thus the abundance of KRPs increases . To test this assumption , we introgressed a previously generated promoter reporter line for FBL17 into cdka;1 heterozygous mutants [23] . However , GUS levels could not reliably be quantified . Thus , we next performed qRT expression analyses from anthers of the fifth and fourth floral buds before the first flower opens , since under our growth conditions anthers of these flowers contain monocellular and bicellular pollen and thus , there was no bias for the number of pollen nuclei between wild-type and mutant plants . Consistent with the prediction of our model , we observed a significant ( unpaired t-test , p<0 . 05 ) reduction in FBL17 expression at both floral stages of heterozygous cdka;1 mutants ( Figure 7C , data not shown ) . Next , we functionally tested this feedback wiring by uncoupling the expression of FBL17 from its regulation by RBR1 . Indeed , expression of FBL17 from the ubiquitin as well as the CDKA;1 promoter in a heterozygous cdka;1 mutant situation could partially rescue the pollen phenotype ( Table 1 ) . Thus , we conclude that the proposed mathematical model captures central aspects of the presented G1/S module .
At the heart of the here-identified G1/S cell-cycle phase regulatory module is the inhibition of CDKA;1 by KRPs . In Arabidopsis , the KRP genes form a gene family of seven members with seemingly highly overlapping functions as suggested by the absence of an obvious mutant phenotype in single krp mutants [33] , [34] . Typically , null mutants for components that function in one regulatory pathway , e . g . here for cdka;1 , fbl17 and several krps , show an epistatic relationship . However , due to the high level of carry over that we identified here for CDKA;1 , but which is likely true for many other cell-cycle as well as developmental regulators , genetic interactions in one pathway , e . g . dosage sensitivity of CDKA;1 toward KRPs , could be revealed . Therefore , the heterozygous cdka;1+/− mutant background , with gradually decreasing levels of maternally provided CDKA;1 , offered a unique opportunity to quantitatively dissect the regulatory cascades at the S-phase entry and to evaluate the effects of the redundantly acting KRPs . The CDK dosage sensitivity became especially apparent when dominant-negative CDK variants were expressed in a cdka;1 mutant background . These dominant-negative CDK versions had been used previously in plants , as well as in yeast and metazoans , and are known to be completely devoid of kinase activity [32] , [40]–[45] . Instead of the expected negative effect on cell-cycle progression , these kinase versions could partially rescue the pollen phenotype of heterozygous cdka;1+/− mutants . Our data suggest that moderate expression levels of dominant-negative CDKs first bind to and titrate CDK inhibitors before sufficiently reducing the levels of active CDK-cyclin complexes . Indeed , it has been found that the biologically important concentration of cyclin lies in a nM range in Xenopus laevis [46] . Although the cyclin concentrations have not been quantified in Arabidopsis , it is well known that Arabidopsis cyclin promoters are very strong , implying that the concentrations are high in plants as well [47]–[49] . In contrast , the abundance of KRPs seems to be very low [50]–[52] . Thus , the moderate expression of ‘dominant-negative’ kinases might be a useful tool to titrate CDK inhibitors outside of pollen in other developmental/physiological contexts . Interestingly , only the mutant CDK variants that could bind to cyclins , i . e . CDKA;1D146N and CDKA;1K33R , partially rescued cdka;1 mutants . In contrast , expression of the CDKA;1PSTAIRE-dead version , which could bind to the CKS cofactor , but not to cyclins , did not rescue the defective pollen in cdka;1 mutants , giving rise to the hypothesis that KRPs may preferentially bind to CDK-cyclin complexes or , at least , have a higher affinity for the dimer than for the individual monomers , consistent with previous interaction assays showing that monomeric CDKs and cyclin were hardly targeted in vitro by KRPs in contrast to a CDK-cyclin complex [39] . Revisiting previously published yeast two-hybrid ( Y2H ) interaction data of ICK1/KRP1 with CDKA;1 and D-type CYCLIN 3;1 ( CYCD3;1 ) also revealed that deletion of the presumptive CDK binding site strongly reduced the interaction of ICK1/KRP1 with CYCD3;1 , further supporting the assumption that KRPs preferentially target a CDK-cyclin dimer [39] , [53] . The CDK inhibitor p27Kip1 from mammals belongs to the class of intrinsically unstructured , also called intrinsically disordered or natively unfolded , proteins [54] , [55] . Kinetic analyses of p27Kip1 have suggested that its folding is induced through binding to the cyclin partner and then reaches the Cdk partner . Consequently , p27Kip1 has the highest affinity for the Cdk-cyclin complex , followed by a preference for the cyclin partner over the isolated Cdk [56] . Thus , it is tempting to speculate that KRPs from plants have very similar structural and kinetic properties , especially because in the above-mentioned Y2H experiments the interaction of ICK1/KRP1 was much stronger with CYCD3;1 than with CDKA;1 , although it is unclear whether CYCD3;1 and CDKA;1 are equally well expressed in the yeast assay [53] . Previously , a cdka;1 mutant version , in which a conserved threonine ( T161 ) in the T-loop that needs to be phosphorylated for full CDK activity was exchanged with a nonphosphorylatable alanine or valine ( CDKA;1T161A and CDKA;1T161V ) , had been reported to be unable to rescue the cdka;1 mutants [8] , [32] . Interestingly , this CDK version could not rescue the cdka;1 mutant pollen , indicating that CDK inhibitors may only target activated CDK-cyclin complexes . In contrast , mammalian p27Kip1 can bind to CDKs irrespective of its T-loop phosphorylation [57] , but due to the different positions of the CDK-cyclin-binding domain in p27Kip1 versus KRPs , both proteins probably make contact to different parts of CDKs and cyclins . Thus , an important next step in the understanding of the KRP action is the unraveling of their crystal structure when bound to plant CDK-cyclin complexes . Because of its importance for growth and development , the S-phase entry has been extensively studied in yeast and metazoans . The crucial aspect of progression into S phase is the activation of the transcription factor E2F that is repressed by Rb in metazoans . Furthermore , Whi5 , an analogous transcriptional repressor in yeast , blocks the activity of the S-phase transcriptional regulator SBF [58] . As representatives of the Plantae supergroup of the eukaryotic kingdom , plants are much more distantly related to fungi and metazoans , which are both in the same supergroup of Opisthokonts [59]; therefore , a comparison of the cell cycle of plants with that of yeast and metazoans will be important to understand the origin and evolution of cell-cycle control in eukaryotes [60] . As the inactivation of the plant homolog RBR1 and , hence , the release of plant E2F homologs are seemingly conserved in plants , early eukaryotes might already have a complex repertoire of cell-cycle control genes [60] , [61] . An important mechanism in metazoans and yeast is the positive feedback regulation of E2F/SBF onto its own activity . An initial phosphorylation of Rb or Whi5 reduces the repression of E2F/SBF and promotes the expression of cyclin E and Cln2 , respectively . The increasing levels of these cyclins fully activate the S-phase kinases Cdk2 in metazoans or CDC28p in yeast , resulting in the complete inactivation of Rb/Whi5 [62] , [63] . Our data show that the principle of this double negative wiring of Rb proteins is conserved in plants . However , the regulators or their relative importance differ in the plant cell cycle: RBR1 represses E2F that activates FBL17 , which , in turn , releases the repression of CDKA;1 that can then phosphorylate RBR1 , presumably leading to its complete inactivation . Thus , rather than liberating a positive factor , such as a cyclin , plants inactivate another repressor , adding one layer of double negative regulation . Based on our simulations , this wiring can give rise to a strong bistable system with hysteresis . The observation that the concomitant loss of CDKA;1 and FBL17 results in a complete arrest of gametogenesis , underlines the crucial importance of this regulation . Nevertheless , due to the still very limited knowledge about plant cyclins , we cannot exclude that , in addition to the FBL17 loop , another pathway leads to the transcriptional activation of cyclins by E2F in plants . Conversely , the transcriptional control of the protein degradation machinery , targeting CDK inhibitors , by E2F and Rb ( or their functional analogs ) appears to be a universal regulatory mechanism . In animals , degradation of the CDK inhibitor p27Kip1 is mediated by the F-box protein Skp2 , which has been found to be a direct target of E2F and Rb regulation [64] . Perhaps a similar regulation is found in fission yeast ( Schizosaccharomyces pombe ) , where the MBF function is executed by the transcriptional regulator Cdc10/Rep2 and its inactivation results in a G1 arrest [65] . Remarkably , this arrest is accompanied by high levels of the Cdk inhibitor Rum1 , but it is currently unclear how Rum1 is regulated in this context . A simple hypothesis is that Cdc10 activates the degradation of Rum1 , possibly through the transcriptional activation of a yet unknown F-box protein or another component of the protein degradation machinery , ultimately hinting at still unexplored parallels between the plant , animal and yeast cell cycles , and revealing general principles and global constraints of cell-cycle control in eukaryotes .
The Arabidopsis thaliana ( L . ) Heynh . plants were all derived from the Columbia ( Col-0 ) accession . Detailed information on mutant lines used can be found in the extended experimental procedures ( Text S1 ) . All genotypes were determined by polymerase chain reaction ( PCR ) with the primers indicated in Table S1 . All seeds were surface-sterilized with chloride gas , sown on 0 . 8% Phyto agar plates ( half-strength Murashige and Skoog ( MS ) salts and 1% sucrose ) and grown under neutral conditions ( 12 h light at 21°C , and 12 h dark at 17°C ) . After germination , plants were transferred to soil and grown under long-day conditions ( 16 h day/8 h night regime at 22°C/18°C ) . For crosses , flowers of the female parent were emasculated 2 days before anthesis and hand-pollinated 2 days later . All manipulations were performed using standard molecular methods , details on the construction of transgenic lines can be found in the extended experimental procedures and Table S1 listing primer sequences . BiFC was assayed as previously described [66] . Co-injection experiments were performed with KRPs in pEXSG-YFP and CKS1 or FBL17 in pEXSG-CFP . Both Gateway compatible vectors were a kind gift of Marcel Wiermer ( AG Romeis , MPIZ , Cologne ) . Only cells with YFP and CFP signals were chosen for measurements . For each measurement single stack images of at least twenty nuclei were taken , all with the same laser settings . The fluorescent intensity of the nuclei was determined with Image J ( http://rsb . info . nih . gov/ij/ ) . Each experiment was performed three times . CKS1 was used as neutral control . For fluorescence microscopy analyses , ovules were dissected from the pistil in 50 mM sodium phosphate buffer pH 7 . 5 . YFP fluorescence of pollen and ovules at different developmental stages was analyzed on confocal microscopes ( Leica TCS SP5 AOBS , Zeiss LSM 510 and Zeiss 710 ) using a BP 530–600 filter . For DIC microscopy , mature ovules and developing seeds were prepared from siliques before and after pollination , respectively , and mounted on microscope slides in a clearing solution of 8∶2∶1 chloral hydrate∶distilled water∶glycerol as described [9] . For DAPI staining , pollen grains were gently released into the DAPI solution ( 2 . 5 µg/ml DAPI , 0 . 01% Tween , 5% dimethyl sulfoxide , 50 mM Na phosphate buffer , pH 7 . 2 ) and incubated at 4°C overnight before observation . Pollen viability was assessed by mounting pollen as described [67] . DNA content was quantified and measured with the software ImageJ ( http://rsbweb . nih . gov/ij ) on images taken with constant settings as described [23] . Differential Interference Contrast ( DIC ) microscopy was done with an Axioimager ( Zeiss ) . For ChIP experiments as described [12] , [68] , 2-week-old seedlings of plants expressing ProRBR1:RBR1:mRFP , kindly provided by Dr . Frédéric Berger [38] , growing on half-strength MS plates were used . Chromatin was sheared by means of a Bioruptor sonicator ( Cosmo Bio ) twice for 15 min with a 50% duty cycle and high-power output to obtain 200-bp to 1000-bp DNA fragments . A DsRed polyclonal antibody ( Clontech ) or an antibody raised against Arabidopsis E2FA protein , as kindly provided by Dr . Lieven De Veylder and described by Heyman et al . [69] , together with Protein A-magnetic beads ( Millipore ) were used for immunoprecipitation . Negative controls were done without antibody . DNA was recovered with Magna ChIP spin filters according to the manufacturer's instructions ( Millipore ) . ChIP DNA ( 0 . 5 µl or 1 µl of a 1/5 dilution ) was analyzed by semi-quantitative PCR or quantitative real-time PCR with gene-specific primers , respectively ( see Table S1 ) . Three biological and three technical replicates were performed for ChIP quantitative PCR with PCNA1 as positive control and primers of the heterochromatic region as negative control . Whole flowers ( Figure 5 ) or dissected anthers ( Figure 7 ) from the fifth and fourth floral buds preceding an open flower ( containing mono- and bicellular pollen ) were immediately frozen in liquid nitrogen and stored temporarily at −80°C . RNA was extracted using NucleoSpin RNA XS Kit ( MACHEREY-NAGEL ) . RNA concentration and purity was tested using nanodrop-photometric quantification ( Thermo Scientific ) . RNA integrity was verified by running 1 µl of total RNA on 1 . 5% agarose TBE-gels to detect the 28S and 16S rRNA bands . 100 ng up to 1 µg of total RNA was processed to obtain cDNA using polyT-primer and SuperScript III RNase H reverse transcriptase . As negative control , all steps were followed in the same manner , except for adding the reverse transcriptase . The resulting cDNA was used for quantitative Real Time-PCR ( qRT-PCR ) using the Roche LightCycler 480 system . Oligonucleotides were designed using either Primer3Plusdesign tool ( http://www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi ) or QuantPrime ( qPCR primer design tool: http://www . quantprime . de/main ) and used in a final concentration of 0 . 5 µM each ( primers are listed in Table S1 ) . Three to four biological with three technical replicates each were processed . Cq calling was done using the Second Derivative Maximum method . Target specific efficiencies were calculated as the mean of all reaction specific efficiencies for a given target . Reaction specific efficiencies were deduced using LinRegPCR 7 . 4 ( http://LinRegPCR . nl ) [70] , [71] . Data was quality controlled , normalized against 2 reference genes and statistically evaluated ( unpaired t-test ) using qbasePLUS 2 . 3 ( http://www . biogazelle . com/products/qbaseplus ) [72] . For the expression analysis of FBL17 and CDKA;1 in heterozygous cdka;1 mutants , ACT2 and EF1 alpha were used as reference genes . For the analysis of FBL17 expression in rbr1–2 mutants , EF1 alpha and TIP41 were used as reference genes ( see Table S1 ) . The molecular network of G1/S regulators was described by ordinary differential equations . Phosphorylations/dephosphorylations and complex formations/dissociations were assumed to be fast , relative to protein synthesis and degradation , and steady-state approximations were used for these steps . The balance curves ( nullclines ) for total KRP and CDKA;1-cyclin complexes were calculated by the XPP/Aut program provided in Dataset S1 . | In order to grow , multicellular organisms need to multiply their cells . Cell proliferation is achieved through a complex order of events called the cell cycle , during which the nuclear DNA is duplicated and subsequently distributed to the newly forming daughter cells . The decision to replicate the nuclear DNA is in many organisms crucial to progress through the entire cell cycle . Alterations of the cell cycle , especially at the entry point , can cause severe developmental defects and are often causal for maladies , such as cancer . Substantial work in yeast and animals has revealed the regulatory steps controlling S-phase entry . In contrast , relatively little is known about the plant cell cycle despite plants being one of the largest classes of living organisms and despite the importance of plants for human life , for instance as the basis of human nutrition . Our work presents a molecular framework of core cell-cycle regulation for entry into the DNA replication phase in the model plant Arabidopsis . We report here the identification of a regulatory cascade that likely functions in many plant cells and organisms . With this , we also provide an important basis for comparative analyses of cell-cycle control between different eukaryotes , such as yeast and mammals . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2012 | A General G1/S-Phase Cell-Cycle Control Module in the Flowering Plant Arabidopsis thaliana |
MicroRNAs are important regulators of gene expression , acting primarily by binding to sequence-specific locations on already transcribed messenger RNAs ( mRNA ) and typically down-regulating their stability or translation . Recent studies indicate that microRNAs may also play a role in up-regulating mRNA transcription levels , although a definitive mechanism has not been established . Double-helical DNA is capable of forming triple-helical structures through Hoogsteen and reverse Hoogsteen interactions in the major groove of the duplex , and we show physical evidence ( i . e . , NMR , FRET , SPR ) that purine or pyrimidine-rich microRNAs of appropriate length and sequence form triple-helical structures with purine-rich sequences of duplex DNA , and identify microRNA sequences that favor triplex formation . We developed an algorithm ( Trident ) to search genome-wide for potential triplex-forming sites and show that several mammalian and non-mammalian genomes are enriched for strong microRNA triplex binding sites . We show that those genes containing sequences favoring microRNA triplex formation are markedly enriched ( 3 . 3 fold , p<2 . 2 × 10−16 ) for genes whose expression is positively correlated with expression of microRNAs targeting triplex binding sequences . This work has thus revealed a new mechanism by which microRNAs could interact with gene promoter regions to modify gene transcription .
MicroRNAs influence a broad spectrum of biological processes and have been extensively characterized as negative regulators of gene function . By pairing with complementary sequences in messenger RNA ( mRNA ) , they are known to down-regulate gene function by enhancing transcript degradation or sequestration , or via suppression of translation . MicroRNAs have also been shown to up-regulate mRNA transcript levels for some genes , but the mechanism ( s ) for increasing gene expression have not been fully elucidated [1–5] . One indirect mechanism by which microRNAs may up-regulate gene expression is via suppression of mRNAs encoding transcriptional suppressors . In addition , there are reports that interactions between microRNAs and gene promoter regions may play a more direct role in regulating the efficiency of gene transcription [1–3 , 6] , for example by mediating de novo CpG methylation [7] . However , it is possible that there are other unidentified or not fully elucidated mechanisms by which microRNAs directly interact with genes to enhance gene transcription . Because double stranded DNA is capable of forming triple-helical structures through interactions with DNA or RNA in the major groove of the DNA duplex , we and others have postulated that microRNA may form triplex structures with duplex DNA via either Hoogsteen or reverse Hoogsteen hydrogen bonds , and thereby directly interacting with target DNA sequences in regulatory regions and gene promoters in the human genome , with the potential to alter gene function [8–10] . Here we provide direct physical evidence that microRNAs of sufficient length and sequence can bind to double stranded DNA to form hetero-triplex structures at specific target sequences in DNA . We computationally show that the human genome , as well as the genomes of multiple other species , contain DNA sequences with properties favoring microRNA triplex formation . We also show that those genes containing sequences favoring microRNA triplex formation are enriched ( 3 . 3 fold ) for genes whose expression is positively correlated with expression of microRNAs targeting triplex binding sequences , indicating this as a potential mechanism via which microRNA can directly enhance gene expression .
To assess the landscape of potential microRNA triplex binding sites in genomic DNA , we developed and implemented a computational algorithm ( ‘Trident’; http://trident . stjude . org ) to identify Hoogsteen and reverse Hoogsteen interactions between single stranded oligonucleotides ( i . e . microRNAs ) and double-stranded oligonucleotides ( i . e . duplex DNA ) . The algorithm identifies Hoogsteen and Reverse Hoogsteen interactions , independently searching for triplex forming units ( e . g . Hoogsteen TA:U and CG:C; Reverse Hoogsteen TA:A and CG:G , in the form XY:Z , where Z represents the microRNA nucleotide ) between stretches of polypurine genomic DNA and either polypurine or polypyrimidine third strand microRNAs , on an individual base level . For each detected triplex binding site ( those sites capable of forming an interaction of multiple units ) , a thermodynamic binding energy and heuristic score was determined , with higher heuristic score and lower thermodynamic energy indicating stronger interaction . To determine the thermodynamic energy for binding , first order free energy calculations were performed to determine the amount of binding energy of each type of interaction . Heuristic score was determined based on the number of triplex forming pairs found between the interacting microRNA and double stranded DNA . Using this computational algorithm , we performed genome-wide binding site analyses on the genomes and microRNAs of several species , as well as randomly generated DNA sequences ( Fig 1 ) . For each genome analyzed , the genome-wide results with heuristic scores greater than or equal to 140 ( ≥7 triplex forming units ) were ranked and categorized on the basis of the number of other identified binding sites with a better energy/score combination ( having lower energy and higher score ) . Those interactions with an energy/score ranking greater than the top 0 . 001% were classified as grade 1 hits , with grades 2–5 being assigned to interactions with energy/score rankings with successive ten-fold lower criteria ( e . g . top 0 . 01% , 0 . 1% , and 1% ) . This analysis revealed a highly significant enrichment of microRNA triplex binding sites in all genomes analyzed ( Fig 1 ) , including Homo sapiens ( Fig 1I ) , Mus musculus ( Fig 1J ) , Rattus norvegicus ( Fig 1N ) , Caenorhabditis elegans ( Fig 1D ) and Arabidopsis thaliana ( Fig 1B ) , when compared to random DNA sequences analyzed for microRNA binding sites . Comparing the log-transformed interpolated frequency of identified binding sites of human genome and randomly-generated DNA sequences , reveals a distinct enrichment ( p-value <2 . 2 × 10−16 ) in low energy and high score hits ( Fig 1M ) . To identify potential novel classes of microRNA that bind to double stranded DNA , we assessed the sequence content of the microRNAs computed to participate in triplex formation . Comparisons of the identified binding site frequencies and microRNA sequence content revealed a marked enrichment ( Fig 2A ) in identified binding sites for microRNAs exhibiting imbalanced purine to pyrimidine content ( e . g . high purine content or high pyrimidine content ) . Notably , the distribution of purine content in known human microRNAs has a mean of approximately 50% , suggesting that those microRNAs with imbalanced purine to pyrimidine content were responsible for a disproportionate number of triplex binding interactions . Indeed , microRNAs with greater than 75% purine or greater than 75% pyrimidine content accounted for 95 . 3% of binding sites ( Grades 1–4 ) identified and this distribution was significantly different than the distribution of purine content in known human microRNAs ( 2-sample test for equality of proportions with continuity correction p-value <2 . 2 × 10−16 ) . This enrichment was significantly different as compared to imbalanced GC ( Fig 2B ) where only 7 . 8% of identified binding sites had microRNAs with greater or less than 25% GC content ( 2-sample test for equality of proportions with continuity correction p-value <2 . 2 × 10−16 ) or imbalanced microRNA GU ( Fig 2C ) content where only 15 . 8% of hits had microRNAs with greater or less than 25% GU content ( 2-sample test for equality of proportions with continuity correction p-value <2 . 2 × 10−16 ) . In addition to purine:pyrimidine content , imbalance being an important determinant of triplex formation , lower than average U content ( Fig 2H ) , higher than average G ( Fig 2F ) or C ( Fig 2G ) content also predicted affinity for double stranded DNA binding . We found no evidence that average A content ( Fig 2E ) was a determinant of microRNA binding to double stranded DNA . To verify that microRNAs are capable of physical interaction and binding to double-stranded DNA , we designed orthogonal methods to directly interrogate binding . A fluorescence resonance energy transfer ( FRET ) based method ( Fig 3B ) to detect triplex formation was designed such that a double stranded DNA intercalating dye ( SYBR Green II ) , when excited at 480 nm , transfers energy to a carboxy-X-rhodamine ( ROX ) molecule covalently coupled to a triplex forming RNA . Decreased emission at 520 nm of the double-stranded DNA intercalating donor dye corresponds with increased emission of ROX acceptor dye at 610 nm . Utilizing this method , we detected the interaction of hsa-miR-483-5p ( a microRNA high in purine content ) and a double stranded DNA ( an identified Hoogsteen binding site in our genome wide screen ) , as evidenced by the decreased SYBR Green emission and increased ROX emission ( Fig 3A ) . Utilizing a complementary surface plasmon resonance ( SPR ) based method , we verified that hsa-miR-483-5p immobilized via biotin based coupling to the detector surface ( Fig 3D ) was able to bind duplex DNA , but neither hsa-miR-1 nor hsa-miR-98 ( microRNAs with mixed purine/pyrimdine content ) was able to bind complementary double stranded DNA ( S1 Fig ) . Kinetic analysis ( Fig 3C ) yielded an association rate constant ( ka ) of 3 . 96 ( ± 0 . 01 ) × 105 M-1s-1 , a dissociation rate constant ( kd ) of 5 . 01 ( ± 0 . 07 ) × 10−4 s-1 and an equilibrium dissociation constant ( KD ) of 1 . 27 ( ± 0 . 02 ) nM . To corroborate our findings by FRET and SPR ( Fig 3 ) , we also performed EMSA experiments to document binding between hsa-miR-483-5p and ROX-labeled 24-bp hairpin duplex DNA , but reasoned that the interaction between duplex DNA and microRNA is likely transient and not readily detectable by EMSA , whereas single stranded purine-rich DNA oligonucleotides were more likely to form stable triplexes that are detectable by EMSA [11] . To test this theory , we performed EMSA experiments which documented that Hoogsteen bond-optimized hsa-miR-483-5p RNA ( 483-opti ) competed with 483-opti DNA oligo with the same nucleotide sequence for binding to duplex DNA , resulting in decreased amounts of triplex DNA and increased amounts of duplex DNA ( Fig 4A , lanes 3–5 ) , providing evidence that Hoogsteen bond-optimized hsa-miR-483-5p ( 483-opti ) binds to duplex DNA . In contrast , because of fewer favorable Hoogsteen bonds , hsa-miR-483-5p ( Fig 4A , lanes 6–8 ) and an RNA oligo with scrambled sequence ( Fig 4A , lanes 9–11 ) did not compete with the DNA oligo for binding to duplex DNA . In addition , EMSA experiments with 11-nucleotide DNA and RNA oligos corresponding to the 5’ and 3’ regions of Hoogsteen bond-optimized hsa-miR-483-5p , did not result in detectable triplex formation with the hairpin duplex DNA ( S2A Fig ) , suggesting that sequence , purine content , and length of the microRNA are important factors influencing binding of microRNA to duplex DNA . These competition EMSA results indicate that microRNA-duplex DNA triplex formation is transient , and better suited for detection by more sensitive methods such as FRET , SPR , and NMR . To corroborate our findings by EMSA , we performed Two-Dimensional ( 2D ) [1H , 1H] NMR of 24 bp hairpin duplex DNA in presence and absence of 22 nucleotide single stranded hsa-miR-483-5p RNA oligo and DNA oligo with the same sequence ( Fig 4B and 4C ) . Overall single stranded RNA ( hsa-miR-483-5p ) and single stranded DNA mixtures with hairpin duplex DNA show similar binding profile , with similar improvement in the peak intensities and chemical shift perturbations with the appearance of new peaks highlighted in blue boxes ( Fig 4B and 4C ) , suggesting that single stranded DNA and single stranded RNA of the same sequence bind to DNA duplex in a similar manner; the major differences ( in red boxes ) are one peak among thymidine cross-peaks ( Fig 4B ) , showing an intermediate change ( peak disappearing ) with single stranded RNA while saturated with hairpin duplex DNA , and one additional peak probably coming from the loop region; two new peaks among cytosine cross-peaks ( Fig 4C ) showing much higher intensities with single stranded DNA , indicating that single stranded DNA binds to duplex DNA duplex with higher binding affinity than RNA , consistent with the results obtained by EMSA . We modeled DNA-microRNA triplex of double stranded DNA and hsa-miR-483-5p in silico by simulated annealing with distance restraints derived from Hoogsteen base pairing , and subsequently simulated the structure by Langevin molecular dynamics in generalized born solvent model . ( Fig 4D ) . The model shows that the hsa-miR-483-5p microRNA strand is binding to the targeted DNA duplex region in an antiparallel mode , and most of the predicted Hoogsteen hydrogen bonds are reasonably well maintained even after the removal of external restraints in the top rated predicted sequence ( Fig 4D-I ) . By comparison , the negative control model of triplex with the reverse RNA sequence cannot maintain Hoogsteen pairs during MD simulation ( Fig 4D-II ) . This binding model is overall consistent with the chemical shift changes observed for the cytosine and thymidine signals of DNA duplex upon single stranded RNA binding ( Fig 4B and 4C ) . Overall , the molecular modeling is consistent with EMSA and NMR results , which suggests that longer purine-rich RNA can form triplex with DNA duplex in an antiparallel manner . In contrast , there was complete overlap between the two TOCSY spectra of free double stranded DNA and that of double stranded DNA combined with a shorter , 11-nucleotide truncated hsa-miR-483-5p oligo ( S2B and S2C Fig ) , indicating that this short RNA was incapable of triplex formation , which is consistent with a previous report that shorter purine-rich RNA cannot form triplex with double stranded DNA[11] . This confirms our EMSA results that , besides the sequence and purine-content , the length of the microRNA is an important factor influencing triplex formation between double stranded DNA and microRNA . To further interrogate our genome-wide assessments of microRNA binding sites , we measured microRNA and mRNA expression levels in primary leukemia cells isolated from two independent cohorts of patients enrolled on either St . Jude Total Study 15 or Study 16 protocol for children with newly diagnosed acute lymphoblastic leukemia ( ALL ) and assessed their correlations . Spearman correlation analysis was performed in each cohort separately , cross comparing every microRNA to every mRNA probe set . Meta-analysis combining the results from both cohorts revealed that for those microRNA-gene pairings with an identified grade 1 Trident binding site ( within 5000 base pairs up and down stream of the gene ) , there was a marked enrichment of significant positive correlations . There were 2639 genes that contained a duplex DNA sequence ( within 5000bp of the gene ) estimated via Trident to have a grade 1 interaction with a microRNA by either Hoogsteen or reverse Hoogsteen interaction . As shown in Fig 5 , of these 2639 genes , there was a highly significant enrichment ( 3 . 3-fold , p<2 . 2 × 10−16 ) for positively correlated mRNA-microRNA pairs ( n = 206 with p-values <0 . 01 ) , compared to negatively correlated microRNA-mRNA pairs ( n = 62 with p-values <0 . 01 ) .
Here we provide multiple lines of direct physical evidence that microRNAs can bind to double stranded DNA to form triplex structures and show that mammalian and non-mammalian genomes are enriched with microRNA triplex binding sites . Regulation of gene expression by microRNA binding directly to messenger RNA is well established . However , several studies have suggested the existence of a mechanism of transcriptional activation by microRNA binding to double stranded DNA , but no definitive mechanism for this microRNA-DNA interaction has been elucidated [1–3 , 12] . Additionally , the presence of microRNA in the nucleus[13] and a molecular mechanism for mature microRNA import into the nucleus[14] underscores the potential for nuclear functions of microRNAs . Indeed , there have been reports of triplex structures involving RNA-RNA interactions[15] , but microRNA-duplex DNA interactions warrant further study[16] . Purine bases have more than one face from which they can form hydrogen bonds , which allows them to simultaneously participate in Watson-Crick pairings and either Hoogsteen or Reverse Hoogsteen pairings . When a run of purines on one strand of the duplex occurs , a third strand of either DNA or RNA with the correct Hoogsteen complementarity can interact with the major groove of DNA to form a triple helix through the formation of Hoogsteen or Reverse Hoogsteen hydrogen bonds . Informatics approaches for identifying homopurine sequences in genomes have been reported previously [8–10 , 15 , 17 , 18] , however these methods did not contextualize these sequences in terms of potential for triplex formation with known microRNA species . Previous studies focused on the identification and interrogation by EMSA of stable interactions between target duplex DNA and purine or pyrimidine rich single stranded DNA or relatively short ( 12–14 mer ) RNA oligonucleotides , with mostly favorable Hoogsteen pairings[11 , 19] . Indeed , our studies ( using either EMSA or NMR ) showed that purine rich short microRNA ( e . g . , 11 nucleotides ) do not form stable triplex structures , consistent with previous reports[11] , whereas longer microRNA ( e . g . , 22 nucleotides ) with the appropriate sequence form triplex structures with duplex DNA as documented by FRET , SPR and NMR . These RNA-duplex DNA triplexes were not sufficiently stable to withstand gel electrophoresis for detection by EMSA , a known limitation of EMSA[20] , but microRNA with appropriate sequence displaced DNA molecules from DNA-DNA triplexes , as documented by EMSA . The development and refinement of more powerful experimental tools such as FRET , SPR , and NMR have made it possible to identify transient interactions that may occur commonly in cell nuclei , and we used each of these methods to document formation of microRNA-duplex DNA triplexes . Analogous to transient protein-protein and DNA/RNA-protein interactions , transient formation of microRNA-duplex DNA triplexes may have as much biological importance as more stable interactions ( reviewed in [16] ) . Interestingly , helicases capable of unwinding intramolecular DNA triplex structures are known [8 , 21] and it is conceivable that this triplex mediated unwinding is a mechanism by which microRNAs can mediate transcriptional activation . Mutations in the human ChlR1 gene , which encodes a triplex-preferring helicase , result in the genetic disorder Warsaw breakage syndrome , characterized by defects in genome maintenance . Cells that were depleted of ChlR1 had increased triplex DNA content and double-stranded breaks[22] . Triplex Structures may promote genome instability by stalling replication forks at ( GAA ) n repeats and inhibiting replication of DNA . Friedreichs ataxia , the most common form of ataxia in humans , is caused by the expansion of a ( GAA ) n repeat in intron 1 of the Frataxin gene , which in turn results in transcriptional silencing , presumably because of the triplex-forming potential of the ( GAA ) n repeat[23] . This suggests that , not only may the formation of DNA triplexes be a well-conserved and essential mechanism to regulate gene transcription , but that stable or prolonged triplex formation may have undesirable consequences . Therefore , it is likely that organisms would have multiple mechanisms to destabilize DNA triplexes . Besides the expression of triplex-specific helicases and potentially other ways to disrupt triplexes , the relatively weak or transient binding of microRNAs to target sites in the genome may constitute another mechanism against DNA-DNA triplex formation . Indeed , our EMSA experiments document that microRNA can disrupt DNA-DNA triplexes in a sequence specific manner , and results from both EMSA and NMR indicate that microRNA-duplex DNA triplexes are relatively transient . The binding to duplex DNA by other microRNAs , and characterization of the effects on gene transcription and downstream phenotypic consequences , merit further study to determine the biological function of such microRNA-DNA interactions . We have shown that DNA sequences that favor microRNA-DNA triplex formation exist throughout the genome of humans and numerous other species . While microRNA-mRNA binding site searches has been done previously [24] , methods presented here represent a novel technique for assessing microRNA-DNA binding through Hoogsteen and reverse Hoogsteen interactions . The Trident algorithm resembles microRNA-mRNA binding site algorithms ( e . g . miRanda ) in its search of binding site pairs , however the algorithm has adapted in it rules for base pair binding . Trident binding rules assign a thermodynamically determined energy to C:G and U:A pairs or G:G and A:A triplex pairs when searching for Hoogsteen and Reverse Hoogsteen binding , respectively . Thermodynamic energies reported by Trident and miRanda differ as well . First order free energy calculations were performed on each possible base pair permutation in a constrained microRNA-DNA triplex and these pair-wise interaction energies are the sum for each base pair in the triplex . Additionally , Trident does not add weighting to base pairs in seed sites due to the symmetric nature of the duplex DNA , microRNA interaction . In addition to the binding site search algorithm , Trident provides a toolkit for analyzing potential triplex structures . Binding site heuristic is developed using a post-processing sequence provided by Trident . For computational efficiency , the entire process was designed and run on a Hadoop cluster . However , each part of the sequence was built to be run as a standalone Python application as well . In addition to statistical analyses , tools are provided to visualize triplex search data , including an interactive web portal ( http://trident . stjude . org ) . While similar websites may be found for microRNA-mRNA binding sites , Trident goes beyond search . Using JBrowse [25] , users can interactively view genome , Trident and gene data , which are all tied to database records . Notably , we have validated the physical interaction of microRNA and duplex DNA using four separate physical methods of triplex detection ( FRET , SPR , EMSA , and NMR ) and used molecular dynamics to model the interaction . These physical interactions are buttressed by empirical measurement of microRNA and mRNA correlations in two separate cohorts of patients with ALL , revealing a marked enrichment ( 3 . 3 fold ) for grade 1 Trident binding near genes whose expression is positively correlated with expression of microRNAs . In conclusion , although intermolecular DNA triplex structures have been detected in cell nuclei , suggesting their possible involvement in gene regulation [26 , 27] , our study provides direct physical evidence of heterotriplex formation involving microRNA and duplex DNA . Moreover , these triplexes involve microRNAs that are either purine or pyrimidine rich ( >75% ) and bind to specific targeted sequences in duplex DNA . We also show that microRNAs that are predicted to form sequence specific triplexes with duplex DNA are enriched for those that are positively correlated with mRNA transcript levels of the targeted genes ( p<2 . 2 x 10−16 ) . The molecular action of these heterotriplexes may include the inducement of conformational changes in the immediately surrounding DNA , including a slight unwinding [28] , a potential mechanism for promoting transcription . Alternatively , triplex specific binding proteins could conceivably alter the topography of gene promoter regions such that transcription factors are able to bind [29] . Our findings provide a platform for discovery of new functions of microRNA in both disease and non-disease states .
Written informed consent was obtained from parents/guardians and assent from patients , as appropriate . The research and use of these samples were approved by the institutional review board at St . Jude Children’s Research Hospital . Total RNA was extracted with TriReagent ( Molecular Research Center , Inc . , Cincinnati , OH ) from cryopreserved mononuclear cell suspensions from patient bone marrow aspirates obtained at diagnosis . All gene expression microarrays were performed by the St . Jude Children’s Research Hospital , Hartwell Center for Bioinformatics & Biotechnology . High-quality RNA was hybridized to the HG-U133A ( GPL96 ) or HGU133 Plus 2 . 0 ( GPL570 ) oligonucleotide microarrays in accordance with the manufacturer’s protocol ( Affymetrix , Santa Clara , CA ) . These microarrays contain 22 , 283 or 54 , 675 gene probe sets , representing approximately 18 , 400 or 47 , 400 human transcripts , respectively . Gene expression data were MAS5 [30] processed using the affy [31] Bioconductor [32] R-project package or using Affymetrix Microarray Suite version 5 . 0 [33 , 34] as previously described [35] . The gene expression data are available via http://trident . stjude . org and http://www . stjuderesearch . org/evans/ . Total RNA was extracted with TriReagent ( Molecular Research Center , Inc . , Cincinnati , OH ) from cryopreserved mononuclear cell suspensions from patient bone marrow aspirates obtained at diagnosis . All microRNA expression microarrays were performed by the St . Jude Children’s Research Hospital , Hartwell Center for Bioinformatics & Biotechnology . High-quality RNA was hybridized to miRCURY LNA 10 . 0 generated from ready to spot probe sets or preprinted 5th generation miRCURY LNA microRNA microarrays in accordance with the manufacturer’s protocol ( Exiqon , Woburn , MA ) . Background subtracted minimum translated data were log2 transformed and then quantile normalized prior to statistical analysis . The microRNA expression data are available for download via http://trident . stjude . org , and http://www . stjuderesearch . org/evans/ . First order free energy calculations were performed on each possible base pair permutation in a constrained microRNA-DNA triplex and these pair-wise interaction energies are then summed for each base pair in the triplex . Derived binding energies are listing in the supplemental material . Restricted geometry optimizations were done via Gaussian03 using B3LYP/6-31g ( p , d ) to obtain the interaction energies of the base pairs . The model systems were constrained such that the nucleic acid ring systems remained co-planar to account for the steric hindrance that would be present in the experimental environment . Solvent effects were modeled using the PCM method [36] with water as the solvent . Differences between the isolated RNA and DNA components were taken to determine the interaction energies on a pairwise basis . Duplex DNA and RNA were manufactured by Integrated DNA Technologies ( Coralville , Iowa ) . Duplex DNA strand 1 ( sense ) : 5’-CTGCTAGCTACTGGGGGAAGAAGAGGGGGCAGAGCTGCTAGCTACT-3’; strand 2 ( antisense ) : 5’-AGTAGCTAGCAGCTCTGCCCCCTCTTCTTCCCCCAGTAGCTAGCAG-3’; synthesized hsa-miR-483-5p: 5’-AAGACGGGAGGAAAGAAGGGAG-3’ . SPR experiments were conducted at 25°C using a Biacore 3000 optical biosensor ( GE Healthcare ) . Streptavidin ( Thermo Scientific ) was covalently immobilized on a polycarboxylate hydrogel-coated gold surface ( HC200m chip; Xantec Bioanalytics ) using routine amine coupling chemistry in immobilization buffer ( 10 mM HEPES pH 7 . 4 , 150 mM NaCl , 0 . 005% Tween20 ) . Carboxyl groups on the hydrogel were activated with N-ethyl-N’- ( 3-dimethylaminopropyl ) carbodiimide ( EDC ) and N-hydroxysuccinimide ( NHS ) , and streptavidin was injected in 10 mM sodium acetate pH 4 . 5 until immobilization levels of 6000 RU were achieved . Remaining active sites were blocked by reaction with ethanolamine . Nucleic acid oligomers were dissolved in TE buffer ( 10 mM Tris pH 8 . 0 , 1 mM EDTA ) and diluted in binding buffer ( 10 mM Tris pH 8 . 0 , 100 mM NaCl , 10 mM MgCl2 , 0 . 02% Tween20 ) before injection over the chip . Biotinylated single stranded RNAs were injected over the streptavidin surfaces until ~30 RU were captured . For manual test injections , data are shown as single-referenced sensorgrams of 20-fold dilutions of the DNA stocks ( final concentrations ~2–5 μM ) injected over the RNA surfaces . For the kinetic analysis , duplex DNA was prepared as a 2-fold dilution series starting at 20 nM and was injected in triplicate at each concentration at a flow rate of 75 μL/min . The chip was regenerated between cycles with a 20 second injection of 1 mM NaOH + 1 M NaCl . The data were processed , double-referenced and globally fit to a 1:1 binding model [37] using the software package Scrubber2 ( version 2 . 0c , BioLogic Software ) . The equilibrium affinity constant ( KD ) was calculated as the quotient of the kinetic rate constants ( kd/ka ) . Duplex DNA and RNA were manufactured by Integrated DNA Technologies ( Coralville , Iowa ) . Duplex DNA strand 1 ( sense ) : 5’-CTGCTAGCTACTGGGGGAAGAAGAGGGGGCAGAGCTGCTAGCTACT-3’; strand 2 ( antisense ) : 5’-AGTAGCTAGCAGCTCTGCCCCCTCTTCTTCCCCCAGTAGCTAGCAG-3’; synthesized hsa-miR-483-5p with a 3’ ROX label: 5’-AAGACGGGAGGAAAGAAGGGAG-ROX-3’ . SYBR Green II ( Life Technologies , Grand Island , NY ) was used as the intercalating dye for duplex DNA . Reaction mixes were plated into 384 well black flat bottom plates and read using a Synergy H4 Hybrid Reader ( Biotek , Winooski , VT ) withGen5 software . SYBR Green II is excited , and an increase in ROX emission is measured to detect binding . Reactions were carried out at physiological pH and temperature . DNA and RNA oligos were manufactured by Integrated DNA Technologies ( Coralville , Iowa ) . A stock solution of 10 μM ROX-labeled hairpin duplex DNA ( ROX-5’-TGGGGGAAGAAGAGGGGGCAGAGATTTTTCTCTGCCCCCTCTTCTTCCCCCA-3’ ) was prepared in 10 mM Tris pH 7 . 4 , heated at 95°C for 5 minutes to fully denature , followed by annealing of the 24-nucleotide sense and antisense regions ( cooling to 22°C at a rate of 0 . 1°C/sec ) . Stock solutions of 200–1000 μM triplex forming oligos ( TFOs ) were prepared in nuclease-free distilled water . The following RNA and DNA TFOs were tested for binding to the duplex DNA: hsa-mIR-483-p ( 5’-AAGACGGGAGGAAAGAAGGGAG-3’ with 16 favorable Hoogsteen bonds ) , Hoogsteen bond-optimized hsa-mIR-483-p ( 483-opti , 5’-GAGACGGGGGAGAAGAAGGGGG-3’ with 21 favorable Hoogsteen hydrogen bonds ) , scrambled microRNA ( 483-scramble , 5’-GGAAGGGCAGGGAGGGGGAAGA-3’ with 10 favorable Hoogsteen bonds ) , truncated hsa-mIR-483-p ( L-11nt-opti , 5’-GAAGAAGGGGG-3’ and R-11nt-opti , 5’-GAGACGGGGGA-3’ , with 11 and 10 favorable Hoogsteen bonds , respectively ) . Binding reactions contained 0 . 1 μM ROX-labeled hairpin duplex DNA , in presence or absence of 5 μM DNA or RNA TFO in 1x binding buffer ( 10 mM Tris pH 7 . 4 , 125 mM NaCl , 6 mM MgCl2 , 0 . 1 mM Spermine ) in a volume of 10 μl; incubated at 22°C for 3 hrs . In competition assays between DNA-TFO and RNA-TFO for binding to duplex DNA , to mixtures of 0 . 1 μM ROX-labeled hairpin duplex DNA and 5 μM 483-opti DNA-TFO , increasing amounts ( 30 , 60 , 150 μM ) of RNA-TFOs were added , and incubated as mentioned above . Reactions were supplemented with 2 μl 6x Gel Loading Solution Type I ( Sigma-Aldrich , Saint Louis , Missouri ) and analyzed by electrophoresis at 50 V on 20% native acrylamide mini gels ( 19:1 acrylamide/bisacrylamide ) in 1xTBE , 125 mM NaCl , 8 mM MgCl2; at 4°C for 16–24 hrs . After electrophoresis the gels were imaged on an Odyssey imager at 600 nm , and duplex and triplex signals were quantified using Image Studio Software ( Li-Cor Biosciences-Biotechnology , Lincoln , Nebraska ) . An algorithm ( ‘Trident’ ) to identify microRNA , genomic DNA binding sites was developed in C and several post-processing pipelines were created ( see Supplement ) . Extending the techniques developed by Betel et al . [38] , the Trident algorithm takes known microRNA transcripts and searches genomic DNA for potential binding sites . MicroRNA sequences were obtained from mirbase http://www . mirbase . org/ version 19 . Genomic DNA sequences for fifteen species ( shown in Fig 1 ) were obtained from National Center for Biotechnology Information , U . S . National Library of Medicine ( NCBI/NLM ) via anonymous file transfer protocol ( FTP ) . Trident performs a search of microRNA—DNA triplex forming sites by assigning both a heuristic score and base pair binding energy to each possible alignment of microRNA and DNA strands . For each alignment location , Trident calculates energy and score values for Direct and Indirect Hoogsteen and Reverse Hoogsteen binding types . If heuristic score and energy exceed specified thresholds , the matching site is reported . Memory usage is directly proportional to the DNA sequence length . Therefore , genome sequences were segmented so that the binding site search could be run in parallel on a compute cluster and an in-house distributed grid [39] . Overlap between each segment is provided to account for the boundary between two neighboring segments . After the binding site search has finished , post-processing is performed on all Trident results to classify relative fitness of matches intra-genomically . To demonstrate relative fitness , heuristic score and energy pairs were classified within each genome based on their relative values . Frequencies for each energy-score pair were analyzed and ranked by percentile , which was used to classify ranks into five match classifications . Linear interpolation was then used to classify arbitrary energy-score pairs . Random DNA sequences were generated by stochastically selecting A , T , G , or C . Although dinucleotide content across genomes may be heterogeneous [40] , we did not adjust the nucleotide frequency for each individual genome , rather used a non-biased frequency of 0 . 25 for each nucleotide for all analyzed genomes . All simulations were performed by AMBER12 with force field ff10 and generalized Born ( GB ) model . The reverse sequence of the selected microRNA , which has less potential to form favorable reverse Hoogsteen pairs , was also constructed as a negative control . The initial conformation of B form DNA duplex and microRNA were generated by 3DNA . The starting complex structures were constructed by simulated annealing with positional restraints of DNA duplex and NMR distance restraints of Hoogsteen hydrogen bonds . The positional restraints of DNA duplex were then removed and a 10ns MD simulation was performed on each system with Watson-Crick pair and reverse Hoogsteen pair restraints . This was followed by a final 10ns MD production run which was performed after gradually removing all the distance restraints in 3ns for each system . Lyophilized RNA ( hsa-miR-483-5p , 5’-AAGACGGGAGGAAAGAAGGGAG-3’ ) , DNA with the same sequence , and hairpin duplex DNA ( 5’-TGGGGGAAGAAGAGGGGGCAGAGATTTTTCTCTGCCCCCTCTTCTTCCCCCA-3’ ) were purchased from Integrated DNA Technologies ( Coralville , Iowa ) and Life Technologies ( Carlsbad , California ) . Hairpin duplex DNA was prepared by heating the duplex DNA oligo at 95°C for 5 minutes to fully denature , followed by annealing of the 24-nucleotide sense and antisense regions ( cooling to 22°C at a rate of 0 . 1°C/sec ) . DNA and RNA oligos were either HPLC-purified or dialyzed . The nucleic acid sample concentration was 250 μM in 15 mM sodium phosphate , 150 mM KCl In 0 . 5 ml of 90% H2O and 10% D2O ( pH = 7 . 5 ) . NMR experiments were measured on a Bruker 600 MHz spectrometer equipped with a 1H and 13C detect , TCI triple resonance cryogenic probe using standard Bruker pulse programs . 2D [1H , 1H] TOCSY ( Total Correlation Spectroscopy ) spectra were acquired with 2048 X 256 points with 80 transients per increment with 70 ms mixing time at 298 K on free DNA duplex and in complex with RNA ( 1:1 . 5 ) and DNA ( 1:1 ) of the same sequence . All the spectra were processed using Topspin 3 . 2 and were analyzed in CARA [41] .
All data are available for download and browsing via http://trident . stjude . org and http://www . stjuderesearch . org/evans/ . | We provide physical evidence , using NMR , FRET and SPR , that purine or pyrimidine-rich microRNAs can form triplexes with complementary purine-rich sequences of duplex DNA and provide an algorithm ( Trident ) to search genome-wide for potential microRNA double-stranded DNA triplex-forming sites . Using this algorithm we document enrichment of microRNA triplex binding sites in mammalian and non-mammalian genomes . We found in primary leukemia cells from patients a significant over-representation of positively correlated microRNA and mRNA expression for genes containing sequences favoring microRNA-duplex DNA triplex formation , suggesting this as a mechanism by which microRNA may enhance gene transcription . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods",
"Data",
"access"
] | [
"chemical",
"compounds",
"electrophoretic",
"mobility",
"shift",
"assay",
"gene",
"regulation",
"organic",
"compounds",
"purines",
"rna",
"stem-loop",
"structure",
"micrornas",
"dna",
"dna",
"structure",
"molecular",
"biology",
"techniques",
"epigenetics",
"chromatin",
... | 2016 | MicroRNAs Form Triplexes with Double Stranded DNA at Sequence-Specific Binding Sites; a Eukaryotic Mechanism via which microRNAs Could Directly Alter Gene Expression |
The development of sensory receptive fields has been modeled in the past by a variety of models including normative models such as sparse coding or independent component analysis and bottom-up models such as spike-timing dependent plasticity or the Bienenstock-Cooper-Munro model of synaptic plasticity . Here we show that the above variety of approaches can all be unified into a single common principle , namely nonlinear Hebbian learning . When nonlinear Hebbian learning is applied to natural images , receptive field shapes were strongly constrained by the input statistics and preprocessing , but exhibited only modest variation across different choices of nonlinearities in neuron models or synaptic plasticity rules . Neither overcompleteness nor sparse network activity are necessary for the development of localized receptive fields . The analysis of alternative sensory modalities such as auditory models or V2 development lead to the same conclusions . In all examples , receptive fields can be predicted a priori by reformulating an abstract model as nonlinear Hebbian learning . Thus nonlinear Hebbian learning and natural statistics can account for many aspects of receptive field formation across models and sensory modalities .
Neurons in sensory areas of the cortex are optimally driven by stimuli with characteristic features that define the receptive field of the cell . While receptive fields of simple cells in primary visual cortex ( V1 ) are localized in visual space and sensitive to the orientation of light contrast [1] , those of auditory neurons are sensitive to specific time-frequency patterns in sounds [2] . The concept of a receptive field is also useful when studying higher-order sensory areas , for instance when analyzing the degree of selectivity and invariance of neurons to stimulus properties [3 , 4] . The characteristic receptive fields of simple cells in V1 have been related to statistical properties of natural images [5] . These findings inspired various models , based on principles as diverse as sparse sensory representations [6] , optimal information transmission [7] , or synaptic plasticity [8] . Several studies highlighted possible connections between biological and normative justifications of sensory receptive fields [9 , 10 , 11 , 12] , not only in V1 , but also in other sensory areas [13] , such as auditory [14 , 15] and secondary visual cortex ( V2 ) [16] . Since disparate models appear to achieve similar results , the question arises whether there exists a general underlying concept in unsupervised learning models [15 , 17] . Here we show that the principle of nonlinear Hebbian learning is sufficient for receptive field development under rather general conditions . The nonlinearity is defined by the neuron’s f-I curve combined with the nonlinearity of the plasticity function . The outcome of such nonlinear learning is equivalent to projection pursuit [18 , 19 , 20] , which focuses on features with non-trivial statistical structure , and therefore links receptive field development to optimality principles . Here we unify and broaden the above concepts and show that plastic neural networks , sparse coding models and independent component analysis can all be reformulated as nonlinear Hebbian learning . For natural images as sensory input , we find that a broad class of nonlinear Hebbian rules lead to orientation selective receptive fields , and explain how seemingly disparate approaches may lead to similar receptive fields . The theory predicts the diversity of receptive field shapes obtained in simulations for several different families of nonlinearities . The robustness to model assumptions also applies to alternative sensory modalities , implying that the statistical properties of the input strongly constrain the type of receptive fields that can be learned . Since the conclusions are robust to specific properties of neurons and plasticity mechanisms , our results support the idea that synaptic plasticity can be interpreted as nonlinear Hebbian learning , implementing a statistical optimization of the neuron’s receptive field properties .
In classic rate models of sensory development [21 , 8 , 6] , a first layer of neurons , representing the sensory input x , is connected to a downstream neuron with activity y , through synaptic connections with weights w ( Fig 1a ) . The response to a specific input is y = g ( wT x ) , where g is the frequency-current ( f-I ) curve . In most models of Hebbian plasticity [22 , 23] , synaptic changes Δw of the connection weights depend on pre- and post-synaptic activity , with a linear dependence on the pre-synaptic and a nonlinear dependence on the post-synaptic activity , Δw ∝ x h ( y ) , in accordance with models of pairing experiments [24 , 10] . The learning dynamics arise from a combination of the neuronal f-I curve y = g ( wTx ) and the Hebbian plasticity function Δw ∝ x h ( y ) : Δw∝xh ( g ( wTx ) ) = xf ( wTx ) ( 1 ) where we define the effective Hebbian nonlinearity f ≔ h ∘ g as the composition of the nonlinearity in the plasticity rule and the neuron’s f-I curve . In an experimental setting , the pre-synaptic activity x is determined by the set of sensory stimuli ( influenced by , e . g . , the rearing conditions during sensory development [25] ) . Therefore , the evolution of synaptic strength , Eq 1 , is determined by the effective nonlinearity f and the statistics of the input x . Many existing models can be formulated in the framework of Eq 1 . For instance , in a classic study of simple-cell formation [8] , the Bienenstock-Cooper-Munro ( BCM ) model [22] has a quadratic plasticity nonlinearity , hθ ( y ) = y ( y − θ ) , with a variable plasticity threshold θ = 〈y2〉 , and a sigmoidal f-I curve , y = σ ( wT x ) . Since the threshold θ is adapted on a time scale sufficiently slow to sample the statistics of 〈y2〉 [28] , and on a time scale faster than the learning dynamics [29] , we may consider it as fixed , and the dynamics are well described by nonlinear Hebbian learning , Δw ∝ x hθ ( σ ( wTx ) ) , with a nonlinearity modulated by θ . More realistic cortical networks have dynamical properties which are not accounted for by rate models . By analyzing state-of-the-art models of cortical neurons and synaptic plasticity , we inspected whether plastic spiking networks can be reduced to nonlinear Hebbian learning . We considered a generalized leaky integrate-and-fire model ( GIF ) , which includes adaptation , stochastic firing and predicts experimental spikes with high accuracy [26] , and we approximate its f-I curve by a linear rectifier , g ( u ) = a ( u − θ ) + , with slope a and threshold θ ( Fig 1b ) . As a phenomenological model of synaptic plasticity grounded on experimental data [27] , we implemented triplet spike-timing dependent plasticity ( STDP ) [24] . In this STDP model , the dependence of long-term potentiation ( LTP ) upon two post-synaptic spikes induces in the corresponding rate model a quadratic dependence on the post-synaptic rate , while long-term depression ( LTD ) is linear . The resulting rate plasticity [24] is h ( y ) = y2 − by , with an LTD factor b ( post-synaptic activity threshold separating LTD from LTP , Fig 1c ) , similar to the classic BCM model [22 , 8] . Composing the f-I curve of the GIF with the h ( y ) for the triplet plasticity model , we have an approximation of the effective learning nonlinearity f = h ∘ g in cortical spiking neurons ( Fig 1d ) , that can be described as a quadratic rectifier , with LTD threshold given by θ1 = θ and LTP threshold given by θ2 = θ+b/a . Interestingly , the f-I slope a and LTD factor b are redundant parameters of the learning dynamics: only their ratio counts in nonlinear Hebbian plasticity . Metaplasticity can control the LTD factor [24 , 30] , thus regulating the LTP threshold . If one considers a linear STDP model [31 , 32] instead of the triplet STDP [24] , the plasticity curve is linear [23] , as in standard Hebbian learning , and the effective nonlinearity is shaped by the properties of the f-I curve ( Fig 2a ) . In the following we consider these rate approximations of STDP and analyze the developmental properties of spiking neurons through their effective nonlinearities . Beyond phenomenological modeling , normative principles that explain receptive fields development have been one of the goals of theoretical neuroscience [33] . Sparse coding [6] starts from the assumptions that V1 aims at maximizing the sparseness of the activity in the sensory representation , and became a well-known normative model to develop orientation selective receptive fields [9 , 12 , 13] . We demonstrate that the algorithm implemented in the sparse coding model is in fact a particular example of nonlinear Hebbian learning . The sparse coding model aims at minimizing an input reconstruction error E = 1 2 | | x − W y | | 2 + λ S ( y ) , under a sparsity constraint S with relative importance λ > 0 . For K hidden neurons yj , such a model implicitly assumes that the vector wj of feed-forward weights onto neuron j are mirrored by hypothetical “reconstruction weights” , W = [w1 … wK] . The resulting encoding algorithm can be recast as a neural model [34] , if neurons are embedded in a feedforward model with lateral inhibition , y = g ( wTx − vTy ) , where v are inhibitory recurrent synaptic connections ( see Methods ) . In the case of a single output neuron , its firing rate is simply y = g ( wTx ) . The nonlinearity g of the f-I curve is threshold-like , and determined by the choice of the sparsity constraint [34] , such as the Cauchy , L0 , or L1 constraints ( Fig 2a , see Methods ) . If weights are updated through gradient descent so as to minimize E , the resulting plasticity rule is Oja’s learning rule [35] , Δw ∝ x y − w y2 . The second term −w y2 has a multiplicative effect on the strength of synapses projecting onto the same neuron ( weight rescaling ) , but does not affect the receptive field shape , whereas the first term x y drives feature selectivity and receptive field formation . Together , these derivations imply that the one-unit sparse coding algorithm can be implemented by an effective nonlinear Hebbian rule combined with weight normalization . Although the plasticity mechanism is linear , Δw ∝ x y , a nonlinearity arises from the f-I curve , y = g ( wTx ) , so that the effective plasticity is Δ w ∝ x g ( w T x ) ( 2 ) This analysis reveals an equivalence between sparse coding models and neural networks with linear plasticity mechanisms , where the sparsity constraint is determined by the f-I curve g . While Oja’s rule is commonly associated with principal component analysis ( PCA ) , developing connections that project the input in the direction of largest variance [35] , this relation is only valid for linear neurons . When nonlinear neurons are considered , Oja’s rule is also sensitive to higher-order statistics , as analyzed below . Similarly , algorithms performing independent component analysis ( ICA ) , a model class closely related to sparse coding , also perform effective nonlinear Hebbian learning , albeit inversely , with linear neurons and a nonlinear plasticity rule [36] . For variants of ICA based on information maximization [7] or kurtosis [36] different nonlinearities arise ( Fig 2a ) , but Eq 2 applies equally well . Hence , various instantiations of sparse coding and ICA models not only relate to each other in their normative assumptions [37] , but when implemented as iterative gradient update rules , they all employ nonlinear Hebbian learning . Since the models described above can be implemented by similar plasticity rules , we hypothesized nonlinear Hebbian learning to be a general principle that explains the development of receptive field selectivity . Nonlinear Hebbian learning with an effective nonlinearity f is linked to an optimization principle with a function F ( u ) = ∫ 0 u f ( u ′ ) d u ′ [19 , 20] . For an input ensemble x , optimality is achieved by weights w ˜ that maximize 〈F ( w ˜ T x ) 〉 , where angular brackets denote the average over the input statistics . Nonlinear Hebbian learning is a stochastic gradient ascent implementation of this optimization process , called projection pursuit [18 , 19 , 20]: w ˜ = m a x w 〈 F ( w T x ) 〉 ⇒ Δ w ∝ x f ( w T x ) ( 3 ) Motivated by results from ICA theory [38] and statistical properties of whitened natural images [5] , we selected diverse Hebbian nonlinearities f ( Fig 2a ) and calculated the corresponding optimization value 〈F ( wTx ) 〉 for different features of interest that we consider as candidate RF shapes , with a whitened ensemble of patches extracted from natural images as input ( see Methods ) . These include a random connectivity pattern , a non-local oriented edge ( as in principal components of natural images ) and localized oriented edges ( as in cat and monkey simple cells in the visual cortex ) , shown in Fig 2b . The relative value of 〈F ( wTx ) 〉 between one feature and another was remarkably consistent across various choices of the nonlinearity f , with localized orientation-selective receptive fields as maxima ( Fig 2b ) . Furthermore , we also searched for the maxima through gradient ascent , so as to confirm that the maxima are orientation selective ( Fig 2c , left ) . Our results indicate that receptive field development of simple cells is mainly governed by the statistical properties of natural images , while robust to specific model assumptions . The relevant property of natural image statistics is that the distribution of a feature derived from typical localized oriented patterns has high kurtosis [5 , 6 , 39] . Thus to establish a quantitative measure whether a nonlinearity is suitable for feature learning , we define a selectivity index ( SI ) , which measures the relative value of 〈F ( . ) 〉 between a variable l with a Laplacian distribution and a variable g with Gaussian distribution [38]: SI = ( 〈F ( l ) 〉 − 〈F ( g ) 〉 ) /σF ( see Methods ) . The Laplacian variable has higher kurtosis than the Gaussian variable , serving as a prototype of a kurtotic distribution . Since values obtained by filtering natural images with localized oriented patterns have a distribution with longer tails than other patterns [5] , as does the Laplacian variable compared to the Gaussian , positive values SI > 0 indicate good candidate functions for learning simple cell-like receptive fields from natural images . We find that each model has an appropriate parameter range where SI > 0 ( Fig 3 ) . For example the quadratic rectifier nonlinearity needs an LTP threshold θ2 below some critical level , so as to be useful for feature learning ( Fig 3a ) . A sigmoidal function with threshold at zero has negative SI , but a negative sigmoid , as used in ICA studies [7] , has SI > 0 . More generally , whenever an effective nonlinearity f is not suited for feature learning , its opposite − f should be , since its SI will have the opposite sign ( Fig 2c ) . This implies that , in general , half of the function space could be suitable for feature learning [38] , i . e . it finds weights w such that the distribution of the feature wTx has a long tail , indicating high kurtosis ( “kurtotic feature” ) . The other half of the function space learns the least kurtotic features ( e . g . random connectivity patterns for natural images , Fig 2b and 2c ) . This universality strongly constrains the possible shape of receptive fields that may arise during development for a given input dataset . For whitened natural images , a learnable receptive field is in general either a localized edge detector or a non-localized random connectivity pattern . While there is no simple description for the class of suitable functions , we may gain some intuition by considering the class of rectified power functions , F ( u ) = u + r , r ∈ ℜ+ . In the case of powers r > 2 , the selectivity index is positive . As a consequence , any supra-linear nonlinearity f ( u ) = u + p with p > 1 should be suitable for feature learning . In Table 1 , we include the appropriate parameter range for various effective nonlinearities . An important special case is an effective linear curve , f ( u ) = u , which arises when both f-I and plasticity curves are linear [21] . Because the linear model maximizes variance 〈 ( wTx ) 2〉 , it can perform principal component analysis [35] , but does not have any feature selectivity on whitened input datasets , where variance is constant ( Fig 2c ) . Symmetric effective nonlinearities , f ( u ) = f ( −u ) , are also exceptions , since their corresponding optimization functions are asymmetric , F ( u ) = −F ( −u ) , so that for datasets with symmetric statistical distributions , P ( x ) = P ( −x ) , the optimization value will be zero , 〈Fasym . ( wTxsym . ) 〉 = 0 . As natural images are not completely symmetric , localized receptive fields do develop , though without orientation selectivity , as illustrated by a cosine function and a symmetric piece-wise linear function as effective nonlinearities ( Fig 2c , bottom rows ) . Sensory neurons display a variety of receptive field shapes [40] , and modeling efforts [41 , 9 , 12] have attempted to understand the properties that give rise to the specific receptive fields seen in experiments . We show here that the shape diversity of a model can be predicted by our projection pursuit analysis , and is primarily determined by the statistics of input representation , while relatively robust to the specific effective nonlinearity . We studied a model with multiple neurons in the second layer , which compete with each other for the representation of specific features of the input . Each neuron had a piece-wise linear f-I curve and a quadratic rectifier plasticity function ( see Methods ) and projected inhibitory connections v onto all others . These inhibitory connections are learned by anti-Hebbian plasticity and enforce decorrelation of neurons , so that receptive fields represent different positions , orientations and shapes [42 , 43 , 44] . For 50 neurons , the resulting receptive fields became diversified ( Fig 4a–4c , colored dots ) . In an overcomplete network of 1000 neurons , the diversity further increased ( Fig 4d–4f , colored dots ) . For the analysis of the simulation results , we refined our inspection of optimal oriented receptive fields for natural images by numerical evaluation of the optimality criterion 〈F ( wTx ) 〉 for receptive fields w = wGabor , described as Gabor functions of variable length , width and spatial frequency . For all tested nonlinearities , the optimization function for single-neuron receptive fields varies smoothly with these parameters ( Fig 4 , grey-shaded background ) . The single-neuron optimality landscape was then used to analyze the multi-neuron simulation results . We found that receptive fields are located in the area where the single-neuron optimality criterion is near its maximum , but spread out so as to represent different features of the input ( Fig 4 ) . Thus the map of optimization values , calculated from the theory of effective nonlinearity , enables us to qualitatively predict the shape diversity of receptive fields . Although qualitatively similar , there are differences in the receptive fields developed for each model , such as smaller lengths for the L0 sparse coding model ( Fig 4c ) . While potentially significant , these differences across models may be overwhelmed by differences due to other model properties , such as different network sizes or input representations . This is illustrated by observing that receptive field diversity for a given model differ substantially across network sizes ( Fig 4 ) . We also studied the variation of receptive field position and orientation . For all five nonlinearities considered , the optimization value is equal for different positions of the receptive field centers , confirming the translation invariance in the image statistics , as long as the receptive field is not too close to the border of the anatomically allowed fan-in of synaptic connections ( Fig 5b ) . Also , all nonlinearities reveal the same bias towards the horizontal and vertical orientations ( Fig 5c ) . These optimality predictions are confirmed in single neuron simulations , which lead mostly to either horizontal or vertical orientations , at random positions ( Fig 5d ) . When the network is expanded to 50 neurons , recurrent inhibition forces receptive fields to cover different positions , though excluding border positions , and some neurons have non-cardinal orientations ( Fig 5e ) . With 1000 neurons , receptive fields diversify to many possible combinations of position , orientation and length ( Fig 5f ) . Natural images have non-uniform spectral properties , with higher variance at low spatial frequencies [39] . Since Hebbian learning is sensitive to second-order correlations , in order to learn receptive fields driven by higher-order statistics , most studies pre-whiten the input , making the variance uninformative [36] . While there is evidence that the early sensory pathway induces decorrelation across neurons [45] , it is unlikely for the input to the visual cortex to be perfectly white . To analyze the impact of residual second-order correlations , we simulated nonlinear Hebbian learning with natural image patches that have been only approximately whitened . Instead of estimating the whitening filter from input correlation matrix , we used the preprocessing filter from the original sparse coding studies [6 , 37] , which assumes that natural images possess an ideal power-law energy spectra ( see Methods ) . In Fig 6 , we show the receptive fields learned for non-white inputs through nonlinear Hebbian learning . For networks with few neurons ( Fig 6a and 6b ) , nonlocal receptive fields develop , with shapes similar to the principal components of natural images [6] . It reflects that when second-order correlations are present , these dominate over higher-order statistics , in which case the models we have considered will not reproduce the development of localized oriented filters . However , when considering an overcomplete network with 1000 neurons , smaller receptive fields are learned ( Fig 6c ) . Our optimization framework provides a new perspective on this phenomena . For non-white inputs , second-order correlation dominate the optimization values , making principal components optima . However , when more neurons are added , competition drives the diversification of receptive fields away from the optima , and localized filters with optimization values driven by higher-order statistics can be learned . We also compare the receptive fields developed for different nonlinearities ( Fig 6d–6i ) . Particularly , the quadratic rectifier appears to develop more elongated filters compared to the linear rectifier network , while the L0 sparse coding network develops shorter ones . However , these differences across nonlinearities are minor compared to the difference to the receptive fields for white inputs ( Fig 4 ) or compared to the differences observed across different network sizes . Thus , our results suggest that efforts to model receptive field shapes observed experimentally [41 , 40 , 9 , 12] should pay particular attention to network size and input preprocessing , which may have a greater effect than the properties of the particular model . Nonlinear Hebbian learning is not limited to explaining simple cells in V1 . We investigated if the same learning principles apply to receptive field development in other visual or auditory areas or under different rearing conditions . For auditory neurons [14] , we used segments of speech as input ( Fig 7a ) and observed the development of spectrotemporal receptive fields localized in both frequency and time [2] ( Fig 7d ) . The statistical distribution of input patterns aligned with the learned receptive fields had longer tails than for random or non-local receptive fields , indicating temporal sparsity of responses ( Fig 7d ) . Similar to our simple cell results , the learned receptive fields show higher optimization value for all five effective nonlinearities ( Fig 7g ) . For a study of receptive field development in the secondary visual cortex ( V2 ) [16] , we used natural images and the standard energy model [46] of V1 complex cells to generate input to V2 ( Fig 7b ) . The learned receptive field was selective to a single orientation over neighboring positions , indicating a higher level of translation invariance . When inputs were processed with this receptive field , we found longer tails in the feature distribution than with random features or receptive fields without orientation coherence ( Fig 7e ) , and the learned receptive field had a higher optimization value for all choices of nonlinearity ( Fig 7h ) . Another important constraint for developmental models are characteristic deviations , such as strabismus , caused by abnormal sensory rearing . Under normal binocular rearing conditions , the fan-in of synaptic input from the left and right eyes overlap in visual space ( Fig 7c ) . In this case , binocular receptive fields with similar features for left and right eyes develop . In the strabismic condition , the left and right eyes are not aligned , modeled as binocular rearing with non-overlapping input from each eye ( Fig 7c ) . In this scenario , a monocular simple cell-like receptive field developed ( Fig 7f ) , as observed in experiments and earlier models [28] . The statistical distributions confirm that for disparate inputs the monocular receptive field is more kurtotic than a binocular one , explaining its formation in diverse models [47] ( Fig 7f and 7i ) . Our results demonstrate the generality of the theory across multiple cortical areas . Selecting a relevant feature space for an extensive analysis , as we have done with simple cells and natural images , may not be possible in general . Nonetheless , nonlinear Hebbian learning helps to explain why some features ( and not others ) are learnable in network models [15] .
Earlier studies have already placed developmental models side by side , comparing their normative assumptions , algorithmic implementation or receptive fields developed . Though consistent with their findings , our results lead to revised interpretations and predictions . The similarities between sparse coding and ICA are clear from their normative correspondence [37] . Nevertheless , the additional constraint in ICA , of having at most as many features as inputs , makes it an easier problem to solve , allowing for a range of suitable algorithms [36] . These differ from algorithms derived for sparse coding , in which the inference step is difficult due to overcompleteness . We have shown that regardless of the specific normative assumptions , it is the common implementation of nonlinear Hebbian learning that explains similarities in their learning properties . Since a given normative model may have very different algorithms , as exemplified by the family of ICA algorithms [36] , this result is not trivial , and it has previously not been clear how sparse coding and ICA models related to each other at the algorithmic level . In contrast to the idea that in sparse coding algorithms overcompleteness is required for development of localized oriented edges [37] , we have demonstrated that a sparse coding model with a single neuron is mathematically equivalent to nonlinear Hebbian learning and learns localized filters in a setting that is clearly “undercomplete” . Thus differences observed in receptive field shapes between sparse coding and ICA models [40] are likely due to differences in network size and input preprocessing . For instance , the original sparse coding model [37] applied a preprocessing filter that did not completely whiten the input , leading to larger receptive fields ( Fig 6 ) . Studies that derive spiking models from normative theories often interpret the development of oriented receptive fields as a consequence of its normative assumptions [11 , 12] . In a recent example , a spiking network has been related to the sparse coding model [12] , using neural properties defined ad hoc . Our results suggest that many other choices of neural activations would have given qualitatively similar receptive fields , independent of the sparse coding assumption . While in sparse coding the effective nonlinearity derives from a linear plasticity rule combined with a nonlinear f-I curve , our results indicate that a nonlinear plasticity rule combined with a linear neuron model would give the same outcome . In order to distinguish between different normative assumptions , or particular neural implementations , the observation of “oriented filters” is not sufficient and additional constraints are needed . Similarly receptive shape diversity , another important experimental constraint , should also be considered with care , since it cannot easily distinguish between models either . Studies that confront the receptive field diversity of a model to experimental data [41 , 40 , 9 , 12] should also take into account input preprocessing choices and how the shape changes with an increasing network size , since we have observed that these aspects may have a larger effect on receptive field shape than the particulars of the learning model . Empirical studies of alternative datasets , including abnormal visual rearing [47] , tactile and auditory stimuli [15] , have also observed that different unsupervised learning algorithms lead to comparable receptive fields shapes . Our results offer a plausible theoretical explanation for these findings . Past investigations on nonlinear Hebbian learning [20 , 38] demonstrated that many nonlinearities were capable of solving the cocktail party problem . Since it is a specific toy model , that asks for the unmixing of linearly mixed independent features , it is not clear a priori whether the same conclusions would hold in other settings . We have shown that the results of [20] and [38] generalize in two directions . First , the effective nonlinear Hebbian learning mechanism is also behind other models beyond ICA , such as sparse coding models and plastic spiking networks . Second , the robustness to the choice of nonlinearity is not limited to a toy example , but also holds in multiple real world data . Our approach of identifying generic principles enables us to transfer results from one model , such as orientation selectivity or optimization of higher-order statistics to other models within the general framework . Therefore our insights may contribute to predict the outcome of a variety of developmental models in diverse applications . Many theoretical studies start from normative assumptions [7 , 9 , 11 , 37] , such as a statistical model of the sensory input or a functional objective , and derive neural and synaptic dynamics from them . Our claim of universality of feature learning indicates that details of normative assumptions may be of lower importance . For instance , in sparse coding one assumes features with a specific statistical prior [9 , 37] . After learning , this prior is expected to match the posterior distribution of the neuron’s firing activity [9 , 37] . Nevertheless , we have shown that receptive field learning is largely unaffected by the choice of prior . Thus , one cannot claim that the features were learned because they match the assumed prior distribution , and indeed in general they do not . For a coherent statistical interpretation , one could search for a prior that would match the feature statistics . However , since the outcome of learning is largely unaffected by the choice of prior , such a statistical approach would have limited predictive power . Generally , kurtotic prior assumptions enable feature learning , but the specific priors are not as decisive as one might expect . Because normative approaches have assumptions , such as independence of hidden features , that are not generally satisfied by the data they are applied to , the actual algorithm that is used for optimization becomes more critical than the formal statistical model . The concept of sparseness of neural activity is used with two distinct meanings . The first one is a single-neuron concept and specifically refers to the long-tailed distribution statistics of neural activity , indicating a “kurtotic” distribution . The second notion of sparseness is an ensemble concept and refers to the very low firing rate of neurons , observed in cortical activity [48] , which may arise from lateral competition in overcomplete representations . Overcompleteness of ensembles makes sparse coding different from ICA [37] . We have shown here that competition between multiple neurons is fundamental for receptive field diversity , whereas it is not required for simple cell formation per se . Kurtotic features can be learned even by a single neuron with nonlinear Hebbian learning , and with no restrictions on the sparseness of its firing activity . Recent studies have also questioned normative explanations for V1 receptive fields by highlighting that these models do not accurately capture the statistics of natural images [49 , 50] . The generative models learned for sparse coding or ICA do not generate qualitatively good samples of natural image patches [50] . In particular , the performance in the quantitative criteria that these models are designed to optimize , such as likelihood of the data [50] or higher-order redundancy [49] , is sometimes only marginally better than that of simpler models . Further studies are necessary to elucidate more complex models going beyond the two-layer model considered here . For instance , models of spiking networks learning spatio-temporal patterns have been proposed based on diverse principles such as reward-modulated plasticity [51 , 52] , novelty-like global factors [53 , 54] and temporal correlations [55 , 56] . It would be interesting to investigate if generality principles can also shed light on such models . Furthermore , top-down inputs form a substantial part of the incoming signal to sensory areas [57] and it is unclear how they might affect learning and representation in sensory networks . Multi-layered models of probabilistic inference may provide ways to integrate these aspects under a coherent framework for sensory development [58 , 59 , 60] . Our arguments can be formulated using Marr’s three levels of analysis: the computational level , the algorithmic level and the implementational level [61] . We have argued that the algorithmic level , through nonlinear Hebbian learning , is fundamental in understanding many current models of sensory development , while being consistent with multiple biological implementations and computational goals . Our results show that the models and experimental evidence considered were not sufficient to conclusively discriminate between normative assumptions , indicating indeterminacy at the computational level . Since ultimately one also wants a normative understanding of sensory networks , our results argue for more experimental evidence to be taken into account , requiring more complex models , which in turn shall be described by , or derived from , precise computational objectives , such as probabilistic inference or efficient coding . The concept of nonlinear Hebbian learning also clarifies the interaction of feature selectivity with input preprocessing . Most studies of receptive field development consider pre-whitened inputs , which may be justified by the evidence that the early sensory pathway decorrelates neural activity [62] . However , we have shown that developmental models are highly sensitive to second-order statistics , and even residual correlations will substantially alter receptive field development . When correlations at low spatial frequencies were present in the input images , nonlinear Hebbian learning models learned nonlocal receptive fields . In this case , additional mechanisms become necessary to reproduce the development of localized receptive fields as observed in the visual cortex . One possibility is that the competition in overcomplete networks drives the diversify of receptive fields away from principal components , so that neurons become sensitive to higher-order statistics [6] . Another explanation is that the restriction on the arborization of input connections is responsible for local properties of V1 receptive fields [63] , in which case localization is not related to higher-order statistics . These considerations demonstrate how alternative input preprocessing can radically change the interpretation of developmental studies , and suggests that more attention should be paid to the preprocessing steps performed in modeling studies . Importantly , it highlights the necessity of more investigations on learning models with robustness to second-order correlations . In studies of spiking networks , the input is restricted to positive rates , possibly through an on/off representation , as observed in the LGN [63] . In such alternative representations , trivial receptive fields may develop , such as a single non-zero synapse , and additional mechanisms , such as hard bounds on each synaptic strength , a ≤ wj ≤ b , may be necessary to restrict the optimization space to desirable features [10] . Instead of constraining the synaptic weights , one may implement a synaptic decay as in Oja’s plasticity rule [35] , Δw ∝ x y − w y2 ( see also [64] ) . Because of its multiplicative effect , the decay term does not alter the receptive field , but only scales its strength . Thus , it is equivalent to rescaling the input in the f-I curve , so as to shift it to the appropriate range ( Fig 3 ) . Similar scaling effects arise from f-I changes due to intrinsic plasticity [11 , 30 , 65] or due to the sliding threshold in BCM-like models , where the effective nonlinearity is modulated by the current weights . Since we have shown that receptive field development is robust to the specific nonlinearity , we expect our results in general to remain valid in the presence of such homeostatic mechanisms . The precise relation between nonlinear Hebbian learning , spiking representations and homeostasis in the cortex is an important topic for further studies . The principle of nonlinear Hebbian learning has a direct correspondence to biological neurons and is compatible with a large variety of plasticity mechanisms . It is not uncommon for biological systems to have diverse implementations with comparable functional properties [66] . Different species , or brain areas , could have different neural and plasticity characteristics , and still have similar feature learning properties [67 , 68] . The generality of the results discussed in this paper reveals learning simple cell-like receptive fields from natural images to be much easier than previously thought . It implies that a biological interpretation of models is possible even if some aspects of a model appear simplified or even wrong in some biological aspects . Universality also implies that the study of receptive field development is not sufficient to distinguish between different models . The relation of nonlinear Hebbian learning to projection pursuit endorses the interpretation of cortical plasticity as an optimization process . Under the rate coding assumptions considered here , the crucial property is an effective synaptic change linear in the pre-synaptic rate , and nonlinear in the post-synaptic input . Pairing experiments with random firing and independently varying pre- and post-synaptic rates would be valuable to investigate these properties [27 , 69 , 70] . Altogether , the robustness to details in both input modality and neural implementation suggests nonlinear Hebbian learning as a fundamental principle underlying the development of sensory representations .
A generalized leaky integrate-and-fire neuron [26] was used as spiking model , which includes power-law spike-triggered adaptation and stochastic firing , with parameters [26] fitted to pyramidal neurons . The f-I curve g ( I ) was estimated by injecting step currents and calculating the trial average of the spike count over the first 500 ms . The minimal triplet-STDP model [24] was implemented , in which synaptic changes follow d d t w ( t ) = A + y ( t ) y ¯ + ( t ) x ¯ + ( t ) − A - x ( t ) y ¯ - ( t ) ( 4 ) where y ( t ) and x ( t ) are the post- and pre-synaptic spike trains , respectively: y ( t ) = ∑f δ ( t − tf ) , where tf are the firing times and δ denotes the Dirac δ-function; x ( t ) is a vector with components x i ( t ) = ∑ f δ ( t − t i f ) , where t i f are the firing times of pre-synaptic neuron i; w is a vector comprising the synaptic weights wi connecting a pre-synaptic neuron i to a post-synaptic cell . A+ = 6 . 5 ⋅ 10−3 and A− = 5 . 3 ⋅ 10−3 are constants , and y ¯ + , x ¯ + and y ¯ - are moving averages , implemented by integration ( e . g . τ ∂ y ¯ ∂ t = − y ¯ + y ) , with time scales 114 . 0 ms , 16 . 8 ms and 33 . 7 ms , respectively [24] . For estimating the nonlinearity h ( y ) of the plasticity , pre- and post-synaptic spike trains were generated as Poisson processes , with the pre-synaptic rate set to 20 Hz . A linear rectifier g ( x ) = a ( x − b ) + was fitted to the f-I curve of the spiking neuron model by squared error optimization . Similarly , a quadratic function h ( x ) = a ( x2 − bx ) was fitted to the nonlinearity of the triplet STDP model . The combination of these two fitted functions was plotted as fit for the effective nonlinearity f ( x ) = h ( g ( x ) ) . A sparse coding model , with K neurons y1 , … , yK , has a nonlinear Hebbian learning formulation . The sparse coding model minimizes a least square reconstruction error between the vector of inputs x and the reconstruction vector Wy , where W = [w1 …wK] , and y = ( y1 , … , yK ) is the vector of neuronal activities , with yj ≥ 0 for 1 ≤ j ≤ K . The total error E combines a sparsity constraint S with weight λ and the reconstruction error , E = 1 2 | | x − W y | | 2 + λ ∑ S ( y k ) . E has to be minimal , averaged across all input samples , under the constraint yj ≥ 0 for all j . The minimization problem is solved by a two-step procedure . In the first step , for each input sample , one minimizes E with respect to all hidden units yj d d y j E = 0 ⇔ w j T ( x − W y ) − λ S ′ ( y j ) = 0 ⇔ w j T x − ∑ k ≠ j ( w j T w k ) y k − | | w j | | 2 y j − λ S ′ ( y j ) = 0 ⇔ y j + λ S ′ ( y j ) = w j T x − ∑ k ≠ j ( w j T w k ) y k ⇔ y j = g ( w j T x − ∑ k ≠ j v j k y k ) ( 5 ) where we constrained the vector wj of synapses projecting onto unit yj by ||wj||2 = 1 , defined the activation function g ( . ) = T−1 ( . ) , the inverse of T ( y ) = ( y+λS′ ( y ) ) , and defined recurrent synaptic weights v j k = w j T w k . For each input sample x , this equation shall be iterated until convergence . The equation can be interpreted as a recurrent neural network , where each neuron has an activation function g , and the input is given by the sum of the feedforward drive w j T x and a recurrent inhibition term −∑k ≠ j vjk yk . To avoid instability , we implement a smooth membrane potential uj , which has the same convergence point [34] τ u d d t u j ( t ) = - u j ( t ) + ( w j T x − ∑ k ≠ j v j k y k ( t ) ) y j ( t ) = g ( u j ( t ) ) ( 6 ) initialized with uj ( t ) = 0 . In the second step , we optimize the weights wj , considering the activations yj obtained in the previous step . Our derivation follows the approach of the original sparse coding study [6] , which is related to the Expectation-Maximization ( EM ) algorithm , in which at this stage the latent variables ( here the activations y ) are treated as constants , so that d y d w j = 0 , and , in particular , d d w j S ( y ) = 0 . We obtain a standard gradient descent implementation of the least square regression optimization , leading to a learning rule Δ w j ∝ d d w j E = ( x − W T y ) y j = x y j − w j y j 2 − ∑ k ≠ j w k y k y j The decay term w j y j 2 has no effect , since the norm is constrained to ||wj|| = 1 at each step . For a single unit y , the model simplifies to a nonlinear Hebbian formulation , Δ w ∝ x g ( w j T x ) . For multiple units , it can be interpreted as projection pursuit on an effective input , not yet represented by other neurons , x j ˜ = x − ∑ k ≠ j w k y k , which simplifies to Δ w j ∝ x ˜ j g ( w j T x j ˜ ) . There are two non-local terms that need to be implemented by local mechanisms so as to be biologically plausible . First , the recurrent weights depend on the overlap between receptive fields , w j T w k , which is non-local . The sparse coding model assumes independent hidden neurons , which implies that after learning neurons should be pair-wise uncorrelated , cov ( yj , yk ) = 0 . As an aside we note that the choice v j k = w j T w k does not automatically guarantee decorrelation . Decorrelation may be enforced through plastic lateral connections , following an anti-Hebbian rule [42 , 12] , Δvjk ∝ ( yj−〈yj〉 ) ⋅ yk , where 〈yj〉 is a moving average ( we use τ = 1000 input samples ) . Thus by substituting fixed recurrent connections by anti-Hebbian plasticity , convergence Δvjk = 0 implies cov ( yj , yk ) = 0 . While this implementation does not guarantee v j k = w j T w k after convergence , neither does v j k = w j T w k guarantee decorrelation cov ( yj , yk ) = 0 , it does lead to optimal decorrelation , which is the basis of the normative assumption . Additionally we constrain vjk ≥ 0 to satisfy Dale’s law . Although some weights would converge to negative values otherwise , most neuron pairs have correlated receptive fields , and thus positive recurrent weights . Second , we ignore the non-local term ∑k ≠ j wk yk yj in the update rule . Although this approximation is not theoretically justified , we observed in simulations that receptive fields do not qualitatively differ when this term is removed . The resulting Hebbian formulation can be summarized as y j = g ( w j T x − ∑ k ≠ j v j k y k ) Δ w j ∝ x y j Δ v j k ∝ ( y j − 〈 y j 〉 ) · y k ( 7 ) This derivation unifies previous results on the biological implementation of sparse coding: the relation of the sparseness constraint to a specific activation function [34] , the derivation of a Hebbian learning rule from quadratic error minimization [35] , and the possibility of approximating lateral interaction terms by learned lateral inhibition [42 , 12] . The optimization value for a given effective nonlinearity f , synaptic weights w , and input samples x , is given by R = 〈F ( wTx ) 〉 , where F ( u ) = ∫ 0 u f ( u ′ ) d u ′ and angular brackets indicate the ensemble average over x . Relative optimization values in Figs 2b and 5 were normalized to [0 , 1] , relative to the minimum and maximum values among the considered choice of features w , R* = ( R − Rmin ) / ( Rmax − Rmin ) . The selectivity index of a nonlinearity f is defined as SI = ( 〈F ( l ) 〉 − 〈F ( g ) 〉 ) /σF , where l and g are Laplacian and Gaussian variables respectively , normalized to unit variance . σ F = σ F ( l ) σ F ( g ) is a normalization factor , with σ F ( . ) = 〈 F ( . ) 2 〉 . The selectivity of an effective nonlinearity f is not altered by multiplicative scaling , f ˜ ( u ) = α f ( u ) , neither by additive constants when the input distribution is symmetric , f ˜ ( u ) = α f ( u ) + β . The effective nonlinearities in Fig 2 included the linear rectifier f ( u ) = { 0 , i f u < θ u − θ , i f u ≥ θ , the quadratic rectifier f ( u ) = { 0 , i f u < θ ( u − θ ) ( u − θ − b ) , i f u ≥ θ , the L0 sparse coding nonlinearity f ( u ) = { 0 , i f u < λ u , i f u ≥ λ , the Cauchy sparse coding nonlinearity f = T − 1 , where T ( y ) = { 0 , i f y < 0 y + 2 λ y / ( 1 + y 2 ) , i f y ≥ 0 , the negative sigmoid f ( u ) = 1 − 2/ ( 1 + e − 2u ) , a polynomial function f ( u ) = u3 , trigonometric functions sin ( u ) and cos ( u ) , a symmetric piece-wise linear function f ( u ) = { 0 , i f | u | < θ | u | − θ , i f | u | ≥ θ , as well as , for comparison , a linear function f ( u ) = u . Natural image patches ( 16 by 16 pixel windows ) were sampled from a standard dataset [6] ( 106 patches ) . Patches were randomly rotated by ±90° degrees to avoid biases in orientation . The dataset was whitened by mean subtraction and a standard linear transformation x* = Mx , where M = RD − 1/2 RT and 〈x xT〉 = RDRT is the eigenvalue decomposition of the input correlation matrix . In Fig 6 , we used images preprocessed as in [6] , filtered in the spatial frequency domain by M ( f ) =fe− ( f/f0 ) 4 . The exponential factor is a low-pass filter that attenuates high-frequency spatial noise , with f0 = 200 cycles per image . The linear factor f was designed to whiten the images by canceling the approximately 1/f power law spatial correlation observed in natural images [39] . But since the exponent of the power law for this particular dataset has an exponent closer to 1 . 2 , the preprocessed images exhibit higher variance at lower spatial frequencies . Synaptic weights were initialized randomly ( normal distribution with zero mean ) and , for an effective nonlinearity f , evolved through w k + 1 = w k + η xk f ( w k T x k ) , for each input sample xk , with a small learning rate η . We enforced normalized weights at each time step , ||w||2 = 1 , through multiplicative normalization , implicitly assuming rapid homeostatic mechanisms [30 , 29] . For multiple neurons , the neural version of the sparse coding model described in Eq 7 was implemented . In Figs 4 and 6 , the learned receptive fields were fitted to Gabor filters by least square optimization . Receptive fields with less than 0 . 6 variance explained were rejected ( less than 5% of all receptive fields ) . In Fig 2b , the five selected candidate patterns are: random connectivity filter ( weights sampled independently from the normal distribution with zero mean ) , high-frequency Fourier filter ( with equal horizontal and vertical spatial periods , Tx = Ty = 8 pixels ) , difference of Gaussians filter ( σ1 = 3 . , σ2 = 4 . ) , low-frequency Fourier filter ( Tx = 16 , Ty = 32 ) , and centered localized Gabor filter ( σx = 1 . 5 , σy = 2 . 0 , f = 0 . 2 , θ = π/3 , ϕ = π/2 ) . Fourier filters were modeled as wab = sin ( 2πa/Tx ) ⋅ cos ( 2πb/Ty ) ; difference of Gaussians filters as the difference between two centered 2D Gaussians with same amplitude and standard deviations σ1 and σ2; and we considered standard Gabor filters , with center ( xc , yc ) , spatial frequency f , width σx , length σy , phase ϕ and angle θ . In Figs 4 and 6 we define the Gabor width and length in pixels as 2 . 5 times the standard deviation of the respective Gaussian envelopes , σx and σy . In Fig 5a , a Gabor filter of size s had parameters σx = 0 . 3 ⋅ s , σy = 0 . 6 ⋅ s , f = 1/s and θ = π/3 . In Fig 5b and 5c , the Gabor filter parameters were σx = 1 . 2 , σy = 2 . 4 , f = 0 . 25 . All receptive fields were normalized to ||w||2 = 1 . In Figs 4 and 6 , the background optimization value was calculated for Gabor filters of different widths , lengths , frequencies , phases ϕ = 0 and ϕ = π/2 . For each width and length , the maximum value among frequencies and phases was plotted . For the strabismus model , two independent natural image patches were concatenated , representing non-overlapping left and right eye inputs , forming a dataset with 16 by 32 patches [28] . For the binocular receptive field in the strabismus statistical analysis ( Fig 7a ) , a receptive field was learned with a binocular input with same input from left and right eyes . As V2 input , V1 complex cell responses were obtained from natural images as in standard energy models [46] , modeled as the sum of the squared responses of simple cells with alternated phases . These simple cells were modeled as linear neurons with Gabor receptive fields ( σx = 1 . 2 , σy = 2 . 4 , f = 0 . 3 ) , with centers placed on a 8 by 8 grid ( 3 . 1 pixels spacing ) , with 8 different orientations at each position ( total of 512 input dimensions ) . For the non-orientation selective receptive field in the V2 statistical analysis ( Fig 7d ) , the orientations of the input complex cells for the learned receptive field were randomized . As auditory input , spectrotemporal segments were sampled from utterances spoken by a US English male speaker ( CMU US BDL ARCTIC database [71] ) . For the frequency decomposition [14] , each audio segment was filtered by gammatone kernels , absolute and log value taken and downsampled to 50 Hz . Each sample was 20 time points long ( 400 ms segment ) and 20 frequency points wide ( equally spaced between 0 . 2 kHz and 4 . 0 kHz ) . For the non-local receptive field in the auditory statistical analysis ( Fig 7g ) , a Fourier filter was used ( Tt = Tf = 10 ) . For all datasets , the input ensemble was whitened after the preprocessing steps , by the same linear transformation described above for natural images , and all receptive fields were normalized to ||w||2 = 1 . | The question of how the brain self-organizes to develop precisely tuned neurons has puzzled neuroscientists at least since the discoveries of Hubel and Wiesel . In the past decades , a variety of theories and models have been proposed to describe receptive field formation , notably V1 simple cells , from natural inputs . We cut through the jungle of candidate explanations by demonstrating that in fact a single principle is sufficient to explain receptive field development . Our results follow from two major insights . First , we show that many representative models of sensory development are in fact implementing variations of a common principle: nonlinear Hebbian learning . Second , we reveal that nonlinear Hebbian learning is sufficient for receptive field formation through sensory inputs . The surprising result is that our findings are robust of specific details of a model , and allows for robust predictions on the learned receptive fields . Nonlinear Hebbian learning is therefore general in two senses: it applies to many models developed by theoreticians , and to many sensory modalities studied by experimental neuroscientists . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"neural",
"networks",
"engineering",
"and",
"technology",
"electronics",
"neuroscience",
"optimization",
"synaptic",
"plasticity",
"mathematics",
"computational",
"neuroscience",
"neuronal",
"plasticity",
"coding",
"mechanisms",
"rectifiers",
"developmental",
"neuroscience",
... | 2016 | Nonlinear Hebbian Learning as a Unifying Principle in Receptive Field Formation |
Hand , foot and mouth disease caused by enterovirus 71 ( EV71 ) leads to the majority of neurological complications and death in young children . While putative inactivated vaccines are only now undergoing clinical trials , no specific treatment options exist yet . Ideally , EV71 specific intravenous immunoglobulins could be developed for targeted treatment of severe cases . To date , only a single universally neutralizing monoclonal antibody against a conserved linear epitope of VP1 has been identified . Other enteroviruses have been shown to possess major conformational neutralizing epitopes on both the VP2 and VP3 capsid proteins . Hence , we attempted to isolate such neutralizing antibodies against conformational epitopes for their potential in the treatment of infection as well as differential diagnosis and vaccine optimization . Here we describe a universal neutralizing monoclonal antibody that recognizes a conserved conformational epitope of EV71 which was mapped using escape mutants . Eight escape mutants from different subgenogroups ( A , B2 , B4 , C2 , C4 ) were rescued; they harbored three essential mutations either at amino acid positions 59 , 62 or 67 of the VP3 protein which are all situated in the “knob” region . The escape mutant phenotype could be mimicked by incorporating these mutations into reverse genetically engineered viruses showing that P59L , A62D , A62P and E67D abolish both monoclonal antibody binding and neutralization activity . This is the first conformational neutralization epitope mapped on VP3 for EV71 .
Human enterovirus 71 ( EV71 ) is a causative agent of hand , foot and mouth disease ( HFMD ) which has become a serious health threat to young children in the Asia Pacific region over the last 15 years . Although HFMD is most commonly caused by members of the coxsackievirus family , which are genetically related to EV71 , infection with EV71 is more often associated with neurological complications in children under 3 years of age and is responsible for the majority of fatalities [1]–[3] . A major concern has been the emergence of a syndrome of rapidly fatal pulmonary edema associated with brainstem encephalitis in the Asian epidemics [4] . In an outbreak of HFMD in 2008 in China , up to half a million cases were reported among children resulting in over 120 fatal cases , which were primarily due to EV71 infection [5] . A recent outbreak in Cambodia led to the deaths of 54 children , most of them under 3 years of age: All samples obtained from fatal cases tested positive for EV71 . Since the nearly complete eradication of polio , EV71 is now regarded as the pre-eminent neurotrophic virus , and a threat to global public health [6] , [7] . To date , there are no specific antivirals or vaccines for clinical use , and prevention is mainly achieved by disrupting virus transmission with improved public hygiene in kindergartens , preschools and daycare centers aided by the temporary closures of affected places . A number of animal studies have shown that neutralizing antibodies stimulated by immunization with inactivated virus , VLPs , or displayed VP1 , are cross-protective against heterologous strains and can passively protect mice and monkeys [8]–[14] . Further , studies on patients have indicated that EV71 infection is cleared by humoral immunity and clinical trials have shown the presence of neutralizing antibodies in the serum of immunized healthy adults and children [13] , [15] , [16] . This significant involvement of neutralizing antibody responses in the control of EV71 infection in humans would render IVIG treatment an ideal therapeutic agent to complement vaccination . Passive immunization by IVIG with pooled sera from convalescent human donors has been pioneered by Behring & Kitasato in the 1890's with the development of anti-diphtheria serum . However , besides the risk of transmitting human pathogens using pooled human sera , necessitating screening and treatment , there are other disadvantages , i . e . the availability of donors , batch to batch variability , and the presence in the serum of virus specific but non-neutralizing antibodies . A solution would be to exploit future passive immunotherapy based on monoclonal antibodies ( mAb ) produced in cell culture . EV71 is a small non-enveloped single positive-stranded RNA virus belonging to the Picornaviridae family , genus Enterovirus , species Enterovirus A . The naked genome is enclosed in an icosahedral capsid composed of the four structural proteins ( VP1 , VP2 , VP3 , and VP4 ) . While VP1-VP3 form the surface of the virion , VP4 is arranged internally [17] , [18] . The rapid mutation rate of RNA viruses results in the emergence of new subgenogroups every few years . To date , 11 EV71 subgenogroups have been identified based on the VP1 capsid protein which is the most variable of the four [19] , [20] . The three genogroups are denoted A , B and C , of which B and C are further divided into subgenogroups B1–5 and C1–5 [19] . Both the co-circulation of different genogroups and the emergence of novel strains have been observed in the Asian pandemics [21]–[24] . Interestingly , it has been shown that this genotyping does not correspond to the virus strain antigenicity which points to the presence of conformational epitopes , independent of sequence homology [25]–[29] . It is surprising then , that only 1 universal neutralizing epitope of EV71 has been described thus far , i . e . a linear epitope on VP1 which encompasses amino acids 215-KQEKD-219 [30] . This region is conserved among all EV71 subtypes - making it a universal neutralizing epitope - and a monoclonal antibody against this epitope is able to protect mouse pups from lethal EV71 infection [30] . Given that other picornaviruses such as poliovirus carry multiple neutralizing epitopes ( including conformational epitopes ) , we expect EV71 to also possess more than one linear neutralization epitope [31] . Indeed , this is the case as we here report the isolation of the first conformational , universally neutralizing mAb against EV71 that is not on VP1 .
All animal experiments were carried out in accordance with the Guidelines for Animal Experiments of the National Advisory Committee on Laboratory Animal Research ( NACLAR ) . Experimental protocols were reviewed and approved by Institutional Animal Care and Use Committee of the Temasek Life Sciences Laboratory , National University of Singapore , Singapore . ( IACUC approval number TLL-IACUC-2013/002 ) . African green monkey cell line ( Vero ATCC number CCL-81 ) and Rhabdomyosarcoma cell line ( RD ATCC number CCL-136 ) were obtained from American Type Culture Collection ( ATCC ) and grown in Dulbecco's modified Eagle medium ( DMEM , Gibco , USA ) containing 10% fetal bovine serum ( FBS , Biowest , France ) at 37°C with 5% CO2 . Wild-type ( wt ) EV71 strains and CVA16 strain ( U05876 ) were obtained from the Host and Pathogen Interactivity Laboratory , Department of Microbiology , Yong Loo Lin School of Medicine , National University of Singapore . The GenBank accession numbers of representative EV71 subgenogroups are listed in Table 1 . Missing subgenogroups B1 ( GenBank AF135901 ) , B3 ( GenBank AF376093 ) and C3 ( GenBank AY125973 ) were constructed using the human RNA polymerase I reverse genetics system by inserting the relevant VP1 genes into the backbone of EV71 C4 strain Fuyang . Anhui . P . R . C/17 . 08/2 ( GenBank EU703813 ) . These viruses were propagated in RD cells cultured in supplemented with 2% FBS . Cell culture supernatants were harvested 4 days post infection ( dpi ) , when 100% cytopathic effect ( CPE ) was observed . After three freeze-thaw cycles and filtration through a 0 . 2 um cut-off filter ( Sartorius , Germany ) , aliquots were stored at −80°C . Virus activity was tested on RD cells in an end-point dilution assay to determine the 50% tissue culture infective dose ( TCID50 ) . For animal immunization EV71-B4 was inactivated with binary ethylenimine ( BEI ) for 48 h at 37°C as described by Bahnemann [32] . Virus was then concentrated 10-fold by ultracentrifugation at 100 , 000 g for 3 h and re-suspended in PBS . Three specific pathogen-free BALB/c mice were immunized subcutaneously on days 0 , 14 and 28 with inactivated EV71-B4 strain in 0 . 1 ml PBS , emulsified with an equal volume of adjuvant ( Seppic , France ) . An intraperitoneal booster of the same virus dose without adjuvant was administered 3 days before the mice were euthanized and their spleen cells harvested . Splenocytes were then fused with SP2/0 myeloma cells as described [33] and the resulting hybridomas were grown in DMEM with 20% FBS containing HAT or HT for 10 days . The hybridomas were screened by IFA of Vero cells infected with EV71-B4 and cells secreting specific antibodies were subcloned by limiting dilution and cultured . Vero African green monkey kidney cells infected with EV71-B4 were used for antibody screening . Cells were seeded overnight into 96 well microtiter plates and infected with a 10−6 dilution of EV71-B4 the next morning . Upon observation of CPE , the cells were fixed in 4% paraformaldehyde ( pH 7 . 4 ) for 20 min , and permeabilized with 0 . 1% Triton-X/PBS for 30 min . The cells were blocked with PBS containing 5% FBS for 1 h at RT and incubated overnight at 4°C with hybridoma cell supernatants or mAb 51 as positive control . Anti-mouse FITC-coupled secondary antibody was then added for 1 h at RT . The ells were washed three times with 0 . 1% PBS-Tween between each step . Results were documented with an inverted microscope ( Olympus ) with Nikon ACT-1 software . IFA of escape mutants and RG viruses was conducted with infected RD cells using the same protocol as above . Reverse-genetically ( RG ) constructed viruses were propagated in RD cell cultures . Harvested supernatants were purified by sucrose gradient centrifugation: cells were pelleted by centrifugation at 8 , 000× g for 40 mins , after which the supernatant was ultracentrifuged at 100 , 000× g for 2 h . The resulting pellet was re-suspended and centrifuged in a 20–60% discontinuous sucrose gradient at 100 , 000× g for 3 h , and the virus band was collected . To concentrate the virus , PBS was added to the sucrose gradient band which was further centrifuged at 100 , 000× g for 1 h , and resuspended in PBS . Protein concentrations were measured by Nanodrop , and all purified viruses were diluted to 0 . 3 mg/ml . 80 µl of RG viruses was then mixed with 20 µl of 10% SDS , boiled for 5 min , and diluted 10 times . 100 µl ( around 30 µg ) of virus was then dotted on a nitrocellulose membrane . The blot was blocked in 5% milk in PBS for 1 h at RT before incubation in primary antibodies 10D3 and 53 for 1 h at RT . After washing with PBS-T the blot was incubated with anti-mouse secondary antibody , then ECL reagent ( GE Healthcare , USA ) , and the image captured by ChemiDoc MP imaging System ( Bio-Rad Laboratories Inc . USA ) . The neutralization titer of mAb 10D3 ( hybridoma cell culture supernatant ) was measured in an in vitro microneutralization assay using RD cells . 100 TCID50 of wild-type , escape mutant , or RG viruses were mixed with an equal amount of 2-fold serial dilutions of mAb 10D3 or mAb 51 as a positive control . The mixtures were incubated for 1 h at RT before adding them in triplicates to the wells of microtiter plates containing 80% confluent RD cells . Presence of CPE was determined after 4 days by examination under the light microscope . The highest dilution of mAb that inhibited virus growth was considered the neutralizing titer and expressed as 2x . Assays were carried out independently three times . The wild-type virus stocks EV71-A , EV71-B2 , EV71-B4 , EV71-C2 , and EV71-C4 were diluted to 50 TCID50×Neutralization titer against accordingly virus [34] . Then incubated in an equal volume of neat mAb 10D3 ( hybridoma supernatant ) for 1 h at room temperature ( RT ) . The mixture was then transferred to 80% confluent RD cells in DMEM with 10% FBS and incubated for 4 days . If no CPE was observed , the supernatants were harvested , subjected to three freeze-thaw cycles and filtered with a 0 . 2 um cut-off before re-infecting a fresh batch of RD cells for 4 days . This was repeated until CPE was observed . 1–3 re-infection cycles were needed for CPE , and hence EV71 escape mutants to develop . The escape mutants were called E1–3/B4 ( three individual experiments using EV71-B4 virus ) , E1–2/B2 ( two individual experiments using EV71-B2 strain ) , E/A ( EV71-A ) , E/C2 ( EV71-C2 ) , and E/C4 ( EV71-C4 ) . TCID50 was measured by end point dilution and IFA as well as microneutralization against mAb 10D3 was conducted to confirm abolishment of antibody binding and neutralization . The viral RNA isolation kit ( Qiagen , Germany ) was used to extract viral RNA from filtered RD cell culture supernatants containing wild-type and escape mutant virus . Typical yields were 80–100 ng/µL as measured by Nanodrop ( ThermoFisher Scientific , USA ) . Reverse transcription was carried out on 500 ng RNA , using gene- and strain-specific primers together with AMV reverse transcriptase ( Roche Applied Science , Germany ) according to the manufacturer's protocol . PCR amplification of 2 overlapping portions of P1 region was then conducted using the primer pairs ( Table 2 ) and the High Fidelity PCR system ( Roche Applied Science , Germany ) . The cycling parameters were as follows: Denaturation at 94°C for 2 min; followed by 10 cycles of denaturation at 94°C for 30 sec , touchdown annealing from 54°C to 45°C in 1°C decrements for 30 sec , extension at 72°C for 2 min; followed by 30 cycles of denaturation at 94°C for 30 sec , annealing at 45°C for 30 sec , extension at 72°C for 2 min+5 sec per cycle increments , and a final extension at 72°C for 7 min . The resulting PCR products were analyzed on a 1% agarose gel and purified by QIAquick gel extraction kit ( Qiagen , Germany ) . A direct sequencing reaction was performed using gene- and strain-specific primers and BigDye terminator cycling at the DNA/Oligonucleotide Synthesis core of Temasek Life Sciences Laboratories , Singapore . Sequences were analyzed using the Lasergene programs ( DNAstar , USA ) . The genome of B4 wild type virus was first amplified by RT-PCR and subjected to human RNA polymerase I promoter as described in the previous paper [35] . The infectious plasmids containing B4 cDNA ( pJET-B4-wt ) were sequenced to confirm their authenticity , and transfected into RD cells to generate RG/B4-wt virus . The mutations were introduced into the pJET-B4-wt plasmid by site-directed mutagenesis ( Stratagene , USA ) using primers ( Table 3 ) . For double mutations in pJET-B4-PE59 , 67LD , pJET-B4-P59L was further mutated by primers B4-E67D-f and B4-E67D-r . The correct mutated plasmids were transfected into RD cells as above to generate the mutants , and designated as RG/B4-P59L , RG/B2-A62D , RG/B2-A62P , RG/B4-E67D , and RG/B4-PE59 , 67LD , respectively . The animal experiments were conducted with two week old AG129 mice . These mice were obtained from B&K Universal ( UK ) . They were housed and bred under specific pathogen-free conditions in individual ventilated cages . To test the protective efficacy of the antibody , these mice were randomly divided into two groups of 10 mice each . Group 1 mice ( prophylactic group ) were injected intraperitoneally with the purified mAb 10D3 antibody ( 0 . 1 ml in 50% glycerol dissolved in PBS ) at a concentration of 10 mg/g of body weight on day one . Group 2 mice ( isotype control group ) were given an equal amount of purified mouse IgM as an isotype control ( eBioscience , USA ) . These two groups of mice were then subjected to a lethal challenge with 107 plaque forming units ( PFU ) of EV71 strain HFM 41 ( 5865/SIN/00009 ) via the intraperitoneal route ( 0 . 4 ml in PBS ) , 24 h post-injection of the immunoglobulins . Survival rates and clinical scores of the mice were monitored daily till 14 days post-infection . Total limb paralysis was used as criterion for early euthanasia [30] , [36] . Brain samples were collected , fixed in formalin , embedded in paraffin blocks , cut at 5 mm thickness ( Leica Microsystems , Germany ) , and attached to coated glass slides . The slides were stained with hematoxylin and eosin ( H&E ) and observed under light microscopy .
To discover novel neutralizing epitopes of EV71 , three BALB/c mice were immunized with 100 µL of inactivated EV71-B4 strain virus in a 1∶1 emulsion with adjuvant ( Seppic France ) . The EV71-B4 strain ( 5865/SIN/000009 ) was propagated in RD cells , the supernatant containing the virus was inactivated with BEI , and concentrated by ultracentrifugation prior to immunization . Boosters were administered at 14 day intervals and the sera were tested 7 days later for the presence of EV71-specific antibodies by IFA of EV71-infected Vero cells . Once the sera exhibited positive IFA signals , an intraperitoneal booster was administered without adjuvant , and the spleens were harvested three days later . Splenocytes were fused with myeloma cells , the resultant hybridomas were cultured in selective medium , and supernatants were screened by IFA of EV71-B4-infected Vero cells . Positive hybridomas were subcloned by limiting dilution , and their supernatants were analyzed for the presence of neutralizing antibody . In this screen , several mAbs were isolated and further characterized but we focused our attention on mAb 10D3 since this mAb reacted positively with all 11 EV71 subgenogroups by IFA but did not cross-react to CVA16 ( Fig . 1 ) . This is a promising feature as the mAb could potentially be applied for differential diagnosis of HFMD to distinguish CVA from EV71 infections . Secondly , the mAb was able to neutralize all EV71 subgenogroups with a neutralization titer of 26 ( genogroups A , B ) to 28 ( genogroup C ) against 100 TCID50 of wild-type virus ( Table 1 ) by using hybridoma cell supernatant . This universal neutralization ability makes mAb 10D3 an ideal candidate for diagnosis and treatment of EV71 infection . Since Western blot analysis against whole virus and overlapping EV71 P1 polyprotein fragments tagged with GST did not result in any bands with mAb 10D3 , we investigated the reactivity of mAb 10D3 with native and denatured viruses in a dot blot assay . Although mAb 10D3 reacted with reverse genetically engineered wild-type B4 virus ( RG/B4-wt ) blotted in its native form ( RG/B4-wt native ) , it did not recognize virus denatured by boiling with SDS ( RG/B4-wt denatured ) ( Fig . 2F ) . MAb 53 , which recognizes a linear epitope on VP1 , was used as a positive control , and could indeed react with both the native and denatured viruses [30] . In conclusion , mAb 10D3 recognizes a conformational epitope . Finally , the mAb 10D3 immunoglobulin isotype was determined as IgM using the mouse monoclonal antibody isotyping kit ( Santa Cruz Biotechnology Inc . , USA ) . The epitope of mAb 10D3 was found to be conformational since this mAb did not react with any capsid protein in a Western blot . Hence , the epitope could not be mapped by the conventional fashion of truncated peptides . Therefore epitope mapping of mAb 10D3 was performed by escape mutant selection . Wild-type EV71 viruses from different subgenogroups ( A , B2 , B4 , C2 , C4 ) were incubated with an excess of mAb 10D3 on RD cells . If no CPE was visible after 4 days , supernatants were filtered and added to fresh RD cells . This process was repeated until CPE was evident . 1 to 3 cycles were necessary to isolate escape mutants for all subgenogroups . The escape mutants were designated E/A ( EV71-A ) , E1–2/B2 ( two experiments using EV71-B2 virus ) , E1–3/B4 ( three individual experiments using EV71-B4 virus ) , E/C2 ( EV71-C2 ) , and E/C4 ( EV71-C4 ) , their TCID50 was measured by end-point dilution assay and they were tested for reactivity with mAb 10D3 by IFA . RD cells were infected with an equal amount of either wild-type virus as positive controls or escaped viruses and observed for 2 days until CPE was visible ( Fig . 2C ) . After incubation with mAb 10D3 , there was a clear fluorescent signal for the wild-type viruses ( Fig . 2A ) , but no signal was detected for any of the corresponding escape mutants ( Fig . 2B ) . To further confirm that the escape mutants have evaded mAb 10D3 binding , a microneutralization assay against 100 TCID50 of escaped viruses was conducted . There was no more virus neutralization by mAb 10D3 of any of the identified escape mutants . However , neutralizing mAb 51 , which recognizes an unaltered epitope on VP1 , was able to neutralize all escape mutants ( Table 4 ) . To delineate the amino acid mutations associated with neutralization escape of the different subgenogroups , the P1 structural gene region of each escape mutant was sequenced and compared to its parental strain . In the eight escape mutants , four mutations were identified in the structural gene VP3 . The mutants E1–3/B4 derived from the parental strain B4 harbored a glutamate to aspartate substitution at amino acid position 67 of VP3 , while the other three mutants E/A , E/C2 , E/C4 derived from A , C2 , and C4 subgenogroups carried a proline to leucine substitution at amino acid position 59 of VP3 . Two separate mutations were discovered at amino acid 62 of the escape mutants from the B2 strain: an alanine to aspartic acid or proline ( Table 5 ) . Since we have discovered only a single amino acid mutation in the capsid proteins of each escape mutant , it can be inferred that these residues are essential for both mAb 10D3 binding and virus neutralization . To test this hypothesis , we engineered an EV71-B4 virus consisting of the EV71-B4 ( 5865/SIN/000009 ) sequence by utilizing a human RNA polymerase I driven reverse genetics system [35] . The four VP3 mutations P59L , A62D , A62P and E67D were then introduced alone ( RG/B4-P59L , RG/B4-A62D , RG/B4-A62P , RG/B4-E67D ) or in tandem ( RG/B4-PE59 , 67LD ) into the wild-type RG virus ( B4 RGV ) by site-directed mutagenesis . The RG viruses were then rescued in RD cells and passage 2 viruses were used in subsequent experiments . The binding ability of mAb 10D3 to the representative mutated RG viruses RG/B4-P59L , RG/B4-E67D and RG/B4-PE59 , 67LD was first tested by IFA ( Fig . 2DE ) , and dot blot ( Fig . 2F ) . RD cells were infected with RG/B4-wt virus as positive control or the mutated RG viruses . The cells were fixed 2 dpi when CPE was clearly observed ( Fig . 2E ) . While the original virus ( RG/B4-wt ) was clearly detected by mAb 10D3 , no fluorescence was visible for the mutated RGVs carrying either a single or double mutations . Additionally , mAb 10D3 was unable to neutralize the mutated RG viruses ( RG/B4-P59L , RG/B4-A62D , RG/B4-A62P , RG/B4-E67D , RG/B4-PE59 , 67LD ) by an in vitro microneutralization assay , while the neutralization titer of mAb 10D3 against RG/B4-wt reached 26 which was the same as for B4-wild-type ( Table 4 ) . As a positive control , mAb 51 against the linear neutralizing epitope KQEKD on VP1 was incubated with the mutated RG viruses . Since the VP1 epitope was unaffected by our mutagenesis , mAb 51 was still able to efficiently neutralize all RG viruses ( Fig . 3 ) . Hence we have demonstrated that the four escape mutations ( P59L , A62D , A62P and E67D ) are sufficient for the abolishment of mAb 10D3 binding to the VP3 protein and neutralization of EV71 virus . Having identified 3 amino acids of VP3 that are essential for mAb 10D3 binding and neutralization , we next investigated whether these residues are conserved in all of the fully sequenced EV71 strains available on GenBank . BLAST analysis of amino acids 59–67 of VP3 revealed a total of 388 EV71 hits which were all 100% identical in the region analyzed , while the amino acid identity was 97% for the full VP3 protein . VP3 is thus more highly conserved between subgenogroups than VP1 ( 93% identity ) , making it an ideal target for a diagnostic or therapeutic mAb . The same region was also compared to CVA16 strains , which exhibited no sequence homology to EV71 ( Table 5 ) . In the view of the recently available 3D crystal structure of EV71-C4 , the epitope of mAb 10D3 could be located by stereographic imaging [17] , [18] . To analyze the location of the VP3 epitope in relation to previously identified EV71 epitopes ( the linear neutralizing epitope KQEKD of mAb 51 on VP1 , and the linear epitope EDSHP of mAb 7C7 on VP2 ) [30] , [37] , we studied stereographic images of EV71 protomers . In Fig . 4 the epitopes of mAb 51 , 7C7 and 10D3 are shown on the virus surface . In Fig . 5 and 6 the EV71 VP3 protein is shown and the escape mutations P59L , A62D , A62P and E67D are indicated . Both sites lie on the major protrusion of VP3 on the capsid surface termed “knob” [18] . In Fig . 5B and 5C , a protomer consisting of one copy each of the viral capsid proteins VP1 ( pink ) , VP2 ( blue ) , VP3 ( brown ) and VP4 ( green ) is shown . The top of the image corresponds to the surface of the virion and the bottom ( where VP4 is located ) to the inside . Two orientations are shown: ( B ) The major groove , formed by VP1 , is visible to the right , while VP2 is in the foreground and VP3 in the back . ( C ) The image was rotated to display the positions of the escape mutations on VP3 more clearly . VP1 , VP3 are now in the foreground and VP2 in the back . Indicated in yellow are the epitopes of some previously identified mAbs of EV71 as well as the conformational neutralizing epitope of mAb 10D3 in the knob region of VP3 ( arrows ) . The three escape mutation sites at amino acid positions 59 , 62 and 67 on VP3 are indicated . As shown in Fig . 6 , the conformational changes generated by the escape mutants were as follows: A loss of the cyclic structure of proline's side chain in the P59L mutant , the addition of a cyclic side chain in the A62P mutant , or of a carboxyl group in the A62D mutant , and the loss of a methylene bridge in the E67D mutant . To test the protective efficacy of mAb 10D3 , two week old AG129 mice were injected intraperitoneally with purified mAb 10D3 or isotype control mAb . One day later , they were challenged with a lethal dose of the virulent EV71 strain HFM41 , and clinical scores as well as survival rates of the mice were monitored daily . In the control animals , which received an isotype antibody , 80% developed severe limb paralysis as early as day 6 post-infection . In contrast , the mice pre-treated with mAb 10D3 did not display any of the disease manifestations , and remained healthy throughout the experiment . Our result thus suggested that the anti-EV71 antibody mAb 10D3 ( administered at a dose of 10 mg/g of body weight ) was able to achieve 100% protection against the lethal EV71 challenge . To confirm the protective efficacy of mAb 10D3 , a histopathologic examination of the mouse brains was conducted . Mice from the isotype control group exhibited neuronal vacuolation and neuronal loss without inflammation in the brain stem ( Figs . 7A , 7B1 , 7B2 and 7B3 ) . In contrast , we did not observe such pathologic changes in mice from the prophylactic group treated with 10D3 . The intact brain morphology ( Figs . 7C and 7D ) suggested that mAb 10D3 was capable of conferring in vivo passive protection against EV71 infection .
Despite the evidence that enteroviruses have major conformational neutralization sites on all capsid proteins , to date the only mapped universal neutralization epitope of EV71 is a linear epitope spanning amino acids 215–219 of VP1 [30] , [38] , whereas the only known conformational epitope is strain-specific and includes amino acid 145 of VP1 [39] . In order to find an optimal mAb for future use as an immunologic therapeutic , either alone or in conjunction with mAb 51 , a larger pool of universally neutralizing mAb candidates would be desired . We therefore undertook the task of isolating such mAbs derived from mice immunized with EV71-B4 ( 5865/SIN/000009 ) . This strain was selected for its virulence , as it consistently attained a higher TCID50 in cell culture and was able to cause disease in older mice pups than any other strain we tested ( unpublished observations ) . This strain was isolated from a fatal case of HFMD with encephalitis during the Singapore outbreak of EV71 in 2000 [40] . We found that mAb 10D3 that efficiently neutralized all EV71 subgenogroups without cross-reaction to CVA16 , making it a highly specific antibody with a potential application in differential diagnosis of HFMD . Since the epitope of 10D3 was found to be conformational , it was crucial to generate escape mutants in order to map the exact epitope . Mutations of amino acids proline to leucine at position 59 , alanine to aspartic acid or proline at position 62 , or glutamate to aspartate at position 67 of VP3 resulted in escape from mAb binding and neutralization . These residues are integral to the “knob” structure of EV71 VP3 protein that protrudes out of the capsid surface , and is completely conserved among all EV71 sequences deposited in GenBank . Hence , we have discovered an additional conformational epitope of EV71 , and provided the first evidence of another capsid protein involved in EV71 virus neutralization besides VP1 . The knob of EV71 VP3 encompasses residues 55 to 69 of VP3 - making it longer than the knob described for coxsackievirus B3 ( CVB3 ) - and contains a major neutralization site for other picornaviruses such as CVB3 , poliovirus 1 , human rhinovirus 14 and hepatitis A virus [41]–[44] . In hepatitis A virus the immunodominant epitope involves residue 70 of VP3 which is close to the mutations identified in our screen [44] . The situation is more complex for poliovirus: while N-AgI is the major neutralizing epitope for the PV-3 serotype ( Sabin ) , N-AgII and N-AgIII are immunodominant for PV-1 ( Mahoney ) [45] . However , neutralizing IgA mAbs , derived from both PV-1 and PV-3 immunized mice , were all predominantly directed against the N-AgIII epitope [46] , [47] which is formed by amino acids 58–59 of VP3 and 286–290 of VP1 [31] . Mapped onto the EV71 virus structure , these residues are in close proximity of the VP3 knob and pass right in front of ( Fig . 5B pink tube at the upper right of the protomer structure ) . It remains to be seen if some 10D3 escape mutants might also have alterations in the C-terminus of VP1 in addition to the VP3 knob , as these residues may contribute to epitope formation . Escape mutants of another enterovirus , coxsackievirus B3 , harbor mutations in both the knob of VP3 and the puff of VP2 . The VP3 mutation was mapped to residue 60 , while the VP2 mutation was located on residue 158 . Stereographic imaging revealed that the two mutations lie in close proximity to one another , forming a conformational epitope [41] . By analogy to CVB3 , we more closely analyze the other identified epitopes of EV71 . Since our previously identified , non-neutralizing , linear mAb 7C7 against EV71 VP2 has a linear epitope quite close to CVB3 VP2 158 , i . e . residues 142–146 [37] , we investigated whether these amino acids might be involved in forming a conformational epitope with VP3 in EV71 as well . As can be readily deduced from the stereographic images , the VP2 puff of EV71 resides much further away from the VP3 knob , and neither residues 142–146 nor 158 are in close proximity to our identified escape mutations . Instead , the VP2 epitope 142-EDSHP-146 is adjacent to the neutralizing VP1 epitope 215-KQEKD-219 indicating that these two linear epitopes may interact in EV71 which could explain the low ( <1∶14 ) neutralizing activity observed for the commercially available mAb979 ( Merck Millipore , Germany ) which recognizes a peptide of VP2 spanning residues 136–150 which encompasses the 7C7 epitope [48] . Neutralizing monoclonal antibodies are specific antiviral agents that can be used for the passive immunization of patients with acute viral infections . They offer a selective advantage over pooled human sera that are more commonly used in IVIG treatment by reducing the risk of transmitting pathogens , and by alleviating batch-to-batch variability , availability of donors , and the presence of non-neutralizing antibodies . Several factors have to be considered when using mAbs instead of polyclonal serum for IVIG , including ( a ) the antigenic variability of circulating strains , i . e . the mAb must cross-neutralize all existing subtypes to be useful; ( b ) The risk of escape mutations , e . g . mutants may emerge under selective pressure such as the presence of a neutralizing antibody . To circumvent this risk , a cocktail of two antiviral mAbs against non-overlapping epitopes can be administered , where escape mutation from a single mAb does not interfere with the neutralizing capability of the second mAb . A combination of synergistic mAbs may also reduce the required dosages [49] , [50] . In conclusion , the protective efficacy of mAb 10D3 was evaluated and verified by an animal challenge experiment using a lethal dose of EV71 . All mice prophylactically treated with mAb 10D3 survived the lethal challenge without showing any disease symptoms . Hence , mAb 10D3 holds promise for being further developed as a prophylactic agent against EV71-associated HFMD . | Over the last decade , EV71 has emerged as a major cause of severe hand , foot and mouth disease in the Asia-Pacific region , occasionally leading to fatal brain stem encephalitis in young children . The rapid progression and high mortality of severe EV71 infection makes it vital to identify neutralization epitopes and putative therapeutic monoclonal antibodies . In this study we mapped the first conformational neutralization epitope on the VP3 protein of EV71 . This epitope was confirmed by introducing the mutations into reverse genetically engineered viruses which abolished neutralization with monoclonal antibody ( mAb ) 10D3 . The importance of this novel neutralization epitope lies in the optimization of putative EV71 vaccines because the VP3 knob could be incorporated together with VP1 into a bivalent subunit vaccine . Further , the universal recognition of a conserved site on EV71 VP3 and not CVA16 makes mAb 10D3 a valuable tool for differential diagnosis of hand , foot and mouth disease . An additional hope is that mAb 10D3 could be used as a therapeutic intravenous immunoglobulin ( IVIG ) . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biology",
"and",
"life",
"sciences",
"immunology",
"microbiology",
"virology"
] | 2014 | A Novel Universal Neutralizing Monoclonal Antibody against Enterovirus 71 That Targets the Highly Conserved “Knob” Region of VP3 Protein |
Rift Valley fever virus ( RVFV ) is an important mosquito-borne veterinary and human pathogen that has caused large outbreaks of severe disease throughout Africa and the Arabian Peninsula . Currently , no licensed vaccine or therapeutics exists to treat this potentially deadly disease . The explosive nature of RVFV outbreaks and the severe consequences of its accidental or intentional introduction into RVFV-free areas provide the impetus for the development of novel vaccine candidates for use in both livestock and humans . Rationally designed vaccine candidates using reverse genetics have been used to develop deletion mutants of two known RVFV virulence factors , the NSs and NSm genes . These recombinant viruses were demonstrated to be protective and immunogenic in rats , mice , and sheep , without producing clinical illness in these animals . Here , we expand upon those findings and evaluate the single deletion mutant ( ΔNSs rRVFV ) and double deletion mutant ( ΔNSs-ΔNSm rRVFV ) vaccine candidates in the common marmoset ( Callithrix jacchus ) , a non-human primate ( NHP ) model resembling severe human RVF disease . We demonstrate that both the ΔNSs and ΔNSs-ΔNSm rRVFV vaccine candidates were found to be safe and immunogenic in the current study . The vaccinated animals received a single dose of vaccine that led to the development of a robust antibody response . No vaccine-induced adverse reactions , signs of clinical illness or infectious virus were detected in the vaccinated marmosets . All vaccinated animals that were subsequently challenged with RVFV were protected against viremia and liver disease . In summary , our results provide the basis for further development of the ΔNSs and ΔNSs-ΔNSm rRVFV as safe and effective human RVFV vaccines for this significant public health threat .
Rift Valley fever virus ( RVFV; family Phenuiviridae , genus Phlebovirus ) was first isolated in 1930 in East Africa [1] and has since caused severe epidemics and epizootics that affects ruminants and humans throughout Africa and the Arabian peninsula [2 , 3] . Human infections result from infected mosquitoes ( Culex , Mansonia and Anopheles mosquitoes appear to be the principal vectors for humans [4] ) or by contact with tissues , blood , or fluids from infected animals . Human cases are typically self-limiting febrile illnesses and recovery occurs without major consequences . Severe cases , which affect around 1–2% of infected individuals , are characterized by acute-onset liver disease , delayed-onset encephalitis , retinitis , blindness , or a hemorrhagic syndrome , with a case fatality ratio of 10–20% in hospitalized individuals [5–7] . Human cases have been reported in much of Africa , Saudi Arabia , and Yemen [8] . The spread of RVFV into other geographic regions is a major global concern . The productive experimental infection of mosquitoes from multiple distinct geographical regions ( including the most widespread vector , Culex pipiens ) reinforces the feasibility of accidental or intentional import of RVFV from endemic regions with subsequent maintenance in nascent vector and host populations [9–12] . The emergence of RVF into new locations has important implications for human health and livestock industries leading to its identification as a notifiable disease by the World Organization for Animal Health [13] and the World Health Organization as a high priority pathogen requiring attention [14] . Furthermore , due to concerns regarding its use as a potential biological weapon , RVFV has been identified as a Category A , high-priority select agent , by the National Institute for Allergy and Infectious Diseases ( NIAID ) , the Centers for Disease Control and Prevention ( CDC ) , and the United States ( U . S . ) Department of Agriculture ( USDA ) . Clearly , RVFV is an important threat to human and animal health for which no specific treatment currently exists . Several RVFV vaccines have been developed [8] , but currently none of these candidates has been approved for human use . The formalin-inactivated vaccine TSI GSD 200 was developed by the U . S . Army to protect at-risk laboratory workers against occupational exposure . However , a significant drawback of this vaccine is it requires three inoculations over a 4-week period , a mandatory boost at 6 months , and many recipients require periodic boosters thereafter [15–17] . To overcome these limitations , several live-attenuated vaccines were developed such as the Smithburn and MP-12 vaccines . The Smithburn vaccine has been used in Africa , but has been associated with teratogenesis and abortions in livestock and retains neurovirulence in non-human primates [18 , 19] . The MP-12 vaccine was developed by the U . S . Department of Defense [20] and has undergone Phase 1 and 2 clinical trials [21 , 22] . Additionally , the MP-12 vaccine is conditionally licensed for veterinary use in the U . S . despite a report that the vaccine may cause teratogenesis or abortions in pregnant ruminants [23] . Furthermore , the MP-12 vaccine lacks a marker for the differentiation of vaccinated from infected animals ( DIVA ) . To overcome some of the limitations of previous live-attenuated vaccines , Bird et al . [24 , 25] used reverse genetics to develop a recombinant RVFV , ΔNSs-ΔNSm rRVFV , which contains complete gene deletions of the 2 known RVFV virulence factors , the NSs and NSm genes [24 , 26–28] . RVFV has a tripartite negative-stranded RNA genome designated Small ( S ) , Medium ( M ) , and Large ( L ) . The S-segment encodes , in an ambisense fashion , the virus nucleoprotein ( NP ) in the genomic ( negative-sense [–] ) orientation and the nonstructural ( NSs ) protein in the antigenomic ( positive-sense [+] ) orientation . The M-segment contains at least four nested proteins in a single open reading frame ( ORF ) : the two structural glycoproteins , Gn and Gc , and two nonstructural proteins , the 14-kDa NSm and the 78-kDa NSm-Gn fusion protein . The L-segment encodes the viral RNA-dependent RNA polymerase . The NSs protein is involved in several functions in infected cells such as inhibition of IFN-β , degradation of protein kinase R ( PKR ) , suppression of host transcription , and interactions with host cell chromosome structures [29–34] . The NSm gene is not as well characterized , but has been implicated to be involved in suppression of virus-induced apoptosis [28] . NSs and NSm are not required in cell culture for efficient virus replication , assembly , or maturation [28 , 34–36] . The rRVFV vaccine candidates containing the insertion of the enhanced green fluorescent protein and the precise deletion of the NSs gene alone ( ΔNSs:GFP rRVFV ) or the NSs/NSm genes in combination ( ΔNSs:GFP-ΔNSm rRVFV ) were described by Bird et al . to be highly attenuated , immunogenic , and efficacious in the rat lethal disease model [24] . Furthermore , the ΔNSs-ΔNSm rRVFV vaccine was demonstrated to be protective and immunogenic in sheep without producing clinical illness in these animals [24 , 25] . Importantly , the vaccine was nonteratogenic in pregnant sheep , which is critical to indicate the safety needed for a veterinary vaccine in a natural RVFV host . While the demonstrated safety and efficacy in a natural target species helps to facilitate the acceptance of a vaccine for human use , it is important to determine immunogenicity and efficacy in a species more closely resembling humans . Thus , we completed an evaluation of the single deletion mutant ( ΔNSs rRVFV ) and double deletion mutant ( ΔNSs-ΔNSm rRVFV ) vaccine candidates in non-human primates ( NHP ) . Rhesus macaques historically have been used to evaluate potential vaccines and therapeutics for RVFV [37] . We previously described the development of a NHP model for RVF using the common marmoset ( Callithrix jacchus ) . Marmosets were more susceptible to RVFV than rhesus macaques and experienced higher rates of morbidity , mortality , and viremia and marked aberrations in hematological and chemistry values . Depending on the route of exposure , these animals exhibited acute-onset hepatitis , delayed-onset encephalitis , and hemorrhagic disease , which are dominant features of severe human RVF [38] . An additional study compared the susceptibility of rhesus macaques , cynomolgus macaques , African green monkeys , and marmosets exposed to RVFV by aerosol [39 , 40] . Cynomolgus and rhesus macaques developed mild fevers , but no other clinical signs were observed and all the monkeys survived . In contrast , African green monkeys and marmosets were found to be highly susceptible to aerosol infection where the majority of animals developed fatal encephalitis [39 , 40] . Collectively , these studies highlight the utility of the marmoset model of RVF to evaluate potential medical countermeasures because of their ability to mimic different features of severe human disease . Here , we demonstrate that both the single deletion mutant ( ΔNSs rRVFV ) and double deletion mutant ( ΔNSs-ΔNSm rRVFV ) vaccine candidates were found to be generally safe and immunogenic in a marmoset model of RVF . The vaccinated marmosets exhibited no signs of clinical illness post-vaccination and post-challenge and developed strong neutralizing antibody titers . Our results provide the basis for further development of the ΔNSs and ΔNSs-ΔNSm rRVFV as safe and effective human RVFV vaccines for this important public health threat .
This work was supported by an approved USAMRIID IACUC animal research protocol ( AP-10-066 ) . Research was conducted under an IACUC approved protocol in compliance with the Animal Welfare Act , PHS Policy , and other Federal statutes and regulations relating to animals and experiments involving animals . The facility where this research was conducted is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care , International and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals , National Research Council , 2011 . Approved USAMRIID animal research protocols undergo an annual review every year . Animals are cared for by a large staff of highly qualified veterinarians , veterinary technicians , and animal caretakers . All personnel caring for and working with animals at USAMRIID have substantial training to ensure only the highest quality animal care and use . Humane endpoints were used during all studies and marmosets were humanely euthanized when moribund according to an endpoint score sheet . Construction of the rRVFV has been previously described [24 , 26 , 35] and a schematic of the rRVFV reverse genetics rescue system and locations of the NSs and NSm gene deletions were described by Bird et al . [25] . Recombinant viral strain ZH501 was rescued as previously described [26] and the exact complete genome sequence confirmed by techniques described by Bird et al . [41] . Strain ZH501 was originally isolated from a fatal human case during the 1977 epidemic in Egypt and the complete genome sequences of the S , M , and L segments used in this work can be found under GenBank accession numbers DQ380149 , DQ380200 , and DQ375406 , respectively . Seventeen healthy adult marmosets ( Callithrix jacchus ) , 1 to 3 years old and ranging in weight from 257 to 398g were obtained from World Wide Primates ( S1 Table ) . None of these primates was exposed to any infectious pathogens in previous studies and all primates were determined to be RVFV naïve by plaque reduction neutralization test ( PRNT; methods below ) before the initiation of the study . For the study design , two groups of six RVFV seronegative marmosets were inoculated subcutaneously with 5 log10 PFU of the ΔNSs or ΔNSs/ΔNSm vaccine candidate and one group of five RVFV seronegative marmosets served as the sham-vaccinated controls . This vaccination dose was determined based on our previous studies in rodents and sheep [24 , 25] . All animals were monitored for weight loss/survival/clinical signs of infection and blood samples were collected from anesthetized animals on days -3 , 0 , 2 , 4 , 7 and once a week thereafter for virological , hematological , immunological , and chemical analyses . Body temperature was monitored rectally . Following vaccination , the marmosets were challenged subcutaneously with 6 . 4 log10 PFU of the virulent strain ( ZH501 ) 35 days post-vaccination . This dose was chosen based on our previous model development study [38] . Following challenge , all animals were monitored for weight loss/survival and blood samples were collected on days 0 , 2 , 4 , 7 and once a week thereafter for virological , hematological , immunological , and chemical analyses . The study endpoint was euthanasia when moribund or 28 days post-challenge ( lethal strain ) /day 63 of study . Marmosets were euthanized by being deeply anesthetized by injection of Telazol followed by exsanguination . Following euthanasia , a full necropsy was performed for collection of tissues . Whole blood was added to an EDTA tube ( Sarstedt , Numbrecht , Germany ) for complete blood count ( CBC ) determination using a Hemavet instrument ( Drew Scientific , Dallas , TX ) according to manufacturer’s instructions . Clinical chemistry analyses were performed by addition of whole blood to a lithium heparin tube ( Sarstedt ) using the comprehensive diagnostic panel analyzed on a Vetscan instrument ( Abaxis , Union City , CA ) according to manufacturer’s instructions . Normal ranges in the chemistry and hematology results of healthy marmosets [42] were used as reference values . The plasma was then collected for viral titer determination by quantitative RT-PCR ( qRT-PCR ) as previously described where the limit of detection ( LOD ) was 0 . 1 PFU [24 , 43] or by standard plaque assay as previously described [44] . A minimum amount of plasma remained ( less than 100 μL ) so the sample was diluted 10-fold prior to the plaque assay yielding a LOD of 100 PFU . For the qRT-PCR assay , a standard curve was generated using serial dilutions of the challenge virus in triplicate on the LightCycler 480 ( Roche Diagnostics , Inc . , Indianapolis , IN ) . The virus titers were calculated using the standard curve and the LightCycler 480 software , and the final PFU equivalents/mL ( PFUe/mL ) calculations were determined based on the sample input volume and the upfront sample dilutions . At the time of necropsy , the following tissues were collected for viral titer determination: liver , cerebrum , spleen , kidney , lung , heart , adrenal gland , inguinal lymph node , axillary lymph node , mesenteric lymph node , duodenum , jejunum , ileum , ovaries/testis , skeletal muscle , bone marrow , and retina . Tissues were collected , weighed , and homogenized in EMEM containing 5% fetal bovine serum and gentamicin . Tissues were homogenized using the Qiagen Mixer Mill 300 ( Retsch , Newtown , PA ) then centrifuged at 9 , 000 x g for 10 min and the supernatant stored at -70°C until further evaluation . Tissues collected at the study endpoint were homogenized according to the methods above and a 1:10 dilution of the supernatant added to 24-well plate of Vero cells in duplicate in a volume of 100 μL for each well . Plates were incubated for 1 h at 37°C with rocking every 15 min . After the incubation , 0 . 5 mL of EMEM was added to each well and incubated for 4 days to monitor for cytopathic effects ( CPE ) . Full necropsies and histological examination were performed by a board-certified veterinary pathologist . The following tissues were collected during necropsy: Axillary , inguinal , submandibular , mesenteric and tracheobronchial lymph node; submandibular salivary gland; haired skin; brachial plexus; sciatic nerve; skeletal muscle; bone marrow ( femur ) ; eyes; brain; pituitary gland; spleen; adrenal gland; kidney; liver; stomach; duodenum; pancreas; jejunum; ileum; cecum; colon; testis/ovary; prostate gland/uterus; urinary bladder; tongue; tonsil; trachea; esophagus; thyroid gland; lung; thymus; and heart . All collected tissues were immersion-fixed in 10% neutral buffered formalin for at least 21 days . The tissues were trimmed and processed according to standard protocol [45] . Histology sections were cut at 5 to 6 μm on a rotary microtome , mounted on glass slides , and stained with hemotoxylin and eosin . For immunohistochemical analysis , serial sections of tissue were cut and stained for RVF antigen using a mouse monoclonal antibody ( 4D4 ) against the glycoprotein Gn [46 , 47] and an immunoperoxidase assay system ( EnVision; DAKO ) . Normal hepatic tissue served as the negative control; the positive control tissue was liver from a known RVF-positive animal . Normal mouse IgG was used as the negative serum control for the control slides . For the immunohistochemistry study , the unstained tissue sections were deparaffinized , rehydrated , subjected to methanol-hydrogen peroxide block , rinsed , and pretreated with Tris/EDTA buffer at 97°C for 30 min . A serum-free protein block ( DAKO ) plus 5% normal goat serum was applied for 30 min . The primary antibody was then applied to the tissue at a dilution of 1:100 and incubated at room temperature overnight . The tissue sections were rinsed and then exposed to the EnVision horseradish peroxidase labeled polymer for 30 min at room temperature . All sections were exposed to DAB permanent chromogen for about 5 min , rinsed , counter-stained with hematoxylin , dehydrated , and applied a coverslip with Permount . Anti-RVFV total IgG ELISA was performed essentially as described previously [24] , with the following modifications necessary for NHP specimens . BHK cell lysate was used rather than Vero E6 cells and the secondary goat anti-monkey IgG horseradish peroxidase-conjugated antibody ( KPL , 074-11-021 ) , which was raised against rhesus macaques and most likely contributed to low adjusted sum optical density ( OD ) values . Neutralizing antibodies were assayed in plasma for marmosets with a 50% PRNT as previously described [48] . Repeated measures ANOVA was used to compare chemistry , viremia , weight , temperature , and antibody response over time and between groups . All analyses were conducted using GraphPad Prism 7 . 00 ( La Jolla , CA ) .
Marmosets ( n = 6/group ) were inoculated subcutaneously with 5 log10 PFU of the ΔNSs or ΔNSs/NSm rRVFV while five marmosets served as the sham-vaccinated controls . All animals were monitored for weight loss , survival , and clinical signs of infection and blood samples were collected for virological , hematological , immunological , and chemical analyses . All of the vaccinated and sham-vaccinated marmosets survived and no animals exhibited clinical signs of illness , experienced significant weight loss ( Fig 1 , S1 Fig ) or temperature changes ( Fig 2 , S2 Fig ) post-vaccination or post-challenge . Additionally , no animals experienced an adverse reaction at the site of vaccination . Viral RNA was detected on day 2 post-vaccination by qRT-PCR in 4/6 animals that received ΔNSs rRVFV and 5/6 animals that received ΔNSs-ΔNSm rRVFV ( Fig 3 , S3 Fig ) . Viral RNA was not detected in any of the animals on day 4 post-vaccination , but was detected for two animals ( one animal that received ΔNSs rRVFV and one animal that received ΔNSs-ΔNSm rRVFV ) on day 7 post-vaccination . These same samples were evaluated for infectious virus by standard plaque assay and no virus was detected . However , a minimum amount of sample volume was left and had to be diluted ten-fold prior to completion of the plaque assay thus reducing the sensitivity to detect infectious virus where the LOD was 100 PFU . Regardless , these results suggest that little to no infectious virus was present in vaccinated marmosets . When marmosets were challenged 35 days post-vaccination , only the marmosets that received the sham inoculation developed viremia as detected by qRT-PCR indicating that the vaccinated monkeys were protected . The samples with the highest viremia as determined by qRT-PCR were evaluated for infectious virus by standard plaque assay and an average of 5 . 4 log10 PFU/mL was detected on day 2 post-challenge . In our previous model development study we observed marked aberrations in hematological and chemistry values from marmosets exposed to RVFV . In particular , the liver enzyme ALT was significantly increased when marmosets were exposed to RVFV subcutaneously . We therefore collected blood samples post-vaccination and post-challenge for hematological and chemical analyses . Overall , no significant change in the hematology and clinical chemistry values of the vaccinated animals was observed post-vaccination or post-challenge . For example , the ALT levels ( Fig 4 ) in vaccinated animals were similar post-vaccination and post-challenge . As expected , the control animals did have an increase in ALT levels on day 2 post-challenge which suggest that the vaccinated animals were protected from RVFV-induced liver disease . We expected to see a change in other hematology and chemistry values in the sham inoculated controls post-challenge , but none were significantly different compared to baseline which is in contrast to what we observed in our previous model development study . All vaccinated animals developed neutralizing antibodies by day 14 post-vaccination ( Table 1 ) . These titers peaked by day 21 post-vaccination , which were slightly higher for animals receiving the ΔNSs rRVFV . The RVFV IgG titers peaked on day 35 post-vaccination and similar to the neutralizing antibody titer results , were slightly higher for animals receiving the ΔNSs rRVFV . However , there was no statistically significant difference in the neutralizing antibody response and anti-RVFV total IgG response for animals receiving the ΔNSs rRVFV vs . ΔNSs-ΔNSm rRVFV ( Table 2 ) . The only statistically significant antibody responses was observed in animals receiving either the ΔNSs rRVFV or ΔNSs-ΔNSm rRVFV vaccine compared to the sham-vaccinated controls . Overall the adjusted sumOD values were low ( including those for RVFV challenged controls ) , which is most likely due to the secondary antibody being raised against rhesus macaques and not marmosets ( a marmoset specific antibody does not exist ) . The neutralizing antibody and RVFV IgG titers increased for all animals post-challenge . The tissues were tested for viral RNA by qRT-PCR and all tissues from the vaccinated animals were negative except for the spleen and axillary lymph node of one animal receiving the rZH501-ΔNSs vaccine and the skeletal muscle and cerebrum of one animal receiving the rZH501-ΔNSs-ΔNSm vaccine ( Table 3 ) . All of the control animals had detectable viral RNA in multiple tissues ( primarily the lymphoid tissues ) . However , these RNA values were low in both the vaccinated and control animals and are most likely insignificant since no tissues were positive by IHC and infectious virus was not detected by cytopathic effect assay . Furthermore , histologic findings directly attributable to RVFV infection were not observed . In summary , both the ΔNSs and ΔNSs-ΔNSm vaccine candidates were found to be safe and immunogenic in the current study . The vaccinated marmosets exhibited no signs of clinical illness post-vaccination and post-challenge and developed strong neutralizing antibody titers . Additionally , minimal viremia as detected by qRT-PCR was observed post-vaccination and no viral RNA was identified in the serum of vaccinated animals post-challenge .
RVFV significantly impacts livestock and human health making it a good target for a one-health prevention approach through animal vaccination . Livestock vaccination during non-epidemic periods or as an early countermeasure against early outbreaks could eliminate one of the main sources of human infection and limit the scope of epidemics . However , previous RVFV outbreaks are generally recognized only after human cases are diagnosed [49 , 50] . Additionally , a human vaccine is still needed to protect veterinarians involved in vaccination programs , slaughterhouse workers and farmers . Human vaccination to protect the general public could be required if efficient spread of RVFV by anthropophilic mosquito species occurs . Finally , since RVFV is a potential agent of bioterrorism , a human vaccine is needed to protect against the threat posed by intentional dissemination . Currently , there is no fully licensed vaccine for veterinary or human use available in non-endemic countries . In endemic countries , there is no clear guidance for livestock vaccinations to prevent RVF outbreaks . Furthermore , previous veterinary vaccines for RVF has been plagued with numerous concerns such as high manufacturing costs , a poorly defined genetic identity , poor efficacy , no capacity to differentiate vaccinated from naturally infected livestock , and the risk of vaccination in pregnant animals due to associated teratogenesis and abortion [2 , 18 , 51 , 52] . Next-generation veterinary vaccines are being developed that overcome many of these limitations . However , regulatory and economic challenges continue to preclude the development of a human vaccine . Clearly , the licensing of both a veterinary and human vaccine is needed for RVFV . A logical strategy is to use a common approach for veterinary and human vaccine development with the goal to reduce development and licensing costs . Bird et al . developed rationally designed vaccine candidates based on the complete deletion of two known RVFV virulence factors , the NSs and NSm genes [24 , 25] . The rRVFV vaccine candidates containing the insertion of the enhanced green fluorescent protein and the precise deletion of the NSs gene alone ( ΔNSs:GFP rRVFV ) or the NSs/NSm genes in combination ( ΔNSs:GFP-ΔNSm rRVFV ) were found to be highly attenuated , immunogenic , and efficacious in the rat lethal disease model [24] . Importantly , a robust antibody response was observed with both vaccine candidates demonstrating that the double-genetic deletions of the entire RVFV NSs and NSm genes does not significantly decrease overall vaccine efficacy compared to the single-genetic deletion of the NSs . The design of a vaccine candidate with attenuating deletions on multiple virus genome segments provides enhanced safety by reducing the possibility of reversion to full virulence via either RVFV polymerase nucleotide substitution or gene segment reassortment with field strains . The insertion of the GFP gene was removed ( due to vaccine licensure concerns containing a foreign gene ) and the double-genetic deletion rRVFV was further evaluated in a natural RVFV host . This ΔNSs-ΔNSm rRVFV vaccine was demonstrated to be protective and immunogenic in sheep without producing clinical illness in these animals [24 , 25] . The vaccine was nonteratogenic in pregnant sheep , which is critical to demonstrate the safety needed for a veterinary vaccine in a natural RVFV host . Additionally , the ΔNSs-ΔNSm rRVFV vaccine was demonstrated to be compatible with a differentiation of infected and vaccinated animals ( DIVA ) enzyme-linked immunosorbent assay ( ELISA ) [24 , 25 , 53] . Here , we expand upon that work and completed an evaluation of the single deletion mutant ( ΔNSs rRVFV ) and double deletion mutant ( ΔNSs-ΔNSm rRVFV ) vaccine candidates in marmosets . We demonstrate that both the ΔNSs rRVFV and ΔNSs-ΔNSm rRVFV vaccine candidates were found to be safe and immunogenic in the current study . The vaccinated marmosets received 5 log10 PFU of virus and exhibited no signs of clinical illness , experienced no significant weight loss , or temperature changes post-vaccination and post-challenge . Additionally , minimal viral RNA was observed post-vaccination and no viral RNA was identified in the serum of vaccinated animals post-challenge . No significant change in the hematology and clinical chemistry values of the vaccinated animals was observed post-vaccination or post-challenge . In contrast , the liver enzyme ALT was significantly increased in sham-vaccinated control animals suggesting that the vaccinated animals were protected from RVFV-induced liver disease . Collectively , these results demonstrate the general safety of these vaccine candidates in NHPs . However , more extensive safety testing such as an assessment of neurovirulence would be necessary for advance development efforts . This would be especially important for the ΔNSs vaccine candidate which was shown to cause a uniform fatal encephalitis after intranasal , but not subcutaneous exposure in C57BL/6 mice [54] . A separate study utilizing another recombinant ZH501 RVFV strain lacking the NSs gene demonstrated that CD1 mice can occasionally develop encephalitis ( 5% mortality was reported ) after intraperitoneal exposure [55] . These studies suggest that additional attenuating mutations other than NSs may be important for the safety of RVFV vaccine candidates . The immunogenicity of the ΔNSs and ΔNSs-ΔNSm rRVFV vaccine candidates was noteworthy . All vaccinated animals developed high neutralizing antibody titers by day 14 post-vaccination , which peaked by day 21 post-vaccination . Antibody titers were slightly higher for animals receiving the ΔNSs rRVFV than animals vaccinated with the double-deletion ΔNSs-ΔNSm rRVFV , but this difference was not found to be statistically significant . The ΔNSs rRVFV may be slightly more immunogenic because of the single deletion in a known RVFV virulence factor compared to ΔNSs-ΔNSm rRVFV , which has two gene deletions . We would expect that the ΔNSs-ΔNSm rRVFV would be more attenuated presumably due to reduced in vivo virus replication and less stimulation of the antiviral immune response . However , we did detect similar levels of viral RNA in the blood on day 2 post-vaccination for both the single and double deletion viruses . It is possible that differences in the kinetics or magnitude of virus replication occur between the single and double deletion viruses that we didn’t detect with the current study design . However , even with the slight reduction in antibody titers all animals were completely protected by both vaccine candidates . Since the double-genetic deletions of the entire RVFV NSs and NSm genes does not significantly decrease overall vaccine efficacy , it makes sense to pursue this as the lead candidate for licensure . The ΔNSs-ΔNSm rRVFV is likely safer due to multiple attenuating lesions leading to a reduced possibility of reversion to full virulence . It is difficult to directly compare antibody titers as an indication of protective immunity to those of previous studies with other RVFV vaccines because of the differences in the candidates/approach , species level differences in immunity , and timing for assessing the response . However , a retrospective study of human volunteers ( n = 598 ) receiving a three-dose regimen ( days 0 , 7 , and 28 ) of inactivated TSI-GSD-200 vaccine reported that subjects developed a mean PRNT80 of 1:237 [17] . The live attenuated MP-12 vaccine was evaluated in rhesus macaques where vaccinated animals demonstrated PRNT80 values of ≥1:640 [19 , 56] . In the current study , the mean PRNT80 ranged from 1:6 , 400 to 1:8 , 267 on day 21 post-vaccination , indicating that the level of neutralizing antibody was substantially higher to that demonstrated in earlier studies of RVFV vaccines in NHP models or in human volunteers . However , it is difficult to directly compare antibody titers between various studies for the aforementioned reasons . The virulent virus challenge dose used in this study ( 6 log10 PFU/mL ) was chosen based on our previous model development effort , which indicated that we would likely see 50% mortality with the sham-vaccinated control animals . Surprisingly , no mortality was observed for sham-vaccinated animals , which may be a result of the age of the animals used in the current study which ranged from 1 to 3 years old . The age of the animals used in our previous model development effort were 2 to 11 years old with the majority of the animals being older ( between 8–11 years old ) . In fact , the animals that succumbed to RVFV by subcutaneous exposure were 10–11 years old [38] . Another possible difference in the model development study vs . the current study is the use of different sources for the animals , which may have resulted in different genetic backgrounds of the marmosets . While this is highly speculative , it is possible that genetics plays a role in the susceptibility to RVFV infection in NHPs , which has been demonstrated in the RVF rat model . For example , Peters and Anderson used breeding experiments to demonstrate that a dominant gene determines resistance to fatal RVFV-induced liver disease [57 , 58] . Clearly , more studies are needed to further characterize the RVF marmoset model and determine the likelihood for the development of severe disease . Despite the lack of mortality , the sham-vaccinated control animals did become viremic as detected by qRT-PCR and experience an increase in the liver enzyme ALT . In contrast , the vaccinated animals did not experience any adverse reactions and viral RNA was not detected in the serum . A previous study of the live attenuated RVFV vaccine MP-12 in rhesus macaques , which is also a non-lethal NHP model , observed post-vaccination viremia detected by plaque assay in 1/3 of vaccinated monkeys and included a slight increase in the liver enzyme AST [56] . Our results suggest that the complete deletion of the NSs and NSs/NSm genes affords a more attenuated phenotype , but still generates a robust antibody response . In summary , both the ΔNSs and ΔNSs-ΔNSm vaccine candidates have many desired features for human vaccine development . No vaccine-induced adverse reactions , signs of clinical illness or infectious virus were detected in the vaccinated marmosets . The vaccinated animals received a single dose of vaccine that led to the development of robust neutralizing antibody titers that provided complete protection against viremia and liver disease . Our results provide the basis for further development of the ΔNSs and ΔNSs-ΔNSm rRVFV as safe and effective human RVFV vaccines for this significant public health threat . | Rift Valley fever ( RVF ) is an important neglected tropical disease that has caused severe epidemics and epizootics throughout Africa and the Arabian Peninsula . Severe outbreaks have involved tens of thousands of both human and livestock cases for which no effective , commercially available human vaccines are available . Vaccine candidates have been developed based on the complete deletion of two known RVF virus virulence factors , the NSs and NSm genes . These vaccines were previously demonstrated to be protective in rats , mice , and sheep . In this study , we expand upon those results and evaluate the vaccine candidates in a non-human primate model for RVF . The animals received a single dose of vaccine that led to the development of a robust immune response . No vaccine-induced adverse reactions , signs of clinical illness or infectious virus were detected in the vaccinated animals . All vaccinated animals that were subsequently challenged with RVF virus were protected against viremia and liver disease . These results demonstrate that the vaccines are safe and effective in non-human primates , which provides the impetus for further development of these candidates for use in humans . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"rift",
"valley",
"fever",
"virus",
"immune",
"physiology",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"vertebrates",
"marmosets",
"animals",
"mammals",... | 2018 | Attenuation and efficacy of live-attenuated Rift Valley fever virus vaccine candidates in non-human primates |
Chagas disease is a major neglected tropical disease with deep socio-economical effects throughout Central and South America . Vector control programs have consistently reduced domestic populations of triatomine vectors , but non-domiciliated vectors still have to be controlled efficiently . Designing control strategies targeting these vectors is challenging , as it requires a quantitative description of the spatio-temporal dynamics of village infestation , which can only be gained from combinations of extensive field studies and spatial population dynamic modelling . A spatially explicit population dynamic model was combined with a two-year field study of T . dimidiata infestation dynamics in the village of Teya , Mexico . The parameterized model fitted and predicted accurately both intra-annual variation and the spatial gradient in vector abundance . Five different control strategies were then applied in concentric rings to mimic spatial design targeting the periphery of the village , where vectors were most abundant . Indoor insecticide spraying and insect screens reduced vector abundance by up to 80% ( when applied to the whole village ) , and half of this effect was obtained when control was applied only to the 33% of households closest to the village periphery . Peri-domicile cleaning was able to eliminate up to 60% of the vectors , but at the periphery of the village it has a low effect , as it is ineffective against sylvatic insects . The use of lethal traps and the management of house attractiveness provided similar levels of control . However this required either house attractiveness to be null , or ≥5 lethal traps , at least as attractive as houses , to be installed in each household . Insecticide and insect screens used in houses at the periphery of the village can contribute to reduce house infestation in more central untreated zones . However , this beneficial effect remains insufficient to allow for a unique spatially targeted strategy to offer protection to all households . Most efficiently , control should combine the use of insect screens in outer zones to reduce infestation by both sylvatic and peri-domiciliated vectors , and cleaning of peri-domicile in the centre of the village where sylvatic vectors are absent . The design of such spatially mixed strategies of control offers a promising avenue to reduce the economic cost associated with the control of non-domiciliated vectors .
Chagas disease , also called American trypanosomiasis , is caused by the protozoan parasite Trypanosoma cruzi , which is primarily transmitted to humans by blood-sucking bugs of the Triatominae subfamily . The disease is endemic throughout Latin America , where it is one of the most important parasitic diseases with large socioeconomic impact . According to various estimates , the prevalence rate in humans varies between 0 . 1 and 45 . 2% ( with an average of 1 . 4% ) , 8 to 15 million people are infected with T . cruzi ( with 40–50 , 000 yearly new cases ) , and 28–75 million individuals are at risk of infection [1]–[3] . The disease causes about 12 , 500 deaths a year , and is responsible for premature disabilities of workers that are estimated to cost 670 , 000 disability-adjusted life years lost [4] . Although international initiatives have been launched to reduce transmission of Chagas disease , especially through vector control and screening of blood or organ donors [5] , there are still large regions with active vector transmission [6] . One of the main explanations for this is the transmission caused by non-domiciliated triatomines [7] . These vectors are not able to reproduce and develop in the domestic habitat , and thus constitute typical ‘sink’ domestic populations sustained by peri-domestic and/or sylvatic ‘source’ populations [8] . Non-domiciliated vectors tend to jeopardize the efficacy of vector control by insecticide spraying in the domestic habitat because of the re-infestation of treated houses [9] , [10] , [11] . This situation has been described for several vector species of triatomines as T . brasiliensis and T . pseudomaculata in Brazil [12] , T . mexicana in central Mexico [13] and T . dimidiata in the Yucatan Peninsula of Mexico and Belize [14] , [15] . Accordingly , the risk of transmission associated with non-domiciliated vectors is now identified as a major challenge for the future of Chagas disease control [16] , [17] , [18] , and a key objective is to evaluate the efficacy of classical or alternative control strategies to reduce their abundance . Identifying optimal strategies can hardly be achieved through laboratory or field experiments , since testing a broad enough number of alternatives would require very large human and financial investments [11] , [19] . Alternatively , mathematical models have proven to be very effective at evaluating the relative merit of various alternative strategies to control parasitic diseases [11; and references therein] . In addition , identifying optimal strategies clearly requires a detailed understanding of the vector spatial and temporal infestation dynamics . Valuable insights into such spatio-temporal dynamics can be gained using the framework of meta-population theory combined with presence/absence data [19]–[21] . Although appealing , the use of more elaborated models that include quantitative information on local population sizes requires even more data than the meta-population model sensus stricto [22] . In previous contributions , we developed spatially explicit population dynamics models that were able to reproduce and to predict the spatial and temporal dynamics of T . dimidiata house infestation observed at the village scale in the Yucatan Peninsula , Mexico . These models provided us with indirect estimates of the origin and characteristics of dispersal of these triatomines [23] , [24] . Individuals found inside houses in the Yucatan Peninsula originated in similar proportions from both sylvatic and peri-domestic habitats , dispersed over rather small distances ( 40–60 m per displacement ) and were strongly attracted to houses [24] . Remarkably , the observed and predicted dynamics showed an heterogeneity in transmission risk both in time , with a peak of vector abundance during March–June [14] , [25] , and in space , with much higher abundance of insects in the periphery of the village reflecting the influence of the sylvatic habitat [11] , [26] . The temporal optimization of insecticide spraying with respect to this pattern has already been investigated at the scale of one house [11] , but the spatial micro-scale heterogeneity suggests that interventions could also be spatially targeted . Such interventions would focus on the periphery of the village , where bugs were found more abundant . While temporal heterogeneity adds constraints on control strategies ( i . e . the timing of intervention has to match the seasonality of house infestation , [11] ) , spatial heterogeneity could have beneficial consequences for control activities as it might allow to reduce the overall surface ( or number of houses ) to be treated and thus allow to reduce the cost associated with control . Properly assessing whether such spatial design is relevant requires evaluating not only the efficacy of control in the treated areas , but also the impact of the control interventions in the untreated areas of the same village . In this contribution , we aimed to build on our understanding of the temporal optimization of control strategies [11] , as well as our previous spatial modelling [23] , [24] to evaluate the potential of several strategies . We first focus on conventional strategies — namely indoor insecticide spraying , use of door/window insect screens and peri-domicile management — that have been used to control vectors of different diseases as well as T . dimidiata [27]–[29] . We further look at the potential of insect lethal traps that are currently extensively investigated for the control of a variety of vector species [30] , [31] . Finally , since we have previously found that T . dimidiata was directly attracted to houses [24] , a control alternative could be to eliminate this house attractiveness , and the potential of such a strategy was also explored .
We aimed to set up a spatial population dynamics model able: ( 1 ) to reproduce and predict the temporal variations of vector abundance in all the houses of one village in the absence of control , and ( 2 ) to spatially represent various control strategies . We adapted previous population dynamic models [23] , [24] , and combined them with a mathematical description of the control strategies that we aimed at evaluating . The resulting model predicts the temporal variations in vector abundance in every house of the village as a function of survival , reproduction and dispersal of the triatomines , and the effect of the above control strategies on the demographic processes at each point of the village . It was then used for the evaluation of the efficacy of spatially targeted interventions based on each of those strategies . Model predictions in absence of control were fitted through a maximum likelihood approach to a first set of spatio-temporal data describing house infestation dynamics by T . dimidiata within a village in the absence of vector control . We tested the predictive value of the resulting parameterized model on a replicate data set , corresponding to the infestation dynamics observed in the same village the following year . The description of the effect of the different control strategies was then added to the model , and the resulting framework was used to explore the efficacy of control interventions whose spatial coverage was progressively increased from the border to the centre of the village . The efficacy of each intervention was evaluated as the percentage of reduction in the yearly abundance of vectors in the village , in comparison with the expected abundance in the absence of control intervention that we evaluated from the model with no control . Efficacy was also related to the consented effort , as measured by the number of households , where control strategies were applied ( either in the domestic or peri-domestic habitats ) . We performed a sensitivity analysis to each survival , reproduction or dispersal parameter of the model to ensure the robustness of our conclusions on the efficacy of the various interventions within the confidence region associated with the maximum likelihood estimate of model parameters . We further conducted a sensitivity analysis to different parameters of the model that described the efficacy of each of the strategies as measured by their impact on the survival , reproduction or dispersal of triatomines . The spatio-temporal pattern of house infestation was observed in the rural village of Teya , Yucatan , Mexico over a two-year period from August 2006 to October 2008 [26] . All houses were identified and geo-referenced with a handheld global positioning system ( GPS ) . Insects were collected by a standardized methodology based on community participation [32] , and data were imported into a geographic information system ( GIS ) database ( ArcView 3 . 2 -Environmental Systems Research Institute , Redlands , CA , USA ) to produce maps of observed triatomine abundance in the houses over 2-week intervals [32] . Participating families provided oral consent prior to their participation , as written consent was waived because the study involved no procedures for which written consent is normally required outside of the research context . Consent was logged in field notebooks . All procedures , including the use of oral consent , were approved by the Institutional Bioethics Committee of the Regional Research Centre “Dr . Hideyo Noguchi” , Universidad Autonoma de Yucatan . We set up a GIS-based Spatially Explicit Model ( GIS-SEM ) as such modelling provides a suitable framework to investigate spatial population dynamics in real landscapes by importing GIS data on a grid representing the area under study [33] . Our GIS-SEM model was based on Cellular Automaton ( CA ) formalism [34] . It consisted of a grid of cells representing the village of Teya , and allowed the calculation of the temporal variations of the vector abundance in cells , referred to as state variables , according to both local rules describing birth and death processes of bugs within cells , and dispersal rules that allow accounting for walking and flight movements between neighboring cells . This model was similar in essence to the models built by Barbu et al . [24] , but with two necessary adaptations . First , the local and dispersal rules were described in a deterministic rather than stochastic manner to reduce the complexity of the model and shorten the simulation time . Second , the time unit of the model was changed from 15 days to a day to allow specifying the effect of control on a daily basis . A deterministic CA such as the one intended here is defined as a quadruple Q = ( A , S , V , f ) , where A is the grid of cells arranged uniformly to represent the studied area; S is the set of values that can be taken by the state variables; V is the neighborhood function that allows identifying the set of neighboring cells V ( c ) that contribute to the change of the state variable of any given cell c by the mapping: ( 1 ) with v denoting the size of the neighborhood; and where f is the function describing the local and dispersal rules and thus specifies how the set of neighboring cells V ( c ) changes the state of the cell c from one time step to another: ( 2 ) with N ( c , t ) , the state variable that tracks the status of cell c at time t . Maximum likelihood estimates ( MLE ) of the parameters of the model with no control were obtained using the spatio-temporal data sets describing T . dimidiata infestation dynamics of the village of Teya between mid-September 2006 and mid-September 2007 . Model predictions were fitted to the observed number of bugs in each cell of the 24 maps describing the average biweekly distribution within the village . The log likelihood ( LLH ) value was then calculated as follows: ( 6 ) where log denotes the natural logarithm , X ( c , t ) is the statistical variable corresponding to the number of adults in cell c , O ( c , t ) the observed abundance in this cell , and θ is a set of parameters of the model . Probabilities were defined assuming a zero-inflated Poisson distribution to take into account an excess of null abundance in the data set [39] , possibly due to the non-participation of a proportion ( w ) of householders , with w = 0 . 7 as before [24] . The parameters θ of the model were identified using a genetic algorithm run at the super-computing centre ‘Institut du Développement et des Ressources en Informatique Scientifique ( IDRIS ) ’ located at Orsay , France ( http://www . idris . fr/ - Project IDRIS 112290 ) . Genetic algorithms search for solutions using techniques inspired by natural evolution . The interested reader can find a detailed description of such methods and the typical terminology we adopted below in [40] . The algorithm considered the 8 parameters of the model ( Sd , Sp , d , Kp , Ks , D , σ and H ) to be estimated as independent quantitative traits with a continuum of alleles representing possible trait values within biologically relevant domains . The fitness function corresponded to the LLH value defined with respect to the GIS-SEM model with no control described above . The fittest individuals were selected to produce offspring through free recombination and unbiased mutations . The variance of the effect of the mutations was dynamically adapted to the variance in the parental population . All codes were written in C/MPI . Confidence intervals were calculated by establishing the profile likelihood for each parameter , and by using these relationships to determine the 1−α confidence region defined as: ( 7 ) where is the MLE of parameter and stands for the ( 1−α ) th quantile of the distribution on 1 degree of freedom [41] . The ability of the parameterized model to predict other infestation dynamics was tested by comparing its prediction to the spatio-temporal distribution of bug abundance in a second year of infestation of the same village . A Poisson regression between observed and predicted abundances was performed after data were pooled over 3-month periods ( starting in mid-September ) and within three distance categories: 0–80 m , 81–200 m and >200 m from the bush area outside the villages [24] , [26] . The McFadden's likelihood ratio index was used as a pseudo R-squared . Because the spatial distribution of bugs follows a spatial gradient with higher abundance at the periphery of the village [24] , [26] , the control strategies were applied to a ring of cells located at the border of the village , the size of this ring increasing progressively until the intervention covered the whole village ( Figure 1 ) . The efficacy of any given spatially targeted strategies was measured in terms of yearly bug abundance both in the whole village and in the different concentric rings . This allowed us to quantify the relationship between the effort in terms of control coverage and the global efficacy , and to simultaneously assess the consequences of interventions in the various parts of the village . The efficacy of intervention was evaluated using the set of parameters' estimates providing the best fit to the data . It was complemented by a sensitivity analysis of the corresponding results to the parameter's estimates . Each parameter was then independently set to the boundary values of its confidence interval , i . e . and , while keeping the others to their MLE . We evaluated the efficacy of five types of control strategies applied individually , including indoor insecticide spraying , door and window insect screens , peri-domicile cleaning , triatomine lethal traps located in the peri-domestic habitat , and housing improvement to reduce house attractiveness to bugs . The effect of each strategy on bug survival , reproduction and/or dispersal was modelled as described below ( see also supplementary methods — Text S1 — for the mathematical changes that were made to the model to include these effects ) . Indoor insecticide spraying was modelled by reducing vector survival in each treated house as before [11] . The control-induced mortality was calculated with respect to the residual dose of insecticide that we adjusted daily , and to the lethality of the dose as expected from a typical sigmoid dose-response relationship . Assuming that the control-induced and natural mortalities act independently ( i . e . to survive one of the two causes of death does not affect the probability to survive the second one ) , we combined them multiplicatively to define the overall survival probability . We considered a spray rate of 50 mg . m−2 of pyrethroid insecticide at the beginning of the infestation season ( since it was previously shown to be the optimal timing for spraying [11] ) , the half-life of the insecticide was set to 38 days , and the lethal doses 50% and 90% were fixed to 32 . 2 mg . m−2 and 182 . 4 mg . m−2 [11] . A sensitivity analysis to insecticide dose was performed predicting the effect of spraying at 100 , 200 and 300 mg . m−2 . Door and window screens were considered as physical barriers impeding the arrival of a proportion of the non-domiciliated vectors into the domestic habitat , and were thus modelled by lowering immigration into the houses by a factor of bug exclusion r set at 85% and constant over time [11] , [42] . Again , a sensitivity analysis was conduced by considering r equals to 70 , 80 and 90% . Because the efficacy of screens is likely to depend on the behavioral response of dispersal bugs failing to enter houses because of screens , and because no information was available in the literature about such a response , we considered three alternative assumptions . Bugs that could not enter into houses were considered: ( 1 ) to stop dispersing and die , or ( 2 ) to stop dispersing for one day before starting again with no learning in their dispersal behavior ( and thus possibly attempting to enter the same house ) , or ( 3 ) to go on dispersing while avoiding the house they could not enter . Peri-domicile cleaning was assumed to eliminate all bug colonies established in this habitat for the rest of the current year . This reduced immigration from the cleaned sites , but did not have any effect on individuals that originated from other areas and may pass through the peri-domiciles where this control strategy was applied . In addition , we performed a sensitivity analysis by considering that cleaning removes only 60% and 80% of insects established in the peri-domestic habitat . Manipulation of houses' attractiveness to bugs was achieved by decreasing H from its estimated value to 1 , the value for which houses are no more attractive than the peri-domestic and sylvatic habitats . This represents the strongest possible effect and allows evaluating the maximal potential for this strategy; a sensitivity analysis for the intermediate values of H was then performed . Triatomine lethal traps in the peri-domestic habitat were assumed to attract and kill triatomines into the cells where they are positioned according to an additional parameter Htrap that measured the trap attraction . As for the study of the control of houses' attractiveness we first wanted to evaluate the maximal potential of this strategy . The density of traps was then fixed at 2 traps per household , and attraction was set to a constant level Htrap = 12 , almost twice the attraction of houses . Sensitivity analysis was then performed for different density of traps , in the range 5 traps per household to 1 trap for 10 households , and trap attraction , in the range 1 to 50 .
The model predictions fitted very well the yearly spatio-temporal dynamics of infestation observed in the village of Teya between mid-September 2006 and mid-September 2007 . The correlation between observed and simulated spatio-temporal data indicated that the model reproduced well both the seasonal variations in triatomine densities , and the spatial spread of bugs from the border to the centre of the village ( Figure 2A , McFadden's likelihood ratio index = 0 . 93 ) . Importantly , the model parameterized with the data on this first year was able to predict the observed spatial and temporal dynamics of bug abundance in the following year ( Figure 2B , McFadden's likelihood ratio index = 0 . 67 ) . We note that while our model tends to predict well high abundances , predictions at lower vector abundances seem less precise . However , this is rather inconsequential since predicting fine variations in space and time at low abundances is of little relevance for our ultimate objective of evaluating control strategies . The convergence of the presented results with a previous study , that used a stochastic model [24] , also showed that the selected local and dispersal rules ( see Definition of the function f including the local and dispersal rules ) were reliable in their ability to both reproduce and anticipate the spatio-temporal dynamics of these non-domiciliated vectors . Likelihood profile confidence intervals gave further information on the estimated parameters of these rules ( Table 1 ) . Those confidence intervals were quite narrow around the MLE . The lower and upper boundaries were typically located at less than 30% of the MLE of each of the parameters , indicating that larger changes in one of the parameter estimates would no longer allow properly reproducing the data . The survival rates in the domestic and peri-domestic habitats were very close to 0 . 2 and 0 . 9 , respectively; the numbers of insects immigrating from the colonies established in the sylvatic and peri-domestic habitats were in the range 150–260 insects for 15 days; there was nearly a 1∶1 ratio between immigration from the sylvatic and peri-domestic habitats; the attraction to the house was always at least 5 times higher than attraction to the peri-domestic area , and the optimal ( and mean ) distance of dispersal was between 50 and 60 meters ( Table 1 ) . All of those results were consistent with and supported our previous conclusions that insects found in houses came in roughly similar proportion from the sylvatic and peri-domestic habitats and that they disperse over rather small distances and with a strong attraction to the domestic habitat [24] . Overall , our spatial model with no control thus offered a good framework where spatially targeted control strategies could be evaluated . We investigated the efficacy of the five strategies considered independently by applying them to concentric rings defined from the border of the village and whose size was increased until a complete coverage of the village was reached . For each strategy , we calculated its efficacy , measured as the post-intervention reduction of bugs' abundance in the whole village , in function of the extent of village zones treated , i . e . the effort in terms of the control intervention ( Figure 3A–E ) . We also calculated the effect of the interventions in each concentric village area , including those without control intervention ( Figure 3F–J ) . Finally , we performed a sensitivity analysis to the parameter values by independently replacing the MLE with the upper and lower values of each profile likelihood based confidence interval ( Table 1 ) . The first key point is that all the results obtained with each of the five strategies were only weakly sensitive to changes in demographic parameters values . Such changes indeed lead to no qualitative change in the form of the relationships ( Figure 3 ) . As expected , parameters with the strongest effect depend on the control strategy considered . Maximal changes were obtained when changing survivals ( Sp , Sd ) for insecticide spraying , immigration rates ( d ) for screens and outdoor traps , houses' attraction ( H ) for the control of houses' attractiveness and the number of individuals leaving colonies ( Kp , Ks ) for peri-domestic cleaning ( results not shown ) . However , these effects were systematically lower than 5% on both treated ( Figure 3A–E ) and untreated areas ( Figure 3F–J ) . The results obtained are thus very robust to variations of the parameters of the model with no control , and we will thus further describe only the results obtained with the MLEs . Indoor insecticide spraying in the whole village allowed the reduction of total bug abundance over a year by about 70% for one year ( Figure 3A ) . The relationship between the proportion of treated houses and global efficacy was a slightly convex diminishing return curve , so that half of the maximal decrease could be obtained by spraying only the first two external zones of the village ( a third of the houses ) . We also evaluated the local efficacy of insecticide in untreated village zones at the forefront of the treated areas . Independently of the number of village areas sprayed , the use of indoor insecticide only reduces the vector abundance in the treated area; it has a negligible effect on neighboring untreated areas ( Figure 3F ) . To increase the dose applied allowed the predicted levels of vector reduction to reach higher levels ( doses of 100 , 200 and 300 mg . m−2 lead to a 79% , 85% and 87% maximal control efficacy , respectively; data not shown ) , with no change in the main conclusion: Insecticide spraying in only the first two outer zones allowed for half of the maximal control efficacy . Door and window insect screens applied to all the houses of the village decreased the total vector abundance by about 80% when bugs that could not enter into houses were assumed to go on dispersing ( assumptions 2 and 3 , the former including possible attempts at entering again the house they just failed to infest ) ( Figure 3B ) . As for insecticide spraying , there was a slightly convex diminishing return between the number of treated zones and efficacy . Accordingly , limiting the intervention to the first two zones at the periphery of the village ( a third of the village houses ) again led to half of the maximal reduction in abundance . Under the two assumptions not including the death of the insects failing to enter the houses [2]–[3] , the analysis of insect screens' local efficacy indicated that while infestation was well controlled in houses with screens , the control had a detrimental effect on the immediate non-equipped neighbor: an increase of up to 40% in vector abundance was estimated in the most proximate untreated village zone ( Figure 3G ) . This negative effect on neighboring areas disappeared for untreated areas more than 3 zones away from the treated one . On the other hand , when the vectors were assumed to die when failing to enter a house ( assumption 1 ) , the effects of screens were significantly different . In this case , vector abundance was reduced slightly further ( up to 90% ) when screens were used in all the houses of the village ( Figure 3B upper dotted black line ) , and the control strategy then had no negative effect on untreated neighboring houses ( Figure 3G upper dotted black line ) . To vary the efficacy of screens produced only small linear changes in the global efficacy . Under assumptions 2 and 3 , a reduction factor r of 70% , 80% and , 90% lead to a 51% , 64% and 80% maximal control efficacy , while under assumption 1 , a reduction factor r of 70% , 80% and 90% led to a 73% , 82% and 91% maximal control efficacy; data not shown . The above conclusions are consequently very robust to variations of r , which is thought to be in the range 80–90% in the field [42] . Peri-domicile cleaning reduced total bug abundance by up to 62% for one year when performed in the whole village ( Figure 3C ) . The increase in efficacy with increasing coverage was a concave relationship with a slightly increasing return . Because of the lower efficacy of peri-domicile cleaning at the periphery of the village , intervention in at least the first 3 zones ( 60% of the village peri-domestic surface ) was required to reach half of the maximal reduction in abundance . Interestingly , when peri-domicile cleaning was performed only in some parts of the village it had an important beneficial effect on untreated neighboring houses . The vector abundance in the two closest non-treated zones was reduced by 40% and 15% respectively ( Figure 3H ) . Lowering the rate of colonies' destruction by peri-domicile cleaning , which was initially set to 100% , lowered the total efficacy in an almost perfectly linear way , but again had no effect on the above qualitative conclusions . Typically , assuming that only 80% or 60% of colonies are removed by cleaning peri-domiciles allowed for a maximal control efficacy of 50% ( ≈62%×80% ) and 37% ( ≈62%×60% ) , and in both cases intervention in the first 3 zones was needed to get half of these outcomes . Manipulation of houses' attractiveness was found 60% effective when applied to the whole village and when such attraction was completely eliminated , so that domestic habitat was no more attractive than the peri-domestic and sylvatic habitats ( H = 1 ) ( Figure 3D ) . Half of the maximal efficacy could be reached by an intervention targeted on the first two zones of the village representing a third of the village houses . However , such strategy had an important negative impact on the abundance of bugs in non-manipulated neighboring houses when applied to parts of the village ( Figure 3I ) . Indeed , the lack of attraction of manipulated houses resulted in an increase of over 50% and 30% in bug abundance in the next two untreated village zones . Importantly , sensitivity analysis of intermediate values of reduction in house attractiveness indicated that efficacy of the intervention was rapidly lost as H was incompletely reduced: the maximal efficacy was of 40% , 17% and less than 5% , for H values of 2 , 4 , and 6 , respectively ( Figure 4 ) . Insect lethal traps were found potentially able to reduce global vector abundance by up to 72% when considering a high density ( two traps per household ) and a high attractiveness ( Htrap = 12 , nearly twice the attractiveness of houses ) and 100% of lethality ( Figure 3E ) . Under these conditions , an important diminishing return was observed since to install traps in the first zone of the village ( 27% of the peri-domestic surface ) allowed to attain half of the maximal efficacy . Furthermore , this strategy had substantial positive effects on the 4–5 neighboring areas without traps , where insects' abundance was decreased by 50% , 30% , 15% and 7% , respectively ( Figure 3J ) . However , reducing the attraction factor of each trap had an important effect at the village scale as the global control went down from 72% to 55% when attraction of individual traps was reduced from Htrap = 12 to Htrap = 5 , a value similar to the attractiveness of houses ( Figure 5 ) . On the contrary , to increase attraction to higher levels had almost no effect whatever the number of traps considered . To lower the number of traps also had a strong detrimental effect , and the reduction of bug abundance due to control was never found larger than 30% when the number of traps was dropped to 1 trap for 10 households ( Figure 5 ) . A nearly 100% control efficacy at the village scale was reached only when more than 2 , 500 traps were used in the village , which represent about 5 traps per household within the village .
Although the elimination of transmission of Chagas disease was targeted by the WHO for the year 2010 [4] , there are still large regions with active vectorial transmission mostly due to non-domiciliated triatomines [6] . These vectors do not constitute permanent colonies inside houses , so that domestic populations actually are typical ‘sinks’ sustained by peri-domestic and/or sylvatic ‘source’ populations [8] . The risk of transmission associated with these non-domiciliated vectors is thus now identified as a major challenge for the future of Chagas disease control [5] , and a key objective is to evaluate the efficacy of classical or alternative control strategies to reduce their abundance . Since non-domiciliated insects infesting houses typically come from the sylvatic and peri-domestic habitat [11] , to evaluate the potential of various strategies requires a good understanding of the village infestation dynamics in absence of control . In this perspective , spatial population dynamic models able to reproduce and predict the dispersion of individuals from these two non-domestic habitats are valuable tools . Taking advantage of previous field and modelling works on well-studied populations of non-domiciliated triatomines in villages of the Yucatan Peninsula , Mexico , we performed the first attempt to evaluate the efficacy of putative control strategies applied spatially . We identified triatomines' dispersal characteristics through a selection model approach based on maximum likelihood estimates [37] , [38] . The best deterministic model and the associated estimates of the dispersal characteristic identified here were found very similar to the ones identified in a similar approach , but based on stochastic models [24] . In addition , just as the previous more complex stochastic model , our deterministic model reproduced and predicted very well the spatio-temporal dynamic of the village infestation . The present study thus confirmed that the selection model approach is a well-adapted strategy to simultaneously obtain indirect estimates of triatomines dispersal , hard to quantify in the field [19] , and robust GIS-based Spatially Explicit Models ( GIS-SEM ) able to reproduce and predict the dynamic of infestation in the absence of control . Such a model is required for the evaluation of the efficacy of putative control strategies; to this end we combined our selected model with a representation of different strategies to evaluate their potential . We found that indoor insecticide spraying and insect screens applied to the entire village were able to reduce yearly vector abundance in the whole village by 70 and 80% . Interestingly , in both cases , half the maximal effect was obtained while interventions were limited to the first two outer zones of the village . This mostly reflected the higher abundance of insects typically found in houses in the periphery of the village , where the vectors dispersing from both the peri-domestic and sylvatic habitats contribute to domestic infestation [23] , [24] , [26] . Although global efficacy was roughly similar for these first two strategies , a possible difference between them could be on their effect on untreated neighboring households . Indeed , insect screens were shown to impose some additional infestation on nearby untreated houses when vectors were allowed to go on dispersing after failing to infest a protected house . However , this negative effect was not present when vectors were assumed to systematically die after their first attempt to infest a protected house . Interestingly , the latter scenario is qualitatively consistent with a field trial conducted in a village of the north of the Yucatan Peninsula , in which the use of impregnated curtains and windows screens in some houses seems to reduce bug abundance in nearby untreated houses [42] . This may be due to some knockout effect of the low dose insecticide used for impregnation , or to a poor energetic status and/or exhaustion of bugs that could prevent re-departure after a flight/walk to intent infest a first house . Particularly in these conditions , and even if more empirical and modelling studies are needed to quantify vector dispersal at the individual scale , our results do support the idea of a spatially targeted use of insect screens to control the higher bug abundance at the periphery of the village as it maximizes the overall reduction in transmission risk at the level of the entire village . The most cost-efficient intervention would then be to treat the houses located in the first two outer zones ( about 33% of the total houses of the village ) to obtain around 50% bug abundance decrease in the entire village . The weak effect of insecticide spraying on the neighboring houses shown in this study is also consistent with a field trial [42] . Treating the first two outer zones would allow obtaining about 40% decrease in total bug abundance in the entire village but with no efficiency on untreated areas . Those results suggest that the cost associated to the temporary effect of insecticide spraying on non-domiciliated vectors demonstrated at the house scale [9] , [11] , can only weakly be compensated for by spatially targeted strategies that would exploit the typical gradient of abundance due to the immigration of sylvatic bugs [26] , [43] . Peri-domicile cleaning appears to be an interesting alternative strategy having the potential to substantially reduce vectors abundance inside the treated zones and to exert a positive influence on untreated areas . By eliminating all the colonies established in the backyards , a perfect cleaning of the peri-domiciles provided a 60% reduction of bug abundance in the village , although it provided a substantially lower efficacy at the periphery of the village , compared to the efficacy of residual insecticides and insect screens . This lower predicted efficacy at the village scale and in the outer zones is due to the absence of impact of this strategy on insects dispersing from the sylvatic habitat . It is also consistent with previous estimates indicating that infesting bugs come from both peri-domestic and sylvatic sites , and that both sources need to be controlled [24] , [44] . Interestingly , the positive effect of this strategy on nearby households with no intervention confirmed results of a previous field trial where peri-domicile cleaning ( elimination of unnecessary objects of the peri-domicile followed by insecticide spraying ) also reduced infestation in neighboring houses without intervention [42] . Accordingly , peri-domicile cleaning could valuably be used to significantly reduce bug abundance , especially in the centre of the village where the majority of non-domiciliated vectors found in houses come from the peri-domestic habitat [24] . Because it targets specifically peri-domestic vectors , such a strategy could lead to a substantial level of control when combined with insect screens in the periphery of the village . The manipulation of house attractiveness was explored here as a potential novel vector control intervention based on the rationale that triatomines were found to be directly attracted to the houses [24] . We found that such an intervention could reduce domestic infestation by up to 60% when applied to the entire village . However , when applied to only a fraction of the houses , we show that it would induce an increased infestation of neighboring untreated areas as bugs no longer attracted to manipulated houses tend to disperse to nearby domiciles . Control intervention based on this strategy should thus preferentially be implemented in all the houses of the village , and feasibility would then rely on the kind of modifications to be done in the domestic habitat to limit attraction . The actual determinants for house attractiveness to bugs are still unknown , but if light is proven to be a key factor [28] , [45] , [46] , the use of devices limiting the diffusion of the light may be considered . Nevertheless , it is important to emphasize that the effect of the intervention is rapidly lost if the reduction in the attractiveness is only partial . This strategy would thus be of little interest if a nearly complete reduction in house attractiveness to bugs could not be achieved . Thorough research on the mechanism and factors of triatomine dispersal toward houses would then be needed to allow the implementation of such a strategy . Triatomine lethal traps were also tested in an attempt to keep bugs away from the houses . Such traps were estimated effective if they were highly attractive and lethal , and used at very high densities; in these conditions they would also have a marked beneficial effect on neighboring houses without traps . The attractiveness of potential traps such as yeast-baited traps is difficult to estimate , but available studies suggest an attractiveness H in the range of 2–3 , i . e . rather less that the attractiveness of houses evaluated at 6–7 [47]–[49] . In such conditions , the use of 5 traps per household , which would represent about 2 , 500 traps in the whole village , only allows for about 30% reduction of triatomines abundance in the village . In addition , traps were assumed to be of constant efficacy in our model , which seems to be highly unlikely in practice , as it would raise the issue of the periodic maintenance/renewal of the traps depending on their half-life . It thus seems that the performance of potential outdoor traps would need to be dramatically improved to become a viable strategy for non-domiciliated triatomine control . Overall , this study has shown that control strategies applied at the periphery of a village can contribute to reduce infestation in untreated , more central houses , but only in limited proportions . Typically , insecticide or insect screens used in the first two outer zones of the village , which represents 33% of the households , would only reduce vector abundance in the whole village by 40–50% . In these conditions , spatial targeting of strategies based on either insecticide spraying or insect screens applied to houses in the two outer zones of a village , combined with peri-domicile cleaning in the centre , would provide optimum vector control at the lowest cost ( Table 2 ) . Essentially , such mixed strategy would remove peri-domestic colonies where they are the major source of vectors , and impede the insects to enter houses where they also come from non-manageable sylvatic colonies . At first , the costs of combining insecticide spraying or insect screens with peri-domicile cleaning seem roughly equivalent . However , the seasonal pattern of house infestation requires , for insecticide spraying , the dispatch of a large number of spraying teams to cover an entire region within 2 months [11] , generating additional costs of transportation and logistics [27] . Thus , a combination of insect screens in the periphery and peri-domicile cleaning in the centre would be the most cost-effective and sustainable strategy to be implemented in the Yucatan Peninsula . The design of such spatially mixed strategies of control offers a promising avenue to reduce the economic cost associated to the repeated intervention intrinsically associated with the permanent re-infestation of houses by non-domiciliated vectors [9] , [11] . | Chagas disease is one of the most important parasitic diseases in Latin America . Since the 1980's , many national and international initiatives have contributed to eliminate vectors developing inside human domiciles . Today's challenge is to control vectors that are non-adapted to the human domicile , but still able to transmit the parasite through regular short stay in the houses . Here , we assess the potential of different control strategies applied in specific spatial patterns using a mathematical model that reproduces the dynamic of dispersion of such ‘non-domiciliated’ vectors within a village of the Yucatan Peninsula , Mexico . We show that no single strategy applied in the periphery of the village , where the insects are more abundant , provides satisfying protection to the whole village . However , combining the use of insect screens in houses at the periphery of the village ( to simultaneously fight insects dispersing from the garden and the forest ) , and the cleaning of the peri-domicile areas of the centre of the village ( where sylvatic insects are absent ) , would provide a cost-effective control . This type of spatially mixed strategy offers a promising way to reduce the cost associated with the repeated interventions required to control non-domiciliated vectors that permanently attempt to infest houses . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"ecology/spatial",
"and",
"landscape",
"ecology",
"infectious",
"diseases/protozoal",
"infections",
"ecology/theoretical",
"ecology",
"ecology/population",
"ecology",
"infectious",
"diseases/epidemiology",
"and",
"contro... | 2011 | Evaluation of Spatially Targeted Strategies to Control Non-Domiciliated Triatoma dimidiata Vector of Chagas Disease |
Rhodnius prolixus is the main vector of Chagas disease in Venezuela . Here , domestic infestations of poor quality rural housing have persisted despite four decades of vector control . This is in contrast to the Southern Cone region of South America , where the main vector , Triatoma infestans , has been eliminated over large areas . The repeated colonisation of houses by silvatic populations of R . prolixus potentially explains the control difficulties . However , controversy surrounds the existence of silvatic R . prolixus: it has been suggested that all silvatic populations are in fact Rhodnius robustus , a related species of minor epidemiological importance . Here we investigate , by direct sequencing ( mtcytb , D2 ) and by microsatellite analysis , 1 ) the identity of silvatic Rhodnius and 2 ) whether silvatic populations of Rhodnius are isolated from domestic populations . Direct sequencing confirmed the presence of R . prolixus in palms and that silvatic bugs can colonise houses , with house and palm specimens sharing seven cytb haplotypes . Additionally , mitochondrial introgression was detected between R . robustus and R . prolixus , indicating a previous hybridisation event . The use of ten polymorphic microsatellite loci revealed a lack of genetic structure between silvatic and domestic ecotopes ( non-significant FST values ) , which is indicative of unrestricted gene flow . Our analyses demonstrate that silvatic R . prolixus presents an unquestionable threat to the control of Chagas disease in Venezuela . The design of improved control strategies is essential for successful long term control and could include modified spraying and surveillance practices , together with housing improvements .
Chagas disease is a chronic parasitic disease transmitted by triatomine bugs ( Reduviidae: Triatominae ) and limited in distribution to the Americas . The causative agent is the protozoan Trypanosoma cruzi . Rhodnius prolixus is the primary vector in Venezuela and Colombia and is one of the main targets of the Andean Pact and Central American initiatives , together with the secondary vectors Triatoma dimidiata in Central America , Rhodnius pallescens in Panama and Rhodnius ecuadoriensis in northern Peru [1] . In Venezuela R . prolixus occurs in all States , where it colonises poor quality housing and exhibits high infection rates with T . cruzi . Significant progress has been made in reducing the incidence of Chagas disease in Venezuela through four decades of triatomine control [2] . Nevertheless , domestic infestations of R . prolixus persist and recent data indicate that transmission of T . cruzi may be increasing [3] . In contrast , in the Southern Cone region of South America the main vector , Triatoma infestans , has been eliminated over large areas following control efforts [1] . Triatoma infestans is considered to be a primarily domestic species , with the exception of Bolivian Andes and Gran Chaco region ( Bolivia and northern Argentina ) where silvatic populations were found [4] . Further studies are needed to evaluate the risk these populations pose to effective control in these regions . In comparison R . prolixus is reported to have a widespread silvatic distribution in Venezuela , found most commonly in palm trees and birds nests and more rarely in other sites such as dry trees [5]–[7] . The reinvasion of sprayed houses by silvatic R . prolixus , together with localised control failures could be maintaining disease transmission in Venezuela [3] . However , the existence of silvatic R . prolixus populations has been questioned due to the identification of the closely related species Rhodnius robustus in palm trees in Venezuela [8] . Rhodnius robustus poses a problem as it is virtually indistinguishable morphologically from R . prolixus but this species it is of minor epidemiological importance as it does not colonise houses , although flying adults may enter domestic areas attracted by light [8] , [9] . Confusion has been fuelled by conflicting results of studies investigating the taxonomic status of R . prolixus and R . robustus , with morphometric and isoenzyme studies failing to detect interspecific differences [10]–[15] . However , recent DNA sequencing analyses has not only supported the validity of R . robustus but also indicated the existence of more than one cryptic species [16]–[18] . Additionally in a preliminary finding for this present study four Rhodnius specimens collected in a palm in Guarico State Venezuela were identified as R . prolixus [17] . Here we investigated the genetic structure of 34 populations of R . prolixus , including five adjacent populations , from silvatic , domestic and peridomestic ecotopes in six Venezuelan States . Our aim was to contribute to the control of Chagas disease in Venezuela , through the provision of information that might allow the design of improved control strategies . We finally resolve this controversy over the existence of silvatic R . prolixus and the interaction between silvatic and domestic populations . Our analyses demonstrate that silvatic R . prolixus presents an unquestionable threat to the control of Chagas disease in Venezuela and that successful long term control could benefit from modified spraying and surveillance practices , together with housing improvement .
For the purpose of this study field work was carried out in 2001–2004 in the Venezuelan States of Lara , Portuguesa , Guarico , Cojedes , Barinas , and Trujillo ( see Figure 1 , Table 1 , Table 2 ) . Fieldwork involved the survey of palms , chicken huts and houses in localities in these States in collaboration with the Ministry of Health field inspectors . Silvatic collections were made with Noireau live bait traps [19] . Palm dissection was also used with the consent of landowners . The palm was cut at the base and cleared from the base up to the crown using a machete , removing and inspecting each layer . Domestic and peridomestic collections were made by the traditional search and capture method , with prior consent of householders . All bugs collected were placed in collection tubes , noting date and place of collection . Specimens were identified using the keys of Lent and Wygodzinsky ( 1979 ) [20] .
Our specific interest , in the context of detecting movement between silvatic and domestic Rhodnius populations , was to genotype adjacent silvatic and domestic populations , before examining the relationship between more geographically distant populations .
Mitochondrial DNA has been used previously in triatomine studies , including the tribe Rhodniini [16]–[18] . Here eighteen haplotypes were detected among the 551 Venezuelan specimens analysed and these were confirmed as both R . prolixus and R . robustus species . Our data detected silvatic R . prolixus in palms in all States , except for Trujillo and Lara . We can therefore unequivocally reaffirm that R . prolixus is present in silvatic habitats in Venezuela . Silvatic R . robustus does also exist and was the only species detected in this study in palms in Trujillo State ( pop 34 ) . In this region the post-spray reinvasion of houses is therefore unlikely , and vector control may be more straightforward . Nevertheless , adult silvatic R . robustus have been implicated in the sporadic transmission of T . cruzi in western Venezuela [9] and the use of insecticide treated curtains may contribute to reducing sporadic cases of Chagas disease in this State [35] . From sequence analysis it is clear that common haplotypes occur across all ecotopes , with palm and house populations sharing five R . prolixus haplotypes . Three of these shared haplotypes were found in domestic nymphs , in addition to domestic adults , thus indicating these silvatic R . prolixus are capable of invading and importantly colonising houses . The incongruence detected between nuclear ( D2 ) and mitochondrial ( mtcytb ) analysis of haplotype 3 confirmed the introgression suspected after the discovery of domestic nymphs of “R . robustus” . Introgression has been recorded previously in triatomine species [38] and other haematophagus insects [39] , [40] . In accord with colonisation behaviour , these “Amazonian R . robustus” are R . prolixus with introgressed R . robustus mitochondrial DNA . Additional support for introgression is the absence of unique microsatellite alleles in these haplotype 3 specimens , in contrast to our single domestic Venezuelan R . robustus adult ( haplotype 16 ) , which revealed four unique alleles . Movement of bugs between silvatic , peridomestic and domestic ecotopes probably occurs both actively and passively . Risk factor analysis detected an association between new thatched palm roofs and infestation [41] . Female R . prolixus glue their eggs to palm fronds suggesting passive transport of bugs into houses on these fronds [6] . Restriction or elimination of palm roofs on dwellings must therefore be a key element of control strategies , although it is important that an appropriate substitute roofing material is readily available to the inhabitants . Active transport can also occur , flying adult triatomine bugs may enter a house attracted to light [9] . Rhodnius prolixus in Venezuela is known to be light attracted [47] . From our data it is clear that silvatic populations of R . prolixus in Venezuela represent a definite threat to successful control of Chagas disease , as suspected but controversially debated since populations of R . prolixus were reported in palm trees [5] . Results indicate that the current control programme in Venezuela is unlikely to achieve the level of success seen in the Southern cone , where T . infestans has been eliminated over large areas [1] . The control programme will have to deal with this continual threat , for example by more frequent spraying of houses , combined with community vigilance for reinfestations as an integral part of the control programme . The additional use of alternative control methods such as insecticide treated curtains [35] or bednets [48] would be beneficial . Increased housing improvements , although expensive , seem vital for long term control , by creating a domestic environment unsuitable for colonisation by silvatic bugs . This study has made a fundamental contribution to the understanding of Rhodnius populations in the context of disease epidemiology and vector control in Venezuela . An important follow-up to this project would be to define population interaction more extensively , particularly in regions of Colombia , where silvatic and domestic Rhodnius populations also occur and reinvasion may be maintaining large domestic colonies of R . prolixus [49] . This would allow prioritisation of control interventions and tailoring of control strategies to regional circumstances . Additionally , modified control strategies to counteract the threat of reinvasion could be assessed , such as widespread provision of ideal low cost roofing , the treatment or removal of palms close to houses , and , improved spraying and surveillance , all with the aim of reducing the burden of Chagas disease in rural areas . | Chagas disease is spread by blood-feeding insects ( triatomine bugs ) that colonise poor-quality houses . Disease control relies primarily on killing domestic bugs by spraying dwellings with residual insecticide . In Venezuela , sustained control has proved difficult despite four decades of campaigns . Considered the main vector in Venezuela , the bug Rhodnius prolixus may also infest palm trees and might repeatedly recolonise houses from palms . A complication is that a morphologically similar species , R . robustus , also infests palms but is of minor medical importance . Therefore , confusion exists as to the true identity of palm bugs and their importance in disease transmission . We applied two molecular methods ( sequencing DNA of the cytochrome b gene , and analysing microsatellites ) to triatomines collected in Venezuela so that we could identify unequivocally the species of palm-dwelling Rhodnius and establish their role in maintaining house infestations . We demonstrated that R . prolixus is indeed present in palms , and that such silvatic populations can colonise houses and are a threat to the successful control of Chagas disease in Venezuela . This finding resolves a longstanding controversy of fundamental epidemiological importance . It is also an example of the application of molecular epidemiology to correct vector identification and successful disease control . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"molecular",
"biology",
"genetics",
"and",
"genomics/population",
"genetics"
] | 2008 | Molecular Genetics Reveal That Silvatic Rhodnius prolixus Do Colonise Rural Houses |
The epithelium of the small intestinal crypt , which has a vital role in protecting the underlying tissue from the harsh intestinal environment , is completely renewed every 4–5 days by a small pool of stem cells at the base of each crypt . How is this renewal controlled and homeostasis maintained , particularly given the rapid nature of this process ? Here , based on the recent observations from in vitro “mini gut” studies , we use a hybrid stochastic model of the crypt to investigate how exogenous niche signaling ( from Wnt and BMP ) combines with auto-regulation to promote homeostasis . This model builds on the sub-cellular element method to account for the three-dimensional structure of the crypt , external regulation by Wnt and BMP , internal regulation by Notch signaling , as well as regulation by internally generated diffusible signals . Results show that Paneth cell derived Wnt signals , which have been observed experimentally to sustain crypts in cultured organs , have a dramatically different influence on niche dynamics than does mesenchyme derived Wnt . While this signaling can indeed act as a redundant backup to the exogenous gradient , it introduces a positive feedback that destabilizes the niche and causes its uncontrolled expansion . We find that in this setting , BMP has a critical role in constraining this expansion , consistent with observations that its removal leads to crypt fission . Further results also point to a new hypothesis for the role of Ephrin mediated motility of Paneth cells , specifically that it is required to constrain niche expansion and maintain the crypt’s spatial structure . Combined , these provide an alternative view of crypt homeostasis where the niche is in a constant state of expansion and the spatial structure of the crypt arises as a balance between this expansion and the action of various sources of negative regulation that hold it in check .
Stem cells have critical physiological roles in both the renewal of healthy tissues and the repair of damage . Intriguingly , while these cells perform the same basic processes as other cells , e . g . growth and division , they are typically associated with a special environment , a “niche” . A common hypothesis for the functional role of such an environment is the regulation of homeostasis [1] . One generic model of homeostatic regulation is the so-called “hand of God” model where external signals regulate stem cell activity . In the intestinal crypt for example , external Wnt signals provided by surrounding tissue have been shown to regulate differentiation [2 , 3] . An alternative ( but not exclusive ) possibility is that stem cells build a niche where internal feedbacks as well as feedbacks between the niche and its environment regulate homeostasis . Stem cells in the olfactory epithelium for example have been shown to interact with their progeny and environment through a complex set of diffusible signals to regulate their own population [4] . Similarly , interactions between stem cells of the hair follicle and their progeny are responsible for the predictable timing of cyclic hair growth [5] . Here we investigate how highly local ( e . g . at the length scale of a single cell ) niche signaling influences the spatial structure of the intestinal crypt and the homeostatic balance between expansion and repression of stem cell populations . The epithelium of the intestinal crypt is an incredibly dynamic tissue , constantly replenishing itself every 4–5 days . This test tube shaped invagination of the intestine is spatially configured with a proliferative compartment at its base with a compartment of differentiated cells above it that perform various physiological functions critical to digestion . The source of this constant replenishment , like with other organs and tissues , is a small pool of cycling intestinal stem cells ( ISCs ) . Early investigations implicated so called “+4” cells ( so named for their position 4 cells up from the base ) as the ISCs [6] . Alternatively , it was suggested that crypt base columnar cells ( CBCs ) interleaved with Paneth cells at the crypt base were the true ISCs [7 , 8] . These investigations however relied on the Lgr5 marker to indicate stem-ness and a functional approach has suggested that only a subset of these Lgr5 cells are actively participating in crypt maintenance at any given time [9] . A more recent theory has suggested there are in fact two populations of ISCs , active CBCs that steadily renew the crypt and quiescent +4 cells that activate and regenerate it after injury [10 , 11] . While the debate about the true identity of ISCs remains , it is clear that the CBCs ( or some subset of them ) at the base of the crypt are responsible for the continual renewal of the crypt epithelium . This constant replenishment is fueled by approximately 15 CBCs [12] . In contrast to canonical renewal processes however , these CBC stem cells divide exclusively symmetrically [13 , 14] and that differentiation is decoupled from division [15] . Furthermore , they do so considerably more quickly than in other tissues , dividing roughly once per day even in healthy tissue [16] . How then is homeostasis of such a dynamic tissue maintained ? Numerous investigations have shown the canonical Wnt / β-catenin pathway to be critical in maintaining homeostasis [2 , 3] . This pathway , which regulates gene transcription and cell fate specification , is required to prevent differentiation of stem cells and maintain the crypt . This is evidenced by the complete depletion of stem cells upon disruption of this pathway [17 , 18] . Interestingly , there are two sources of Wnt signaling in the crypt [2] . The mesenchyme that surrounds it produces graded expression ( highest at the base ) of a number of Wnts including Wnt2b , Wnt4 , and Wnt5a . Additionally , Paneth cells , which are interleaved with the CBCs at the base and commonly referred to as niche cells , also produce Wnt3a . Surprisingly , genetic deletion of this “local” , Paneth cell derived Wnt source does not impair stem cell populations in the in vivo crypt [19] , suggesting the global Wnt gradient is sufficient for homeostasis . However , in vitro studies of “mini-guts” grown from CBCs have shown that Paneth derived Wnt3a alone is also sufficient to maintain crypt structure in the absence of the other exogenous Wnt sources [20 , 21] . While Wnt signaling is crucial to crypt homeostasis , there are other important regulatory pathways that are also required . Notch lateral inhibition creates a toggle switch that leads to the salt and pepper organization of stem / Paneth cells in the base of the crypt . This pathway is also responsible for a similar arrangement of secretory ( Goblet ) and absorptive ( enterocyte ) lineages further up the crypt walls [22 , 23] . This salt and pepper arrangement in particular is critical in maintaining the niche structure at the crypt base since contact with a Paneth cell is required to prevent stem cell differentiation [24] . Additionally , Eph / ephrin signaling interactions generate repulsive forces that drive Paneth cells to migrate down the crypt wall while all other cells passively migrate upward from the base , driven by proliferative pressure [25] . Bone morphogenic proteins ( BMPs ) , which form a gradient opposing that of Wnt [26 , 27] , are also known to influence crypt homeostasis by suppressing proliferation of stem cells [28] , and their inhibition leads to crypt fission [29] . How do these signaling components contribute to maintaining the spatial structure of the crypt and how do they interact ? In addition to experimental interrogation , extensive computational modeling has been employed to address this and related questions . Using optimal control theory , it was shown that a “bang bang” growth process is responsible for crypt formation [30] . Numerous compartment models , which consider the crypt to be spatially well mixed and focus on temporal dynamics , have been used to investigate the processes that promote homeostasis and drive tumorigenesis in crypts [31–34] , see [35] for a recent review . Continuum spatial models have similarly been used to investigate the formation and regeneration [26] of crypts as well as mutation acquisition [36] in them . Each intestinal crypt however contains on the order of tens of stem cells and hundreds of total cells and is thus a highly stochastic entity . Further , the spatial arrangement of stem and Paneth cells at the crypt base has an important role in niche homeostasis . Discrete models accounting for individual cell dynamics and interactions have been developed to account for these features . In [37] , it was shown that the geometry of the crypt could affect organ aging and susceptibility to cancer . Wong et al . [38] demonstrated that under certain conditions , Eph / ephrin mediated differential adhesion is required for proper crypt organization . In [39 , 40] , discrete modeling methods were used to show that the basement membrane has a critical role in defining the crypt geometry , which is crucial for proper function . In [41] , the geometry was further shown to have a significant impact on the time it takes for neutral drift to drive a crypt to mono-clonality , which has implications to mutation acquisition and fixation . In [42 , 43] , it was shown that Wnt and Notch signaling are critical to organizing crypt architecture and that under the assumption of reversible cell fate specification , this architecture is extremely robust to perturbation . Extensive use of agent based modeling [44] has been used in this domain as well . Bravo et al . [45] constructed a 2D agent-based crypt model that was calibrated to human biopsy data to accurately account for the number of cells of different types as well as the variance of those numbers . They then used this as an in silico test platform to determine the efficacy of different cancer therapy protocols . For further review of the extensive discrete crypt modeling literature , see [46] . Most of these investigations have however been directed at understanding the physical structure of the crypt and how it influences function , rather than the role of niche signaling on homeostasis . Those that have incorporated signaling have thus far primarily focused on the influence of external signals , exogenous Wnt signaling in particular [42] . Here , we extend these investigations to investigate the role of local production of Wnt by Paneth cells as well as negative regulation via BMP . Toward this end , we build a comprehensive discrete model accounting for both the physical structure of the crypt and these signaling interactions . This model is then probed to determine the influence of these different signaling components and the implications of their deletion . A number of 2D [38 , 47] and 3D [42 , 48] models have been utilized to investigate various aspects of crypt dynamics in the past . Here , we utilize the relatively new sub-cellular element method ( SSEM ) [49 , 50] to treat individual cells as discrete , deformable objects . The SSEM provides a natural framework to describe the mechanical force interactions between cells , which is important in this application . This method has been previously employed to model both multi-cellular systems and single cell dynamics [51] . In the multi-cellular context , it was used to describe the dynamics of epithelial sheets [49] , primitive streak formation in the chick embryo [52] , the influence of Notch signaling on regulatory networks controlling cell division [53] , the dynamics of bacterial swarms [54] , and the dynamics of epithelial layer formation [55] . Here we use this framework to construct a model of the 3D , dynamically evolving crypt . Within this model , we account for cell-cell force interactions , Ephrin mediated repulsion of Paneth cells , cell-cell Notch signaling , exogenous Wnt and BMP signaling , as well as local Paneth cell derived Wnt signaling . The latter of which requires a substantial augmentation of the SSEM to account for the presence of diffusible signals . By interrogating this model computationally , we show that indeed , multiple sources of Wnt signaling can act redundantly to maintain the crypt . A crucial implication of this redundancy however is that the stem cell niche is in a constant state of expansion , which if left unchecked would lead to the niche cannibalizing the entire crypt . Further results however indicate that inhibition of proliferation by BMP can constrain this expansion and promote homeostasis . This view also points to a different interpretation of the function of downward motion of Paneth cells . We find that this motion is not required to maintain the niche as might be expected ( given the role of Paneth cells in niche signaling ) , but rather it is again needed to constrain niche expansion . Taken together , these results suggest that different Wnt sources have significantly different influences on niche homeostasis , and that negative regulation is required to balance the expansive influence of Paneth derived Wnt signaling .
Here we describe a three-dimensional , multi-scale model of the dynamically evolving crypt . This model combines 1 ) a subcellular element formalism describing physical / mechanical aspects of cellular dynamics , 2 ) a chemical diffusion-reaction model for endogenously produced signaling , and 3 ) a stochastic differential equation model of lineage regulation for each individual cell . For further details , see Materials and Methods . For all parameters listed , see Table 1 . We utilize an extension of the subcellular element method to describe each discrete cell of the evolving crypt as a deformable object . In this formulation , a cell is described by a collection of N sub-cellular elements [49] that interact pairwise according to user defined forces . These forces encode short-range repulsion , which endows each element with a “volume” , and medium-range attraction , which causes all elements to form a coherent cell . In the absence of external forces , energy minimization will cause these cells to round up to a preferred spherical shape and volume . Additionally , we model direct cell-cell interactions by specifying forces between different cells that mimic contact and adhesion . The resulting model forms a large system of differential equations that describe the evolution of all elements ( and hence the cells themselves ) in time . The flexibility of this method further allows specification of different properties based on a cell’s identity . One such difference we include is the tendency for Paneth cells to move downward toward the base of the crypt , while all other cell types move passively up the crypt wall in response to proliferative pressures . See Materials and Methods for further details . In addition to cell-cell interactions , forces between each cell and its environment are also prescribed . Each cell is assumed to adhere to a rigid , test tube shaped basement membrane . Additionally , a drag force between cells and the membrane is imposed , mimicking the friction caused by the need to break and re-form bonds as the cells move . When cells reach the upper bound of the crypt domain , they are removed . Similarly , if detachment of a cell from the basement membrane is detected , it is removed . Four primary cell types are considered: stem , Paneth , enterocyte , and Goblet . The former two are well known to occupy the base of the crypt while the latter two comprise the upper crypt epithelium ( Fig 1A ) . Since we are primarily interested in how signaling in the stem cell niche influences crypt dynamics and stability , we simplify the system by assuming only stem cells are proliferative and do not include the transient period during which enterocytes and Goblet cells further up the crypt divide . Following a previous study [42] , we assume Wnt- and Notch-signaling jointly regulate fate specification ( Fig 1B ) . For both pathways , we assume cells above or below the threshold ( THNotch for Notch and THWnt for Wnt ) take different fates . INotch denotes the notch activation level of a cell , which is determined by the activation of its neighbors . IWnt represents the Wnt level that a cell is exposed to , which is the sum of contributions from the external Wnt gradient and the Wnt produced by nearby Paneth cells . Fig 1B indicates the combination of these signaling levels that determine each cell’s fate . As in [42] , we assume that cells can reversibly transition between Goblet and enterocytes fates , depending on Notch levels . Paneth cells are further assumed to terminally differentiate , after which they enter a long-lived quiescent state . We assume two sources of Wnt influence cell differentiation: a global gradient derived from the surrounding mesenchyme , and an additional contribution being produced by Paneth cells in the niche . The global gradient is assigned to be highest at the crypt bottom and to decrease gradually along the crypt axis . Given the lack of in vivo concentration information and the fact that the relative levels of a morphogen determine the spatial cues , this quantity is non-dimensionalized in the range [0 , 1] . Initially , we will assume this gradient is deterministic but later will consider the influence of noise superimposed on that gradient , which we assume to be uncorrelated ( in space and time ) multiplicative noise . Each Paneth cell is further assumed to secrete Wnt at a constant rate . This signal field is modeled using the chemical reaction-diffusion equations: ∂c∂t=DcΔc+δcnPaneth−dcc , ( 1 ) where c denotes Wnt concentration at a given location in space and time . The second term on the right hand side δcnPaneth represents the total secretion rate , δc is the secretion rate of an individual Paneth cell , and nPaneth is a measure of the local density of Paneth cells at each grid location . This mapping is used to spread the production of each Paneth cell over the grid nodes that each Paneth cell occupies . Dc is the diffusion coefficient , and dc is the decay rate . For simplicity , we assume that the signal cannot diffuse across the basement membrane or into the crypt lumen . Rather than define no flux boundary conditions on complex surfaces , we instead extend the computational domain beyond the domain containing the cells and assign the chemical diffusion coefficient to be Dc = 0 on the extended domain . In addition to simplifying boundary conditions , this also allows the use of a box shaped domain , which is simpler computationally . This chemical field is simulated on the regular grid , and a reverse mapping is used to determine the value of this Wnt field that each cell is exposed to . The contribution of the global Wnt gradient is then added to this local value to generate the total Wnt field . The Notch activity is calculated via direct cell-cell contact analysis . A cell is Notch-activated by direct neighboring cells expressing Notch-ligands according to INotch=∑cellδ ( i ) NP . Here , the sum runs over all neighboring cells of the target cell . δ ( i ) is equal to one if cell i is in contact with the target cell ( which is determined by proximity ) , otherwise it is zero . The degree of activation by a single cell ( NP ) depends on the cell type . NP is assumed to be larger than zero for Paneth and Goblet cells and zero for all other cells . A cell changes its fate if its Notch-activity crosses the threshold THNotch . We include the influence of BMP signaling on cell proliferation . Similar to Wnt gradient , a global BMP gradient is applied along crypt axis . This gradient opposes the Wnt gradient however , with low levels at the base that increase as you move up the crypt ( again in the range of [0 , 1] ) . Multiplicative noise is again applied to mimic the stochasticity of this gradient . Threshold regulation is similarly assumed , so that if a cell is exposed to BMP levels above THBMP , proliferation is inhibited . Cells grow at a constant rate; an element is added to each cell at regularly scheduled intervals , resulting in a volume increase . Only stem cells are capable of proliferation , with a cell cycle of approximately 24 hours [56] . This is taken as the mean value of a normal distribution ( with standard deviation of 4 hours ) , truncated to the interval 20–28 to ensure reasonable values . Each cell is assigned a division cycle length and an internal timer . At beginning , the internal timer is set to zero , and increments every time step . If the inner time counter exceeds the assigned division cycle length , that cell undergoes division . When division occurs , a plane perpendicular to the crypt wall with a randomly chosen angle is assigned as the division plane to split the cell . This is done to ensure both daughter cells maintain contact with the wall after division . The elements of the cell are then divided so that the daughter cells have an equal number of elements ( plus or minus one ) . For each daughter cell , the internal clock counter is set to zero . Paneth cells are assumed to enter a quiescent phase after differentiation . We thus assume that after differentiating , each Paneth cell undergoes apoptosis . The lifetime of Paneth cell is normally distributed with mean 8 weeks [21] and standard deviation 2 weeks ( this distribution is truncated to the 6 to 10 week range to ensure reasonable values are chosen ) .
The murine intestinal crypt epithelium is one of the most dynamic organs in the body , completely replenishing itself every 4–5 days . This quick turnover improves epithelium integrity in the intestinal environment where cells are under constant assault from toxins , gastric acids , and microorganisms . The speed of this replenishment and the fact that this continually occurs over the life of an organism however raises the question , how are size and structure robustly maintained . It is well known that a pool of approximately 10–15 fast cycling crypt base columnar “stem” cells ( CBCs ) at the base of each crypt is responsible for constant renewal [7 , 8 , 12] . But while a number of molecular regulators that influence CBC proliferation and differentiation dynamics have been identified , it remains unclear how these regulators coordinate to maintain homeostasis . Here , we construct a discrete , multiscale model of the evolving crypt and interrogate the role of different hypothesized regulators on homeostasis . This model uses a subcellular element formalism [49] to describe the structure of cells , their interactions , and their interactions with the crypt wall . On top of this formalism , which primarily describes physical aspects of the system , we include the dynamics of cellular commitment , cell-cell signaling ( Notch signaling in particular ) , and the presence of diffusible signals that influence cellular commitment . We do note that there are a number of aspects of crypt biology that we do not account for . In particular , there are at least seven different cell types present in the crypt , some of which we do not include . We do not account for the polarized nature of cells [58 , 59] , the resulting function of those cells ( e . g . transport of material into and out of the crypts lumen ) , or systemic responses to damage [60] . Similarly , we do not account for density dependent inhibition of proliferation or damage-induced effects such as activation of +4 cells [10 , 11] . The goal of this exposition is to investigate the role of different regulatory mechanisms in maintaining homeostasis in healthy crypts . Toward this goal , we incorporate the features of crypt biology most germane to this context and leave the inclusion of these additional features for future work . In particular , we use this model as a platform to investigate the influence of 1 ) Paneth cell derived Wnt , 2 ) BMP signaling , and 3 ) Ephrin mediated repulsion of Paneth cells on homeostasis as well as 4 ) the influence of signaling noise on crypt structure . It is well established that continual activation of the canonical Wnt signaling pathway is required to prevent stem cell differentiation . However , there are multiple sources of Wnt signaling [2] . First , the mesenchyme surrounding the crypt generates a Wnt gradient that is highest at the base . It has been shown previously that this gradient in concert with Notch lateral inhibition and Paneth cell migration ( driven by Ephrin signaling ) can maintain a crypt with the proper structure [42] . However recent evidence suggests that Wnt secreted by Paneth cells , which are interleaved with CBCs at the base , is sufficient to both generate and maintain crypts [20] , suggesting redundancy . Our results suggest that these two Wnt sources have functionally different effects on crypt homeostasis , and that the influence of this additional Wnt source depends critically on its rate of production . When that rate is below a critical level , it is not sufficient to maintain the niche and exogenous Wnt is required . In this case , it has no influence on homeostasis . If on the other hand , the production rate is above this critical level , the crypt becomes robust against perturbations or even complete removal of the exogenous Wnt source . This redundancy however comes at an expense . This source of Wnt combined with the mutually reinforcing feedback between stem and Paneth cells creates a positive feedback that drives uncontrolled expansion of the niche , at the expense of the remainder of the crypt . Further results however show that BMP signaling , which forms an opposing gradient to the exogenous Wnt gradient and inhibits proliferation [26 , 28] constrains this expansion and promotes homeostasis . Results also suggest a different possible explanation for the downward migration of Paneth cells . It is commonly held that this motion is required to maintain the niche . Simulation of a number of different model variants however suggests that abrogation of this migration does not destroy the niche . On the contrary , removal of this motion in some circumstances leads to its uncontrolled expansion , even when BMP inhibition is present . Thus rather than being required to maintain the niche , this motion may instead be required to constrain its expansion and maintain the remainder of the crypt . There is a caveat to this result of course . In this investigation we have assumed that Paneth cell function is independent of location in the crypt . That is , they can secrete Wnt at any location in the crypt . It is possible that Paneth cell function is inherently tied to location though , i . e . they can only secrete Wnt if they are at the base . This would however create an entire different regulatory mechanism . In this case , Paneth cells would essentially act as amplifiers of the external Wnt gradient . In this way , they would not be participating in an auto-regulatory feedback loop but rather would be an intermediary of a purely external regulatory scheme . We do not reject this possibility , but do suggest it is seemingly inconsistent with in vitro “mini guts” results . In that setting , there is no external tissue or signal telling Paneth cells where they are . Furthermore , they are the only source of Wnt , suggesting they act as more than a passive amplifier of external Wnt signals and do participate in an auto-regulatory feedback . We further investigated the robustness of a niche regulated in this manner to noise in exogenous Wnt and BMP signals . Results show that as expected , noise has the effect of introducing a small amount of variability in the size of the niche . More unexpectedly , we find that the introduction of noise actually increases the size of the niche . In most cases , this is not significant and for all practical purposes , the niche is robust against these noise sources . In extreme circumstances however , where both noise levels and the rate of Wnt production are quite large , this noise actually destabilizes the niche causing uncontrolled expansion . These extreme circumstances however appear to be outside the physiological regime , and we thus conclude that there is a sizable operating regime where a regulatory system of this form creates a homeostatic environment that is relatively insensitive to noise in Wnt and BMP . The essential biological feature that gives rise to these results is that Paneth cells create a “mini niche” surrounding them . While this is well characterized experimentally [21] , the implication of this to crypt stability and homeostasis , which is the topic of this investigation , is relatively less understood . The central result here is that these “mini niches” form a local , auto-regulatory feedback loop that supplements external Wnt signaling to redundantly reinforce crypt renewal . This hypothesis leads to a few predictions . First , when Paneth derived Wnt levels are sufficient to maintain the niche , removal of BMP is predicted to lead to aberrant expansion of the niche . Previous results have indeed shown that deletion of BMP signaling leads to crypt fission [29] and the formation of ectopic crypts [61] , which is one potential effect of this expansion . An alternative prediction is that the in vivo crypt can be maintained in the absence of mesenchyme derived Wnt . Testing this prediction would require deleting mesenchyme derived Wnt without perturbing Paneth cell derived Wnt or the β-catenin required to transduce Wnt signals . Fevr et al . [17] demonstrated that Wnt deletion leads to terminal differentiation of all crypt stem cells . This investigation however deleted cell’s ability to transduce all Wnt signals . It has been verified in vitro that external Wnt signals are not required for crypt development or maintenance , however these results are confounded with the presence of additional Wnt regulators , chiefly R-spondin . A sounder test of this regulatory scheme requires specific removal of the exogenous Wnt signal in vivo , without perturbing the proposed auto regulatory feedback . These results paint a somewhat different picture of small intestinal crypt homeostasis from the existing view . In the canonical view , Wnt is a master regulator that serves as a morphogen of sorts , creating a road map that links cell properties to their locations . This however implicitly assumes the niche is a delicate environment that needs to be supported ( by Wnt ) . Results here suggest that instead , auto-regulation pushes the niche into a constant state of expansion and that various forms of negative regulation constrain that expansion . This hypothesis is consistent with observations in other systems , such as the olfactory epithelium or the hair follicle niche , where negative regulation is a critical component of niche dynamics . More generally , these results are in line with observations in other systems showing that cells of a niche actively participate in the maintenance of their own microenvironment , rather than being slaved to external regulation .
Each individual cell is represented by a set of N connected elements . N is chosen to be 20 in simulation to balance between the flexibility of modeling cellular dynamical activity and computational costs . The dynamics of each element is determined by biomechanical forces , which consist of intracellular interaction among the elements of the same cell , intercellular interaction between elements of different cells , and external force due to environmental cues . The equation of motion of the position vector Yαi of element αi for cell i is dYαidt=−∇αi∑αi≠βiVintra ( |Yαi−Yβi| ) −∇αi∑i≠j∑βjVinter ( |Yαi−Yβj| ) −∇αiFexternal ( Yαi ) , where Vintra is a pairwise force interaction between elements αi and βi of the same cell i , Vinter is a pairwise force interaction between elements αi of cell i and βj of cell j , and Fexternal is an external force representing membrane adhesion and other environmental interactions . All elements within a cell interact according to the spring potential [47 , 48] Vintra=μ ( rij−r0 ) 22 , where rij is the distance between element i and element j of the same cell and r0 is a rest length . In the absence of external forces , the intra-cellular forces will scatter the inner elements to the minimum energy configuration with a roughly spherical shape of preferred size . That size is determined by the rest length r0 for Vintra , defining a volume of sorts for the cell . The inter-cellular force interactions are described by Lennard-Jones type potentials [47 , 48] Vinter=ε ( ( σ|rij| ) 12− ( σ|rij| ) 6 ) where rij is the distance between element i and element j . The parameter ε determines the strength of interaction . σ is the equilibrium separation where the inter-element potential is zero and two elements are at relative balance position . If the distance between two elements is smaller than σ , they experience a repulsion force to prevent overlap of the cell bodies . When the distance between the elements is greater than σ , but less than a cutoff value , an attraction exists between the elements . Beyond this cut-off value , we assign zero interaction between cells . These medium range interactions are designed to represent the surface interactions of cadherin-mediated cell-cell adhesion . We consider a simplified crypt structure with a test tube geometry ( similar to [42 , 48] ) , which is a cylinder attached above a semi sphere . The tube is chosen to have 16 cell diameter in height and 6 cell diameter in diameter . The adherent force between cell elements and the crypt wall is defined by Fexternal ( Yαi ) =εexternal|ri| . Here , εexternal is the strength of external force , and ri is the distance between element i and the crypt wall . This force has a cut-off distance of half the rest diameter of a cell , to ensure that only the elements “attached” to the wall experience the attraction . The friction between cells and the basal membrane is modeled as a linear drag with the equation Fd=bzv⇀ , where bz is the linear drag constant in z-axis and v⇀ is the velocity of the cell . The force of drag is approximately proportional to velocity , but opposite in direction to mimic the rupture of many ligand bonds distributed on the cell membrane . To couple cell dynamics and signaling pathway , a regular , rectangular grid for chemical diffusion is superimposed on the subcellular element model domain . Each simulation time step for the evolution of the full system consists of a substep of subcellular element model followed by a substep evolving the state of the chemical field according to the reaction diffusion PDE . During the substep of the cell-based subcellular element model , cells move to a new location , undergo growth and division , make lineage decisions , and produce Wnt signals , which modifies the local Wnt field . Each Paneth cell serves as a Wnt source . This is implemented by having each Paneth cell element secrete Wnt at a rate δc . This production is extrapolated from the element’s position to the nearby grid elements , so that this production acts as a distributed source in the chemical diffusion PDEs . Each cell in turn reads the Wnt concentration at its location to make fate commitment decisions . This is implemented using standard trilinear interpolation where the concentration value at each element position is determined from the chemical field on the regular grid . Subsequently , the value of the external Wnt field is added to this value to produce the total Wnt concentration . For each cell , a linear combination of the Wnt concentration at each of its elements is then used to determine the net Wnt concentration that cell is exposed to . During the substep of chemical diffusion evolution , not only do Paneth cells serve as sources of Wnt production , but all cells serve as barriers for diffusion . If a well-defined boundary of each cell were determinable , diffusion could simply be prohibited within those boundaries . However using this methodology , cells do not have a well-defined boundary but instead are made up of elements . To mimic the restriction of diffusion to the exterior of the cells , we make the rate of diffusion dependent on the local number density of elements . That is , for each grid node , the number of elements in neighboring grid spaces is computed , with the rate of diffusion decreasing as this quantity increases . Diffusion is also restricted to the epithelium itself under the assumption that the basement membrane is impermeable . To implement this , we expand the computational grid beyond the crypt itself so that a regular , rectangular grid can be employed . We then assign a zero diffusion coefficient at nodes beyond the crypt walls . To solve Eq ( 1 ) , we apply a second-order central difference for the spatial derivatives , and a forward Euler scheme to the temporal discretization . Step size in space is chosen to be 1 μm . For each step of chemical diffusion evolution , the chemical field is updated for 1000 times , giving dt = 0 . 0036s for chemical equation updates . To build up the initial configuration of a crypt , we begin with a test simulation where only two stem cells placed at the crypt bottom . These cells ( which have no identity at this point ) are allowed to proliferate and move until the daughter cells cover up the whole crypt to form a compact cell packing . Cell Notch levels and identities are then initialized with the canonical spatial cell distribution: stem and Paneth cells interleaved with each other at the bottom of crypt with enterocytes and Goblet cells residing in the top of the crypt . The inner time counter for each cell is then chosen as a random number smaller than its pre-assigned life cycle to initialize each cell’s cycle length . The chemical diffusion-reaction equation is given by Eq ( 1 ) in the Results section . Estimates for D , d were obtained from [26 , 62] ( Table 1 ) . However we do not have an estimate for either the Wnt concentration in vivo or the rate of Wnt secretion rate by Paneth cells . It is known however that in this system , Wnt is lapidated and thus has a short diffusion length scale [2] . We thus assign a base value of the Wnt production rate as that rate at which the Wnt threshold THWnt is achieved at a distance of 12 . 5 microns from the source ( the equivalent of 1 . 25 cell diameters ) . This is chosen to represent a type of threshold production rate . Above this rate , local Wnt concentrations will presumably be sufficiently large to maintain a stem cell in near direct contact . Below this rate , it will not . We compute this critical rate δc by considering a simplified setting where a single Paneth cell is placed at x = 0 in a one-dimensional domain . The concentration field is then simulated for a range of values and the critical value at which this condition is achieved is recorded . The most computationally intensive component of this method is computation of the forces between elements that drive cell motions . This is an N-body simulation where the force between all pairs of elements must be computed . This scales as O ( n2 ) where n is the number of subcellular elements in the system . Fortunately , this computation is also highly parallel and suitable for GPU implementation . We followed [49] to provide a parallel implementation of the subcellular element model , and include memory layout of data structures and functional decomposition for efficient implementation . In this implementation , the highly-parallel parts like force and dynamic computations , which do not require dynamic updating of data structures , were carried out on the GPU using OpenCL . The less intensive computations such as cell division and growth , which require updating data structures , were carried out on the CPU using C . To minimize memory transfer between the GPU and CPU , a fixed number of cell position updates are iteratively computed on the GPU and the data is subsequently shipped to the CPU where a single growth / division / differentiation update is performed . The data is then transferred back to the GPU where subsequent updates are performed and this process is iterated over the length of the simulation . | The small intestinal epithelium , like our skin , is constantly being renewed . In the intestine however , this epithelium is exposed to the harsh digestive environment , necessitating much more rapid renewal . Remarkably , the entire epithelium is renewed every 4–5 days . This raises the question , how can the size and structure of this tissue be maintained given this pace . Motivated by recent experimental observations , we construct a three-dimensional , hybrid stochastic model to investigate the mechanisms responsible for homeostasis of this tissue . We find that there are redundant signals created by both the epithelium itself and surrounding tissues that act in parallel to maintain epithelial structure . This redundancy comes at a price however: it introduces the possibility of explosive stem cell population growth . Additional results suggest that other signals along with choreographed motion of cells are responsible for repressing this expansion . Taken together , our results provide a novel hypothesis for how robust but fast renewal of the crypt is achieved: as a balance between expansion , which drives fast renewal and repression , which holds that expansion in check to maintain the crypt’s structure . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | The Interplay between Wnt Mediated Expansion and Negative Regulation of Growth Promotes Robust Intestinal Crypt Structure and Homeostasis |
Short-term changes in illumination elicit alterations in thylakoid protein phosphorylation and reorganization of the photosynthetic machinery . Phosphorylation of LHCII , the light-harvesting complex of photosystem II , facilitates its relocation to photosystem I and permits excitation energy redistribution between the photosystems ( state transitions ) . The protein kinase STN7 is required for LHCII phosphorylation and state transitions in the flowering plant Arabidopsis thaliana . LHCII phosphorylation is reversible , but extensive efforts to identify the protein phosphatase ( s ) that dephosphorylate LHCII have been unsuccessful . Here , we show that the thylakoid-associated phosphatase TAP38 is required for LHCII dephosphorylation and for the transition from state 2 to state 1 in A . thaliana . In tap38 mutants , thylakoid electron flow is enhanced , resulting in more rapid growth under constant low-light regimes . TAP38 gene overexpression markedly decreases LHCII phosphorylation and inhibits state 1→2 transition , thus mimicking the stn7 phenotype . Furthermore , the recombinant TAP38 protein is able , in an in vitro assay , to directly dephosphorylate LHCII . The dependence of LHCII dephosphorylation upon TAP38 dosage , together with the in vitro TAP38-mediated dephosphorylation of LHCII , suggests that TAP38 directly acts on LHCII . Although reversible phosphorylation of LHCII and state transitions are crucial for plant fitness under natural light conditions , LHCII hyperphosphorylation associated with an arrest of photosynthesis in state 2 due to inactivation of TAP38 improves photosynthetic performance and plant growth under state 2-favoring light conditions .
Owing to their sessile life style , plants have to cope with environmental changes in their habitats , such as fluctuations in the incident light . Changes in light quantity or quality ( i . e . , spectral composition ) result in imbalanced excitation of the two photosystems and decrease the efficiency of the photosynthetic light reactions . Plants can counteract such excitation imbalances within minutes by a mechanism called state transitions , which depends on the reversible association of the mobile pool of major light-harvesting ( LHCII ) proteins with photosystem II ( state 1 ) or photosystem I ( PSI ) ( state 2 ) ( reviewed in [1]–[5] ) . In detail , the accumulation of phosphorylated LHCII ( pLHCII ) , stimulated in low white light , or by light of wavelengths specifically exciting PSII ( red light ) , causes association of pLHCII with PSI ( state 2 ) , thus directing additional excitation energy to PSI . Conditions like darkness or light of wavelengths specifically exciting PSI ( far-red light ) , as well as high intensities of white light , stimulate pLHCII dephosphorylation and its migration to PSII ( state 1 ) , thus redirecting excitation energy to PSII . LHCII phosphorylation and state transitions have been extensively studied in the green alga Chlamydomonas reinhardtii and the flowering plant Arabidopsis thaliana [2] , [4]–[6] . In C . reinhardtii , the impact of state transitions on interphotosystem energy balancing and on promoting cyclic electron flow is well established [2] , [5] . In flowering plants , however , the physiological significance of state transitions is less clear , because their mobile LHCII pools are significantly smaller than those in green algae [7] , [8] . Thus , A . thaliana mutant plants impaired in state transitions are only marginally affected in their development and fitness [9]–[11] , even under fluctuating light or field conditions [12] , [13] . However , when Arabidopsis state transition mutants are perturbed in linear electron flow , effects on plant performance and growth rate become evident [14] , indicating that also in flowering plants , state transitions are physiologically relevant . The protein kinase responsible for phosphorylating LHCII is membrane bound and activated upon reduction of the cytochrome b6/f ( Cyt b6/f ) complex via the plastoquinone ( PQ ) pool under state 2-promoting light conditions ( low white light or red light ) [15] , [16] . PQ oxidizing conditions induced by state 1-promoting light conditions ( dark or far-red light ) inactivate the LHCII kinase and result in association of pLHCII with PSII ( state 1 , reviewed in [4] , [5] ) . The LHCII kinase activity , however , is also inactivated under high white light conditions , when the stromal reduction state is very high . In vitro and , more recently , in vivo studies suggest that suppression of LHCII kinase activity might be mediated by reduced thioredoxin [17] , [18] . In C . reinhardtii and A . thaliana , the orthologous thylakoid protein kinases Stt7 and STN7 , respectively , are required for LHCII phosphorylation and state transitions [12] , [19] . Coimmunoprecipitation assays showed that the Stt7 kinase interacts with Cyt b6/f , PSI , and LHCII [17] , suggesting that Stt7 ( and STN7 in Arabidopsis ) directly phosphorylates LHCII , rather than being part of a Stt7/STN7-dependent phosphorylation cascade . Under PQ oxidizing conditions when the LHCII kinase becomes inactivated , pLHCII is dephosphorylated by the action of an as-yet unknown protein phosphatase , thus allowing the association of the mobile fraction of LHCII with PSII ( state 1 ) [5] , [7] . For many years , attempts were undertaken to elucidate the characteristics and to identify the LHCII protein phosphatase ( s ) . By means of biochemical approaches [20]–[22] , it was shown that protein phosphatases of different families must be involved in the reversible phosphorylation of thylakoid phosphoproteins . A PP2A-like phosphatase was postulated to be responsible for the desphosphorylation of the PSII core proteins [23] , whereas the LHCII phosphatase activity was shown to be dependent on the presence of divalent cations and not to be inhibited by microcystin and okadaic acid [21] , [22] . These findings strongly suggested an involvement of a PP2C-type phosphatase in pLHCII dephosphorylation [24] . Here , we show that the thylakoid protein phosphatase TAP38 is required for pLHCII dephosphorylation and state transitions . In plants with markedly reduced TAP38 levels , hyperphosphorylation of LHCII is associated with enhanced thylakoid electron flow , resulting in more rapid growth under constant low-light regimes . Together with the results of an in vitro dephosphorylation assay , our data indicate that TAP38 dephosphorylates pLHCII directly .
To identify the LHCII phosphatase , we systematically isolated loss-of-function mutants for known chloroplast protein phosphatases and assessed their capacity to dephosphorylate pLHCII ( see below for details ) . However , none of the nine protein phosphatases At3g52180 ( DSP4/SEX4 ) , At4g21210 ( AtRP1 ) , At1g07160 , At3g30020 , At4g33500 , At1g67820 , At2g30170 , At3g10940 , or At4g03415 , demonstrated to reside in the chloroplast [25]–[29] , qualified as the LHCII phosphatase ( unpublished data ) . Next , we extended our search to protein phosphatases tentatively identified as chloroplast proteins by proteomic analyses in A . thaliana [30] , [31] . Of those , the serine/threonine protein phosphatase At4g27800 turned out to be the most promising candidate . Proteins with high homology to At4g27800 exist in mosses and higher plants , but not in algae or prokaryotes . Furthermore , At4g27800 and its homologs share a predicted N-terminal chloroplast transit peptide ( cTP ) , a putative transmembrane domain at their very C-terminus and a protein phosphatase 2C signature ( Figure 1 ) . For A . thaliana , three At4g27800 mRNAs are predicted ( Figure S1A ) . To verify their existence and to distinguish between the different splice forms , reverse-transcriptase PCR analyses were performed . Only At4g27800 . 1 , and much less At4g27800 . 2 , were detectable in leaves , whereas for the At4g27800 . 3 splice variant , no signal could be obtained ( Figure S1B ) . In protoplasts transfected with At4g27800 . 1 fused to the coding sequence for the red fluorescent protein ( RFP ) [32] , the fusion protein localized to chloroplasts ( Figure 2A ) . Chloroplast import assays with the radioactively labeled At4g27800 . 1 protein confirmed the uptake into the chloroplast with concomitant removal of its cTP . Mature At4g27800 . 1 has a molecular weight of ∼38 kDa ( Figure 2B ) . Immunoblot analysis using a specific antibody raised against the mature At4g27800 . 1 protein ( Figure 2C ) detected the protein in thylakoid preparations but not in stromal fractions . It is noteworthy , that the putative translation products At4g27800 . 2 and At4g27800 . 3 ( ∼32 kDa ) were undetectable ( Figure 2C ) . At4g27800 . 1 is therefore the major isoform in leaves , and was renamed TAP38 ( Thylakoid-Associated Phosphatase of 38 kDa ) . Two tap38 insertion mutants , tap38-1 ( SAIL_514_C03 ) [33] and tap38-2 ( SALK_025713 ) [34] , were obtained from T-DNA insertion collections ( Figure S1A ) . In tap38-1 and tap38-2 plants , amounts of TAP38 transcripts were severely reduced , to 10% and 13% of WT levels , respectively ( Figure 3A ) . Conversely , in transgenic lines carrying the TAP38 coding sequence under control of the 35S promoter of Cauliflower Mosaic Virus ( oeTAP38 ) , levels of TAP38 mRNA were much higher than in wild type ( WT ) ( Figure 3A ) . TAP38 protein concentrations reflected the abundance of TAP38 transcripts: tap38-1 and tap38-2 thylakoids had <5% and ∼10% of WT levels , respectively , whereas the oeTAP38 plants displayed >20-fold overexpression on the protein level ( Figure 3B ) . TAP38 protein levels were also determined under light conditions relevant for state transitions ( see Materials and Methods ) . In WT plants , TAP38 was constitutively expressed at similar levels under all light conditions applied ( Figure 3C ) . To determine whether TAP38 is involved in state transitions , chlorophyll fluorescence was measured in WT , tap38 , and oeTAP38 leaves ( Figure 4A ) . Plants were exposed to light conditions that stimulate either state 2 ( red light ) or state 1 ( far-red light ) [35] , [36] , and the corresponding maximum fluorescence in state 2 ( FM2 ) and in state 1 ( FM1 ) values were determined . Because the light intensity chosen to induce state transitions did not elicit photoinhibition ( as monitored by measurements of the maximum quantum yield [FV/FM] ) , changes in FM , the maximum fluorescence , could be attributed to state transitions alone . This allowed us to calculate the degree of quenching of chlorophyll fluorescence due to state transitions ( qT ) [36] . In the tap38 mutants , qT was markedly decreased ( tap38-1 , 0 . 01±0 . 003; tap38-2 , 0 . 03±0 . 001; WT , 0 . 10±0 . 001 ) . In tap38-1 plants complemented with the TAP38 genomic sequence ( including its native promoter ) , qT values were normal , confirming that state transitions require TAP38 . Interestingly , oeTAP38 plants exhibited qT values of about 0 . 01±0 . 001 , indicating that both absence and excess of TAP38 interfere with the ability to undergo reversible state transitions . To determine the antenna sizes of PSII and PSI , 77K fluorescence emission spectra were measured under state 1 ( exposure to far-red light ) and state 2 ( low light ) conditions as described [11] , [12] , [37] ( Figure 4B ) . The spectra were normalized at 685 nm , the peak of PSII fluorescence . In WT , the transition from state 1 to state 2 was accompanied by a marked increase in relative PSI fluorescence at 730 nm , reflecting the redistribution of excitation energy from PSII to PSI . In contrast , in tap38 leaves , the PSI fluorescence peak was relatively high even under state 1-promoting conditions , implying that the mutants were blocked in state 2—i . e . , pLHCII should be predominantly attached to PSI . Additionally , under state 2-promoting light conditions , the PSI antenna size ( expressed as F730/F685 ) was larger in tap38 mutants than in WT ( tap38-1 , 1 . 47; tap38-2 , 1 . 45; WT , 1 . 38; see also Table S1 ) , arguing in favor of the idea that in tap38 plants , a larger fraction of the mobile pool of LHCII can attach to PSI . On the contrary , in oeTAP38 plants , the relative fluorescence of PSI hardly increased at all under conditions expected to induce the state 1→state 2 shift ( Figure 4B; Table S1 ) . This behavior resembles that of stn7 mutants , which are blocked in state 1 , i . e . , with LHCII permanently attached to PSII [12] . It is generally accepted that state transitions require reversible phosphorylation of LHCII [2] , [4] , [5] . Therefore , the phosphorylation state of LHCII was monitored under light conditions that favor state 1 ( dark or far-red light treatment ) or state 2 ( low light ) . Plants with abnormal levels of TAP38 , and WT plants were dark adapted for 16 h ( state 1 ) , then exposed to low light ( 80 µmol m−2 s−1 , 8 h ) ( state 2 ) , and then to far-red light ( 4 . 5 µmol m−2 s−1 , 740 nm ) for up to 120 min to induce a return to state 1 . Thylakoid proteins were isolated after each treatment , fractionated by sodium dodecyl sulfate ( SDS ) -PAGE , and analyzed with a phosphothreonine-specific antibody ( Figure 5 , left panels ) . WT plants showed the expected increase in pLHCII during the transition from state 1 ( dark [D] ) to state 2 ( low light [LL] ) , followed by a progressive decrease in pLHCII upon exposure to far-red light ( FR ) . In tap38 mutants , levels of pLHCII were aberrantly high at all time points , whereas the oeTAP38 plants again mimicked the stn7 phenotype [9] , [12] , displaying constitutively reduced levels of pLHCII . To directly visualize how alterations in LHCII phosphorylation in lines lacking or overexpressing TAP38 affect the distribution of the mobile LHCII fraction between the two photosystems , we subjected thylakoid protein complexes of plants adapted to state 1 ( dark and far-red light treatments ) or state 2 ( low-light treatment ) to nondenaturing Blue-native ( BN ) PAGE [38] ( Figure 5 , right panels ) . In this assay , a pigment–protein complex of about 670 kDa , which represents pLHCII associated with the PSI-LHCI complex [14] , [38] , [39] , can be visualized . Whereas in WT thylakoids , the 670-kDa complex was only observable under state 2 conditions ( Figure 5A , right panel ) , as previously reported [14] , [39]; the constitutive phosphorylation of LHCII in the tap38 mutants was associated with the presence of a prominent band for the 670-kDa complex under all light conditions ( Figure 5B , right panel ) . The formation of the 670-kDa complex was totally prevented in oeTAP38 plants with a block in state 1 and highly reduced levels of pLHCII ( Figure 5C , right panel ) . Two-dimensional ( 2D ) PA gel fractionation confirmed that the pigment–protein complex consists of PSI and LHCI subunits , together with a portion of pLHCII that associates with PSI upon state 1→state 2 transition in WT plants ( Figure 6; [14] ) . Additionally , quantification of the different PSI complexes on 2D PA gels showed that the number of PSI complexes associated with LHCII was increased in the tap38 mutants ( Figures 6B and 6C ) , supporting the findings obtained from the 77K fluorescence analyses . An in vitro dephosphorylation assay was established to assess the capability of TAP38 to directly dephosphorylate pLHCII . To this purpose , an N-terminal His-tag fusion of the TAP38 phosphatase was expressed in Escherichia coli and purified ( see Materials and Methods ) . Solubilized thylakoids from tap38-1 mutant plants were then fractionated by sucrose gradient ultracentrifugation , and the protein fraction enriched in pLHCII was isolated . Subsequently , the pLHCII pigment–protein complex was incubated at 30°C for 2 h either in the presence or absence of the recombinant TAP38 phosphatase . At the end of the incubation period , the reaction mixture was fractionated by SDS-PAGE and subjected to immunoblotting using a phosphothreonine-specific antibody ( Figure 7 ) . Clearly , the addition of the recombinant TAP38 decreased the level of LHCII phosphorylation by about 50% ( relative to the untreated pLHCII sample ) . In the presence of the phosphatase inhibitor NaF , TAP38 addition did not markedly alter the phosphorylation level of LHCII . Taken together , these findings suggest that TAP38 is able to directly dephosphorylate pLHCII . When kept under low-light intensities ( 80 µmol m−2 s−1 ) that favor state 2 , tap38 mutants grew larger than WT plants ( Figure 8A ) , whereas oeTAP38 plants behaved like WT ( unpublished data ) . Detailed growth measurements revealed that the tap38 mutants exhibited a constant growth advantage over WT plants , starting at the cotyledon stage ( Figure 8B ) . Because this difference might be attributable to altered photosynthetic performance , parameters of thylakoid electron flow were measured . The fraction of QA ( the primary electron acceptor of PSII ) present in the reduced state ( 1-qP ) was lower in tap38-1 ( 0 . 06±0 . 01 ) and tap38-2 plants ( 0 . 07±0 . 01 ) than in WT ( 0 . 10±0 . 01 ) , when both genotypes were grown as in Figure 8A and chlorophyll fluorescence was excited with 22 µmol m−2 s−1 actinic red light . Comparable differences in the redox state of the primary electron acceptor persisted up to 95 µmol m−2 s−1 actinic red light ( Figure 8C ) , indicating that the tap38 mutants can redistribute a larger fraction of energy to PSI , in accordance with the increase in its antenna size under state 2 light conditions ( see Figure 4B; Table S1 and Figure 6 ) . This idea was supported by measurements of the maximum ( FV/FM ) and effective ( ΦII ) quantum yields of PSII . FV/FM remained unaltered in mutant plants ( see Figure 8D , dark-adapted plants , photosynthetically active radiation [PAR] = 0 ) , indicating WT-like efficiency of mutant PSII complexes . However , ΦII was increased in tap38-1 ( 0 . 75±0 . 01 ) and tap38-2 ( 0 . 73±0 . 02 ) relative to WT ( 0 . 72±0 . 01 ) , suggesting that electron flow through the thylakoids was more efficient in tap38 mutants ( Figure 8D ) . The improvement in photosynthetic performance of the tap38 mutants was most pronounced under low and moderate illumination ( Figures 8C and 8D ) , as expected from their growth phenotype .
How does TAP38 control LHCII dephosphorylation ? Three possibilities appear plausible: TAP38 ( 1 ) negatively regulates the activity of STN7 ( e . g . , by dephosphorylating it [40] ) , ( 2 ) dephosphorylates LHCII directly , or ( 3 ) forms part of a phosphorylation/dephosphorylation cascade that controls the activity of the LHCII kinase or phosphatase . The observation that oeTAP38 plants , although showing a >20-fold increase in TAP38 levels , still exhibit residual LHCII phosphorylation ( see Figure 5C ) , argues against the idea that TAP38 does inhibit STN7 by dephosphorylation . Differences in TAP38 levels resulted in a clear change in pLHCII levels: although in tap38 mutants a strong reduction in TAP38 led to a constantly high level of pLHCII and an increase in the amount of the PSI-LHCI-LHCII complex , strong overexpression of TAP38 ( oeTAP38 ) caused the complete disappearance of pLHCII attached to PSI , although pLHCII was still present . Taking these observations together , it appears that the TAP38 phosphatase acts specifically on pLHCII associated to PSI-LHCI complexes . Indeed , the dephosphorylation of pLHCII still attached to PSII under state 2-inducing light conditions seems unfavorable in terms of energy efficiency . Interestingly , in WT where pLHCII levels can vary dramatically depending on the light conditions [9] , [12] ( see also Figure 5A ) , TAP38 seems to be constitutively expressed under the different light conditions applied ( see Figure 3C ) . A plausible explanation for this is that TAP38 is constitutively active and directly responsible for the dephosphorylation of pLHCII . For that , TAP38 would need to be present in a certain concentration range ( as it is the case for WT ) to constantly dephosphorylate pLHCII . In agreement with that , thylakoid protein phosphatase reactions have been described as redox independent , leading to the conclusion that the redox dependency of LHCII phosphorylation is a property of the kinase reaction [41] . This , together with the observation that Stt7 levels increase under prolonged state 2 conditions ( favoring LHCII phosphorylation ) and decrease under state 1 conditions ( favoring dephosphorylation of LHCII ) [17] , argues in favor of the hypothesis that the LHCII kinase is the decisive factor in controlling the phosphorylation state of LHCII . Despite the obvious TAP38 dosage dependence of pLHCII dephosphorylation ( see Figures 5 and 6 ) , TAP38 activity could be regulated on other levels than only its abundance . However , the strong decrease or increase of TAP38 levels in tap38 mutant and oeTAP38 plants might interfere with other types of regulation in these genotypes . As outlined above , the dependence of LHCII dephosphorylation upon TAP38 dosage—when comparing tap38 mutants , WT , and TAP38 overexpressors—strongly suggests that TAP38 dephosphorylates pLHCII directly , particularly when it is associated with the PSI-LHCI complex . Alternatively , TAP38 could act in a phosphorylation/dephosphorylation cascade that controls the activity of the LHCII phosphatase . Although the latter hypothesis cannot be totally excluded , a set of evidences point to a direct role of TAP38 on LHCII phosphorylation . Indeed , our in vitro dephosphorylation assay clearly indicated that TAP38 can dephosphorylate pLHCII directly ( see Figure 7 ) . Moreover , as in the case of STN kinases , extensive efforts searching to identify other LHCII phosphatase candidates failed: knockout lines for all the protein phosphatases demonstrated to be located in the chloroplast [25]–[29] did not show any alteration in LHCII phosphorylation . Additionally , extensive biochemical studies did not reveal the existence of a complex network of phosphatases involved in LHCII dephosphorylation , but postulated the involvement of only two distinct chloroplast protein phosphatases from different families in the dephosphorylation of thylakoid phosphoproteins [20]–[23] , [42] . Our data support this notion , as shown by the absence of major alterations in the phosphorylation pattern of CP43 , D1 , and D2 subunits in tap38 mutant plants ( see Figure 5 ) . Moreover , pLHCII dephosphorylation was suggested to be catalyzed by only two independent protein phosphatases , a membrane-bound one and a stromal protein phosphatase [42] . In contrast to this , our results clearly show that TAP38 , a thylakoid-associated phosphatase , alone is responsible for LHCII dephosphorylation . Thus , although slightly leaky , the tap38-1 mutants show a large fraction of LHCII in the phosphorylated state under all investigated conditions ( see Figure 5 ) . If a second LHCII phosphatase with redundant function would operate in chloroplasts , one would expect some residual dephosphorylation of pLHCII . A plausible explanation for the previously shown stromal pLHCII dephosphorylation activity [22] might be that during the preparation of stromal extracts , a significant portion of TAP38 was released from the thylakoid membrane into the stroma . Interestingly , TAP38 appears to influence also the phosphorylation levels of other thylakoid proteins , as shown by the higher phosphorylation of the CAS protein in tap38-1 thylakoids ( see Figure 5 ) . Taking these observations together , it appears that , as in the case of the STN kinases , two distinct phosphatases are needed to dephosphorylate LHCII and PSII core proteins . TAP38 , similar to the STN7 kinase , seems to have a high specificity for pLHCII associated with PSI-LHCI complexes as substrate . The counterpart of STN8 [9] , [43] , the PSII core–specific phosphatase , remains to be identified . However , as in the case of the STN7 and STN8 kinases , some degree of substrate overlap seems to exist also between the phosphatases , as shown by the more rapid dephosphorylation of PSII-D1/D2 subunits in the TAP38 overexpressor lines exposed to far-red light conditions ( see Figure 5C ) . Additionally , it is noteworthy that the activity of TAP38 does not seem to be restricted to STN7 substrates , as shown by its influence on CAS protein phosphorylation , previously reported to be a substrate of the STN8 kinase [44] . It is known that an increase in the relative size of the reduced fraction of the plastoquinone pool ( PQH2 ) enhances phosphorylation of LHCII [1] , [5] , [39] , [45] . Depletion of TAP38 in tap38 mutants , however , increases both LHCII phosphorylation ( see Figure 5B ) and PQ oxidation ( see 1-qP values in Figure 8C ) . This discrepancy can be resolved by assuming that the enhanced oxidation of PQ caused by the increase in PSI antenna size ( and LHCII phosphorylation ) in tap38 plants is not sufficient to down-regulate the LHCII kinase to such an extent that it can compensate for the decline in LHCII dephosphorylation . The enhanced photosynthetic performance indicated by an increase in ΦII and a decrease of 1-qP ( see Figure 8C and 8D ) , as well as the growth advantage of the tap38 mutants under constant moderate-light intensities that stimulate LHCII phosphorylation and state 2 , can be attributed to the redistribution of a larger fraction of energy to PSI . This is in accordance with the increase in PSI antenna size in tap38 mutants when compared to WT plants ( see Figure 4B , Table S1 , and Figure 6 ) . Therefore , it is straightforward to speculate that the enhanced PSI antenna size provides the tap38 mutants with a more robust photosynthetic electron flow under conditions that preferentially excite PSII and induce state 2 . As a consequence of the more balanced light reaction , the photosynthetic efficiency is improved resulting in an increased growth rate . However , the fitness advantage will revert under conditions that induce state 1 , or under more natural conditions with fluctuating light; here , it can be expected that tap38 mutants will perform less efficiently than the WT with respect to photosynthesis and growth , very similar to what has been observed for the stn7 mutant [12] , [13] . Taken together , future analyses should clarify which protein phosphatase is involved in the dephosphorylation of PSII core proteins and which are the counterparts of higher plant phosphatases , including TAP38 , in Chlamydomonas ( which apparently lacks a TAP38 ortholog ) . Additionally , further biochemical evidences that TAP38 ( and STN7 ) uses pLHCII as a substrate will be very important for the complete molecular dissection of state transitions .
Procedures for plant propagation and growth measurements have been described elsewhere [46] . The tap38-2 insertion line ( SALK_025713 ) was identified in the SALK collection [34] ( http://signal . salk . edu/ ) , whereas insertion line tap38-1 ( SAIL_514_C03 ) originated from the Sail collection [33] . Both lines were identified by searching the insertion flanking database SIGNAL ( http://signal . salk . edu/cgi-bin/tdnaexpress ) . To generate oeTAP38 lines , the coding sequence of TAP38 was cloned into the plant expression vector pH2GW7 ( Invitrogen ) . For complementation of the tap38-1 mutant , the TAP38 genomic DNA , together with 1 kb of its natural promoter , was ligated into the plant expression vector pP001-VS . The constructs were used to transform flowers of Col-0 or tap38-1 mutant plants by the floral dipping technique as described [47] . Transgenic plants , after selection for resistance to hygromycin ( oeTAP38 ) or Basta herbicide ( complemented tap38-1 ) , were grown on soil in a climate chamber under controlled conditions ( PAR: 80 µmol m−2 s−1 , 12/12 h dark/light cycles ) . The T2 generation of the oeTAP38 plants was used for the experiments reported . Successful complementation of tap38-1 mutants was confirmed by measurements of chlorophyll fluorescence and LHCII phosphorylation levels under light regimes promoting state transitions . The full-length coding region of the TAP38 gene was cloned into the vector pGJ1425 , in frame with , and immediately upstream of the sequence encoding dsRED [32] . Isolation , transfection , and fluorescence microscopy of A . thaliana protoplasts were performed as described [48] . The coding region of TAP38 was cloned into the pGEM-Teasy vector ( Promega ) downstream of its SP6 promoter region , and mRNA was produced in vitro using SP6 RNA polymerase ( MBI Fermentas ) . The TAP38 precursor protein was synthesized in a Reticulocyte Extract System ( Flexi; Promega ) in the presence of [35S]methionine . Aliquots of the translation reaction were incubated with intact chloroplasts , and protein uptake was analyzed after treatment of isolated chloroplasts with thermolysin ( Calbiochem ) as described previously [49] . Labeled proteins were subjected to SDS-PAGE and detected by phosphorimaging ( Typhoon; Amersham Biosciences ) . Total RNA was extracted with the RNeasy Plant Mini Kit ( QIAGEN ) according to the manufacturer's instructions . cDNA was prepared from 1 µg of total RNA using the iScript cDNA Synthesis Kit ( Bio-Rad ) according to the manufacturer's instructions . For semiquantitative reverse-transcriptase PCR , cDNA was diluted 10-fold , and 3 µl of the dilution was used in a 20-µl reaction . Thermal cycling consisted of an initial step at 95°C for 3 min , followed by 30 cycles of 10 s at 95°C , 30 s at 55°C , and 10 s at 72°C . For real-time PCR analysis , 3 µl of the diluted cDNA was mixed with iQ SYBR Green Supermix ( Bio-Rad ) . Thermal cycling consisted of an initial step at 95°C for 3 min , followed by 40 cycles of 10 s at 95°C , 30 s at 55°C , and 10 s at 72°C , after which a melting curve was performed . Real-time PCR was monitored using the iQ5Multi-Color Real-Time PCR Detection System ( Bio-Rad ) . All reactions were performed in triplicate with at least two biological replicates . Total protein extracts and proteins from total chloroplasts , thylakoids , and the stroma fraction were prepared from 4-wk-old leaves in the presence of 10 mM NaF as described [48] , [50] . Immunoblot analyses with phosphothreonine-specific antibodies ( Cell Signaling ) or polyclonal antibodies raised against the mature TAP38 protein were performed as described [45] . For BN-PAGE , thylakoid membranes were prepared as described above . Aliquots corresponding to 100 µg of chlorophyll were solubilized in solubilization buffer ( 750 mM 6-aminocaproic acid; 5 mM EDTA [pH 7]; 50 mM NaCl; 1 . 5% digitonin ) for 1 h at 4°C . After centrifugation for 1 h at 21 , 000g , the solubilized material was fractionated by nondenaturing BN-PAGE at 4°C as described [38] . For 2D-PAGE , samples were fractionated in the first dimension by BN-PAGE as described above and subsequently by denaturing SDS-PAGE as described previously [51] . Densitometric analysis of the stained gels was performed using the Lumi Analyst 3 . 0 ( Boehringer ) . State transitions were measured by pulse amplitude modulation fluorometry ( PAM ) [35] , [36] and 77 K fluorescence emission analysis [12] , [37] . Plants adapted to state 1 conditions were obtained by incubation either in darkness or far-red light , whereas state 2 was induced by either red- or low-light illumination . Both state 1 and state 2 light-inducing conditions were used in different combinations , since they resulted in identical effects on state transitions . Additionally , there was no major reason to prefer one light setting to the other , except for the fact that the PAM fluorometer is equipped with red and far-red lights . For state transition measurements , five plants of each genotype were analyzed , and mean values and standard deviations were calculated . In vivo chlorophyll a fluorescence of single leaves was measured using the Dual-PAM 100 ( Walz ) . Pulses ( 0 . 5 s ) of red light ( 5 , 000 µmol m−2 s−1 ) were used to determine the maximum fluorescence and the ratio ( FM−F0 ) /FM = FV/FM . Quenching of chlorophyll fluorescence due to state transitions ( qT ) was determined by illuminating dark-adapted leaves with red light ( 35 µmol m−2 s−1 , 15 min ) and then measuring the maximum fluorescence in state 2 ( FM2 ) . Next , state 1 was induced by adding far-red light ( maximal light intensity corresponding to level 20 in the Dual-PAM setting , 15 min ) , and recording FM1 . qT was calculated as ( FM1−FM2 ) /FM1 [36] . For 77 K fluorescence emission spectroscopy , the fluorescence spectra of thylakoids were recorded after irradiating plants with light that favored excitation of PSII ( 80 µmol m−2 s−1 , 8 h ) or PSI ( LED light of 740 nm wavelength , 4 . 6 µmol m−2 s−1 , 2 h ) . Thylakoids were isolated in the presence of 10 mM NaF as described [11] , and 77 K fluorescence spectra were obtained by excitation at 475 nm using a Spex Fluorolog mod . 1 fluorometer ( Spex Industries ) . The emission between 600 and 800 nm was recorded , and spectra were normalized relative to peak height at 685 nm . Data frequency was of 0 . 5 nm with an integration time of 0 . 1 s . pLHCII was obtained from fractionation of tap38-1 thylakoids by sucrose gradient ultracentrifugation as previously described [45] . The cDNA sequence of mature TAP38 was cloned into pET151 ( Invitrogen ) , and recombinant TAP38 ( recTAP38 ) was expressed in the E . coli strain BL21 with a N-terminal-6x His-tag . recTAP38 was purified under denaturing conditions following a Ni-NTA batch purification procedure according to the manufacturer's instructions ( Qiagen ) . After protein precipitation in 10% trichloroacetic acid ( TCA ) followed by three washing steps with absolute ethanol , around 500 µg of TAP38 protein were resuspended in 500 µl of 1% ( w/v ) lithium dodecyl sulfate ( LDS ) , 12 . 5% ( w/v ) sucrose , 5 mM ε-aminocaproic acid , 1 mM benzamidine , and 50 mM HEPES KOH ( pH 7 . 8 ) , as previously described [52] . Subsequently , TAP38 protein was boiled for 2 min at 100°C and then transferred for 15 min at 25°C . Then , dithiothreitol ( DTT; 75 mM final concentration ) was added , and the solution was subjected to three freezing-thawing cycles ( 20 min at −20°C , 20 min at −80°C , 20 min at −20°C , thawing in a ice-water bath , and 5 min at 25°C ) . After completion of the three freezing-thawing cycles , octyl-glucopyranoside ( OGP; 1% [w/v] final concentration ) was added , and the solution was kept on ice for 15 min . Afterwards , KCl ( 75 mM , final concentration ) was added to precipitate the LDS detergent . After centrifugation at 16 , 000g at 4°C for 10 min , the supernatant containing the refolded TAP38 in the presence of 1% ( w/v ) OGP was collected . Subsequently , 1 µl of phosphatase was incubated together with pLHCII corresponding to 2 µg of total chlorophyll . The dephosphorylation reaction was performed in 50 µl containing 0 . 06% ( w/v ) dodecyl-ß-D-maltoside , 5 mM Mg-acetate , 5 mM DTT , 100 mM HEPES ( pH 7 . 8 ) , at 37°C for 2 h as previously described [22] . The reaction mixture was loaded on a SDS-PAGE and immunodecorated with a phosphothreonine-specific antibody , as described above . | Plants are able to adapt photosynthesis to changes in light levels by adjusting the activities of their two photosystems , the structures responsible for light energy capture . During a process called state transitions , a part of the photosynthetic complex responsible for light harvesting ( the photosynthetic antennae ) becomes reversibly phosphorylated and migrates between the photosystems to redistribute light-derived energy . The protein kinase responsible for phosphorylating photosynthetic antenna proteins was identified recently . However , despite extensive biochemical efforts to isolate the enzyme that catalyzes the corresponding dephosphorylation reaction , the identity of this protein phosphatase has remained unknown . In this study , we identified and characterized the thylakoid-associated phosphatase TAP38 . We first demonstrate by spectroscopic measurements that the redistribution of excitation energy between photosystems that are characteristic of state transitions do not take place in plants without a functional TAP38 protein . We then show that the phosphorylation of photosynthetic antenna proteins is markedly increased in plants without TAP38 , but decreased in plants that express more TAP38 protein than wild-type plants . This , together with the observation that addition of recombinant TAP38 decreases the level of antenna protein phosphorylation in an in vitro assay , suggests that TAP38 directly acts on the photosynthetic antenna proteins as the critical phosphatase regulating state transitions . Moreover , in plants without TAP38 , photosynthetic electron flow is enhanced , resulting in more rapid growth under constant low-light regimes , thus providing the first instance of a mutant plant with improved photosynthesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"biology/plant",
"biochemistry",
"and",
"physiology",
"plant",
"biology/plant-environment",
"interactions",
"plant",
"biology"
] | 2010 | Role of Plastid Protein Phosphatase TAP38 in LHCII Dephosphorylation and Thylakoid Electron Flow |
Despite extensive research on the mechanisms of HLA-mediated immune control of HIV-1 pathogenesis , it is clear that much remains to be discovered , as exemplified by protective HLA alleles like HLA-B*81 which are associated with profound protection from CD4+ T cell decline without robust control of early plasma viremia . Here , we report on additional HLA class I ( B*1401 , B*57 , B*5801 , as well as B*81 ) , and HLA class II ( DQB1*02 and DRB1*15 ) alleles that display discordant virological and immunological phenotypes in a Zambian early infection cohort . HLA class I alleles of this nature were also associated with enhanced immune responses to conserved epitopes in Gag . Furthermore , these HLA class I alleles were associated with reduced levels of lipopolysaccharide ( LPS ) in the plasma during acute infection . Elevated LPS levels measured early in infection predicted accelerated CD4+ T cell decline , as well as immune activation and exhaustion . Taken together , these data suggest novel mechanisms for HLA-mediated immune control of HIV-1 pathogenesis that do not necessarily involve significant control of early viremia and point to microbial translocation as a direct driver of HIV-1 pathogenesis rather than simply a consequence .
The factors underlying HIV-1 pathogenesis are a complex interplay between the host and virus . We have previously shown that viral characteristics , such as the replicative fitness of the transmitted variant , are significant predictors of early immune activation and further disease progression in HIV-1 infected individuals [1] . With respect to host factors , a series of genome-wide association studies ( GWAS ) have identified the HLA I locus to be the primary site of polymorphisms significantly affecting HIV-1 disease outcome [2–4] . Variations at this locus have also been linked to the efficiency of T cell control of viremia [5 , 6] , and viremia is a strong predictor of HIV-1 disease progression [7–10] . Specific HLA class I alleles , such as B*57 and B*27 , are consistently enriched in HIV-1 controllers , reinforcing the role of these genes in viral suppression [11] . However , the role of HLA may extend beyond viral control; in a multi-country HIV cohort study , the HLA class I allele , B*81 , was shown to be associated with protection from CD4+ T cell decline without significant control of plasma viremia [12] , suggesting that there are additional mechanisms of HLA-mediated CD4+ T cell protection distinct from early control of plasma viral load . We propose that additional immunogenetic factors with this protective phenotype exist , and furthermore , that these factors alter early events post infection , before the establishment of a viral load set point . The most potent clinical indicator of HIV-1 pathogenesis and the severe immune depletion of AIDS is the CD4+ T cell count in the peripheral blood . CD4+ T cell decline during HIV-1 infection has been shown to be more accurately predicted by levels of chronic immune activation than viral load , underlining the importance of both host and virus in pathogenesis [13–15] . The determinants of this chronic immune activation are multiple , and include viral activity along with microbial translocation [16–18] . Microbial translocation is a major contributor to immune activation and disease progression in HIV-1 infection [19] . The early loss of gut mucosal integrity with the infection and depletion of CD4+ T cells locally is thought to facilitate the transit of bacterial components into the bloodstream [20] . Studies have linked both microbial DNA and LPS levels to T cell activation and amplification of viremia [21] . However , there are differences in baseline levels of detectable microbial translocation in specific populations , with evidence of ongoing immune activation in HIV seronegative MSM [22] , suggesting that some of this pathogenesis may have host determinants . Further work has demonstrated that the composition of the microbial community in the female genital tract has a significant effect on the local inflammatory environment [23] . In addition , specific HLA alleles have been linked to particular microbial community phenotypes , and in turn to inflammatory disease susceptibility [24–27] . Taken together , there is strong evidence for the importance of microbial translocation in immune activation and HIV-1 pathogenesis , and that the composition of the microbiome is partially determined by host immunogenetics . To elucidate novel mechanisms of immune protection , we sought to identify additional HLA alleles associated with the phenotype of CD4+ T cell preservation , irrespective of their associations with plasma viremia . This approach allows us to uncover alternate mechanisms by which the host cellular immune response alters disease trajectory following acute infection . We identified 4 HLA class I alleles and 2 HLA class II alleles significantly associated with protection from rapid CD4+ T cell decline even when correcting for early set point viral load , sex of the infected individual , and viral replicative capacity ( vRC ) . Furthermore , the HLA class I alleles we identified were associated with a reduction in levels of circulating LPS during acute infection , and appear to be associated with CD8+ T cell responses targeting more conserved regions of Gag . Early LPS levels were a significant predictor of CD4+ T cell decline independent of plasma viral load , and were further associated with cellular immune activation , specifically in the T cell compartment . Collectively , these data suggest that microbial translocation is a driver of immune activation rather than a consequence , establishing early gut damage and microbial translocation as a potential target for therapeutic intervention at both the acute and chronic stages of HIV infection .
To identify the contribution of protective immunogenetic factors to disease progression in a Zambian heterosexual transmission cohort , we performed HLA class I and class II typing to 4-digit resolution in a group of 127 acutely infected subjects ( S1 Table ) . Detailed clinical data was collected and plasma viral load and CD4+T cell counts measured longitudinally starting from the time of HIV-1 diagnosis ( median of 44 days post estimated date of infection ) and continuing at 3-month intervals thereafter up to 6 years post infection . We have previously defined viral replicative capacity ( vRC ) for the transmitted viruses in this cohort , as measured by an in vitro replicative fitness assay [1 , 28] and that data was integrated into our analysis . We employed Cox proportional hazards models coupled with a stepwise backward variable selection approach , to identify both HLA class I and II alleles associated with protection from significant CD4+ T cell decline , which we defined as the time to a CD4+ T cell count less than 300 cells/mm3 . Sex of the infected participant and vRC of the transmitted virus , which represent additional host and viral characteristics previously shown to significantly affect HIV-1 pathogenesis [1 , 29–31] , were added to the model as static covariates during the backwards selection process . In a multivariable Cox proportional hazards model , 4 HLA class I alleles ( B*1401 , B*57 , B*5801 , and B*81 ) and 2 HLA class II alleles ( DQB1*02 and DRB1*15 ) were found to be independent protective factors delaying the loss of CD4+ T cells ( Table 1 ) . As is implied by their independent predictive nature , HLA alleles affecting CD4+ T cell decline were additive in their effects and formed distinct profiles even in the absence of sex and vRC as covariates ( Fig 1A ) . Furthermore , these HLA alleles generated additive scores when divided into HLA class I or class II alleles only ( S1 Fig ) . In contrast , a significant additive effect of the protective alleles was not observed on set point viral load ( Fig 1B ) , suggesting that these alleles may affect immunopathogenesis without a substantial impact on early plasma viremia in this cohort . To further interrogate this observation , we sought to isolate the effects from the influence of viral load . Plasma viremia , and specifically viral load set point , is a consistent and strong predictor of CD4+ T cell decline in this and multiple other HIV infection cohorts [1 , 32 , 33] . Classically , HLA alleles that are protective in the context of HIV-1 pathogenesis have been associated with significant control of plasma viremia , especially early in infection [4 , 34] . As the initial analyses ( Table 1 ) assessed HLA alleles for their effect on CD4+ T decline without controlling for set point viral load we built a multivariable Cox proportional hazards model where set point viral load was added as a covariate . We found that all 6 HLA alleles remained significant predictors with set point viral load in the model ( S2 Table ) . Indeed , only B*57 was associated with a statistically significant decrease in early set point viral loads when analyzed in isolation ( S2 Fig ) , an association which has been described previously [35] . These data suggest that these particular HLA alleles are exerting their protective effect on CD4+ decline via an alternative mechanism and not simply through the well-described control of early plasma viremia . Inflammation is an alternative driver of disease progression , independent of viral load [36] , and is frequently associated with gut damage and bacterial translocation [37] . We therefore investigated associations between the protective HLA alleles identified in this study and inflammatory markers linked to pathogenesis . We found a striking association between the identified protective HLA class I alleles ( B*1401 , B*57 , B*5801 , and B*81 ) and lower plasma LPS levels at the earliest time of sampling ( median of 44 days post estimated date of infection; Fig 2A ) . This difference in LPS levels was specific for class I alleles; we observed no association between protective HLA class II alleles and circulating LPS ( Fig 2B ) . The grouped protective class I alleles were also associated with a moderate 0 . 31 log reduction in plasma viral loads at the time of LPS sampling ( Fig 2C ) . However , in a generalized linear model , carriage of protective class I alleles is a significant predictor of LPS levels at seroconversion ( p = 0 . 008 ) , whereas plasma viral loads are not ( p = 0 . 66 ) ( Fig 2D ) . Protective HLA class I alleles were also associated with reduced levels of LPS , soluble CD14 , and intestinal fatty acid-binding protein ( I-FABP ) at 6 months post seroconversion ( Fig 3A–3C ) in a subset of individuals with samples available for analysis ( n = 30 ) . These data indicate that the LPS association is durable and is reinforced by other markers of gut integrity and microbial translocation . Furthermore , individuals carrying these protective HLA-I alleles , but not those lacking them , demonstrated a significant reduction in IL-10 levels over time ( Fig 3D ) , consistent with published literature linking the presence of circulating microbial products with increased production of IL-10 by monocytes [38 , 39] . Other factors such as excessive alcohol use can significantly affect gut integrity and bacterial dysbiosis leading to increased levels of microbial translocation [40–42] . In this cohort , data on alcohol consumption that was previously collected [43 , 44] was used to generate scores identifying individuals who consumed alcohol excessively and those that did not . We observed a significant elevation in plasma LPS at seroconversion in individuals that consumed alcohol excessively ( S3 Fig ) . Importantly , in a generalized linear model , excessive alcohol consumption and protective HLA class I alleles were independent predictors of plasma LPS levels near seroconversion ( S3 Fig ) , suggesting the association between HLA class I and LPS is not significantly confounded by alcohol consumption in this cohort . The MHC gene locus contains additional genes that play key roles in the innate and adaptive immune systems[45] . Many of the polymorphisms in this region are in strong linkage disequilibrium ( LD ) [46] , raising the possibility that the observed associations between these 4 HLA class I alleles and microbial translocation may be attributed to other genes in this locus . To rule out this possibility , we analyzed high resolution ImmunoChip data to look for other polymorphisms in the MHC locus in strong LD with these four HLA haplotypes . We observed no SNPs with r2 values >0 . 8 ( S4 Fig ) . This suggests that the observed reduction in microbial translocation is likely attributable to the presence of these 4 protective HLA class I alleles and not other immune-related genes in the MHC locus . We next assessed whether a direct relationship between LPS levels and CD4+ T cell decline existed in this cohort . While translocation of microbial products , such as LPS , has been linked to immune activation and disease progression during chronic HIV-1 infection [19] , we sought to specifically establish a link between very early microbial translocation and subsequent CD4+ T cell loss . In a Kaplan-Meier survival analysis with an endpoint defined as CD4+ T cell counts falling below 300 cells/mm3 , individuals in the lowest 50th percentile of LPS levels at seroconversion exhibited significantly slower CD4+ T cell decline ( Fig 4 ) . This association was independent of viral load at the time of LPS sampling ( S3 Table ) . Thus , this measure of microbial translocation early during HIV-1 infection is directly associated with the kinetics of longitudinal CD4+ T cell decline . In chronic HIV-1 infection , systemic LPS levels are correlated with increased cellular immune activation , specifically with an increased frequency of activated CD38+/HLA-DR+ CD8 T cells [19] . The mechanisms of this effect may include both direct stimulation of T cells by microbial products [47] , as well as secondary effects mediated by innate recognition and inflammatory cytokine production [22 , 48] . In order to determine the effect of early microbial translocation on subsequent T cell activation in this cohort , cryopreserved PBMCs available from a subset of individuals collected 3 months after the estimated date of infection ( median 97 days post estimated date of infection and median 65 days post systemic LPS measurements ) were assessed by flow cytometry for markers of T cell activation and antigen experience/exhaustion . We found that LPS levels at seroconversion directly correlated with PD-1 expression on central memory and effector memory CD4+ T cells an average of two months later ( Fig 5A and 5B ) . Similarly , in the CD8+ T cell compartment , increased LPS in the periphery also correlated with the percentage of CD38 and HLA-DR double positive cells , increased percentages of Ki67+ cells , and CD38 expression levels ( assessed by MFI ) ( Fig 5C–5E ) . Furthermore , the associations between LPS and these activated T cell phenotypes were independent of viral loads at the time of PBMC collection ( S4 Table ) , providing further evidence for early microbial translocation as an independent predictor of immune activation and a potential driver of immunopathogenesis . In order to gain insight regarding the quality of the CD8+ T cell response restricted by these protective HLA class I alleles , we analyzed IFNγ ELISpot data for a subset of individuals in this cohort for which early cryopreserved PMBCs were available . When PBMCs were interrogated with potential T cell epitope ( PTE ) Gag pools [49 , 50] , we observed no difference in the magnitude of responses between those with or without protective HLA class I alleles , even when controlling for the relatedness of the transmitted gag sequence to the peptides contained in the PTE Gag pools ( Fig 6A ) . Furthermore , we observed no differences in the breadth of Gag targeting when analyzing the number of specific epitopes targeted between HLA-I groups ( Fig 6B ) . However , when peptide pools were optimized for sequence conservation , we observed that individuals with the protective HLA class I alleles identified in this study exhibited significantly higher IFNγ ELISpot responses to a pool of highly conserved Gag peptides ( Fig 6C ) [51–53] . Targeting more conserved regions of Gag , which are unlikely to rapidly escape due to fitness constraints [54–58] , may help to explain the sustained protective effect of these alleles with the respect to longitudinal CD4+ T cell preservation .
In this study , we have utilized detailed clinical , virologic , and immunogenetic data from a well-defined heterosexual transmission cohort [1] to identify protective HLA alleles associated with attenuated CD4+ T cell decline irrespective of reduced early set point viral loads . Selected HLA class I ( HLA-I ) alleles were associated with significantly lower levels of systemic LPS during the early stages of infection , suggesting that this may be an unrecognized mechanism of protection associated with cellular immune responses restricted by favorable HLA alleles . This association was durable throughout the early stages of infection , as protective HLA class I alleles were favorably associated with additional markers of gut integrity and microbial translocation at subsequent time points . The discovery of HLA alleles associated with protection from rapid CD4+ T cell decline without significant control of early plasma viremia is not without precedent . HLA-B*81 has been shown to be highly protective in this and in other African cohorts without significant control of early set point viral loads [12] . The importance of set point viral loads not withstanding [32 , 33] , the study by Prentice et al . , as well as the current study , serve to highlight other critical determinants of longitudinal pathogenesis as measured by CD4+ T cell counts . We have previously demonstrated that higher viral replication capacity was linked with increased inflammation and immune activation early in infection , setting the scene for subsequent pathogenesis [1] . We concluded that elevated immune activation , linked to transmission of viruses with higher replication capacity , was responsible for more rapid CD4+ T cell decline . In this present study , we observe a similar phenomenon , where the identified HLA-I alleles were protective with respect to CD4+ T cell decline independent of early set point viral loads and viral replicative capacity . These alleles were also associated with lower levels of microbial translocation as measured by systemic LPS . This leads us to hypothesize that these protective alleles may be exerting their effects of preserving CD4+ T cell counts by limiting early immune activation , perhaps by controlling in situ viral replication in key tissues , such as the gastrointestinal tract , a site of early HIV replication post transmission [59–61] . In contrast to the class I alleles identified in this study , the protective HLA class II alleles ( DQB1*02 and DRB1*15 ) were not associated with reduced levels of LPS in early infection . The protective mechanism of HIV-specific CD4+ T cells is less clear than CD8+ T cells , especially since HIV-specific CD4+ T cells are preferentially infected and depleted by HIV during infection [62] . However , Ranasinghe et al . have previously identified one of these same alleles , DRB1*15 , as significantly associated with control of plasma viremia in a subtype B infected cohort [63] . In that study , DRB1*15 alleles were shown to present a greater range of HIV-derived peptides , suggesting that these class II alleles may prime broader CD4+ T cell responses against HIV . It is not clear why we did not observe a similar association between DRB1*15 and control of viremia , but it is most likely due to the distinct nature of the cohorts studied ( subtype B infected and of European descent compared to subtype C infected African populations ) as well as differences in the 4-digit typing of the alleles identified , where DRB1*1503 was the most common variant in this Zambian cohort and DRB1*1502 was most prevalent in the cohort studied by Ranasinghe et al . Translocation of bacterial-derived products into the systemic circulation was first shown to be associated with pathogenesis in HIV-1 infection by Brenchley et al . , specifically in the chronic stages of infection [19] . It has been proposed that specific loss of Th17 CD4+ T cells in gut-associated lymphoid tissue ( GALT ) contributes to structural damage seen at the mucosal barrier , allowing subsequent translocation of bioactive bacterial products into the periphery [37 , 64 , 65] . Bacterial products such as LPS have been associated with increased immune activation , a driver of HIV pathogenesis that is independent of plasma viral loads . Although depletion of total CD4+ T cells ( including Th17 ) in the gut lamina propria occurs early during acute infection , it has been difficult to unravel the cause/effect relationship between microbial translocation and immune activation in the context of ongoing viral replication . Several in vitro and in vivo studies have attempted to understand this relationship . In vitro studies have shown that LPS alone , as well as other TLR agonists , can drive central and effector memory CD4+ T cells into cell cycle and lead to expression of activation markers on CD8+ T cells [47] . Moreover , in vivo studies of a high-risk HIV-negative MSM cohort have shed light on the specific effects of microbial translocation and systemic LPS on cellular immune activation [22 , 48] . In these studies , HIV-negative individuals with detectable but subclinical levels of endotoxemia exhibited significantly elevated levels of proinflammatory cytokines such as TNFa , IP-10 , and MCP-1 , as well as significantly lower CD4/CD8 T cell ratios . Fluctuations in naturally occurring endotoxemia in the referenced HIV-negative cohort were also associated with altered CD4+ T cell proliferation profiles and altered monocyte function . Specifically , elevated LPS was associated with a reduced proliferative capacity in the CD4+ T cell compartment , which has clear implications for CD4+ T cell decline in HIV infection . It is not known if the findings presented in the current study can be generalized to other populations , such as MSM , living with HIV . Innate immune cell persistence in the gut is linked to the composition of the microbiome [66] and specific microbiota features have been shown to enhance inflammation [67] underlining the significance of gut barrier maintenance and a healthy microbiome . Moreover , there is evidence that both sex and sexual preferences can influence the composition of the microbiome , leaving open questions about the intersection between gut dysbiosis , HIV infection , and systemic inflammation[68–71] . Nevertheless , in our study cohort of both men and women , the findings linking LPS to outcomes were independent of the sex of the participants , suggesting a more generalized phenomenon potentially linked to inflammation . We demonstrate in this current study that LPS levels in the plasma measured at or near seroconversion predict longitudinal CD4+ T cell decline , with individuals in the lowest 50th percentile of LPS values protected from rapid loss of CD4+ T cells . In addition , we observed correlations between circulating LPS levels at seroconversion and markers of T cell activation ( CD38 MFI and CD38/HLA-DR co-expression ) , antigen experience/exhaustion ( PD-1 expression , specifically on CD4+ memory T cells ) , and proliferation ( Ki67 expression ) at a subsequent 3-month time point , suggesting a causal link between circulating LPS and immune activation , which itself is strongly correlated with disease progression irrespective of viral loads [36] . The associations between circulating LPS and cellular immune activation are largely unaffected by the inclusion of plasma viral load at the time of PBMC sampling in a multivariable general linearized model . These data suggest that , at some of the earliest time points post infection , LPS is a driver of immune activation and establishes LPS as a predictor of CD4+ T cell decline that is independent of plasma viremia . Interestingly , we did not observed any difference between individuals in terms of magnitude or breadth of epitope recognition as measure by IFNγ ELISpot response between those that expressed or lacked CD4-protective HLA class I alleles when using broad potential T cell epitope ( PTE ) Gag pools [49 , 50] . However , we did observe a more robust response to conserved regions of the Gag protein [51 , 52] in individuals expressing CD4-protective HLA-I alleles . This data demonstrates that the quality of the cytotoxic T lymphocyte response restricted by these HLA-I alleles may be responsible for their efficacy . Collectively , these data suggest that CD4-protective HLA-I alleles , perhaps through their ability to target biologically relevant epitopes , can initiate cellular immune responses capable of maintaining gut integrity and limiting the translocation of microbial products into the systemic circulation . This is possibly via robust control of viral replication at specific tissue sites , at a time when we do not observe significant control of plasma viremia . This hypothesis is supported by a study by Altfeld et al . , where the authors observed a striking disconnect between cytotoxic T lymphocyte ( CTL ) responses measured in the peripheral blood and responses measured in the lymph nodes after structured treatment interruptions [72] . Indeed , a subset of the measured CTL responses were found exclusively in the lymph node , suggesting the cellular immune response can be compartmentalized . In addition , the influence of HLA genotype on the composition of the gut microbiome may contribute; with specific HLA types supporting the growth of more favorable microbial communities [25] . In summary , these data definitively link the presence of microbial products in the circulation during acute/early infection with subsequent CD4+ T cell decline and cellular immune activation . The levels of bacterial products found in the periphery , as measured by plasma LPS levels are modulated by protective HLA class I alleles at both seroconversion and 6 months post infection . These findings demonstrate an unappreciated mechanism of protection by favorable HLA alleles , beyond simply controlling systemic viral load , and further establish microbial translocation as a driver of HIV pathogenesis . Moreover , they offer additional supporting data that efforts to ameliorate gut damage and dysbiosis , and the subsequent immune activation this dysregulation initiates in HIV-infected patients , will have positive impacts on disease outcome .
All participants in the Zambia Emory HIV Research Project ( ZEHRP ) discordant couples cohort in Lusaka , Zambia were adults at the time of participation and enrolled in human subjects protocols approved by both the University of Zambia Research Ethics Committee and the Emory University Institutional Review Board . Prior to enrollment , individuals received counseling and signed a written informed consent form agreeing to participate . The subjects included in this study were selected from the ZEHRP cohort based on being recently infected with HIV-1 . All subjects were initially seronegative partners within serodiscordant cohabitating heterosexual couples , who knew their HIV status and were provided counseling and condoms , but subsequently seroconverted . All subjects were antiretroviral therapy naïve and were identified a median of 44 ( IQR = 33–49 ) days after the estimated date of infection ( EDI ) . The algorithm used to determine the EDI has been previously described [73] . Additional demographic and cohort data can be found in S1 Table . All subjects were infected by HIV-1 subtype C viruses . Early set point viral load ( VL ) for newly infected individuals was defined as the earliest stable nadir VL value measured between 3 and 9 months post infection and which did not show a significant increase in value within a 3–4 month window . HIV plasma VL was determined at the Emory Center for AIDS Research Virology Core Laboratory using the Amplicor HIV-1 Monitor Test ( version 1 . 5; Roche ) . CD4+ T cell counts were based on T-cell immunophenotyping , with assays done using the FACScount System ( Beckman Coulter Ltd . , London , United Kingdom ) in collaboration with the International AIDS Vaccine Initiative . Genomic DNA was extracted from whole blood or buffy coats ( QIAamp blood kit; Qiagen ) . HLA genotyping relied on a combination of PCR-based techniques , involving sequence-specific primers ( Invitrogen ) and sequence-specific oligonucleotide probes ( Innogenetics ) , as described previously [74] . Ambiguities were resolved by direct sequencing of three exons in each gene , using kits ( Abbott Molecular , Inc . ) designed for capillary electrophoresis and the ABI 3130xl DNA Analyzer ( Applied Biosystems ) . SNP genotyping with the Illumina ImmunoChip was processed at a genomics core facility ( University of Alabama at Birmingham ) ; SNP alleles were inferred using the joint calling and haplotype phasing algorithm implemented in BEAGLECALL[75] . We completed a series of data cleaning and quality control procedures for SNPs in the xMHC region , excluding SNPs based on the following criteria: ( i ) duplication , ( ii ) missingness ( call rate<98 . 5% ) , ( iii ) minor allele frequency <0 . 025 in SCs and <0 . 015 in SPs and ( iv ) deviation from Hardy–Weinberg equilibrium ( P<10−6 ) . Data processing and quality control procedures have been described previously for this data set[46] . Gag-MJ4 chimeras were generated from frozen plasma isolated at the seroconversion time point for 127 subjects as previously described[1] . Briefly , viral RNA was extracted from 140 ul of plasma using the Qiagen Viral RNA extraction kit ( Qiagen ) . Combined RT-PCR and first round PCR were performed in a single reaction , and gag genes were amplified using a nested second round PCR . Patient-derived gag genes were joined with the MJ4 long terminal repeat portion via splice-overlap-extension PCR , and gag-LTR amplicons were cloned into the MJ4 proviral vector using NgoMIV and BclI endonuclease restriction enzymes . The in vitro replication assay used to generate viral replication capacity ( vRC ) values has been described extensively [76] . Briefly , replication competent virus was made via transfection of 293T cells ( American Type Culture Collection ) . Virus supernatants were harvested , titered on the TZM-bl indicator cell line , and used to infect the GXR25 T cell line ( gift from Dr . Mark Brockman , Simon Fraser University , Burnaby , BC , Canada ) at a constant multiplicity of infection . Supernatants from infected cultures were harvested every 2 days , and cultures were split to maintain healthy cell confluency . The extent of viral particles released into the supernatant was quantified via a radiolabeled reverse transcriptase assay . vRC values were generated by dividing the slope of replication of each virus by the slope of replication of wild-type MJ4 . Cox proportional hazards models with a backwards variable selection strategy were used to identify protective HLA alleles . All HLA alleles represented by at least 3 or more individuals in the cohort were included in the initial model . Sex of the seroconvertor as well as viral replication capacity of the transmitted founder virus ( based on the gag gene alone ) were added as static covariates present in every iteration of the model . HLA alleles contributing least significantly to the model ( highest p-value in the model ) were iteratively excluded in a stepwise fashion until all variables in the model had p-values < 0 . 05 . Linkage disequilibrium between specific HLA alleles was accounted for during the variable selection process . HLA scores presented in Fig 1 were generated by adding the number of HLA class I ( B*1401 , B*57 , B*5801 , B*81 ) and HLA class II ( DQB1*02 , DRB1*15 ) alleles identified to be CD4-protectve in this cohort for each individual . In S1A Fig and S1B Fig , only CD4-protective HLA class I alleles or CD4-protective HLA class II alleles were summed , respectively . Scores of 0 , 1 , 2 , or 3 refer to any individual having that number of CD4-protective HLA class I and/or class II alleles . Measurement of LPS levels was performed using the LAL Chromogenic Endotoxin Quantification kit ( American Diagnostica ) . Levels of sCD14 ( R&D systems ) were measured using a standard ELISA based assay according to the manufacturer’s instructions . Plasma levels of IL-10 were measured using MILLIPLEX Human Cytokine/Chemokine detection kit ( EMD Millipore ) . Samples were run in duplicate with all individuals on the same plate and wells with low bead count or coefficient of variance >30% were excluded from subsequent analysis . Plates were read on the Bio-Plex 3D Suspension Array System ( Bio-Rad ) . Intestinal fatty acid binding protein ( I-FABP ) was measured using a commercially available ELISA DuoSet assay ( R&D Systems , Minneapolis , MN , USA ) according to the manufacturer's instructions with minor adjustments . Plasma samples were diluted to 10% in diluent from the R&D Systems soluble CD14 ELISA kit ( DC140 ) and plates were blocked with Sigma Blocking Buffer . In order to maintain normal distributions and to lessen the effect of outliers , extreme positive values for each analyte measured in plasma were Winsorized to the 90th percentile . Cryopreserved PBMCs were stained for flow cytometric analysis in a manner identical to that described in [1] . The following antibodies were used to distinguish T cell populations and activation phenotypes: CD3-APC/Cy7 clone SP34-2 , HLA-DR-PerCP/Cy5 . 5 clone G46-6 , CCR5-PE clone 3A9 , Ki67- AlexaFluor700 clone B56 , CCR7-PE/Cy7 clone 3D12 ( BD Biosciences ) ; CD4- APC clone OKT4 , CD8-BV605 clone RPA-T8 , PD-1-BV421 clone EH12 . 2H7 ( BioLegend ) ; CD45RO-PE/Texas Red clone UCHL1 , CD27-PE/Cy5 clone 1A4CD27 ( Beckman Coulter ) ; CD38-FITC clone AT-1 ( Stemcell ) . All flow cytometry data was collected on an LSRII ( BD Biosciences ) cytometer using FACSDiVa version 6 . 3 . 1 software . Analysis of the data performed using FlowJo version 9 . 7 . 5 software ( TreeStar ) . Data on alcohol consumption was collected for individuals in the ZEHRP cohort as described previously [43 , 44] . In order to categorize individuals into those that frequently drank alcohol to excess and those that did not , we made use of data collected from questions regarding how often in the last year individuals got drunk . Initial scores were denoted as follows: 1 = never , 2 = less than monthly , 3 = monthly , 4 = weekly , 5 = daily or almost daily . For the purpose of this study , individuals with scores 2 or less were denoted as “drunk less than monthly” and individuals with scores 3 or greater were denoted as “drunk monthly or more” . Cryopreserved PBMCs collected at the first study visit post infection ( median 45 post estimated date of infection ) from 33 individuals were thawed , rested overnight , and interrogated with two PTE Gag peptide pools ( HIV-1 PTE Gag Peptide Pool from NIAID , DAIDS ) and a pool of 31 15-mer peptides spanning 3 highly conserved regions in Gag [51] via an IFNγ ELISpot assay as previously described [77 , 78] . Spot forming units per million PBMCs for the two PTE pools were summed . Correction factors for sequence conservation of the transmitted Gag amino acid sequence was calculated for each of the 33 individuals , specific for either the PTE or conserved Gag peptide pools . For the PTE pool correction factor , all 320 PTE peptides were aligned to full length Gag amino acid sequences derived from viral RNA isolated from the seroconversion time point ( median of 45 days post estimated date of infection ) for all 33 individuals . Scores were subsequently calculated by summing the number of PTE peptides that were at least 90% conserved with respect to the individual’s Gag sequence . For the conserved epitope pool correction factor , Gag amino acid sequences from infected individuals were aligned to the 3 distinct regions covered by this pool of 31 15-mer peptides [51] . Percent similarity to these 3 regions was calculated by extracting the pairwise identity based on the sequence alignment performed using Geneious bioinformatics software ( Biomatters , Aukland , NZ ) . The nucleotide sequences corresponding to the amino acid sequences used for these calculations have previously been deposited under GenBank accession nos . KP715723–KP715849 . Peptide specific IFNγ ELISpot responses to measure breadth were performed on a separate subset of cryopreserved PBMCs collected a median of 339 days post infection for a total of 18 individuals in the cohort [79 , 80] . PBMCs were thawed and expanded ex vivo using a bi-specific CD3/CD4 antibody [81] . Expanded cells were then used to interrogate responses to a matrix of peptide pools with subsequent deconvolution to single peptides [82] . All statistical analysis was performed using JMP Pro , version 12 ( SAS Institute Inc . , Cary , NC ) . All bivariate continuous correlations were performed using standard linear regression . One-way comparison of means was performed using the Student’s t-test , and one-tailed p-values are reported . Standard ANOVA was used for the comparison of means between more than two variables . MANOVA was used to assess changes between means of two groups over time . Generalized Linear Models were used to test the predictive nature of two or more continuous or categorical variables against a single continuous outcome . Kaplan-Meier survival curves and Cox proportional hazards models were performed using an endpoint defined as a single CD4+ T cell count reading less than 300 cells/μl , unless otherwise specified , and statistics reported for survival analyses are generated from the log-rank test . | During acute HIV infection , there exists a complex interplay between the host immune response and the virus , and the balance of these interactions dramatically affects disease trajectory in infected individuals . Variations in Human Leukocyte Antigen ( HLA ) alleles dictate the potency of the cellular immune response to HIV , and certain well-studied alleles ( HLA-B*57 , B*27 ) are associated with control of HIV viremia . However , though plasma viral load is indicative of disease progression , the number of CD4+ T cells in the blood is a better measurement of disease severity . Through analysis of a large Zambian acute infection cohort , we identified HLA alleles that were associated with protection for CD4+ T cell loss , without dramatic affect on early plasma viremia . We further link these favorable HLA alleles to reduction in a well-known contributor to HIV pathogenesis , the presence of microbial products in the blood , which is indicative of damage to the gastrointestinal tract , a process which accelerates disease progression in HIV infected individuals . Ultimately , these results suggest a new mechanism by which the cellular immune response can combat HIV-associated pathogenesis , and further highlight the contribution of gut damage and microbial translocation to accelerating disease progression , even at early stages in HIV infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"blood",
"cells",
"hiv",
"infections",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"immune",
"activation",
"pathogens",
"immunology",
"microbiology",
"retroviruses... | 2019 | Protective HLA alleles are associated with reduced LPS levels in acute HIV infection with implications for immune activation and pathogenesis |
Chromosome termini form a specialized type of heterochromatin that is important for chromosome stability . The recent discovery of telomeric RNA transcripts in yeast and vertebrates raised the question of whether RNA–based mechanisms are involved in the formation of telomeric heterochromatin . In this study , we performed detailed analysis of chromatin structure and RNA transcription at chromosome termini in Arabidopsis . Arabidopsis telomeres display features of intermediate heterochromatin that does not extensively spread to subtelomeric regions which encode transcriptionally active genes . We also found telomeric repeat–containing transcripts arising from telomeres and centromeric loci , a portion of which are processed into small interfering RNAs . These telomeric siRNAs contribute to the maintenance of telomeric chromatin through promoting methylation of asymmetric cytosines in telomeric ( CCCTAAA ) n repeats . The formation of telomeric siRNAs and methylation of telomeres relies on the RNA–dependent DNA methylation pathway . The loss of telomeric DNA methylation in rdr2 mutants is accompanied by only a modest effect on histone heterochromatic marks , indicating that maintenance of telomeric heterochromatin in Arabidopsis is reinforced by several independent mechanisms . In conclusion , this study provides evidence for an siRNA–directed mechanism of chromatin maintenance at telomeres in Arabidopsis .
Telomeres safeguard the stability of eukaryotic chromosomes by protecting natural chromosome ends from triggering DNA damage responses . Chromosome termini consist of telomeric and subtelomeric repeats that are bound by a specific set of telomere binding proteins as well as nucleosomes that exhibit features of pericentric heterochromatin [1] . These regions are usually devoid of functional genes , and transgenes integrated in the vicinity of telomeres are subjected to transcriptional silencing , a phenomenon known as telomere position effect [2] . Studies in mammals indicate that telomeric heterochromatin plays an important function in chromosome end protection and telomere length regulation . Inactivation of the SIRT6 histone deacetylase in human cells causes hyperacetylation of telomeric histone H3 , telomere dysfunction and premature cell senescence [3] . Deficiency in histone methyltransferases or the retinoblastoma tumor suppressor leads to disruption of telomeric heterochromatin and aberrant telomere elongation in mouse cells [4]–[6] . Another important hallmark of heterochromatin in mammals is DNA methylation . Although vertebrate telomeric DNA does not appear to be methylated due to the lack of canonical CG sites , subtelomeric repeats are heavily methylated [7] . Interestingly , inactivation of DNA methyltransferases in mouse cells decreases 5-methylcytosine at subtelomeres and leads to increased telomeric recombination , without a concomitant change in histone modifications [7] . These data indicate a functional interaction between subtelomeric and telomeric chromatin . Heterochromatin was thought to be transcriptionally inactive , but this view has been challenged by discoveries of numerous non-coding ( nc ) transcripts derived from heterochromatic loci . Some of these transcripts directly contribute to the assembly of heterochromatin at defined chromosomal domains and their biogenesis is vital for processes such as X chromosome inactivation , genomic imprinting , transposon silencing and centromere function [8] . Thus , it is not surprising that although telomeres possess marks of repressive heterochromatin , they are not transcriptionally silent . Recent studies revealed the presence of telomeric repeat-containing RNAs ( TERRA ) that are transcribed from subtelomeric regions in yeast and vertebrates [9]–[11] . TERRA are removed from telomeres either through Rat1p-dependent degradation in budding yeast or through non-sense mediated RNA decay ( NMD ) in human; deficiencies in these RNA processing pathways have dramatic effects on telomere maintenance [9] , [10] . Hypomethylation of subtelomeric regions in mammalian cells lacking DNA methyltransferases leads to the overproduction of TERRA [11] , [12] . This suggests that the epigenetic status of subtelomeres and telomeres influences TERRA expression . The discovery of TERRA raised the question of whether ncRNAs contribute to the establishment of telomeric heterochromatin . This hypothesis gained support in a recent study in which downregulation of TERRA by exogenous short interfering RNAs ( siRNAs ) in human cell lines led to depletion of histone heterochromatic modification from telomeres [13] . In many organisms , RNA-mediated chromatin silencing relies on small RNA molecules that guide effector complexes to target sites [8] , [14] . However , involvement of small RNAs in chromatin formation at canonical telomeres has not been shown yet . In this study , we investigate chromatin organization and transcription at chromosome ends in the model plant Arabidopsis thaliana . We detect the presence of transcripts containing telomeric repeats and show that some of these transcripts are processed into ∼24 nt siRNAs . These transcripts are produced from telomeres as well as from intrachromosomal telomeric loci that are mainly located at centromeres . The 24 nt siRNAs are generated through the RNA-dependent DNA methylation ( RdDM ) pathway , which is a plant-specific mechanism that utilizes siRNAs to guide DNA methyltransferases to asymmetric cytosines ( CNN ) [15] , [16] . We demonstrate that RdDM is responsible for methylation of telomeric DNA that contains cytosines exclusively in asymmetric sequence contexts and hence for reinforcement of heterochromatic marks at telomeres .
Gene organization at chromosome ends in Arabidopsis appears to be unique . In contrast to the majority of organisms with known telomere/subtelomere sequences , 8 of the 10 Arabidopsis subtelomeres have no repetitive DNA , and predicted genes are annotated in the immediate vicinity of telomeres [17] ( Figure 1A ) . We experimentally confirmed that sequences annotated as chromosome ends are indeed associated with telomeres for 7 chromosome arms with the exception of the right arm of chromosome 3 [18] . The two remaining chromosome termini contain clusters of ribosomal RNA genes ( NORs ) [19] . We performed reverse transcription ( RT ) PCR analysis to verify that all the predicted terminal genes are expressed and that they do not represent pseudogenes ( Figure 1B ) . The genes showed distinct tissue-specific expression patterns and the size of the RT-PCR products corresponded to the predicted size of the spliced mRNAs . There was no obvious correlation between the level of expression and promoter distance from telomeres , and even the At2g48160 gene , with a promoter immediately adjacent to telomeric DNA , was robustly expressed . These data indicate that , in contrast to yeast and mammals , Arabidopsis telomeres do not silence genes located in their vicinity . The high transcriptional activity near telomeres raised questions about the chromatin structure of chromosome termini in Arabidopsis . We investigated the distribution of histone modification marks typical for plant euchromatin ( tri-methylation of histone H3 at Lys4 , H3K4me3 ) and heterochromatin ( di-methylation of H3 at Lys9 , H3K9me2; and mono-methylation of H3 at Lys27 , H3K27me1 ) at telomere-associated regions by chromatin immunoprecipitation ( ChIP ) . The ∼600 bp region immediately adjacent to the telomere on the right arm of chromosome 2 ( 2R ) represents the promoter of the At2g48160 gene ( Figure 2A ) and carries typical euchromatic histone marks ( Figure 2B ) . The H3K4me3 euchromatin mark was also dominant at the promoter of the At1g01010 gene that is located ∼3 . 5 kb from the telomere on the left arm of chromosome 1 ( region 1L-3 , Figure 2A and 2B ) , although we could detect a weak H3K27me1 signal that is usually typical of heterochromatin . Histone heterochromatic marks ( H3K9me2 and H3K27me1 ) became more pronounced at the 1L-2 and 1L-1 regions that are located on the same arm ∼1 . 5 kb and 1 kb from the telomere , respectively ( Figure 2A and 2B ) . The 1L telomere contains a recent 104 bp insertion of mitochondrial DNA embedded within the centromere-proximal region of telomeric repeats [20] ( Figure 2A ) . Using this insertion to design primers that span the centromere-proximal part of the 1L telomere ( 1L-0 , Figure 2A ) , we were able to demonstrate that this region also displays heterochromatin marks ( Figure 2B ) . Nevertheless , the 1L-0 region still possessed clearly detectable H3K4me3 , which is atypical of classical heterochromatin where the H3K4me3 modification is strongly reduced in comparison to H3K27me1 and H3K9me2 . A similar histone-modification pattern was also observed in telomere-adjacent regions of five other chromosome arms ( Figure 2B ) . To further examine chromatin at telomeres , we analyzed ChIP fractions by dot-blot hybridization with a telomeric probe ( Figure 2C ) . The Arabidopsis genome is enriched for intrachromosomal degenerated telomeric repeats that are mainly localized at centromeres ( Figure S1 ) . To specifically assay for chromatin at telomeres , we used stringent hybridization conditions at which the centromere-derived signal is eliminated to less than 2% of the total telomeric signal ( Figure S1 ) . We readily detected H3K27me1 and H3K9me2 modifications , and a weaker but still clearly detectable H3K4me3 signal . This hybridization pattern was reminiscent of the results obtained by ChIP analysis of telomere-adjacent regions by PCR ( Figure 2B ) . Thus , our ChIP data show that Arabidopsis telomeres form chromatin that is enriched for H3K9me2 and H3K27me1 heterochromatic marks , but still retains the euchromatic H3K4me3 modification . We found that the heterochromatin marks extend ∼1 . 5 kb into the subtelomeric region of 1L . A survey of a high-resolution genome-wide map of H3K9me2 distribution indicates that H3K9me2 also spreads up to 1 . 5 kb from telomeres at chromosome arms 1R , 3L , 4R and 5L [21] ( http://epigenomics . mcdb . ucla . edu/H3K9m2/ ) . However , detecting the prominent H3K4me3 signal side by side with the heterochromatic marks ( Figure 2B and 2C ) strongly indicates that Arabidopsis telomeres exhibit features of intermediate heterochromatin that is characterized by retention of opposing histone H3 methylation marks [22] . We next asked whether Arabidopsis telomeres are transcribed by assaying for the presence of TERRA by Northern hybridization with a CCCTAAA probe . We readily detected two types of TERRA: heterogeneous transcripts which ranged from high molecular weight strands that migrated at the limits of gel resolution to hundreds of nucleotides , and several distinct bands ( Figure 3A ) . We also detected antisense telomeric transcripts ( ARRET ) that gave a similar hybridization pattern as the TERRA by the complementary TTTAGGG probe ( Figure 3A ) . These signals disappeared after pretreatment of the samples with RNaseA ( Figure 3B and data not shown ) demonstrating that they do not represent remnants of DNA in RNA preparations . Expression of TERRA varied between RNA samples extracted from different tissues of Arabidopsis ( Figure 3C ) . Interestingly , remarkable variation in expression was also detected between different Arabidopsis accessions , as the levels of TERRA in seedlings of Zur and Ws ecotypes were almost two orders of magnitude higher than in Col and Ler ( Figure 3C ) . Arabidopsis TERRA and ARRET can originate at telomeres or arise from transcription of degenerated intrachromosomal telomeric sequences localized at centromeric regions ( Figure S1 ) . The bulk of centromeric DNA consists of 177–179 bp satellite repeats ( CEN180 ) , a subset of which is transcribed [23] . Sequential hybridization of a Northern blot with probes detecting TERRA and CEN180 resulted in an almost identical hybridization pattern , characterized by five distinct bands ( Figure 3A ) . Hybridization of the blots with probes detecting sequences immediately adjacent to telomeres did not produce any detectable signal ( data not shown ) . These results suggest that TERRA and ARRET transcripts detected by Northern analysis mainly arise from centromeric regions that contain remnants of telomeric DNA and not from the transcription of telomeres . To examine whether telomeres are transcribed at levels non-detectable by Northern hybridization , we analyzed expression of subtelomeric regions adjacent to telomeric DNA by strand-specific RT-PCR in flowers . We could distinguish expression of TERRA and ARRET by using either telomeric or subtelomeric arm-specific primers for reverse transcription ( Figure 3D ) . We detected expression of both TERRA and ARRET at four out of eight analyzed chromosome arms . We failed to detect any transcription at chromosome arms 1R and 5R . Interestingly , only the TERRA but not ARRET transcripts were detected at 1L . The RT-PCR data demonstrate that at least five Arabidopsis telomeres are indeed transcribed , albeit at a low level . To gain further insights into telomere transcription , we cloned a ∼500 nt promoter of the At2g48160 gene , which is located next to the telomere ( Figure 1 ) , in front of a reporter β-glucuronidase ( GUS ) gene in both sense and antisense orientations . We could detect GUS transcripts in transgenic plants carrying both constructs , although the expression in the antisense direction was much weaker than in the sense orientation ( Figure S2 ) . This experiment further supports the idea that telomere adjacent regions can drive transcription into a telomere . The presence of centromeric and telomeric TERRA and ARRET indicated that telomeric transcripts are able to form partially double stranded ( ds ) intermediates that could be processed by a Dicer into siRNA . In support of this hypothesis , siRNAs corresponding to both strands of telomeric DNA were detected in wild-type plants ( Figure 4A ) . We estimate the size of the telomeric C-rich strand siRNAs ( C-siRNA ) to be 24–25 nt , and the size of G-siRNAs to be 23–24 nt ( Figure S3 ) . The formation of 24 nt siRNAs in Arabidopsis is mediated by RNA-processing enzymes of the RdDM pathway [24] . This pathway is specific to plants and mediates methylation of cytosine residues in an asymmetric sequence context ( CNN ) . The absence of telomeric 23–25 siRNAs in plants lacking RNA-dependent RNA polymerase 2 ( RDR2 ) , Dicer-like 3 ( DCL3 ) or subunits of RNA Polymerase IV ( NRPD1 or NRPD2 ) and their reduction in two other RdDM mutants ( drd1 and nrpe1 ) further demonstrated that telomeric siRNAs belong to the category of 24 nt heterochromatic siRNAs ( Figure 4A ) . These siRNAs are usually derived from heterochromatic loci and form the most abundant fraction of plant small RNAs [25] , [26] . They typically associate with Argonaute 4 ( AGO4 ) that is part of the effector complex that , together with Polymerase V , mediates CNN methylation [27] , [28] . To determine whether telomeric siRNAs associate with AGO4 , we surveyed published datasets containing ∼600 , 000 Argonaute ( AGO1 , AGO2 , AGO4 and AGO5 ) -bound small RNAs [29] . We identified a total of 133 small RNAs containing at least 12 nucleotides with a perfect telomeric repeat ( Table S1 ) . As expected , the majority of these small RNAs were associated with AGO4 ( Figure 4B ) . Surprisingly , the AGO4-associated telomeric siRNAs were almost exclusively G-siRNAs and only a few C-siRNAs containing no more than 14 nt of the CCCTAAA repeat sequence were found in the dataset ( Figure 4C ) . Since the levels of total G- and C-siRNAs are similar ( Figure 4A ) , this bias may be caused by a selective incorporation of the G-siRNAs into the AGO4 complex . As TERRA transcripts are produced from telomeres as well as from centromere-located telomeric DNA , the siRNAs may be of either telomeric or centromeric origin . To determine whether telomere-derived transcripts are processed into siRNAs , we aligned Argonaute-associated siRNAs with telomere-adjacent sequences . We found abundant siRNAs corresponding to both strands of subtelomeric DNA at chromosome arms 1L , 1R , 3L , 4R and 5L ( Figure 5 , Table S2 ) . Since these regions are formed by unique sequences , the origin of the siRNAs can be unambiguously traced to these loci . Interestingly , AGO4-associated siRNAs were particularly enriched at the chromosome ends that also exhibited expression of TERRA and ARRET ( 1L , 3L , 4R , 5L; Figure 5 ) . These data strongly argue that telomeric TERRA and/or ARRET are processed into siRNAs . Plants can methylate cytosines in all sequence contexts , and DNA methylation at asymmetric positions relies largely on 24 nt siRNAs and on the RdDM pathway . The presence of telomeric siRNAs prompted us to ask whether telomeric DNA , which contains cytosines exclusively in the CNN context , can be methylated . We took advantage of the unique insertion in the 1L telomere that allowed us to design primers spanning 13 CCCTAAA repeats located in the centromere-proximal part of the 1L telomere ( region 1L-0'; Figure 2A ) . Bisulfite sequencing of the 1L-0' region in wild-type plants revealed that over 40% of cytosines in these telomeric repeats are methylated ( Figure 6 ) . In contrast , the 1L and 2R subtelomeric regions are devoid of DNA methylation ( Figure S4 ) . The telomeric methylation in 1L-0' is non-randomly distributed , with preferential enrichment at the third cytosine in the CCCTAAA sequence ( Figure 6A and 6B ) . A similar observation was recently made through whole genome bisulfite sequencing that also revealed methylation of telomeric repeats , albeit at a lower total frequency than reported here [30] . The level of 5-methylcytosine in all sequence contexts was dramatically reduced in rdr2 mutants , arguing that methylation of the 1L-0' region primarily depends on the RdDM mechanism ( Figure 6A and 6C ) . We next examined whether cytosine methylation and its dependence on the RdDM pathway is a general feature of telomeric DNA . We sequentially hybridized bisulfite-treated total genomic DNA to oligonucleotide probes that first detected fully converted telomeric DNA ( probe AAAATTT ) , then unconverted , and thus completely methylated DNA ( probe TTTAGGG ) , and finally the complementary cytosine-free strand ( probe CCCTAAA ) as a control for loading ( Figure 6D ) . A strong hybridization AAATTTT signal suggested that the bulk of telomeric DNA is only weakly methylated . Nevertheless , a portion of wild-type DNA was resistant to bisulfite conversion as hybridization with the TTTAGGG oligo probe showed a signal that was ∼4-fold higher than a background signal from a corresponding amount of non-methylated bisulfite-converted telomeric DNA cloned in a plasmid ( Figure 6D and 6E ) . These data further indicate the presence of some heavily methylated CCCTAAA sequences in wild-type plants . Importantly , this CCCTAAA signal was reduced to a background level in rdr2 and nrpd2a mutants ( Figure 6D and 6E ) . To further investigate whether methylation occurs at telomeres , we performed high-stringency hybridization of the bisulfite-converted samples with a long telomeric TTTAGGG probe ( Figure 6C ) . Under these conditions , converted plasmid-cloned telomeric DNA produces a high background hybridization signal that is likely caused by sufficiently stable interactions between longer fragments of the ( TTTTAAA ) n converted telomeric DNA and the ( TTTAGGG ) n probe . Nevertheless , wild-type DNA samples still produced a signal that was significantly higher than the background hybridization ( Figure 6F ) . These data , together with the bisulfite sequencing of the 1L-0' telomeric region , strongly argue that DNA methylation is a general characteristic of Arabidopsis telomeres and that its maintenance requires the RdDM pathway . Loss of DNA methylation is often accompanied by chromatin remodeling . However , the decrease in telomeric DNA methylation did not result in a significant loss of heterochromatic histone marks , and both H3K9me2 and H3K27me1 remained enriched at the bulk of telomeric DNA in rdr2 mutants ( Figure 7A ) . However , analysis of histone modifications at the 1L-0' locus by ChIP and quantitative PCR ( Figure 7B and 7C ) showed a decrease in H3K9me2 and H3K27me1 ( Figure 7C ) in rdr2 mutants . These data indicate that although the RdDM-dependent mechanism is not solely responsible for heterochromatin formation at telomeres , it contributes to its maintenance by mediating methylation of telomeric DNA , thereby reinforcing heterochromatic histone modifications . Disruption of telomeric heterochromatin or demethylation of subtelomeric sequences leads to increased telomere elongation and recombination in mouse [7] . Our analysis of telomere length and intrachromatid recombination at chromosome ends did not reveal any differences between RdDM mutants and wild-type plants ( Figure S5 and Figure S6 ) . This observation further corroborates our finding that despite reduced DNA methylation , the bulk of telomeric chromatin in rdr2 mutants still retains heterochromatic features .
Heterochromatin is a universal characteristic of chromosome termini in a variety of organisms , including yeast , flies and mammals . Subtelomeric regions in these organisms are gene-poor and enriched for middle to highly repetitive sequences that contribute to the formation of a heritably repressed chromatin structure at chromosome termini that shares similarities with pericentromeric heterochromatin [1] , [31] , [32] . Nevertheless , some aspects of chromatin organization appear to be unique at telomeres as telomeric chromatin in humans and plants display unusually short nucleosomal spacing ( ∼160 nt ) in comparison with the ∼180 nt periodicity at the bulk of chromatin [33]–[35] . In contrast to many other organisms , telomeres in Arabidopsis are directly adjacent to transcriptionally active genes . This situation is more similar to silenced transposons inserted in gene-rich regions than to pericentromeric heterochromatin . This is also reflected in the organization of telomeric chromatin that exhibits features of intermediate heterochromatin that is characterized by the presence of both active and repressive histone H3 marks . Such chromatin was described to be associated with some Arabidopsis transposons and transgenic loci [22] , [36] . Chromatin analysis of the 1L subtelomere demonstrates that repressive histone H3 modifications are most pronounced immediately next to telomeres and that their presence gradually recedes with growing distance from telomeres . Data on whole-genome distribution of H3K9me2 indicate that this also holds true for telomere-associated regions of several other chromosome arms [21] . These data infer that repressive histone marks are primarily established at telomeres and spread only a limited distance within adjacent subtelomeric sequences . The existence of such relatively small clusters of repressive chromatin ( 2–5 kb ) next to otherwise large gene-rich regions suggests a functional importance for the heterochromatization of telomeres in Arabidopsis . It further suggests the existence of mechanisms that specifically maintain repressive histone modifications at telomeres . Assembly of heterochromatin at chromosome ends in budding yeast is partially dependent on tethering Sir proteins to telomeres via the Rap1 telomere-binding protein [31] . Human SIRT6 histone deacetylase preferentially associates with telomeres , although how it is recruited to chromosome termini is not known [3] . A recent study in mice overexpressing TRF2 indicates that , similar to the situation in yeast , heterochromatin formation at telomeres in mammals may also involve telomere-binding proteins [37] . The discovery of TERRA provides another attractive model that involves targeting the chromatin remodeling machinery to chromosome termini through ncRNA [38] , [39] . This suggestion was recently corroborated by the finding that downregulation of TERRA by RNAi in human cells causes a decrease in histone H3K9 methylation [13] . It was proposed that TERRA facilitates heterochromatin formation by stabilizing interactions between heterochromatin factors and telomeric DNA . In this study , we demonstrate expression of telomeric transcripts in Arabidopsis and describe a mechanism by which telomeric repeats-containing RNAs affect telomeric chromatin through siRNA . In contrast to the situation in mammals , where only UUAGGG telomeric transcripts were detected [10] , [11] , both telomeric strands appear to be transcribed from some telomeres in Arabidopsis . This indicates that canonical telomeric DNA may , under certain circumstances , act as a promoter and initiate transcription . Two lines of observations further corroborate the link between transcription and telomeric DNA in Arabidopsis . Firstly , short stretches of a telomeric sequence were found in numerous Arabidopsis promoters and it has been shown that these interstitial telomere motifs are required for transcription [40] . Secondly , several transcription factors have been identified in Arabidopsis that specifically bind to telomeric DNA in electromobility shift assays ( reviewed in [41] ) . Thus , it is possible that some of these transcription factors localize to telomeres and promote their expression . In addition to transcripts that originated at telomeres , we detected TERRA and ARRET that are apparently generated by transcription of centromere-associated telomeric loci . We cannot currently determine the exact identity of telomere- or centromere-derived TERRA/ARRET that is processed by DCL3 and degraded to telomeric siRNAs . The requirement of RDR2 for siRNA formation indicates that the predicted dsRNA intermediate is not a simple annealing product of complementary TERRA and ARRET , but is dependent on additional RNA-dependent RNA synthesis . Thus , even relatively low level transcripts can yield significant amounts of siRNA . In fact , direct detection of precursor transcripts in the RdDM pathway has been so far reported only in a special transgene system [42] . In plants , heterochromatic siRNAs serve to guide DNA methylases to specific asymmetric CNN positions in a mechanism that relies on AGO4 [28] . Interestingly , AGO4 appears to retain telomeric G-siRNAs , and not the complementary C-siRNAs , although these data should still be verified by Northern analysis of AGO4 co-immunoprecipitated siRNAs . It is unknown whether the bias towards G-siRNAs is of biological significance , but it is interesting that the AGO4 complex appears to specifically retain siRNAs complementary to the telomeric strand to be methylated . Our data , showing methylation of bulk telomeric DNA as well as heavy methylation of the centromere-proximal region of the 1L telomere , together with data from whole genome bisulfite sequencing [30] , argue that telomeric heterochromatin in Arabidopsis is not only defined by histone modifications , but also by DNA methylation . Although mammalian telomeres lack CG sites , and are , thus , believed to be unmethylated , at least two proteins linked to DNA methylation ( SMCHD1 , MBD3 ) have been found in purified fractions of human telomeric chromatin [43] . Additionally , the recent discovery of CNN and CNG methylation in human embryonic stem cells warrants the re-examination of DNA methylation at human telomeres [44] . We demonstrate that the maintenance of telomeric DNA methylation depends , to a large extent , on heterochromatic siRNA and the RdDM machinery . Intriguingly , loss of telomeric DNA methylation only has a slight effect on histone methylation at bulk telomeres , indicating that assembly of Arabidopsis telomeric heterochromatin relies on several reinforcing mechanisms that recruit histone methyltransferases such as SUVH4 to telomeres [45] . Loss of DNA methylation has a more profound effect on histone methylation at the centromere-proximal part of the 1L telomere . This indicates that RdDM may play a role in maintaining heterochromatin at the boundary between telomeres and adjacent euchromatic genes . The involvement of siRNA in modulation of telomeric heterochromatin may not be restricted to plants . Our data in Arabidopsis are reminiscent of the situation in fission yeast where heterochromatin in subtelomeric regions is established by two independent pathways , one of which relies on the telomere-binding protein Taz1 , while the other involves RNA-induced transcriptional silencing ( RITS ) [46] . However , in contrast to the situation in Arabidopsis where siRNA targets canonical telomeric repeats , RITS in fission yeast is directed at centromere-like sequences that are located ∼15 kb from telomeres . In humans , TERRA has been proposed to act as a scaffold , reinforcing interactions between telomere-binding proteins and heterochromatin factors such as ORC1 and HP 1 [13] . Nevertheless , human TERRA could also promote heterochromatin formation through an siRNA-mediated pathway . This notion is supported by the observation that enrichment of Argonaute-1 at human telomeres is correlated with increased H3K9 methylation and HP1 association [47] , and by the discovery of telomere-derived human siRNAs [48] .
Arabidopsis mutants carrying the following alleles were used in this study: dcl3-1 ( dcl3 ) , rdr2-1 ( rdr2 ) , nrpd1a-4 ( nrpd1 ) , nrpd1b-1 ( nrpe1 ) , sgs2-1 ( rdr6 ) , drd1-1 ( drd1 ) and nrpd2a-1 ( nrpd2 ) . Plants were grown in soil under long-day conditions ( 16 h light/8 h dark ) at 22°C . Total RNA was extracted using TriReagent solution ( Sigma ) . For Northern blot analysis , 10 µg aliquots were separated on 1 . 2% formaldehyde agarose gels , blotted onto a nylon membrane and hybridized with [32P] 5′ end-labeled ( TTTAGGG ) 4 ( TTTAGGG probe ) or ( TAAACCC ) 4 ( CCCTAAA probe ) oligonucleotides . Oligo hybridizations were carried out at 55°C as previously described [49] . Centromeric transcripts were detected by hybridization with a [32P]-labeled CEN180 repeat unit amplified from Arabidopsis genomic DNA using primers CEN1 and CEN2 ( Table S3 ) . For RT-PCR analyses , ∼2 µg of total RNA was reverse transcribed by using oligo dT for gene expression . The ( CCCTAAA ) 3 oligo or subtelomere-specific primers ( Table S3 ) were used for RT of TERRA and ARRET , respectively . The respective cDNAs were amplified by 25–35 cycles of PCR with specific primers ( Table S3 ) . Small RNAs were isolated from inflorescences using the mirVana miRNA isolation kit ( Ambion ) , separated on 15% polyacrylamide gels and electroblotted onto a nylon membrane . Telomeric siRNAs were detected by hybridization with either ( TTTAGGG ) 4 or ( TAAACCC ) 4 oligo probes in ULTRAhyb-Oligo hybridization buffer ( Ambion ) at 42°C . The artificial 25 and 23 nt siRNAs were synthesized by in vitro transcription using T7 RNA polymerase ( MBI ) . The T7-TOP oligonucleotide ( 10 µM ) was annealed to a template oligonucleotide ( 10 µM ) as indicated in Figure S3 . In vitro transcription was carried out with 30U of T7 RNA polymerase ( MBI ) and the annealed oligos ( 0 . 5 µM ) in 50 µL of 1× Transcription buffer ( MBI ) supplemented with NTPs ( 10 mM ) and RiboLock RNase inhibitors ( MBI ) for 60 min at 37°C . 25 µL of the reaction was separated on a 15% polyacrylamide gel , electroblotted onto a nylon membrane and analyzed by Southern hybridization . Genomic DNA was extracted from 4 week old plants with the DNAeasy Plant Maxi Kit ( Qiagen ) . Bisulfite modification was performed using the EpiTect Bisulphite Kit ( Qiagen ) according to the manufacturer's instructions . The completeness of the conversion was tested by PCR amplification of a non-methylated genomic region [50] . Modified DNA was used as a template for PCR amplification with the primers indicated in Table S3 . The PCR products were cloned into the pCR2 . 1 TOPO cloning vector ( Invitrogen ) and sequenced using a BigDye terminator and an ABI310 sequencer ( Applied Biosystems ) . The sequence of the clones was analyzed with the software CyMATE [50] . The efficiency of cytosine conversion in the 1L-0' region in these samples was further controlled by either spiking genomic DNA with a bacterial plasmid containing a region that partially overlaps with 1L-0' or by sequence analysis of other genomic loci that are devoid of 5-methylcytosines . For methylation analysis at bulk telomeric DNA , bisulfite-modified genomic DNA was transferred onto a nylon membrane by vacuum-blotting . As a control , a bisulfite-modified plasmid containing 750 bp of plant non-methylated telomeric DNA was blotted onto the membrane in an amount that roughly corresponded to the total amount of telomeric DNA present in genomic samples ( 1 ng of the plasmid contained telomeric DNA equivalent to ∼260 ng of genomic DNA ) . The membrane was hybridized with the [32P] 5′ end-labeled ( TTTAAAA ) 4 oligo ( AAAATTT probe ) in a standard hybridization buffer [49] at 40°C . The membrane was washed twice for 10 min at 40°C in 2× SSC followed by a 40 min wash in 1× SSC at 40°C . The membrane was exposed to a Kodak Phosphor screen ( Biorad ) and scanned with Molecular Imager FX ( Biorad ) . The membrane was then stripped and sequentially rehybridized with the TTTAGGG and CCCTAAA oligo probes at 55°C as described [49] . The final rehybridization was performed at 65°C with a strand-specific ( TTTAGGG ) n probe that was obtained by labeling of a 750 bp fragment of telomeric DNA with α-[32P]-GTP . The signals were quantified using QuantityOne software ( Biorad ) . Chromatin isolation and immunoprecipitation were performed as described [51] using antibodies against histone H3 ( Abcam; cat . no . ab1791 ) , H3K9me2 ( Abcam; cat . no . ab1220 ) , H3K4me3 ( Abcam; cat . no . ab8580 ) and H3K27me1 ( provided by Thomas Jenuwein ) . The DNA was column-purified from immunoprecipitated chromatin and concentrated in 50 µl of elution buffer . For dot-blot analysis , 40 µl of the DNA was blotted onto a nylon membrane and analyzed by hybridization with a [32P]-labeled 750 bp ( TTTAGGG ) n probe . For PCR analysis , 1 µl of the eluted DNA was amplified by 30 cycles of PCR with the primers specified in Table S1 . Quantitative PCR analysis of the 1L-0 region was performed using the iQ5 Real Time PCR detection system ( Biorad ) and a 2× SensiMix Plus SyBR Kit ( PeqLab ) . The sequences of Argonaute-associated siRNAs were retrieved from the NCBI ( accession number GSE10036 ) . The individual AGO datasets were searched for the presence of siRNAs containing a string of at least 12 nucleotides of Arabidopsis telomeric repeats of any possible permutation . Telomeric siRNAs were copied into an Excel table and manually annotated . Subtelomeric siRNAs were identified by attempting to align all Argonaute-associated siRNAs to an ∼15 kb sequence from the ends of each Arabidopsis chromosome using the publicly available program SOAP [52] . The subtelomeric sequences were derived from the sequences of whole chromosomes available in TAIR , and from cloned fragments of telomere-associated sequences deposited in the Gene Bank ( AB033278 and AM177017 ) . Perfectly matching siRNA alignments were retained , and plotted using R . Mitotic chromosomes prepared from pistils of wild-type plants were subjected to fluorescence in situ hybridization ( FISH ) with a Cy3-conjugated ( CCCTAAA ) 2 PNA probe ( Metabion ) as previously described [53] . Chromosomes were examined using a Zeiss Axioscope fluorescence microscope equipped with a CCD camera . The PETRA assay was carried out with genomic DNA extracted from a fifth generation tert mutant plant [54] according to the published protocol [18] . Terminal restriction fragment analysis was performed as described [49] , [55] . Analysis of intrachromatid telomeric recombination was performed by the t-circle amplification assay [56] . DNA extracted from Arabidopsis ku70 [57] mutants was used as a positive control . | Telomeres are protein–DNA structures that protect the ends of eukaryotic chromosomes . A failure in this protective structure can lead to chromosomal instabilities and contribute to cancer and aging . The protective nature of telomeres relies on complex interactions between repetitive telomeric DNA and associated proteins . One major question is how telomeric proteins , including telomere-associated nucleosomes , are modified in order to achieve this protection . In this study , we have discovered that Arabidopsis telomeric nucleosomes contain a unique mixture of both active and inactive chromatin marks . Additionally , the telomeric DNA itself is modified by methylation of cytosines within the telomeric repeat . Regulation of DNA methylation is achieved by telomeric repeat–containing small RNAs , which are derived from the processing of telomeric transcripts by the RNA–dependent DNA methylation pathway . From these data , we infer that the formation of a proper telomere structure is partly regulated by non-coding telomeric RNAs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/epigenetics",
"genetics",
"and",
"genomics/nuclear",
"structure",
"and",
"function",
"genetics",
"and",
"genomics/plant",
"genetics",
"and",
"gene",
"expression",
"genetics",
"and",
"genomics/chromosome",
"biology"
] | 2010 | siRNA–Mediated Methylation of Arabidopsis Telomeres |
Long-term disease surveillance data provide a basis for studying drivers of pathogen transmission dynamics . Dengue is a mosquito-borne disease caused by four distinct , but related , viruses ( DENV-1-4 ) that potentially affect over half the world's population . Dengue incidence varies seasonally and on longer time scales , presumably driven by the interaction of climate and host susceptibility . Precise understanding of dengue dynamics is constrained , however , by the relative paucity of laboratory-confirmed longitudinal data . We studied 10 years ( 2000–2010 ) of laboratory-confirmed , clinic-based surveillance data collected in Iquitos , Peru . We characterized inter and intra-annual patterns of dengue dynamics on a weekly time scale using wavelet analysis . We explored the relationships of case counts to climatic variables with cross-correlation maps on annual and trimester bases . Transmission was dominated by single serotypes , first DENV-3 ( 2001–2007 ) then DENV-4 ( 2008–2010 ) . After 2003 , incidence fluctuated inter-annually with outbreaks usually occurring between October and April . We detected a strong positive autocorrelation in case counts at a lag of ∼70 weeks , indicating a shift in the timing of peak incidence year-to-year . All climatic variables showed modest seasonality and correlated weakly with the number of reported dengue cases across a range of time lags . Cases were reduced after citywide insecticide fumigation if conducted early in the transmission season . Dengue case counts peaked seasonally despite limited intra-annual variation in climate conditions . Contrary to expectations for this mosquito-borne disease , no climatic variable considered exhibited a strong relationship with transmission . Vector control operations did , however , appear to have a significant impact on transmission some years . Our results indicate that a complicated interplay of factors underlie DENV transmission in contexts such as Iquitos .
Dengue is a mosquito-borne disease common throughout the tropics and sub-tropics [1] , [2] . It is caused by infection with any of four antigenically-distinct , but related , dengue viruses ( DENV-1 , 2 , 3 , and 4 ) in a human-mosquito transmission cycle . The anthropophilic mosquito , Aedes aegypti , is the predominant vector [3] , [4] . The long-term patterns of dengue incidence have been studied at numerous endemic sites , especially in Southeast Asia [5]–[12] and the Americas [8] , [13]–[15] . Results highlight intra-annual ( seasonal ) and inter-annual ( across multiple years ) signatures in transmission intensity [8] , [10] , [16] , [17] , as well as occasional abrupt shifts in the age of people with clinically apparent illness [18] . Conclusions from these studies are mixed , although in aggregate they highlight that dengue occurs across a diverse array of conditions and that the key drivers of transmission similarly vary across those different contexts [8] , [10] , [16] . Continued , detailed documentation of these temporal dengue patterns in different , endemic populations is useful for improving our understanding of DENV transmission and testing the link of key variables like temperature to components of the virus transmission cycle [10] , [19]–[21] . With this goal in mind , here we examined the temporal patterns of laboratory-confirmed dengue cases over a 10-year period encompassing the introductions of two novel serotypes into the Amazonian city of Iquitos , Peru . Despite their informational value , long-term disease data sets often lack detail because of the costs associated with detection of potential cases and laboratory-based diagnosis [22] . Furthermore , the symptoms associated with dengue fever are non-specific and can lead to misdiagnosis [23] , [24] . Nevertheless , many surveillance systems report suspected cases with confirmation of only a small fraction . While severe , hospitalized cases are less prone to misdiagnosis and are usually laboratory confirmed , they typically represent only a small proportion of the total number of people infected [25] . Moreover , severe disease outcomes are influenced by a variety of intrinsic factors ( e . g . , virus virulence , host exposure history ) [26] and not necessarily external drivers , such as climate conditions . Limitations of many long-term dengue datasets analyzed to date [9] , [13] , [15] , [e . g . 27]–[30] , in addition to variation in reporting methods , increase the difficulty and reduce confidence in defining universal properties of dengue transmission dynamics [8] . Johansson et al . [16] concluded that results of these analyses are sometimes biologically implausible and confusing , such as a negative effect of increasing temperatures on transmission [see references in 16] . Because transmission is seasonal , it will correlate with other seasonal patterns even though there is no mechanistic link . Thus , any statistical analysis should be rigorously scrutinized from a biological perspective and , preferably , cross-validated with additional data . A recent study analyzed seasonal dengue in Ecuador using linear mixed models incorporating entomological , epidemiological , and climate data [15] . The investigators found important influences of climate and entomological indices on monthly dengue case counts . Nevertheless , even using new and improved modeling approaches , in aiming to fit a particular statistical model to temporal disease data–which is often aggregated–to predict transmission patterns over time , the analysis potentially obscures other features of the time series that might generate hypotheses about underlying mechanisms . Here , we examined the seasonal patterns of dengue over a 10-year period in relation to climatic factors and citywide vector control efforts . Our analysis focused on laboratory-confirmed dengue fever cases reported to a surveillance network based in multiple health-care facilities in Iquitos , Peru . During the period of study , two novel DENV serotypes invaded Iquitos , which was already endemic for DENV . In response to the invasions and subsequent epidemics , the local ministry of health conducted citywide house-to-house insecticide fumigation campaigns to kill adult mosquitoes and reduce virus transmission . Our analyses indicate that , although climatic variables correlate weakly with variation in transmission intensity , mosquito control efforts do appear to curtail epidemics when properly applied .
Iquitos is a city of ∼377 , 000 inhabitants that sits at the confluence of the Nanay , Itaya , and Amazon Rivers in the department of Loreto in northeast Peru . Iquitos has been thoroughly described in previous publications [23] , [31]–[35] . In 2000 , as part of a collaborative effort between the Peruvian Ministry of Health and the U . S . Naval Medical Research Unit No . 6 , a surveillance network was established in public and military hospitals and clinics throughout Iquitos . For most years , 12 or 13 health centers participated , representing predominantly urban and peri-urban areas in and around Iquitos . A core of 3 hospitals and 6 clinics consistently provided samples throughout the study . A few health centers discontinued participation mid-study but were replaced by other health centers from the same geographic area . Additional details are described in Forshey et al . [23] . All data collection was conducted under study protocol NMRCD . 2000 . 0006 , approved by the Naval Medical Research Center Institutional Review Board ( Bethesda , MD ) in compliance with all U . S . Federal regulations governing the protection of human subjects . In addition , the study protocol was reviewed and approved by health authorities in Peru ( Dirección General de Epidemiología ) . Written consent was obtained from participants 18 years of age and older . For participants younger than 18 years , written consent was obtained from a parent or legal guardian . Additionally , written assent was obtained from participants between 8 and 17 years of age . Prior to analysis , all data were de-identified and aggregated into weekly case counts . Details of the surveillance system , including inclusion criteria and laboratory assays are detailed in Forshey et al . [23] . Briefly , consenting participants ( ≥5 years old ) provided an acute blood sample on the day they visited the health care facility for laboratory confirmation of DENV infection . Laboratory procedures included RT-PCR and virus isolation to identify acute infections and IgM ELISA to detect anti-DENV antibodies consistent with a recent infection . Convalescent samples collected 10 days to 4 weeks later were tested for anti-DENV IgM by ELISA . We identified the infecting serotype when possible ( 55% ) ; positive diagnosis was generically defined a “DENV” infection when based solely on IgM assay results ( Table 1 ) . In response to dengue outbreaks in Iquitos over the period of study , the Loreto Regional Health Department ( LRHD ) conducted large-scale vector control interventions ( Table 1 ) . In these , they sprayed inside houses with an ultra low volume ( ULV ) , non-residual insecticide ( deltamethrin [2002–2006] , cypermethrin [2006–2008] , or alpha-cypermethrin [2008–2010] ) three times over approximately a three-week period . The LRHD attempts to treat all houses within designated sectors of the city , which are chosen based on epidemiological information ( Table 1 ) . These citywide efforts usually treated ∼40% of all houses in Iquitos , which total ∼80 , 000 houses . Data on interventions were provided by the LRHD . For our analyses we identified weeks when fumigation was conducted in the city and examined whether treatments were associated with reductions in dengue incidence within and across years . Case data were restricted to the period between 1 July 2000 and 30 June 2010 . Positive cases were those with evidence of virus ( RT-PCR or virus isolation ) or immunologic evidence of recent infection ( acute or convalescent IgM ELISA titer>1∶100 ) . We combined all DENV+ cases into weekly totals for use in correlation and wavelet analyses ( see below ) . Generally , people visiting health centers were received for 5–7 hours a day , 5 days a week , although there was some variability in rates across seasons and clinics . A major exception was a 2-week period of 2004 when surveillance in one hospital was extended to 24 hours a day due to the large number of dengue cases they were receiving . To correct for this extended effort , we rescaled the number of cases captured in these 2 weeks by the ratio of the maximum number of negative cases observed in the remainder of the time-series to the number of negative cases observed during those particular weeks ( approximately 1∶5 ) . For disaggregated analyses , the data were randomly thinned in these two weeks based on the same scaling factors . Using the corrected time-series , we conducted autocorrelation analysis to characterize the temporal structure of the case data . Subsequently , we used wavelet analysis to identify temporal variation in the periodicity of dengue case reports . Our analysis was conducted on the square-root transformed and normalized ( by standard deviation ) time series using the Morlet wavelet transform and implemented in Matlab using the algorithm of Torrence and Compo [36] . Daily climate data for Iquitos was acquired from a US National Oceanic and Atmospheric Administration ( NOAA ) weather station located at the Iquitos airport . Reported variables include: mean , maximum , and minimum temperatures; precipitation; air pressure; wind speed; and dew point . From these data we generated several derived variables , including: daily temperature range ( DTR; max - min ) , degree-days ( DD ) , relative humidity ( RH; 100 - 5* ( Temp_mean - dewpoint ) ) , and precipitation events ( per week ) . We calculated degree-days using the triangle method and a 24°C threshold temperature for virus replication [DD 24]; [ 37 , 38] . We considered the river depth of the Amazon River as a covariate , because this variable changes dramatically over the course of the year as a function of rainfall in the Andes Mountains . At high river levels , fringe areas of Iquitos have occasionally flooded , which could have had an impact on mosquito populations . It is more probable , however , that river depth serves as an indicator of broader scale climate patterns that might correspond with conditions suitable for DENV transmission . River depth data for the Amazon River in meters above sea level was provided by the Servicio Nacional de Metereologia e Hidrologia , Peru . Seasonal and annual climate patterns were summarized graphically using a loess smoother , which summarizes the data by fitting a local polynomial [39] . The degree of smoothing desired is controlled by the parameter α , where large values indicate more smoothing . We heuristically chose values of α to emphasize short-term and long-term temporal patterns in the data . Because climate variables are highly collinear , interpretation of the relationship between any single variable and epidemiological patterns could be misleading . Maximum and minimum temperatures , for instance , should correlate in time . To address this issue , we conducted principal components analysis ( PCA ) on the climate variables . Briefly , PCA reorients a set of n covariates into n principal components ( PCs ) based on their covariation structure . The first PC ( PC1 ) always captures the largest proportion of the covariance between the covariates , with successive PCs explaining less and less of the remaining variation . With more correlation among covariates , fewer PCs are required to capture most of the variation in the dataset . By definition , the resulting principal components are orthogonal with each other ( i . e . , they do not correlate ) and the set of n PCs exactly encapsulates all covariation among the covariates . Within a PC , loading values describe the relative contribution of each original covariate . Higher loadings indicate greater correlation and high loadings on the first few PCs indicates the overall importance of that covariate in the covariation structure of the dataset . We examined the relationships between weekly DENV cases and climate variables using cross-correlation maps [CCMs; 40] . For each variable , maps were generated by varying the temporal lag and the period over which the variable was aggregated . Briefly , cases in week t0 were correlated with each covariate aggregated over a range of weeks prior to t0 , defined by the interval [t0-a , t0-b] . We evaluated the mean , median , and maximum values of the covariate , and , in some cases , the sum for each period . We present results for the median unless the sum was more appropriate . For example , it is possible that rain influences dengue cases 4–8 weeks later because of effects on mosquito population dynamics . In that case , we set a = 4 , b = 8 and looked at the correlation of maximum rainfall over that interval with the number of dengue cases a month in the future . To limit the identification of spurious correlations , we did not investigate lags more than half a year before the cases were observed ( 27 weeks ) . We believe , however , that effects most likely to have biological relevance on transmission would occur within a lag of 17 weeks ( 1 trimester ) . To investigate both linear and monotonic associations between climatic variables and cases we calculated Pearson and Spearman correlations . We categorized the correlation coefficient , r , as follows: |r|<0 . 1 , no correlation; 0 . 1≤|r|<0 . 2 , very weak , 0 . 2≤|r|>0 . 3 , weak; 0 . 3≤|r|<0 . 4 , weak moderate; 0 . 4≤|r|<0 . 5 , moderate; |r|≥0 . 5 , moderate strong to strong . Because of the large number of tests conducted ( each CCM equates to 338 correlation tests ) , we did not calculate p-values and rather focused on the relative strength of correlations . Unless otherwise stated , all analyses were conducted with R 2 . 13 .
Over the 10 years of study , 12 , 602 febrile participants were enrolled , 40% of whom were laboratory diagnosed as having acute or recent DENV infection ( Table 1 ) . Although very few dengue cases were detected at the beginning of the study ( consistent with serology data [34] ) , after late 2001 , outbreaks occurred on an annual basis ( Figs . 1 , 2 ) . Overall , weekly case reports fluctuated seasonally ( i . e . , DENV positive and negative cases together; Fig . 1 ) . Cross-correlation analysis showed that the number of cases diagnosed as something other than dengue ( DENV negative cases ) mirrored the number of dengue positive reports ( i . e . , the best lag was 0; Fig . 1c ) . Wavelet analysis indicated that the annual periodicity in transmission was particularly strong from the 2004–2005 season forward ( Fig . S1 ) . A longer , ∼3 year periodicity was also suggested by the analysis , but the 10 year time-series was too short to place confidence in this result . Over all years , 75% of DENV cases were reported between the 37th week of the preceding year and 13th week of the subsequent year , peaking on average in the last week of December ( Fig . 2 ) . We thus define the dengue season in Iquitos as occurring between September and April ( between trimester III and trimester I of the subsequent year ) . Over the 10 dengue seasons single serotypes accounted for the majority of all cases . DENV-1 was dominant in the first season , followed by the emergence of DENV-3 in 2001 , [34] , [genotype III;41] , and the emergence of DENV-4 in 2008 [genotype II; 42] ( Fig . 1 , Table 1 ) . DENV-2 ( lineage I of American/Asian genotype ) was only detected in a few study participants in 2001–2002 . DENV-1 appeared at low levels in 2002–03 and 2005–06 when DENV-3 was dominant . Although transmission intensified on an annual basis , the magnitude and timing of the peaks varied across seasons . Temporal autocorrelation of the number of weekly DENV cases indicates a strong positive auto-correlation at a lag around 2 years and a negative correlation around a lag of 1 . 5 years ( Fig . 2 ) . This result is consistent with an apparent shift in the timing of peak transmission from year to year ( Fig . 2 ) . In other words , the inter-epidemic period fluctuated between approximately 8 and 16 months . Climatic variables demonstrated seasonality in Iquitos , although the magnitude of variation was small ( Table 2 , see SI ) . Maximum and mean weekly temperatures were warmest in trimesters III and I ( between November and April ) , coinciding with the timing of detection of most dengue cases ( Figs . S2 , S4 ) . Mean and minimum temperatures showed a gradual increasing trend over the 10 years , culminating in a ∼1°C increase between 2000 and 2010 ( Fig . S4 , S6 ) . Cumulative weekly DD24 largely mirrored trends in mean and maximum temperatures , peaking in late trimester III ( November [28 . 04°C•days]; Fig . S8 ) and bottoming in trimester II ( June [19 . 27 C°•days] ) . The 10-year trend in DD24 was highly non-linear , lowest in early 2008 and increasing rapidly to its highest levels in 2009 and 2010 ( Fig . S8 ) . Precipitation occurred throughout the year , but it was usually lower in the later part of trimester II ( July [3 . 77 cm•week−1] and August [4 . 66 cm•week−1]; See Table 2 , Fig . S12 ) . Over all years , rainfall amounts were highest between 2003 and 2008 , dropping significantly in later years , although the number of precipitation events remained the same ( Fig . S14 ) . Additional climatic variables are shown in the SI . Because climate variables correlate , we conducted principal components analysis ( PCA ) to simplify the data and identify subsets of highly collinear drivers . The results of the PCA identified three components that described 79% of the variation among the climate variables ( Table 3 ) . The first , PC1 , related most strongly to temperature variables and humidity . PC1 increased with increasing humidity and decreased with increasing temperatures . The second , PC2 , captured variability in temperatures . PC2 decreased with increasing minimum temperature and river level and increased with larger DTR . The third component , PC3 , decreased with precipitation and wind speed and increased with river level ( Table 3 ) . All three principal components exhibited seasonal periodicity , although this was attenuated for PC3 in later years . ( Fig . S21 , S23 , S25 ) . Taken together , conditions in Iquitos can be described by three seasons: In trimester I , temperatures are warm , rainfall is elevated , the level of the Amazon river is increasing and dengue cases subside; in trimester II , conditions are relatively cooler and drier , the river begins to subside , and there are few dengue cases; in trimester III temperatures are their warmest and precipitation increases , the river subsides to its lowest levels , begins to rise again , and dengue transmission picks up . We related weekly reported dengue cases to climate variables using temporal cross-correlation maps ( CCMs; Fig . 3; see Methods and SI ) . Because pair-wise relationships to individual climate variables can be misleading and conflated by collinearity between climate variables , we first examined CCMs of the three principal components described earlier . We subsequently considered specific individual variables commonly associated with DENV transmission . In all instances , we produced CCMs for the whole year and for trimesters I and III ( Fig . 3 ) , when most DENV transmission took place ( see above ) . Overall , CCMs showed that there was a correlation between most climatic variables or their components and reported dengue cases , although the correlations—especially on an annual basis—were often weak ( |r|<0 . 3; Fig . 4 ) . For each CCM , we identified the maximum absolute r and plotted weekly case reports against the climate covariate to characterize the nature of the relationship ( linear , non-linear; Fig . 3 ) . We first examined the relationship between weekly dengue reports and the first three principal components , which consolidate highly collinear variables into orthogonal components ( Table 3 ) . The first component , PC1 , which associated negatively with temperature variables and positively with RH , correlated weakly and positively with dengue cases when aggregated over a broad period from 17 to 1 week earlier ( Figs . 3 , 4 , S23 ) . This means that a period of relatively lower temperatures and elevated RH preceded high case counts . When we focused only on trimester III , the correlation was weaker ( 0 . 18 ) and the lag was greater ( [−26 , −21]; Figs . 3 , 4 ) . In trimester I , the correlation was stronger ( 0 . 29 ) and the lag was less ( [−9 , −6]; Figs . 3 , 4 ) . The second component , PC2 , aggregated over 26 to 12 weeks prior , correlated more strongly ( weak moderate ) with cases on an annual basis ( −0 . 34 , [−26 , −12]; Figs . 4 , S25 ) . In the principal components analysis , PC2 correlated most strongly with minimum temperatures and DTR , thus when minimum temperature was high and DTR was small 3–6 months previous , case counts were elevated ( Table 3 ) . In trimester III , PC2 again correlated negatively ( −0 . 32 ) with cases , but at a smaller lag ( [−6 , −2]; Figs . 4 , S25 ) . The PC2 correlation and lag for trimester I was similar to the annual pattern ( −0 . 3 , [−26 , −6]; Figs . 4 , S25 ) . Finally , PC3 , which correlated most strongly with wind speed and river level , showed a weak correlation with cases on an annual basis ( 0 . 12 , [−7 , −6]; Figs . 4 , S27 ) . In trimester III , PC3 correlated weakly and positively at a large lag ( 0 . 22 , [−26 , −24] ) . In trimester I , the correlation was negative and strongest at a large lag as well ( −0 . 33 , [−26 , −20]; Figs . 4 , S27 ) . PC3 also correlated positively with cases at shorter , biologically relevant lags in this trimester ( Fig . S27 ) . Pearson and Spearman correlations for PC1 and PC2 were similar ( Fig . S28 ) . PC3 , however , differed markedly in annual and trimester III CCMs ( Figs . S27 , S28 ) . Within what we considered a biologically relevant window of 17 weeks , PCs 1 and 2 correlated with cases on an annual basis . By trimester , only PC2 correlated significantly in trimester III and both PC1 and PC2 correlated in trimester I , although the relationship with PC2 was distributed over a broad range of lags . Examination of scatterplots relating components to weekly cases revealed distinct non-linear patterns . The number of cases increased more rapidly with increasing PC1 [−17 , −1] than expected of a linear relationship ( Fig . 3 ) . There was a considerable increase in variation in the number of cases each week at higher values of PC1 ( i . e . , at lower maximum/mean temperatures and increasing humidity ) . Thus , few cases should be expected when PC1 is low 1 to 17 weeks earlier , but it is uncertain how many cases will result when PC1 is elevated over the same period . The patterns by trimester were mostly similar . Conversely , the number of cases decreased more rapidly than expected ( linear ) in relation to increasing PC2 ( Fig . S25 ) . The scatterplot of cases against the best PC2 lag shows a decrease in both the mean and variance of cases as PC2 increases , indicating that the weeks of highest incidence occurred when PC2 was very low ( high minimum temperature , low DTR; Table 3 ) between 26 and 15 weeks before . As with the relationship between PC1 and cases , due to heteroskedasticity , high values of PC2 always correspond to few cases . The patterns were similar by trimester , except the lag was much less in trimester III . Finally , the scatter plot of cases relative to PC3 showed a distinct humped pattern with most transmission occurring when PC3 was between −0 . 5 and 0 . 5 , suggesting that there is a stronger association between this component and cases than that measured with simple correlation ( Fig . S27 ) . Partitioning this analysis by trimester partly resolved this non-linearity: in trimester III the relationship is positive and linear while in trimester I it is negative and linear ( Fig . S27 ) . Mosquito development and virus replication in the mosquito are temperature dependent [43] , so ambient temperatures are often thought to play an important role in DENV transmission [37] , [44] . Precipitation , too , is often thought to be a key local variable influencing DENV transmission because mosquitoes require aquatic habitats for larval development [8] , [43] . Relative humidity combines aspects of temperature and precipitation and is probably directly important for mosquito survival because it influences desiccation rates . All of these variables naturally correlate with each other and for this reason we focused on the analysis of principal components . When considering individual variables , however , we found that correlations on an annual basis were mostly weak ( Fig . 4; See the SI for results , figures S8—S27 ) . The number of precipitation events and relative humidity correlated strongest at relatively large lags ( Fig . 4 ) . Several individual variables , temperature related variables in particular , correlated with case reports within a 17-week lag ( Fig . 4 ) . In trimester III , maximum temperature and DD24 showed moderate negative correlations , but at very large lags . Within our biologically relevant window of 17 weeks , only minimum temperature , RH and river level showed appreciable correlations in this trimester ( Fig . 4 ) . In trimester I , precipitation events , RH , wind , and river level were most strongly correlated with weekly case numbers , but at large lags . Only precipitation and wind speed correlated within a lag of 17 weeks ( Fig . 4 ) . There was evidence of non-linear relationships and heteroskedasticity in many instances ( see , for example , mean temperature in Fig . S5 ) . These were occasionally resolved when portioning the analysis by trimester . That is , a positive relationship in trimester III changed to a negative relationship in trimester I . On an annual basis , results for Spearman correlations were largely similar to those for Pearson correlations , although the correlations were stronger and extended over a longer period for temperature covariates ( Fig . S28 ) . The one exception was DTR , which correlated positively in Pearson tests , but negatively in Spearman tests at a shorter lag—although in both cases the correlation was very weak and may not be important ( Fig . S11 ) . On a trimester basis , several variables correlated well with weekly DENV cases within a 17-week lag . These were , for trimester III , minimum temperature ( 0 . 39 [−2 , 0] ) , DTR ( −0 . 44 [−8 , 0] ) , and RH ( −0 . 42 [−15 , −4]; Fig . S28 ) . In addition to climatic variation , city-wide efforts to fumigate households with insecticide to curtail transmission hold large potential for shaping inter and intra-annual patterns of transmission in Iquitos . Using data provided by the LRHD on their vector control efforts , we assessed the potential effect of citywide interventions on the number of reported dengue cases by plotting cases in week t0 with the total number of cases in the subsequent 3 weeks . We split the data by whether an intervention was taking place in week t0 and by trimester ( Fig . 5 ) . As indicated above , in trimester III dengue outbreaks were usually beginning and so the relation between cases this week and cases over the following three weeks was approximately 1∶1 or greater ( compare black and red lines in Fig . 5 ) . In seasons when an intervention was conducted in trimester III ( blue points ) , however , the relation was less than 1∶1 , which indicates a reduction in the rate new cases were captured . Conversely , in trimester I transmission was subsiding and the relationship was usually less than 1∶1 even in the absence of vector interventions . Moreover , there did not appear to be any impact of interventions when they were conducted in trimester I ( compare black and blue lines ) . That is , when interventions were conducted in trimester I any reduction in transmission was masked by the natural decline in the number of new cases reported . Over the full 10 year study period , when transmission and interventions both occurred in trimester III there appeared to be lower transmission in the subsequent trimester I ( Fig . 5 ) . We did not observe any seasons with high trimester III transmission without any intervention activities .
Dengue was not reported in Iquitos from the late 1970s—the end of the hemisphere-wide campaign to eradicate Ae . aegypti from the Americas—until a DENV-1 outbreak in 1990 [45] . Continuous DENV transmission has been detected since that time . DENV-2 American was detected in 1995 [46] . Over the period of this study , 2000–2010 , DENV-3 [34] and then DENV-4 [42] invaded the city . DENV-3 was dominant over 6 transmission seasons until it was replaced by DENV-4 in 2008 [17] . Virus transmission dynamics in Iquitos have , therefore , been largely due to single serotypes and marked by annual periodicity , suggestive of seasonal forcing . The magnitude and timing of outbreaks were variable from year to year . Because of the obvious seasonality of dengue in Iquitos and elsewhere , we examined the role of climatic drivers in transmission dynamics . Our descriptive analysis of temporal variation in dengue cases in relation to climate did not , however , resolve clear relationships . The magnitude of seasonal climatic variation in Iquitos was quite small and at least low-level transmission was detected year-round . On an annual basis , almost all of the climatic variables we considered correlated weakly ( |r|<0 . 3 ) with the number of dengue cases reported each week , with a few exceptions that were only slightly better correlated ( e . g . , relative humidity ) . Partitioning the analysis by trimester revealed stronger relationships , but most of these were distributed over very long lags ( >20 weeks ) , suggesting that the observed correlation was due to the phase difference between seasonal signals and not a mechanistic link . Principal components analysis facilitated interpretation of the observed patterns , but generally highlighted that the relationship between climate and dengue in a place like Iquitos—where climate conditions may be suitable for transmission year-round—is complex , with no single dominant climate driver . Finally , citywide vector control efforts targeting adult mosquitoes—depending on their timing—appeared to reduce transmission . In many regions of the world , particularly Southeast Asia , dengue epidemiology is characterized by co-circulation of multiple serotypes [47] . Serotype co-circulation complicates analysis of disease dynamics because the different virus serotypes interact immunologically at the level of the host and may be differentially transmitted by local mosquito vectors [25] , [48]–[50] . After examining laboratory-confirmed dengue cases reporting to a network of clinics and hospitals in Iquitos , Peru , we provide a different perspective on the dynamics of this disease from that reported for other contexts . In this isolated population of ∼400 , 000 , transmission has largely been dominated by single serotypes . On a few occasions , a small fraction of cases were due to other serotypes . Because of its population size [13] , circulation of a single serotype ( and genotype of a serotype ) over multiple years and at least one confirmed dengue case in the majority of weeks ( 81% ) , we conclude that DENV is endemic and persists in Iquitos year-round . Dengue is not hyper-endemic ( i . e . , stable , year-to-year , co-circulation of multiple serotypes [47] ) , probably because of limited connectivity to other dengue endemic areas . Occasionally , new virus strains are amplified in other parts of Peru , Colombia and/or Brazil , from which they are introduced and become established in Iquitos . Indeed , the molecular epidemiology and timing of DENV-3 and DENV-4 emergence [41] , [42] suggests that those viruses arrived to Iquitos via the Peruvian cities of Pucallpa to the south ( population ∼120 , 000; DENV-3 ) and Yurimaguas to the southwest ( population ∼48 , 000; DENV-4 ) both separated from Iquitos by a short flight or multi-day boat ride . There are no roads connecting those cities to Iquitos . Our data show that dengue incidence in Iquitos follows a clear , seasonal pattern with the number of dengue cases peaking around December ( calendar year trimesters III and I in this analysis ) . The timing of this peak varied year to year such that a short inter-epidemic period appeared to be followed by long inter-epidemic period . While this pattern is intriguing , our time series was too short to determine whether it is real and not a coincidence . Wavelet analysis suggests a 3-year cycle in incidence similar to that reported for hyper-endemic settings [27] , but , again , 10 years is insufficient data to confirm this result statistically . We find it compelling that transmission was distinctly seasonal , especially after 2004 , even though the magnitude of seasonal variation in climate was very small . When looked at on an annual basis , PC2 , which aggregated minimum temperature , DTR , and river level , was the best linear covariate . This correlation , however , was distributed over large lags and so may simply be the result of the phase difference between two seasonal signals . PC1 , which aggregated temperature variables and RH , showed some correlation with cases and over shorter lags , but it was weak . PC3 showed a very weak linear correlation , but scatterplots indicated that the actual relationship was highly non-linear . PC3 aggregated precipitation and wind speed . When we partitioned the analysis by trimester , we observed that PC2 correlated with cases in a biologically reasonable time frame in trimester III and PC1 did so in trimester I . Neither of these correlations was very strong . Also , the non-linearity in the relationship between PC3 and weekly cases was partly resolved , i . e . , the correlation was positive in trimester III and negative in trimester I . Altogether , the analysis of principle components with CCMs suggests , at best , weak climatic forcing of dengue transmission in Iquitos . This is confounded by the impacts of vector control ( see below ) , herd immunity [17] , [34] , [51] , and non-linearities in the relationships—in addition to the caveats associated with our analysis ( see below ) . Both RH and minimum temperatures have been cited elsewhere as strong correlates of DENV transmission [10] , [13] , [28] , [29] . Precipitation , too , is commonly observed to drive transmission [15] . On an annual basis , temperature-related variables predominantly correlated with dengue cases within a 17-week lag . On a trimester basis , minimum temperature , RH , and river level stand out in trimester III ( Fig . 4 ) . Precipitation and wind speed stand out in trimester I . Spearman correlations highlighted minimum temperature and RH , but also DTR . Elevated minimum temperature could accelerate larval development and reduce the DENV extrinsic incubation period . Although RH has been shown to correlate positively with transmission [10] , within the range of values we observed , it correlated negatively with cases . This is likely due to the relationship between RH and temperature ( see Table 3 ) . River level , which is driven by precipitation in the Andes mountains and not in Iquitos , probably serves as a proxy for some other proximate factors influencing local mosquito populations or transmission because it has limited impact in the areas of the city where dengue is most common . When river levels are high , transport times are significantly reduced ( AC Morrison , personal communication ) and Ae . aegypti abundances on boats are highest in October ( Guagliardo et al . in review ) . Similarly , although wind speed might affect mosquito behavior , it seems more probably that wind proxies for other environmental conditions . The range of temperatures experienced each day ( DTR ) may modify Ae . aegypti life history traits and Ae . aegypti-DENV interactions [44] , [52]–[54] . In Thailand , large daily fluctuations corresponded with less transmission . On an annual basis , we found DTR to be weakly correlated with dengue cases . DTR did load heavily on PC2 , which was more strongly correlated with transmission . This latter relationship indicates that high DTR over a period 3–6 months earlier correlated with high current case counts ( Fig . 4 ) . In trimester III , however , DTR was one of the strongest correlates over short lags in Spearman tests ( Figs . S11 , S28 ) . While this result was not apparent in Pearson tests , it suggests that DTR may be epidemiologically important for DENV transmission in Iquitos , as suspected for parts of Thailand . Overall , it appears that climatic conditions in Iquitos always hover near to a critical threshold for transmission . For instance , a small difference in temperatures could allow female mosquitoes to become infectious after only 2 gonotrophic cycles , as opposed to 3 or more , which would be expected to markedly increase vectorial capacity [55] . Clearly , though , other undefined factors are playing important roles in determining the temporal patterns of DENV transmission in Iquitos [17] . Although mosquito abundances must be important , we do not think that dengue seasonality ( especially the increase in transmission ) is uniquely driven by fluctuations in Ae . aegypti populations . Aedes aegypti is found in Iquitos year-round and , although population size fluctuates , is relatively abundant when dengue transmission is low [Reiner et al . unpublished 32] . This may contrast with other contexts where climatic variables , especially precipitation , vary more than in Iquitos [15] . On the other hand , our results indicate that vector control efforts targeting adult mosquitoes in large portions of the city were effective , accelerating virus fade-out when the intervention was applied early in the dengue season . In addition to truncating lifespan and killing infected and incubating mosquitoes , these control efforts may transiently reduce the vector population below a threshold density necessary for sustaining epidemic transmission . Health authorities in Iquitos have responded to a number of outbreaks since 2000 with the intent to kill infected and/or infectious adult Ae . aegypti and reduce mosquito abundance ( Table 1 ) . Their interventions usually involved three cycles of non-residual , intra-domicile ULV space spraying with an adulticide ( deltamethrin , cypermethrin , or alpha-cypermethrin ) . Spraying was organized by spatial units defined by the ministry of health and was typically guided by epidemiological information in order to prioritize areas with the largest number of cases . A large number of domiciles were usually treated over a period of several weeks to months . Our analyses indicate that these responses were effective at reducing transmission , which is most easily detected when cases peaked in trimester III and an intervention was conducted in this same period , i . e . , early in the transmission season ( Fig . 5 ) . Later , in trimester I , it was more difficult to detect a reduction in the number of cases caused by fumigation efforts , presumably because transmission intensity was fading for reasons other than vector control . We assume conditions become less suitable for transmission , but cannot say whether this is due to the effects of temperature on virus replication , a natural reduction in the vector population ( although vector abundances remain high in trimester I; [Reiner , et al . Unpublished] ) , increasing herd immunity or some combination of these and other factors . We deliberately focused on characterizing the temporal patterns of dengue case reporting in Iquitos in relation to commonly studied covariates , namely climate variables . One of our major goals was to inform the development of mechanistic models . In doing so , we made two methodological observations . First , CCMs are a useful tool for describing the nature of a linear correlation between two covariates . In our case we used them to find the ‘best’ periods of correlation , but found also that the maps were often very “flat . ” This simply indicates that the correlation was similar across a range of lags and periods . In other instances , there were clearly multiple possible solutions; i . e . , there were several different ‘best’ lags . Second , when plotting the scatter plot of cases against covariates at the best lag and period , we found many non-linear patterns . PC3 , which had no linear correlation with dengue case counts at any lag on an annual basis , exhibited a distinct humped relationship . Together , these observations bring into question interpretation and use of classical , linear modeling methods for fitting case data without first doing careful exploratory data analysis . Mixed modeling approaches incorporating appropriate lags and confounds might then prove appropriate tools for modeling and predicting transmission [15] , [56] . Nevertheless , where the shape of relationships is uncertain , a priori , non-parametric methods such as general additive models would be more useful . New tools are needed for exploratory analysis , however , to search across lags in order to identify the periods when covariates are most strongly associated with the variable of interest , which will guide model development . Regardless , it is critical that we develop an improved understanding of the relationship between virus transmission dynamics [e . g . 17] , [57] , [58] , per se , and disease . Although the surveillance program that generated our data was largely uniform across the years of study , changes in personnel , protocol modifications , and variation in transmission intensity likely affected the number of cases captured on a daily basis by the system . Moreover , our surveillance only covered approximately 40% of the Iquitos population , participation rates were far short of 100% , and participation was only sought during the day . We specifically addressed one period of a large increase in surveillance effort during a particularly intense dengue outbreak , but otherwise did not attempt to correct for variation in case capture efficiency . We acknowledge this limitation and in our correlation analyses used a non-parametric method ( Spearman correlations ) that–for the most part–confirmed results from the Pearson correlations . Reporting rates probably varied over the 10 years as a function of disease severity and other factors influencing individual care-seeking behavior . Although each year there was an increase in the number of dengue cases each year , awareness both in the medical community and the general public would be expected to lag actual transmission . We speculate that care-seeking behavior may change during the course of a dengue outbreak . Initially , during the increase in DENV transmission , people may be more likely to report to a clinic or hospital at the first signs of a fever or other symptom . After a period of transmission and the recognition that dengue cannot be cured with a drug , people may self medicate mild disease with an antipyretic and , thus , be less likely to visit a clinic or hospital . We acknowledge that although all of the cases in our data set were laboratory confirmed , factors not associated with transmission per se likely influenced the patterns we observed and so we considered these patterns only indicators of the true transmission dynamic . In contrast to the seasonal patterns described above ( i . e . , transmission typically peaking in late December ) , we note that the DENV-1 outbreak in 1990 peaked in May [45] and the DENV-2 outbreak in 1995 peaked in August [46] . While climatic averages may have changed some since then , the seasonality has not , which begs explanation . In light of the weak , direct relationships between climatic variables and dengue case totals we measured and the observation that conditions in Iquitos may always support some level of transmission [10] , we posit that other factors that we did not measure are important for determining the timing of intra-annual fluctuations and seasonal peaks in transmission . Both the 1990 and 1995 outbreaks were associated with novel virus introductions and we found no record of attempts during those times to perform citywide fumigation campaigns such as those begun in 2003 . Because Ae . aegypti , is present year-round in Iquitos , herd immunity and the timing of virus introduction emerge as key determinants of when outbreaks occur . Consistent with this idea , DENV-3 transmission remained high in April/May of 2002 , later than all other seasons in our analysis ( Fig . 2 ) . DENV-4 , on the other hand , peaked in October of 2008 . After invasion , as herd immunity rises , variation in mosquito abundances and the suitability of environmental conditions for transmission , should play more of a role determining transmission dynamics . We speculate that the timing and intensity of mosquito interventions to control mosquito populations influenced dynamics in subsequent seasons through their effect on herd immunity [59] . Our future work will focus on testing these hypotheses using mechanistic models [e . g . 21] . We emphasize that while climate plays a key role in DENV transmission at broad spatial scales [8] , [10] , there remain significant uncertainties regarding its specific role and importance when weighed against other drivers at local , fine scales . In different geographic contexts , climate could play a greater role in DENV transmission than in Iquitos , highlighting that DENV ecology is complex and context dependent . Nevertheless , the patterns we document here provide valuable material for the development of mechanistic models that can be used to explore alternative hypotheses about transmission drivers in addition to climate . Importantly , our results indicate that vector control efforts , albeit intensive , can reduce transmission if timed and placed properly . This indicates that vector control can be an effective tool for preventing dengue . | Description of long-term temporal patterns in disease occurrence improves our understanding of pathogen transmission dynamics and facilitates predicting new epidemics . Dengue , the most prevalent mosquito-borne , viral disease of humans , typically varies seasonally and on longer , inter-annual time scales . In most studies of these patterns , however , only a fraction of putative dengue cases are confirmed with laboratory diagnostics . Here we analyzed 10 years of fully confirmed dengue cases reported to a sentinel surveillance system in Iquitos , Peru . We describe the inter and intra-annual patterns of weekly case counts and relate these to climate and local vector control efforts . We show that dengue case counts vary seasonally in Iquitos despite very little variation in key climatic conditions , such as temperature and humidity . Overall , transmission correlated poorly with climate regardless of time lag . In seasons when vector control was conducted early , there was an apparent decline in cases later that season . We speculate that the relationships between climatic conditions and transmission of DENV in Iquitos are complex and non-linear , and that other factors , such as herd immunity , virus diversity , and vector control efforts , play key roles determining the timing and intensity of transmission . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"viral",
"transmission",
"and",
"infection",
"infectious",
"disease",
"epidemiology",
"emerging",
"viral",
"diseases",
"population",
"dynamics",
"tropical",
"diseases",
"microbiology",
"viral",
"vectors",
"emerging",
"infectious",
... | 2014 | Long-Term and Seasonal Dynamics of Dengue in Iquitos, Peru |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.